30th International PLEA Conference SUSTAINABLE HABITAT FOR DEVELOPING SOCIETIES Choosing the way forward December 16 -18, 2014 Proceedings Vol.1 Sustainable Habitat for Developing Societies: Choosing the way forward PLEA 2014 Conference 16-18 December, 2014 Conference Chairman Nimish Patel Principal, Abhikram, Ahmedabad Principal Consultant, Panika, Ahmedabad Technical Conference Chairman Prof. N K Bansal CEPT University, Ahmedabad Organizing Committee Nimish Patel, Rajan Rawal, Sanyogita Manu, Agam Shah, Yash Shukla, Keyur Vadodaria Editorial Team Rajan Rawal, Sanyogita Manu, Agam Shah, Divya Batra Conference Office Asha Joshi, Manish Salvi Hosted by: CEPT University, Ahmedabad, India Venue: Knowledge Consortium of Gujarat, Ahmedabad, India Edited by: Rajan Rawal, Sanyogita Manu, Nirmala Khadpekar ©Copyright: CEPT University, Center for Advanced Research in Building Science & Energy, Ahmedabad First Published 2014 Publisher: CEPT UNIVERSITY PRESS Centre for Documentation & Publications CEPT University Kasturbhai Lalbhai Campus University Road, Navrangpura Ahmedabad 380009, Gujarat, India Phone: +91 79 26302740/ 26302470 Fax: +91 79 26302075 Email: publications@cept.ac.in Website: www.cept.ac.in | www.plea2014.in This book was prepared from the input files supplied by the authors. The publisher is not responsible for the use which might be made of the information. II Scienti c Committee Adil Sharag-Eldin, United States Amar Bennadji, United Kingdom Amjad Almusaed, Denmark Arnaud Evrard, Belgium Denise Helena Silva Duarte, Brazil Dru Crawley, United States El-Hadi Bouguerra, Algeria Feng Yang, China Heide G. Schuster, Germany Jessen Page, Switzerland José Eduardo Vázquez-Tépox, Mexico Khandaker Shabbir Ahmed, Bangladesh Li Wan, Hong Kong Lucelia Taranto Rodrigues, UK Maria Solange Fontes, Brazil Marwa Dabaieh, LTH, Sweden Minal Pathak, India Paula Cadima, United Kingdom Ruchi Choudhary, United Kingdom Shanthi Priya Radhakrishnan, India Shoichi Kojima, Japan Silvia de Schiller, Ecuador Takashi Asawa, Japan Tetsu Kubota, Japan Thanos Stasinopoulos, Turkey Veronica Soebarto, Australia Vishal Garg, IIIT, India Werner Xaver Lang, Germany Aalok Deshmukh, India Andre Omer Desjarlais, United States Mario Cucinella, Italy Mike Barker, South Africa P C Thomas, Australia Parag Rastogi, Switzerland Prasad Vaidya, United States Santosh V Philip, United States Tanmay Tathagat, India Alpana Jain, India Dana Raydan, United Kingdom Kristina Kiesel, Austria Leena Elizabeth Thomas, Australia Melissa Smith, India Roshni Udyavar, , India Alessandro Rogora, Italy Andre de Herde, , Belgium Antje Junghans, Norway Arvind Krishan, India Bharat Dave, Australia Brian Ford, United Kingdom Byungseon Sean Kim, Korea Claude-Alain Roulet, , Switzerland Dale Clifford, United States Emauele Naboni, Denmark Emmanuel Rey, Switzerland Flavio Celis D’Amico, Spain Francois Garde, , Réunion (France) Harvey Brayn, United States Hasim Altan, United Arab Emirates Jaffer AA Khan, India Joana Carla Soares Goncalves, Brazil Jose Ripper Kos, Brazil Luc Adolphe, France M Susan Ubbelohde, United States Magali E. Bodart, Belgium Manuel de Arriaga Brito Correia-Guedes, Portugal Mary Guzowski, United States N K Bansal, India Rajat Gupta, United Kingdom Richard De Dear, Australia Robert Mark Dekay, United States Sadhan Mahapatra, India Shady Attia, Belgium Simos Yannas, United Kingdom Thomas Spiegelhalter, United States Uta Pottgiesser, Germany Vincent Buhagiar, Malta Vivian Loftness, United States Yuichiro Kodama, Japan III S pporting eviewers to the Scienti c Committee Aalok Arvind Deshmukh, India Adil Sharag-Eldin, United States Alessandro Rogora, Italy Alpana Jain, India Amar Bennadji, United Kingdom Amjad Almusaed, Denmark André De Herde, Belgium Andre Omer Desjarlais, United States Antje Junghans, Norway Arnaud Evrard, Belgium Arvind Krishan, India Bharat Dave, Australia Brian Ford,United Kingdom Byungseon Sean Kim, South Korea Catherine Semidor, France Chitrarekha Kabre, India Claude-Alain Roulet, Switzerland Dale Clifford, United States Dana Raydan, United Kingdom Denise Helena Silva Duarte, Brazil Dru Crawley, United States El Hadi Bouguerra, Algeria Emauele Naboni, Denmark Emmanuel Rey, Switzerland Flavio Celis D’Amico , Spain Francois GARDE, Reunion Harvey Bryan, United States Hasim Altan, United Arab Emirates Heide G. Schuster, Germany Hom Bahadur Rijal, Japan Isaac A. Meir, Israel Jaffer AA Khan, India Jessen Page, Switzerland Joana Carla Soares Goncalves, Brazil José Eduardo Vázquez-Tépox, Mexico Jose Ripper Kos, Brazil Keyur Vadodaria, India Khandaker Shabbir Ahmed, Bangladesh Kristina Kiesel, Austria Leena Elizabeth Thomas, Australia Li WAN, Hong Kong Luc Adolphe, France Lucelia Taranto Rodrigues, U.K. M Susan Ubbelohde, United States IV Magali E. Bodart, Belgium Manuel de Arriaga Brito Correia-Guedes, Portugal Maria Solange Fontes, Brazil Mario Cucinella, Italy Marwa Dabaieh, Sweden Mary Guzowski, United States Melissa Katheryn Smith, India Mike Barker, South Africa Minal Pathak, India Narendra Bansal, India Nimish Patel, India Parag Rastogi, Switzerland Paula Cadima, United Kingdom PC Thomas, Australia Poorva Ujwal Keskar, India Prasad Vaidya, United States Radha krishnan Shanthi Priya, India Rajan Rawal, India Rajat Gupta, United Kingdom Robert Mark Dekay, United States Roshni Udyavar, India Ruchi Choudhary, United Kingdom Sadhan Mahapatra, India Santosh V Philip, United States Sanyogita Manu, India Shady Attia, Belgium Shoichi Kojima, Japan Silvia de Schiller, Ecuador Simos Yannas, United Kingdom Smita Chandiwala, India Takashi Asawa, Japan Tanmay Tathagat, India Tetsu Kubota, Japan Thanos Stasinopoulos, Turkey Thomas Spiegelhalter, United States Uta Pottgiesser, Germany Veronica Soebarto, Australia Vincent Buhagiar, Malta Vishal Garg, India Vivian Loftness, United States Werner Xaver Lang, Germany Yash Shukla, India Yuichiro Kodama, Japan Advisory Committee Paula Cadima, United Kingdom Bimal Patel, India N K Bansal, India Brian Ford, UK Ashok Lall, India Satish Kumar, India Dru Crawley, Bentley Systems Simos Yannas, London Mike Barker, South Africa Edward Ng, , Hong Kong Denise Duarte, Brazil Ing. Heide Schuster, Germany Mario Cucinella, Italy George Baird, New Zealand V Supported and Sponsored by Ministry of New and Renewable Energy Bayer Material Science Gujarat Energy Development Agency (GEDA) Vibrant Gujarat VI Table of Contents Forewards Paula Cadima, President of Passive and Low Energy Architecture............................................................................................................................................................. IX Nimish Patel, Chairman of Passive and Low Energy Architecture 2014.................................................................................................................................................... X Keynotes Ashok B Lall.................................................................................................................................................................................................................................................................... XI Prof.Richard de Dear................................................................................................................................................................................................................................................. XII Dr. Ardeshir Mahdavi............................................................................................................................................................................................................................................... XIII Dr. Chandrashekar Hariharan ............................................................................................................................................................................................................................. XIV Session 1 (Day1, December 16, 11:30 - 13:10) Session 1A: Passive Design Effect of courtyard height and proportions on energy performance of multi-storey air-conditioned desert buildings ......................................................... 1 Improving ventilation condition of labour-intensive garment factories in Bangladesh ....................................................................................................................... 9 Impact of native evergreen trees on the visual comfort in an office space in Ahmedabad, India .................................................................................................. 17 Session 1B: Low carbon cities and neighborhood development Achieving Best Practice Net-Zero-Energy Building Design Instruction Methods ................................................................................................................................. 25 A low energy community? A comparative study of Eco Villages around the world ............................................................................................................................. 34 Spatial Structure of City Blocks with Vacant Lands in Edo, Early Modern Tokyo - Introducing the Appropriate Wind into Outdoor Living Spaces ...... 42 Optimization for Passive Design of Large Scale Housing Projects for Energy and Thermal Comfort in a Hot and Humid Climate .................................... 50 Investigations about a Scale of Correlation for the Relationship between Urban Physical Dimensions and Wind Cp ........................................................... 59 Session 1C: User behavior, thermal comfort & energy performance Daylighting for Visual Comfort and Energy Conservation in Offices in Sunny Regions ...................................................................................................................... 67 Improving Outdoor Urban Environments: Three Case Studies in Spain ................................................................................................................................................ 75 Thermal Comfort in Naturally Ventilated Classrooms .................................................................................................................................................................................. 83 Comparison of Strategies improving Local Energy Self-sufficiency at Neighborhood Scale Case study in Yverdon-les-Bains (Switzerland) .................. 91 Session 1D: Tools and methods/ framework PROSOLIS: A Web Tool for Thermal and Daylight Characteristics Comparison of Glazing Complexes ....................................................................................... 99 Improving the Energy Efficiency of the Building Stock: A Bottom-up Model and its Application in an Online Interactive Portal ....................................... 107 Comparing deterministic and probabilistic non-operational building energy modelling ............................................................................................................... 115 CBD greening and Air Temperature Variation in Melbourne ................................................................................................................................................................... 123 Session 2 (Day1, December 16, 11:30 - 13:10) Session 2A: Passive Design Sustainable habitat for emerging economies.................................................................................................................................................................................................133 A Comparative Study of Design Strategies for Energy Efficiency in 6 High-Rise Buildings in Two Different Climates.............................................................141 The potential for natural ventilation as a viable passive cooling strategy in hot developing countries .......................................................................................149 Shop-window lighting: the use of sun to improve visual appeal and reduce energy demand ......................................................................................................159 Developing bio-climatic zones and passive solar design strategies for Nepal ....................................................................................................................................167 Session 2B: Low carbon cities and neighborhood development Improving pedestrian thermal comfort by pavement-watering during intense heat events ........................................................................................................ 175 Sunlight availability for food and energy harvesting in tropical generic residential districts .......................................................................................................... 183 The Cooling Effect of Green Strategies Proposed in the Hanoi Master Plan for Mitigation of Urban Heat Island ................................................................ 191 Tall buildings and the urban microclimate in the city of London ............................................................................................................................................................ 199 Baseline Scenario of Energy Consumption of Urban Multi- Storey Residential Buildings in India ............................................................................................. 207 Session 2C: User behavior, thermal comfort & energy performance Cool spots in hot climates: a means to achieve pedestrian comfort in hot climates........................................................................................................................ 215 The Effect of Natural Ventilation and Daylighting on Occupants’ Health in Malaysian Urban Housing ..................................................................................... 223 Thermal Comfort in Offices in India: Behavioral Adaptation and the Effect of Age and Gender ................................................................................................. 231 Factors influencing window-opening behavior in apartments of Indonesia ...................................................................................................................................... 239 Session 2D: Tools and methods/ framework An Operational Indicator System for the Integration of Sustainability into the Design Process of Urban Wasteland Regeneration Projects …......... 247 A simplified approach to integrate energy calculations in the Life Cycle Assessment of neighbourhoods ............................................................................. 255 VII Session 3 (Day1, December 16, 16:05 - 17:45) Session 3A: Lessons from vernacular architecture The ‘Teatinas’ of Lima: Energy Analysis and Possibilities of Contemporary Use ................................................................................................................................. 263 Influence of Greenery in Cooling the Urban Atmosphere and Surfaces in Compact Old Residential Building Blocks: A Building Morphology Approach ................................................................................................................................................................................................................................................................... 271 Design optimisation of vernacular building in warm and humid climate of North-East India ....................................................................................................... 279 The Climate Design in Chinese Vernacular Courtyard House settlement – Wind Environment Simulation .......................................................................... 287 Demystifying vernacular shop houses and contemporary shop houses in Malaysia; A Green-Shop Framework ................................................................ 295 Session 3B: Innovative technologies Design Strategies on Heat Recovery of Cooking Stove in Rural Houses of China ............................................................................................................................. 304 Efficient building design model generation and evaluation: The SEMERGY Approach ................................................................................................................... 311 Integration of Outdoor Thermal andVisual Comfort in Parametric Design ........................................................................................................................................ 319 Session 3C: Control techniques for energy management Energy demand, thermal and luminous comfort in office buildings: a computer method to evaluate different solar control strategies .................. 329 Tuning a house through building management systems .......................................................................................................................................................................... 337 Adaptive comfort and control protocols for natural ventilation .............................................................................................................................................................. 345 Effective natural ventilation in modern apartment buildings ................................................................................................................................................................... 353 Session 3D: Tools and methods/ framework Numerical Simulation of Passive Cooling Strategies for Urban Terraced Houses in Hot-Humid Climate of Malaysia .......................................................... 361 Thermographic Study on Thermal Performance of Rural Houses in Southwest China .................................................................................................................. 369 Aggregating building energy demand simulation to support urban energy design .........................................................................................................................378 Built Environment Sustainability Assessment of Poor Rural Areas of Southwest China ................................................................................................................ 386 Session 4 (Day2, December 17, 08:30 - 10:10) Session 4A: Passive Design Climate-responsive Vernacular Swahili Housing .......................................................................................................................................................................................... 394 ‘The Open Air Office’ - Climatic adaptation of the office building typology in the Mediterranean ................................................................................................ 402 Session 4B: Low carbon cities and neighborhood development Solar radiation availability in forested urban environments with dry climate. Case: Mendoza metropolitan area, Argentina ........................................... 410 Incremental Housing as a method to the Sustainable Habitat ................................................................................................................................................................ 418 An ENVI-met Simulation Study on Urban Open Spaces of Dhaka, Bangladesh ................................................................................................................................. 426 Session 4C: User behavior, thermal comfort & energy performance Thermal comfort in residential buildings for the elderly under climate changes context .............................................................................................................. 432 Shifting the Norm - Towards Effective Mixed Mode Buildings ................................................................................................................................................................. 440 Integration of a comprehensive stochastic model of occupancy in building simulation to study how inhabitants influence energy performance . 448 Assesing Pedestrian Thermal Comfort within The Buenos Aires Climatic Context ........................................................................................................................ 456 The Catalyst Role of School Architecture in Enhancing Children’s Environmental Behavior ........................................................................................................ 464 Session 4D: Tools and methods/ framework A new subjective-objective approach to evaluating lighting quality: A case study of concert lighting for Cambridge King’s College Chapel ................. 476 Characterization and valorization of shading devices: proposition of a simple and flexible model .......................................................................................... 484 Testing a method to assess the thermal sensation and preference of children in kindergartens............................................................................................. 492 Analysis of the contribution of the building elements for improving the airtightness in residential buildings ...................................................................... 500 DOE sensitivity analysis of urban morphology factors regarding solar irradiation on buildings envelope in the Brazilian tropical context ................. 508 Session 4E: Material technology Analytical computation of thermal response characteristics of homogeneous and composite walls of Building and Insulating materials used in India ........................................................................................................................................................................................................................................................................ 516 Investigation on the performance of alternative walling materials in an affordable housing unit situated in warm humid climate ............................. 524 Parametric design for technological and “smart” system. Adaptive and optimized skin ............................................................................................................. 532 The Influence of Insulation Styles on the Air Conditioning Load of Japanese Multi - Family Residences ................................................................................ 540 Contemporary use of earthen techniques in Colombia: Thermal performance of domestic and non-domestic building typologies ........................... 548 VIII Foreword Founded in 1981, PLEA (Passive and Low Energy Architecture) is the oldest organisation that has played a pivotal role in bringing Sustainable Architecture to the mainstream. As an autonomous world-wide non-profit network of professionals, academics and students, PLEA International is engaged in a discourse on sustainable architecture and urban design (http://plea-arch.org). PLEA conferences are held annually and address a wide variety of contemporary and imminent themes, relevant to the local hosting regions. It was with great enthusiasm that the PLEA Board of Directors welcomed the challenge for the 30th PLEA International Conference to be hosted in Ahmedabad and organised by the Centre for Advanced Research in Building Science and Energy, CEPT University. This year’s focus on sustainable habitat for developing societies and emerging economies will challenge the urgent need to reduce energy use in new and existing buildings in cities that are witnessing rapid growth and urbanization. India is today one of the world’s fastest growing economies due to industrialisation, urbanisation and economic development together with people’s expectations of improved living standards. India’s urban landscape is changing at unprecedented rates due to the immense increase in population and migration to its cities. Some cities are expanding and others are being built afresh, while new architectural and urban solutions are being shaped. Energy consumption in the building sector is very high and it is expected to increase further due to change of life style, typology of the building and climate change. With further growth there is an urgent need to reduce energy use in new and existing buildings in cities. Hot and dry climates, like those in several Indian regions, encourage greater use of air conditioning system. Today new Indian architecture in India shows an increase number of enclosed, gated communities, glass boxes, high-rise buildings and skyscrapers that produce self-contained environments. Buildings fully dependent on mechanical air-conditioning are still on the rise, with little connection to the outside and no control over individual environmental conditions. We also see a movement towards a globalised, far from locally generated, architecture. However, we never had so many buildings as before claiming to be “green” or “sustainable”, but there is little knowledge about how they performed and about its inhabitants’ perspective. Detailed and reliable measurements are quite scarce. We need to know more about how buildings are used and how they adapt to their environment. Labelling and certification can be an indication of certain environmental qualities of a building, but they tell little about how people feel and enjoy living inside it. Measurements and data from surveys while buildings are being occupied are fundamental for architecture to evolve and turn more sustainable. There are many interesting examples out there worth having data about the inhabitants’ experiences. It is the only way we designers can learn how well our concepts and our designs are doing. Concerns over global warming and the need for sustainable development, have led to large investment into the research and development of technologies and design approaches which will reduce our dependency on fossil fuels. The application of these techniques to building projects has spread around the world and is slowly becoming part of the mainstream, but there is a perceived knowledge and skills gap among construction professionals. The global appeal of bioclimatic approach has been promoted worldwide by the PLEA expert network through its international conferences. During the course of the PLEA 2014 conference in Ahmedabad, debates will address various dimensions of architectural and design science to help realising buildings, neighbourhoods and cities that have minimal impact on natural resources whilst satisfying the comfort requirements and aspirations of a fast developing society. Under this central theme, the conference will propose a diverse range of topics to understand the role of architectural practice, research and education towards addressing the issues of energy conservation, efficiency and management through design, construction and operational stages of buildings, neighbourhoods and cities. Will Indian architectural practices come up with the right vision and lead Asia to move towards a new approach to deal with the challenges of climate change, technological development and urban growth, in an alternative way, less dependent on fossil fuels? I hope the current transformations India is facing are a great opportunity for new moves towards the development of more pleasant cities and sustainable buildings. I expect that PLEA 2014 is a chance for international and local sharing of current practices, research and knowledge and that the event generates dynamic discussions and innovative approaches to future challenges in architecture and urban design in India and worldwide. I would like to thank all those who contributed to the elaboration of this book of proceedings: the Organising Team for their relentless dedication during the preparations; the peer reviewers who helped with the challenging task of selecting and reviewing almost 950 abstracts and 380 full papers, providing critical and constructive comments and ensuring scientific quality; and all the authors for their contributions and for sharing their projects and findings without which this book of Proceedings and the 30th PLEA International Conference would not be possible. Paula Cadima, President of PLEA IX Foreword ‘CHOICES WE HAVE, AND CHOICES WE NEED TO MAKE’ The Twentieth century has been a century of ‘Consumption’, a concern, which a majority of thinkers, decision makers, professionals, activists, and a few leaders of the world have already brought to the forefront. Nonetheless, there is a growing consensus worldwide, that the Twentyfirst century needs to be the century of ‘Frugality’ and ‘Responsible Decision Making’. I am convinced that making judicious choices is no longer an option for any of us, irrespective of the nature, the scale, the complexity, or the constraints, of the issues, instead, it must be imbibed in our psyche. ‘Information era’ as it has also been called, has been marked by an onslaught of unprecedented mass communication tools, decreasing the physical distances, and increasing the range of choices for every conceivable problem or an issue. It appears that technology backed answers or solutions spring up in no time being labeled as ‘Innovation’ and ‘Development’. It is not my argument that these innovations and developments are undesirable, however, the questions is, have these innovative choices, in reality, solved the problems, and resolved the issues, without creating another set of problems and issues? Has the freedom offered by multiple choices at our finger tips, led to improving our decision making processes in the long term interest of the society at large, of our globalized but diverse worlds? Quite often, I wonder, if the time we consume in taking a decision from the many choices out there, leave us with any time to think about ‘how to make a choice’? Does the pace of the growth of the society, and the pressures of making timely decisions leave us with any space for considering, what will be the long term impact, or consequences, of that choice? For any given problem, or an issue, or a situation, do the choices we make reflect a holistic thought process, do they curb the creation of new unforeseen issues to be resolved, or do they just follow the market driven, novelty focused materials, technologies, and innovations? Why are the ‘frugal innovations’ not seen to be emerging as fast as the prevalent unsustainable scenarios demand them to be? In my observation, application of the knowledge of the ‘Traditional Wisdom’ and the freely available ‘Common Knowledge & Common Sense’ has eluded their use in our decision making processes. The question is, is this by choice, or by default? The ‘Traditional Knowledge & Wisdom’ needs to be seen as a community asset, which have emerged over centuries of development, and have contributed significantly to make the world far more sustainable than what we find ourselves in today. These appear to have almost disappeared from our choices today, as if they are irrelevant. These applications, however, are not only relevant, but are available and utilised in the fringe & regional pockets of most emerging economies, and need to be taken cognizance of. PLEA 2014 is the opportunity for all of us to deliberate some of these issues threadbare and explore further the appropriate avenues & directions to make the world we live in more responsible, more sustainable, and of course more livable. This is certainly possible when 164 papers and 65 posters representing 42 countries come together as more than 350 delegates exchange thoughts, ideas, and their divergent views towards a common cause, over three intense days. It is time to think of the ‘Choices we have and Choices we need to make’, to become more responsible citizens of the world we intend to leave behind for the coming generations. Nimish Patel Chairman PLEA 2014 X Widening Horizons and Evolving Practice of Sustainability in India – A Case for Convergence Ashok B. Lall, Principal, Ashok B. Lall Architects ablarch@gmail.com In today’s context of a developing economy undergoing rapid urbanization the coming decade will, arguably, be the determining decade for the quality of life and environmental fate of our cities. Over the last decade, with a growing awareness of climate change and environmental sustainability, the professions of the built environment have come to recognize the criticality of the ways in which buildings are to be built and the patterns of urban development that are adopted for the growth of cities. One can see increasing activity at the governmental and institutional levels with a primary focus on Climate Change and, more recently, a concern for protection of the environmental commons and bio-diversity. In this largely science-and-technology approach, which has its roots in the formulation of the sustainability theory in the developed West, what is clearly missing is an understanding of the relationship between sustainable architecture and urbanism in its developmental dimension –as a strategic method for improving the quality of life founded on environmental security, from a platform of limited resources. A more serious concern is the confusion that pervades the culture of architectural practice with respect to sustainability. This paper attempts to establish the social and environmental context of the present developmental condition in India which the practice of sustainability needs to engage with. It briefly traces the historical processes that have thrown up divergent attitudes to sustainability in the practice of architecture today. It argues for bringing together the various strands of knowledge and practice, and for a theory for the practice of sustainability specific to the situation of the developing South. XI Thermal counterpoint in the phenomenology of architecture – A Phsychophysiological explanation of Heschong’s ‘Thermal Delight’ Prof. Richard de Dear Professor Dr. Richard de Dear is the Director of the Indoor Environmental Quality Lab and Head of Architectural Science Discipline at the Faculty of Architecture, Design and Planning, The University of Sydney, Sydney, Australia. Faculty of Architecture, Design and Planning, The University of Sydney, Sydney, NSW 2006, Australia Typically about half of a commercial building’s energy input is allocated to the pursuit of thermally neutral indoor environments. In developed countries we find ourselves spending more than 90% of our daily lives inside built environments, most of which are now sealed off from the outside world and fully air-conditioned such that their indoor climate typically hovers within a single degree some theoretical optimum - circa 22°C. Yet despite the prodigious energy costs of thermally neutralising buildings, large thermal comfort field-studies consistently report overall levels of occupant thermal satisfaction rarely getting above 80%. In this paper I’ll take another look at Heschong’s (1979) eloquent diatribe against thermal neutrality. Using a phenomenological rather than scientific analysis, Heschong’s ‘Thermal Delight in Architecture’ succinctly put the case that architecture was profoundly impoverished when it outsourced to engineers responsibility for the thermal realm of buildings. Heschong contends that, under certain combinations and sequences, the elements of indoor climate can infuse our total sensory experience of space with layers of affect, emotion, even delight, in ways that other dimension of the built environment can’t. But these opportunities are squandered when thermal design falls into the hands of those whose stated mission is to neutralise buildings. Although she didn’t know it at the time of writing her book, the phenomenon Heschong describes as thermal delight has a name in contemporary physiology – alliesthesia. It refers to situations in which a given thermal stimulus can be subjectively experienced as either pleasant or unpleasant, depending on whether it is likely to restore or perturb the milieu interior’s target set-points. The hallmark of positive alliesthesia is pleasure, which is made available across all our senses through contrast, transience, non-steady-state, light and shadow. So in the thermal context of buildings, the HVAC engineer’s design objective of eliminating all thermal sensation from the occupant’s experience of the space efficiently precludes thermal delight. When we engineer spatial and temporal thermal uniformity into a space the building ceases to have any physiological significance or meaning for its occupants – thermal affect and hedonics are extinguished when a building is neutralised. The paper will conclude by illustrating the phenomenon of alliesthesia with Gujarat’s architectural treasures. The AdalajStepwell is a magnificent water cistern close to Ahmedabad in the Indian state of Gujarat. Since the structure is five storeys deep, people wishing to draw water from the well experience a pronounced thermocline upon descending from the heat of the sub-continental climate at ground level, down to the luxurious subterranean coolth at water-level. The thermal textures and counterpoint along this trajectory beautifully exemplifies the psychophysiological principle of alliesthesia, reinforcing and amplifying the visual delight of the step-well’s carved sandstone with another exquisite delight of the thermal variety, engendering a deeply poetic sense of place. Keywords: Thermal comfort, isothermal, transient, thermoreceptor, alliesthesia, topophilia, step-well. XII The ill-tempered urban environment Ardeshir Mahdavi, PhD Professor Dr.ArdeshirMahdavi is the Director of The Department of Building Physics and Building Ecology and Chair of the Graduate Studies in Building Science and Technology, Vienna University of Technology, Vienna, Austria. Department of Building Physics and Building Ecology Vienna University of Technology, Vienna, Austria To call cities complex entities would be a major understatement. A multitude of environmental, economical, and social disciplines, approaches, and studies have only started to scratch the surface of intricate life and evolution of structures we call cities. In this context, the present keynote address focuses on the challenging and consequential topic of the urban microclimate. In recent history, cities have grown in number and size, emerging thus as massive anthropogenic interventions in the planetary environment. They contribute to climate change and are affected by it. The urban microclimate with its temporal and spatial variance can significantly influence the performance of buildings and the well-being of the city dwellers. In light of this, there is a critical need for a deeper understanding of the tightly intertwined feedback loops between the local, regional, and global climate and their consequences for urbanism and architecture. XIII Am I My Brother’s Keeper. Dr. Chandrashekar Hariharan Chairperson, ZED Group, Bangalore ‘I am my brother’s keeper’ is shorthand for an ideal self-sacrifice and service to the larger group. If helping your brother is a ‘social purpose’, then how does one see the bizarre spectacle of individuals, companies, and senior bureaucrats and politicians forcing helpless and mute millions to give up or sacrifice their lands, their forests, their rivers, and to suffer untold misery in the name of helping one’s brother? This is not social responsibility. Can you move from such supply/side thinking, which are insensitive to ecosystems and to their vulnerable people, and move firmly toward demand-side approaches where you tell yourself that the only solution for energy deficiency, is not energy generation, but is energy efficiency. How do you move from these market-led central solutions for energy, water, and waste and move towards federal and local solutions within your home, neighborhood or office block? How do you move from central infrastructure to self-reliance with from government agencies for energy, water and waste? XIV Session 1A : Passive Design PLEA 2014: Day 1, Tuesday, December 16 11:30 - 13:10, Auditorium - Knowledge Consortium of Gujarat Effect of Courtyard Height and Proportions on Energy Performance of Multi-Storey Air-Conditioned Desert Buildings. Khaled El-Deeb, PhD Ahmed Sherif, PhD Abbas El-Zafarany, PhD [Alexandria University, Egypt.] Km_eldeeb@yahoo.com [The Ameican University in Cairo] asherif@aucegypt.edu [Cairo University, Egypt] elzafarany@hotmail.com ABSTR ACT Courtyard buildings have been always recommended as a passive architectural technique in desert environments in order to maintain indoor thermal comfort. Nowadays, an increasing number of buildings are air-conditioned. The importance of using passive techniques, then, becomes to reduce energy consumption. A previous study, however, showed that in desert environments, the energy performance of two-storey residential courtyard buildings proved less efficient than other solid forms, even when attached to neighbouring buildings from three sides in a compact urban fabric. Their performance was relatively better in mild desert climates than in extreme hot ones. The study was limited to a single family house with “thin” depth of zones surrounding the courtyard. In multi-storey courtyard buildings, the courtyard results in more height and self-shading on the facades overlooking the courtyard. This will have a direct effect on the energy consumed for cooling and heating, as well as on that consumed by artificial lighting This study questions the effect of courtyard height proportions and thickness of the built area surrounding it on the energy consumption in multi-storey air-conditioned courtyard buildings and tracks that effect under different desert climates. Courtyard buildings of 1-10storey-height were modelled using the DesignBuilder software and simulated using EnergyPlus simulation engine for the desert climates of Khargah, Cairo, Alexandria and for the temperate climate of Berlin for comparison. All cases were compared to the corresponding solid building forms of the same built area. Air-conditioned courtyard houses has not shown a significant improvement in energy savings in desert environments, buildings with bigger depth surrounding the courtyard had a much better performance than thinner buildings, giving small energy savings with building depth exceeding 12m Keywords: multi-storey courtyard, air-conditioned, desert buildings, energy performance simulation. INTRODUCTION Courtyard buildings have been always recommended as a passive architectural technique in desert environments in order to maintain indoor thermal comfort. Nowadays, an increasing number of buildings are air-conditioned. The importance of using passive techniques, then, becomes to reduce energy consumption. A previous study, however, showed that in desert environments, the energy performance of two-storey residential courtyard buildings proved less efficient than other solid forms, even when attached to neighbouring buildings from three sides in a compact urban fabric [1]. Their performance was relatively better in mild desert climates than in extreme hot ones. The study was limited to a single 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 1 family house with “thin” depth of zones surrounding the courtyard. Neither the effect of change in building depth (BD) surrounding the courtyard was not studied, nor the effect of courtyard height proportions (HP), while both are still questionable. Review of recent literature demonstrated that the performance of a courtyard as a passive cooling strategy was discussed in numerous publications. The effect of a naturally ventilated courtyard on thermal performance was studied in hot arid, tropical and warm humid tropical climates [2, 3, 4]. Results showed that a courtyard building with controlled natural ventilation, of specified opening time improved thermal performance. However in hot arid climate, the thermal performance resulting from continuous day and night natural ventilation was worse than keeping the building closed without natural ventilation [2]. The shading effect of different courtyard forms [5] and that of courtyard proportions [6] were studied. It was found that in Rome, courtyards with deep proportions were recommended over shallow ones. However, in both studies the tested buildings were solid with no windows, and thus both the effect of transmitted solar radiation and the energy needed for artificial lighting were not considered. The passive effect of courtyard with plants and water pool on energy consumed for heating and cooling was studied [7]. It was found that passive features alone could not maintain comfort during hot summer times in Tehran, and that similar effects could be obtained through envelope components such as insulation and double glazing. However, the energy needed for artificial lighting that compensates for the effect of shading was not accounted for. A study of energy performance of courtyard buildings in different climatic conditions showed that better performance was achieved in hot-dry and hot-humid climates rather than in cold and temperate ones [8]. The study was limited to zones overlooking the courtyard and ignored the influence of the external perimeter walls and zones. The impact of integrating deep courtyards in mid-rise housing buildings in Dubai was evaluated, showing that a six-storey courtyard building achieved up to 6.9% savings [9]. The addressed heights ranged from 4 to 10 stories high, while two-storey low-rise residential buildings that are common in some countries like Saudi Arabia were not considered. Some studies addressed the effect of orientation on thermal performance for non-air-conditioned buildings in a hot-humid tropical climate [10] and the implications of orientation on thermal energy efficiency of passive buildings in mild temperate climate [11]. Literature showed that the combined effect of building depth surrounding the courtyard and the courtyard’s height proportions on energy consumed in heating, cooling and lighting of air-conditioned buildings in the desert needs more investigation. OBJECTIVES This research aims at exploring the effect of courtyard on energy consumption of heating, cooling and lighting in air-conditioned multi-storey residential desert buildings based on two parameters: height proportions, and the depth of built area surrounding the courtyard. METHODOLOGY Six courtyard buildings with fixed courtyard plan dimensions (12X12m) were tested for energy performance in the following cases of thickness of built area surrounding the courtyard: 4, 6, 8, 10, 12, 20m. Values of thicknesses 14, 16, 18m were interpolated.Each case was tested in building heights of 1,2,4,6,8 and 10 floors. These represented courtyard section length-to-height proportions of 1:0.25, 1:0.5, 1:1, 1:1.5, 1:2, and 1:2.5 respectively. Then, for each of the six main cases, a solid square building of the same built area and height (but with no courtyard) was tested for comparison. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 2 The energy use intensity (EUI) of each of the courtyard BD cases was compared in different HPs in order to detect how the building’s height affects the overall energy consumption per square meter. This overall value in each courtyard building case was compared to its corresponding solid square case to detect which form was more efficient in energy consumption. COURTYARD and SOLID SQUARE Building Cases. 4m Building Depth (BD) 6m 8m 10m 12m 20m Height Proportions (HP) 2 1:2.5 1:2 1:1.5 1:1 1:0.5 1:0.25 Figure 1: Tested courtyard height and building depth proportions and solid square buildings. All courtyard and solid building cases were modelled using Design Builder software and simulated using EnergyPlus. Even at small building thickness, it was assumed the thickness is divided into two zones one facing the courtyard and the other on the external perimeter. Table 1: Simulation parameters of tested Buildings SIMULATION PARAMETERS BUILDING Form CONSTRUCTION Square External walls 20cm concrete block + 2cm cement plaster each side Internal walls 10cm concrete block + 2cm cement plaster each side WWR 20% fixed for all forms Roof Insulated with 10 cm polystyrene foam Occupancy 0.13 person/m2 Internal slab 20cm concrete + 10cm flooring + 2cm plaster Schedule Residential Windows Type Double-glazed clear Type: Fluorescent Suspended Daylightin g control Illuminance: 300 lux Dimming: On/off Courtyard Dimensions 12X12m HVAC Cooling LIGHTING 23 Heating 21 Type Split Sensor Height: 0.8m Simulations were performed for four cities: Alexandria, Cairo, and Khargah located in Egypt, and classified as hot arid according to koeppen-Geiger classification [12]. For comparison; Berlin, a temperate city with warm summer was simulated. Despite being classified as desert, the first three cities represent three different cases: Alexandria is a Mediterranean coastal city, Cairo is inland 220 km south of Alex., Khargah lies in the sahara 600km south of Alex. Figure 2 shows the difference in climate. Khargah is the highest in temperature, and out of comfort level for nearly all the year. Cairo is less in temperature than Khargah, yet higher than Alexandria. Berlin is the lowest. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 3 45 45 40 40 35 35 30 30 25 45 45 40 40 35 35 35 35 30 30 30 30 25 25 25 25 25 20 20 20 20 20 20 15 15 15 15 15 15 10 10 10 10 10 10 5 5 5 5 5 5 0 0 0 1 -5 45 Khargah Mean Daily Temperatures o C 40 -5 1 2 3 4 5 6 MAX 7 8 9 10 11 12 -5 Cairo 0 0 21 32 43 54 65 76 87 918 1029 11 3 12 4 12 5 6 7 10 11 -5 -5 Mean Maximum MAX MAX MIN MIN MAXDaily MIN 45 Alexandria 40 Berlin 0 81 92 103 114 125 Mean MIN 6 7 8 9 10 11 12 -5 1 2 3 DailyMAX Minimum MIN 4 5 6 7 MAX 8 9 10 11 12 MIN Figure 2: Mean daily maximum and minimum temperatures in tested cities for each month. RESULTS AND DISCUSSION Performance simulations showed a clear difference across the tested cities. Height proportions: 350 300 250 200 In desert climates, results showed that EUI for all forms increased by increasing height in all cities. For example, In Khargah city, dominated by cooling loads, a courtyard building with BD 8m in different cases of floor HPs showed that the ground floor was always of the least consumption, then the second floor consumed more cooling energy, and starting from the third one the cooling energy at each floor were nearly constant, then it increased again at the top floor, Figure 3. This indicated that the ground floor was significantly lower in consumption due to the heat sink to the ground, while the top floor was higher but with a small difference than the preceding floor despite being subjected to the solar radiation 350 due to the thermal insulation of the roof by 10cm. Thus in low-rise cases, the positive effect of heat sink on minimizing the overall EUI was significant. This effect became less as height increased. 300 As the tested buildings are fully air-conditioned with no natural ventilation, they were not directly affected by air temperature inside the courtyard. These results differed from what is expected in naturally ventilated250courtyard buildings where height promotes natural convection, and stack ventilation. The decreased direct solar radiation at bottom floors increased lighting energy consumption, and its radiant 200 so, minimized the expected savings in cooling loads resulting from self-shading. fraction; and 300 150 150 100 100 250 200 150 50 50 100 50 0 0 m 12m 14m 4m 16m 6m 18m 8m 20m10m 4m12m 6m14m 8m16m 10m18m 12m20m 14m4m 16m6m 18m8m 20m10m 4m12m 6m14m 8m16m 10m18m 12m 14m4m 16m6m 18m8m 20m 4m12m 6m14m 8m16m 10m 12m 14m4m 16m6m 18m 20m 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m 12m 14m 16m 20m 10m 18m 20m 8m 1 1: 41.0 1: 61.5 10 ALEX FLOOR 08 FLOOR 07 FLOOR 06 FLOOR 05 CAIRO FLOOR 04 10 FLOOR 03 FLOOR 02 10 ALEX FLOOR 01 FLOOR 06 FLOOR 05 FLOOR 04 CAIRO FLOOR 03 10 FLOOR 02 FLOOR 01 FLOOR 04 FLOOR 03 CAIRO KHARGAH KHARGAH 1 1: 20.5 Courtyard Height 1: 0.25 Proportions FLOOR 02 10 FLOOR 01 FLOOR 02 10 FLOOR 01 KHARGAH 10 Floor Number FLOOR 01 0 10 10 10 BERLIN ALEX 1: 82.0 COURT COURT ON KHARGAH 10 cm Sum of BUILDING LIGHTING (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BUILDIN Sum of BUILDING LIGHTING (Kwhr/M2) SumLIGHTING of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING (Kwhr/M2) Lighting Sum of BUILDING (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING (Kwhr/M2) Cooling Heating COURT COURT 1:1 - 8m KHARGAH Figure 3: EUI per floor inFLOOR courtyard buildings, with total height Sum 1,2,4,6 and (Kwhr/M2) 8 Floors in Khargah Sum of LIGHTING (Kwhr/M2) Sum of FLOOR HEATING (Kwhr/M2) of FLOOR COOLING city representing courtyard height ratios 1:0.25 to 1:2. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 4 300 250 The reflectance of wall paint was assumed to be 50% as the color of the accumulated dust of the desert 200 usually overrides the color of light paints. In Berlin, dominated350by heating loads, the contrary occurred. The heat sink lead the ground floor to 350 350 150 consume more heating, making the building of 1 or 2 floors consume more than a the 4-storey one, then 300 300 a gradual increase occurred as height increased, Figure 4. 300 350 300 250 100 200 50 150 100 250 250 250 200 200 200 150 150 150 100 100 100 0 16m 18m 8m 20m10m 4m12m 6m14m 8m16m 10m18m 12m 14m4m 16m6m 18m8m 20m10m 4m12m 6m14m 8m16m 10m 12m 14m4m 16m6m 18m8m 20m 4m12m 6m14m 8m16m 10m 12m 14m 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m m 18m 20m 10m 18m 20m 50 506m 50 20m 1: 2.5 1: 1.5 1: 2.0 1: 0.5 2 4ALEX 6 8 10 1 2 CAIRO 2 KHARGAH 4ALEX6 8 10 1 2 COURT KHARGAH ALEX 8m 8m BUILDING COURT LIGHTING 10 10 1: 1.0 1: 2.5 1: 0.25 1: 1.5 10 1: 2.0 1: 1.0 1 1 1: 0.5 8 4 6 8 BERLIN CAIRO COURTCAIRO CAIRO 1: 2.5 4 6 CAIRO 0 10 10 1: 0.25 2 2 1: 1.5 1 1 1: 1.0 10 10 1: 0.5 1: 2.5 8 2 KHARGAH 4ALEX6 8 ALEX KHARGAH 1: 0.25 1: 1.5 4 6 2 KHARGAH 8m 10 10 1: 2.0 1: 0.5 1: 1.0 2 1: 0.25 1 10 0 4 6 KHARGAH 8 10 1 Courtyard 2 4 6Height 8 10 1 CAIRO 0 1 Proportions ALEX CAIRO 1: 2.0 50 10 10 0 4 6 8 4 6 8 CAIRO BERLIN BERLIN BERLIN 10 10 1 2 ALEX 4 6 8 10 1 2 BERLIN KHARGAH 8m 4 BERL COURT COURT COURT Sum Sum of (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BUILDING LIGHTING (Kwhr/M2) of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING (Kwhr/M2) Lighting Sum of BUILDING LIGHTING (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING ( Cooling Heating Sum of BUILDING LIGHTING (Kwhr/M2) COURT Sum of BUILDING LIGHTING (Kwhr/M2) of BUILDING LIGHTING (Kwhr/M2) Sum of BUILDING Sum HEATING (Kwhr/M2) LIGHTING (Kwhr/M2) (Kwhr/M2) Sum of BUILDING BUILDING HEATING HEATING (Kwhr/M2) Sum of BUILDING BUILDING COOLING LIGHTING (Kwhr/M2) Sum of BUILDING Sum HEATING (Kwhr/M2) of BUILDING COOLING (Kwhr/M2) Sum of COOLING (Kwhr/M2) (Kwhr/M2) 2 Sum of BUILDING HEATING (Kwhr/M2) building with BD 8m, in case of different building heights Figure 4: EUI in Kwhr/m forSumcourtyard of BUILDING COOLING (Kwhr/M2) in the four cities. Sum of BUILDING COOLING (Kwhr/M2) In Cairo and Alexandria the same pattern occurred as in Khargah, however, consumption values differed according to climate. Figure 4 shows that increasing HPs lead to an increase in the overall energy consumption in all tested desert cities. 350 300 250 200 150 Depth of Building Surrounding the Courtyard: 350 Forms with BD alternatives 4m-20m surrounding the courtyard were tested. Changing BD while fixing courtyard dimensions means that the exposed surface area-to-built volume ratio (S:V) was also 300 changed. This ratio was also changed by changing HP at each BD. Results showed that in both extreme hot and cold climates of Khargah and Berlin, the BD was a 250 factor, Figures 5,6. In Khargah, as BD increased, total energy consumption decreased, in determinant spite of the increase in lighting energy that was overcome by greater savings in cooling loads. On the other hand, 200 in Berlin, the increase in BD lead to a large decrease in heating loads due to both the increased internal area protected from external conditions, as well as the increased lighting energy and its emitted thermal loads that help decrease heating loads, while increase cooling loads in summer. The 150 result was350a decrease in the overall consumption. 300 100 100 250 50 50 200 150 0 0 100 12m 14m 4m 16m 6m 18m 8m 20m10m 4m12m 6m14m 8m16m 10m18m 12m20m 14m4m 16m6m 18m8m 20m10m 4m12m 6m14m 8m16m 10m18m 12m20m 14m4m 16m6m 18m8m 20m 4m12m 6m14m 8m16m 10m 12m 14m4m 16m6m 18m8 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m 12m 14m 16m 18m 20m 4m 6m 8m 10m 12m 14m 10m 18m 20m 10 ARGAH 10 50 10 CAIRO KHARGAH KHARGAH 0 Building Depth 4m 6m 10 10 10 CAIRO ALEX CAIRO 10 8m 10m 12m 14m 16m 18m 20m 4m 6m KHARGAH KHARGAH 8m 10m 12m 14m 16m 18m 20m 4m COURT CAIRO CAIRO 6m 8m 10m 12m 14m 16m 18m 20m 4m COURT ALEX ALEX 6m 10 10 10 ALEX BERLIN ALEX 8m 10m 12m 14m 16m 18m 20m COURT BERLIN BERLIN 10 Sum of BUILDING LIGHTING (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BU Sum of BUILDING LIGHTING (Kwhr/M2) SumLIGHTING of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING (Kwhr/M2) Lighting Sum of BUILDING (Kwhr/M2) Sum of BUILDING HEATING (Kwhr/M2) Sum of BUILDING COOLING (Kwhr/M2) Cooling Heating COURT Figure 5: Energy in of Kwhr/m2 for courtyard proportion 1:2.5 Sum of BUILDINGuse LIGHTINGintensity (Kwhr/M2) Sum BUILDING HEATING (Kwhr/M2) Sum of BUILDING height COOLING (Kwhr/M2) (building height 10 floors), in case of different building depths in tested cities. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 5 In all cities, BD 4m was of the highest EUI followed by BD 6m, as their S:V ratio were much higher than the other BDs, thus they were more liable to be affected by the outdoor climate. The S:V ratio for BDs 4 and 6m at HP 1:1 for example were 58%, 42% respectively, while BDs 8 to 20m ranged from 18-33% only. In Alexandria, the climate is moderate and close to comfort levels for long annual periods. Cooling loads were not as high as the extreme environment of Khargah because the difference in temperature between indoor and outdoor is relatively small. For that, the effect of BD was the lowest of the four cities. Forms of different BDs other than 4m were of close EUI values. Lighting consumption increased uptill BD 8m then became nearly constant. Savings occured in lighting at low HPs, up to 1:1 (4 floors) and small BDs. The courtyard building with BD 8m was the lowest in consumption at all tested building heights. EUI of BD 6m was nearly similar to the rest of BDs starting from HP 1:0.5 (2 floors). The BD 4m case was of the highest consumption until HP 1:1.5 (6 floors), then became of similar values to BD 10-20m Sum cases. ... Sum ... Sum ... Sum ... Sum ... Sum o... Sum ... Sum ... Sum ... Sum ... Sum ... Sum o... Kwhr/m2 320 320 KHARGAH 300 300 280 280 260 260 240 240 220 CAIRO 220 Sum ... Sum Sum ... Sum ... Sum ... Sum ... Sum ......Sum ... Sum ......Sum ... Sum o... ... Sum o... ... Sum o... Sum Sum Sum ... Sum ... Sum 200 200 Sum ... Sum Sum ... Sum ... Sum ... Sum ... Sum ......Sum ... Sum ......Sum ... Sum o... ... Sum o... ... Sum o... Sum Sum Sum ... Sum ... Sum Sum ... 240 Sum Sum ... 180 Sum ... Sum ... Sum ... Sum ......Sum ... Sum ......Sum ... Sum o... ... Sum o... ... Sum Sum Sum ... Sum ... Sum 240 220 220 300 120 220 1:0.25 Courtyard Height Sum of 1 280 Proportions 200 200 180 180 240 200 CITY Kwhr/m2 260 160 140 140 1:0.5 Sum of 2 1:1of 4 Sum 200 1:1.5 Sum 200of 6 1:2 Sum 200 of 8 KHARGAH - COURT 1:1 - 10m 160 160 1:1 - 14m KHARGAH - COURT 160 KHARGAH - COURT 1:1 - 16m KHARGAH - COURT 1:1 - 18m KHARGAH - COURT 1:1 - 140 20m 160 120 140 100 140 300 200120 280 140 260 160240 220 140 120 100 100 Sum of 1 220 220 200 200 180 CITY 180 FORM 1:0.25 Sum of 1 180 Val... 1:0.5 Sum of 2 1:1of 4 Sum CAIRO - COURT 1:1 - 4m BERLIN 160 160 - COURT 1:1 - 8m CAIRO 140 200 120 120 100 Sum120 of 1 1:2.5 Sum of 10 KHARGAH - COURT 180 180 1:1 - 6m 140 180 Val... 240 320 140 180 120 100 220 220 KHARGAH - COURT 1:1 - 160 12m 200 120 220 240 160 Sum ... Sum ... Sum ... Sum ... Sum ... Sum o... FORM KHARGAH - COURT 1:1 - 180 4m ALEX 180 KHARGAH - COURT 1:1 - 8m 220 160 240 240 320 140 220 240180 Sum o... 240 240 160 Sum ... Sum ... Sum ... Sum ... Sum ... Sum o... 1:1.5 Sum of 6 1:2of 8 Sum 1:2.5 Sum of 10 CAIRO - COURT 1:1 - 6m CAIRO - COURT 1:1 - 10m CAIRO - COURT 1:1 - 12m CAIRO - COURT 1:1 - 14m 140 - COURT 1:1 - 16m CAIRO CAIRO - COURT 1:1 - 18m CAIRO - COURT 1:1 - 20m 120 120 120 100160 100 100 Sum of 1 Sum ofSum 1 Sum of 2 of Sum 2 Sum of 4 of Sum 4 Sum of 6 of Sum 6 Sum of 8 of Sum 8 of 10 ofSum 1 ofSum 2 ofSum 4 ofSum 6 ofSum Sum 8 of 10 100 140 Sum ofSum 1 Sum of 2 of Sum 2 Sum of 4 of Sum Sum 4 Sum of 6 FORM Sum 6 FORM of 8 of Sum ofSum 8 of 10 ofSum 1 ofSum 2 ofCITY ofCITY 4 ofSum Sum 6 FORM Sum 8 of 10 Sum of 10 CITY 120 4m1:1- COURT 4m 4m 1:1 - ALEX 4m 1:1 - ALEX 6m ALEX -Depth COURT 1:1--COURT 4mALEX -ALEX 4m - COURT 1:1--COURT 6mALEX -ALEX 6m - COURT 1:1--COURT 8mALEX - 8m 1:1 - 8m 6m1:1- COURT 6m 6m Sum ofSum 1 Sum of 2 of Sum 2 Sum of 4 of Sum Sum 4 Sum of 6 FORM Sum 6 FORM of 8 of Sum ofSum 8 Depth of 10 ofSum of 10 1 ofSum 2 ofCITY ofCITY 4 ofSum Sum 6 FORM Sum 8 Depth Sum of 10ALEX 8m1:1- COURT 8m 8m CITY Courtyard Height Sum 1:0.25 1:0.5 1:0.25 1:0.5 1:1of 4 Sum 1:1.5 1:2of 8 Sum 1:1of 4 Sum 1:1.5 1:2of 8 Sum 1:2.5 1:2.5 Sum of 1 Sum of 2 Sum of 6 Sum of 10 of 1 Sum of 2 Sum of 6 Sum of 10 10m 12m 14m ALEX -8m COURT 1:1 - ALEX 10m ALEX 1:1 - ALEX 12m 10m 10m 12m 12m 14m 14m ALEX - COURT ALEX 1:1 --COURT 1:1 -ALEX 10m - COURT 1:1 --COURT 1:1- COURT -ALEX 12m- COURT 1:1--COURT 14mALEX 1:1- COURT - 14m 1:1 - 14 4m 4m 4m ALEX COURT 1:1 - ALEX 4m ALEX - COURT 1:1 -10m 6m ALEX - COURT 1:1 -12m 8m ALEX COURT ALEX 1:1 COURT 4m 1:1 ALEX 4m - COURT 1:1--COURT 6m -ALEX 6m - COURT ALEX 1:1 --COURT - 8m 6m1:1 6m 6m 8m1:1 Depth 8m 8m Depth Depth CITY FORM CITY FORM CITY FORM Proportions CITY FORM CITY FORM 4m1:1- COURT 4m DEPTH 4m 1:1 - ALEX 4m ALEX - COURT 1:1 -10m 6m ALEX -1:1 COURT 1:1 -12m 8m -Depth COURT ALEX 1:1 --COURT 4mALEX -ALEX 4m - COURT 1:1--COURT 6m -ALEX 6m -1:1 COURT ALEX 1:1 -1:1 -COURT 8m - 8m ALEX -14m COURT 1:1 - ALEX 16m ALEX 1:1 - ALEX 18m ALEX 10m 12m 14m ALEX - COURT - ALEX 10m ALEX - COURT 1:1 -16m 12m ALEX - COURT 1:1 -18m 14m 10m 10m 12m 12m 14m 14m ALEX - COURT ALEX 1:1 --COURT 1:1 -ALEX 16m - COURT 1:1 --COURT 1:1- COURT -ALEX 18m- COURT 1:1--COURT 20m 1:1- COURT - 20m 1:1 - 20 ALEX - COURT ALEX --COURT -ALEX 10m - COURT 1:1 --COURT 1:1 -ALEX 12m - COURT ALEX 1:1 --COURT 1:1 - 14m 6m1:1 6m 6m 8m1:1 8m 8m Depth DepthALEX 16m 16m 16m 18m 18m 18m 20m 20m 20m BERLIN 1:1--ALEX 4m BERLIN COURT - 6m ALEX COURT 1:1 -10m 4m- COURT ALEX -12m COURT - ALEX 6m 10m 12m 14m - COURT 1:1 - ALEX 10m ALEX -1:1 COURT 1:1 -16m 12m ALEX -1:1 COURT 1:1 -18m 14m 10m- COURT 10m 12m 12m 14m 14m ALEX ALEX 1:1---COURT 10mALEX 1:1 -ALEX 1:1--COURT 1:1 -ALEX 12m -1:1 COURT ALEX 1:1 -1:1 -COURT 14m 1:1 - 14m ALEX - COURT - ALEX 16m ALEX 1:1 18m ALEX 1:1 1:1 - 20m ALEX - COURT --COURT -ALEX 16m - COURT 1:1 --COURT 1:1-- COURT -COURT ALEX 18m- COURT 1:1--COURT 20m 1:1- COURT - -20m 16m 16m 16m 18m 18m 18m 20m 20m 20m Val... Val... Val... ALEX COURT 1:1 -18m 8m- COURT ALEX -18m COURT - 10m - COURT 1:1 - ALEX 16m ALEX COURT 1:1 - ALEX 18m ALEX 20m - COURT 1:1 - 8m ALEX ALEX 1:1---COURT 16mALEX 1:1 -ALEX 16m 1:1 --COURT 1:1-1:1 -ALEX 18m - COURT 1:1--COURT 20m 1:1- COURT - 20m 1:1 - BERLIN 16m 16m- COURT 16m 18m 18m 20m 20m 20m Val... Val... Val... Val... BERLIN - COURT 1:1 - 10m 2 BERLIN - COURT - 12m BERLIN - COURT 1:1 - 14m and ALEX - COURT 1:1 intensity - 12m ALEX 1:1 - 14m for courtyard buildings Figure 6: Energy use in- COURT Kwhr/m of 1:1 different building depths Val... Val... BERLIN - COURT 1:1 - 16m BERLIN - COURT 1:1 - 18m ALEX - COURT 1:1 - 16m ALEX - COURT 1:1 - 18m height proportions. BERLIN - COURT 1:1 - 20m ALEX - COURT 1:1 - 20m Val... Val... In Cairo, whose climate showed higher temperatures than Alexandria and lower than Kharga, results showed some similarity to either cities. As in Khargah, BD was inversely proportional to consumption, however, the differences in consumption between BD cases were much smaller than the corresponding values in Khargah, while larger than those in Alexandria. At BDs 10-20m, a slight decrease in cooling loads occured while lighting loads were nearly the same. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 6 Courtyard or Solid Building: In order to evaluate whether the performance of multi-storey courtyard buildings achieve savings in comparison with solid ones without courtyard, each of the tested building forms was compared to a solid square form with the same built area and no courtyard, Figure 7. In Khargah, results showed that courtyard buildings did not achieve savings in any case of HPs for BD 4-14m, moreover, it lead to a significant increase in consumption. Only at BD 16m, minor savings occurred in case of HP 1:0.25 and 1:0.5 only. Also, minor savings were achieved in BD 18m at HPs upto 1:1. The only case where saving were achieved at all heights was in the BD 20m, especially at up to 1:1 height ratio, while up to 1:2, savings were very small. In Berlin, courtyard buildings did not show any improvement compared to the solid square until BD 16m, at which saving were achieved in nearly all floors. Savings increased as BD increased. In most cases it caused a high increase in consumption that reached 40% in some cases. In Cairo, minor savings were achieved at and some cases of BDs 8m and 16m. At BDs 18-20m, savings upto 6-8% were achieved at low HPs. The courtyard building consumed more energy than its corresponding solid building not exceeding 5% in most of the other cases except for BD 4m at low HPs. In Alexandria, courtyard building achieved energy savings compared to their corresponding solid ones in the majority of cases. In the cases that did not achieve Total savings, the increase in consumption was was less than 4% except for BD 4m at low HPs. This indicated that Courtyard building is more liable to 40% be used in the moderate climate of Alexandria than in other tested cities. 30% KHARGAH 20% Total 10% Total 40% 0% 30% -10% 20% -20% Total CAIRO 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 40% 10% COURT COURT COURT COURT COURT COURT COURT COURT COURT 4m 6m 8m 10m 12m 14m 16m 18m 20m 30% 0% Total KHARGAH 20% -10% 10% -20% ALEX Total Total 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 0% 40% 40% -10% 30% 40% 30% -20% COURT COURT COURT COURT 4m 6m 8m 10m COURT COURT COURT COURT COURT 12m 14m 16m 18m 20m Total Total CAIRO BERLIN 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 20% 30% 20% 10% 20% 10% COURT COURT COURT COURT COURT COURT COURT COURT COURT 4m 6m 8m 10m 12m 14m 16m 18m 20m ALEX Total Total 0% 10% 0% Total -10% 0% -10% -20% 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 1: 0.25 1: 0.5 1: 1.0 1: 1.5 1: 2.0 1: 2.5 -20% -10% 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 Courtyard 1 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 1 2 4 6 8 10 Proportions COURT COURT COURT COURT COURT COURT COURT COURT COURT Building-20% Depth COURT COURT COURT COURT COURT COURT COURT COURT COURT 4m 6m 8m 10m 12m 14m 16m 18m 20m 1 2 44m6 8 10 1 2 46m6 8 10 1 2 48m6 8 10 1 2 10m 4 6 8 10 1 2 12m 4 6 8 10 1 2 14m 4 6 8 10 1 2 16m 4 6 8 10 1 2 18m 4 6 8 10 1 2 20m 4 6 8 10 Courtyard Building BERLIN COURT COURT COURT COURT COURT COURT COURT COURT COURT BERLIN 4m 6m 8m 10m 12m 14m 16m 18m 20m Figure 7: Percentage of change in energy consumption of courtyard buildings compared to its BERLIN corresponding solid square. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 7 CONCLUSION Height proportions had a lower effect than building depth which was a key factor in the cities with extreme cold and hot climates, Khargah and Berlin. EUI values decreased significantly by the increase in depth due to the decrease of exposed surface area with respect to the indoor air-conditioned volume. This BD effect was less in Cairo and nearly insignificant in Alexandria where temperature differences between indoor and outdoor is small, thus decreasing heat transfer by conduction. For a fixed depth, a courtyard with lower height proportions consumed less energy in desert cities due to the effect of the heat sink to the ground which became of less impact as height increased, leading to an increase in EUI accompanied by the increase in artificial lighting and its consequent cooling loads. This nearly cancelled the self-shading effect of the courtyard. The opposite effect occurred in Berlin. Compared to the corresponding solid square, the courtyard building achieved significant savings in the moderate climate of Alexandria especially in case of medium height proportions (1:1) at small BD and in low height at large BD. In khargah and Cairo, that are more hot cities, significant saving were only achieved at large BD (18m-20m) and low height proportions (1:0.25 to 1:1) while a significant increase in consumption occured especially at small BD and higher height proportions in most cases. Further research is required to quantify the effect of courtyard house with more proportions. ACKNOWLEDGEMENTS This research is financially supported by King Abdullah University of Science and Technology (KAUST) as part of the Integrated Desert Building Technologies Project IDBT (Award no.UK-C0015). REFERENCES [1] El-Deeb K., El-Zafarany A., Sherif A. (2012), Effect of building form and urban pattern on energy consumption of residential buildings in different desert climates. PLEA2012 - 28th Conference, Opportunities, Limits & Needs Towards an environmentally responsible architecture Lima, Perú. [2] Al-Hemiddi, N. A., & Megren Al-Saud, K. A. (2001). The effect of a ventilated interior courtyard on the thermal performance of a house in a hot-arid region.Renewable Energy, 24(3-4), 581-595. [3] Sadafi, N., Salleh, E., Haw, L. C., & Jaafar, Z. (2011). Evaluating thermal effects of internal courtyard in a tropical terrace house by computational simulation.Energy and Buildings, 43(4), 887893. [4] Rajapaksha, I., Nagai, H., & Okumiya, M. (2003). A ventilated courtyard as a passive cooling strategy in the warm humid tropics. Renewable Energy, 28(11), 1755-1778 [5] Muhaisen, A. S., & Gadi, M. B. (2006). Shading performance of polygonal courtyard forms. Building and Environment, 41(8), 1050-1059. [6] Muhaisen, A. S., & Gadi, M. B. (2006). Effect of courtyard proportions on solar heat gain and energy requirement in the temperate climate of Rome. Building and Environment, 41(3), 245-253. [7] Muhaisen, A. S. (2006). Shading simulation of the courtyard form in different climatic regions. Building and Environment, 41(12), 1731-1741. [8] Safarzadeh, H., & Bahadori, M. N. (2005). Passive cooling effects of courtyards.Building and Environment, 40(1), 89-104. [9] Aldawoud, A. (2008). Thermal performance of courtyard buildings. Energy and Buildings, 40(5), 906-910. [10] Al-Masri, N., & Abu-Hijleh, B. (2012). Courtyard housing in midrise buildings: An environmental assessment in hot-arid climate. Renewable and Sustainable Energy Reviews, 16, 1892-1898. [11] M. Al-Tamimi, N. A., Syed Fadzil, S. F., & Wan Harun, W. M. (2011). The Effects of Orientation, Ventilation, and Varied WWR on the Thermal Performance of Residential Rooms in the Tropics. Journal of Sustainable Development, 4 (2) 142-149. [12] Morrissey, J., Moore, T., & Horne, R. E. (2011). Affordable passive solar design in a temperate climate: An experiment in residential building orientation.Renewable Energy, 36(2), 568-577. [13] Kottek M, Greicer J, Becck C, Rudolf B, Rubel F. (2006).World Map of the Köppen-Gieger Climate Classification Updated. MeterologischeZeitschrift; 15(3): 259-63. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 8 Improving Ventilation Condition of Labourintensive Garment Factories in Bangladesh Md Mohataz Hossain, MArch Brian Ford, RIBA FRSA Benson Lau, RIBA The University of Nottingham mdmohataz.hossain@nottingham.ac.uk The University of Nottingham The University of Nottingham ABSTRACT The ready-made garment (RMG) sector of Bangladesh is based on the productions from the garment factories where workers are engaged in textile sewing activities, ironing and operating machines. Due to the generally poor quality working environment, these factory workers suffer discomfort and a range of health problems. It is widely known that the thermal environment of workspaces has a direct impact on physical comfort and hence productivity. In the context of a tropical climate, flushing-out the unwanted heat in these deep-planned production spaces is always a major challenge. Mechanical means that annually consume significant amount of energy are usually applied to resolve the ventilation issue. Potentially, passive ventilation strategies within the garment factory buildings may not only enrich the indoor working environment but also reduce carbon emissions. However, research has not yet demonstrated that passive ventilation strategies are viable in this sector. This paper describes an approach that may passively improve ventilation conditions in the existing garment factories of Bangladesh in terms of indoor air quality, thermal comfort; and, potentially, emergency smoke removal. These studies suggest that a methodology to develop passive ventilation strategies within existing garment factories is feasible in this tropical climatic context. INTRODUCTION RMG sector plays an essential role in the economy of Bangladesh, accounting for more than 80% of the total export earnings (Rahman et al, 2008) and nearly 10% of GDP (IFC, 2007). The production space (cutting, sewing and finishing sections) of this sector is usually human labour intensive. The workers’ health, comfort and performance can be influenced by the quality of the production space (NAP, 2010). Hence, optimal working environment is necessary to maximise productivity (Prokaushali Sangsad Limited, 2007). Poor indoor environment has harmful impacts on workers’ health (Wilson and Corlett, 2005) resulting in a high incidence of illness (Zohir and Paul-Majumder, 2008). The most frequent incidences are headache (98%), respiratory problem (36%), vomiting (28%), fatigue (28%) and fainting (18%) (Mridula et al., 2009). These are likely to result from the humid indoor conditions and lack of ventilation of the factories. After the ‘Rana Plaza tragedy’ in April 2013, new ‘Alliance’ and ‘Accord’ between RMG factories in Bangladesh and International organisations have been formed to ensure fire and structural safety in the buildings. However, improving the indoor workspace environments for workers’ safety and comfort is also important. There is a significant amount of heat gain inside the building from the artificial luminaires, workers’ body temperature and constantly in-use equipment (e.g. sewing machines, iron machines, etc.) (Hossain, 2011; Naz, 2008). The resultant gained heat is usually trapped at indoor due to lack of air changes. The factory owners use mechanical means to keep the indoor environment comfortable consuming a significant portion of energy. Local regulatory frameworks (e.g. ‘Bangladesh National Building Code 2006’) generally guide about Md Mohataz Hossain is a PhD student, Faculty of Engineering, The University of Nottingham, UK. Professor Brian Ford is the Chair in Bioclimatic Architecture, Faculty of Engineering, The University of Nottingham, UK. Benson Lau is a Chartered Architect, MArch Environmental Design Course Director and Lecturer in the Institute of Architecture, Faculty of Engineering, The University of Nottingham, UK. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 9 window-floor areas for buildings which may not apply to the deep-planned one and need contextualisation (Ahmed, 2011). About 414 garment workers were killed in 213 factory fires between 2006 and 2009; and workers lost their lives in 2010 (Clean Clothes Campaign, 2012). During fire incidents, indoor trapped smoke is one of the main issues of fatalities (Akther et al., 2010) that correlated to ventilation efficiency. In 2005, Ali showed that workspaces with light-wells of ‘National Assembly Building’ in Dhaka had optimum ventilation performance. Courtyards buildings also have advantages of increased incidence of natural ventilation (Ali, 2007). Increasing openings, soft surfaces and vegetation on facades were indicated as the possible solutions in Ahmed and Roy’s study of 2007, while adding ventilation shafts in residential apartment buildings is a common practice. Even in a still outside air condition, required air flow rate can be achieved by changing opening size and location (Ahmed at el., 2006). Though cross ventilation is suggested in fully humid tropical context (Bay & Ong, 2006), these deep-planned buildings have no provisions of cross-ventilation. Hence, to get a passive solution in existing buildings, main possible solutions are to alter the fabric of the building, to add shaft or atria, to optimise space utilisation and to install control systems (Lush and Meikle, 1988). However, no research has been done prescribing any passive design solution for improving the existing RMG factory buildings in the tropical climatic context of Bangladesh. OBJECTIVE The main objective of the paper is to propose a feasible design approach that may passively improve ventilation conditions in the existing multi-storied RMG factories in context of Bangladesh in terms of indoor air quality, thermal comfort and potentially emergency smoke removal. METHODOLOGY Building selection method As per recent database (May 2014) of the Bangladesh Garment Manufacturers and Exporters Association (BGMEA), a total of 5708 member garment factories are located in Dhaka region: Dhaka, Savar and Gazipur (74.7%), Narayanganj (17.9%) and Chittagong (10.8%). Approximately above 80% of the factory buildings, listed under the recently developed alliance and accord, are multi-storied. Hence, considering the existing building stock scenario, it was justified to choose a multi-storied RMG building within Dhaka region to establish a tangible and replicable outcome. In reference to previous studies (Naz, 2008; Hossain, 2011 and Fatemi, 2012), the major archetype of multi-storied RMG buildings was of ‘shoebox’ shape (either rectangular-oblong or tapered). Hence, after getting shortlisted buildings according to selection criteria, a typical shoe-box shape building within Dhaka region has been selected for the pilot study. Empirical data and physical viability testing method In the site-micro climate analysis, the local meteorological data and updated weather file of Dhaka region along with computer aided tools (i.e. ‘Autodesk weather tool, 2011’ and ‘Climate consultant 4’) have been utilised. ‘Ecotect Analysis 2011’, an established validated tool in previous academic M.Arch and PhD research, has been used for the shadow and solar radiation study only. As a part of Hossain’s research in 2011, a HOBO scientific ‘data logger’ with Dry Bulb Temperature (DBT), Relative Humidity (RH) and Air Velocity sensors (placed in the centre of the 1st floor at 2.1 m height level) was moderately used. Other Information (i.e. numbers of workers, activity types, equipment etc.) have been collected during Hossain’s previous field study in 2011. A calculation tool ‘Opti-VENT’ (developed by Brian Ford & Associates), with contextualising the input data (e.g. deploying the solar radiation data from the Ecotect analysis, design DBT target from Fatemi’s study, 2012), was accomplished to test the physical viability. ‘Bentley Tas Simulator V8i’ has been applied to validate the logged-data and evident the thermal improvement of the workspace. Questionnaire survey and practical viability testing method An online questionnaire survey was conducted to get feedback from the owners and directors of RMG (18 respondents) factories in Bangladesh. The 11 structured questions were formulated to understand their perception on natural ventilation and energy cost, refurbishments and to identify possible constraints. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 10 PROPOSING A PASSIVE VENTILATION APPROACH Site and Micro-climate Analysis The impact of local climate: Plotting the local climatic data reveals that the local DBT varies between 6~37°C within different periods of the year. Hence, the local seasons can be classified into major three categories (figure 1a): warm-dry (DBT 28.08°C), warm-humid: monsoon and post-monsoon (28.08°C and 26.6°C) and cool-dry (19.9°C). During the occupied period of the factories, both outside DBT and solar radiation are relatively high (figure 2b) which can be utilised or controlled for passive ventilation. Figure 1: a) Seasonal variation of DBT b) Daily DBT profile c) Psychometric charts showing boundary of natural ventilation (Source: Climate Consultant 4 and Autodesk weather tool 2011) Noli-plan and possible wind regime Shaded area of the site Figure 2: Location of the building, local wind regime with future development and shadow pattern analysis (source: Google map, Ecotect analysis 2011 and Autodesk weather tool 2011) Phychrometric chart analysis: According to ASHRAE comfort range, comfort can be achieved in at least 9.3% period of the year utilising the natural wind speed (figure 1c). It can be extended by reducing RH or increasing air flow. However, in 2012, Fatemi proposed the garment workers’ higher comfort range of 28.5-33°C BDT and 56-72% RH if the air velocity is 0.8-1.5 ms-1. Hence, passive ventilation in the studied building may still deliver comfortable air temperature for the workers covering more period of the year. Physical context and wind regime: Heavy traffic road at west side (Figure 2) is a source of polluted air and noise. Wind with higher velocity usually approaches from the south, south-east and north side towards the building site during the warm-dry and cool-dry periods. However, considering the future development, wind of reduced velocity may be able to reach to south and north building-facades where major operable openings are also located. Considering sun-path and shadow analysis (Figure 2), the north facade’s wind regime, usually shaded, can be the source of cooler air during daytime working hours (Ford, 2010). Figure 3 also illustrates that the north façade and the ground level area adjacent to a five storied building can potentially deliver cooler air. In contrast, the south façade and roof have higher solar radiations. Hence, these facades require solar control (Akbari, 2007) to avoid external heat gain (e.g. 1000~1800Wh solar radiations). Effect of Street pattern, Vegetation and Urban Heat Island (UHI): The existing street pattern and vegetation reveals that west-side air must be avoided, while south and south-east vegetation is the source of fresh air. Since the site is 26km far from Dhaka city, local temperature can be less affected by UHI effect. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 11 Solar radiation at front road North part has less solar radiation (Cool air source) Direct Solar Radiation at roof level Direct Solar Radiation Indirect Solar Radiation Direct Solar Radiation Figure 3: Solar radiations in the hottest day of a year (source: Ecotect Analysis 2011) Based on the finding in the microclimate analysis and literature review on ventilation principles (table 1), it can be proposed that free running ‘natural ventilation’ can be applicable in cool dry and partially hotdry seasons (40% of the year), while night ventilation and evaporative cooling is also partially applicable in these seasons. Other seasons may need dehumidification due to high level of RH. Table 1: Summary proposal from the findings of micro climate analysis Climatic seasons Months Potential natural ventilation approaches (options) Ventilation principal Hot-dry Mar-May Natural ventilation Night ventilation (Thermal mass) Wind forces Thermal forces Wind from the south West and South façade Warm-humid: Monsoon Jun-Sep Dehumidified cooling Thermal forces Wind from any direction (except the west side) West façade and South facades Warm-humid: Post monsoon Oct-Nov Dehumidified cooling Thermal forces Wind from any direction (except the west side) West and South facades Cool dry Dec-Feb Natural ventilation (with control) Evaporative cooling (Limited) Wind forces Thermal forces Wind from the north (with control strategy) West and South facades 4th floor 3nd floor 2nd floor 1st floor Mezzanin e flo Ground or flo or Exploitation of fresh air and wind regime Required Solar Control Canteen, sewing, finishing. Store Sewing and finishing Cutting section Sewing, finishing and knitting Section Administration Office and Store Dyeing and storage, delivery Figure 4: Functions, heat gains and occupancy density of studied building (source: field survey 2009) Building baseline condition and environmental aspect analysis Considering the work-type, workers’ number, equipment and above all the artificial lighting configuration, it reveals that 1st and 3rd floors have higher heat gain and 1st ~4th floors have high density (figure 4). Moreover, the top floor has higher conductive external gain from roof. To resolve the ventilation issue, ceiling fans and extractors are partially added. The logged-data (figure 5a) clearly reveals that even after having mechanical ventilation, the internal DBT is high during occupied period. Moreover, plotting the field measured DBT in compare to local meteorological data (figure 5b); it can be observed that the heat was trapped inside the production space with a maximum 11degC of indoor-outdoor temperature difference (∆T). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 12 The trapped heat also implies that there was not enough air change rate available in the workspace during the cool-dry season (namely the month of December). An empirical data of Naz (2008) showed that sewing and ironing section could have high DBT of 35°C-39°C with minimum ∆T of 3-5 degC in warm-humid season. Lunch break Unoccupied hours Occupied hours High ∆T due to indoor trapped Higher heat and lack of ventilation outlet Lunch break Occupied hours Unoccupied hours rapped hea t Figure 5: a) field logged data of the 1st floor and b) comparative diagram of DBT in 1st floor (Source: Hossain 2011, previous field survey, weather tool generated data and TAS output data) To sum up, the micro climate analysis and existing empirical evidence shows that thermal principles (considering ∆T and cooler air sources) may be utilised and ‘stack induced ventilation’ can be proposed as a robust solution in this pilot surveyed building. TESTING THE PHYSICAL VIABILITY OF THE PROPOSED APPROACH For effective stack ventilation, three concerned variables are: effective area of the inlets and outlets (A), ∆T and stack height (H), where indoor air flow rate is directly proportional to these variables. Estimated air flow rates can be compared with target design flow rates required for fresh air and comfort cooling; while required air changes are 1~2 ACH and 12~15 ACH respectively (Baker, 2013). For more flow rate, the outlet size and/or the stack height need to be higher. For calculating solar gain, roof surface absorbance, roof UValue, roof external surface conductance are assumed as 0.65, 1.15 W/m2K and 8.5 W/m2K respectively. Figure 6: a) Preliminary design and b) Comparative air flow rates (source: calculation by Opti-vent) Testing by a preliminary shaft design: The outlet sizes can be determined from the ‘required air flow rates’. However, to keep it simple in preliminary design, initial ventilation shaft’s size was determined following the existing beam-column layout (Lomas, 2007). In the preliminary design step, a modular size of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 13 9.2mx7.4m (3.78% of each floor area and 7.4mx2m each of four modules) has been selected and located centrally within the building (figure 6a) believing it may help to equally remove the warm air from all surrounding indoor area. Thus the maximum allowable area of the shaft in each floor is 9.2X7.4=68.08m2 and maximum perimeter is 2X9.2+2X7.4=33.2m. The inlet sizes are determined from the existing opening (35% effective). Considering average outdoor BDT in three seasons (figure 1); three cases were preliminarily considered where assumed ∆T were 5.2, 5.6 and 13.4 degC. An initial shaft height was also assumed (figure 6a) with a stack height of maximum 25.3m in the ground floor and minimum 6m in the top floor. Figure 6b shows that the proposed shaft has met the fresh air flow targets in all seasonal cases. However, the 3rd and 4th floors have not met the cooling targets in warm-dry season so as the 4th floor in the warm-humid case. In cool-dry case, it has been gained in all floors. Increasing stack height may improve the condition. Table 2: Estimation of structural and effective outlets size to test physical viability Structural Case 1: Hot seasons while average ∆T=5 degC (represents warm-dry and warm-humid seasons) inlet (existing Required Effective Effective outlet structural window in structural outlet at shaft-top (50% outlet at N+S sides) (50% shaft top outlet for structural) m2 m2 cooling, m2 structural) m2 Floor 97 97 97 97 49 4th 3rd 2nd 1st Gr+Mez 152.4* 37.7 17.2 29.2 14.7** 76.20* 18.85 8.60 14.60 7.35 125.6* Case 2: Cool seasons while average ∆T=11 degC (represents cool-dry season) 251.2* Required outlet for cooling m2 Effective outlet (50% structural) m2 Effective outlet at shaft top (50% structural) m2 structural outlet at shaft top m2 19.6 10.2 5.1 8.2 4.1 9.8 5.1 2.55 4.1 2.05 23.6 47.2 Required max. shaft perimeter (outletheight is 1.2 m) m 127.0* 31.4 14.3 24.3 15.35** *In 4th floor, the required effective outlet area and shaft perimeter are not feasible to achieve due to high shaft perimeter requirement. **In Ground-Mezzanine floor, the outlet would be a horizontal opening in the ceiling (figure 7). Hence, perimeter has been calculated directly from required structural area (14.7m) assuming area 14.7m2 =4mx3.68m and perimeter 15.35m=2x4m+2x3.68m. Hence, 1.2m height of outlet is not applicable here. TAS modelling of the existing building TAS modelling with the stack-shaft Thermal performance of case 1 Thermal performance of case 2 45 45 40 40 35 35 30 30 25 25 20 20 15 External Temperature (°C) in 9-10 June Dry Bulb (°C) in the 1st floor Dry Bulb (°C) in the 1st floor with the stack-shaft External Temperature (°C) in 20-21 December Dry Bulb (°C) in the 1st floor with the stack-shaft 22:00 20:00 18:00 16:00 14:00 12:00 10:00 08:00 06:00 04:00 02:00 00:00 22:00 20:00 18:00 16:00 14:00 12:00 10:00 08:00 06:00 04:00 02:00 10 00:00 22:00 20:00 18:00 16:00 14:00 12:00 10:00 08:00 06:00 04:00 02:00 00:00 22:00 20:00 18:00 16:00 14:00 12:00 10:00 08:00 06:00 04:00 02:00 15 00:00 10 Dry Bulb (°C) in the 1st floor Figure 7: a) Proposed passive ventilation approach to improve ventilation condition b) Improvement of Thermal performance with proposed shaft (source: TAS simulation) Testing by effective area (A) of outlet: An increased stack height has been assumed where maximum stack height is 30.3m in the ground-mezzanine level and minimum 11m in the top floor. The effective outlets of the shaft (50% of the structural openings in each floor and at the top of the shaft) were actually the free areas to drain warm and stale air. Required cooling flow during warm seasonal conditions always determines the effective free area of opening required (Lomas, 2007). Hence, at this stage, only two cases have been considered, where case 1 and case 2 represent the hotter (∆T = 5 degC) and cooler (∆T = 11 degC) seasons respectively. As the area for the outlet available in the perimeter, rather than the cross sectional area of the shaft, determines its effective area (Thomas, 2007); for variability the maximum effective outlets need to meet two criteria: to achieve cooling flow target in case-1 (38.14~13.39m3/s) and perimeter of the shaft 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 14 would not exceed 33.2m per floor. From Table 2, it can be noted that the outlet size of the 4th floor does not meet the criteria. The primary reasons behind this situation are excessive heat gain (12.9 W/m2), workerdensity (5.6 m2/person) and lower stack height (11m). Apart from the top floor in case-1, the calculations clearly demonstrate that the sizes of the effective outlets are easily achievable to incorporate in the existing building with minimum shaft perimeters between 15.35m and 31.4m (as shown in table 2 and figure 7b). Thermal analysis with ‘TAS’ simulator: Figure 6b reveals that ‘TAS dynamic thermal modelling’ output data has acceptable deviations (maximum 2°C) with the logged-data, which also establishes its validity. Incorporating the effective outlet shafts and achieved air change rate in the pilot studied building (figure 7a), it is revealed that this passive design approach can reduce 2~5°C and 2~7°C DBT (figure 7b) in Cases 1 and 2 (table 2) respectively in the studied first floor. Considering all analyses, it is evident that it is physically feasible to design a central supply route as a shaft of sufficient cross-sectional area to achieve a presumed design air flow rate and reduce indoor temperature as an improvement of the indoor workspace environment within this existing studied building. VIABILITY OF IMPLEMENTING THE PROPOSED STRATEGY From the questionnaire survey, it has been found that the majority of the factories consume 10002000KWh (33%) and above 2000KWh (61%) of electrical energy with an annual average expenditure of US$2500-6250 (39%) and above US$6250 (56%). All the factories are using mechanical ventilation system, though they know it consumes significant amount of electricity. 50% of the stakeholders claimed that they have emergency smoke removal system. The other results (figure 8) reveal that about 72% of stakeholders are inclined to adopt a passive ventilation strategy and 78% would like to undertake refurbishments to improve ventilation condition. 72% may invest US$6250-12500 to implement the strategy subject to an assured return on investment within 5 years. Survey result deployed that the possible challenging issues of execution are construction, disruption of production, existing functional layout and reduction of floor area. Figure 8: Stakeholders’ feedback on adopting passive ventilation strategies and any refurbishment CONCLUDING REMARKS Stack driven ventilation and night cooling can be utilised to improve the air flow rate in all seasons, with a free running period of about 40% of the year, in the studied building in Bangladesh and to save a significant energy cost. A common shaft can be incorporated at every floor to naturally remove the trapped hot air from the production floors. Effective outlet for cooling may always meet the target of air quality. However, improved thermal condition at the top floor is difficult to achieve in warm seasons due to conductive heat gains of roof. Relocating functional zones with less internal gains towards top floor within a building may potentially optimise ventilation performance and shaft-outlet size. Moreover, the inlets may need dehumidified cooling and bio-climatic solar control to ensure cool air inflow in warm seasons. Preliminary shaft size should be determined from the modular structural layout and strength, equipment and work-lane dimension, etc. for efficient space usage. Based on relationship between ∆T and air flow rate, during any fire hazard, the temperature of that floor automatically increases which eventually increase air volume flow speed of the stack ventilation due to the raised ∆T. Furthermore, the hot air containing smoke will naturally travel through the shaft subject to effective air back flow control. This may reduce indoor trapped smoke and workers may, therefore, get an additional time to evacuate from the fire incident floor. RMG factory owners and directors are affirmative about adopting passive strategies and refurbishments to improve ventilation condition. Hence, methodology of implementing proposed passive strategy may be practically viable subject to it is within owners’ budget and payback plan (i.e. cost-benefit analysis) through energy saving and increased productivity with minimum disruption during execution of the refurbishment. An extended field investigation can be developed to observe workers’ adaptive comfort strategy. Emerging passive design cases and associate cost estimations can be part of the extended research with some 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 15 extended questionnaire survey and interviews. This paper has attempted to demonstrate a feasible passive design approach to improve the ventilation condition in existing RMG factories. The authors look forward to pursuing further extended research with more field evidences, sophisticated analyses and larger samples. ACKNOWLEDGEMENTS The authors acknowledge the Commonwealth Scholarship Commission, UK and the University of Nottingham for their continued support. REFERENCE Ahmed, K.S., Haq, A., and M. Moniruzzaman. 2006. Potentials of Window Design in Inducing Air-change in Still Air Condition. Proceedings of PLEA 2006, Geneva, Switzerland. Ahmed, Z.N. 2011. Contextualizing International standards for compliance in Factories. Proceedings of PLEA 2011, Louvain-la-Neuve, Belgium. Akbari, H. 2007. Opportunities for Saving Energy and Improving Air Quality in Urban Heat Islands. Building, Energy and Solar Technology: Advances in Passive Cooling. London, UK: Earthscan. Akther, S., Salahuddin, A., Iqbal, M., Malek, A. and N. Jahan. 2010. Health and Occupational Safety for Female Workforce of Garment Industries in Bangladesh.Journal of Mechanical Engineering,41(1),65-70. Ali, Z.F. 2007. Comfort with Courtyards in Dhaka Apartments. BRAC University Journal, IV (2), 1-6. Ali, Z.F. 2005 ‘Kahn in the Tropics’, International Solar Energy Society Solar World Congress, Orlando. Alliance for Bangladesh Worker Safety 2013. Alliance Factory Profile, November 15, 2013. Retrieved December 2013, 2013, from www.afbws.org Baker, N. 2013. Natural Ventilation: Stack Ventilation. The Royal Institute of British Architects. Retrieved on March 2014. http://www.architecture.com/SustainabilityHub/Designstrategies/Air/1-2-1-2 Naturalventilation-stackventilation.aspx Bangladesh Accord Foundation. 2013. Accord Factory List. Retrieved on March 2014, 2014, from http://www.bangladeshaccord.org/factories/ Bay, J.H., and B.L. Ong. 2006. Tropical Sustainable Architecture: Social and Environmental Dimensions (First Edition ed.). Italy: Architectural Press of Elsevier Ltd. Clean Clothes Campaign. 2012. Hazardous workplaces: Making the Bangladesh Garment industry safe, European Union, retrieve from http://www.cleanclothes.org/resources/publications/2012-11hazardousworkplaces.pdf Accessed on April 2014. Fatemi, M.N. 2012. Study of Thermal Environment in Relation to Human Comfort in Production Spaces of Ready Made Garments Factories in the Dhaka Region. (Unpublished M.Arch thesis), Bangladesh University of Engineering & Technology (BUET), Dhaka, Bangladesh. Ford, B. 2010. The Architecture & Engineering of Downdraught Cooling: A Design Sourcebook. UK: PHDC Press. Hossain, M.M. 2011. Study of Illumination Condition of Production Spaces With Reference To the Ready Made Garments Sector of Dhaka Region. (Unpublished M.Arch thesis), BUET, Dhaka, Bangladesh. IFC. 2007. Ready-made Garments: Challenges in Implementing a Sector Strategy. Dhaka: Monitor, International Finance Corporation, Bangladesh. Lomas, K.J. 2007. Architectural design of an advanced naturally ventilated building form. Energy and Buildings, 39(2), 166-181. Lush, D. and J. Meikle. 1988. Industrial, Retail and Service Buildings- Options for Passive Solar Design. Passive Solar Energy in Buildings. London: Elsevier Applied Science Publishers. Mridula, S.M. and K.A. Khan. 2009. Working Conditions and Reproductive Health Status of Female Garments Workers of Bangladesh, Study Report, Health and Environment Foundation, Dhaka. Munim, J.M.A., Hakim, M.M., and M. Abdullah-Al-Mamun. 2010. Analysis of energy consumption and indicators of energy use in Bangladesh. Economic Change and Restructuring (43): 275–302. NAP. 2010. Executive Summary by ‘The National Academies Press’. Retrieved on December 2013, from http://www.nap.edu/openbook.php?record_id=11233&page=1. Naz, Farah. 2008. Energy Efficient garment factories in Bangladesh. PLEA 2008 – 25th Conference on Passive and Low Energy Architecture, 22nd to 24th October 2008, Dublin, Ireland. Prokaushali Sangsad Limited. 2007. Identification of eco-Efficiency Measures for the Readymade Garments Factories in Bangladesh. working paper no-2, GTZ Progress, Bangladesh. Rahman, M., Bhattacharya, D. and K. G. Moazzem., 2008, Bangladesh Apparel Sector in Post MFA Era: A Study on the Ongoing Restructuring Process. Dhaka: Centre for Policy Dialogue. Zohir, S.C., and P. Paul-Majumder. 2008. Garment Workers in Bangladesh: Economic, Social and Health Condition. Dhaka: Bangladesh Institute of Development Studies. Wilson, J.R and N. Corlett. 2005. Evaluation of Human Work. 3rd ed.: Taylor and Francis. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 16 Impact of native evergreen trees on the visual comfort in an office space in Ahmedabad, India Ankit Bhalla Sanyogita Manu [GRIHA Council, TERI] [CARBSE, CEPT University] ABSTRACT This study investigated the impact of native evergreen trees on the daylight availability in office spaces in Ahmedabad, India. An evergreen tree, native to the hot and dry climate of Ahmedabad, was selected and its impact on daylighting in interior spaces is analyzed compared to a no-tree scenario. The distance of the tree from the window was varied to examine parameters such as contrast and brightness at the task plane for the equinox and solstice days. Desktop Radiance 2.0, which is a backward ray tracing daylight simulation software, was used, followed by a calibration study. Uniform and sunny sky conditions based on Ahmedabad climate data were considered. The results indicate that trees can be very effective in achieving visual comfort in conditions of harsh sunshine outdoors. The type of tree is of more importance for visual comfort than the distance between the tree and the window. The evergreen tree performed well to mitigate visual discomfort. Careful selection of the tree type and its positioning on the southern facade reduced illuminance levels but helped improve visual comfort by almost 50%. This study also explains in detail the method used for determination of Leaf Area Index and Leaf Area Density used for calculating the crown density of the tree, which may help future work attempting to study the impact of vegetation on the thermal or visual performance of building envelope. INTRODUCTION Daylight is considered the best source of light for good color rendering. It gives a sense of cheeriness and brightness which is known to have a significant positive impact on people. Therefore, people desire good natural lighting in their living environments (Li DHW et al, 2006). Ahmedabad is located at 23°N latitude and 72°E longitude, in close proximity to the Tropic of Cancer. It, therefore, falls in a region that receives the highest annual rate of solar radiation. In such a harsh climate characterized by high levels of solar radiation and intense sunlight, appropriate design of windows is critical to minimizing direct sunlight by means of shading and providing diffused daylight reflected from the ceiling. Previous studies recommend using systems that can help to redistribute and filter daylight coming from windows and skylights. Shrubs and trees can achieve this in addition to providing other benefits such as pleasing aesthetics, noise reduction, and passive cooling (Khaled & Ahmed, 2012). Another study states that the shade from the trees reduces not only the direct solar heat gain through the building envelope but also helps to diffuse the light reflected from the sky and surrounding surfaces (Lechner, 2002). A recent study indicates that plants and trees can provide solar shading in the same manner as the jalis, chajjas, awnings, louvres while improving the quality of daylight by scattering direct sunlight and moderating glare the bright sky (Khaled & Natheer, 2009). Trees provide summer shade yet Ankit Bhalla is a Project Officer at GRIHA Council, The Energy & Resources Institute, New Delhi. Sanyogita Manu is an Assistant Professor at CEPT University, and a Senior Research Associate at the Centre for Advanced Research in Building Science and Energy (CARBSE ), CEPT University, Ahmedabad, India. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 17 allow winter access. The best locations for evergreen trees are on the south and south-west side of the building. When trees drop their leaves in winter the sunlight can reach inside to heat the interiors (Kamal, 2012). The penetration of daylight into a building depends on many factors, including the depth of the room from the window wall, ceiling height, internal reflectance value of the room surfaces, window orientation, shape and size, and the optical properties of the glazing. However, the most significant factor is the availability of daylight outside the building, which can be seriously affected by external obstructions like neighboring buildings or trees (Capeluto, 2003). Some studies such as of done by (Manglani, 2001; Gates, 1979; Reinhart & Jakubiec, 2012; Laband & Sophocleus, 2009; Yates & McKennan, 1988) have examined the effect of trees on the heating and cooling loads in buildings in various climatic types. These include quantitative and qualitative analysis of the effects of tree shading, evapotranspiration and wind control. The external radiative exchanges that took place between one tree and a west wall were studied. The study executed by (Manglani, 2001) proposed a methodology for collecting, analyzing and evaluating relevant data for the study of vegetation shade as a means of attenuating the incident solar radiation. The methodology consisted in collecting the values of solar radiation (incoming and outgoing), the air temperature, surface temperatures of the trees and the wall, both in direct sunlight and shade through field measurements and calculating the long-wave radiation flux. Quantification and measurement of the role of trees in scattering sunlight and providing quality daylight in buildings is an area of research that has not been examined closely in the earlier studies. This paper aims to emphasize on the importance of tree shading and provides a methodology to analyze the effect of tree shading on daylight performance and lighting quality. METHODOLOGY Simulation Model For the study, a room of 20m х15m, with fully glazed window of 3m х 20m on the southern facade and placing a mature evergreen tree (Te) at 6m, 9m and12m one by one and comparing it with a no-tree scenario (Tn). Maximum distance of the tree from the window wall is calculated such that the highest point of the tree canopy makes a 45 degree angle with the center of the window sill. Diameter of the tree roots determines the minimum distance of the tree from the window wall. The minimum, maximum distances, and a mid-point between the two, are considered as three points for varying the location of the tree vis-à-vis the wall. For glare analysis, a computer screen is considered as the reference point positioned at the center of the room at task level (0.7m from the floor), facing south (the vertical task screen faces the window while the user faces the wall) in one scenario and facing north (the screen faces the wall while the user faces the window) in the other. A calibration study was also done using a 1:10 scale physical model of the office space, and the lux levels were logged on Mar 13, 2013 at hourly intervals from 0900-1800 hours at the 20 grid points shown in Figure 1. It was observed that from 0900 to 1800 hours the average deflection in the measurement ranges from 6-10%, with the Figure 1: Plan of the calibration model measured readings being higher than the simulated values (Table 1). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 18 Table 1 Percentage difference between simulated and measured values Sensor Points Time (hours) 0900 1000 1100 1200 1300 1400 1500 1600 1700 1800 A1 10% 9% 13% 12% 13% 13% 11% 14% 13% 14% A2 7% 8% 3% 3% 12% 13% 14% 6% 9% 12% A3 12% 3% 3% 8% 0% 1% 1% 4% 1% 4% A4 4% 3% 4% 4% 7% 2% 0% 2% 4% 3% B1 13% 11% 14% 13% 11% 13% 10% 13% 10% 13% B2 9% 7% 4% 6% 13% 4% 3% 7% 12% 10% B3 11% 1% 3% 1% 8% 3% 6% 11% 3% 12% B4 6% 1% 0% 2% 6% 0% 5% 3% 1% 4% C1 12% 14% 15% 14% 12% 12% 10% 14% 11% 15% C2 11% 9% 6% 8% 11% 7% 0% 5% 11% 13% C3 11% 4% 1% 6% 6% 2% 5% 10% 5% 14% C4 3% 3% 6% 1% 7% 0% 9% 5% 3% 3% D1 13% 15% 14% 13% 12% 12% 11% 13% 9% 14% D2 8% 10% 9% 9% 12% 5% 3% 5% 12% 11% D3 2% 5% 4% 6% 5% 5% 4% 13% 6% 13% D4 0% 1% 2% 5% 5% 2% 3% 4% 4% 4% E1 11% 13% 14% 14% 11% 13% 10% 12% 11% 15% E2 11% 10% 13% 7% 10% 8% 5% 8% 10% 14% E3 4% 4% 3% 7% 3% 4% 6% 15% 3% 14% E4 4% 2% 6% 3% 1% 4% 4% 8% 4% 5% (a) (b) Figure 2: (a) Tn (= No Tree) view of the office layout with the analysis grid, (b) Te (= Evergreen Tree) view of the office layout with the tree placed at three distances (6m, 9m and 12m) from the building envelope Desktop Radiance 2.0 was used for daylighting simulation taking from similar studies (Khaled & Ahmed, 2012; Khaled & Natheer, 2009; Gandhi, 2011). Radiance uses accurate ray-tracing technique for generating an image by tracing the path of light through pixels in an image plane and simulating the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 19 effects of its encounters with simulated objects. In the first phase of the study, from a list of all native evergreen trees, one is selected. A simulation model was then developed with tree placement and task location, along with the appropriate material properties for a typical office space. Tree canopy density input was based on the analysis of actual trees. The precision of the 3D models was maintained by matching the canopy density between actual and modeled trees, and using the Leaf Area Density formula (Fahmy et al, 2010; Cantin & Dubois, 2011). Finally, the daylight simulations were run and the impact of tree type and its distance from the window was studied on illuminance, illuminance range and visual quality. (a) (b) Figure 3: (a) Tn (= No Tree) interior view of the office layout with the analysis grid, (b) Te (= Evergreen Tree) interior view of the office layout with the tree placed at a distances of 6m from the building envelope. The simulations were run for sunny sky conditions for Ahmedabad to account for extreme conditions. Test times are selected as representative of conditions during the year. The use of all or any two days of the equinox or two days of solstice are adopted widely by previous studies in daylighting research (Khaled & Ahmed, 2007; Khaled & Natheer, 2009; Hongbing et al, 2010). For this study, two equinox days of Mar 21 and Sep 22, and two solstice days of Jun 21 and Dec 27 are studied in detail. The standard daily office working hours in Ahmedabad are 0900-1700 hours. Simulations were run for 0800, 1000, 1200, 1500 and 1700 hours. Simulation inputs for building material properties and selected trees are described in Table 2 below. Table 2 Material properties for simulation inputs Objects Properties Walls, Window, Ceiling White paint; Reflectance: 70% Floor White paint; Reflectance: 50% Outside Exposed Ground Green Grass; Reflectance: 34% Glazing 10 mm Single pane clear glass with aluminum frame; VLT: 73% Table and Wooden brown laminate; Reflectance: 30% LCD computer screen Single pane black glass; Reflectance: 95% Neem Tree Canopy density: 60% (Mar 21), 70% (Jun 21), 80 % (Sep 22 ), 50% (Dec 27) Height: 12m; Crown diameter: 9m; Reflectance: 31% Selection of Trees References from literature (TCPO, 1980; Krishen, 2006) helped in developing specific criteria for selection of the appropriate tree type as follows:  The tree should have the potential to scatter sunlight and improve lighting quality in indoor spaces; it should not block or significantly reduce illumination levels inside the space. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 20 It should be able to grow and withstand harsh climatic conditions of Ahmedabad with the maximum day temperature reaching as high as 50°C.  It should be able to grow well in areas with mean annual rainfall varying from 400-1200 mm.  The tree roots should not cause damage to building foundations, or, in other words, not extend beyond 5m of the tree spine for the purpose of this study The tree selected for the study was Neem (Azadirachta Indica, henceforth referred to as Te). Tₑ belongs to fabaceae family. It is a slow growing evergreen tree; the average height of the tree is 10-15m. The average canopy density ranges from 80-90% through the year. The tree is found in areas with mean annual rainfall as low as 300 mm (TCPO, 1980; Krishen, 2006).  RESULTS Luminance contrast in the field of view should be comfortable and should improve the visual performance. The following luminance ratios within the field of view were used as the basis of evaluation: 3:1 between task and darker surrounding and 10:1 between task and remote darker surfaces (IESNA, 2000; CIBSE, 2008; Khaled, 2010; Khaled & Natheer, 2009; Khaled & Ahmed, 2007). From Figures 4 through 7, it can be observed that Te at a distance of 6m from the building envelope on Mar 21 (Figure 4) allowed UDI in a range of 62-94% under sunny sky conditions whereas on Jun 21 (Figure 5) the UDI range increased to 73-100%. UDI range on Sep 22 (Figure 6) was 75-95% and dropped to 34-86% on Dec 27 (Figure 7). After the Te was placed at a distance of 9m from the building envelop, a slight decrease in the UDI range (49-94%) on Mar 21 was noted of useful daylight illuminance from 49%-94% under sunny sky condition, but increased to 72-100% on Jun 21. It was stable at 84-90% on Sep 22 and dipped again to 44-86% on Dec 27. When the distance between the tree and the building is increased to 12m, a UDI range of 65-94% was observed on Mar 21, 79-93% on Jun 21, increased to 86-100% on Sep 22. On Mar 21 UDI was observed to be in the range of 0 -39% under sunny sky conditions when there was no tree in front of the window (Tn scenario) while on Jun 21 it to 0-42%. The UDI range on Sep 22 was 54-73% and Dec 27 was 30-61%. Planting an evergreen tree in front of an office window in Ahmedabad Te demonstrated about 50% higher UDI in the space for all the 4 days (Mar 21, Jun 21, Sep 22 and Dec 27) as compared to the notree scenario. It is further observed that in case of Tn the daylighting levels are above 2000 lux for about 90% of the time in the month of March and June and almost 65% in the month of September and December, this leads to visual discomfort for the building users, on other hand plantation of tree Te at multiple distances 6m, 9m, & 12m showcased that the daylighting levels are achieved well within the range of 100-2000 lux confirming visual comfort in the space, the unwanted solar radiations were reflected by almost 67%. UDI graphs determines that trees can provide sun shading and improve the quality of daylight entering through windows by scattering direct sunlight and reducing its intensity while moderating glare coming from the bright sky and confirms that sunlight can be filtered and softened by plantation of trees in front of the fenestration of the building envelope. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 21 Figure 4: Useful Daylight Index, Te on Mar 21 Figure 5: Useful Daylight Index, Te on Jun 21 Figure 6: Useful Daylight Index, Te on Sep 22 Figure 7: Useful Daylight Index, Te on Dec 27 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 22 From Figure 8 (the upper bar “yes” is the time when visual comfort was attained and the below bar “no” indicates the situation when visual comfort was not achieved), it is observed that Te when placed at a distance of 6m from the building envelope demonstrated visual comfort for most of the time on Mar 21 and Sep 22 as compared to Jun 21 and Dec 27. Te at a distance of 9m demonstrated visual comfort for further more hours on Mar 21, Jun 21 and Sep 22 and when Te was placed at 12m distance from the building envelop revealed that under sunny sky conditions the visually comfortable hours increased only for Mar 21 and Sep 22. The scenario where the user was facing the window (and the work screen was facing the wall) performed better on all four days as compared to that where the user was facing the wall. For the Tn scenario in Figure 9, visual comfort was attained for very only a few hours on Sep 22 and Dec 27 under sunny sky conditions. For the rest of the time, both orientations of the work screen lead to discomfort. It is evident from Figures 8 and 9 that Te helped achieve better visual comfort and the desired contrast on the task screen as compared to the Tn scenario due to reduction in luminance distribution area between the luminance iso-contours. The luminance ratio between the task, nearby surroundings and remote darker surface satisfied the recommendations when the user was facing the window. Thus it can be concluded from the figures 8 and 9 that trees have the potential to mitigate acute brightness of the sky perceived through the fenestration of the building envelope and reduce the contrasting luminance ratios, particularly at certain view angles to achieve visual comfort. Yes No Figure 8: Visual comfort for Te scenario evaluated on the basis of the target luminance ratios Yes No Figure 9: Visual comfort for Tn scenario evaluated on the basis of the target luminance ratios CONCLUSION Daylight analysis (luminance ratio and illuminance levels) for a typical office space in Ahmedabad brought to the fore three sources of visual discomfort: acute contrast in luminance between the task surface and background surfaces, high brightness from the windows, and uneven distribution of daylight in the space. Accordingly, it is suggested to plant native evergreen trees at a distance of 9m in front of the southern window to maintain the UDI within the range of 70-75% and provide visual comfort. In future, a study of other types of trees, tree arrangements, and building orientation may be added for further study. It is also important to compare simulation results against actual measurements to validate the model, and develop more appropriate metrics to quantify visual comfort. ACKNOWLEDGMENTS The authors would like to acknowledge Prof. Rajan Rawal, Prasad Vaidya, Jalpa Gandhi, Wayne C. Zipperer, Prof. Khaled A. Al-Sallal, Dr. A. S. Sidhu, Dr. Santan Barthwal and Dr. Priyabrata Santra for their support and guidance. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 23 NOMENCLATURE Te = Evergreen Tree Tn = No Tree UDI = Uniform daylight illuminance REFERENCES Capeluto, G. (2003). The influence of the urban environment on the availability of daylighting. Building and Environment 38 (5), 745 – 752. Cantin, F., & Dubois, M. C. (2011). Daylighting metrics based on illuminance, distribution, glare and directivity. Lighting Research and Technology, 43 (3), 291–307. CIBSE. (2008). CIBSE (Chartered Institution of Building Services Engineers). London, UK: Page Bros (Norwich) Ltd., Norwich, Norfolk NR6 6SA. Fahmy, M., Sharples, S., & Yahiya M. (2010). LAI based trees selection for mid latitude urban developments: A microclimatic study in Cario, Egypt. Building and Environment, 42(2), 345-357. Gandhi, J. (2011). Thermal effect of vegetation on urban microclimate in hot-dry region: Taking a case of Gandhinagar Central Business District. Ahmedabad. Unpublished Thesis, Faculty of Design, CEPT University, Ahmedabad. Gates, R. (1979). Energy Conservation as a Passive Solar System. Proceedings of Fourth National Passive Solar Conference. Hongbing, W., Qin, J., Hu, Y., & Dong L. (2010). Optimal tree design for daylighting in residential buildings. Building and Environment, 45 (12), 2594–2606. IESNA. (2000). The IESNA Lighting Handbook, Ninth Edition. Retrieved May 28, 2013, from http://www.scribd.com/doc/46634221/IESNA-Lighting-Handbook Kamal, M. A. (2012). An Overview of Passive Cooling Techniques in Buildings: Design Concepts and Architectural Interventions. Acta Technica Napocensis: Civil Engineering & Architecture, 55(1), 8497 Khaled A. Al-S & Ahmed, L. (2007). Improving Natural Light in Classroom Spaces with Local Trees: Simulation Analysis under the Desert Conditions Of the UAE, Proceedings: Building Simulation, 1168-1174. Khaled A. Al-S & Ahmed, L. (2012). Improving Natural Light in Classroom Spaces with Local Trees: Simulation Analysis under the Desert Conditions of the UAE. Retrieved October 5, 2012, from http://www.ibpsa.org/proceedings/BS2007/p027_final.pdf Khaled A. Al-S & Natheer A. (2009). Effcts of Shade Trees on Illuminance in Classrooms. Architectural Science Review, 52(4), 295-311. Khaled A. Al-S. (2009). Practical Method to Model Trees for Daylighting Simulation Using Hemispherical Photography. Eleventh International IBPSA Conference Glasgow, Scotland, 280285. Khaled A. Al-S. (2010). Daylighting and visual performance: evaluation of classroom design issues in the UAE. International Journal of Low-Carbon Technologies, 201-209. Krishen, P. (2006). Trees of Delhi: A Field Guide. London, UK: Dorling Kindersley. Laband, N. & Sophocleus, P. (2009). An Experimental Analysis of the Impact of Tree Shade on Electricity Consumption. Arboriculture & Urban Forestry, 35(4), 197–202. Lechner, N. (2002). Heating, Cooling, Lighting: Design Methods for Architects. New York: Nostrand Reinhold. Li DHW, Wong S., Tsang C., & Cheung G. (2006). A study of the daylighting performance and energy use in heavily obstructed residential buildings via computer simulation techniques. Energy and Buildings, 38(11),1343-1348. Manglani, P. (2001). Shading Effects of Trees On Building Surfaces: A Radiative Exchange Analysis. Unpublished Thesis, Arizona State: College of Architecture and Environmental Design, Arizona State University. Reinhart, C. & Jakubiec, J. (2012). The ‘adaptive zone’ – A concept for assessing discomfort glare throughout daylit spaces. Lighting Research & Technology, 44 (2), 149-170. TCPO. (1980). Guide on Plant Materials for Landscaping in India. Town and Country Planning Organization New Delhi: Government of India. Yates, D. & McKennan, G. (1988). Solar architecture and light attenuation by trees: conflict or compromise . Landscape Research, 13(1), 19-23. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 24 Session 1B : Low carbon cities and neighborhood development PLEA2014: Day 1, Tuesday, December 16, 2014 11:30 - 13:10, Compassion - Knowledge Consortium of Gujarat Achieving Best Practice Net-Zero-Energy Building Design Instruction Methods Thomas Spiegelhalter, MArch Shahin Vassigh, MArch [Florida International University, Miami] tspiege@fiu.edu [Florida International University, Miami] svassigh@fiu.edu ABSTRACT The United Nation's climate panel has published the third part of its long-awaited report on strategies for greenhouse gases (GHG) mitigation in 2014. The document by the Intergovernmental Panel on Climate Change (IPCC) considers the options for limiting or preventing GHG emissions and enhancing wide ranging activities that remove them from the atmosphere. For the building sector, numerous energy efficiency and GHG reduction market changes, new design and learning algorithms for more efficient simulation tools and benchmarking procedures have been developed. For example, the mandatory E.U. ‘nearly Net-Zero-Energy-Building 2018 regulation’ for all new public and privately owned buildings is now set up to help minimizing carbon emissions and reverse the negative impact. In the U.S., the American Institute of Architects (AIA) adopted the 2030 Challenge as a voluntary program, where participating buildings aim to achieve a 90% fossil fuel reduction by 2025, and carbon-neutrality by 2030. The following paper presents the outcomes from a funded project by the U.S. Department of Education under the topic of Building Literacy: the Integration of Building Technology and Design in Architectural Education. The funds supported the interdisciplinary development of a hybrid educational platform comprised of software and a hard copy textbook for advancing Net-Zero-Energy Building design. The most significant challenge was to select the best practice design variables for landscape and climate, building orientation and occupancy types, passive-active energy and climate control systems and their dynamic impact on each other. The paper will critically discuss and analyze the project implementation and the diverse feedback of multiple users from the profession and academia for further improvements for the second edition. HYBRID PROJECT APPROACH The instruction methods for both the book and the DVD “Best Practices for Sustainable Building Design” are developed for students, faculties, architects, engineers and everyone who is interested to apply best practice principles for Net-Zero-Energy Building designs (Figure 1). For instance, in today’s struggling economies practicing architects are faced with lower in-service training costs. For that reason with better prepared graduating students, architecture firms will not have to incur the costs of providing technical training to new hires. This book with the DVD supports and promotes a self-directed form of learning, which not only is more effective than the traditional method, because it also offers greater flexibility and links to other online animation tools. The book and the DVD are structured into seven learning modules: landscape/climate and building form, building structure and envelopes, climate building control, renewable energy and lighting. The combined text and animation modules allow the user to quickly grasp the various, but interconnected concepts of passive and active strategies influenced by microclimatic conditions, building form, envelope materials and environmental Authors are professor at the College of Architecture + The Arts at Florida International University, Miami, Florida, USA 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 25 control systems, and other elements, ensuring an accessible step-by-step learning process. The book and DVD animation project also addresses a number of critical international issues as it aims to significantly improve the effectiveness of teaching sustainable and net-zero energy efficient design within architecture programs. In particular, the originally funded U.S. - Building Literacy pedagogy is designed to increase students’ comprehension, problem-solving capacity, and most importantly, ability to apply learned principles to carbon neutral design. The following diagrams are DVD animation excerpts about different Net-Zero-Energy building typologies in different climate zones within the Climate Control Module (Fig.1). Figure 1. Excerpt of the Interactive Software: two Zero-Net-Energy building analysis tools for average hot and cold climates with passive and active building integrated renewable systems and their prospective Energy and CO2 per m2 a year performance. Source: Spiegelhalter/Ozer/Vassigh. DESIGN VARIABLES AND TYPICAL SELECTION CHALLENGES The book and software components of the educational framework have been developed as an immersive and integrated learning environment, delivering the content in a combined interactive format. Harnessing the capabilities of other advanced media, such as dynamic parametric modeling with other regular software tools, animations, interactive diagrams, and hypertext, the book DVD’s software generated environment helps to visualize and engage concepts that may be difficult to grasp in traditional learning formats. The interactive content aims to engage and compel users to participate actively in the learning process. The inclusion of the software component also responds to the proclivity of the new generation of students who are more accustomed to accessing information in such environments. The software content is organized under a graphical user interface system that serves as a vehicle for learning on demand, linking to proper information quickly and easily. All the content of the book is cross-referenced to the accompanying software with graphic icons at important reference points. The icons are used to inform the user that there are interactive diagrams, charts, and animations explaining the subject in greater depth in the accompanying software (Figure 2). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 26 The biggest challenge is posed by the selection of the design variables in context to the interconnectivity of building systems and their impact on each other. For example, how the selection of a certain type of cooling and dehumidification system will lead to maximum efficiency when combined with a particular structural thermal mass system, or how the choice of a specific structural enclosure system can affect the levels of natural lighting within the space and therefore impact the building’s netzero-energy profile. Since there are many building systems choices and a great number of combinations possible, leading students to learn about optimized solutions with so many variables and without a complex parametric computation seemed a difficult task to achieve. To face this challenge the authors had to limit many variables and choose strategically only those for general demonstration purposes (Figure 2). Another significant challenge was the complexity of the architectural design process itself. Designing a building refers normally to a wide range of socio-cultural, aesthetic, and economical constraints. For example, to address the challenge that a sustainable building could be among one of the worst socially and culturally conceived buildings and may not work for the occupants at all. Although the authors had to make a decision to deliberately exclude all these other factors and limits the project expression of architectural form within the tool. However, these constraints still evolve and engage many of the authors and book/DVD user’s discussions. The selection, development and organization of the entire content from various discipline areas of architecture, engineering and landscape architecture under one umbrella and providing instantaneous access to the vast amount of information occupied a significant amount of the project efforts. It is clear that not all topics could be covered as forementioned. Figure 2. Screen shots from animations showing passive system choices on the left and one example of encapsulated phase changing materials as a pop up animation menu. Source: Spiegelhalter/Ozer/Vassigh. BUILDING FORM AND ORIENTATION Building form has a critical impact on the well-being of the users, resource use and on the water and energy consumption. The choice of form includes the shape, volume, mass, scale, and configuration of a building. Form should also support the requirements for the users’ activities. Although building form may often be determined based on a number of other concerns outside energy efficiency and sustainability, selecting the proper form is one of the most important steps in net-zero-energy design. A properly conceived building form can mitigate the external climate in order to provide comfortable interior conditions, thereby reducing cooling, heating, ventilation, and electrical lighting demand. In 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 27 particular, the building’s environmental performance in relation to its formal configuration can be determined based on another number of factors including plan geometry, surface area to volume ratio, orientation, access to natural light, natural ventilation, and the location of the structural core and circulation spaces (Figure 3). The book software interface addresses these issues in each of the climatic zones of 1) Hot and Humid, 2) Hot and Arid, 3) Temperate and 4) Cool. Figure 3. Screen shot showing thermal performance of a C-shaped building in a hot and humid climatic zone. Source: Spiegelhalter/Ozer/Vassigh. BUILDING ENVELOPE AND THE LIFE CYCLE OF MATERIALS The building envelope is the primary interface of a building with the exterior surroundings. As a result, the building envelope plays a critical role for the thermal comfort of the users and in the energy management and greenhouse gases emission mitigation. The proper selection of walls, roof and floor systems, construction materials and a rigorous, detailed development of connections and structural joinery are important components of thermal comfort, and water and energy saving strategies. In addition, thermal and moisture control, sound and fire insulation, and natural lighting strategies can significantly reduce dependence on mechanical climate control systems. . Figure 4. Screen shots showing the environmental impact in the use of materials for construction. The diagrams include embodied energy and water and the carbon footprint. Source: Spiegelhalter/Ozer/Vassigh. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 28 The life cycle of materials entails the consumption of energy and water in their removal, manufacturing, transportation, maintenance and recycling and produce hazardous emissions during these processes. Approximate indicators of a material’s environmental impact are embodied energy, embodied water and carbon footprint. Designing buildings with improved environmental performance should go beyond decreasing the operational energy, and aim at reducing embodied energy, embodied water and carbon footprint during the life cycle of building materials. BUILDING STRUCTURES Selection of structural materials and systems for buildings is often based on material efficiency and reducing the structural components to the smallest possible size without compromising safety. Although this process promotes effective use of natural resources, other strategies that utilize materials with highrecycled content and reduced impact on the environment are significant ways in which the structure can become more effective in reducing the buildings’ carbon footprint. The concepts of structures are summarized to 1) Structural materials and their properties, 2) Horizontal Spanning Systems, and 3) Vertical Spanning Systems (Figure 5). Figure 5. Screen shot showing a wooden structural materials and frame systems. Source: Spiegelhalter/Ozer/Vassigh. CLIMATE CONTROL SYSTEMS Designing resource efficient buildings with active, passive or hybrid means of achieving comfort requires a thorough understanding of climatic conditions, building occupancy types, and the availability of resources. Although there is a wide range of mechanical means for controlling the interior conditions of buildings, they present significant drawbacks to the natural resources and the environment. While it may be unrealistic from case to case to completely move away from the active methods for climate control, investigating passive means of ventilating, daylighting, heating and cooling buildings are critical (Figure 6). The book texts and software categorizes the passive climate control systems into: 1) Natural Systems, 2) Solar Heating, 3) Passive Cooling, 4) Phase Change Materials, and 5) Building Insulation. The study of active climate control systems includes: 1) Production Systems, 2) Distribution Mediums, 3) Distribution Methods and 4) Recovery Systems. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 29 Figure 6. Screen shot of an animation showing a latent heat process and energy exchange cycle. Source: Spiegelhalter/Ozer/Vassigh. RENEWABLE ENERGY Fossil fuels are non-renewable as they draw on finite resources that are diminishing. These fuels are becoming increasingly more expensive and produce irretrievable damage to the environment, and with that impacts, as well, human health and the survival in climate change threatened societies. In contrast, renewable energy resources are constantly replenished and their capacity to replace conventional fuels has significantly increased during the past decade at the global scale. In its various forms, renewable energy sources include natural elements such as sunlight, wind, biomass, tides and geothermal heat. Energy harnessed from these elements can be used to produce electricity, heating and cooling energy for buildings operations. The book software investigates selected renewable energy forms in five modules of 1) Solar Thermal Convection, 2) Photovoltaic Systems, 3) Wind Energy, 4) Geothermal Energy, and 5) Energy Storage Systems. Biomass is not included and will be integrated in the next book edition. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 30 Figure 7. Screen shots from animations showing, a) solar façade integrated photovoltaic system. b) heat pipe conductor. Source: Spiegelhalter/Ozer/Vassigh. NATURAL AND ARTIFICIAL LIGTHING Since thousands of years architecture has embraced natural lighting as an important component of spiritful, healthy and inspiring aspect of buildings. Combined with daylight control systems, effective daylighting also saves energy and avoids greenhouse gases. It is well-known but not often practiced and implemented in building designs that bright, ambient living space or workplace can improve quality of life, improve user productivity and reduce buildings’ lifecycle cost, while minimizing the adverse impacts on the environment. However, using natural light as the only source of illumination is not always possible and various levels of artificial lighting are often required per occupancy type and specific building code requirements. The use of artificial lighting in buildings can account for a significant portion of the buildings’ electric energy consumption. Using artificial lighting strategies with occupancy sensor infrastructure technology, dimmable efficient lighting systems and control devices, can reduce the electric energy demand significantly. This section therefore divides the study of light into natural lighting and artificial lighting. The book software organizes the study of natural lighting into seven modules; 1) Day Lighting, 2) Glazing, 3) Climate zones, 4) Side Lighting, 5) Top Lighting, 6) Light Shelves, and 7) Light Redirection Systems. The Artificial lighting modules include; 1) Light sources, and 2) Zoning by light. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 31 Figure 8. Screen shot showing the concept of sun light angles and paths with in relation to window sizes, room geometries and surface reflectances of the choosen materials. Source: Spiegelhalter/Ozer/Vassigh. LANDSCAPE DESIGN AND SYSTEMS Landscape design should be considered as an essential and integral component of a holistic approach to sustainable net-zero-energy building design. Landscape elements and systems have a major impact on water, energy and resource management. Remarkable reductions in non-renewable resources consumption can be realized within a building and in its surrounding site by simply using landscape elements for the avoidance of heat islands, mitigation of heat transfer through shading and increase of natural ventilation strategies through soft cape designs with temperature and pressure differences. The book software interface organizes landscape systems into four major segments: Concepts, Thermal Efficiency, Hyrological Efficiency and Case Studies. The segment Thermal Efficiency concentrates on the effective use of landscape elements to mitigate climatic conditions in building projects. Figure 9. Screen shot showing the impact of urban heat islands in different climate zones and different surface materials and plants. Source: Spiegelhalter/Ozer/Vassigh. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 32 CONCLUSION AND OUTVIEW The project outlined describes a step-by-step pedagogical platform designed to teach basic resource conservation and concepts for designing carbon-neutral or nearly net-zero-energy buildings. Perhaps the real strength of the platform is the combination of book and interactive software formats. The funded project by the US Department of Education required a summative and formative project evaluation. This was accomplished through a component collected responses from the project team, faculty, and participants during the project development phase and beta testing. The information was used to provide feedback to the authors in order to improve the project. The summative evaluation measured the effectiveness of the software by analyzing comparative student performance at the State University at Buffalo (UB) and Florida International University (FIU) from 2009 to 2012. In all those tests, the animation DVD helped to visualize and engage with the exposed concepts that may otherwise be too ambiguous, too boring or too difficult to comprehend in only a text-image book format. In summary, the results showed all experimental groups exposed to the software pedagogy and tool displayed statistically significant improvement between entry and exit test scores, while the control groups not exposed to the tool showed no significant improvement. The future work to improve the next edition will include more interdisciplinary testing methods and educational games with interactive quizzes. Another emerging question for improvement is how the future of academic and professional education will change when it comes to the increased usage of text/web-based software for tablets versus textbooks. The benefits are clear, that text/web-based software for tablets can store and process more learning materials than textbooks. Most tablets today have memories between 20GB and 100GB, which can hold hundreds of thousands of textbooks. This means a single tablet is more than capable for holding all the textbooks a learner needs plus animations, quizzes and homework. This poses the critical question if we then still need physical text books in the future? Will this lead to a fundamental paradigm change in education and practice? What are the disadvantages of digital learning with tablets and other handheld devices? Some say there are a number of health issues caused by the heavy usage of tablets: eyestrain, blurred vision and headaches, to name a few. These are symptoms which are collectively known as Computer Vision Syndrome. Or others observe disadvantages in the use of tablets that students who use them tend to get too distracted, as opposed to those who use textbooks. Some state that distractions are related to digitally connected students who are simultaneously into games, videos, emails and countless entertainment applications while they are in a learning environment. In summary, all the fore-mentioned and perhaps many more constraints would be worthwhile to further investigate to improve the pedagogy of “Achieving Best Practice Net-Zero-Energy Building Design Instruction Methods.” REFERENCES Vassigh Shahin, Ozer E., Spiegelhalter Thomas (2012). Best Practices in Sustainabe Building Design, J. Ross Publishing, Fort Lauderdale, Har/DVD first edition. All new buildings to be zero energy from 2019 (Nov. 16, 2012). Retrieved from http://www.europarl.europa.eu/sides. AIA Agenda 2030 Initiative (June 15, 2011). Retrieved from: http://www.architecture2030.org/. Rachke K. (2003). The digital revolution and the coming of postmodernist University, Routlege Falmer, London, p. 38. Schlaich, J. (1991). Practices Which Integrate Architecture and Engineering, In Bridging the Gap: Rethinking the Relationship of Architect and Engineer, Webster, A. (Coord. Ed), Van Nostrand Reinhold, New York, p. 109-122. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 33 A LOW ENERGY COMMUNITY? A comparative study of Eco-Villages around the world Nilesh Bakshi Prof Robert Vale Prof Brenda Vale Victoria University of Wellington ABSTRACT This paper investigates what is being done to produce sustainable community developments to minimize ecological footprint. Five international case studies were compared with the Govardhan Eco Village in Maharashtra, India. The study describes each case study and then looks at how various sustainable principles have been integrated into the community. Each case study was compared to an appropriate set of sustainability indicators to see which parameters were addressed. In order to establish the fundamental sustainable design focus of each case study, whether technology or human behaviour, the analysis looked at the types of parameters governing each project. Results showed the parameters incorporated in the case studies did not obviously change with time. Further scrutiny of the parameter matrix for all case studies suggested two distinctly different trends in the ‘eastern’ and ‘western’ examples. The Indian example appears to show true sustainable development, relying less on technology and more on human capital. INTRODUCTION As identified by Holling (2000) and substantiated by Ewing and Moore (2010) there is growing awareness of the massive changes that human societies are causing to the environment, implying a need for a fundamental change in lifestyles (Holling 2000). In 1987 the Brundtland Report, Our Common Future, first defined Sustainable Development as human development that "meets the needs of the present without compromising the ability of future generations to meet their own needs” (United Nations General Assembly, 1987) and supported policies such as the adoption of the Kyoto Protocol; the founding of Agenda 21 (agenda for the 21st century) in 1992; and the 1997 formation of the Global Reporting Initiative (GRI). A more recent populist influence was the film, An Inconvenient Truth, by former American Vice President Al Gore (Gore, 2006). Most energy consumption and pollution originates in the lifestyles of the citizens of developed countries (Mithraratne, 2013) and this has caused some members of the architectural profession to rethink design principles, resulting in new „sustainable‟ developments, as discussed here. The 1992 Earth Summit in Rio de Janeiro led the way by outlining a set of actions for sustainable development, Agenda 21(United Nations Environment Programme, 1992). To implement this, the United Nations instituted a “set of indicators of sustainable development” to monitor progress (Bell and Morse 2008). Robert Vale & Brenda Vale are professors in the School of Architecture, Victoria University of Wellington, New Zealand. Nilesh Bakshi is a PhD candidate in the School of Architecture, Victoria University of Wellington, New Zealand. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 34 PROBLEM STATEMENT AND OBJECTIVES Reducing energy use in new and existing buildings is an urgent necessity. Energy consumption in the building sector is more than one-third of national energy use in India (CARBSE 2014), and with further growth in this sector, India, along with other developing countries, faces a formidable challenge in reducing its dependence on fossil fuels. Biocapacity, referred to as the new wealth of nations (Ewing and Moore, 2010), is the ability of ecosystems to provide the resources people need and absorb the wastes they create. At 58.2 million global hectares (gha), New Zealand has the world‟s largest biocapacity, and uses only 39.4% of it. With an ecological footprint of 22.9 million gha (Ewing and Moore, 2010), New Zealand is also unhappily responsible for one of the world‟s larger ecological footprints of 5.17 gha per person (shown in figure 1). As shown in figure 1 the global „fair share‟ footprint, which all people could have without overtaxing resources, is 1.7 gha per person as opposed to the average of 2.7 the world is using currently, made possible by using finite resources such as coal, oil and uranium (WWF, 2010). Should India, whose biocapacity is 630 million gha and whose ecological footprint is only 0.9 gha per person, (Global Footprint Atlas 2014) be following the example of countries like New Zealand, which take far more than their fair share of the Earth‟s resources? This study is a comparison of current sustainable design practice in India and the developed world to identify if, and how, India might learn from the successes and failures of „developed‟ countries. The aim is not only to determine how varying nations identify and assess sustainable design but also whether there has been a shift in focus in sustainable developments. The best example of a sustainable development in India (Govardhan Eco Village) will be compared with similar privately funded examples from New Zealand (Earthsong), Germany (Solarsiedlung), the United Kingdom (BedZED), Japan (Fuji Eco-park Village), and Australia (Crystal Waters), countries which have roughly similar ecological footprints, all well above the fair share (Figure 1). Figure 1 Case study selection based on footprint (USA included for comparison only). Source: Global footprint network (2010) The developments were selected based on self-reported claims of sustainability that purport to follow guidelines determined by Agenda 21 (United Nations Environment Programme, 1992) as discussed by Bell and Morse (2008) and by Bakshi (2009). CASE STUDIES Table 1 identifies the self-reported case study parameters ordered by their occurrence in the six developments to give an overview of the sustainable design emphasis in each case. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 35 Table 1: Parameters in occurrence order Case study one: Earthsong Eco-Neighborhood, New Zealand (2008) Earthsong is based on the three principal areas of sustainable design (Environmental, Social and Economic) as established in „Agenda 21‟, (Bell and Morse, 2008). Environmental consideration at Earthsong relates to the overall design of buildings and landscape into a coherent whole; orientation of all buildings on principles of passive solar design and natural climate control; building materials choice considering embodied energy, durability, toxicity, recyclability and environmental impact; collection and re-use of rainwater and waste water; renewable energy; solar hot water systems; and clustering of buildings creating productive open spaces (Earthsong 2014). The „village‟ arrangement balances the needs of individuals and community. Physical spaces encourage a diverse range of social interaction; cars are confined to the outer regions of the site; varying dwelling sizes support a wide range of ages and household types; a centrally located common house is a focus for community activities; and resident management occurs through the body-corporate. A self-sustaining economy to allow creation of on-site work and wealth was achieved through shared workshop and office facilities; leasable multi-purpose workspaces (stage 2); reduced domestic energy costs from energy efficient and passive solar design; reduced commuting costs through the site being 500 metres from a railway station. The Earthsong Eco-Neighborhood‟s incorporation of some onsite renewable energy and permaculture makes it a more sustainable co-housing project. Sustainable materials were also considered important, with careful management of non-renewable resources. As stated (Earthsong Econeighbourhood 1999) this required; • • • • Specifying sustainably grown timbers as an integral part of each building; Using coating systems/carpets/thermal insulation with reduced environmental impact; Incorporating some major recycled building products. Encouraging better management of water as a precious resource through rainwater harvesting and having water efficient devices and grey-water recycling. Case study two: Solarsiedlung (The Solar Village), Germany (2008) Solarsiedlung GmbH is a sustainable scheme that incorporates „energy-plus‟ initiatives to produce more energy than it consumes (EEG, as of 1st April 2000). Apart from terraced housing it has an office/housing block based on German Passivhaus and Plus Energy House directives. Environmental consideration comes from material selection, appliance choice, energy plus initiatives and limiting cars to the outskirts of the site. The scheme has four main elements (Behling & Schindler 1996); 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 36 1. Use of environmentally friendly materials (wood from regional forests, PVC-free, environmentally friendly insulation) 2. Highly insulated thermal building shell (heat requirement: 11-14 kWh/m² per year) 3. Active ventilation with heat recovery 4. Photovoltaic panels on the roofs The distance between buildings allows passive solar heating and solar electricity generation. Solar electricity is fed into the grid and extracted as needed. Any additional energy required in the winter is provided by a local wood-chip fuelled combined heat and power station (Solar Architecture 2014). Rainwater passes through a “biotope” for purification and to relieve pressure on local storm water drains, with the aim of recharging the city groundwater. Some rainwater is captured for garden irrigation and toilet flushing. The buildings have large south-facing glazed openings to maximize solar gain and small openings to the north to minimize heat loss. Vacuum insulation is used on opaque portions of the facade. Triple glazing reduces heat loss through windows and glazed facades (Behling & Schindler 1996). Because this project produces more energy (from renewable sources) than it consumes it can be considered to be sustainable design. Case study three: Crystal Waters Eco-village, Australia (1988) In 1996, Crystal Waters in Australia received a United Nations World Habitat Award for its “pioneering work in demonstrating new ways of low impact, sustainable living” (Barton 2013). There were six design objectives at Crystal Waters: 1. 2. 3. 4. 5. 6. Clean air, water and soil (thus food) Freedom in spiritual belief To work towards a guarantee of meaningful activity for all To create a place for healthy play and safe recreation Active social interaction Healthy shelter Crystal Waters accommodates up to 300 people with 80% of the land set aside for agriculture, and steeper areas given to forestry, recreation and natural habitat.The houses are built of a variety of materials such as straw bale, rammed earth, poles and mud domes, and avoid rainforest timbers and toxic chemicals. The EcoCentre (for education) has rammed earth walls, photovoltaic power, rainwater collection, and a biolytic waste system. The 83 residential lots are arranged in clusters to encourage interaction, co-operation and a sense of belonging. Scheme aspects show careful consideration for the environment, social needs and the economy. Food is grown onsite and most residents maintain home gardens and orchards. The scheme has „home occupation‟ zoning, which creates business zones within the residential areas (Barton 2013). The permaculture design (Mollison and Slay 1994), which has a diversity of plants and wildlife uses knowledge of eco-systems to create a more sustainable way of life. The hydrological balance was maintained, ensuring the quality and quantity of the water downstream was not negatively affected by the development (Mollison and Slay 1994). Seventeen multi-purpose dams provide a flood mitigation strategy by absorbing runoff, with the overflow directed into the rivers via swales (Barton 2013). They are also a source of emergency water. A long term sustainable approach is taken to forestry. Trees were planted to provide habitat and moderate environmental extremes, as well as for various timber end uses. Buildings make extensive use of renewable materials with particular emphasis on passive solar design (Barton 2013). Case study four: Fuji Eco-park village, Japan (2000) Fuji Eco-park village is a project with a perfected “Permaculture Master Plan” (World Permaculture projects 2014). The project was scrutinized for its environmental and financial selfsufficiency; and its ability to be self-replicating and because it achieved all these it was included in a 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 37 global list of sustainable permaculture projects (World Permaculture projects 2014). It was also recognized as an example of sustainable living in the Eco-village conference in Japan 2005 (GEN 2014). The Fuji Eco-park village is an eco-housing development with a single main dwelling, the Centre House, built on passive solar principles on 7.5 acres of land. Environmental consideration is given to transportation options, appliance choice, materials selection, energy consumption, and construction methods. The scheme has four main elements (Permaculture News 2014); 1. 2. 3. 4. Renewable agriculture Renewable energy Sustainable design for the built environment A water conservation scheme The implementation of all four aspects makes this project unique. The layout is based on wind and solar orientation for renewable agriculture and renewable energy collection. There are solar PV panels and onsite wind generated electricity. The site layout has a main area for vegetables, an animal farm and a wind turbine farm. Because the scheme revolves around a single Centre House there was ample space for PV panels at ground level, making them easier to maintain. The wind farm and PV panels are said to produce more than enough energy for the single dwelling. To achieve renewable agriculture 90% of the land was devoted to food self-reliance (World Permaculture projects 2014). This led to the creation of a potato field, blueberry farm, pumpkin field, sweet corn field, herb gardens, earth worm farm, bee keeping farm, compost area and the orchard (Permaculture News 2014). Case study five: BedZED (Beddington Zero Energy Development), United Kingdom (2002) The BioRegional committee defines BedZED as the “UK‟s largest mixed use sustainable community” (Bio-regional 2009). It further states the scheme is one of the most successful projects to follow Agenda 21 in the United Kingdom. The project was aimed to be a zero energy design, only relying on energy from onsite renewable sources (Desai and Riddlestone 2002), and was designed to incorporate techniques to utilize resources sustainably. The three main principals of design were (Desai and Riddlestone 2002); 1. Zero fossil energy 2. Maximise passive solar energy 3. A “pedestrian first” policy After its evaluation by the BioRegional committee (BioRegional 2009), the scheme was shown to be as not as successful in performance as intended. BedZED is the only one of the six case studies for which there is a detailed record of performance in practice. Identified in the report by BioRegional (2009), BedZED‟s design reduces space-heating requirements by 88%, one of the major design features that was intended to contribute to this is the multicolored wind funnels that provided passive ventilation (BioRegional 2009), but these had been rendered inoperative by rust a few years after the building opened (Vale 2014). Hot-water consumption was 57% less then the UK average; the electrical power used, at 3 kWh per person per day, was 25% less than the UK average; 11% of this was produced by solar panels; mains-water consumption was reduced by 50%; and the residents' car mileage is 65% less than the UK average. There are 777 m2 of solar PV panels. Tree waste was to fuel the development's cogeneration CHP plant to provide district heating and electricity, although this system is currently not working and electricity comes from the grid (BioRegional 2009). All dwellings face south to take advantage of solar gain, incorporate triple glazing, and have high thermal insulation (Desai and Riddlestone 2002). Rain water is collected and reused. Appliances are water-efficient and use recycled water. Low-impact materials were specified and these come from renewable or recycled sources within 35 miles of the site, to minimize embodied energy. This scheme incorporates collection facilities to support recycling. To encourage eco-friendly transport, electric and liquefied-petroleum-gas cars have 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 38 priority over petrol and diesel cars and there are parking spaces for charging electric cars (Desai and Riddlestone 2002). All these design considerations mean people living in BedZED have a reduced ecological footprint (identified by BioRegional as 4.7 gha compared to the UK average of 5.5 gha) and this makes the project somewhat more sustainable (BioRegional 2009). Another very important aspect of BedZED is the fact that it also strongly promotes behaviour and lifestyle changes such as joining the car club which reduces the carbon emissions of occupants‟ whole lifestyle by 50% (BioRegional 2009). Encouraging behavioural change through governance and estate management, make BedZED one of the earlier eco-villages that is beginning to identify the importance of behaviour. However, communal elements like the CHP plant and “Living Machine” system to treat waste water have failed whereas those at a household scale (PV, gardens, passive solar, water) have worked (BioRegional 2009). Case study Six: Govardhan Eco Village, India (2003) Govardhan Eco Village aims for “Simple living and High thinking”, and demonstrates practical ways of achieving a sustainable lifestyle. The scheme revolves around 5 elements (Govardhan 2014); 1. 2. 3. 4. 5. Natural Buildings Sewage Treatment Organic Farming Alternative energy from solar energy and biogass Education for sustainable behaviour Govardhan Eco Village is a farm community spread over 60 acres at Galtare, Wada, 110 km North of Mumbai that incorporates use of alternative energy, eco friendly construction, and sustainable living. Alternative energy sources such as biogas and PV panels reduce dependency on the national grid (GEV 2013). The green building scheme considers existing ecologies, ensuring the constructional choices do not impact the surrounding site negatively. This is achieved by having separate building and planting zones. Use of materials 90% sourced from within a 100 kilometre radius (GEV 2013b), coupled with zoned construction, hinder unsustainable behaviour by limiting material choices possible in the built architecture. There is also a waste management system, linked with biogas generation on site, and water harvesting (GEV 2013c). Finally organic farming is the sole source of food making this project, like Fuji Eco-park in Japan, self-sufficient in food. As shown in table 1, most if not all the parameters for Goverdhan Eco Village focus on the behaviour and impact of the occupants, with less reliance on innovative technologies for generating energy. Everything from choices made during construction to the way occupants eat and sources of food are governed by local resources. FINDINGS Table 2: Parameters based on Agenda 21 sustainability indicators in date order 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 39 In Table 2 the order of the parameters has been changed to place those parameters that can be considered „technology based‟ towards the top and those that can be considered „behaviour based‟ towards the bottom. The trend line in table 2, which follows the mid-point of the parameters in each case study shows that no obvious change based trends in the parameters are observable over time. This suggests the focus on technology versus behaviour is arbitrary. In the next step (Tables 3 and 4) the case studies were segregated and grouped. Grouping the „eastern‟ case studies (Japan and India) seperately from the „western‟ case studies (England, Germany, Australia and New Zealand) created two distinct trends, showing a definite shift in focus. Table 3: Eastern Model Table 4: Western Model In Table 3, the Eastern Model, there is a shift over time towards incorporating more parameters for sustainable behaviour, education and food self-sufficiency. The Western Model (Table 4) shows an increasing reliance on technology. Crystal Waters is most similar to the eastern models and also one of the oldest western case studies (26 years old). Like the former it is self-sufficient in food but unlike the eastern case studies (and the other western ones) it does not have good access to public transport and other local facilities, suggesting sustainable settlement design in the west might only follow the eastern model if the site is rural and remote. This also suggests that the differences may be to do with having a suburban rather than a rural location. However, this would require further research and an extended list of case studies, as well as more detailed studies of measured performance. The trends depicted in the findings of this paper are established through a very small sample, especially for the eastern set. In addition, it would be worth carrying out much more detailed studies to see the measurable degree of sustainability offered by each settlement, perhaps using ecological footprint, as for BedZED, the only one of the six which is relatively fully documented or per capita energy consumption. CONCLUSION This research shows that most (if not all) the case studies appear to consider themselves in alignment with Agenda 21. However this „tick the boxes‟ approach to assessment does not highlight a more fundamental difference in the approach to sustainable development as shown by the Eastern and Western Models. The former, with less reliance on technology and more on human capital, could be viewed as more sustainable than the Western Model. It has been shown by other research that significantly greater reductions in ecological footprint and therefore increased sustainability can be achieved by behavioural change than by technology (Vale and Vale, 2013). Further detailed investigation of ecological footprint and other measured aspects of performance in use, such as per capita energy demand, would show if this is indeed the case. Given the state of the world, it is perhaps the case study from India that is showing the way to true 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 40 sustainable development, especially for developing countries, with the focus on sustainable education, behaviour and food self-sufficiency. As BedZED shows, reliance on „innovative‟ technology can be misplaced, whereas where the community is responsible for its own food and dealing with its own wastes, based on human capital, there is less chance of failure. REFERENCES Bakshi, N. (2009). Our Sustainable Side. School of Architecture. Victoria University of Wellington. NZ. Bell, S., & Morse, S. (2008). Sustainability indicators: Measuring the immeasurable?. London: Earthscan. BioRegional (2009). BedZED seven years later. http://www.bioregional.com/files/publications/bedzedseven-years-exec-summary.pdf Retrieved 24 May 2014 Barton, H. (2000). Sustainable communities: The potential for eco-neighbourhoods. London: Earthscan. Behling, S., & Schindler, B. (1996). Sol power: The evolution of solar architecture. Munich: Prestel. CARBSE (2014). Center for Advanced Research in Building Science & Energy (CARBSE). http://cept.ac.in/179/center-for-advanced-research-in-building-science-energy-carbse- Retrieved 24 May 2014 Desai, P., & Riddlestone, S. (2002). Bioregional solutions for living on one planet. Totnes, Devon: Green Books for the Schumacher Society. EEG. (2000). Renewable Energy Sources Act. http://www.erneuerbare-energien.de/en/topics/acts-andordinances/renewable-energy-sources-act/eeg-2012/ Retrieved 05 June 2014 Earthsong Eco-neighbourhood. (1999). Earthsong Database resource section: Design Brief. http://www.earthsong.org.nz/docs/DesignBrief.pdf. Retrieved 24 May 2014 Earthsong Eco-neighbourhood. (2014). Earthsong Information section: About Earthsong. http://www.earthsong.org.nz/about.html Retrieved 24 May 2014 Ewing B. & Moore D. (2014). Ecological Footprint Atlas 2010, Oakland CA: Global Footprint Network. Fujieco. (2014). Fuji Eco-park home page. http://www.fujieco.co.jp/ Retrieved 24 May 2014 Global footprint network. (2010). Ecological footprint atlas. Retrieved 05 June 2014. http://www.footprintnetwork.org/en/index.php/GFN/page/ecological_footprint_atlas_2010 GEN. (2014). How Sustainable is Your Community? Community Sustainability Assessment. Global Ecovillage Network. Retrieved 25 May 2014 GEV. (2013). Alternative energy. http://ecovillage.org.in/Case-Studies/Alternative%20energy.pdf Retrieved 24 May 2014 GEV. (2013b). Green Building Technology. http://ecovillage.org.in/CaseStudies/Green%20Building%20Technology.pdf Retrieved 24 May 2014 GEV. (2013c). Waste Management. http://ecovillage.org.in/Case-Studies/Waste%20Management.pdf Retrieved 24 May 2014 Gore, A. 2006. „An Inconvenient Truth’. Pennsylvania: Rodale Press. Global Footprint Network. 2009. Ecological Footprint Standards 2009. In J. Kitzes (Ed.). Oakland, CA. Govardhan 2014. Govardhan Eco Village website. http://www.ecovillage.org.in/ Retrieved 24 May 2014 Holling, C.S. 2000 „Theories for sustainable futures’. Conservation Ecology, 4. Kachadorian, J. (1997). The passive solar house. White River Junction, Vt: Chelsea Green Pub. Co. Mollison, B. C., & Slay, R. M. (1994). Introduction to permaculture. Tyalgum, Australia: Tagari Publications. Permaculture news. (2014). Permaculture Research institute of Australia. http://permaculturenews.org/ Retrieved 24 May 2014 Solar Architecture. (2014). The Solar Settlement in Freiburg. http://www.rolfdisch.de Retrieved 24 May 2014 United Nations General Assembly. (1987). „Report of the world commission on environment and development: Our common future, chapter 2: Towards sustainable development’. United Nations Environment Programme. (1992). Agenda 21 - 10.1 Integrated Approach to the Planning and Management of Land Resources.: United Nations http://www.unep.org/Documents.Multilingual/Default.asp?DocumentID=52&ArticleID=58. Retrieved 24 May 2014 Vale R. and Vale B. (eds) (2013) Living within a fair share ecological footprint. Abingdon, Oxon: Routledge. Vale R. (2014). Personal communication September 2, 2014. World Permaculture projects (WPP) (2014) Japan – Fuji Eco-Park Village http://www.permacultures.com/wp/2010/05/22/japan-fuji-eco-park-village/ Retrieved 24 May 2014 WWF 2010. Living Planet Report 2010. Gland, Switzerland: World Wide Fund for Nature. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 41 Spatial Structure of City Blocks with Vacant Lands in Edo, Early Modern Tokyo - Introducing the Appropriate Wind into Outdoor Living Spaces – Masahito TAKATA, PhD Akira HOYANO, PhD [Kumamoto University, Japan] [The Open University of Japan, Japan] Email: takata.m@arch.kumamoto-u.ac.jp ABSTRACT One of the main factors of urban heat island phenomena is surface temperature, and surface temperature is mainly affected by spatial geometry and material of cities. The townsmen’s areas in the city of Edo, early modern Tokyo, were totally different from the present day metropolis in these factors. A series of papers evaluated the summer thermal environment of these townsmen areas using numerical simulation, in order to acquire knowledge for designing an environmentally symbiotic city in Asia. In the previous paper, summer surface temperature and sensible heat flux from surface in a typical Machiyashiki (= city block) was evaluated using numerical simulation. The following result was obtained; the nighttime heat flux was negative, indicating that the townsmen residential areas in Edo were never subject to the nighttime heat island phenomenon. Although the residential areas were notorious for their densely populated and low-rise wooden buildings, some vacant lands were scattered in Machiyashiki temporally, according to historical materials. In this paper, summer thermal and wind environment in several city blocks with vacant lands, located next to the previous target site, was calculated using coupled analysis of heat balance and airflow, concerning the influence of south wind from the bay. The result showed that, in outdoor living spaces, people could attain moderate wind, which was blowing down to the south-north alley through vacant lands and gardens, in the evening. From this, it was confirmed that some vacant lands and spatial structure of Machiyashiki in Edo made it possible to introduce the appropriate wind into outdoor living spaces under the comfort thermal radiant environment on a specified time section. The daily changes of sensible heat flux from surface in several city blocks were also calculated, and the result showed that these vacant lands had effect on reducing daytime heat island phenomena. 1. INTRODUCTION The urban heat island phenomenon has become a serious modern-day problem. Surface temperature is one of the main factors that determine this phenomenon, and mainly affected by spatial geometry and building materials used in urban areas. The city of Edo was a metropolis that existed from the 1790s to the 1860s, after which it was renamed as Tokyo. The urban residential areas in Edo differed greatly from those in present-day Tokyo in terms of their spatial geometry and building materials. The objective of this study was to acquire a fundamental understanding of how to design a better summer thermal environment based on the Edo townsmen areas for an environmentally symbiotic city in Asia. [1], [2] In previous studies , the summer surface temperature of Machiyashiki (city block) in Edo Masahito TAKATA is an Assistant Professor in Department of Architecture and Building Engineering, Kumamoto University, Japan. Akira HOYANO is a Professor in Faculty of Liberal Arts, The Open University of Japan, Japan. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 42 [3] townsmen residential areas was calculated by numerical simulation using the information gleaned from literature and historical materials of that time. The following result was obtained; the nighttime heat flux load was lower, indicating that the townsmen residential areas in Edo were never subject to the heat island phenomenon. Although the Edo townsmen residential areas were notorious for their densely populated and low-rise wooden buildings, some vacant lands were scattered in Machiyashiki temporally because of frequent fires, according to historical materials (Fig. 1). In addition to the constant south wind from the bay, these vacant lands might have affected the thermal and wind environment of outdoor living spaces in these city blocks. In this study, the summer thermal and wind environment in several Machiyashiki was calculated using the coupled analysis of heat balance and airflow. The influence of vacant land in city blocks was also evaluated. N
[m]
40
0
②/ ③/ ⑨/ st/ e7Cho/1 inab Minam ⑫/ ⑦/ Takiyama7Cho/ ⑩/ ⑥/ ④ ⑤/ ⑪/ ⑧ h/ ut
So ind W Vacant/Land
Owned/Vacant/Land
Alley
Gardens
Greens
Longhouse
Townhouse
Warehouse
①/ Minaminabe7Cho/2nd/ Previous/Target/siteMoriyama(Cho+ Figure 1. Spatial composition of the Target Area and Major Outdoor Living Area 2. SELECTING MODELS AND ARRANGING INFORMATION FOR SIMULATION 2-1. Selecting Simulation Models To reproduce the thermal and wind environment in outdoor living areas, the reproduction model calculated three components: the distribution of (1) wind velocity, (2) surface temperature, and (3) air temperature. For complicated spatial structures in Machiyashiki, the coupled analysis of heat balance and [4] airflow, proposed by Hoyano (2007) , was suited for this study. This analysis consisted of two [3] and generalized CFD software simulation models: a 3D-CAD-based thermal simulation tool (STREAM). The 3D-CAD-based thermal simulation tool could calculate the detailed distribution of the surface temperature in consideration of the airflow distribution. The turbulence model used in the generalized CFD software carried out a steady-state analysis in a high Re k-ε turbulence model, which had enough precision on account of the average airflow in Edo Machiyashiki. In this method, the output condition of one simulation model is made the input condition of the other model until the calculation converges with sufficient precision. 2-2. Arranging Information for the Coupled Analysis [4] According to the proposed method , the necessary and sufficient information for the coupled analysis was arranged and the information was classified as known and unknown. The insufficient information for the coupled analysis, especially related to CFD, was identified (Table 1). In this section, this unknown information is presented as obtained from previous studies and historical materials. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 43 Table 1. Necessary Information for the Coupled Analysis '--$!$.$*) /$'$)"- *1)#*/- /$'$)" '*)"$)"- )*1)
)!*,(.$*) *((*)
$.
')
 -/,  $.
')
$)

$-.*,. 
$.  ')
' ( ).
**(
2*/. ')
 ''$)"
+  ' 0.$*)
.,/./, 
$( )-$*) *!
#
( ( ,
$.#
**! . ,$' /,!
. ,$'
''
**! /,!
. ,$'

,*-,*--
 .$*)
# ,('
#2-$'  .$*)

 ''$)"
+  ,*+ ,.$ - +.$'
*,(  -/, 
*,(
 ,2$)"
,  #*+
/,.$)- . ,$' , ) *((*) $'$.$ - */.
, --/,
*--
*/),2
#*+ /,.$)- *((*)
$.
')
 -/,  ')
' ( ).
' 0.$*) .,/./, 
$.#
**!  /,!
. ,$'
,*--
 .$*) # ,('
#2-$'
,*+ ,.$ -    ')
' ( ).  . ,$' /,!
. ,$'
,*--
 .$*) # ,('
#2-$'
,*+ ,.$ -  ,*/) /,! +.$'
*,( ')
' ( ). . ,$' /,!
. ,$'
,*--
 .$*) /,!
. ,$'
,*--
 .$*) # ,('
#2-$'
,*+ ,.$ # ,('
#2-$'
,*+ ,.$ ., .-
'' 2- ).
')- , + $ - *((*)
*$' . *((*)
 '' /-.$) /$'$)" . ,$' */.
 1"
/.. , +.$'
*,(

/! +.$'
!*,( , ) ,*/)
/,! ' ( ).
*!
, )- /,!
. ,$'
,*--
 .$*) /,!
. ,$'
,*--
 .$*) # ,('
#2-$'
,*+ ,.$ , )-  +.$'
*,(  -/, 
*,(
$")*,- /,!
. ,$'
,*--
 .$*) /,!
. ,$'
,*--
 .$*) # ,('
#2-$'
,*+ ,.$ # ,('
#2-$'
,*+ ,.$  ,2$)"
, 
#*+
/,.$)- $")*, +.$'
*,( ')
 -/, 
*,( . ,$' % .- )&)*1)
)!*,(.$*) +.$'
*,( */.
+.$'
*,( ** )
. *)
.#
., .- . ,$' */.
** )
! ) ')
' ( ).  ,  ,
$"#.
,*1)
$.#
,/)&  $$"#.  ,"
* !!$$ ).
*!
, , )
*0 ,"  $.
')
 -/,  /,!
. ,$'
,*--
 .$*) 
+ $- - 
 ,"
* !!$$ ).
*!
..$ +.$'
*,( $.
')
 -/,  . ,$' /,!
. ,$'
,*--
 .$*)  2-2-1. Signboards in the Streets [5] Ito (2003) classified the placement and frequency of different objects in the arterial streets of the [6] Edo townsmen area from the pictures in Kidai-Shoran . Based on this study, the objects were classified into six groups from a total of 89 townhouses and three signboards were selected as having a high number. In this study, eaves, signboards, and standing signboards were selected as located signboards [7] (Fig. 2). The size of the signboards was set as 1212 mm depending on Edofunai-Ehonfuzokuourai , which describes the life and customs of townsmen in Edo. Eaves' Standing' Located' Species' signboard' signboard' signboard
%
22%
13%
9%
Roof' Head9on' Leaning' Without' signboard' 'signboard''signboard''signboard' 47%
5%
3%
1%
Images'of'Signboard' in'Kidaishouran+ Reference:'Unknown','Kidaishouran+,+Berlin'Orient'art'museum'possession','1804' Figure 2. Objects in the Arterial Streets (e.g., Signboards) 2-2-2. 2. Ground Covers in Vacant Lands [8] According to Ito (1986) , vacant lands in the city blocks were used as drying areas, scrapyards, or open space. In areas where it was hard to identify the things placed in the field, the use for vacant lands [9] was classified as open space. Based on previous studies in weed science and the summer weather on a [1] particular day , it was determined that the vacant lands were covered in weeds. In this study, vacant lands were classified as open space with green fields. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 44 2-2-3. Greens in the Townhouse Garden Kaso-sho, which was a textbook on house physiognomy, was very popular in Japan at that time. In this study, the information contained in the textbook was used to identify the species and placement of [10] greens in townhouse gardens . The placement of greens was decided s follows: divide the garden into nine sections which was same way as at the time, and decide which greens are suitable for each section according to good and bad luck suggested in the kaso-sho (Fig. 3). The size of the greens was decided by depending on studies and documents obtained from the Japanese Institute of Landscape Architecture. Plant1Peach
Ware1 2house
Townhouse
Alley
1
m
0
(m)
2.5
Street
2.5
m
Cannot1Plant1 :1Next1to1Building
Division1into19
N
Shrubbery
Don’t1plant1in1center
Figure 3. Placement of Greens in the Garden according to Kaso Kaso-sho sho 2-2-4. 4. Residents’ Summer Living Activities in Vacant Lands For the evaluation of outdoor living areas, the summer living activities on vacant lands during each time section at that time were identified from previous studies and historical materials. According to Ito [8] (1986) , vacant lands in the Edo townsmen area were used as drying areas during the day and used to [7] enjoy the evening breeze during the evening. In Edofunai-Ehonfuzokuourai , there was a description about vacant lands which stated that, “many vendors and food stalls had been opened up from evening till night, and all the people live in Machiyashiki got together to enjoy cool evening breeze.” From this information, the vacant lands in the target site could be set as outdoor living areas that some residents occupied during the morning, and all residents occupied to enjoy cool breeze during the evening. The identified information was added into the table of outdoor living space of residents, proposed in the [2] previous study (Table 2). Table 2. Outdoor Living Space of Residents (Including Vacant Land) 1)
,*/)! )-&!"


 &0&)$
,")!**,.*1)%*/-"
 *)$%*/-"6  
4 *,)&)$
*,)&)$

4 **)
 **)
 **)
0")&)$
&$%.
4  
4  
4  
4   
4  
4  
4              5 5
 3  5 5 *((/)'
,"-  3    5 5  5 5  ).
)!-    5 5   
3  3  .,"".- 5 5 5 5 5   
3  3  5         5 ''"2- /.-&!"
 "-&!").-
'--
", %).7", %).-
1&#"7,#.-()7)/'
*,",7,#.-()-
1&#"
3
-+ "-
,*/)!
*((*)
1"''
3 -+ "-
,*/)!
*/.!**,
") %"-580 ).
3. PRECISION OF REPRODUCTION MODELS AND BOUNDARY CONDITIONS 3-1. Concerning Precision of Reproduction Models Considering the width of streets and the lengths of the members of buildings, which were different depending on the type of outdoor living spaces, the precision of the reproduction model was set. The buildings in the city of Edo were measured using the Japanese measuring system. The simulations in this study were calculated in a structured grid. As a result, the minimum width of each mesh in CFD was 300 mm for streets and 100 mm for alleys. Finally, a model of the target site was created for this simulation (Fig. 4). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 45 Warehouse
e
hous Town Standing%Signboard% 0.3m×0.3m
Tree
Drying%Area
Shop%Curtains% height%0.3m
Located%Signboard% 0.6m×0.6m
Inner%Wall
3.6m
Tatami%mat% 1.5m
height%0.3m
Inner%Wall
.2m
Street
9.0m
6.0m ay
Walkw .8m
1 1.2m ace
lling%Sp Garden
Se 
Shop%curtains%and%located%signboards%were%exchanged% for%wooden%shuKers%at%night%(including%early%evening).
Figure 4. 3D-CAD Model for the Simulation 3-2. Setting Boundary Conditions on Building Parts The coupled analysis needed to calculate not only values of thermal physical properties but also boundary conditions of building members. Although the thermal physical properties of building [1] members had been prepared in the previous study , boundary conditions of building members had to be considered. As the predominant wind direction in the streets was arranged vertically with respect to the wooden lattice gate, the gate served as a substitute for a grid-like panel in CFD, and the pressure loss [11] boundary of the gate was determined depending on the architectural design data corpus . The parameters (coefficient, LAE, green coverage ratio) necessary for calculating the effect of trees in CFD [12] were set depending on Yamada (1982) . 4. EVALUATING THE INFLUENCE OF VACANT LANDS ON THERMAL AND WIND ENVIRONMENT IN EDO MACHIYASHIKI In this chapter, the distribution of surface temperature, wind velocity, and air temperature of the [1] [4] target site was calculated on a clear summer day using the coupled analysis , and the influence of vacant lands in Machiyashiki on the outdoor thermal and wind environment was evaluated. 4-1. Calculation of Surface Temperature, Air Temperature, and Wind Velocity The surface temperature at the target site on a clear summer day was calculated using the 3D-CAD[3] based thermal simulation tool . The room and air temperatures were maintained at the same value because the buildings at that time were not airtight. Depending on the average vertical thermal distribution calculated previously, the ground temperature at time 0:00 was set as the underground thermal boundary condition. The simulation was calculated over five days and was run for the first 4 days under the conditions mentioned above. The distribution of surface temperature on the fifth day was regarded as the periodical steady state adapted to CFD as input data in the coupled analysis. Airflow was calculated using generalized CFD software (STREAM). The nested grid method was adapted to the airflow analysis to incorporate the influence of surroundings. Table 2 lists the boundary conditions used in this calculation. First, in the wide area, only the airflow analysis was adopted with a minimum mesh size of 900 mm. The wind velocity and direction were considered as boundary conditions (inflow and outflow) for the middle area. Next, in the middle area, the coupled analysis was adopted with a minimum mesh size of 300 mm. The air temperature, wind velocity, and wind direction were considered as boundary conditions for the middle area. Finally, in the narrow area, coupled analysis was adopted with a minimum mesh size of 100 mm. The boundary conditions were the same as those for the middle area (Fig. 5). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 46 Table 2. Boundary Conditions of the Coupled Analysis 2&0;6.65&2,* .))0*&5*& %.)*&5*& 81'*53+*6- := ;= < :=;= *6-".<*1.2  11 #85'80&2713)*0 "7&2)&5)/>3)*0 0,35.7-1 "!*,80&51*7-3) .++*5*27"(-*1* 2+039'382)&5; .++*5*27.2*&(- < ? 11 11 $ %.2)).5*(7.32"387%.2)"4**) 16"7&2)&5)-*.,-7 1 :432*27  /  1 6 >   1 6 87+039'382)&5; "85+&(*5*6685* & "/;".)*'382)&5; +5**60.4 Fiigures(are( nested(from(the( analyses(in(Middle( area(and(Narrow( Area
"30.)685+&(* 2360.4 "85+&(*#*14*5&785* &0(80&7*)+.,85*6'; '&6*)7-*51&06.180&7.327330 Wide
N
z
&5539&5*&  1= 1= 1  1=  1= 1 Inflow
Middle
Wide
y
Narrow
Target Site x
Surroundings 54m 1080m Middle
6.1m/s ( height 15m : South ) Wind Direction & Velocity (Current of Air Distribution : Wide Area) Narrow
Wind Direction & Velocity Air Temperature 1080m (Current of Air Distribution : Wide Area) Outflow All: Surface Pressure Boundary (0Pa) Figure 5. Analysis Range (Wide, Middle, Narrow) 4-2. Surface Temperature and Heat Flux from Target Site Both during the day and at night, the surface temperature of the roof in long houses (Japanese [1] cedar) and townhouses (clay tile roofing) had the same tendency as was found in the previous study . Because of the low thermal conductivity and thermal capacity, the surface temperature of Japanese cedar and clay tile roofing was more than 55°C (up to 65°C) during the day and 1–3°C lower than the air temperature at night. The surface temperatures of vacant lands fell from 35°C to 25°C at night. Although there were trees inside Machiyashiki, the trees had little effect on creating shade because of their size. The sensible heat flux from each of the 17 Machiyashikis in the target site was evaluated from the [13] heat island potential (HIP) , an index that indicates the amount of heat transferred to the surrounding air from an area on a typical day in terms of the heat island phenomenon (fine weather and low wind speeds). Note that all the vertical unevenness of the surfaces is also regarded to be part of the horizontal surface area. The index was calculated using the equation below (Eq. (1)). …  (1)   (1)     Ts: surf surface temperature of a minute plane inside   the area of interest (ºC), Ta: air temperature in the area for interest (ºC), A: horizontal projected area of the target site (m 2 ), ds: area of a minute plane (m2) The difference in the spatial structure of each Machiyashiki was considered from the daily change in the HIP (Fig. 6). The sensible heat flux from all Machiyashikis were extremely high during the day and low at night, which means that most of the solar radiation at the target site was immediately radiated back toward the surrounding air. Therefore, during the summer, the heat island phenomenon never developed in the city blocks of Edo at night with or without vacant lands. However, the daytime HIP of Machiyashiki with vacant lands was getting lower in inversely proportional amounts to the area of vacant land during the day (5–15°C lower than usual one). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 47 index

50
include Vacant land
includeAlley (South-North)
40
HIP (Heat Island Potential)
Densest Area of Building
includeAlley (East-West)
45
Previous Study (Moriyama-cho)
35
Dense Area of Building with Garden 30
25
20
Asphalt 15
Modern blocks
10
5
Lawn
0
HIP=0
'5
0
4
Sunrize1 8
4:451 12
Time
16
Sunset120
19:001 Figure 6. 6 Daily Change of HIP in each Machiyashiki 4-3. Thermal and Wind Environment in Outdoor Living Areas At Noon 7 (during 16:20 to 18:39), which was Japanese local time section in 19th century, most of the residents in Edo Machiyashiki stayed in the streets and alleys to enjoy the evening breeze (Table 2). The influence of vacant lands on the thermal and wind environment in the residents’ living areas was evaluated during this period. 4-3-1. Street (east-west) The distribution of mean radiant temperature (MRT) at the living height (1.5 m) in the vacant areas facing the east-west street was lower than the air temperature because the surface temperatures were the same as the air temperature and the area was open to the sky and was able to get nocturnal radiation cooling. Whereas wind blew into the street from the vacant land at 2.5 m/s, the air temperature, owing to heat from nearby building roofs, was 30.5°C, which was 0.5°C higher than the wind temperature (Fig. 7a). In this case, the spatial structure of Edo Machiyashiki had a negative effect on the thermal and wind environment, and it was difficult for the residents to enjoy cool air in the evening. 4-3-2. Street (north-south) Facing Large Vacant Lands As the wooden gate reduced the wind velocity, the wind in the large area of vacant land facing the north–south street was getting mild (1.5–2.0 m/s). The distribution of MRT at the living height (1.5 m) in the vacant area was 0.5°C lower than the wind temperature (about 29.5°C). The air temperature at the same height was also the same as the wind temperature (Fig. 7b). Concerning the living activities during [2] this period of time , it is considered a possible reason for the ability to enjoy the evening cool air in the vacant land. 4-3-3. Alley (north-south) inside Machiyashiki Considering heat balance and airflow, the distribution of MRT was 27–28°C and was 2°C lower than the wind temperature because of the spatial structure of the north-south alleys and thermal capacity [1] [2] of building members, as was the result in the previous studies . Although it was expected that there would be little wind in the north-south alley with a dead end, wind collided with the roof of a townhouse and down flow flew into the alley at about 2.0 m/s next to the common area (Fig. 7c). The air temperature in the alley with the down flow of air heated by the wooden roofs of townhouses is about 30°C, and a little higher than MRT. Consequently, the environment in the north-south alley with vacant lands, gardens or common areas in Machiyashiki is good both in terms of thermal and wind conditions, and it was suited for residents to enjoy the cool air in the evening. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 48 Vacant+ land
Sign+ Mboard
OMOTE NAGAY A
Roof
30.5
29
aAlley+(No.+⑦)
N
Street
OMOTENAGAYA
Garden
N
bStreet+(No.+①)
30.5
31
Street
Down+flow+:+ +23m/s
30.5
33
Index
OMOTENAGAYA
Garden
URA+ NAGAYA
29.5
27
Vacant+ land
URANAGAYA 29.5

29
Ta
MRT
Living+ height
URA+ NAGAYA
[m/s]
4
3+ Alley
CAlley+(No.+⑦)
N
Figure 7. Distribution of Wind Direction and Velocity in the outdoor living spaces 5. CONCLUSION 2
1
0
In this study, the townsmen residential area of Edo, early modern Tokyo, in the 1790s to the 1860s was the focus, and the summer thermal and wind environment over several city blocks was calculated using the coupled analysis of heat balance and airflow. As a result, the influence of south wind from the Edo bay and vacant lands was evaluated and the influence of vacant lands in the townsmen residential areas were determined quantitatively. One influence is that townsmen residential areas in Edo with vacant lands had never become susceptible to hot summer nights because the spatial structure, vacant lands, and gardens in Edo Machiyashiki contribute to reducing HIP during the day as well. The other influence is that it is clear that some vacant lands works well with the spatial structure of Machiyashiki to introduce the appropriate wind into outdoor living spaces under the better thermal environment during a specified period of time, and this seemed to work well with the residents’ living activities at that time. One of the future subjects is the analysis of the indoor thermal and wind environment. REFERENCES [1] M. Takata, A. Hoyano, “The City without a Nighttime Heat Island: Reproduction and Evaluation of Summer Thermal Environment in Urban Residential Areas in Early Modern Tokyo Using Numerical Simulation”, PLEA2011, 2011.7, pp. 465-470 [2] M. Takata, A. Hoyano, “Evaluating the Relationship among Spatial Structure, Surface Temperature, and Residents’ Behavior in the Alleys of Early Modern Tokyo”, PLEA2012, 2012.9, DVD-ROM [3] T. Asawa, A. Hoyano, and K. Nakaohkubo, “Thermal Design Tool for Outdoor Spaces based on Heat Balance Simulation using a 3D-CAD System”, Building and Environment, vol. 43, pp. 21122123, 2008 [4] S. Yamamura, A. Hoyano, T. Asawa, “Study on Coupled Simulation System of Heat and Air Flow in Outdoor Space for the Computer Aided Design Tool”, Journal of Architecture and Planning (Transaction of AIJ), Vol. 560, 2002, pp. 73-82 [5] T. Ito, “The World of the Kidai Shoran – A Picture Scroll of Edo-period Nihonbashi”, Kodansha Ltd., 2003, pp. 82-86 [6] Unknown, “Kidai-Shoran”, 1804 [7] K. Kikuchi, “Edofunai-Ehonfuzokuourai”, Heibonsha Ltd., 1976 [8] K. Ito, “Streets in Edo”, Heibonsha Ltd., 1986 [9] H. Hagimoto, “Reflection on the Definition of a Weed and the Role of Weed Science”, Study of Weed, Vol. 46(1), 2001.3, pp. 56-59 [10] H. Maruyama, “'Kaso-sho', Textbook of House Aspects (Art of Divination), and 'Kichi or Kyo' (good or ill luck) of Landscape Gardening in the late Edo Period”, Journal of the Japanese Institute of Landscape Architecture, Vol. 61(5), 1998.3, pp. 379-384 [11] Architectural Institute of Japan. “Architectural Design Data Corpus”, Architectural Institute of Japan, 1986 [12] Yamada T. “A Numerical Model Study of Turbulence Air Flow in and above a Forest Canopy”, Journal of the Meteorology Society, Japan, Vol. 60, 1982, pp. 439-454 [13] A. Iino, A. Hoyano, “Development of a Method to Predict the Heat Island Potential using Remote Sensing and GIS data”, Energy and Building, vol. 23, 1996, pp. 199-205 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 49 Optimization for Passive Design of Large Scale Housing Projects for Energy and Thermal Comfort in a Hot and Humid Climate Tam Nguyen Van*, Damien Trigaux, Karen Allacker and Frank De Troyer Architectural Engineering, Architecture Department, Faculty of Engineering Science, KU Leuven, Belgium ABSTRACT Rapid urbanization in emerging economies, like Vietnam, is commonly realized via the multiplication of stereotype projects consisting of high-rise apartment blocks, terraced, semi-detached and detached houses. For these types of dwellings, in the hot and humid Vietnamese climate, individual air conditioning systems are typically used. This requires a large share of the country energy resources. Until recently comfortable traditional housing has however been built without using such energy intensive cooling. This paper focuses on the energy consumption and thermal comfort in residential buildings if only natural ventilation is used, taking into account the urban environment. Via a parametric simulation several building types are optimized, looking at the urban layout, building orientation and size, window design and internal wind permeability. A two-step procedure is followed for the analysis. Firstly, a simplified model is used to analyse a large range of design alternatives and secondly, dynamic energy and thermal comfort simulations with EnergyPlus are made for a selected number of design options. Results reveal that the average life cycle cost of the optimal cases is 34% lower than the reference cases. The window sizes, building layouts and urban forms are crucial parameters to compensate with orientations at the early design stages. One optimization procedure was developed to make maximum use of passive design measures in large scale housing projects. Key words: EnergyPlus, GenOpt, Natural ventilation, Method for cost control, Parametric Wind pressure coefficient. INTRODUCTION Energy efficiency due to energy price and thermal comfort expectation in dwellings in urban areas lead to passive designs at the early stage. For passive design in hot and humid climates natural ventilation and thermal mass are crucial aspects. Natural ventilation, even when coupled with air conditioning, plays an important role when optimizing the life cycle cost and thermal comfort of buildings (A. T. Nguyen & Reiter, 2013). Studies in the context of Malaysia and Singapore proved that ventilation can provide good thermal comfort (Kubota, Chyee, & Ahmad, 2009). Moreover airflow through building openings is a critical factor influencing heat and moisture exchange between thermal Frank De Troyer and Karen Allacker are professor; Tam Nguyen Van and Damien Trigaux are PhD researcher in the Division of Architectural Engineering, Department of Architecture, KU Leuven, Kasteelpark Arenberg 1/ 2430, 3001 Heverlee, Belgium. *Corresponding author: tam.nguyenvan@asro.kuleuven.be , nvtam@ctu.edu.vn (Cantho University, Vietnam) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 50 zones and the outdoor environment (Hens, 2002). The wind pressure is therefore a boundary condition for airflow network models (Cóstola, Blocken, & Hensen, 2009). Until recently vernacular housing has been built without using energy intensive cooling. In Vietnam, vernacular dwellings cannot provide perfect comfort, but they can be fairly well adapted to the local climate by using low energy design principles (A.-T. Nguyen, Tran, Tran, & Reiter, 2011). Therefore, buildings would benefit from low-energy mechanical systems and occupants’ adaptive responses such as changing clothing, opening windows and switch on mechanical ventilation. In urban areas with high building density, such as in the Mekong Delta, cooling systems have however been increasingly installed in bedrooms. This is a consequence of inappropriate design of buildings and urban layouts, such as building orientation and density, window sizes and overhang depths. Those cooling systems require a large share of the country energy resources. There exist however only a few studies to optimize both building geometry and urban layout by parametric simulation and which consider effects of natural ventilation and solar radiation on energy cost, thermal comfort and human responses. The aim of this research project is to predict building energy consumption and discomfort, in an urban context, if only natural ventilation is used and supplemented, if necessary, by cooling. Via a parametric simulation building types are optimized, looking at the urban layout, building orientation and size, window design and internal wind permeability. A two-step procedure is followed for the analysis. Firstly, a simplified model is used to analyse the effect of urban layouts by comparing six reference cases and optimising building geometry. Secondly, the parameters, defined on the neighbourhood and building scale are considered together. EnergyPlus is used for the dynamic energy calculations and GenOpt is used to search for an optimum out of the multiple design options. METHODOLOGY This study includes five steps. (1) Urban layouts, consisting of a number of terraced houses, are simplified and key parameters are selected. (2) For one reference terraced house, the wind pressure coefficients (Cp) on the roof and facades are estimated for 36 orientations and different urban parameters. (3) The Fanger model for thermal comfort evaluation is implemented. (4) A stepwise strategy to strive for occupant comfort is defined, including the following measures: adapt clothing manage natural ventilation - switch on forced air circulation - start the cooling system. (5) An optimization method and an objective function are used to calculate the energy cost, investment cost and 60-year life cycle cost of the analysed options. All design parameters related to both the building and urban scale are varied and sent to EnegyPlus by the optimization tool GenOpt. Simplifying urban forms and parameters selection Figure 1 simplified urban fragment, representative for large scale housing projects in Vietnam. The simplified urban layout model, as shown in Figure 1, represents large scale housing projects, which can be found in many suburban areas in Vietnam. The terraced house type was selected because 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 51 this type constitutes the majority of the dwelling units in housing projects. This urban layout is described with following key parameters: the height of the surrounding buildings, road width, building setback and house width and depth. For this simplified urban fragment, the percentages of circulation, residentialopen and residential-built-up are calculated, using the element method for cost control (De Troyer, 2008); (Allacker et al., 2011). The terraced house geometry As shown in Figure 2, a simple terraced house model was defined, including seven thermal zones: a stair case, a living room, a kitchen and four bedrooms. The outside doors are modelled as windows. The floor height is three meters and the depth of the circulation zone is four meters. A balcony, functioning as an overhang for shading, is provided along the whole house width. The geometric parameters of the analysed terraced house, which is representative for the Vietnamese city of Cantho, are shown in Table 1. This housing unit is simulated with different construction elements including external walls, internal walls, glazing and roof elements. Table 1 Numerical variables and their design options (continuous variables). Design parameters Urban layout: Height of surrounding buildings Width of roads Back garden depth Terraced row depth Terraced row length Terraced house geometry: Terraced house width Front façade overhang Rear façade overhang Window width Abbreviation Initial values Range (m) Step size (m) H1 to H14 Rw, Rl Bg W L 9 12 12 16 40 1 to 36 12 to 24 12 to 24 10 to 20 40 to 120 3 4 4 2 10 Bw ov1 ov2 win1, win2, wwr 6 1.5 1.5 1.5 5 to 10 0.5 to 2 0.5 to 2 1 to 5 1 0.5 0.5 0.5 Figure 2 Terraced house section. Zone 1 and zone 4 are living room and kitchen. Zone 2, 3, 5 and 6 are bedrooms. Zone 7 includes the stair case and circulation area. Wind pressure coefficient (Cp) The TNO Cp-generator, developed in the Netherlands, is used to calculate the wind pressure coefficient (Cp) values on facades and roofs of block-shaped buildings. This generator is based on finite element calculations and has been verified with wind tunnel experiments (Nicolas Heijmans & Peter Wouters, 2003). The TNO Cp-generator was validated by measured data (B. Knoll, J.C. Phaff, & W.F. de Gids, 1995). This validation showed a rather good agreement between measured and calculated results. A similar approach, using this tool, was applied in other studies to obtain the Cp value for a large 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 52 urban fragment (Sun et al., 2014). Multiple linear regressions for wind pressure coefficient (Cp) values For a specific wind orientation the Cp value in the middle of the windward façade of the terraced row, situated in the middle of the urban fragment, depends on 19 independent parameters. An overview of those parameters is given in Table 1: terraced row length (L), terraced row depth (w), width of road parallel with terraced row (Rw), width of road perpendicular to terraced row (Rl), building setback (Bg) and the height of the fourteen surrounding buildings (Hi) (with i = 1 to14, thus 5+14= 19). The same dependency is true for the leeward façade and the roof. The theoretical combinations that can be derived by varying each parameter are very large. Therefore in a first step the “Latin Hypercube Sampling method” is applied to generate 200 combinations of 19 independent parameters. In a second step the Cp values of those combinations are calculated based on the TNO Cp-generator. In a third step, via multiple linear regression analysis, the coefficients of the equation, predicting the Cp values based on the neighbourhood parameters, are derived. For example, one regression function for the front façade is: Cpfront1 = af1i*Hi+ af2*L + af3*W + af4*Rw + af5*Rl + af6*Bg + intercept. The model that has to be considered in order to obtain stable results consists of five building rows in depth direction, three buildings rows in width direction and an air volume height that is equal to 5 times the building height. In this study all 14 surrounding buildings were assumed to have the same height, but the model could also be used for buildings with different heights. In a next step, the same approach can be followed for another wind direction. In order to keep the model manageable only 36 orientations are considered, thus varying in steps of 10°. Figure 3 compares the Cp values obtained from the TNO Cp-generator with those predicted by the linear regression function. The results of one geometric variant are shown for the front and back façade, and the roof, for the 36 wind directions. For all other variants a similar good fit could be found. As the “test reference year” (TRY) in EnergyPlus contains for each hour the wind speed and wind direction, linear approximations are used to derive the driving forces for natural ventilation. Based on opening characteristics and wind-permeability of the dwelling the inside air velocity is calculated. This air speed is then used for predicting thermal comfort based on the Fanger model. Figure 3 Wind pressure coefficient values of multiple linear regression functions (…Cp) and calculation with the by CpGenerator TNO (…CpTNO), (Nguyen Van, Miyamoto, Trigaux, & De Troyer, 2014). Comfort evaluation and strategy based on dynamic schedules for ventilation and cooling Fanger (Fanger, 1970) developed a thermal load index, consisting of a “predicted mean vote” (PMV) on a 7 points scale from “cold” (-3) over “neutral” (0) up to “hot” (+3), based on the heat balance between the body and the environment. His work was the basis for different thermal comfort standards, such as the ASHRAE Standard 55-92 (ASHRAE, 2004) and ISO 7730 (ISO 7730, 2005). In this method, the thermal comfort is defined as the condition when the PMV is between -0.5 and +0.5, which correspond with 90% of users satisfied. The PMV can be used to simulate control actions from “passive” 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 53 to full HVAC (Fadzli Haniff, Selamat, Yusof, Buyamin, & Sham Ismail, 2013). In this study, this approach was implemented in the following way: (1) The PMV values for a zero air velocity and minimum cloth-values (0.36), which are representative for the domestic habits in Vietnam, are calculated via the “Energy management System” in EnergyPlus. If the PMV value is higher than +0.5, fans are switched on. In the living areas, fans with three speed levels generate a wind speed of respectively 0.2, 0.4 and 0.6m/s. In the bedrooms, fans with two speed levels (0.2 and 0.4 m/s) are provided. When the maximal fan velocity is set and comfort is not yet reached, windows are closed and the cooling system is switched on with a set point of 27°C. During the simulations, activity levels of 1.2 met for the living room and 1.0 met for the bedrooms are considered. The living room, kitchen and staircase are assumed to be used from 5:00 to 9:00 and from 17:00 to 23:00 during working days. In the weekend, the living room is used from 5:00 to 23:00. The bedrooms are used from 23:00 to 5:00. All lights in the model are controlled by time schedules and luminance levels using two sensors in each zone. The luminance levels for the living room, kitchen, bedrooms and circulation area are respectively 300 lx, 500 lx, 300 lx and 150 lx. 69 lm/W LED lights are considered in all zones. An example of input (occupancy schedule) and output (PMV, inside air velocity) is given in Figure 4. Figure 4 Example of a dynamic schedule for one week Optimization method and objective function To optimize the building costs with thermal comfort constraint, a parametric simulation, based on the dynamic energy simulation software, EnergyPlus 8.01, was used. Then EnergyPlus was coupled with the optimization tool GenOpt to minimize the results of the objective function (Wetter, 2011). The life cycle cost is present value which is calculated including construction, maintain, operation, land and infrastructure costs. The operation and maintain costs were estimated with 5% inflation rate and 10% nominal discount rate for the energy cost for 60 years life span. The unit costs were used from local data at Cantho and land cost based on (Nguyen Van & De Troyer, 2013). The optimization of building design alternatives is a non-linear multi-objective process. Hence, it often requires a trade-off between conflicting design criteria, such as the initial construction cost and the operating cost (Wright, et al., 2002). The most common approach to optimize such conflicting criteria is to apply the concept of Pareto optimization in which a set of trade-off solutions, called Pareto front, is obtained. This approach results in many Pareto-optimal design options, out of which designers or users can select the most preferred one, based on their specific preferences. The study minimizes the sum of the initial cost and operational cost for energy, replacements and maintenance within the constraint that total discomfort hours should be smaller than 10% of the hours that people are present. The number of hours that cooling is required or that people feel dis-comfortable if no cooling is available and the required cooling energy based on a system with 100% efficiency is reported. CASE STUDY Varying all selected parameters with fixed north orientation In Figure 5, as an example, input and output for one day in April are represented for one bedroom, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 54 in order to illustrate the variations in PMV values, operation of fans and cooling. Additional settings are: sleeping period from 23 to 5. As a rule, when the indoor air temperature is higher than the outdoor air temperature, the windows are opened for natural ventilation. As shown in Figure 5, the PMV values are between -0.5 and +0.5 during the occupation period from 0 to 5 am, thanks to the natural and forced ventilation. From 5 am nobody is present in the rooms and fans or air conditioners are stopped. At the same time, the increase in outdoor temperature results in a higher indoor air temperature and the PMV surpasses 0.5. From 17:00, the outdoor temperature is again below the indoor temperature, so that cooling via ventilation is possible. From 23:00 the bedroom is occupied and additional cooling is required during one hour. Figure 5 Ventilation volume, outdoor air temperature, zone air temperature, PMV values, zone air velocity and cooling set point, during one day in April, in zone 2 (bedroom). The reference terraced house, using the initial values defined in Table 1, was analysed in six extreme urban layouts, considering north-oriented front façades, as shown in Figure 6. The results for the energy, construction and life cycle cost are reported for the reference cases and further optimised by changing window, glazing types and overhang sizes. Figure 6 Six urban layouts with different built-up ratio and building densities: Case 1, 2, 3: L =120m; W = 16m; Rl = Rw = Bg =12m; Built-up Ratio = 52%. Case 4, 5, 6: L = 40m; W = 16m; Rl = Rw = Bg =24m; Built-up Ratio =25%. Figure 7 Energy cost of the reference terraced houses and the optimal solutions of 6 cases. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 55 To evaluate the effects of solar radiation and natural ventilation, the 6 urban layouts, described in Table 1 were analyzed by varying window sizes glazing types and overhang depths. The results show that the life cycle cost of the optimal cases is 34% lower than the reference cases, as shown in Figure 7. Moreover, we can see the impact of the urban layout on available solar radiation and natural ventilation: the models with a higher building density (narrow streets or higher buildings) require a higher energy cost for fans and cooling. When only considering a fixed north orientation (A) for the front façade, as shown in Figure 8 or varying all parameters including (B) orientation, as shown in Figure 9, the lowest building life cycle costs (including land and infrastructure costs) are respectively 245 and 242 ($/sqm floor) and energy costs are respectively 13.1 and 13.8 ($/sqm floor). As shown in Table 2, important differences are found for the optimal window dimensions, depending on the orientation. Moreover, the optimal height of surrounding buildings is 34.5 m for an orientation of 120 degrees, compared to 9.4 m for the north orientation because of the predominating south-east wind direction in summer time. Figure 8 Pareto fronts of about 7300 design alternatives with north-oriented front façades. Table 2 Optimization results for the neighbourhood model with north-oriented front façade (A) and varying orientation (B). Case L (m) W (m) Rw (m) Rl (m) Bg (m) H (m) Bw (m) ww1 (m) ww2 (m) wwr (m) ov1 (m) ov2 (m) orient (degree) (A) (B) 120 120 20 20 12 12 12 12 6 6 9.4 34.5 10 10 2.8 3.5 2.5 3.7 0.97 1.09 0.5 0.5 0.5 0.5 0 120 Varying all selected parameters with orientation from 0 to 360 degree 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 56 Figure 9 Pareto fronts of the neighbourhood model with varying orientation from 0 to 360. Construction elements of optimal solutions of both cases with the North and 120 degree orientation are shown in Table 3. Pyrolytic clear glass is a reflective glazing type with low solar transmittance at normal incidence. This glazing type is appropriate to reduce heat gains in hot and humid climates with strong solar radiation. Elements Roof Internal wall External wall Glazing Floor Table 3 Optimization results for the construction elements Material layers 6mm asphalt; 10cm reinforced concrete; 1.5 cm ceiling mortar 1.5cm mortar; 10cm clay brick; 1.5cm mortar 1.5cm mortar; 10cm clay brick; 1.5cm mortar 6mm Pyrolytic clear glass Floor tile; 1.5cm mortar; 10cm reinforced concrete; 1.5 cm ceiling mortar DISCUSSION The results of the optimization process show that adaptations on three levels should be considered in the early design stage: materials, building geometry and urban layout. First, reflective glazing can reduce solar gain. Moreover external walls with a clay brick layer and two layers of mortar provide a good thermal performance for a low cost. Second, when windows are protected from solar gains with overhangs, the window size should be maximized for improving natural ventilation. The cross ventilation can be increased because of the buoyancy wind flows street canyon (Allegrini, Dorer, & Carmeliet, 2014). Furthermore, road width, building setback and obstacle heights effect solar gains and wind circulation importantly. Hence, natural ventilation can be improved by window size, orientation and urban density. Not considered in this analysis is that trees and plants can reduce incident solar radiation, (Villalba, Pattini, & Córica, 2014), but also wind permeability. As a conclusion, the model can be adapted based on many design parameters at different scale levels. In reality, selecting a main orientation for an urban area depends on the existing infrastructure, but a disadvantaged orientation can be compensated by changing the building geometry. CONCLUSIONS In this paper one optimization procedure is developed to make maximum use of passive design measures in large scale housing projects, while guaranteeing a pre-defined thermal comfort. This procedure takes into account following aspects: solar radiation (including reflection), wind pressure coefficients for natural ventilation, building geometry, urban layout and a stepwise strategy adapting (1) clothing, (2) open or close windows, (3) vertical or ceiling fans, and (4) cooling. The implemented dynamic schedule provides comfort, in the hot and humid Vietnamese climate, during more than 90% of the occupation time. This strategy requires sustainable occupant behaviours or automatic control systems to follow the dynamic schedule or careful intervention by occupants. This model can be extended to other urban geometries, such as towers and urban street blocks, other energy management systems, other occupancy and activity schemes, other construction technologies and other economic scenarios (growth rate of costs, interest rates and life span. ACKNOWLEDGEMENTS This study was supported by the Ministry of Education and Training of Vietnam. We thank to Dr Anh Tuan Nguyen, Faculty of Architecture, the University of Da Nang, Viet Nam, who kindly gave many instructions to couple EnergyPlus to GenOpt engine. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 57 REFERENCES Allacker, K., Troyer, F. D., Trigaux, D., Geerken, T., Debacker, W., Spirinckx, C., … Putzeys, K. (2011). Sustainability, Financial and Quality evaluation of Dwelling Types - SuFiQuaD - FINAL REPORT. Brussels. Allegrini, J., Dorer, V., & Carmeliet, J. (2014). Buoyant flows in street canyons: Validation of CFD simulations with wind tunnel measurements. Building and Environment, 72, 63–74. doi:10.1016/j.buildenv.2013.10.021 ASHRAE. (2004, February 24). ASHRAE - Std 55-2004 Thermal Environmental Conditions for Human Occupancy. B. Knoll, J.C. Phaff, & W.F. de Gids. (1995). Pressure Simulation Program. Presented at the The 16 th AIVC Conference, Palm Springs, USA. Cóstola, D., Blocken, B., & Hensen, J. L. M. (2009). Overview of pressure coefficient data in building energy simulation and airflow network programs. Building and Environment, 44(10), 2027–2036. doi:10.1016/j.buildenv.2009.02.006 De Troyer, F. (2008). BB/SfB-plus - Een functionele hiërarchie voor gebouwen. Leuven: ACCO. Fadzli Haniff, M., Selamat, H., Yusof, R., Buyamin, S., & Sham Ismail, F. (2013). Review of HVAC scheduling techniques for buildings towards energy-efficient and cost-effective operations. Renewable and Sustainable Energy Reviews, 27, 94–103. doi:10.1016/j.rser.2013.06.041 Fanger, P. O. (1970). Thermal comfort: analysis and applications in environmental engineering. Danish Technical Press Copenhagen. Hens, H. (2002). Heat, air and moisture transfer in insulated envelope parts: performance and practice, International Energy Agency (Final Report No. Annex 24, vol. 1Acco). United Kingdom: International Energy Agency. ISO 7730. (2005). ISO 7730, Ergonomics of the thermal environment — Analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria. Kubota, T., Chyee, D. T. H., & Ahmad, S. (2009). The effects of night ventilation technique on indoor thermal environment for residential buildings in hot-humid climate of Malaysia. Energy and Buildings, 41(8), 829–839. doi:10.1016/j.enbuild.2009.03.008 Nguyen, A. T., & Reiter, S. (2013). Passive designs and strategies for low-cost housing using simulation-based optimization and different thermal comfort criteria. Journal of Building Performance Simulation, 0(0), 1–14. doi:10.1080/19401493.2013.770067 Nguyen, A.-T., Tran, Q.-B., Tran, D.-Q., & Reiter, S. (2011). An investigation on climate responsive design strategies of vernacular housing in Vietnam. Building and Environment, 46(10), 2088–2106. doi:10.1016/j.buildenv.2011.04.019 Nguyen Van, T., & De Troyer, F. (2013). Deriving Housing Preferences from advertising on the web for improving decision making by Economic and Social actors. Presented at the At home on the housing market: RC43 conference book of proceedings, Amsterdam University, Netherlands. Nguyen Van, T., Miyamoto, A., Trigaux, D., & De Troyer, F. (2014). Cost and comfort optimisation for buildings and urban layouts by combining dynamic energy simulations and generic optimisation tools (pp. 81–92). Presented at the ECO-ARCHITECTURE V, Harmonisation Between Architecture and Nature, SIENA, Italy: WIT Press. Nicolas Heijmans, & Peter Wouters. (2003). Impact of the uncertainties on wind pressures on the prediction of thermal comfort performances (No. IEA ECBCS Annex 35). Retrieved from http://www.hybvent.civil.aau.dk/puplications/Technical%20Reports/TR23%20WindCp.pdf Sun, Y., Heo, Y., Tan, M., Xie, H., Jeff Wu, C. F., & Augenbroe, G. (2014). Uncertainty quantification of microclimate variables in building energy models. Journal of Building Performance Simulation, 7(1), 17–32. doi:10.1080/19401493.2012.757368 Villalba, A. M., Pattini, A. E., & Córica, M. L. (2014). Urban trees as sunlight control elements of vertical openings in front façades in sunny climates. Case Study: Morus alba on north façade. Indoor and Built Environment, 1420326X14543506. doi:10.1177/1420326X14543506 Wetter, M. (2011, December 8). GenOpt(R), Generic Optimization Program, User Manual, Version 3.1.0. Lawrence Berkeley National Laboratory,. Retrieved from http://SimulationResearch.lbl.gov 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 58 Investigations about a Scale of Correlation for the Relationship between Urban Physical Dimensions and Wind Cp Luciano Caruggi de Faria, PhD Phil Jones, Professor Don Alexander, Senior Lecturer Jonas Birger Arquitetura luciano.caruggi@gmail.com Cardiff University- WSA Cardiff University – WSA ABSTRACT This paper presents the results of an investigation focused on a scale for predicting the potential for naturally ventilate buildings in the urban environment. This scale is based on the Pearson r model of correlation coefficients between urban physical dimensions with the resulting wind-driven pressure coefficients (Cp). Numerical simulations using computational fluid dynamics (CFD) were performed to a large number of urban prototypes with simplified volumetric shape. These urban prototypes were originally based on ratios of actual urban areas. The systematic variation of the volumetric urban aspect ratio of these prototypes and the simulation for three wind directions allowed finding the relationship between the urban fabric and the ∆Cp. The range in the urban prototype shapes covered as many types of urban fabric as possible, from high to low density, from low building centres to downtown skyscraper areas. Eventually, two actual urban areas and buildings, the Cardiff University Cathays Campus, in Cardiff- Wales, and the Paulista Avenue, in São Paulo- Brazil, were also simulated and contrasted to the urban prototypes scale. A relationship was found between the urban aspect ratios and the ΔCp results. This was demonstrated by statistical methods using the data on the variables concerned, thus verifying the strength of the correlation between them. Strong correlation was found between the investigations into similar scenarios of the urban prototypes and the two case studies as regards both the aspect ratios and the ΔCp results. On the other hand, low correlation for the same variables was identified when contrasting dissimilar urban prototype scenarios. KEY WORDS urban environment; aspect ratio; wind pressure; cfd; k-e; and Cp. BACKGROUND TO STUDIES ON AIRFLOW IN URBAN AREAS Flow patterns around isolated bluff bodies are well-known and the basis of knowledge for the calculation of pressure coefficients across buildings and wind loads on structures (Olgyay, 1973; MacDonald 1975; Awbi, 2003; Cook 1985; Holmes 2001). In contrast, the wind field in the urban environment is more complex and less predictable, notably below the canopy height of high-density city centres. According to Cook (1985), when the surface roughness is large and packed, as in towns, airflow detaches at roof height and is channelled in several directions, and will be related to the local neighbourhood buildings shape. General aspects of wind patterns in the urban environment, as compared to those of undisturbed wind, are: mean speed reduction, turbulence intensity increase and greater incidence of weak winds (Ghiaus and Allard, 2005), thought coupling between free and channelized winds is also observed. Urban canyon areas are created by the corridors lying between buildings and are formed by the cavities between the road surface and its flanking buildings, up to roof-top level (Vardoulakis et al., 2003) and the air volume within an urban canyon plays an active role in the definition of the surrounding urban micro-climate and its interaction with the meso-scale climate (Nakamura and Oke, 1988). Studies on the effects of airflows within urban canyons are focused on natural ventilation, the dispersal of air 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 59 pollution concentration and urban noise, and descriptions for parallel, orthogonal and skewed airflow in simple canyons of infinite length are found in the literature (Wedding et al., 1977; DePaul and Shieh, 1986; Hoydysh and Dabberdt, 1988; Oke 1988; Hunter et al., 1991; Sini et al., 1996; and Santamouris et al, 1999). Effects of slanted flows in intersections and urban areas are reported by Yamartino and Wiegang (1986), DePaul (1986), Hoydysh and Griffiths (1987, in Ahmad et. al, 2005), and Georgakis and Santamouris (2004). Givoni (1976) states that studies with simple shapes based on urban form give an indication of reality such as avoids the interference of other factors in the outcomes, serving as a parameter for other similar, but more complex, urban arrangements. Hunter et al. (1991) describe the important role that the urban geometry plays in the near-surface airflow in urban centres. Several geometric parameters are employed, which are based on linear dimensions, areas and volumes. The proportionality between the building and/or block height (H) and building and/or block length (L) and the road width (W) identifies the built aspect ratio and the type of volumetric canyon within it. It is expected that the resultant airflow speed and direction below the canopy height should be connected to variations in these aspect ratios. These urban ratios are given by the relation between the: building height to road width (H/W) or length (L/H), plan-area density (a= Aroof/ Aurb), and built-area density (b= Abuilt/ Aurb). An urban canyon can be considered uniform or regular when its cross-sectional H/W ratio approximates to 1.0, deep or narrow when this ratio increases to 2.0 and wide or shallow when it drops to 0.5. Also, the canyon length L/H ratio is considered short, medium or long for respective ratios of 3.0, 4.5 and 6.0 (Nakamura and Oke, 1988; Vardoulakis et al., 2003). THE RESEARCH METHOD With the aim to verify with which extent the airflow in urban environment results from a combination of built density and the undisturbed wind, and if it can be translated into a matrix of correlation, the research method comprised two steps (de Faria, 2013): 1. The simulation on computational fluid dynamics (CFD) of fifty three urban prototypes with simplified volumetric shape based on eighteen urban arrangements for three wind directions: parallel (0°), orthogonal (90°), and oblique (45°); and 2. The assessment of the airflow in two real urban centres (Cardiff Cathays campus area, Wales; and Paulista Avenue, São Paulo, Brazil) carried out via two (wind tunnel- WT and CFD) or three (WT, CFD, and field measurement- FM) techniques combined and for 8 cardinal and intermediate wind directions. The groups of urban prototypes The definition of the proposed urban prototypes covered a variety of urban landscapes, from high to low density, from low building centres to downtown skyscrapers, and was based on the aspect ratios of five areas in the cities of: Cardiff, London, Paris, São Paulo and Hong Kong. The approach covered an area equivalent to that of a circle 500m in diameter from the target building. The systematic variation of the volumetric urban aspect ratio of these prototypes allowed observing the relationship between the built environment and the airflow speed and direction and the wind-driven pressure. Furthermore, these sets of prototypes are not intended to be generally valid or applicable since they have limitations and were created specifically to answer the hypothesis set out in this research. The prototypes were divided into four types: ‘A’, ‘B’, ‘C’, and ‘D’, in accordance with the H/W aspect ratio, and then into four sub-types: 1, 2, 3, and 4, with decreasing plot occupancy density (see Figure 1). While the first two aspect ratios refer to the respective canyon’s linear dimension, the last two refer to areas of several blocks within a pre-established urban perimeter area. From ‘A’ to ‘C’ the scenarios were symmetrical, the height of the blocks was kept constant at 30m, and the division among the types took into account the H/W aspect ratio and the roof and built areas. The length of the blocks also varied from 180m to 30m. The type ‘B-Step’ was a variation of the ‘B-2’ in which half of the blocks had their height doubled in order to assess the impact of step-up and step-down airflows in canyons. Type ‘D’ was also based on the previous sets ‘A’ and ‘B’, but it presented random asymmetry due to height variation of up to three times the previous ones in some of its blocks. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 60 This set sought to represent a more heterogeneous urban scenario. Finally, the ‘D-4’ scenario presented several detached blocks of 30, 60 and 90m height, thus resembling an urban landscape with high-rise buildings. Figure 1 Urban prototype scenarios top views and cross sections and the wind directions. Table 1. Definition and Characteristics of the Urban Prototype Models and their Equivalence to the Real Urban Canyon Assessed1 . Several links between the prototypes and the urban areas may be made (see Table 1). When these links are related to one aspect alone there is a weak connection between them. For instance, if the H/W aspect ratio is considered alone, four urban areas, London, Paris, São Paulo and Hong Kong, have an H/W ratio around 1.0. Conversely, when associated with other criteria, for instance plot occupancy; the first two cities are closer to prototype B1, and the last two to D4, since there is another link as well. In addition, the respective examples present visual compatibility in their urban landscape. In order to confirm whether the built aspect ratio links can be transferred to the results in terms of airflow pressure and velocity Several H/W and L/H ratios can be found in the D1, D2, D3, and D4 prototypes since the geometry and volumes are asymmetrical and heterogeneous, and an averaged value based on the several dimensions in the model is used for calculating the related urban aspect ratios. 1 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 61 decrease within these urban areas, two of these sites, Cardiff and São Paulo, were selected for further investigation. Both of these places could provide essential information to verify the accuracy of the proposed method. Further, neither Cardiff nor São Paulo matched accurately a prototype in all three criteria. This may help to bring out whether one of the criteria is stronger than the other in the relation between built mass and the resultant airflow field. The case studies The two urban areas selected for the case study were: Park Place on the Cardiff University Cathays Campus area (Figure 3); and the Paulista Avenue, in São Paulo (Figure 4). As case studies, both areas were simulated by CFD and wind tunnel, while field measurements (FM) were only performed in the former2. While the Cathays Campus neighbourhood is considered a low-density area with mostly threefloor low buildings close to open areas such as Alexandria Gardens and Bute Park, in contrast, the urban site and immediate surroundings of Paulista Avenue, located on a hill-crest at the core of the City of São Paulo, is characterized by high-density land occupation and high-rise buildings, with this avenue being one of the most important financial poles in Brazil. The computational fluid dynamics (CFD) simulations The CFD programme used in this investigation was a research version of the ANSYS FLUENT 6.2 and the 3D models were built and meshed in the Gambit 2.0 software. Although CFD results are susceptible to uncertainties and approximations, the achievement of consistency and reliability in the outcomes is related to the control of a number of input and calculating parameters, which is achieved by following standard procedures and performing pre-test simulations for calibration, verification and validation of the results3. These procedures and tests, and the steps taken during the CFD stages of modelling, pre-processing, solving and post-processing are described in de Faria (2008) and de Faria (2013). This practice was used to ensure consistency in the modelling for all the three groups of CFD calculations undertaken: the calibration of the CFD input and modelling parameters itself; the investigation of the urban prototypes; and the assessment of urban areas approached as case studies. In the CFD preprocessing stage the 3D model input is specified, which involves decision making about the domain discretization, size and verifying the impact that the boundaries specification, the mesh type and size, the mesh adaption, the fluid properties, the cell blockage and other aspects of the problem description may have on the results. The solution to this imposed problem is calculated during the CFD solving stage. The turbulent viscosity model adopted for all the CFD simulations was the k-e RANS standard. Several steps involving the solution control parameters, such as the choice of the time mode; thermal mode; turbulence model; solution controls; relaxation factors; monitoring solution progress; and residual plot thresholds, may interfere in the quality of the simulation and, in consequence, in the reliability of the results. In order to analyse the CFD results, airflow pathlines were used for visualizing the airflow field through the blockages. The quantitative data from the CFD output were extracted either from lines (data about wind velocity (magnitude, ‘x’, ‘y’, or ‘z’ vectors) or from surfaces (Cp values). The data were exported from the export panel as comma delimited in the ASCII format and imported into the Excel software. Correlation coefficients and the scale of significance for urban aspect ratios The correlation coefficient identifies the number of relationship between two sources of quantitative variable data, thus ascertaining the statistical strength between them. The Pearson r model provides a scale of significance for correlation coefficients. This scale is a linear association between standard productmoment sources of data. The correlation coefficient ‘r’ equation, based on series of data ‘x’ and ‘y’ and Only the CFD results are presented in this paper. Further information is available in de Faria (2013). The calibration, verification and validation of the parameters used in the CFD investigations were attained by calculating the flow field around two parallel rectangular bricks and contrasting the results with those of the wind tunnel physical model (de Faria, 2013). 2 3 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 62 employed here, is given by Barrow (2009). The values on this scale range from +1.00 to -1.00, on which zero means absence of correlation (Warner, 2008; Barrow, 2009; Kottegoda and Rosso, 2009; Croft and Davidson, 2010). Several correlation analyses were employed in this investigation in order to reveal a number of associations between different models, such as: the level of diversity among the several urban prototype’s physical dimensions and aspect ratios adopted; the variety of ΔCp results among the several urban prototypes; the urban shape/ aspect ratio’s similarity strength between the several urban prototypes and each case study investigated; and the ΔCp results’ similarity strength between the several urban prototypes and each case study investigated. By comparing the correlation coefficients for the urban prototype’s Groups 01 to 06 it was possible to observe that for those from the same group (e.g. A1, A2 and A3; B1, B2 and B3...) a correlation relationship from 1.00 to 0.94, while the ones from an adjacent group (e.g. A1 and B1, A2 and B2; A3 and B3...) present a correlation relationship from 0.94 to 0.91. This shows that the systematic variation of the aspect ratios for these simplified scenarios was obtained in a balanced gradient between prototypes both intra and inter-group. Conversely, when comparing dissimilar scenarios, such as A1 and D4, and A4 and C1, the relationship was 0.66 and 0.45, respectively. By comparing the urban prototype aspect ratios and the two case studies aspect ratios, a scale of significance for correlation coefficient strength of physical dimensions in the urban environment is presented (Figure 2). Although Pearson’s model is frequently applied in civil and environmental engineering investigations and in the field of the Sciences of Technology, including models for spatial correlation (Kottegoda and Rosso, 2009), a scale such as would determine the strength of the correlation which is compatible with this investigation could not be found in the referenced literature reviewed (Warner, 2008; Barrow, 2009; Croft and Davidson, 2010). Further, the correlation strength scales provided by the Social or Biological Sciences literature are not appropriate for this application4, since they are specific to those fields and thus do not match the scale of results from this research area. On the other hand, the urban prototypes proposed in this investigation were based on urban aspect ratios of actual urban areas and that several links between them were identified. For instance, when observing the urban landscapes of the two case studies investigated in depth it is possible to associate the Cathays Campus with the urban prototypes A and C, while the Paulista Ave. is more closely similar to the urban prototypes D3 and/ or D4. Therefore, for the five cities assessed in the urban area analysis, the Cathays Campus would be positioned on one side of the scale, characterized as a low-height built-up area, whilst the Paulista Ave. would sit on the other side, as a high-rise built-up area, with both landscapes representing the extremities of this scale. Based on this hypothesis, it is to be expected that results between Cathays Campus and prototypes A and/ or C will present a strong correlation while results between Cathays Campus and prototypes D3 and/ or D3 will present a weak correlation, with the opposite occurring with the Paulista Ave. Indeed, the correlation coefficient between the Cathays Campus aspect ratios and the urban prototypes previously related to have similarities in their landscapes, such as A1; A2; A3; B1; B2 and B3 showed correlation coefficients of 0.94; 0.87; 0.91; 0.85; 0.80 and 0.82, respectively. In contrast, the correlation coefficient between the Cathays Campus aspect ratios and the urban prototype with opposite landscape features (D4), on the other edge of the scale, was the lowest found: 0.51. The correlation coefficients between the Paulista Ave. and the urban prototypes aspect ratios seen on Table 4 showed a relationship of 0.95 and 0.90 with the urban prototypes D4 and D3, respectively, which belong to the Group 6 prototypes scenarios and were previously described as having the most similar urban landscape features. Once more, the urban prototype previously defined as the opposite one to this high-rise building urban landscape presented the lowest correlation coefficient: 0.53 (A1). This lowest result was followed by the ones obtained with C1 (0.52) and B1 (0.58). Again, the similar and the dissimilar urban landscapes were positioned in opposite edges of the scale. Therefore, the correlation coefficient found between these will serve as a standard for the scale of significance for this exercise. Based on the findings the scale of significance for assessing and comparing 4 For instance, De Vaus (2002) ranks the correlation coefficients for Social Science researchers as follows: 1.00= perfect; 0.99 to 0.90= near perfect; 0.89 to 0.80= very strong; 0.79 to 0.70= strong; 0.69 to 0.50= substantial; 0.49 to 0.30= moderate; 0.29 to 0.10= low; 0.09 to 0.00= trivial; while a negative result implies in a reverse correlation, in De Vaus, D. 2002. Analyzing Social Science Data. London: Sage. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 63 urban landscapes’ physical aspect ratios correlation coefficient strength proposed in Figure 2 provides four ranks for the Pearson’s correlation scale of significance: strong, substantial, moderate, and low. For the urban aspect ratios analysis, this means a range of correlation coefficient ranging from 0.95 to 0.51. In the following topic, this scale will serve as a reference for ranking the ΔCp level of association and correlation coefficient strength. It is expected the ΔCp results to follow the same sequence as this scale, though not of the same order of magnitude. a) 1 Scale of significance for urban prototype aspect ratio’s and Cardiff/ Paulista aspect ratio’s correlation coefficient (r) strength 0,9 0,8 b) 0,7 0,6 0,5 A1 1 B1 2 C1 3 D1 4 A2 5 B2 6 C2 7 D2 8 A3 9 B310 C3 11 D312 A4 13 B4 14 C4 15D4 16 Cardiff São Paulo The scale of significance for urban prototype aspect ratio’s and Cardiff/ Paulista aspect Figure 2 ratio’s correlation coefficient (r) strength (a) and illustrative graphic (b). RESULTS Correlation coefficients and the scale of significance for ΔCp results Here this investigation seeks to identify if the correlation coefficients found for the aspect ratios between the Cathays Campus (targeting the Law School Building external façade, situated on the Museum Ave. and the Park Place) and the Paulista Ave. (CYK Tower) with the Groups 01 to 06 of the urban prototypes, may also be translated into ΔCp results. The Cardiff University Cathays Campus The ΔCp results for Cathays Campus, when contrasted to the Prototype A1 (Figure 3), whose correlation for aspect ratio was of 0.94, showed also strong correlation of 0.90 for S winds (at 45o to the Museum Avenue external side) and substantial correlation for SE (0.83) and NW (0.78) - both at 0o. In 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 64 contrast, the wind directions from E (0.53), N (0.36) and W (0.33) - the three at 45o - presented moderate to low correlation, while the ones from NE (-0.30) and SW (-0.68) - both at 90o - showed low to moderate reverse correlation. Both the equivalences and disparities found between the aspect ratio and the ΔCp results correlation coefficients may occur due to the ‘V’ shape of the Law School Building which forms a courtyard at an angle of 45o with side high-rise buildings has an impact on both the airflow velocity and the ΔCp results. On the other hand, when contrasted to the Prototype A2, whose correlation for aspect ratio was of 0.87, the correlation for the NE and SW winds - orthogonal to the façade – raised to 0.22 and 0.55, respectively. And when compared to the contrasting D2 prototype, whose correlation for aspect ratio was of 0.68, the correlation found for the E wind raised to 0.62. a) b) The arrows point the wind direction and show: strong (white), low (lines) or reverse Figure 3 (black) correlation strength between (a) the Cardiff Cathays area and (b) the Prototype A1 ΔCp results. CYK Tower and the Urban Prototypes correlation assessment Strong correlation was found between the Paulista Ave. and the urban prototype D4 on both the aspect ratio (0.95) and the ΔCp: 0.91 for NE (0o), 0.90 and 0.94 for N and S (90o), and 0.92 and 0.93 for NW and SE (45o) winds (Figure 4). Further, at least two out five wind directions with the same rank for the other prototypes from the Group 6 (D1, D2, and D3) also showed strong statistical similarity to the Paulista Ave. The prototype scenarios A1, A3, C3, and C4 showed substantial statistical similarity, while all the other prototype scenarios showed from moderate to low relationship levels. a) b) The arrows point the wind direction and show: strong (white), low (lines) or reverse Figure 4 (black) correlation strength between (a) the Paulista Ave. and (b) the Prototype D4 ΔCp results. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 65 CONCLUSION A large number of CFD simulations involving urban prototypes and two case studies was performed aiming to verify the strength on the relationships between the built environment and the ∆Cp distribution in several urban scenarios for parallel, perpendicular and oblique wind directions. Both the H/W and the Abuilt/ Aurb aspect ratios on the windward side of the target area are mandatory on the definition of the airflow field and the Cp on the buildings envelope. A comparison of Cathays Campus ΔCp results showed strong correlation between this actual urban area and the comparable prototype A1 for three wind directions, while other three showed moderate to low correlation. Contrastingly, a stronger link was found with overall dissimilar prototypes but whose urban shape is related to this real urban area on its windward side. The combined analysis between the Paulista Ave. and the D4 prototype showed strong statistical strength between both the physical aspect ratios and the ΔCp results for five wind directions assessed. This is consistent with the hypotheses and the objectives of this investigation. Based on these findings it may be affirmed that the relationship between the various physical dimensions which characterize the urban environment in terms of its urban aspect ratios have proved to be related to the resultant ΔCp in buildings when associated with air flow data. Therefore, it seems possible to create an empirical scale that permits to estimate the ΔCp results. Although such method requires further research and validation before its application as a practical tool, such scale would be helpful for architects, building engineers and urban planners on designing naturally ventilated buildings. REFERENCES Awbi, H. B. 2003. Ventilation of Buildings. 2 ed. London: Spon Press. Barrow, M. 2009. Statistics for economics, accounting and business. 5th ed. Essex: Pearson, p 475. Cook, N. J. 1985. Designer's Guide to Wind Loading of Building Structures. London: Butterworths. Croft, A. and Davison, R. 2010. Foundation Maths. 5 ed. Harlow, England: Pearson. De Faria, L. C. 2013. Airflow in the urban environment. An evaluation of the relationship between urban aspect ratios and patterns of airflow, wind velocity and direction in urban areas, and coefficient of pressure distribution on building envelopes. Thesis (PhD). WSA, Cardiff University. Cardiff. Available at: http://orca.cf.ac.uk/45307/ [Retrieved in: 15 November 2013]. DePaul, F. 1986. Measurements of wind in a street canyon, Atmospheric Environment 20, pp. 455–459. DePaul, F. T. and Sheih, C. M. 1986. Measurements of wind velocities in a street canyon. Atmospheric Environment 20, pp. 445-459. Georgakis, Ch. and Santamouris, M. 2004. On the air flow in urban canyons for ventilation purposes. The International Journal of Ventilation 3(1), pp. 53-66. Ghiaus, C. and Allard, F. 2005. The physics of natural ventilation. In: Ghiaus, C.; Allard, F. Natural Ventilation in the Urban Environment. URBVENT. London: Earthscan, pp. 36-80. Givoni, B. 1976. Man, Climate and Architecture. 2 ed. London: Applied Science Publishers. Holmes, J. D. 2001. Wind Loading of Structures. New York: Spon Press. Hoydysh, W. G. and Dabberdt, W. F. 1988. Kinematics and dispersion characteristics of flows in asymmetric street canyons. Atmospheric Environment 22, pp. 2677-2689. Hunter, L. J., Watson, I. D. and Johnson, G.T. (1991) Modelling air flow regimes in urban canyons. Energy and Buildings 15-16, pp. 315-324. Kottegoda, T. N., and Rosso, R. 2009. Applied statistics for civil and environmental engineers 2nd ed. Oxford: Wiley-Blackwell, pp 736. MacDonald, A. J. 1975. Wind Loading on Building. London: Applied Science Publisher. Nakamura, Y. Oke, T. 1988. Wind, temperature and stability conditions in an east-west oriented urban canyon. Atmospheric Environment 22(12), pp. 2691-2700. Oke, T. 1988. Street design and urban canopy layer climate. Energy and Buildings 11, pp. 103-113. Olgyay, V. 1973. Design with Climate. Bioclimatic Approach to Architectural Regionalism. NJ: Princeton University Press. Santamouris M, et. Al. 1999. Thermal and airflow characteristics in a deep pedestrian canyon under hot weather conditions. Atmospheric Environment. 33, pp. 4503-4521. Sini, J.-F. et al. 1996. Pollutant dispersion and thermal effects in urban street canyons. Atmospheric Environment 30, pp. 2659–2677. Vardoulakis, B. et al. 2003. Modelling air quality in street canyons: a review. Atmospheric Environment 37 (2), pp. 155-182. Warner, R. 2008. Applied Statistics: From Bivariate Through Multivariate Techniques. LA: Sage. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 66 Session 1C : User behavior, thermal comfort & energy performance PLEA2014: Day 1, Tuesday, December 16 11:30 - 13:10, Grace - Knowledge Consortium of Gujarat Daylighting for Visual Comfort and Energy Conservation in Offices in Sunny Regions Evyatar Erell, PhD Eran Kaftan, PhD Yaakov Garb, PhD Ben-Gurion University of the Negev erell@bgu.ac.il Ben-Gurion University of the Negev Ben-Gurion University of the Negev ABSTRACT Office buildings in regions with abundant sunlight may still fail to make effective use of daylight: the difficulty in controlling variations in natural illumination, which may be substantial, often results in extensive use of artificial lighting. A solution to this paradox was sought by means of a controlled experiment designed to investigate the effect of several strategies to reduce glare and to achieve visual comfort in a test room configured to represent a typical side-lit office. Subjects performed office tasks such as reading or operating a computer, and completed a detailed questionnaire about their work environment, whose physical parameters were monitored in great detail. The study showed that if the window is exposed to direct sunlight, the use of tinted glass may not be an adequate response. Internal Venetian blinds, if deployed correctly, may prevent glare and provide visual comfort to workers near the window – but they require frequent adjustment and reduce the depth at which daylighting may still be enjoyed. A light shelf with an exterior part to shade the view pane from direct sunlight in summer and an interior part to reflect light to the ceiling resulted in superior daylighting and better visual comfort in all room configurations. It is suggested that since windows in offices fulfil multiple roles (daylighting, natural ventilation and a view outdoors), their functioning could be improved by subdividing them into panes to optimize their provision. INTRODUCTION Lighting comprises a significant part of the energy used in office buildings: Estimates range from 35% in Adelaide, which has a mild sunny climate (Blanchard, 2005), to 23% in cooler, overcast London, where heating requirements are greater (Majoros, 1998). Because office buildings are occupied mostly during the daytime, the potential for energy savings through substituting daylighting for electric lights is high. Simulation studies of typical office spaces in Belgium (Bodart and De Herde, 2002) show that approximately 40% of this energy can be saved by automatic dimming of artificial lighting when sufficiently illumination can be achieved by daylighting alone. Demonstration buildings have shown that energy consumption for lighting and HVAC can be reduced by 50% (Voss et al., 2005), but savings depend on the proportion of occupants that have access to external windows and on the ratio of window area to floor area (Krarti et al., 2005). The predominant strategy for increasing daylight utilization has been to design extensive glazed The first and third authors are faculty members of the Dept. for Man in the Desert at the Jacob Blaustein Institutes for Desert Research. The second author was a PhD student at the Dept. and is now an independent consultant and director of Research and EcoDesign. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 67 facades. However, when the sun is out, a person near the window may be exposed to high, asymmetrical radiant loads and very high levels of illumination: mean radiant temperatures of up to 47oC and illumination levels in excess of 30,000 lux were measured indoors near a clear window on a sunny winter day (Erell et al., 2004). To counter this, the occupants of such buildings may require curtains or blinds to avoid glare or overheating. Unfortunately, they typically leave their blinds deployed at all times, and employ electric lighting even when external illumination levels are sufficient to provide adequate illumination. In warm sunny locations this behaviour results in a double penalty: First, because in summertime external gains from a glazed area are greater than from an opaque wall; and second, because the use of electric light increases internal gains and increases the load on the A/C system. Visual discomfort caused by excessive light is referred as ‘glare’, defined by the CIE as the “condition of vision in which there is discomfort or a reduction in the ability to see details or objects, caused by an unsuitable distribution or range of luminance, or extreme contrasts”. Evaluation of glare is sometimes not straightforward, as most traditional indices for visual comfort encounter difficulties either at extreme luminance levels or in the evaluation of large sources of light or sources not mounted close to the ceiling plane, such as vertical windows (Osterhaus, 2005). If the potential sources of glare cover a significant part of the visual field of the observer, the adaptation of the eye to higher luminance reduces the glare sensation and contrast effect. When the glare source is small, however, the observer’s adaptation level, which is determined by the luminance of the background, is virtually independent of the light source. The Daylight Glare Probability (DGP) is probably the index best-suited for highly luminous environments. It is based on the vertical illuminance at the eye as well as on the luminance of the glare sources, their solid angle and their position index. The DGP, which was calibrated empirically on the basis of controlled experiments at Freiburg and Copenhagen, has values that range from 0 to 1, and is calculated as follows (Wienold and Christoffersen, 2006):  L2,i   s    0.16 DGP  5.87 10 5  Ev  9.18 10 2  log1   1s.87 2   E P i v i   where Ev is the vertical illuminance at eye level [lux], Ls is the luminance of the ith source contributing to the glare [cd m-2], s is the solid angle subtended by the source and P is the dimensionless Guth position index. An index incorporating a probability is preferable for most rating schemes, since there is a large variation of responses when comparing visual comfort, especially with respect to glare Osterhaus (2005). Thus, a 100-fold increase in luminance may be required to arrive at the same subjective glare rating between the least sensitive and most sensitive subjects (Osterhaus and Bailey, 1992). Furthermore, responses of individuals are often inconsistent when assessing the same environment on different occasions. This may be due to acclimatization to current daylight levels outdoors, which vary greatly on a daily basis but also between locations. For example, illuminance levels of over 75,000 lux occur on nearly two-thirds of the days in Tel Aviv, but barely on one day in ten in Berlin. The search for a universal index is further complicated by the fact that there are apparently persistent cultural differences in illuminance preferences (Belcher, 1985). Indeed, Veitch and Newsham (1996) suggest that the perception of lighting quality is affected by behavioural factors that are not accounted for in any of the existing indices. Nonetheless, current standards for lighting – artificial or natural – make no allowance for such disparities among countries, or for other sources of individual and contextual variability. Much of the research on daylighting has been carried out in overcast locations, and the current study seeks to complement this body of knowledge in a highly luminous environment. Its aims are to evaluate several daylight control systems in such locations, and to evaluate the use of the DGP for subjects acclimatized to these kinds of lighting conditions. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 68 METHODOLOGY A controlled experiment was carried out to investigate occupant preferences with respect to several façade designs for better visual comfort and glare control. The test subjects were asked to perform several tasks representing typical office work, such as reading from paper and typing to a computer, in a room furnished to resemble a normal office. Measurement of light – luminance and illuminance – was carried out concurrently in a second adjacent room where conditions were almost identical. The position of the work station and its orientation with respect to the window were varied according to a predefined schedule. To reduce the possibility of inadvertent bias because of minor differences between the rooms, the roles of the two rooms were alternated. The subjective responses to a questionnaire were analysed and compared to prevailing conditions during the test sessions. Test rooms The test rooms were 2.7m by 3.5m wide and 3.05m high, with white walls and ceiling (reflectance 0.75), and a terrazzo floor (reflectance 0.45). An aluminium-framed window 1.34m wide by 1.76m high was located near the middle of the (long) south-facing wall. The windows comprised a 'view pane' consisting of horizontal sliders 114cm in height beginning 95cm above the floor, and a fixed 'daylighting pane' 62 cm high above them. An external roll-down shutter remained open for the duration of the tests. Both test rooms were furnished with identical computer work stations installed on small tables fitted with wheels, to allow easy repositioning by test subjects. Each station included a personal computer with a 19" LCD monitor, keyboard and mouse. Walls were decorated with coloured posters to enhance the visual environment and to reduce glare from uniform white surfaces, simulating a real office. Daylight control strategies Three daylight control strategies were tested, singly and in combination: 1. Tinted glass: The use of tinted glass is widespread in office buildings in most sunny locations, both to reduce solar gains (so-called ‘solar control’ glazing) and to reduce glare near the windows. The view pane of the window was fitted with either clear double-glazed panes (VT=0.79) or with similar panes equipped with a tinted foil with a total light transmittance of 0.47 (VT47). 2. Venetian blinds: Venetian blinds are the most common internal shading device found in offices. The window was equipped with standard white blinds with curved slats (25mm wide, 1.5mm curvature) that covered the entire glazed area. After installation of the light shelf (see below), these blinds covered the view pane only. 3. Light shelf: A ‘portable’ light shelf comprising an internal element and an external one was attached to the window at a height of 2.1 meters above the floor (Figure 1a and b). The shelf had a curved section (convex on the interior, concave on the exterior), was 50 cm deep and extended 20cm beyond each side of the window. Light shelf installed for the experiment, seen from interior (left) and exterior (middle); Figure 1 setup for HDR photography used to establish luminance (right). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 69 Monitoring internal environmental conditions     While subjects were carrying out the test procedure, the following parameters were monitored: Indoor air temperature, relative humidity and CO2 concentration Indoor illuminance: 4 horizontal measurements of general room lighting - 10 cm above desk as well as half a meter from wall centres (TES-1332A light meter), each of which received different amounts of light; Illuminance at the task area (vertically next to the computer screen and horizontally on the keyboard); Vertical illuminance at the eye; Total exterior illumination received at the window (on a vertical plane) and the net flux transmitted to the interior, adjacent to the window pane (or 10 cm behind the venetian blinds when these were in place). Indoor luminance was evaluated from HDR images taken from the subjects’ eye position looking towards the task area and the window (with a Coolpix 5400 camera and FC-E8 fish eye lens, Figure 1c). Images were calibrated using spot measurements (Minolta luminance meter LS-110). Outdoor horizontal illuminance (TES-1332A light meter) Questionnaire The subjective sensation of glare experienced by the participants in the controlled daylighting experiment was recorded by means of a questionnaire consisting of 4 sections: a. Personal (demographic) questions b. Subject assessment of the rooms and quality of the visual environment c. The subject’s explanation of their individual preferences in setting up the subject-defined test environment d. Questions on subject’s perception of indoor climate within the room Experimental procedure Experiments in the test rooms at the Sde Boqer campus of Ben-Gurion University (30.8 N 35.1E) took place between October and March at 10:30-15:00. Sessions were conducted only on sunny, cloudless days, characterized as a CIE Standard Clear Sky (Type 12), with a (vertical) illuminance on the window in excess of 50,000 lux. Each subject was asked to carry out a sequence of tasks and to fill in questionnaires to obtain a subjective rating of the visual comfort of the simulated office environment. While the subjects were thus occupied, research staff carried out measurement of the visual environment in the second adjacent test room, which was identically oriented and equipped. The subjects were requested to follow instructions given in an interactive PowerPoint presentation. This provided the basic structure of the experiment, as follows: 1. After carrying out typing tasks requiring them to copy text from both the screen and from paper, in order to become acquainted with the procedure, subjects were asked to fill in a questionnaire with basic demographic information. 2. The test room was then arranged in the first test configuration, according to a predetermined schedule established to test all possible combinations of desk location relative to the window and of the daylight control strategy. The subjects then performed the first test unit: They watched one of three short videos, to allow them to become acclimatized to the new visual conditions, then performed two typing tasks and finally answered a questionnaire to give their assessment of the visual environment. 3. The test room was then arranged in a second configuration by changing the glazing type. The subjects then performed the second test unit, which was identical to the first one. 4. After completing the second test unit, subjects were asked to arrange the test room so as to maximize comfort. They were allowed to choose either type of glazing, to select any location in the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 70 room for the computer work station, and to manipulate the venetian blinds to any position. Following this, they performed the third test unit. 5. After completing all three test units, the subjects answered one final questionnaire to give their overall assessment of the work environment in general. RESULTS A total of 59 subjects completed the entire survey procedure. Subjects were mostly students and faculty at the Blaustein Institutes for Desert Research, between the ages of 17 and 48 (average age 31), equally divided by gender and from diverse ethnic backgrounds. One third of the subjects wore glasses. Effect of Tinted Glazing 100% 75% Tinted (VT 0.47) All Categories Percentage of subjects (by glass type) The total number of surveys for clear and tinted glass was the same. Questionnaire findings indicated that installation of tinted glass improved overall satisfaction with the visual environment. However, as the mosaic plot in Figure 2 shows, in spite of much lower levels of illumination in the room, the majority of respondents (as indicated by the width of the two right-most bars) still rated the office as either ‘very uncomfortable’ or ‘somewhat uncomfortable’, irrespective of the type of glass (indicated by colour – red for clear and blue for tinted). The main contribution of the tinted glass appears to have been to mitigate the extreme condition somewhat: Only one third of subjects who rated the office very uncomfortable did so when tinted glass was installed – but almost 60% of those who rated the office ‘somewhat uncomfortable’ did so in spite of the presence of tinted glass. The Daylight Glare Probability, estimated using digital HDR images of the test office with tinted glass, is significantly lower than for clear glazing (Figure 2, right). However, even though the reduction in the visible light transmission of the glazing is 40 percent, compared to the standard clear glazing used as a reference – the median value of the DGP index was still over 40%, indicating that close to half of the occupants would be likely to suffer from glare in such conditions. In other words: In luminous climates such as Israel, even very substantial reductions in the visible light flux are not, in themselves, sufficient to prevent glare. 50% 25% Very Comfortable Clear (VT 0.79) Very Uncomfortable How suitable are the lighting conditions in this room for office work? Effect of tinted glass on visual comfort responses of subjects. Left: Mosaic plot of Figure 2 subjective responses from questionnaires. Right: DGP predicted by analysis of luminance from HDR images of subjects' field of view when facing the computer display. Effect of Venetian Blinds Subjects were offered the option of deploying Venetian blinds either in conjunction with other light control means such as tinted glazing (with or without a light shelf) or as the only method of controlling 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 71 1 75% Tinted (VT 0.47) 50% 25% Clear (VT 0.79) 0% Very Comfortable DGP facing work task 100% All Categories Percentage of subjects (by use of blinds) exposure to daylight. The degree of exposure to daylight was determined by the subjects, who were allowed to deploy the blinds according to their personal preference, manipulating both the angle of the slats and the proportion of the window shaded (by lowering or raising them). After the position of the blinds was fixed by the subject, the illumination levels in the test room were measured by a technician, who then manipulated the position of the blinds in the reference room to obtain identical illumination levels throughout the room. The subjects were allowed to begin their evaluation of the specific configuration of blinds only after this calibration procedure was completed satisfactorily, and the full set of measurements could be carried out in the reference room. The responses of the subjects who were allowed to deploy venetian blinds were in very good agreement with the predicted evaluation of the visual environment given by the DGP index. Analysis of the questionnaire indicated that when the venetian blinds were deployed, subjects were in fact almost always satisfied with the resulting visual environment, with only a very small proportion still rating conditions as either ‘somewhat uncomfortable’ or ‘very uncomfortable’ (Figure 3). Although the combination of tinted glass and venetian blinds was slightly more likely to produce ‘very comfortable’ conditions than Venetian blinds alone, the contribution of tinted glass to obtaining merely ‘comfortable’ conditions was negligible. Subjects who gave this evaluation were equally likely to have clear glass as tinted glazing. 0.8 0.6 0.4 0.2 0 Very Uncomfortable Use of blinds How suitable are the lighting conditions in this room for office work? Effect of Venetian blinds in combination with clear or tinted glass on visual comfort Figure 3 responses of subjects. Left: Mosaic plot of subjective responses from questionnaires. Right: DGP predicted by analysis of luminance from HDR images of subjects' field of view when facing the computer display. Effect of Light Shelf The role of a light shelf is to redirect sunlight to modify its distribution in the room, and to provide partial shading in hot weather. The experiment evaluated the effect of a light shelf comprised of both an external element and an internal component on the light distribution in the room, with the venetian blinds either deployed or not. The subjects’ evaluation of the resulting illumination was recorded as part of the questionnaire they were asked to fill. When the Venetian blinds were not deployed the light shelf reduced the light level on the desk significantly. In this mode, its primary role was as a shading device, blocking part of the direct sunlight impinging on the window. As Figure 4 (right) shows, the light shelf reduced median illumination in the room from over 13,000 lux to about 3,500 lux. However, in spite of the reduction in illumination caused by the light shelf, for much of the time light levels were still well above recommendations for visual comfort, both on the desk and on the computer display. Thus, although the light shelf reduced the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 72 3,000 2,500 2,000 1,500 1,000 500 0 Horizontal illuminance at task (lux) Horizontal illuminance at task (lux) probability for glare substantially, it by no means eliminated it: the median value of the DGP was 39%, compared to 57% with no light shelf. In sunny conditions when solar elevation is low enough to allow substantial penetration of direct sunlight, additional shading (such as blinds or curtains) may be required for the lower part of the window. When Venetian blinds were deployed, the primary role of the light shelf was to improve light distribution in the room, providing substantially higher levels of illumination in areas not adjacent to the window (Figure 4, left). More importantly, although the light shelf led to median light levels that were twice as high as for window equipped with Venetian blinds only, the probability for glare, as measured by the DGP indicator, remained very low. 60,000 50,000 40,000 30,000 20,000 10,000 0 Effect of light shelf on horizontal illuminance of work surface. Left: with Venetian Figure 4 blinds deployed on view pane of window. Right: without blinds. DISCUSSION AND CONCLUSIONS The effect of tinted glazing on visual comfort was quite modest. The almost universal use of such glass in office buildings in sunny locations, and its wide application even in overcast ones, should not obscure the fact that even low levels of light transmission do not guarantee comfort. Venetian blinds are known to be an effective means of controlling penetration of direct sunlight to the office interior, and the ease with which occupants can manipulate their position is a great advantage compared with most other forms of solar control. This suggests that if occupants of small offices take sufficient care in adjusting the position of the blinds, taking into account both the extent of deployment (or proportion of the window left fully exposed) and the angle of the slats – they can enjoy the benefits daylight without being exposed to glare. However, Venetian blinds are not a panacea: maintaining visual comfort with natural light requires frequent adjustment of the blinds in response to changing quantity and quality of external light: variations may be the result of changing weather or cloud cover, but also because the diurnal path of the sun that means a window may be illuminated by direct sunlight for only several hours of each day. The findings of a field survey (Erell and Kaftan, 2011) suggest, however, that such continuous adjustment is rarely carried out in practice. Furthermore, while work stations adjacent to the window may be well-served by Venetian blinds, work areas located further away might require artificial light if the blinds are deployed in response to conditions near the window. This is exacerbated if so-called 'solar control (tinted or reflective) glazing is installed to reduce overheating. As this study has shown, having a light shelf is beneficial both when the venetian blinds are deployed, and when they are not. In the former case, they reduce illumination levels which might otherwise be excessive, thus reducing glare. In the second instance, they increase illumination levels, especially in parts of the interior not adjacent to the window – without increasing the probability for glare. A light shelf with blinds deployed below it provides high quality daylight: reducing glare in the working area adjacent to the window while enabling higher illuminance levels deeper in the office. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 73 Therefore, a fundamental daylighting solution for buildings in sunny climates may consist of an upper daylight window, a lower view window, a light shelf, and daylighting control systems (such as blinds). Although it is beyond the scope of the present paper, it may be noted that the experiment also demonstrated that the relation of the working position to the window also has a great bearing on visual comfort. The worst orientation is facing the window, whereas lighting from the side or diagonally results in less glare. However while the desk position may, on its own, mitigate glare to some extent, additional means such as daylighting control systems were still required in sunny conditions. Daylighting design is frequently concerned with obtaining sufficient light, typically measured by metrics such as illuminance of a horizontal work surface or a daylight factor. These indicators ensure that a minimum level of natural light is obtained, contributing to health and alertness. However, these measures are insufficient when it comes to predicting visual comfort – which is better assessed by means of the Daylight Glare Probability. The calculation of this last metric requires detailed information and appropriate computer software, but demonstrated a very good correlation with subject responses in this experiment. ACKNOWLEDGMENTS The research was supported by funding from the Israel Ministry of Energy Water and National Infrastructure under contract 2006-8-44/26-11-013. Wolfgang Motzafi-Haller installed the monitoring equipment, supervised the subject test sessions and assisted with data processing. The authors are especially grateful to Jan Wienold for providing the EVALGLARE computer tool and for advice on setting up the software. REFERENCES Belcher M. 1985. Cultural aspects of illuminance levels. Lighting Design and Application, 15(2):49-50. Blanchard, C. 2005. Submission to the Energy Efficiency Inquiry, Australia. Bodart M. and De Herde A. 2002. Global energy savings in office buildings by the use of daylighting. Energy and Buildings, 34:421-429. Erell E., Etzion Y., Carlstrom N., Sandberg M., Molina J., Maestre I., Maldonado E., Leal V. and Gutschker O. 2004. SOLVENT: Development of a reversible solar-screen glazing system. Energy and Buildings, 36:468-480. Erell E., Kaftan E. and Motzafi-Haller W. 2011. Daylighting for visual comfort and energy conservation in offices in sunny locations. Final research report to the Israel Ministry of National Infrastructures. (Partly in Hebrew) Krarti M., Erickson P. and Hillman T. 2005. A simplified method to estimate energy savings of artificial lighting use from daylighting. Building and Environment, 40:747-754. Majoros, A. 1998. Daylighting. PLEA International and Queensland University, Brisbane, 76 pp. Osterhaus W. 2005. Discomfort glare assessment and prevention for daylight applications in office environments. Solar Energy 79:140-158. Osterhaus W. and Bailey I. 1992. Large area glare sources and their effect on discomfort and visual performance at computer workstations. Proceedings of the IEE Industry Applications Society Annual Meeting, Houston, Texas 4-9 October, pp. 1825-1829, and LBNL Report No. 35037. Veitch J. and Newsham G. 1996. Determinants of lighting quality II: Research and recommendations. Proceedings of the 104th Annual Convention of the American Psychological Association, Toronto, Canada. Voss K., Herkel, S., Pfafferott, J., Löhnert, G. and Wagner A. 2005. Energy efficient office buildings with passive cooling – Results and experiences from a research and demonstration programme in Germany. Proceedings of the Conference on Passive and Low Energy Cooling of the Built Environment, Santorini, Greece. Wienold, J. and Christoffersen, J. 2006. Evaluation methods and development of a new glare prediction model of daylight environments with the use of CCD cameras and RADIANCE, Energy and Buildings 38(7):743-757. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 74 Improving Outdoor Urban Environments: Three Case Studies in Spain Patricia Martin del Guayo, MArch Simos Yannas, DiplArchEng AADiplGrad(Hons) PhD [Architectural Association, London UK] Patricia.Martin@aaschool.ac.uk [Architectural Association, London UK] ABSTRACT Public life is as intrinsically linked to the physical and material settings as it is to the climatic environment. The physical properties of a space, such as materiality, temperature, or light, can enhance or inhibit people from using and enjoying it. When architects and urban designers deal with the physical properties of a space, and therefore modify its material, thermal, and lighting characteristics, they influence the social environment as well. This paper describes a field study undertaken in Spain. Its aim was to understand the implications of climate-responsive design in urban public spaces over the physical, climatic, and social environment. Three case studies in Spain were selected and the resulting climatic and social environment analyzed. The three structures are: Metropol Parasol in Seville, Ecobulevar in Madrid, and Plaza Pormetxeta in Barakaldo. Each of the three case studies is located in a different climatic region within the Spanish geographical context. The study included on-site measurement of climatic conditions and observations of people behaviour in the public spaces. These were correlated and compared to nearby public spaces to assess the success of the structures in creating lively urban areas. The study suggested that, while climatic conditions play a key role in the success of public spaces, climate-responsive design must also consider other aspects of urban design such as social activities, accessibility, preconceptions of the space, and visual delight. INTRODUCTION During the last decade, Spanish cities have witnessed the proliferation of climate-responsive structures, such as canopies or wind towers, in their outdoor public spaces. The main purpose of such structures was to adapt these spaces to existing microclimatic conditions and promote outdoor comfort. In many Spanish cities, staying outdoors in summer at midday can be too uncomfortable and even dangerous. This is due to high temperatures and intense solar radiation that are intensified by the Urban Heat Island effect and climate change. Some studies have analyzed the effects of the structures over the climatic conditions around them (Soutullo et al., 2007). However, no research has addressed their effects on the use of the public space by citizens. What have not been assessed yet are the implications of climate-responsive structures for citizens’ everyday lives. By studying people’s perceptions and reactions to these structures, climate-responsive design could be improved to fulfil its ultimate aim: to promote the use of public spaces. Architects and urban designers should follow an integrative approach to the design of public spaces, one that combines the search for a comfortable environment through climate-responsive design, with a lively social environment that improves the quality of life within cities. The study described in this paper analyses the resulting climatic and social environment of three climate-responsive structures in Spanish urban public spaces. The three structures are: Metropol Parasol in Seville, Ecobulevar in Madrid, and Plaza Pormetxeta in Barakaldo. Each of the three case studies is located in a different climatic region within the Spanish geographical context. The structures were selected, taking into account the objectives of the design, techniques applied, and location. The study 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 75 followed a uniform approach for the three structures. It reviewed and analyzed the original authors’ documents, as well as any other documentation and opinions published in the media. Each of the case study sites was visited, and one-time on-site measurements of climatic conditions were recorded. In addition, while climatic temperatures were recorded, interviews with people using the space were carried out and observations of people’s behavior noted. Measurements were also taken simultaneously in nearby public spaces, in order to compare and assess the strategies used. The analysis of the case studies was divided into two groups: climatic environment—temperature, relative humidity, wind speed, and solar radiation; and the social environment—number of people using the public space and people’s perceived comfort. Table 1 presents the selected case studies, their location, and the strategies used. Table 1. Case Studies City Project year Geographic coordinates Climate Area of the public space Technique Metropol Parasol Seville 2005-2010 37°22′ N, 5°59′ W Mediterranean 32,605 m2 Canopy, raised plaza, water fountains, low heat storage materials Ecobulevar Madrid 2004-2007 40°23′ N, 3°43′ W Continental 27,500 m2 Solar shields, wind towers, vegetation, lightcolored materials Plaza Pormetxeta Barakaldo 2003-2010 43°17′ N, 2°59′ W Oceanic 27,102 m2 Canopy, wind shields Pictures of the three case studies. From left to right: Metropol Parasol in Seville, Figure 1 Ecobulevar in Madrid, and Plaza Pormetxeta in Barakaldo. CASE STUDY 1: METROPOL PARASOL, SEVILLE Metropol Parasol is located in the historic center of Seville, in a large void of the dense medieval fabric. The structure consists of an extensive canopy of 150 by 70 meters 25 meters above street level, supported by six gigantic columns. The public space that was the object of study, Plaza Mayor, is located underneath the canopy on a platform raised 5 meters above street level. It has a total surface of 10,600 m2 and is approximately 85 m wide and 140 m long. It is furnished with four concrete semi-circular benches, three small fountains on the borders, and a playground. The materials used are clear granite for the pavement, timber for the canopy, and concrete for the structure, which becomes visible at the bases of the pillars. The purpose of the canopy was to create a comfortable environment in the plaza, where people congregate and big public events take place. The canopy protects the space from direct solar radiation, and the plaza is raised above street level to increase air flow. In addition, the selected light tones of the materials are appropriate for reflecting solar radiation. Moreover, the use of fountains, although scarce, helps to decrease air temperature by evaporative cooling. The construction work of Metropol Parasol started in 2005 but soon encountered technical problems that forced the construction system to change and delayed its completion to April 2011, also doubling its estimated costs. Considering the current economic crisis that Spain, and especially the region of Andalucía, is going through, the delays and cost overruns of the building have resulted in much public controversy. Nevertheless, the project has been widely published in specialized magazines and journals and generally supported by the professional community, which has awarded the structure with several international prizes of architecture. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 76 Climatic Environment Temperature (˚C) The main issue when dealing with outdoor comfort in Seville is its high temperatures and solar radiation in summer, which are among the highest in Europe. On July 25, 2012, on-site measurements of temperature, humidity, and wind speed were recorded every hour from 10:00 a.m. to 10:00 p.m. at Plaza Mayor. Simultaneously, on-site measurements were taken at Plaza del Cristo Burgos, a highly vegetated park with two playgrounds located 200 meters west of Metropol Parasol. Temperatures recorded at Plaza del Cristo Burgos were very similar to those at Plaza Mayor between 10:00 a.m. and 1:00 p.m. and between 6:00 p.m. and 10:00 p.m. However, between 2:00 p.m. and 5:00 p.m., when the sun is higher, temperatures were cooler in Plaza del Cristo Burgos. Temperatures in both public spaces were also compared to the climatic data registered at the airport meteorological station. Figure 2a shows the comparison among temperatures in these three points of the city. As shown, temperatures at the airport, located in a non-urbanized area outside of the city, are much lower than in the city during the night, indicating the presence of the Urban Heat Island. In addition, spot temperature measurements were taken at different public spaces across the city, where devices such as canopies were in use to improve pedestrian outdoor comfort. Temperature measurements at these Seville public spaces were taken at different times of the day. Figure 2a shows these measurements as dots to compare them to temperatures at Plaza Mayor. It can be inferred from the graph that Plaza Mayor performed better climatically than most of the other spaces, decreasing temperatures and creating a more comfortable urban environment. (a) Temperatures measured at Metropol Parasol compared to temperatures measured at Figure 2 other public spaces and (b) People recorded staying at Metropol Parasol and at Plaza del Cristo Burgos. Social Environment The number of people present at Plaza Mayor was recorded in parallel to climatic measurements and compared to those at Plaza del Cristo Burgos. As shown in Figure 2b, the number of people at Metropol Parasol was lower than that at Plaza del Cristo Burgos. In addition, Plaza Mayor did not attract as many people as other public spaces of the city where measurements were taken. A total of 25 interviews were undertaken; questions focused on two topics: the reasons for staying at Plaza Mayor and people’s thermal perceptions. Thermal perception was further evaluated by asking interviewees to compare their thermal comfort with the comfort they expected to have in adjacent streets. According to the results, the main reasons for staying at Metropol Parasol were: first, more comfortable climatic conditions; second, attractiveness of the space; and third, the playground. Moreover, all those interviewed found the environment “warm” or “too warm”, but most of them stated that the temperature at Plaza Mayor was more comfortable than that in adjacent outdoor spaces. In fact, those who perceived a more comfortable climatic environment were those who nominated the attractiveness of the space as the reason to be there, suggesting an influence of the visual environment on thermal perceptions. All those interviewed had a preconceived idea of Metropol Parasol derived from the exhaustive coverage of the building in the local media as well as in the tourist guides and other city information books. In the case of Seville inhabitants, these preconceived ideas were related to its excessive costs. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 77 a) Figure 3 b) (a) Plaza Mayor beneath Metropol Parasol (b) Plaza del Cristo Burgos. Findings The climatic techniques applied at Metropol Parasol decreased temperatures by over 3˚C compared to adjacent open areas. When temperatures were compared to those in other public spaces of the city, Plaza Mayor performed better than many, providing a more comfortable climatic environment. Nevertheless, the public space at Metropol Parasol did not draw as many people as other public spaces. Raising the plaza above the street level resulted in more air flow and decreased temperatures. However, the separation of the space from the street, shops, bars, restaurants, and pedestrian and vehicular flows discouraged people from going there. Lack of benches, activities and attractions left the space empty most of the day. In addition, the fieldwork suggested that the powerful visual environment of Metropol Parasol influenced thermal perception. Preconceived ideas, mainly based on economic and political issues, influenced the assessment of the climatic environment. The experience of the thermal environment could not be isolated from issues such as the visual qualities of the structure and its economic and political implications (Nikolopoulou & Steemers, 2003). CASE STUDY 2: ECOBULEVAR, MADRID Located in the suburban development of Vallecas in Madrid, Ecobulevar is the redefinition of an existing 550 m by 50 m boulevard according to two objectives: social and environmental. The design team, Ecosistema Urbano, considered trees to be the perfect tools to achieve both objectives (Ecosistema Urbano, 2004). However, according to the project brief, a tree could take 25 years to satisfy these social and climatic needs. Ecosistema Urbano proposed the construction of three artificial trees that would function climatically and socially from the start. These are composed of cylindrical metallic structures that are around 18 m high with 25 m diameters. Each structure improves climatic conditions using different strategies. The evaporative tree comprises 16 cylindrical wind towers surrounding the principal space formed by a larger cylindrical metallic structure. The wind towers, which are oriented to catch the prevailing winds in the area, inject atomized water to the air flow that passes through them. At the bottom, six nozzles drive the cooled air into the inner space. The second tree, the vegetal tree, is covered by vegetation to provide shadow and decrease temperatures. Finally, the recreational tree is enclosed by an inner screen that, while shading the interior, can be used as a TV screen for different activities. All three trees are located over a modified topography that confines the space and protects it from wind flows. Climatic Environment On-site temperature, relative humidity, and wind speed measurements were taken on July 19, 2012 between 12:00 p.m. and 9:00 p.m., both inside and outside the artificial trees. The outdoor temperatures measured on the boulevard that day reached 41 °C, wind flew below 2m/s, and relative humidity levels remained at around 10%. Figure 5a shows the temperatures that were measured inside each tree and at the street hourly. The temperatures remained lower inside the trees than outside them over the course of the visit (see Figure 5a). The greatest differences were found inside the evaporative tree, which is 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 78 surrounded by 16 wind towers. The maximum temperature difference between the interior of this tree and the outdoor temperature equaled 2.4 °C at 4:00 p.m. On average, the evaporative tree created a thermal environment that was 1.42 °C cooler than outside. The average temperature measured inside the vegetal tree was merely 0.42 °C lower than the exterior temperature, and the highest difference registered was 0.6 °C at 6:00 p.m. The recreational tree presented an average temperature of 1.05 degrees lower than the exterior temperature. The peak in thermal difference was recorded at 4:00 p.m., with an environment that was 1.5 °C cooler inside. Inside the artificial trees, air velocity remained lower than outside, with a maximum value of 1.2m/s, while outside the maximum value recorded was 3.0m/s. a) Figure 4 b) c) (a) Evaporative Tree (b) Vegetal Tree and (c) Recreational Tree (a) Temperatures measured inside and outside the artificial trees at Ecobulevar and (b) Figure 5 People recorded staying inside and outside the artificial trees at Ecobulevar. Social Environment The number of people staying inside each artificial tree and in the space in between them was recorded in parallel to climatic measurements. Due to the high temperatures that day, the number of people staying outdoors was minimal. Figure 4b shows the number of people staying inside each tree and outside along the boulevard. As seen in the graph, the total number of people increased significantly after 6:00 p.m. and reached its peak at 8:00 p.m., with 27 people or groups of people sitting there. Among the artificial trees, the vegetal tree was preferred. It held more people in five out of the six measurements, with a total of 40 people or groups. The evaporative tree, however, hosted only five people or groups during the six hours of fieldwork and remained empty for most of the time. These measurements contrast radically with the climatic measurements that were recorded, where the evaporative tree provided the most comfortable thermal environment and the vegetal tree the less comfortable. The recreational tree remained vacant until 6pm, but was then used by a total of nine people or groups until 9pm. Similarly, the outdoor spaces between the trees were not occupied until 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 79 7pm, when a total of eight people or groups used the seating available there for the next two hours. In order to understand individual environmental perceptions at Ecobulevar, a total of 25 people were interviewed inside and outside the artificial trees. The interviews followed the same questionnaire as in the previous case study. In this case, the main reasons for staying at Ecobulevar were: first, attractiveness of the space; second, more comfortable climatic conditions; and third, the amenities offered in the space. All of those interviewed found the environment to be “too warm” or “comfortably warm”. People inside the artificial trees were asked how temperatures were comparing to outside. 90% of those who liked the artificial trees stated that the climatic conditions inside them were better than outside. Conversely, 75% of people who found the space inside the artificial trees to be unattractive stated that the climatic conditions inside were equal to those outside. Findings The fieldwork revealed that the artificial trees did provide a more comfortable thermal environment than that experienced outside of them, but the effects of the strategies varied, depending on the technique applied. Specifically, the evaporative tree was the most effective for decreasing air temperatures and protecting from solar radiation, while the vegetal tree barely modified the existing climatic conditions. However, when the study analyzed the social performance of the artificial trees, the vegetal tree was the most effective. The number of people staying inside the vegetal tree was considerably higher than those inside the other two trees. Based on the conversations and interviews carried out, people chose the vegetal tree for its facilities, specifically the benches and swings. Moreover, the direct connections between its interior and exterior in all directions and without architectural barriers offered a constant visual and physical relation with the outside, which facilitated the access and use of the tree. CASE STUDY 3: PLAZA PORMETXETA, BARAKALDO Plaza Pormetxeta is located in the city of Barakaldo, Spain, which is part of Bilbao’s metropolitan area. The plaza connects the town center with the river through a series of walkways that overcome a 20meter height difference. These walkways are made of steel plates and paved with hexagonal ceramic tiles. At some points, the steel plates fold over the walkways, providing protection from direct solar radiation. The space between the walkways forms a plaza of 6,500 m2 furnished with benches and playgrounds beneath tree-shaped canopies. This space constitutes the public space object of study, as the tree-shaped canopies aim to create a more comfortable space for citizens by providing shade. The canopies, which are called Stone Trees in the project, cover a 750 m2 area and are 11.5 meters high. They are constituted by a steel structure of pillars and beams imitating the trunks and branches of a group of trees. On top of the structure, another box structure made of steel holds a metallic mesh and the stones that form the top cover for the canopies. According to the project brief, the Stone Trees “act as an atmospheric device that balances the exuberant natural surroundings” (MTM Arquitectos, 2013). Climatic Environment Barakaldo has an oceanic temperate climate with low temperature variations through the year. Summers and winters are both mild seasons with no extreme temperatures. Precipitation is very abundant in Barakaldo, and clouds cover the sky on 330 days per year on average. Moreover, Barakaldo is located in the area with the lowest solar radiation values of Spain, with a mean daily solar global radiation value of 3.54 KWh/ m2. On August 21, 2012, on-site measurements of climatic conditions were taken every hour from 11:00 a.m. to 7:00 p.m. beneath one of the Stone Trees. Climatic conditions were unusually warm, with maximum temperatures around 30°C, covered skies, relative humidity levels around 60%, and low wind flows. The maximum temperature recorded beneath the artificial trees was 27.2°C at 1:00 p.m. and the minimum was 24.8 at 7:00 p.m. The maximum humidity level measured was 65.2 at 6:00 p.m. and the minimum was 58.3 at 11:00 a.m. Wind speeds remained similar during the entire measurement period, with maximums around 2 m/sec. These data were compared with climatic measurements registered in Plaza del Desierto, located 150 m from Plaza Pormetxeta. The design for Plaza del Desierto did not consider any strategy for adapting the space to the local microclimate or for generating a more comfortable thermal environment. Figure 6a shows the temperatures measured in both spaces. As shown, temperatures at Plaza Pormetxeta were consistently lower than those in Plaza del Desierto. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 80 (a) Temperatures measured hourly at Plaza Pormetxeta and Plaza del Desierto on Figure 6 August 21, 2012 and (b) People recorded staying in both public spaces at the time climatic measurements were recorded. a) Figure 7 b) (a) Plaza Pormetxeta (b) Plaza del Desierto Social Environment The number of people staying on the plaza, the spaces they occupied, the number of people crossing the space, and the routes they chose were noted. These measurements were recorded simultaneous to climatic measurements. During the nine hours of field study, the total number of stays at Plaza Pormetxeta was 9. Conversely, the number of stays recorded at Plaza del Desierto was significantly higher, reaching 28 by the end of the visit (see Figure 6b). This plaza is surrounded by several cafes, restaurants, and shops that draw citizens to the public space. Although the number of stays in Plaza Pormetxeta was low, it was observed that the number of people that crossed the space was considerably higher. Many people used the pathways to go from the city center to the new urban area by the river and vice versa. But the difficult access to the plaza by labyrinthine pathways and the lack of services and attractions dissuaded people from sitting and remaining there. Despite the scarce number of people staying under the stone trees at Plaza Pormetxeta, some interviews were carried out to assess the environmental perception of the space. One woman interviewed pointed out that the geometry of the space and the materials used in the construction of the plaza looked dangerous. Specifically, she noted the use of heavy stones suspended on metallic meshes as solar protection as a threatening system that produced an uncomfortable environment. Besides, the enclosed and irregular space formed by the pathways to protect from prevalent winds generated numerous hidden corners and dark spaces, producing a feeling of vulnerability. These statements were also supported by neighbors and the local media (Llamas, 2010). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 81 Findings On-site measurements proved that Plaza Pormetxeta’s canopies and shields decrease temperatures up to 2 ˚C on hot days. However, these elements produced other effects such as darkness and threat that dissuaded people from staying at the plaza. Some residents defined the perceived environment as “dark” and “cold,” referring not literally to the thermal environment, but meaning “unfriendly” and “unwelcoming.” As a result, the climatic environment generated influenced negatively the built and the social realm. The public space was rarely used and remained empty for most of the time. People stayed in nearby public spaces and used Plaza Pormetxeta as a pedestrian pathway to move through the city but did not stay there. Consequently, Plaza Pormetxeta did not fulfil its role as a public space for the community. CONCLUSION Metropol Parasol in Seville decreased temperatures beneath the canopy, providing a more comfortable climatic environment. However, other spaces, such as Plaza del Cristo Burgos, decreased temperatures further with humbler and simpler techniques such as vegetation. In addition, the Metropol space was not used by citizens as much as other nearby plazas due to its lack of social facilities. Ecobulevar in Madrid generated three different microclimates at three points spread along a boulevard, all of them more comfortable environmentally than that found in nearby public spaces. The vegetal tree was the most popular, although it did not provide the most comfortable climatic environment. Finally, Plaza Pormetxeta offered a less comfortable climatic environment, influencing negatively the social environment and discouraging people from remaining there. The study suggested that, while a comfortable climatic environment is necessary to generate successful urban public spaces, it needs to be combined with other physical and social aspects of the design. The microclimatic environment of a specific urban space has the potential to attract people to or repel people from it. However, it has to be understood as an integrant of the entire design and needs to be treated together with other requirements of the social environment. In fact, the study indicated that the perception of the climatic environment by citizens is not only determined by its physical properties. A further understanding of the socio-cultural and psychological factors influencing outdoor comfort will help producing successful urban public spaces that foster integration and improve quality of life in cities. ACKNOWLEDGMENTS This research paper was made possible by the guidance of my supervisors, Prof. Simos Yannas and Prof. Paula Cadima, to whom I am very grateful. I would like to show also my gratitude to the Departamento de Educación, Universidades e Investigación del Gobierno Vasco for sponsoring my research. REFERENCES AEMET Agencia Estatal de Meteorología. 2013. Valores climatológicos normales. Sevilla Aeropuerto. Retrieved 05 14, 2013, from http://www.aemet.es/es/serviciosclimaticos/datosclimatologicos/ valoresclimatologicos?l=5783&k=and Ecosistema Urbano. (2004). ECO-BOULEVARD. Retrieved June 27, 2013, from http://ecosistemaurbano.com/portfolio/eco-boulevard/ Llamas, S. 2010. El Correo.com. Retrieved 09 26, 2013, from http://www.elcorreo.com/vizcaya/ v/20101230/margen-izquierda/barakaldo-inaugura-nueva-plaza-20101230.html MTM Arquitectos. (2013). Plaza Pormetxeta. Madrid: MTM Arquitectos. Nikolopoulou, M., & Steemers, K. (2003). Thermal comfort and psychological adaptation as a guide for designing urban spaces. Energy and Buildings, pp. 95-101. Sancho Ávila, J., Riesco Martín, J., Jiménez Alonso, C., Sánchez de Cos Escuin, M., Montero Cadalso, J., & López Bartolomé, M. (2012). Atlas de Radiación Solar en España utilizando datos del SAF de Clima de EUMETSAT. Madrid: Agencia Espanola de Meteorología. Soutullo, S., San Juan, C., Olmedo, R., Enriquez, R., Palero, S., Ferrer, J. A., et al. (2007). Refrigeration in open spaces by means of evaporative systems. 2nd PALENC Conference and 28th AIVC Conference on Building Low Energy Cooling and Advanced Ventilation Technologies in the 21st Century, (pp. 1000-1004). Crete island, Greece. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 82 Thermal Comfort in Naturally Ventilated Classrooms Plabita Baruah, M.Tech. Manoj Kumar Singh, PhD Sadhan Mahapatra, M.Tech. Department of Energy, Tezpur University, Tezpur 784028, Assam, India; Department of Electrical Engineering, GIMT, Tezpur 784501, Assam, India Local Environment Management and Analysis (LEMA), Université de Liège, Chemin des Chevreuils, 1 - 4000 Liège, Belgium;Integrated Research and Action for Development (IRADe), C-80, Shivalik, Malviya nagar, New Delhi 110017, India Department of Energy, Tezpur University, Tezpur, 784028, Assam, India ABSTRACT Thermal comfort study is very important because it correlates occupants comfort in built environment to the functioning of the building and energy consumption. PMV-PPD method works fairly well for conditioned buildings. However, this method does not provide expected results when applied to naturally ventilated buildings. Naturally ventilated buildings are much more dynamic compared to conditioned buildings in terms of thermal environment and occupant’s behaviour in the built environment. In this study, questionnaire based thermal comfort survey has been carried out in naturally ventilated classrooms of Tezpur University during the months of February and May 2013 i.e. at the end of the winter season and the beginning of summer. Thermal sensation and preferences of 228 students are recorded on ASHRAE thermal sensation scale. Various associated parameters like indoor and outdoor air temperature, humidity, clothing and metabolic rate are also measured. The results reveal that the subjects did not feel extreme levels of thermal discomfort during this period. It has been observed that there is a large variation in the clothing pattern (0.83 to 1.52 clo in winter and 0.43 to 0.68 clo in summer) in both the seasons which justify the behavioural, physiological and psychological adaptation of the respondent. It is also found that the other adaptive means like use of fans, closing or openings of windows etc are used quite often. This study concludes that the comfort temperature range varies from 22 to 23.5 °C in winter month and 27.3 to 30.7 °C in summer month. It also concludes that most of the objects recorded cool thermal sensation and preferred a warmer climate in winter and warm thermal sensation and preferred a cooler environment in summer. INTRODUCTION Thermal comfort is defined by ASHRAE as “state of mind that expresses satisfaction with existing environment” (ASHRAE 55, 2013). This definition is subjective because state of mind is widely driven by perception as well as expectations of the person in question. It also can be mentioned that the dissatisfaction can be associated with warm or cool sensation of the habitants in general and it is expressed by PMV (Predicted Mean Vote) and PPD (Predicted Percentage of Dissatisfied) indices (Fanger, 1986). Hence, it is not possible to specify an environment that will satisfy everybody’s thermal comfort. Considering the discreatness of thermal comfort, it can be stated that the same thermal environment may be perceived differently by different people or different people may perceive same thermal comfort at different thermal environments (ASHRAE 55, 2013). However, it may be possible to specify environments to be predicted acceptable, if at least 80% of the occupants feels comfortable . 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 83 (Fanger, 1986). At present, the definition of thermal comfort can be approached in two different ways, each one with its own advantages and limitations: the heat-balance approach and the adaptive approach (Singh et al., 2011). The PMV-PPD model (laboratory based) established by Fanger was based on heat balance model (ISO 7730, 2005). The subjects considered in laboratory experiment were European and American students and experiments were conducted in a controlled climate chamber. This method of evaluating comfort is best suitable for conditioned buildings and deviates largely in case of naturally ventilated buildings. The interactions between occupant and immediate environment in a naturally ventilated building are much more dynamic and the occupant’s behavioural, physiological and psychological adaptations are more wide compared to conditioned buildings (Alfano et al., 2013; Singh et al., 2011; Singh et al., 2015). Singh et al. developed theoretical adaptive thermal comfort models explaining the reason behind the deviation of PMV to that of Actual Mean Vote (AMV) for same set of environmental parameters (Singh et al., 2011). Alfano et al. also reported that Fanger’s thermal comfort model can be made effective in naturally ventilated environments by adding the right expectancy factor with the model (Alfano et al., 2013). Wong and Khoo conducted thermal comfort survey in classrooms which are mechanically ventilated by fans in Singapore (Wong and Khoo, 2003). It is found that the occupants’ acceptable temperature range lies beyond the comfort zone of ASHRAE standard 55. Corgnati et al. carried out surveys in two University classrooms in Turin, Italy applying both objective and subjective surveys confirming that thermal comfort condition and high energy performance are complimentary to each other (Corgnati et al., 2009). Jung et al. investigated subjective responses of thermal comfort of students in a University in Korea (Jung et al., 2011) This study found that the mean Thermal Sensation Vote (TSV) of respondents is almost neutral when the PMV in the classroom moves to neutral and slightly cool, and the TSV is almost ‘+1.5’ when the PMV moves to slightly warm. It is also reported in this study that the acceptability ratio of thermal environment is slightly different from ASHRAE Standard 55-2004. It is found from thermal comfort survey at school that children are more sensitive to changes in their metabolism than adults, and their preferred temperature is lower than that predicted by the standard models (Teli et al., 2012; Yun et al., 2014). Wang et al. study on thermal environment of University classrooms and offices suggested that the neutral temperature varies with the indoor temperature variations (Wang et al., 2014). This study also concludes that the indoor environment has influences on human adaptability, and this determines different neutral temperatures in winter and spring. Mishra and Ramgopal have done a thermal comfort survey inside a naturally ventilated laboratory in the tropical climatic region of India (Mishra and Ramgopal, 2014). This study found that large number of respondent found their indoor thermal environment to be acceptable. The comfort temperatures obtained in the study are used to develope adaptive comfort equation. This equation shows satisfactory results with the predictions from similar equations in comfort standards. Raja et al. studied the use of controls to modify the surrounding environment and how thermal sensation varies with application of these controls (Raja et al., 2001). Pellegrino et al. did a small-scale field survey on occupant’s comfort and related perceptions in two University buildings in Calcutta, India and found that occupants in naturally ventilated schools show acceptability to a wider range of environmental conditions than specified by ASHRAE and ISO standards (Pellegrino et al., 2012). Hwang et al. investigated the adaptive model of thermal comfort for naturally ventilated school buildings in Taiwan and found that the main reason behind discomfort in the classrooms was because most students have to thermally adapt in a naturally ventilated environment when attending school because most of the families in Taiwan have airconditioners in their household (Hwang et al., 2009). Thermal comfort assessments of classrooms are important because extreme discomfort conditions may affect the learning ability of students. Since the classrooms thermal environment requirement is completely different to that of residential and office environment, so it demands a separate thermal environment assessment study to be carried out. In this study, thermal comfort survey through questionnaire has been carried out in naturally ventilated classrooms of Tezpur University during the months of February and May 2013 i.e. during the end of the winter season and the beginning of summer. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 84 The thermal sensation and preference of 228 students are taken into account, in terms of the ASHRAE scale and various parameters like indoor and outdoor air temperature, humidity, clothing and metabolic rate are measured. The subjects chosen for this survey were all university students, both male and female belonging to the age group 20 to 26 years. The thermal sensation votes recorded on the ASHRAE 7 point scale during comfort survey is considered as actual mean vote (AMV). These AMV values along with other set of indoor environmental conditions are used to calculate PMV values using ASHRAE 55 and ISO 7730 standard. METHODOLOGY Thermal sensation is primarily related to the thermal balance of the body. This balance is influenced by the physical activity and clothing pattern of the habitants. Along with these two variables, the environmental parameters like air temperature, mean radiant temperature, air velocity and relative humidity also has an effect on thermal sensation. Thermal sensation of the occupants can be predicted, if all the above parameters are known. Hence, it is important to find out the response of the occupants about the indoor thermal environment. It has to be kept in mind that judgment of the occupant depends on his perception and expectation about thermal comfort. During field study, questionnaire is administered to subjects and simultaneously other micro-climatic parameters are measured. The subjects were asked to express their level of thermal sensation characterized in ASHRAE thermal sensation scale as shown in Table 1. It is also important to understand that the habitants are always active to the changes in existing thermal environment and always try to adapt themselves to changing environmental conditions to feel thermally comfortable. In naturally ventilated buildings, occupant’s preference and expectations about comfortable thermal environment keep on changing with the change in outdoor conditions or seasons (Singh et al., 2010, Singh et al., 2011). During the comfort survey, respondent were advised to sit idle for 20 minutes, and the activity of 1.2 met is considered for the analysis. Clothing insulation is measured in terms of ‘clo’ unit, and is used to estimate the insulating properties of clothing by using the tables provided in ISO 7730 standard (ISO 7730, 2005). The clothing value is determined based on an occupant’s garment checklist in the questionnaire. Table 2 represents the details of the thermal comfort survey. Table 1 ASHRAE Thermal sensation scale Value Sensation +3 Hot +2 Warm +1 Slightly warm 0 Neutral -1 Slightly cool -2 Cool -3 Cold Table 2 Thermal comfort survey details Climatic zone Warm and humid Number of subjects 228 (114 in winter and 114 in summer) Age group of the subjects 20 – 26 years Range of clothing (during summer) 0.4 – 0.7 clo Range of clothing (during winter) 0.8 - 1.5 clo Number of male respondent 46 (winter); 46 (summer) Number of female respondent 68 (winter); 68 (summer) The thermal sensation votes recorded during comfort survey is considered as actual mean vote (AMV). Environmental, activity and clothing level data collected during comfort survey are used to calculate the PMV-PPD by using ISO 7730 standard calculation procedure. The PMV value is calculated by using equation 1 provided in ISO standard (ISO 7730, 2005). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 85 PMV  [0.303  e 0.036 M  0.028]  L (1) Where L is the thermal load difference between the internal heat production and the heat loss to the actual environment and M is the metabolic rate. The PPD is calculated by using the equation 2. 4 2 PPD  100  95  e  (0.03353 PMV  0.2179 PMV ) (2) The PMV and PPD results are cross checked using the CBE/Berkley PMV-PPD calculation tool (CBE/Berkley, 2013). PMV values calculated by these methods over estimate the thermal condition in summer season and under estimate the thermal condition in winter season. This may be due to the model fails to consider the adaptive opportunities, preferences and expectations of the habitants in naturally ventilated buildings. It has been tried to use adaptive thermal comfort model which is combination of physics of the body’s heat balance plus local climatic behaviour, preference and expectations, past thermal experiences, social and cultural practices to overcome this discrepancy. Hence, it is important to calculate the adaptive coefficient ( = factor for adaptation) that needs to be added to PMV to make the result close to AMV. Singh et al. 2011 proposed the following relation to calculate the cPMV for naturally ventilated buildings of North-East India. PMV (3) cPMV  1    PMV The adaptive coefficient is positive means the indoor temperature is greater than comfort temperature. This case is generally common in summer for naturally ventilated buildings. It also can be concluded that at this situation, the value of cPMV is lower corresponding to the PMV or cPMV is giving cooler feeling than PMV, i.e. cPMV votes are towards comfort to that of same PMV. Similarly, the adaptive coefficient is negative means; the indoor temperature is lower than comfort temperature.This condition occurred in winter season for naturally ventilated buildings. In this situation, it is observed that cPMV is giving warmer feeling than corresponding PMV. RESULTS AND DISCUSSION The thermal comfort survey was carried out among the students in naturally ventilated classrooms of six departments of Tezpur University during the months of February and May 2013 i.e. during the end of the winter season and the beginning of summer. The thermal sensation and preference of the students are taken into account, in terms of the ASHRAE 7 point scale. The indoor temperature was in the range 22°C to 23.5°C during Februray and 27.3°C to 30.7°C during May. The indoor humidity during the winter season ranged from 56% to 63% and it was higher in the summer season ranging from 77% to 84%. The outdoor air temperature was found to be slightly higher than the indoor air temperature. It was observed that factors like building orientation and shading affected the indoor temperature. Clothing level adjustment is one of the important and most effective adaptation processes to maintain the comfort at different temperatures. Figure 1 represents the relationship between outdoor temperature and clothing pattern. It has been found from the comfort survey that the clothing values are largely scattered and varies from 0.43 to 0.68 clo in summer and 0.83 to 1.52 clo in winter. The outdoor temperature variation in winter is from 21.9 to 24 °C and 28.5 to 32°C in summer. It is observed from the Figure 1, that there are two distinct clothing profiles in these two seasons. In summer, the clothing profile decreases, as the outdoor temperature increases and vice versa in the winter season. It can be concluded that there is a strong relation between the clothing pattern and outdoor temperature. The dependence of clothing pattern with the outdoor temperature has been examined through linear and polynomial regression and presented in equation 4 and 5 (where T0 is the outdoor temperature). The coefficient of regression (CC) is low as this analysis is based on only two seasons of the year. clo  0.091T0  3.2941 CC: 0.7449 (4) clo  0.0081T02  0.5253T0  9.0534 CC: 0.7721 (5) It is observed that when the clothing level is less than 0.8, the thermal sensation lies from 0 to 2. This shows that during summer when the subject is feeling warm they tend to lessen their clothing 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 86 insulation to attain comfort. During winter when the temperature is low and the respondent feel cool or cold thermal sensations, the clothing level is high. The respondents wear more clothes to keep themselves warm but in some cases it is observed even though when the temperature is low the students felt warm and the less clothing level is observed. This is justifying the physiological adaptation of the respondent which is resulted from long-term exposure to certain thermal environment which made the respondents habituated. It is found that the CC value is less in the present study than what observed in case of naturally ventilated residential buildings in same climatic zone (Singh et al., 2011). This happens because of restrictions in clothing pattern in University classrooms and sitting positions (sitting near or away from window with varying temperature). However, it can be observed from Figure 1 that the polynomial regression curve bents inwards suggesting the adjustments a subject undergoes and make themselves adapted to reduce discomfort created by high clothing insulation level even at relatively high temperature. This also put forth the argument that in naturally ventilated buildings, the relation between clothing level and outdoor temperature is not linear. 1.6 Clothing pattern (Clo) 1.4 1.2 y = ‐0.091x + 3.2941 R² = 0.7449 1.0 0.8 y = 0.0081x2 ‐ 0.5253x + 9.0534 R² = 0.7721 0.6 0.4 0.2 0.0 20 22 24 26 28 30 Outdoor temperature (0C) 32 34 Figure 1 Relationship between outdoor temperature and clothing pattern 2.5 2.0 Thermal sensation 1.5 1.0 0.5 0.0 ‐0.5 ‐1.0 ‐1.5 ‐2.0 ‐2.5 20 22 24 26 28 30 Outdoor temperature (0C) 32 34 Figure 2 Relationship between outdoor temperature and thermal sensation Figure 2 represents the thermal sensation profile against outside temperature. It can be observed from the Figure 2 that there are two distinct profiles for two different seasons of the year. In case of winter month, thermal sensation varies from -2 to +2 with the variations of outdoor temperature from 21.9 to 24 °C. Similarly, in summer month, thermal sensation varies from 0 to 2 with the temperature variation from 28.5 to 32 0C. Perception and expectation about comfort differ from person to person (behavioural, physiological and psychological adaptation). Hence, it can conclude that the same temperature perceived different thermal sensations by the occupants or different occupants perceived same thermal sensation at different temperatures. This also justifies from the clothing level variation in 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 87 winter and summer months. It is also found from the comfort survey that the AMV is as high as +2 for few respondents. This may be due to the past experiences of cooler thermal environment of these respondents.In this study, inside classrooms the comfort temperature range is found to be 22-23.5 °C in winter and 27.3 -30.7 °C in summer (PMV lies between -1 to 1, or more than 80% of the people satisfied in this temperature range). This comfort temperature range is closely agreed with range of comfort temperature in naturally ventilated buildings reported in different studies (CBE/Berkley, 2013; Hwang et al., 2009; Pellegrino et al., 2012; Raja et al., 2001; Singh et al., 2011). Predicted Mean Vote (PMV) predicts the mean thermal sensation vote on a standard scale for a large group of people. Predicted Percentage of Dissatisfied (PPD) index provides the number of people dissatisfied at a particular environmental condition. The PMV and PPD values are calculated using the calculation procedure provided at ISO 7730 standard and CBE/Berkley PMV-PPD tool. Figure 3 presents the PMV/PPD values obtained through ISO 7730 (equation 1 and 2) and also by using CBE, Berkley tool. It is observed from the Figure 3, that the PMV-PPD profile complies with the standard PMV-PPD graphs. In this figure, only one side of the profile is observed, as our thermal comfort survey is limited only to two seasons. The experimental results obtained through field measurement (calculated by using equation 1 and 2) are validated by using CBE, Berkley tool. In an attempt to incorporate adaptive comfort model, the PPD upper limit is increased to 20% i.e. -1 to +1 sensation which is completely in agreement with our results at +1 thermal sensation, the PPD value is near to 20 %. It shows that there is a slight variation in PMV/PPD values obtained by these two methods. 100 CBE/Berkley Method 90 ISO Method 80 70 PPD 60 50 40 30 20 10 0 ‐3.0 ‐2.5 ‐2.0 ‐1.5 ‐1.0 ‐0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 PMV Figure 3 Relationships between PMV and PPD In case of naturally ventilated buildings, PMV deviates widely from AMV values due to inherent limitation in assessing thermal comfort. To minimize this deviation in PMV values, Singh et al. proposed cPMV relation, which accommodates behavioural, physiological and psychological adaptation to calculate the adaptive coefficient (Singh et al., 2011). Equation 3 is used to calculate cPMV values. Figure 4 represents the relation between adaptive mean vote (AMV) and corrected mean vote (cPMV) with respect to PMV. The plot concludes that cPMV provides better indoor thermal sensation as this includes the adaptation of the occupants of a naturally ventilated building. The closer the cPMV to AMV mean it is assessing more correctly the real indoor thermal environment from occupant’s perspective. The differences of adaptive coefficients in different seasons present the extent of adaptation of the subject. The adaptive opportunities which are available to the occupants of a naturally ventilated building actually shift the neutral temperature as well as the range of comfort temperature. It is observed from the Figure 4 that the cPMV values (-1.32 to 1.45) come closer to AMV values (-2 to +2) whereas PMV values are distributed between -1 to 3. This adaptation processes through different adaptive opportunities help the respondent to achieve required thermal comfort at a relatively lower indoor temperature in winter or higher temperature in summer month.The positive adaptive coefficient means the corrected mean vote is giving cooler feeling than the predicted mean vote. This kind of situation would occur in warm months, when the indoor temperature is higher than the comfort temperature. The reverse situation would also occur for winter months. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 88 The thermal comfort survey has been carried out to the University students of both male and female in the age group 20 to 26. It is important to note that the age of respondent does not have any signification variations on the thermal sensation. However, as the comfort survey has been done only for two seasons, it will not be wise to make any generalized comment on this. The outdoor temperature variations during the winter days of the survey were 21.9 to 24 °C. Most of the subjects recorded cool thermal sensation and preferred a warmer climate. However, 25% voted in the neutral range and it is observed that a few subjects felt warm thermal sensation in this temperature range. Thermal comfort survey during the summer, the outdoor temperature variations recorded was 28.5 to 32 °C. Most of the subjects voted +1 or +2 i.e. slightly warm and warm thermal sensation and preferred a cooler environment. It is also observed that there is a drop of clo value as the temperature increase in summer months in comparison to winter months. Change in clothing pattern is a significant adaptive measure adopted by the students to increase their level of comfort. The students increase or decrease their layers of clothing in winter and summer respectively to adjust with the environment. 2.5 2.0 1.5 AMV / cPMV 1.0 0.5 0.0 ‐0.5 ‐1.0 ‐1.5 AMV ‐2.0 cPMV ‐2.5 ‐1.0 ‐0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 PMV Figure 4 Relationship between PMV, AMV and cPMV The thermal comfort temperature recommended by ASHRAE and obtained from the survey is presented in Table 3. Ranges of comfort temperature are derived from the comfort survey and corresponding measurements that were carried out during survey. The range of temperature represents the temperature corresponding to thermal sensation -1 to +1 (according to adaptive thermal comfort model, occupants can make themselves comfortable in this range by utilizing adaptive opportunities in naturally ventilated buildings). It is observed from Table 3 that the acceptable limit of comfort temperature recommended by ASHRAE is closely similar to the survey results for winter months. However, the summer months acceptable limit does not agree with the ASHRAE recommended value. This can be due to limited respondent and also for only two seasons have been covered in the thermal comfort survey. It is highly desirable that the thermal comfort survey to be done throughout the year with more respondent to get the generalized comfort temperature range. Table 3 Comfort temperature recommended by ASHRAE and obtained from survey Conditions Summer (light clothing) Winter (warm clothing) ASHRAE recommended acceptable operating temperature (°C) Humidity range if 30% : 24.5 – 28.0 Humidity range if 60% : 23.0 - 25.5 Humidity range if 30%: 20.5 - 25.5 Humidity range if 30%: 20.0– 24.0 Comfort temperature obtained from survey (°C) Humidity (%) : 77 - 82.2: 27.3-30.7°C Humidity (%) : 55 - 63: 22-23.5 °C CONCLUSION This study is based on the responses of the questionnaire based thermal comfort survey of 228 students of Tezpur University. The survey was carried out in six naturally ventilated classrooms at various departments located inside the University. The survey has been carried out during two different 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 89 seasons of the year. It has been found that adaptation of the respondents is clearly visible in the clothing pattern which has a strong dependence on outdoor temperature. The comfortale thermal sensation (acceptable) has been observed for the temperature range from 22 to 23.5 °C in winter and 27.3 to 30.7 °C in summer month. Clothing level varies from 0.83 to 1.52 clo in winter and 0.43 to 0.68 clo in summer month. Most of the subjects recorded cool thermal sensation and preferred a warmer climate in winter. Most of the subjects voted +1 or +2 i.e. slightly warm and warm thermal sensation and preferred a cooler environment in summer. It is observed from the comfort survey that clothing pattern is a significant adaptive measure adopted by the students to increase their level of comfort. It can be concluded that the deviation in AMV to that of corresponding PMV is due to various adaptation processes used by the students to make themselves comfortable in the indoor environment. Different values of adaptive coefficient provide a better understanding about the impact of various adaptive factors on an individual to attain thermal comfort. It is felt that thermal comfort survey should be done throughout the year with more respondent to get the generalized comfort temperature range. REFERENCES Alfano, F. R. d’A., Ianniello, E., & Palella, B. I. 2013. PMV-PPD and acceptability in naturally ventilated schools. Building and Environment, 67, 129-137. ASHRAE 55. 2013. Thermal environmental conditions for human occupancy, American Society of Heating, Refrigerating and Air-conditioning Engineers Inc. CBE/Berkley Thermal Comfort Tool for ASHRAE-55, http://cbe.berkeley.edu/comforttool/, Accessed on 1005-2013 Corgnati, S. P., Ansaldi, R., & Fillipi, M. 2009. Thermal comfort in Italian classrooms under free running conditions during mid seasons: Assessments through objective and subjective approaches. Building and Environment, 44(4), 785-792. Fanger, P. O. 1986. Thermal environment - human requirements. The Environmentalist, 6, 275-278. Hwang, R. L., Lin, T. P., & Chen, C. P. 2009. Investigating the adaptive model of thermal comfort for naturally ventilated school building in Taiwan. International Journal of Biometeorology, 53(2), 189-200. ISO 7730, 2005. Moderate thermal environment - Determination of PMV and PPD indices and specifications of the conditions for thermal comfort. International Organization for Standardization, Geneva, Switzerland. Jung, G. J., Song, S. K., Ahn, Y. C., Oh, G. S., & Im, Y. B. 2011. Experimental research on thermal comfort in the university classroom of regular semesters in Korea. Journal of Mechanical Science and Technology, 25 (2), 503-512. Mishra, A. K., & Ramgopal, M. 2014. Thermal comfort in undergraduate laboratories – A field study in Kharagpur, India. Building and Environment, 71, 223-232. Pellegrino, M., Simonetti, M., & Fournier, L. 2012. A field survey in Calcutta: Architectural issues, thermal comfort and adaptive mechanisms in hot humid climates. The changing context of comfort in an unpredictable world Cumberland Lodge, Windsor, UK, 12-15 April 2012. London, Network for comfort and energy use in Buildings 2012. http://nceub.org.uk. Raja, I. A., Nicol, J. F., McCartney, K. J., & Humphreys, M. A. 2001. Thermal comfort: use of controls in naturally ventilated buildings. Energy and Buildings, 33(3), 235-244. Singh, M. K., Mahapatra, S., & Atreya, S.K. 2010. Thermal performance study and evaluation of comfort temperatures in vernacular buildings of North-East India. Building and Environment, 45(2): 320 - 329. Singh, M. K., Mahapatra, S., & Atreya, S. K. 2011. Adaptive thermal comfort model for different climatic zones of North-East India. Applied Energy, 88(7), 2420 - 2428. Singh, M. K., Mahapatra, S., & Teller J. 2015. Development of thermal comfort models for various climatic zones of North-East India, Sustainable Cities and Society, 14, 133-145. Teli, D., Jentsch, M. F., & James, P. A. B. 2012. Naturally ventilated classrooms: An assessment of existing comfort models for predicting the thermal sensation and preference of primary school children. Energy and Buildings, 53, 166- 182. Wang, Z., Li, A., Ren, J., & He, Y. 2014. Thermal adaptation and thermal environment in universityclassrooms and offices in Harbin. Energy and Buildings, 77, 192-196. Wong, N. H., & Khoo, S. S. 2003. Thermal comfort in classrooms in the tropics. Energy and Buildings, 35(4), 337-351. Yun, H., Nam, I., Kim, J., Yang, J., Lee, K., & Sohn, J. 2014. A field study of thermal comfort for kindergarten children in Korea: An assessment of existing models and preferences of children. Building and Environment, 75, 182-189. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 90 Comparison of Strategies improving Local Energy Self-sufficiency at Neighborhood Scale. Case study in Yverdon-les-Bains (Switzerland) Sophie Lufkin, PhD Emmanuel Rey, Prof. [Ecole polytechnique Fédérale de Lausanne] [Ecole polytechnique Fédérale de Lausanne] sophie.lufkin@epfl.ch ABSTRACT Within a context of growing efforts to develop sustainability strategies, one of the main challenges is promoting value creation while using fewer resources. In this perspective, how can we design attractive urban neighborhoods generating endogenous economic activity and fostering socio-cultural dynamics, while moving towards local energy self-sufficiency? Answering that question requires major changes in the way we consider energy in the construction sector, by thinking beyond the scale of a single building and by including a greater number of design parameters. Filling this gap in current research, the Symbiotic Districts project examines dimensions influencing energy self-sufficiency at neighborhood scale by integrating parameters related to buildings, infrastructure, mobility, food, goods and services. The present paper analyzes the results of a case study on an urban sector in the city of Yverdon-lesBains (Switzerland). Taking lifestyles as a starting point, the project explores three scenarios (technological, behavioral and symbiotic) for the future development of this neighborhood for 2035. The scenarios test different design strategies related to industrial symbioses, production, storage, transportation or urban agriculture. In order to calculate an estimated global balance, an energy flow analysis allows the assessment and comparison of the energy performance of each scenario. In parallel, an urban form adapted to the proposed vision evaluates how architectural and urban design is likely to foster the necessary behavior changes towards the expected energy turnaround. 1 INTRODUCTION Within a context of a growing efforts to create sustainable development strategies, a wide array of research programs are being conducted on energy-related issues in the built environment. And for good reason: over 40% of worldwide energy consumption can be attributed to the construction sector (Wallbaum, 2012). In Switzerland, a landscape dense with urban development, total energy expenditures associated with buildings account for no less than half of total energy consumption (Zimmermann, Althaus, & Haas, 2005). Ambitious objectives to reduce renewable and non-renewable energy consumption are now being set by several European countries, following the example of the 2000-Watt Society concept developed in Switzerland or the political vision of phasing out nuclear energy over the medium-term (Jochem, 2004; Previdoli, 2012). Author A is a research fellow at Laboratory of Architecture and Sustainable Technologies (LAST), Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Suisse. Author B is a professor at LAST (EPFL). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 91 At the same time, the projected end of abundant low-cost fossil fuels, geopolitical tensions around the issue of natural resources and the vulnerability of electrical power grids are all factors that encourage finding more secure energy supply strategies, particularly by making use of local resources. With this objective in mind, working towards local energy self-sufficiency – and specifically a balance between the energy consumption of a territory and its ability to meet its own demand through sustainable production – will allow us to minimize environmental impacts while at the same time generating endogenous economic activity and promoting a social and cultural dynamic in which the users can become involved. This type of approach requires a significant reduction in demand (moderation), the widespread use of renewable energy (local production), and an effort to achieve complementarities between operation (industrial symbiosis) and on-site energy storage (Grospart, 2009). Taking these matters into account requires major changes in the way we consider energy in the construction sector, firstly by clearly transcending the scale of the single building in order to address urban reality at neighborhood scale (Rey, Lufkin, Renaud, & Perret, 2013). This intermediate scale reveals some surprising information. On one hand, it is broad enough to address themes that transcend the single building, opening up possibilities for studying interactions between these entities. On the other hand, unlike city scale, on which most of the current research on local energy self-sufficiency focuses, it is restricted enough to design, test and examine concrete and operational initiatives (Rey, 2011). Therefore, this approach allows taking into account multi-functionality, considering certain industrial activities and urban or suburban agriculture activities near residential areas, while keeping them to the appropriate scale for the most strategic approaches to urban development (e.g. master plan). Secondly, a greater number of design parameters needs to be included in the reflection, moving well beyond basic issues related to the buildings’ heat and electricity consumption. The observation of the traditional neighborhood highlights the limitations of its urban flows operational scheme. This urban metabolism, which can be described as linear, requires large amounts of external inputs, largely stemming from non-renewable sources, and generates a high level of non-valorized rejects (waste, greenhouse gas, dissemination into the environment or liquid effluents). In addition, interactions between the functions are very limited. This system increases the neighborhood’s ecological footprint and could potentially challenge its very existence over the long term. To work towards greater sustainability, new modalities are therefore needed to increase both the self-sufficiency and efficiency of urban environments. In reaction, the Symbiotic District project was conceived in order to promote a "syntropic" urban system, i.e. a mature ecosystem capable of fostering cities’ economic and sociocultural development, while making the best use of imported resources and limiting waste production thanks to a circular metabolism. Concretely, such an approach embraces industrial ecology principles (Erkman, 1998) and aims at transposing them to the built environment in general, and the Swiss urban context in particular. The Symbiotic District project simultaneously examines scientific, technical, urban development and architectural aspects of local energy selfsufficiency at neighbourhood scale by integrating issues related to buildings, infrastructure, mobility, goods, services and food (Lufkin, Rey, & Erkman, 2014). The approach relies on three complementary optimization strategies: increasing the city’s intrinsic efficiency, valorizing renewable energy sources and implementing urban symbioses (Lufkin, Rey, & Erkman, 2013). The research also aims at identifying the most relevant levers to reduce energy consumption lifestyles, technology or urban form - and studying interactions between these lines of action. Indeed, in spite of increased consciousness about energy issues, private or public stakeholders find it difficult to commit to a responsible behavior due to the absence of sufficiently accurate information. Establishing a reliable basis to address energy issues in future sustainable urban neighborhoods (in the horizon 2035), the approach provides a systematic exploration of the links, still to be created, between strictly quantitative aspects related to energy self-sufficiency (stemming from industrial ecology) and qualitative and operational aspects related to their implementation into urban and architectural projects (Erkman, 1998). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 92 2 STATE OF THE ART The idea of considering city as an ecosystem is not new. It was introduced in the sixties, in particular by biologists, who started drawing their inspiration from the theory of ecosystems in order to deal with the complexity of the environment and to understand it in a more systematic way. Deriving from these reflections, urban metabolism provides sound methodological and practical tools to analyze urban resources and flows (Baccini, 1996; Newman, 1999). Applying this approach to cities, researchers started highlighting a number of dysfunctions: high dependency towards fossil energy, low efficiency due to linear processes, inefficiency of sectoral policies and "end of pipe" solutions, etc. (Barles, 2008; Dobbelsteen, Keeffee, Tillie, & Roggema, 2012). Urban metabolism is a very efficient approach to assess a region's or a city's level of sustainability and to identify resources and waste potentially reusable at regional scale (Codoban & Kennedy, 2008). However, territorial or urban scale remains too large to transpose the results from such a model to strategic operational processes. To date, attempts to apply urban metabolism at neighborhood scale are few and very recent. Indeed, the parameters usually considered in research and practice rarely go beyond the building's energy consumption (heating, domestic hot water, electricity, grey energy). A limited number of experiences try to include aspects related to the inhabitants’ transportation and food in a broad perspective, addressing energy supply as both an energy consideration – power supply accounts for a significant portion of the total energy balance per inhabitant (Rey, 2006) – and from an urban development standpoint – urban agriculture, for instance, is becoming an increasingly popular consideration with regard to achieving urban sustainability (Gorgolewski, Komisar, & Nasr, 2011; Jourdan & Mirenowicz, 2011). These examples include the REAP methodology in Rotterdam (Tillie et al., 2009), the Amsterdam Guide to Energetic Urban Planning (Tillie, Kürschner, Mantel, & Hackvoort, 2011), the Urban Harvest Concept in Kerkade West (Agudelo-Vera, Leduc, Mels, & Rijnaarts, 2012) and the New Stepped Strategy (Dobbelsteen, 2008). These references speak to the benefits of combining different functions within the same neighborhood or even within the same building, thus revisiting a "fine-grained" functional mix. All these pilot projects are still at experimentation or planning phases, none of them has yet been realized. Today, the main challenge is their integration into a consistent and realistic reflection in order to positively influence local energy and resource self-sufficiency at neighborhood scale. 3 METHODOLOGY In reaction, the case study presented in this paper focuses on the Gare-Lac sector in Yverdon-lesBains (Switzerland). The site is currently a large urban wasteland of about 23 hectares, strategically situated between the railway tracks and Lake Neuchâtel, in very close proximity to the station and the city center. The local master plan (PDL) (Bauart Architectes et Urbanistes SA, 2010), currently under validation, was used as a basis for the present case study. The research is conducted in four stages: 1 - Energy cadaster Making an inventory of available local resources, the first stage establishes a regional energy cadaster. Renewable energy production installations and supply projects situated within the perimeter of the urban region of Yverdon are listed. The resulting local energy mapping takes into consideration resources such as biomass, sun, wind, waste heat, geothermal potential, lake, etc. 2 - Scenarios Based on this cadaster and on the recent PDL, three radical scenarios (technological, behavioral and symbiotic) are developed in a 2035 perspective. Enriched by prospective reflections on the evolution of European lifestyles (IDDRI, 2012), the scenarios embody a specific positioning to meet sustainability concerns. Set by the PDL, the human density (number of inhabitants and jobs per hectare) is the same for all scenarios, i.e. 3’810 inhabitants and 1’260 jobs. The built density (gross floor area, GFA), however, varies from one scenario to the other, mainly because of the variation of average per capita living space and the type of activity. 3 - Energy flow analyses For each scenario, several hypotheses are then formulated. They are structured into five domains, which contain a variable number of categories and sub-categories: 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 93 Table 1. Summary List of the five domains and their respective categories Buildings   Mobility   Infrastructure   Food   Goods  and  services   Construction Domestic hot water Heating/Ventilation Lights and devices Car Airplane Train Other Neighborhood / municipal facilities External installations Other Agriculture Transformation Packaging Distribution Clothes Furniture Restaurant Hotel / Leisure Three indicators are calculated in order to analyze the energy consumption of each scenario: Total Primary Energy (TPE), Non-renewable Primary Energy (NRPE) and Global Warming Potential (GWP). The first step is the evaluation of the current situation, which serves as reference point. Users behavior and habits are based on the current Swiss average. Each (sub)-category value is then adapted according to the scenario’s specific hypotheses. 4 - Urban form In parallel, an urban form is proposed for each scenario (Fig. 1-3). It is developed according to the lifestyle assumptions on which the scenario is based and provides a visualization of the future neighborhood. Indeed, each lifestyle reflects distinct uses, which correspond to specific needs in terms of spatial, functional and sensitive qualities (Thomas, 2011). This transposition of conceptual assumptions into an urban form also assesses the extent to which urban and architectural quality is likely to promote behavioral changes necessary for a transition towards a more sustainable society. 4 RESULTS 4.1 Local resources The identified local resources are attributed to the Gare-Lac neighborhood according to a principle of territorial representativeness. For instance, if a resource is shared by the whole urban region (or the city) of Yverdon-les-Bains, only 7% (respectively 14%) of this potential is allocated to the site. This percentage corresponds to the ratio between the site's population and that of the considered territory. Waste heat The public baths of Yverdon, whose water is heated to 32°C, and the water treatment plant are the major installations likely to contribute to local symbioses through a process of waste heat recovery. The combination of these two sources could potentially produce 3.5 thermal MW. Due to the proximity of these installations to the site, it was decided to allocate 30% of this potential to the new neighborhood. Geothermal potential According to information provided by the Commune of Yverdon-les-Bains, a geothermal cogeneration project should be completed in 2017. It represents a potential of 5 electric MW and 60 thermal MW, of which 14% are allotted to the new neighborhood. Biomass It was estimated that organic waste produced by the inhabitants of the neighborhood and the animals living in the vicinity could produce as much as 114 electric kW thanks to heat-power coupling generated by agriculture biogas. In addition, a wood-energy plant is being studied. The latter could potentially produce 17 thermal MW, of which 14% would be allotted to the Gare-Lac sector. Solar potential The SEY have planned a photovoltaic supply growth of 0.5 MW per year until 2035, i.e. a total growth of 13 MW including the existing capacity. In a similar way as biomass and geothermal potential, 14% of the stock is attributed to the future neighborhood. Wind potential Two major wind power projects are being considered in the Northern Vaud region, totaling 39.5 electrical MW. Considering this scale, only 7 % of this energy will be distributed to the neighborhood, i.e. 2.8 electric MW. Lake Neuchâtel In spite of the important potential of the lake, it has not been considered as a thermal resource for the sake of realism. For environmental reasons, this option is not desired by cantonal and communal authorities of the urban region of Yverdon-les-Bains. The following table provides a summary of the available local resources, which are then allocated to the future neighborhood according to the hypotheses formulated by each scenario: 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 94 Table 2. Summary List of the local resources available Resource   Waste  heat   Geothermal   Biomass   Sun   Wind   Thermal [MW] 3.5 60 17 - - Electric [MW] - 5 0.1 13 39.5 30 % 14 % 14 % 14 % 7% Percentage attributed to site 4.2 Performance of the scenarios Technologic scenario Using the most advanced technologies to reduce energy consumption, this scenario doesn’t imply any modification of the user’s behavior. Overall, the environmental impacts decrease thanks to the improvement of the devices’ efficiency, but this effect is counterbalanced by the general higher consumption. Main characteristics of this scenario are: Minergie A standard for all buildings, heavy construction mode, integration of renewable energies, increased average per capita living space (60 m2 per person instead of the current 50 m2), stabilization of travelled kilometers, hydrogen-powered cars (for 50% of the users), imported, transformed and conditioned food, etc. Table 3. Synthetic chart of the performance of the technologic scenario Domain   TPE  [W/pers]   NRPE  [W/pers]   GWP  [kg  CO2eq/year]   Buildings 642 428 914 Mobility 2’299 1’199 2’238 Infrastructure 551 491 603 Food 720 663 1'638 Goods and services 750 690 1'012 4'962 3'471 6'405 Total Figure 1 Visualization of the technologic scenario. Roofs’ volumetries have been optimized in order to integrate photovoltaic panels (in green). • Inhabitants: 3'810; • Jobs: 1’260 2 • GFA: 260’100 m Behavioral scenario This scenario takes the opposite view and relies mainly on a change in users’ behavior towards more frugal consumption, sobriety, simplicity, reduced consumerism and decelerating lifestyles. Thus, the driving force of the energy transition is mostly the demand reduction thanks to the modification of current social practices (energy supply by biogas and wood-energy plant, light wooden constructions, pooling of facilities, diminution of the average per capita living space to 40m2 per person, increased soft mobility, increased car occupancy through car sharing, urban farming, natural treatment for public spaces, vegetarian, local and organic diets, autonomous production of goods and services, etc.) Table 4. Synthetic chart of the performance of the behavioral scenario Domain   TPE  [W/pers]   NRPE  [W/pers]   GWP  [kg  CO2eq/year]   Buildings 651 449 400 Mobility 1'483 1'353 2'515 Infrastructure 508 479 578 Food 406 373 923 Goods and services Total 638 587 860 3'685 3'240 5'275 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 95 Figure 2 Visualization of the behavioral scenario. Large garden dedicated to urban farming are also special meeting places for the inhabitants (in green). • Inhabitants: 3'810; • Jobs: 1’260; 2 • GFA: 209’100 m Symbiotic scenario This scenario promotes urban and industrial symbioses opportunities to reduce the environmental impact of the neighborhood. Energy exchanges are implemented at all scales (building, group of buildings, neighborhood and between the neighborhood and its surrounding perimeter). The symbiotic scenario implies changes in behavior, but not as radical as the ones required by the behavioral scenario: users take responsibility toward sustainability and foster network and partnership dynamics. The main features of this scenario include: energy mainly supplied by heat recovered from the public baths and water treatment plant (3,4 thermal [MW] and 0.1 electrical [MW]), Minergie P standard for new constructions, recycled materials, heat recovery on waste domestic water and ventilation, stabilization of the average per capita living space (50m2 per person), significant functional diversity (crafts and non-polluting industries), smaller and lighter vehicles, biodiesel for cars, development of the public transport network, diminution of air travels, healthy and responsible diet (reduced meat consumption, local and seasonal products), recyclable or repairable goods). Table 5. Synthetic chart of the performance of the symbiotic scenario Domain   TPE  [W/pers]   NRPE  [W/pers]   GWP  [kg  CO2eq/year]   Buildings 778 690 731 Mobility 1’307 1'187 2'015 Infrastructure 543 513 611 Food 653 600 1'485 Goods and services 675 621 911 4'005 3'654 5'775 Total Figure 3 Visualization of the symbiotic scenario. Large urban blocks encourage short-distance exchanges at building and neighborhood scale. • Inhabitants: 3'810; • Jobs: 1’260; 2 • GFA: 262’320 m 4.3 Discussion of the results First of all, the energy consumption of all three scenarios is lower than that of the current situation (Figure 4). Besides, the ranking is rather immediate: for all indicators, the behavioral scenario appears as the most performing, and the technologic scenario is the most energy intensive (except for NRPE). However, these results need to be contrasted with the fact that only energy-related aspects were integrated into the assessment. Other criteria, in relation to social (acceptance), economic (costs of the implemented technologies) or environmental impacts, would be necessary in order to establish a more complete and global evaluation of the scenarios. For instance, the radical and very restrictive vision 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 96 embodied by the behavioral scenario is not fully realistic because of all the constraints imposed on the inhabitants’ individual freedom. This remark highlights the relevance of holistic approaches, which form the core concept of sustainable development. Nevertheless, the three indicators bring to light certain interesting phenomena. Concerning buildings, for instance, results are counter-intuitive: the buildings with the lowest energy consumption – those of the behavioral scenario – comply with the least strict construction standard. The explanation is provided by two factors: the average per capita living space and the construction modes. In the behavioral scenario, the reduction of living space to 40 m2 per person (compared to the current 50 m2) leads to a diminution of approx. 30% of the necessary GFA in the neighborhood, which significantly influences construction and operation energy. In addition, light wooden constructions have a positive impact on the buildings’ grey energy (as opposed to the heavy constructions of the technologic scenario). Mobility also plays an important role in the energy balance of the scenarios. In the technologic scenario, mobility represents approximately half of the TPE. Its impact decreases a lot for the NRPE, thanks to the use of hydrogen-powered cars. In the behavioral scenario, the use of conventional cars penalizes the balance in spite of the absolute reduction of kilometers travelled. The symbiotic scenario offers the most convincing solution by encouraging simultaneously the use of collective transports and of biodiesel and electric cars (produced from renewable sources). Regarding food, the most significant levers are related to reduced meat consumption. However, the impact of this change of eating behavior on the global balance remains low. From a strictly energetic point of view, this effort is of minimal benefit while a transition towards a vegetarian diet implies a high level of commitment of the inhabitants. 5 CONCLUSIONS AND FUTURE PERSPECTIVES In order to put this reflection in perspective with long-term sustainability objectives pursued by several countries, Switzerland in particular, the results were confronted to the intermediary goals of the 2’000 Watts society concept for 2035 (Jochem, 2004). Figure 4 shows that none of the radical scenarios meets the targets for all three indicators. TPE values of both the behavioral and symbiotic scenarios are below the threshold of 4'400 [W/pers], while the technologic vision exceeds the limit. For NRPE, results of the three scenarios are all slightly above the objective of 3'300 [W/pers]. However, considering the uncertainties affecting some of the data, it can be considered that these orders of magnitude are roughly equivalent. CO2 emissions of the three scenarios, by contrast, clearly exceed the intermediary target of 3,2 [tons CO2eq/year]. This can be explained in part by the pessimism of the assumptions on which calculations of the flow analyses were made. In the next few years, technological innovations can be expected to improve efficiency and reduce losses of the systems. In relative terms, greenhouse gases emissions should therefore decrease more than absolute fuel consumption. However, these complex developments are difficult to forecast and would require a more in-depth analysis, which goes beyond the scope of the present research. Figure 4 Histograms showing the comparative results of the current situation and the three scenarios, as well as the confrontation to the 2'000 Watts society objectives. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 97 This confrontation to the 2’000 Watts society targets illustrates that no single strategy can work on its own to move societies towards a global energy transition. The behavioral scenario, which appears at first view as the better candidate, raises a series of questions in terms of acceptance and future oppositions. The energy consumption of the technologic scenario remains too high – not to speak about unresolved problems of economic feasibility. The same doubts can be expressed with respect to the symbiotic scenario, which would require intense political backing and educational support in order to translate the opportunities offered by urban symbioses into real achievements. Accordingly, a balance needs to be achieved by merging these strategies. It is precisely the objective of the future stages of the present research, which will explore this middle way approach by developing integrated scenarios, combining technological innovation, changes of lifestyle and short-distance exchanges. REFERENCES Agudelo-Vera, C. M., Leduc, W. R. W. A., Mels, A. R., & Rijnaarts, H. H. M. (2012). Harvesting urban resources towards more resilient cities. Resources, Conservation and Recycling, 64(0), 3–12. Baccini, P. (1996). Understanding Regional Metabolism for a Sustainable Development of Urban Systems. ESPR Environ. Sci. & Pollut. Res., 3(2), 108–111. Barles, S. (2008). Comprendre et maîtriser le métabolisme urbain et l’empreinte environnementale des villes. RESPONSABILITÉ & ENVIRONNEMENT, N° 52, 21–26. Bauart Architectes et Urbanistes SA. (2010). Commune d’Yverdon-les-Bains, PDL Gare-Lac. Codoban, N., & Kennedy, C. A. (2008). Metabolism of Neighborhoods. Journal of Urban Planning and Development, (March), 21–31. Dobbelsteen, A. van den. (2008). Towards closed cycles - New strategy steps inspired by the Cradle to Cradle approach. PLEA 2008 – 25th Conference on Passive and Low Energy Architecture. Dublin. Dobbelsteen, A. van den, Keeffee, G., Tillie, N., & Roggema, R. (2012). Cities as Organisms. In R. Roggem (Ed.), Swarming Landscapes: The Art of Designing For Climate Adaptation (pp. 196–206). Dordrecht: Springer Science+Business Media. Erkman, S. (1998). Vers une écologie industrielle. (C. L. Mayer, Ed.). Paris. Gorgolewski, M., Komisar, J., & Nasr, J. (2011). Carrot City: Creating Places for Urban Agriculture. (The Monacelli Press, Ed.). Grospart, F. (2009). Autonomie énergétique locale Sommaire. Vendôme. IDDRI Modes de vie et empreinte carbone (2012). Jochem, E. (2004). Steps towards a sustainable development a white book for R & D energy-efficient technologies. Novatlantis. Jourdan, S., & Mirenowicz, J. (2011). L’agriculture regagne du terrain dans et autour des villes. La Revue Durable, 43. Lufkin, S., Rey, E., & Erkman, S. (2013). Quartiers symbiotiques: augmenter le potentiel d’autonomie énergétique à l’échelle locale. Vers la ville symbiotique? Valoriser les ressources cachées. Tracés. Lufkin, S., Rey, E., & Erkman, S. (2014). Quartiers symbiotiques. Stratégies innovantes pour favoriser l’autonomie énergétique à l’échelle du quartier par l’intégration des enjeux relatifs aux bâtiments, aux infrastructures, à la mobilité et à l’alimentation. CROSS, Final report. Lausanne. Newman, P. W. G. (1999). Sustainability and cities  : extending the metabolism model. Landscape and Urban Planning, 44(February), 219–226. Previdoli, P. (2012). Energiestrategie 2050. In B. für E. BFE (Ed.), . Bundesamt für Energie BFE. Rey, E. (2006). Integration of energy issues into the design process of sustainable neighborhoods. PLEA2006 - The 23rd Conference on Passive and Low Energy Architecture. Geneva. Rey, E. (2011). Concevoir des quartiers durables. In OFEN / ARE (Ed.), Quartiers durables. Défis et potentialités pour le développement urbain. (OFEN / ARE., pp. 15–24). Bern. Rey, E., Lufkin, S., Renaud, P., & Perret, L. (2013). The influence of centrality on the global energy consumption in Swiss neighborhoods. Energy and Buildings, 60(0), 75–82. Thomas, M.-P. (2011). En quête d’habitat  : choix résidentiels et différenciation des modes de vie familiaux en Suisse. EPFL. Tillie, N., Kürschner, J., Mantel, B., & Hackvoort, L. (2011). The Amsterdam guide to energetic urban planning. Management and Innovation for a Sustainable Built Environment. Amsterdam. Tillie, N., Van Den Dobbelsteen, A., Doepel, D., Joubert, M., De Jager, W., & Mayenburg, D. (2009). Towards CO2 Neutral Urban Planning: Presenting the Rotterdam Energy Approach and Planning (REAP). Journal of Green Building, 4(3), 103–112. Wallbaum, H. (2012). Mainstreaming energy and resource efficiency in the built environment – just a dream  ? In IED public lecture series (Ed.), . IED public lecture series. Zimmermann, M., Althaus, H.-J., & Haas, a. (2005). Benchmarks for sustainable construction. Energy and Buildings, 37(11), 1147–1157. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 98 Session 1D : Tools and methods/ framework PLEA2014: Day 1, Tuesday, December 16 11:30 - 13:10, Trust - Knowledge Consortium of Gujarat PROSOLIS: a Web Tool for Thermal and Daylight Characteristics Comparison of Glazing Complexes O. Dartevelle, M. Sc. Arch. A. Deneyer, M. Sc. Arch. M. Bodart, Prof. Université catholique de Louvain (U.C.L.) olivier.dartevelle@uclouvain.be Belgian Building Research Institute Université catholique de Louvain ABSTRACT Since these last years, the application of the European Energy Performance Building Directive (EPBD) has led to a higher interest in summer comfort issue. In this context, the design of glazing complexes (glazing and solar shading) is a key issue since it directly influences the thermal and visual perception of interior spaces. Glazing complex determine the view and the opening to the outside, determining solar gains and penetration of natural light, but is also responsible for heat loss and can cause overheating and glare. The choice of an adequate glazing complex should therefore be done considering all of these aspects. This paper presents a free web tool realised within the frame of the PROSOLIS research project. Based on a set of results obtained by the advanced use of BSDF functions for optical properties description of solar shading in specific thermal and daylight simulation software’s (WINDOW 7, EnergyPlus 8, Lighttools 8.0), the PROSOLIS web tool helps to evaluate the impact of the glazing complex choice on both thermal and visual comforts in residential and office buildings. This web tool, dedicated to building designers, proposes a multi-criteria approach for comparing accurately the most current types of glazing complexes. It considers internal and external screen fabrics and venetian blinds, combined with five different types of glazing and informs designers on energy and light performance levels of the selected combinations. From this information, designer should be able to easily choose glazing complexes fitting with their needs. INTRODUCTION PROSOLIS is a tool designed to compare the energy and light performance levels of different glazing and solar shading combinations. Users can therefore study and compare the behaviour of different combinations of glazing and of solar shading parallel to glazing for a wide range of configurations depending on the position of the solar shading, window orientation and use of the studied building. The tool is divided into 6 main screens: use, orientation, glazing, solar shading, concise and detailed results. It also integrates a word index defining all technical terms used in the tool. The help section shows how to use the tool and presents the hypothesis on which the simulations are based. The tool is available in French and English on www.prosolis.be. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 99 USE SELECTION In the first screen of the tool, the user can choose from three room types: living room (residential building), sleeping room (residential building) and individual office space (non residential building). This choice determines the simulation conditions and hypotheses behind the contextualized energy property results (cooling needs of the room, heat balance of the complex (see after)). To provide valid thermal behaviour, two whole buildings have been modeled (Figure 1): one residential building and one office building. Both models include 13 thermal zones. The first covers all everyday functions of a single-family home (kitchen, sleeping room, living room, washroom, etc.) and the second covers those of an office building (office space, corridors, etc.). All thermal simulation hypothesis regarding geometry, construction types, internal gains, hours of occupancy are described in the help section of the tool. Figure 1 Left: residential building - Right: office building Thermal modellings. ORIENTATION SELECTION The user can choose among eight different orientations (North, North-East, East, South-East, South, South-West, West and North-West) determining the conditions for the room being studied. GLAZING SELECTION Five different types of glazing (presented in Table 1) are proposed: clear double glazing; clear double solar control glazing; double glazing with enhanced solar control; reflecting double glazing; triple glazing. Energy properties Thermal transmittance factor Solar factor Solar transmittance Solar reflectance Solar absorption Light properties Light transmittance Colour rendering index Table 1. Properties of glazing Clear double Clear solar Double double control with enhanced glazing glazing solar control Ug Reflecting double glazing Triple glazing 1.2 1.1 1.1 1.0 0.7 g [-] τe[%] ρe [%] αe [%] 0.62 53 23 24 0.41 37 28 35 0.35 31 32 37 0.30 27 33 40 0.60 53 25 23 τv [%] 77 69 61 50 72 Ra [%] 98 97 95 97 99 [W/m²K] SOLAR SHADING SELECTION As presented in Figure 2, the user can specify: the potential absence of solar shading; the position of the solar shading: interior or exterior; the type of solar shading: blinds or screens; the properties of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 100 the selected solar shading (optical properties, tint, fabrics type, etc.). For screens, the user can choose among 6 screens of different colours (black, grey, white) and fabric types (natte or serge). For blinds, the user can choose among 4 slats of different colours (dark or light grey) and reflexion types (diffuse or specular). The solar shading devices were selected to represent the products found in practice for standard solar shading applications. Their properties are presented in Table 2 for screens and in Table 3 for venetian blinds. For blinds with metal slats, the slats are fixed and inclined at a 30°angle in relation to the horizontal position. Screen for shading device selection Figure 2 Energy properties Solar transmittance Solar reflectance Light properties Light transmittance Other properties Openness factor Energy properties Solar reflectance Light reflectance Light properties Light transmittance Table 2. Properties of screens Serge Serge Serge White Grey Black Natte White Natte Grey τe [%] ρe[%] 20.5 66.5 7.1 19.0 3.7 5.9 27.1 61.6 15.1 19.0 τv[%] 19.9 5.4 3.7 26.4 13.1 O.F. [%] 4.3 4.2 3.3 12.1 11.4 Table 3. Properties of slats Diffuse Specular reflection reflection Light grey Light grey ρe,n-h [%] ρv,n-h [%] τv[%] Diffuse reflection Dark grey Specular reflection Dark grey 52.7 59.2 58.8 59.4 20.8 21.1 39.0 44.5 0.0 0.0 0.0 0.0 CONCISE RESULTS This section of the tool allows users to easily compare the behaviour of different glazing complexes (up to four) for the following criteria: Overheating protection; Daylight harvesting; Glare protection. Overheating protection This criterion compares the efficiency of the chosen glazing complexes (glazing and solar shading) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 101 regarding their impact on the reduction of the cooling needs of the room (here, the reduction is seen as the difference of “Annual cooling needs of the room” (see detailed results after) between the configuration with double clear glazing (and no solar shading) and the configuration with the selected glazing complex). On the right side of the screen (see Figure 3), these values are displayed on a scale defined by the minimum and maximum values of coolings needs reduction obtained in the tool (for the selected use). This permits a precise comparison of the selected combinations. Also, in the center of the screen, symbols are used to describe in a simple way the impact of each selected combination: ‘/’ for negligible protecion; ‘+’ for low protection; ‘++’for medium protection; ‘+++’ for high protection. These categories were calibrated by qualifying all internal shading devices for north orientation as negligible protection and all external shading devices for south as high protection. Daylight harvesting This criterion compares the daylight penetration through the glazing complex. It is based on the “Daylight harvesting” criterion presented in the detailed results section of the tool (see after). For the selected use and orientation, it expresses the daylight penetration through the glazing and solar shading combination (mean for summer and winter conditions) in relative terms in relation to the maximum value obtained for the double clear glazing configuration (without solar shading). The scores obtained depend on the position of the chosen solar shading on a scale defined by the minimum (0%) and the maximum (100%). As already seen for the “overheating protection” criteria, these values are displayed on a scale on the right side of the screen to ease precise comparison. Also, the following symbols are used to describe the behavior of each selected combination: ‘+’ for poor daylight supply (0 to 10%); ‘++’ for moderate daylight supply (10 and 23%); ‘+++’ for good daylight supply (23 to 100%). Theses boundaries were calibrated to highlights the best cases with shading devices. Glare protection This criterion compares the impact of the chosen glazing complex on protecting against glare. It is based on the “Glare protection” criterion (see detailed results). The scores obtained depend on the category of solar shading for this criterion. The following symbols are used: ‘/’ if no solar shading is present, ‘+’ for low protection; ‘++’ for medium protection; ‘+++’ for high protection. a b c d a c c b d b d a Concise results obtained for the comparison of the following combinations: (a) double Figure 3 clear glazing; (b) reflecting double glazing; (c) double clear glazing with internal screen (serge grey); (d) double clear glazing with external venetian blinds (dark grey slats with specular reflection). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 102 DETAILED RESULTS This screen (Figure 4 and 5) is used to compare more in details the thermal and visual characteristics of multiple (up to four) different combinations of glazing and solar shading selected by the user. It includes 5 different sections: Summary of choices; Glazing properties; Solar shading properties; Solar energy properties of the combination glazing and solar shading; Light properties of the combination glazing and solar shading. Detailed results - Solar shading properties- obtained for the comparison of the Figure 4 following combinations: (a) double clear glazing; (b) reflecting double glazing; (c) double clear glazing with internal screen (serge grey); (d) double clear glazing with external venetian blinds (dark grey slats with specular reflection). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 103 Detailed results – Solar energy and light properties of the combination “glazing and Figure 5 solar shading” - obtained for the comparison of the following combinations: (a) double clear glazing; (b) reflecting double glazing; (c) double clear glazing with internal screen (serge grey); (d) double clear glazing with external venetian blinds (dark grey slats with specular reflection). Summary of choices This section (see Figure 4) summarises the selections made by the user: use, orientation, glazing, solar shading, type, position, colour, type of reflectance, fabric type. Glazing properties This section (see Figure 4) presents the property values of the glazing selected by the user: Solar factor; Thermal transmittance factor (U-value); Light transmittance. Solar factors and light transmittance factors of the glazing were determined by the glazing manufacturer in compliance with the standard NBN EN 410:1998. Thermal transmittance factors of the glazing were determined by the glazing manufacturer in compliance with the standard NBN EN 673:2011. Solar shading properties First, this section (see Figure 4) covers the detailed values of the properties of the solar shading 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 104 selected by the user. For screens the following properties are described: Openness factor; Solar transmittance (normal-hemispherical, normal diffuse, normal-normal); Reflectance and Absorption; Light transmittance (normal-hemispherical, normal diffuse, normal-normal). For blinds, Solar and Light reflectances are described. The solar and light transmittance, reflectance and absorption values were determined in compliance with the standard NBN EN 14500:2008. The screen openness factor was based on a physical measurement of the proportion of holes in compliance with Annex B of this standard. The classifications for protection against total heat transfer, protection against direct transmittance, protection against secondary heat transfer, opacity, glare control, night privacy, visual contact with the outside and daylight supply are then described according to the standard NBN EN 14501:2005. This standard establishes classifications for these properties ranging from 0 to 3 or 4: 0 being a property resulting in very little effect and 4 being a property resulting in a very good effect. Closing this section of the detailed results, outwards view in daytime conditions and inwards views in nighttime conditions are given. These images were determined for standard daytime and nighttime observation conditions. They were taken in a laboratory under controlled lighting conditions, in particular with regard to background contrast differences, thus generating an accurate reproduction of internal and external views. The observation distance was set to 80 cm from the solar shading device. Under daytime conditions, a lighting contrast with a ratio of 1 to 300 was generated between the vertical plane on which the solar shading device is positioned and the background. Under nighttime conditions, this lighting contrast was maintained at a ratio of 1 to 4000. Energy properties of the combination glazing and solar shading First, this section (see Figure 5) covers properties of the glazing and solar shading combination at normal incidence: Solar factor (gtot); Direct (te) and secondary heat transfer factor (qi); Shading factor (Fc). These values were obtained using the Window 7.2 software (LBNL, 2013). The calculation was made taking into account on the one hand glazing data issued by manufacturers, in compliance with the standard NBN EN 410:2011 and on the other hand the spatial behavioural properties of the solar shading (using Bidirectional Scattering Distribution Functions (BSDF) (Deroisy et al, 2013)). Then, this section presents the energy properties characterising the glazing and solar shading combination in the context (use and orientation) defined by the user: heat balance of the complex; cooling needs of the room; impact of solar shading on cooling needs. All of these results were calculated by dynamic thermal modelling using EnergyPlus V8.1 software (DOE, 2013). These simulations were performed by series of 5-minutes time intervals over a standard year in Brussels (ASHRAE, 2001). The simulations took place for the geographic location of Uccle. The results for this criterion must therefore be considered for this geographic position (Latitude 50.8°N). All of the optical properties of the glazing and solar shading combinations used in the thermal simulation were introduced based on prior detailed modelling results (integrating BSDF measurements of solar screen properties) derived from the WINDOW 7.2 software via a BSDF formalism (Dartevelle et al., 2013). An automated solar shading management system was modelled. This was based on a criterion of 150 W/m² of total solar radiation on the window to trigger the closing of the solar shading device. For more representative results, a solar shading device was modelled on all windows with the orientation selected by the user. No shading caused by the outdoor environment was taken into account. The heat balance is obtained by the sum of the monthly gains and losses of the room by conduction, radiation and convection through the glazing and solar shading combination (DOE, 2013(2)). The monthly cooling needs of the room analyzed are calculated based on an indoor temperature of 25°C that must not be exceeded during occupancy. The impact of solar shading on cooling needs is quantified based on the difference in cooling needs between the considered case and the same case without solar shading (all other considered assumptions (use, orientation, glazing) being identical). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 105 Light properties of the combination glazing and solar shading This section (see Figure 5) covers on the one hand the light properties of the combination glazing and solar shading at normal incidence: Light transmittance. The light transmittance values of the different glazing and solar shading combinations were obtained using the Window 7.2 software tool. On the other hand, this section covers the properties characterising the glazing and solar shading combination in the context (use and orientation) defined by the user: Glare control, Daylight supply (summer and winter) according to the selected orientation. Information regarding glare control was collected by collating the results from advanced computer simulations performed using LightTools 8.1 software (OSG, 2013), integrating precise data on the properties of the materials used to constitute the solar shading devices (BSDF data) and brightness measurements for solar shading devices under real external exposure conditions calculated by Photolux 3.2 (SE, 2012) software based on High Dynamic Range (HDR) images. The direction of observation of the solar shading is perpendicular and the distance of the observer from the solar shading is such that with an average solar altitude, direct view of the sun is impossible. Three categories have been created, distinguishing the mean amount of light perceived through glazing and solar shading combinations: Low glare control (mean brightness greater than 3000 cd/m2); Medium glare control (between 1000 and 3000 cd/m2); High glare control (less than 1000 cd/m2). The information regarding "daylight harvesting" was also generated by computer simulations using LightTools 8.1 software and integrating BSDF data measured for the materials used to constitute solar shading devices. It represents the total light flow through the combination with the solar shading device extended where applicable and for a perfectly clear sky. It was calculated for each orientation on a vertical reference plane located behind the glazing complex and exposed to a cumulated clear summer (15 June) and winter (15 December) sky. This criterion is expressed in relative terms compared to the maximum value obtained for all considered configurations. CONCLUSION The PROSOLIS web tool proposes an original multi-criteria approach for comparing precisely performance levels of common types of glazing complexes (glazing and shading devices). It permits to easily obtain and compare their detailed and contextualized energy and light characteristics. In this way, this tool should help designers to choose glazing complexes corresponding to their needs. REFERENCE ASHRAE. 2001. International Weather for Energy Calculations (IWEC). Atlanta: American Society of Heating Refrigeration and Air Conditioning Engineers, Inc. Dartevelle, O., Lethé, G., Deneyer, A., Bodart, M.. 2013. The use of bi-directional scattering distribution functions for solar shading modelling in dynamic thermal simulation: a results comparison. Lausanne: CISBAT 2013. Deroisy, B., Deneyer, A., Lethé, G., Flamant, G.. 2013. Performance analysis of common solar shading devices: experimental assessment and ray-tracing calculations using bi-directional scattering distribution data. Cracovia: Lux Europa 2013. DOE. 2013. EnergyPlus 8 simulation software. Washington DC: U.S. Department of Energy. DOE. 2013(2). EnergyPlus 8 Engineering Reference Document.Washington DC: U.S. Department of Energy. LBNL. 2013. Window 7 Simulation Tool. Berkeley: Lawrence Berkeley National Laboratory. NBN EN 14500. 2008. Blinds and shutters - Thermal and visual comfort - Test and calculation methods. Bruxelles: Bureau for Standardisation. NBN EN 14501. 2005. Blinds and shutters - Thermal and visual comfort - Performance characteristics and classification. Bruxelles: Bureau for Standardisation. NBN EN 410. 1998. Glass in building - Determination of luminous and solar characteristics of glazing. Bruxelles: Bureau for Standardisation. NBN EN 673. 2011. Glass in building - Determination of thermal transmittance (U value) - Calculation method. Bruxelles: Bureau for Standardisation. OSG. 2013. LightTools 8.1 software. Pasadena: Synopsys’ Optical Solutions Group. SE. 2012. Photolux 3.2 software. Ecully:Soft Energy SARL. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 106 Improving the Energy Efficiency of the Building Stock: A Bottom-up Model and its Application in an Online Interactive Portal Simon Cuvellier Anne-Françoise Marique, PhD Sigrid Reiter, Prof Université catholique de Louvain Architecture et Climat Université de Liège LEMA Université de Liège LEMA ABSTRACT There is an urgent need to reduce energy uses in new and retrofitted buildings. In Europe, energy consumption in the building sector still represents more than 40% of the final energy use. Emerging countries are also concerned by such issues at even wider levels because of the huge demographic growth they are witnessing. Numerous research studies have highlighted the need to produce more efficient buildings, but also to retrofit the existing building stock. However, research methods and tools that allow a precise quantification of energy uses in buildings and energy savings related to various actions (insulating the roofs, changing the glazing, behavioral changes, etc.) are mainly dedicated to trained professional users, thus neglecting the huge potential energy savings that is linked to individual actions undertaken by citizens in their dwellings. In this context, the main aim of our research is to raise awareness of energy efficiency in residential buildings and encourage positive changes to the energy efficiency of the building stock, starting at the individual scale. This paper first presents the methodology that allows a precise energy assessment (heating, cooling, ventilation, lighting, appliances and cooking) of buildings (at the house, neighborhood, city and region scales) on the basis of a “bottom-up” approach. This methodology uses a typological classification of buildings, thermal simulations and local surveys. In this paper, this methodology is applied to the Walloon (Belgium) building stock. Many parameters are defined and taken into account to capture the specificities of numerous types of buildings (e.g., the number of floors, common ownership, orientation, thermal performances, ventilation, etc.). Several occupation modes are modelled to capture the impact of occupants’ behavior on energy consumption. To take into account the impact of urban form, correction factors are defined and applied according to the type of neighborhoods in which the buildings are located. All things considered, 250,000 individual results are obtained and stored in a huge database. Linear extrapolations and correction factors are used to extrapolate and apply these results to any type of residential building in Wallonia. This methodology is then used to develop an online portal that aims to strengthen citizens’ awareness of the necessity for ecological changes in the building sector and encourage individual actions to improve the energy efficiency of buildings. This tool allows for a transfer of the main results of a two-year scientific research effort to citizens in a very simple and intuitive way. Although the results presented in this paper are focused on Wallonia (Belgium), the research is easily reproducible to other territories by adapting local parameters. INTRODUCTION There is an urgent need to reduce energy uses in new and retrofitted buildings. In Europe, energy consumption in the building sector still represents more than 40% of the final energy use. Emerging countries are also concerned by such an issue at even wider levels because of the huge demographic growth 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 107 that they are witnessing. Numerous research efforts have highlighted the need to produce more efficient buildings, but also to retrofit the existing building stock, especially in Europe where the renewal rate of buildings is quite low. Moreover, several research and empirical results have demonstrated the significant impact of the behavior of housing occupants on energy consumption (e.g., de Meester et al., 2012; Santin et al., 2009). However, research methods and tools that allow a precise quantification of energy uses in buildings and energy savings related to various actions (insulating the roof, changing the glazing, behavioral changes, etc.) are mainly dedicated to trained professional users, thus neglecting the huge potential energy savings linked to individual actions undertaken by citizens in their dwellings. Citizens are, in fact, the first actors who can concretely act to alter the energy consumption in residential buildings. However, although an increasing number of households are paying attention to their energy consumption and are motivated to undertake light or heavy renovation work, they do not know what action to choose and are unaware of the impacts of renovation in terms of comfort, energy savings, etc. In fact, efforts to promote energy efficiency remain concentrated on the general guidelines; in particular, user-friendly assessment tools dedicated to a non-specialized audience (local authorities, developers, citizens) are lacking (Tweed and Jones, 2000). Most existing BPS (Building Performance Simulation) tools (e.g., TRNSYS, Comfie+Pleiades, Energyplus, phpp) are designed by engineers for use by other trained engineers, which make them too complicated to quickly evaluate the performance of different design concepts or strategies (Attia et al., 2012). Amongst the existing simplified evaluation tools, Gratia and De Herde (2002a and 2002b) developed a simple design tool for the thermal study of dwellings and office buildings. The calculation of the energy consumption is mainly dedicated to architects and based on the results of dynamic thermal software. Performing an assessment by using a simple design tool is much easier than performing an assessment while using this thermal simulation software, but the values of many parameters that are not often known by households and local authorities are still required. In this paper, we argue that the implementation of energy efficiency measures into concrete policies and the popularization of academic research to the general public (citizens, local authorities, policy makers, etc.) are crucial to ensure a more sustainable development of our territories and to reduce energy consumption in buildings. The main aim of our research is, thus, to encourage positive changes to the energy efficiency of the building stock, starting at the individual scale, by transferring the main results of a two-year research to a non-specialized audience. The need for this type of research lies in the fact that its dissemination is for “normal people,” for them to have the necessary information to conduct themselves and their homes more energy efficiently. To this end, this paper first presents the methodology that enables a precise assessment of energy uses in buildings. This methodology is applied to the Walloon building stock and a huge database comprised of more than 250,000 individual results is produced. Then, the methodology and the database are used to develop an online portal that aims to raise public awareness on energy efficiency in buildings. METHODOLOGY: A BOTTOM-UP MODEL A methodology was developed to assess energy uses (energy requirements for space heating, cooling, ventilation, electrical appliances, cooking and domestic hot water) in residential buildings at an individual scale. This methodology must allow researchers to precisely assess energy uses at the individual building scale, but also to draw trends, at the neighborhood, city and regional scales. This methodology combines several research methods and tools, including a typological classification of buildings, dynamic thermal simulations, surveys, etc. The energy consumption levels related to heating, cooling and ventilation are derived from dynamic thermal simulations. The energy consumption levels related to domestic hot water, electrical appliances and cooking are based on regional empirical surveys and are linked to the number of inhabitants in each dwelling. In this paper, this methodology is applied to the Walloon (Belgium) building stock. However, this work is also reproducible to any other territory by adapting local parameters. Parameters taken into account to develop the bottom-up model The parameters that were taken into account to build the bottom-up model are related to (A) the environment in which the building is located, (B) the characteristics of the building, (C) the thermal 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 108 performances of the envelope and the systems. They are explained below and summarized in Figure 1. As far as the environment in which the building is located is concerned, 1,347 possibilities were defined to cover the variation of the climate (see column A.1 in Figure 1) (in comparison with Brussels, a temperate climate in the northern part of Europe) in Wallonia. A coefficient based on degree-day was attributed to each location and then applied to thermal simulation results (performed with Brussels’ climate). Eight main types of residential neighborhoods (A.2) were defined (dense urban core, continuous urban, semi-continuous urban, homogeneous semi-continuous and social housing, villages and rural cores, suburban neighborhoods, isolated rural, great sets) on the basis of a typological classification of the whole Walloon building stock (Marique and Reiter, 2013). Simulations were performed on 24 selected representative neighborhoods with Townscope software (Teller and Azar, 2001): 500 points were randomly defined on the facades and roofs of each neighborhood and an assessment of solar gains was performed in order to define corrections factors according to the density of the neighborhood. These correction factors were stored in the database and then applied to the results of thermal simulations of buildings, in order to take into account the diminution of solar gains on facades and roofs, according to the built density of the neighborhood in which the considered building is located. A. ENVIRONMENT 1. Climate 2. Neighborhood B. TYPOLOGY 1. Housing type Dense downtown 2. Number of floors C. THERMAL PERFORMANCES AND SYSTEMS 3. Common ownership 4. Orientation Detached or semi-detached North 1. Wall type Wide house Continuous urban North or East Detached or semi-detached North 3. Wall insulation [cm] 0 0 3 3 North or East Detached or semi-detached North 5. Glazing 6. Ventilation 7. Thermostat 16 2 Terraced 4. Roof insulation [cm] 0 1,5 Terraced 2. Slab insulation [cm] 3 16 Simple, double-old or double-new 18°C A 8. Heating system 9. Fuel type Hot water boiler condensing Natural gas Hot water boiler non condensing Gazole Electric resistance heating Propane Heat pump Butane Micro-CHP LPG stove Coal Radiator or electric heater Wood (4 forms) Electric storage heater Electricity Solid Semi-continuous urban Choice amongst 1347 locations and their related degree days Homogeneous semi-continuous or social city Wide crossing apartment Narrow crossing apartment 1 1 Three fronts apartment 1 One front apartment 1 "Great Sets" Corner apartment 1 6 Terraced North or East 16 Detached or semi-detached North 10 Double-old, double-new or triple Terraced North or East 20 Double-new or triple Semi-detached North 15 Double-old, double-new or triple Terraced North or East 30 Double-new or triple Terraced North or East 20 Double-old, double-new or triple 30 Double-new or triple 25 Double-old, double-new or triple 35 Double-new or triple 30 Double-old, double-new or triple 35 Double-new or triple 10 3 4 Isolated rural 6 2 Narrow house Village or rural nucleus Suburban subdivision 6 1,5 15 Top, intermediate or ground Top, intermediate or ground Top, intermediate or ground Top, intermediate or ground Top, intermediate or ground 20 10 15 20 North or East Cavity North or East 25 25 North North, East, South or West 30 North, East, South or West 30 20°C C or D 20-16°C Summary of the parameters related to (1) the environment in which the building is Figure 1 located, (2) the characteristics of the building, (3) the thermal performances of the envelope and the systems. As far as the characteristics of the buildings are concerned, 60 main types of dwellings were defined to cover the whole Walloon residential building stock on the basis of the main characteristics of Walloon dwellings, as identified in previous research (Evrard et al., 2012; Kints, 2008). The typology is comprised of houses and apartments. For houses, the characteristics that were taken into account to build the typology are the plan orientation (B.1) (perpendicular or parallel to the street), the number of floors (B.2) (one and half, two, three or four), the common ownership (B.3) (detached, semi-detached or terraced house) and the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 109 orientation (B.4) of the building (north, east, south, west). For apartments, the characteristics were the plan configuration (B.1) (wide crossing, narrow crossing, three fronts, one front or corner apartment), the position (B.3) of the apartment in the building (ground floor, intermediate floor or top floor apartment) and the orientation (B.4) of the building (north, east, south, west). Each of the 60 types has a fixed heated surface area. Extrapolations are then performed to extend the results obtained in the thermal simulations to similar types of buildings presenting different heated surface areas (see below). As far as the thermal performances of the housing are concerned, two types of wall (C.1) are defined (solid or cavity). The insulation in the slabs and the walls (C.2 and C.3) vary from 0 to 30 centimeters. The insulation in the roofs (C.4) varies from 0 to 35 centimeters. The glazing type (C.5) may be simple, doubleold, double-new or triple. The glazing surface area on each façade is defined according to the housing type. For all types of dwellings, the ceiling height is worth 2.4m; no attachments are modeled as such, but they can be taken into account by including them in the total area of the dwelling; the basement floor is not taken into account. Three ventilation modes (C.6) are defined: natural ventilation (type A), ventilation with mechanical extraction (type C) and double flow ventilation with heat recovery (type D). For the ventilation of type A, the windows are opened for an internal temperature higher than 25°C and are closed when this temperature drops down to 23°C. In the case of ventilation of types C and D, air flows are defined according to the Belgian requirements (NBN, 2008). The heat recovery system efficiency is set at 85%. Three types of thermostat (C.7) are defined (18°C constant during day and night, 20°C constant during day and night and 20°C with a reduction to 16°C in a daily work-pattern and during night. No weekly or annual profile has been defined.). Eight types of heating systems (C.8) and 11 types of fuel (C.9) were taken into account. For each type, a correction coefficient was used to integrate the efficiency of the heating system (production, distribution and emission). These coefficients come from the Belgian regulation (PEB, 2008 and 2012). A couple of rules of combinations were finally defined to eliminate nonrealistic cases (for example, a building with 20 cm of insulation and simple glazing). In all, 250,000 types of buildings were defined. Assessment of energy uses in buildings The TRNSys thermal simulation software was then used to perform an energy consumption analysis of space heating needs and electricity needs for ventilation systems and solar gains for each of the 250,000 cases in the Belgian context (the climate of Brussels without any surroundings buildings). Cooling was not considered in the analysis because cooling needs are minimal in Belgium. In these simulations, internal gains were defined according to Massart and De Herde (2010), such as 70W/person for occupation and 6W/m² and 4W/m² for nominal power lighting and appliances, respectively. They are functions of the dwelling’s surface and of the number of occupants which are set in function of the type of housing. Internal gains are also set according to a daily and weekly schedule. In addition, three standards defined by the European Energy Performance of Buildings Directive (EPBD) were added: the low-energy standard, the very low-energy standard and the passive standard, which correspond to annual heating requirements lower than 60 kWh/m².year, 30 kWh/m².year and 15 kWh/m².year, respectively. The energy consumption related to appliances and cooking are respectively evaluated at 1,000 kWh per person and per year and 165 kWh per person and per year in Belgium on the basis of a local survey of energy uses by households (ICEDD, 2008). The energy consumption related to domestic hot water is assumed to be dependent on the number of inhabitants. We consider that each inhabitant needs 100 liters of cold water (10°C) and 40 liters of hot water (60°C) per day in accordance with the regional trends (ICEDD, 2008). The number of inhabitants is dependent on the surface area of the dwelling. Storage of the results The results of the energy assessments (space heating, cooling, ventilation, appliances, cooking and hot 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 110 water) were stored in a huge database comprised of seven parts: Part 1 is dedicated to the based degreedays coefficient, Part 2 stores the solar factors depending on the built density of the neighborhood in which the building is located, Part 3 is dedicated to space heating requirements and electricity needs for the ventilation system, Part 4 relates to the characteristics of the heating systems, Part 5 addresses domestic hot water requirements, Part 6 relates to cooking requirements and Part 7 is dedicated to electrical appliances requirements. Part 3 (space heating energy needs and electricity needs for the ventilation system) is comprised of seven columns, as illustrated in Table 1. The first column includes a unique numeric code that allows one to directly and easily identify the corresponding building and its characteristics: each item of the code corresponds to a specific variation of a parameter and provides the identity card of the tested case. Table 1. Example of the database for the space heating and ventilation needs Code mq pq ms ps mv pv 1_3_2_1_c_l_g_3_2_1 1_3_2_1_c_l_g_3_2_2 1_3_2_1_c_l_g_3_2_3 1_3_2_1_c_l_g_3_3_1 1_3_2_1_c_l_g_3_3_2 1_3_2_1_c_l_g_3_3_3 41.9 54.1 49.1 24.7 32.4 30.2 38.6 142.3 105.5 558.9 792.5 681.6 8.1 13.9 8.1 8.1 11.3 8.1 486 768.9 486 486 646.2 486 2.2 2.2 2.2 6.9 6.9 6.9 -23.1 -23.1 -23.1 -73 -73 -73 Columns 2 and 3 are used to store the slope (mq) and the intercept (pq) of the linear extrapolation used to generalize space heating energy needs obtained through thermal simulations to any similar building that presents a different heated surface area. Columns 4 and 5 (ms and ps) store parameters related to solar gains. The two last columns (mv and pv) are used to calculate the electrical needs of the ventilation’s fans. The coefficients that are used to transform energy needs into fuel consumptions, primary energy consumptions, CO2 emissions and into an estimation of the annual cost are also stored in the database. These coefficients are used with an identification number that depends on the systems used in the simulations. Extrapolation and final results Energy needs for space heating, cooling and ventilation are obtained by using the surface area of the dwelling (S) and the data from the linear extrapolation (mq and pq) stored in the database. Afterward, a first correction is applied to energy need for space heating, cooling and ventilation to take into account the neighborhood in which the dwelling is located and the loss of solar gains related to the density of the neighborhood (Fs, ms and ps). A second correction factor (C) is applied to take into account the location of the dwelling on the territory and the corresponding degree-days in comparison with Brussels’ climate. Finally, the annual space heating consumption (QSH) – expressed in kWh/year – is obtained by multiplying the space heating need by the efficiency of the whole heating installation (SHn), as shown in equation 1. QSH = [(mq * S + pq) + (ms * S + ps) x (1-Fs)] * C * SHn (1) Electricity consumption of the ventilation system (QVT) – in kWh/year – is calculated on the basis of the surface area (S) and the data of a linear extrapolation (mv and pv) from the heating database, as shown in equation 2. QVT = (mv * S + pv) (2) Energy needs for domestic hot water, electric appliances and cooking depend on the number of inhabitants in the dwelling (N). Such statistics also incorporate data from regional surveys, as shown in 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 111 equations 3, 4 and 5. The energy need for hot water (QHW) takes into account the quantity (liter) of hot water consumed annually (Lliter/year), the difference between the hot and cold water temperatures (ΔT) and the water heat capacity (Cv), as shown in equation 3. As for space heating consumption, energy consumptions for domestic hot water, appliances and cooking – expressed in kWh/year – are obtained by multiplying the respective energy need by the corresponding yield coefficient (HWn, EAn and CKn respectively). QHW = (Lliter/year * N * ΔT * Cv) * HWn (3) QEA = (-40 * N² + 550 * N + 1765) * EAn (4) QCK = (200 * N) * CKn (5) Finally, the household annual consumptions can also be converted into primary energy consumptions, CO2 emissions and euros by applying the conversion coefficients stored in the database. Validation and relevance of the results The methods and data used to build the database were presented extensively in previous papers (Marique and Reiter, 2012; Marique et al., 2014). The software used in the analysis has namely been validated by the International Energy Agency Bestest. We used the database to calculate the energy consumption of the whole building stock of Wallonia and compared this result with an in-situ survey (“annual thermal survey”) carried out by ICEDD (2008) on the basis of the real consumption of Walloon households. Differences between our simulations and figures from ICEDD are worth a maximum of 8.2%, which was considered to be acceptable. AN ONLINE INTERACTIVE PORTAL DEDICATED TO CITIZENS The model developed to assess energy uses in Walloon buildings and the numerous results stored in the database were then used to develop several types of applications that benefit different types of users (citizens, architects and urban planners, local or regional authorities, etc.). Due to the restricted length of this paper, in this section we will only present the online interactive portal that we developed for citizens. The main aim of this portal is to raise awareness of current energy issues and offer concrete solutions to reduce energy uses in buildings by arguing that citizens are the first actors who can concretely act to improve energy efficiency in residential buildings (the building stock is mostly private in Belgium). However, citizens often face huge difficulties in highlighting key parameters and strategies in the energy efficiency of their dwellings and in identifying the most efficient retrofitting work to perform in each particular case. To this end, the online portal that we developed makes available, in a very simple and intuitive way, more than 250,000 results of dynamic thermal simulations to a non-specialized audience. This knowledge aims to help them to assess energy uses in their dwellings and to make the best choices to improve its energy efficiency. Thus, it addresses one major shortcoming of existing simulation tools (BPS tools): the accessibility to citizens and local stakeholders. The online portal comprises three different evaluation tools, two of which are specifically dedicated to citizens. The simplified evaluation allows an individual user or household to assess building energy consumption on the basis of limited information. Completion of the questionnaires is very simple to allow the user to complete them quickly and without specific data and technical knowledge. The detailed evaluation allows an individual user or household to assess building energy consumption more precisely than the simplified evaluation. The questionnaires are more complex, but the results are closer to the real situation of the user and can be strongly personalized. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 112 To ensure a wide diffusion of the portal, the questionnaires used in the evaluation tools are simple, intuitive and easy to complete, as illustrated in Figure 2. The results are also expressed in a very simple form, as seen in Figure 3, for energy uses in the considered dwellings. Several strategies to improve the energy efficiency of the tested building are then provided (such as the insulation of the roof, the change of the glazing, the insulation of the whole building’s envelope, behavioral changes) to the user. They are personalized according to the characteristics of its dwelling. The quantification (in kWh/year and in %) of the potential energy savings linked to each strategy is also provided on the basis of the results stored in the database and the unique code used to store the results (Table 1). In addition to the results presented in this paper that are focused on energy uses in buildings, additional indicators are also provided in the online portal. In particular, it is possible to take into account the impact of the location of the dwelling on the daily mobility of inhabitants by assessing the energy consumption for daily mobility. The use of renewable energy is also included in the portal on the basis of the methodology developed by Marique et al. (2013) and Marique and Reiter (2014). The final version of this portal is online as of the end of August 2014 at www.solen-energie.be (only in French for the moment). In addition to the development of this online portal, numerous actions have been – and will beundertaken by the research teams to promote this initiative and to extensively raise awareness of the importance of improving energy efficiency in buildings, starting from the individual scale. These actions are dedicated to a wide range of actors: students in architecture and urban planning, researchers, local and regional stakeholders, private developers, architects, and, of course, citizens. It is too early to provide the results of usability testing since the final version of the tool has only been online for a few weeks, but the first simplified version of the portal that is solely comprised of suburban types of dwellings (Marique et al., 2012 and 2014) was launched between 2012 and August 2014. We have registered approximately 400 visits per month on the website. Direct interviews and workshops have shown positive feedback from users of the online interactive portal. Last, but not least, in May 2014 the online portal and the research project that allowed its development were awarded the “Energy Globe Award” for Belgium, one of the world’s most prestigious environmental awards (see also http://www.energyglobe.info/belgium2014?cl=english). Figure 2 Figure 3 Example of the questionnaires provided in the online portal (left) Example of results provided in the online portal (right) CONCLUSIONS This paper presented a methodology that enables one to precisely assess energy needs and energy consumption for space heating, cooling, ventilation, lighting, appliances and cooking within individual dwellings on the basis of a “bottom-up” approach. This methodology was based on a typological classification of buildings, thermal simulations, local surveys and linear extrapolations to cover a wide range of buildings and parameters. Among others, the impact of the location in which a dwelling is located 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 113 is taken into account via the application of correction factors based on the built density of the neighborhood. This methodology was applied to the Walloon (Belgium) residential building stock to build a huge database that included more than 250,000 individual results. This database was then mobilized to build an online interactive portal that aims to raise public awareness of energy efficiency in buildings. Thus, citizens are able to easily assess the sources of energy consumption for buildings. They may also compare these different energy consumption sources in order to determine relevant and personalized recommendations with which to reduce their energy consumptions. This interactive online portal represents the main results of an important two-year scientific research project dedicated to energy efficiency in Wallonia that is accessible to a large non-specialized audience, which is crucial in the scope of sustainable development. ACKNOWLEDGMENTS This research was funded by Wallonia-DGO4 (Belgium) under the “SOlutions for Low Energy Neighborhoods” (SOLEN) project. We thank J. Winant for the web development and the design of the portal and the numerous people who participated in the test sessions. REFERENCES Attia, S., Gratia, E., De Herde, A., and J.L.M. Hensen. 2012.Simulation-based decision support tool for early stages of zero-energy building design. Energy and Buildings, 49: 2-15. Evrard, A., Hernand, C. and A. De Herde. 2012. Vade-mecum – Outils EPEEH, Evaluation du potentiel d’économie d’énergie par type d’habitat wallon. Architecture et Climat – UCL. Gratia, E., and A. De Herde. 2002a. A simple design tool for the thermal study of dwellings. Energy and Buildings, 34: 411-420. Gratia, E., and A. De Herde. 2002b. A simple design tool for the thermal study of an office building, Energy and Buildings, 34: 279-289. ICEDD. 2008. Bilan énergétique wallon 2008. Consommations du secteur du logement 2008. MRW.Conception et réalisation ICEDD asbl. Kints, C. 2008. La rénovation énergétique et durable des logements wallons. Architecture et Climat – UCL. Marique, A.-F., and S. Reiter. 2012. A Method to Evaluate the Energy Consumption of Suburban Neighbourhoods. HVAC&R Research, 18(1-2), 88-99. Marique, A.-F., de Meester, T., and S. Reiter. 2012. An online interactive tool for the energy assessment of residential buildings and transportation. Proceedings of PLEA 2012, Lima, Peru. Marique, A.-F., and S. Reiter. 2013. Solar buildings and the urban environment. Paper presented at The 3rd New Energy Forum-2013. From Green Dream to Reality, Xian, China. Marique, A.-F., Penders, M., and S. Reiter. 2013. From zero-energy building to zero-energy neighbourhood: urban form and mobility matter. Proceedings PLEA 2013, Munich, Germany. Marique, A.-F., de Meester, T., De Herde, A., and S. Reiter. 2014. An online interactive tool to assess energy consumption in residential buildings and for daily mobility. Energy and Buildings, 78C: 50-58. Marique, A.-F., and S. Reiter. (2014). A simplified framework to assess the feasibility of zero-energy at the neighbourhood / community scale. Energy and Buildings, 82, 114-122. Massart, C., and A. De Herde. 2010. Conception de maisons neuves durables, Elaboration d’un outil d’aide à la conception de maisons à très basse consommation d’énergie. Architecture et Climat – UCL. de Meester, T., Marique, A-F., De Herde, A., and S. Reiter. 2013. Impacts of occupant behaviours on residential heating consumption for detached houses. Energy and Buildings, 57: 313-323. NBN. 2008. NORME NBN D50-001, Dispositifs de ventilation dans les bâtiments d’habitation. Brussels. PEB. 2008. Arrèté du Gouvernement wallon, Annexe I and Annexe III . PEB. 2012. Arrèté du Gouvernement wallon Annexe III . Santin, O.G., Itard, L., and H. Visscher. 2009. The effect of occupancy and building characteristics on energy consumption for space and water heating in Dutch residential stock. Energy and Buildings, 41: 1123-1232. Teller, T., and S. Azar. 2001. TOWNSCOPE II - A computer system to support solar access decisionmaking. International Journal of Solar Energy, 70(3): 187-200. Tweed, C., and P. Jones. 2000. The role of models in arguments about urban sustainability. Environmental Impact Assessment Review, 20: 277-287. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 114 Comparing deterministic and probabilistic non-operational building energy modelling Bengt Cousins-Jenvey, MSc Pete Walker, PhD Andrew Shea, EngD University of Bath University of Bath University of Bath Judith Sykes, MPhil Anders Johansson, PhD Useful Simple Projects University of Bristol ABSTRACT There is a lack of consensus about whether researchers and practitioners should use a deterministic single value or probabilistic distribution of values for each input when modelling the life cycle of a building. This study produces a direct comparison of the two approaches by modelling the nonoperational life cycle energy of three buildings deterministically and probabilistically to explore whether the two approaches produce different conclusions. A detailed method describes the model - its structure, formulae and the best-case, typical and worst-case inputs - with supporting references. This detail provides a thorough explanation of why probabilistic modelling suggests that non-operational energy could be 28-44% lower or 48-283% higher than the original values and why a significant shift in the distribution of non-operational life cycle energy is possible. When used deterministically, the model suggests non-operational energy use is greatest during the product phase. However, when used probabilistically, the model highlights the risk that short component lives and long distance transport by road can significantly increase non-operational energy during the use, construction and end of life phases. The study discusses how future modelling should address a number of uncertainties so that it is more useful for researchers and practitioners. INTRODUCTION Reducing occupant demands for the operational energy used to heat, cool, light and ventilate buildings is currently the narrow focus of building regulations all over the world. However, there is increasing interest in a better understanding of the ‘non-operational’ environmental impacts of buildings that occur between the extraction of raw materials and their disposal, incineration or recovery. One technique that researchers and practitioners use for this purpose is life cycle assessment (LCA) and there are now emerging standards that adapt the general principles of LCA (detailed in ISO 14040) to buildings. In theory, these standards should ensure greater consistency and fairer comparisons of compliant studies, but they still require users to make a number of important choices. One such choice is whether to use a deterministic single value or probabilistic distribution of values for each input when Bengt Cousins-Jenvey is an engineering doctorate candidate, Pete Walker is a professor and Andrew Shea is a lecturer – all in the Department of Architecture and Civil Engineering, University of Bath, Bath, UK. Judith Sykes is a director of Useful Simple Projects, London, UK. Anders Johansson is a senior lecturer in the Department of Civil Engineering, University of Bristol, Bristol, UK. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 115 modelling the life cycle of a building. Studies can currently adopt either of these approaches and still be compliant with the standard, which reflects the current lack of consensus among researchers and practitioners about whether the two approaches are equally valid. A deterministic approach to LCA and specifically life cycle energy is currently the norm. This is indicated by the precise results discussed by a 2010 review of building life cycle energy (Ramesh, Prakash, & Shukla, 2010), but also by the fact that the latest LCA software for practitioners only allows users to input a single value for multiple assumptions. However, a probabilistic approach to LCA also has its advocates that argue a deterministic approach is unrealistically precise (Fawcett, Hughes, Krieg, Albrecht, & Vennström, 2012). They cite how life cycle costing (LCC) uses probabilistic (or ‘stochastic’) models (which is evidenced by (Korpi & Ala-Risku, 2008)), but this does not seem to have influenced the approach to LCAs of buildings. The lack of consensus among life cycle assessors suggest that a direct comparison of the deterministic and probabilistic approaches to building life cycle assessment modelling would also be a useful contribution to knowledge and would have numerous precedents. Others have undertaken many deterministic and probabilistic modelling comparisons for a variety of phenomena. Aim The aim of this study is to undertake a direct comparison of the deterministic and probabilistic approaches to building life cycle modelling by probabilistically modelling three building LCAs that were originally modelled deterministically. The focus is non-operational life cycle energy (primary) as it is consistently and comprehensively dealt with by existing building LCAs (Khasreen, Banfill, & Menzies, 2009). However, the following method could apply to other environmental impacts too. METHOD The method comprised four steps. The first step was to design a comprehensive, but concise modelling process that was compatible with the relevant building LCA guidance as well as both deterministic and probabilistic approaches. The second step was to select three deterministic building LCAs and reproduce their results, thereby establishing confidence in the model as well as clarifying the original values used for each model input. The third step was to select reasonable best-case, typical and worst-case values as an approximate probability distribution for each model input. The fourth and final step was to compare the one result - produced with the set of original values - of the deterministic approach and the three results - produced with the sets of best-case, typical and worst-case values - of the probabilistic approach. Designing the model - its structure, formulae and scope The model is process-based and can accept a three-point estimation of the probability distribution for each input. It breaks down the building life cycle into phases and modules according to the standard EN 15978, and the building fabric into components based on the New Rules of Measurement (NRM) (RICS, 2012). The way it calculates non-operational life cycle energy is also compatible with EN 15978 guidance, but several other sources were also consulted to understand emerging conventions not currently covered by the standard (Adalberth, 1997; EeBGuide, 2012; Moncaster & Symons, 2013). More specifically, the model generates the non-operational building life cycle energy of a building from the sum of the product phase energy, construction phase energy, use phase energy and end of life energy for the n different materials in the building: (1) The model defines each of the terms in Eq. (1) as follows: 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 116 (2) (3) (4) (5) (6) (7) Note that it is crucial to assign materials to different components so that it is possible to establish the quantity of recurring material from the quantity of initial material using Eq. (7). Note also that Eq. (7) rounds up the number of times each component is maintained, repaired, replaced and refurbished during the life of the building (which EN 15978 refers to as modules B2-B5). Eq. (4) enables a comprehensive assessment of the non-operational energy used during all four of these processes between obtaining the recurring materials and treating them at the end of life. The obvious omission from Eq. (4) is the use of machinery during installing and uninstalling, which requires further study. Selecting deterministic studies and reproducing them with the model Analysis of the building life cycle energy studies summarised by two recent review articles (Ramesh et al., 2010; Yung, Lam, & Yu, 2013), highlighted 20 cases for closer examination. This examination revealed that it was not possible to reproduce 15 of the cases, because they did not provide a detailed breakdown of material quantities. Of the remaining five, three offered the most comprehensive scope and explicit assumptions for reproduction by this study. Table 1 briefly summarises basic information about these three chosen cases and highlights the different building types and locations. Table 1. Basic Information about the Three Cases No. 1 2 3 * Reference (Kofoworola & Gheewala, 2009) (Scheuer, Keoleian, & Reppe, 2003) (Gustavsson, Joelsson, & Sathre, 2010) Type Office University Apartments Location Thailand USA Sweden Area (m2) 60,000 7,300 3,374 Life (years)* 50 75 50 Case 3 also modelled an ‘actual’ life of 100 years, but 50 years is a more likely ‘design’ life Table 2. Life Cycle Phases and BS 15978 Modules Considered by the Three Cases Phase Product Construction Construction Use End of Life End of Life End of Life Beyond * EN 15978 Modules A1-A3 A4 A5 B2-B5 C1 C2 C3&C4 D Case 1 Yes Yes Yes B4-Partial* Yes Yes Partial Partial Case 2 Yes Yes Yes B4-Partial* Yes Yes Partial No Case 3 Yes No Yes No Partial No Partial Partial Transport from factory gate to site and site to waste treatment facility not considered 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 117 Table 3. Building Components and NRM Elements Considered by the Three Cases Components Envelope External Works Internal Finishes Fixtures/Fittings Services Substructure Superstructure Temporary Works NRM Elements 2.5 & 2.6 8 3 4 5 1 2.1,2.2 & 2.3.1 - Case 1 Yes Partial No Partial No Yes Yes Unclear Case 2 Yes No Yes Yes Yes Yes Yes Unclear Case 3 Yes No Yes Yes No Yes Yes Unclear Table 2 and Table 3 provide more detail about the scope of each study in relation to EN 15978 modules and NRM components. Case 1 and Case 2 combine the end of life modules C3-C4 (processing and disposing of waste) with module D (benefits beyond the life of the building). This means that they consider the energy saved by recycling or incinerating materials as a reduction in the life cycle energy of the building, which is actually contrary to EN 15978. In order to avoid confusion, this study does not model modules C3, C4 or D. Consequently, Cdis,m is zero in Eq. (4) and Eq. (5). Table 4 and Table 5 identify and compare the original values used by the three cases. Table 4 summarises values that apply to all of the n materials. Table 5 summarises a few of the values for the production phase coefficient Cpro of specific materials used by the three cases. The table also compares them with typical values from the University of Bath (UoB) Inventory of Carbon and Energy (ICE) (Hammond & Jones, 2011) used subsequently for the probabilistic modelling. Table 4. Original Values the Three Cases Used Term Lbuild Ccon Cdec Ctra Dcon Dend W * Full Description Assumed life of the building Coefficient for installing energy Coefficient for uninstalling energy Coefficient for transporting energy Distance from factory to site Distance from site to waste facility Waste generated during construction Unit Years MJ/m2 MJ/m2 MJ/kg/km km km % Case 1 50 300* 52* 0.0027 50* 50 6* Case 2 75 337 350 0.0023* 75 50* 5* Case 3 50 288 36 5* Inferred approximate or average values rather than values explicitly stated in the case studies Table 5. Original Production Phase Coefficients Cpro (in MJ/kg) from the Three Cases and ICE Material Aluminium Brick Concrete Glass Sand Steel Wood * Case 1 216.50 1.86 1.30 17.10 0.10 11.10 - 22.10 - Case 2 207.00 2.70 Unclear 6.80 0.60 12.30 - 30.60 10.80 Case 3* 1.45 14.84 29.43 7.28 ICE Typical Value 155.00 3.00 0.75 15.00 0.10 9.40 - 35.40 7.11 ICE Range (%) +/- 20 +/- 30 +/- 30 +/- 30 Unclear +/- 30 Unclear Inferred from end-use energy in kWh/tonne assuming an energy conversion efficiency of 0.5 Selecting the sets of best-case, typical and worst-case values A literature review produced the best, typical and worst values in Table 6 used for all of the n materials in all three cases. (Although the model could accept a specific value for each material, it would then become hard to document it concisely.) Table 7 shows the best, typical and worst values 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 118 used for calculating the replacement period Yp for each component c. The values used for Cpro were specific for each material m and are the “typical” values from the ICE (Hammond & Jones, 2011) unless a more specific value was obviously appropriate. A percentage modifier Ppro in Eq. (2) enabled modelling of the general range associated with ICE values. Table 6. Values this Study Uses Term Ccon Cdec Ctra Dcon Dend Mp Ppro W Unit 2 MJ/m MJ/m2 MJ/tonne km km km % % % Best Value 20 10 0.0004 50 50 100 70 0 Typical Value 300 300 0.0010 300 300 100 100 5 Worst Value 750 900 0.0020 500 500 100 130 10 Supporting Reference(s) (Gustavsson et al., 2010; Scheuer et al., 2003) (ATHENA, 1997; Gustavsson et al., 2010) (IPCC, 1996) (EeBGuide, 2012) (EeBGuide, 2012) (Hammond & Jones, 2011) (Adalberth, 1997) Table 7. Values this Study Uses for the Replacement Periods Yp (in years) of each Component c (Supporting Reference (BCIS, 2006)) Component Envelope External Works Internal Finishes Fixtures/Fittings Services Substructure Superstructure NRM Elements 2.5 & 2.6 8 3 4 5 1 2.1,2.2 & 2.3.1 Best Value 50 40 30 30 30 90 90 Typical Value 40 30 20 20 20 60 60 Worst Value 30 20 10 10 10 30 30 With values for all inputs identified, it was possible to model the three cases with the original, best, typical and worst values. Note that all of the modelling was faithful to the scopes of the three cases (Table 2 and Table 3) and did not attempt to standardise them. RESULTS Table 8 summarises the results of reproducing the three cases. Figure 1 compares modelling the non-operational life cycle energy of the three cases with the original, best, typical and worst values. Table 8. Deterministic Results of the Three Cases and their Reproducibility with the Model Case Phase 1 1 1 1 1 1 1 2 2 2 2 2 2 2 Product Construction Construction Use End of Life End of Life End of Life Product Construction Construction Use End of Life End of Life End of Life EN 15978 Modules A1-A3 A4 A5 B2-B5 C1 C2 C3&C4 A1-A3 A4 A5 B2-B5 C1 C2 C3&C4 Original Result (TJ) 375.0 Unclear Unclear Unclear 3.1 6.8 Unclear 39.3 2.5 2.4 6.9 Unclear Unclear - 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Reproduced Result (TJ) 375.8 21.4 18.0 60.8 3.1 21.4 40.7 2.5 2.5 5.0 2.6 1.6 - Model Simplifications and Plausible Explanations for Discrepancies Modelled using one value for W See the end of the discussion section Approximated Mini Modelled using one value for Ctra Rounded up 2.46 TJ to 2.5 TJ Approximated Mrec Modelled using one value for Ctra 119 2 2 3 3 3 3 3 3 3 Other Other Product Construction Construction Use End of Life End of Life End of Life C1&C2 A1-A3&B4 A1-A3 A4 A5 B2-B5 C1 C2 C3&C4 4.0 46.3 10.9 1.0 0.1 - 4.2 45.7 15.0 1.0 0.1 - Ignored materials under 1 tonne See the end of the discussion section Out of original scope Out of original scope Out of original scope Out of original scope 700 600 MJ/m2/year 500 End of Life Use Construction Product 400 300 200 100 0 O B T W Case 1 Figure 1 O B T Case 2 W O B T W Case 3 A non-operational life cycle energy comparison of the three cases (1-3) modelled using the original (O), best (B), typical (T) and worst (W) values DISCUSSION Figure 1 shows the difference between the one result - produced with the original values - of a deterministic approach and the three results - produced with the best, typical and worst values - of a probabilistic approach. Modelling with the original values appears optimistic in the context of the results produced with the other three sets of values and it is clear that the best and worst values highlight a significant opportunity and risk associated with the minimum and maximum quantities of nonoperational energy. Compared to the original value results, the best value results are 28-44% lower while the worst value results are 277-283% higher for Case 1 and Case 2 and 48% for Case 3. The use phase non-operational energy for Case 1 and Case 2 notably increases dramatically with the worst values (by a factor of 13 and 23 respectively). The worst value results suggest that the total non-operational energy could represent as much as approximately 8% and 17% of the total life cycle energy of Case 2 and Case 1 respectively (based on actual operational energy data) and 12-22% of Case 3 (based on modelled operational energy scenarios and even though the use phase is out of scope). Based upon the original values, the discussions and conclusions of Case 1 and Case 2 logically focussed on energy use during the operation and product phases while Case 3 focussed specifically on energy sources during the operation and construction phases. However, it is noticeable that all three cases suggested only a few practical actions to reduce non-operational energy use. Case 1 and Case 2 mentioned increasing recycled content, while Case 3 concentrated on substituting biomass for fossil fuel. Only Case 2 highlighted the possibility that frequent renovations could increase and shift the distribution of life cycle energy. The probabilistic modelling undertaken here suggests that practitioners should also consider the risk of short component replacement periods during the use phase and long distance transport by road during the construction and end of life phases. Probabilistic modelling is clearly useful to highlight the model inputs that produce significant uncertainty, but it is important that future research 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 120 considers how an understanding of these sensitive model inputs translates into robust design decisions. Close collaboration between researchers and practitioners could be necessary. Probabilistic modelling is obviously less useful when there is high certainty about the sensitive inputs. However, it is important to note that even though Case 1, 2 and 3 were all assessed after they were constructed, they mention numerous challenges of obtaining every detail about a building and how it was built. This suggests assessors should be cautious not to overestimate certainty. They could find probabilistic modelling useful not just for prospective, but also retrospective assessment. Exploring uncertainties with future studies The choice of scope in all three cases appears more important with probabilistic modelling than with deterministic modelling. Case 3 ignored the use phase completely while Case 1 and 2 ignored modules B2, B3 and B5 (the processes of maintaining, repairing and refurbishing) as well as all transportation. The results suggest that building LCAs should aim to be complete and imprecise rather than incomplete and precise. There is an argument that the standard EN 15978 should mandate a minimum scope of life cycle phases, modules and building components to reduce the variations in Table 2 and 3. While adopting this comprehensive scope, future building LCAs should also model different building types, constructed from different materials with context specific model inputs. These LCAs should continue to question transport distances from factory gate to site D, the building life Ybuild, all use phase process periods Yp and material replacement quantities Mp. They could test the logic of the different versions of Eq. (7) found in the literature and consider if and how different possible futures (changing practices, regulations, technologies) could affect the values in Table 6. This will require a more complex probabilistic modelling process and an appropriate approach to sensitivity analysis. Learning from reproducing the three cases A challenge of adopting a more complex modelling process is to still provide the information required to enable others to examine, reproduce, critique and improve it. (Although researchers could consider publishing whole models not just their research data.) The method of this study contains a number of tables that others could find useful to concisely provide the detailed description necessary, but they must also include (or reference) material quantities by component (see the matrix included in Case 3). Without this information, the reasons for the small discrepancies in Table 8 are unclear, but it is likely that they result from the original studies using specific values for each material rather than one average value for all materials unless otherwise stated. The significant discrepancies that resulted when modelling Case 1 Module C2 (transport from site to treatment facility) and Case 3 Modules A1-A3 (the whole product phase) are easier to explain. Case 1 originally assumed that the building generated only 0.845 t/m2 of waste, but the model assumes the transportation of all material to a waste treatment facility. Case 3 was only explicit about the production phase energy coefficients Cpro used for a few materials, so the model had to adopt appropriate typical values from the ICE instead. CONCLUSION A comprehensive, but concise building life cycle model was introduced and used to reproduce the non-operational energy of three buildings deterministically (based on the original information or assumptions) and probabilistically based upon a literature review of best, likely and worst values. When used deterministically, the model suggests non-operational energy use is most significant during the product phase. However, when used probabilistically, the model highlights the risk that short component lives and long distance transport by road can significantly increase non-operational energy during the use, construction and end of life phases. The formulae and literature review of plausible ranges for each model input presented here should facilitate probabilistic modelling in the future. They provide the basis for others to model different building types, constructed from different materials with context specific probability distributions for each model input that are based upon a better understanding of what happens in reality. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 121 ACKNOWLEDGMENTS The authors gratefully acknowledge the support of the EPSRC funded Industrial Doctorate Centre in Systems (Grant EP/G037353/1), Useful Simple Projects Ltd and Expedition Engineering Ltd. They would also like to thank Dr Pete Winslow and Clement Thirion at Expedition Engineering for their contributions to early versions of this study. NOMENCLATURE A C D E = = = = floor area of the building conversion factor distance non-operational energy M P W Y = = = = = = = = m n p pro rec un use = = = = quantity of material production phase ‘modifier’ waste life or period of time Subscripts build c con des dis end in ini building component construction phase design disposal end of life phase installing initial = = = = = = = material number of materials use phase process product phase recurring uninstalling use phase REFERENCES Adalberth, K. (1997). Energy use during the life cycle of buildings: a method. Building and Environment, 32(4), 317–320. ATHENA. (1997). Demolition Energy Analysis of Office Building Structural Systems. BCIS. (2006). BMI Life Expectancy of Building Components. EeBGuide. (2012). Guidance Document Part B: Buildings. Fawcett, W., Hughes, M., Krieg, H., Albrecht, S., & Vennström, A. (2012). Flexible strategies for longterm sustainability under uncertainty. Building Research & Information, 40(5), 545–557. Gustavsson, L., Joelsson, A., & Sathre, R. (2010). Life cycle primary energy use and carbon emission of an eight-storey wood-framed apartment building. Energy and Buildings, 42(2), 230–242. Hammond, G. P., & Jones, C. I. (2011). Inventory of Carbon and Energy (ICE), V2.0. Retrieved January 10, 2014, from www.circularecology.com/ice-database.html IPCC. (1996). Second Assessment Report (p. 693). Khasreen, M. M., Banfill, P. F. G., & Menzies, G. F. (2009). Life-cycle assessment and the environmental impact of buildings: a review. Sustainability, 1(3), 674–701. Kofoworola, O. F., & Gheewala, S. H. (2009). Life cycle energy assessment of a typical office building in Thailand. Energy and Buildings, 41(10), 1076–1083. Korpi, E., & Ala-Risku, T. (2008). Life cycle costing: a review of published case studies. Managerial Auditing Journal, 23(3), 240–261. Moncaster, A. M., & Symons, K. E. (2013). A method and tool for “cradle to grave” embodied carbon and energy impacts of UK buildings in compliance with the new TC350 standards. Energy and Buildings, 66, 514–523. Ramesh, T., Prakash, R., & Shukla, K. K. (2010). Life cycle energy analysis of buildings: an overview. Energy and Buildings, 42(10), 1592–1600. RICS. (2012). Methodology to calculate embodied carbon of materials. Scheuer, C., Keoleian, G. A., & Reppe, P. (2003). Life cycle energy and environmental performance of a new university building: modeling challenges and design implications. Energy and Buildings, 35(10), 1049–1064. Yung, P., Lam, K. C., & Yu, C. (2013). An audit of life cycle energy analyses of buildings. Habitat International, 39, 43–54. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 122 CBD greening and Air Temperature Variation in Melbourne Elmira Jamei, PhD [Deakin University] Harsh Sachdeva, MArch [Deakin University] Priyadarsini Rajagopalan, Senior Lecturer [Deakin University] ABSTR ACT Melbourne, the second most populous city in Australia, is growing rapidly. To accommodate this growth, CBD area is undergoing dense urban development and consequently the microclimate of the city is affected. Understanding how greenery can influence the air temperature is very important in urban planning. This study presents a simulation approach to examine the impact of CBD greening on the air temperature variation, during a summer day in Melbourne. The numerical simulation system, ENVI-met was used to examine the impact of vegetation on the air temperature in CBD area, under various scenarios. Three scenarios were applied; without any vegetation, with the existing trees (2%), and enhanced number of the trees (6%). The simulation results showed significant lower air temperatures in both greening scenarios compared to the base case scenario without any vegetation. Increasing 4% tree coverage in the study area, led to 0.2 C° reduction in the air temperature. The study also found that the maximum cooling effect, occurs at mid afternoon. The outcomes of this study could be used to assist urban planners in developing policy suggestions for improving Melbourne’s microclimate and offsetting the likely temperature impacts from increasing urban densities. INTRODUCTION Consistent population growth and urban development have triggered high demand for built environment and migration of people from rural to urban areas. According to Australian bureau of statistics, population of Melbourne has been increased by 9.7% from 2006 to 2011 (Australian Bureau of Statistics 2000). To accommodate this population growth, natural landscapes and vegetated areas have been replaced by impervious surfaces of buildings and pavements, leading to an alteration in the radiative, thermal and aerodynamic characteristics of the urban surfaces (Morris, Simmonds & Plummer 2001). One of the most significant consequences of the urbanization, is the temperature difference between the urban and rural areas, known as “urban heat island” (Oke 1984). The phenomenon, is mainly a result of high thermal capacity and heat storage of urban surfaces, anthropogenic heat, caused by human activities and reduced rate of evapotranspiration in urban areas (Oke 1988). The temperature rise in cities might be beneficial during winters, but it increases the energy demand and health risks during the summer (Yu & Hien 2006). Melbourne, with a population over 3.6 million, features UHI consistently throughout the year (Morris, Simmonds & Plummer 2001). In 1992, an automobile transect across the city monitored 7.1C° temperature difference between the central business district (CBD) and surrounding suburban areas, with smaller peaks in industrial areas and the medium-density terrace housing in the inner northern suburbs (Torok et al. 2001). According to CSIRO (2011), the average daytime air temperature in Melbourne tends to rise from15.7 C° to 18.5 C° by 2070. Consequently, the number of the days with maximum air temperature will be increased. A study by Lynch et al. (2011) states that, the mortality rate is likely to be doubled by the latter part of the current century. Four days heat wave in Melbourne was resulted in 374 excess deaths in January 2009. The mortality rate is often maximum among among elderly and people with respiratory diseases (Victorian Department of Human Services 2010). Therefore, the process of urbanization and temperature rise in Melbourne city, presents a clear issue for public health, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 123 sustainability of urban environments and thermal condition of the city, particularly during hot seasons. Microclimatic benefits of vegetation have been extensively investigated in previous researches (Avissar 1996; Huang et al. 1987; Jauregui 1991; Oke et al. 1989; Shashua-Bar & Hoffman 2000). Vegetation not only provides pedestrians with pleasurable visual scenes, but also provides shading, improves air quality, reduces the noise levels and contributes to the mitigation of the urban heat island effect (Dimoudi & Nikolopoulou 2003). Over 25-50% mitigation of heat island intensity can be achieved through greening strategies (Rowntree, Sanders & Stevens 1982). Various greening strategies have been used to reduce the air temperature and the level of air and noise pollution, such as, green roofs, urban parks, trees, shrubs and grass (Oliveira, Andrade & Vaz 2011). The cooling effect of vegetation occurs through the process of shading, evapotranspiration and changing the wind pattern. The average cooling effect of vegetation is between 1 to 4.7 C°, that can be extended by 100 to 1000 meter radius around the vegetated area. The cooling effect also highly depends on the available water for irrigation (Schmidt 2006). Over the last decade, several studies have been conducted in various climatic condition, to investigate the detailed relationship between different greening scenarios and urban microclimate (Lin, Matzarakis & Hwang 2010; Ng et al. 2012; Shashua‐Bar et al. 2010; Wong et al. 2007), but the number of the number of these studies in Australia is lacking. Therefore, this study aims to examine the effect of different greening scenarios on the air temperature variation in Melbourne, by using a numerical modeling system, ENVI-met. Many methods have been applied to investigate the effect of vegetation on microclimate, such as numerical modelling (Avissar 1996; Pearlmutter, Krüger & Berliner 2009; Spronken-Smith & Oke 1999), empirical analysis, on site measurement (mobile traverse, weather station data) (Jonsson 2004; Sani 1987; Upmanis, Eliasson & Lindqvist 1998) and satellite images (Ooka 2007). But, numerical modeling has become more popular than on-site field measurements during recent years. Because researchers have greater control over modeling in regards to the time and resources (Arnfield 2003). Additionally, numerical models are capable of coping with the complexities and non-linearities of urban structures. Some recent studies used three-dimensional numerical model, ENVI-met to simulate the effect of vegetation on microclimate (Ali-Toudert & Mayer 2007; Fahmy & Sharples 2009; Fahmy, Sharples & Eltrapolsi 2009; Spangenberg et al. 2008; Yu & Hien 2006). ENVI-met, simulates the microclimatic changes within urban environments in a high spatial and temporal resolution (Bruse & Fleer 1998). It can also calculates all important meteorological parameters, such as the solar radiation, air temperature, relative humidity, wind speed, as well as the mean radiant temperature. Buildings, overhangs, galleries and setbacks can be illustrated via ENVI-met (Hedquist et al. 2009). This study, uses numerical simulation model, ENVI-met (Bruse M 2011a) to generate the air temperature data for the central business district area, and to examine the effect of different greening scenarios on the air temperature variation in a typical urban environment in Melbourne’s CBD. Three scenarios are examined; a base case scenario without any vegetation, scenario “1” with existing trees in the site which cover 2% of the study area and scenario “2” with uniformly enhanced trees, placed at the fixed distance from each other, which cover 6% of the study area. Table 2 lists the detailed characteristics of each scenario. METHODOLOGY Melbourne, is the capital and most populous city in the state of Victoria, and is the second most populous city in Australia. Geographical coordinates of Melbourne are (37°49′S, 144°53′E) and according to Köppen climate classification, Melbourne has a moderate oceanic climate. The city has been well reputed for its unstable weather condition (Sturman & Tapper 2006). Melbourne summers are notable for the occasional days of extreme heat (Bureau of Meteorology 2009). The highest temperature recorded in Melbourne city was 46.4 °C (115.5 °F), on 7 February 2009. To investigate the impact of different greening scenarios on the air temperature, a typical urban environment was selected in the central business district area in Melbourne. The Hoddle Grid with the dimensions of 1.61 by 0.80 km, forms the center of Melbourne's central business district. Most of the buildings in this area are 8 to 12 storey and streets have 15 and 30 meter width. Figure 1, shows the boundary of the selected site. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 124 Most of the studies using ENVI-met, conducted comprehensive field measurements to validate the outputs of the software with on-site measurement. Thapar and Yannas (Thapar & Yannas 2007), used field measurements to validate the findings of ENVI-met on the air temperature and wind variation around specific urban forms. Hedquist et al. (2009), used the software along with CFD and field measurements to report the temperature variation in high density areas. Boundary of the study area in central business district of Melbourne, Australia, Red Figure 1. color circles indicate the spot points for field measurements (Left) 3D view of the study area (Right) In this study, three points were selected in the study area, to verify the results of the simulation with the on-site measurements. Hobo data loggers were used to monitor and compare the air temperature variation at the selected points. The location of each point is shown in Figure1. An ENVI-met model was first created, according to the exact urban geometry and vegetation coverage of the site. Climatic data, such as the initial air temperature, relative humidity, cloud cover and wind speed (19 December 2013) were initially given to the model, according to the data obtained from the “Australian Bureau of Meteorology”. The model was then run for 24 hours, starting from 6 am and ending at 6 am the following day. The simulation results for the air temperature (Ta) were extracted, plotted and compared with the air temperature measured at the site, for four different times of the day; 9 am, 12 noon, 3 pm and 6 pm. These periods of the time were selected, because they cover the times when the temperature is minimum, when it reaches to its maximum and when the temperature begins to drop. Comparison between the measured values and simulated data was made by conducting a regression analysis. Figure 2, shows a reasonable agreement between the measured and simulated data, with R value of 0.8. The usefulness of ENVI-met in predicting the air temperature variation, in Melbourne’s urban environment during summertime was therefore confirmed. The verified simulation settings are given in Table 1. These settings were then applied in simulating different scenarios. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 125 Comparison of the averaged measured and simulated air temperatures (Ta) in 19 Figure 2 December 2013 Time 19 December 2013 Initial Temperature 300.5 K Table 1. Verified ENVI-met simulation settings Start Relative Wind Wind Time Humidity Direction Speed at at 2 m (%) 10 Meter level (m/s) 6 :00 am 49% North 15 Albedo of the Roofs Albedo of the Walls 0.3 0.2 One approach in urban climate modeling, is simplifying the complex urban structures into generic urban layouts, in order to understand the effect of altering a certain parameter in a system (Robinson et al. 2007). This method is also applicable in modelling the global climate, economic or ecological systems, as well as in biology and health studies. Therefore, to understand the impact of trees on the air temperature variation, a generic layout of the selected site was created. Climatic data and geographical features of the site, such as the average building height (30 meter) and the widths of the streets (15, 30) were given to the model. Mature 20m dense distinct crown trees were used in the model. A snapshot of the generic urban layout of the study area was created by ENVI-met in Figure 3. The detailed inputs for configuration file is also presented. Figure 3 (Left) Aerial view of the study area (right) Snapshot of the ENVI-met model 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 126 Ariaepel % ---- Basic Configuration File for ENVI-met Version 3 --------------% ---- MAIN-DATA Block ------------------------------------------------Name for Simulation (Text): = Base Case Scenario Input file Model Area =C:\Users\User\Desktop\paper 2\Base Case Scenario.in Filebase name for Output (Text): =model 1 base Output Directory: =C:\Users\User\Desktop\paper 2 Start Simulation at Day (DD.MM.YYYY): =19.12.2013 Start Simulation at Time (HH:MM:SS): =06:00:00 Total Simulation Time in Hours: =24.00 Save Model State each? Min =60 Wind Speed in 10 m ab. Ground [m/s] =15 Wind Direction (0: N...90: E...180: S...270: W...) =0 Roughness Length z0 at Reference Point =0.1 Initial Temperature Atmosphere [K] =300.5 Specific Humidity in 2500 m [g Water/kg air] =7 Relative Humidity in 2m [%] =49 Database Plants =C:\ENVImet31\sys.basedata\Plants.datFirst numbered item Base case scenario, Without any tree (0% tree coverage) Table 2. Different scenarios Scenario 1, Existing trees (2% tree coverage) Scenario 2, (6% tree coverage) RESULTS AND DISCUSSIONS A typical summer day, 19 December 2013 was simulated, using verified settings of the model (Table 1). The simulations were run on a core 2 quad processor 8 and GB of RAM. Each run took about 7 to 8 days. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 127 Figure 4 Variation of the air temperature in different scenarios The results of the simulations for 9 am, 12 noon, 3 pm and 6 pm were extracted and analyzed. Figure 4, shows the variation of the air temperature in each scenario and Table 3 illustrates the LEONARDO images of the outputs for various scenarios, in different times of the day. Some preliminary findings can be derived from Figure 4 and Table 3. The main finding is that, both greening scenarios (existing trees and uniform enhancement of the trees) significantly modify the air temperature. The maximum level of modification occurs at 12 noon. Compared with the base case scenario without any tree, scenario 1, with the existing trees in the site contribute to 4.8 °C, 7.5 °C, 6.1 °C and 5.5 °C reduction in the air temperature at 9am, 12 noon, 3 pm and 6 pm respectively. The existing tree case (scenario1), can provide more shade, therefore, from the LEONARDO images, it can be seen that, under the trees, the reduction of the air temperature is more intense. Uniform enhancement of the tree (scenario 2) also decreases the air temperature, compared to the base case scenario. As Figure 4 shows, 4.9 °C, 7.8°C, 6.5 °C and 5.9 °C temperature difference is recorded between the base case scenario and scenario Table 3. The spatial distribution of the air temperature in different scenarios at 9 am, 12 noon, 3pm and 6 pm (From top to bottom) Base case scenario, Without Scenario 1, Existing trees (2% Scenario2, Uniform greening any tree (0% tree coverage) tree coverage) (6% tree coverage) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 128 These results are in accordance with the findings of similar studies on the impact of city greening on the air temperature (Bowler et al. 2010; Yu & Hien 2006). In regards to the air temperature reduction caused by greening scenarios, the result of this study is comparable to Taipei study, which showed 0.81K temperature reduction caused by urban parks (Chang, Li & Chang 2007) and Singapore study, which monitored 1.3 K decrease in the air temperature, due to urban greening (Yu & Hien 2006). A general conclusion can be achieved that, urban greening in the forms of trees can provide cooling effect to the urban environment. LEONARDO images in Table 3, show that, the air temperature in deep canyons is slightly lower than the air temperature in shallow canyons. More shading in deep canyons and less exposure to the direct sun are the plausible explanation of slightly lowered ambient temperature in narrow canyons for most of the locations. Explanation for the smaller temperature difference between base case scenario and scenario 1 compared to the base case scenario and scenario2, relates to the lower level of tree coverage in scenario 1. Because in scenario1, the existing trees are quite sparse, in comparison with scenario 2 with uniform trees planted through the site. As Figure 5 indicates, the maximum temperature difference between the base case scenario and greening scenarios was monitored 12 noon. Figure 5 Temperature difference between greening scenarios at 9am, 12 noon, 3 pm and 6 pm The findings of this study serve as a proof of concept to show how numerical microclimatic modeling can help to incorporate the urban greening schemes into CBD planning, urban development and visualize the potential cooling benefits of various greening and design scenarios. This study was limited to the impact of urban greening in the form of trees on the air temperature. However , the study 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 129 aims to include the impact of urban parks, Australian native trees and leaf area index will be studied on the air temperature variation in Melbourne. Furthermore, the effect of building layouts, street orientation and urban layouts on the air temperature will be also addressed. CONCLUSION This paper presented a preliminary study on the cooling effect of street trees in the CBD of Melbourne. Numerical modeling system, ENVI-met was verified through conducting field measurements. Verified settings were applied to the model, to simulate the air temperature variation in a generic urban layout of the CBD. Three scenarios were simulated; a base case scenario without any vegetation, scenario “1” with the existing trees in the site (2% tree coverage) and scenario “2” with uniform tree enhancement (6% tree coverage). It is found that both greening scenarios, contribute to the significant air temperature reduction. The maximum cooling effect of trees was detected at 12 noon. The study also revealed that 4% increase in tree coverage would lead to 0.2 °C reduction in the average air temperature. This study demonstrates how the simulation approach can help urban planners to better understand, visualize and analyze the potential cooling effect of urban greening and design related strategies. REFERENCES Ali‐Toudert, F., & Mayer, H. (2006). Numerical study on the effects of aspect ratio and orientation of an urban street canyon on outdoor thermal comfort in hot and dry climate. Building and Environment, 41(2), 94‐108. Ali‐Toudert, F., & Mayer, H. (2007). Effects of asymmetry, galleries, overhanging facades and vegetation on thermal comfort in urban street canyons. Solar Energy, 81(6), 742‐754. doi: DOI 10.1016/j.solener.2006.10.007 Arnfield, A. J. (2003). Two decades of urban climate research: a review of turbulence, exchanges of energy and water, and the urban heat island. International Journal of Climatology, 23(1), 1‐26. Australian Bureau of Statistics. (2000). Australian social trends: Australian Bureau of Statistics. Avissar, R. (1996). Potential effects of vegetation on the urban thermal environment. Atmospheric Environment, 30(3), 437‐448. Bowler, D. E., Buyung‐Ali, L., Knight, T. M., & Pullin, A. S. (2010). Urban greening to cool towns and cities: A systematic review of the empirical evidence. Landscape and Urban Planning, 97(3), 147‐155. Bruse M. (2011a). About ENVI‐met ; General idea. Bruse M. ((2011h)). ENVI‐met. Overview of model layout. doi: www.envi‐met.com Bruse, M., & Fleer, H. (1998). Simulating surface–plant–air interactions inside urban environments with a three dimensional numerical model. Environmental Modelling & Software, 13(3), 373‐384. Bureau of Meteorology. (2009). Climate statistics for Australian locations. Chang, C.‐R., Li, M.‐H., & Chang, S.‐D. (2007). A preliminary study on the local cool‐island intensity of Taipei city parks. Landscape and Urban Planning, 80(4), 386‐395. Dimoudi, A., & Nikolopoulou, M. (2003). Vegetation in the urban environment: microclimatic analysis and benefits. Energy and Buildings, 35(1), 69‐76. Fahmy, M., & Sharples, S. (2009). On the development of an urban passive thermal comfort system in Cairo, Egypt. Building and environment, 44(9), 1907‐1916. Fahmy, M., Sharples, S., & Eltrapolsi, A. (2009). Dual stage simulations to study the microclimatic effects of trees on thermal comfort in a residential building, Cairo, Egypt. Paper presented at the 11th International IBPSA Conference. Glasgow, Scotland. Hedquist, B., Di Sabatino, S., Fernando, H., Leo, L., & Brazel, A. (2009). Results from the Phoenix Arizona Urban Heat Island Experiment. Paper presented at the The seventh International Conference on Urban Climate. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 130 Huang, Y., Akbari, H., Taha, H., & Rosenfeld, A. H. (1987). The potential of vegetation in reducing summer cooling loads in residential buildings. Journal of climate and Applied Meteorology, 26(9), 1103‐1116. Jauregui, E. (1991). Influence of a large urban park on temperature and convective precipitation in a tropical city. Energy and Buildings, 15(3), 457‐463. Jonsson, P. (2004). Vegetation as an urban climate control in the subtropical city of Gaborone, Botswana. International Journal of Climatology, 24(10), 1307‐1322. Lin, T.‐P., Matzarakis, A., & Hwang, R.‐L. (2010). Shading effect on long‐term outdoor thermal comfort. Building and environment, 45(1), 213‐221. Lynch, M., Kirkwood, R., Mitchell, A., Duignan, P., & Arnould, J. P. (2011). Prevalence and significance of an alopecia syndrome in Australian fur seals (Arctocephalus pusillus doriferus). Journal of mammalogy, 92(2), 342‐351. Morris, C., Simmonds, I., & Plummer, N. (2001). Quantification of the influences of wind and cloud on the nocturnal urban heat island of a large city. Journal of Applied Meteorology, 40(2), 169‐182. Ng, E., Chen, L., Wang, Y., & Yuan, C. (2012). A study on the cooling effects of greening in a high‐ density city: an experience from Hong Kong. Building and Environment, 47, 256‐271. Oke, T. (1984). Towards a prescription for the greater use of climatic principles in settlement planning. Energy and Buildings, 7(1), 1‐10. Oke, T. (1988). Street design and urban canopy layer climate. Energy and Buildings, 11(1), 103‐113. Oke, T. R., Crowther, J., McNaughton, K., Monteith, J., & Gardiner, B. (1989). The micrometeorology of the urban forest [and discussion]. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 324(1223), 335‐349. Oliveira, S., Andrade, H., & Vaz, T. (2011). The cooling effect of green spaces as a contribution to the mitigation of urban heat: A case study in Lisbon. Building and environment, 46(11), 2186‐2194. doi: http://dx.doi.org/10.1016/j.buildenv.2011.04.034 Ooka, R. (2007). Recent development of assessment tools for urban climate and heat‐island investigation especially based on experiences in Japan. International Journal of Climatology, 27(14), 1919‐1930. Pearlmutter, D., Krüger, E., & Berliner, P. (2009). The role of evaporation in the energy balance of an open‐air scaled urban surface. International Journal of Climatology, 29(6), 911‐920. Robinson, D., Campbell, N., Gaiser, W., Kabel, K., Le‐Mouel, A., Morel, N., . . . Stone, A. (2007). SUNtool–a new modelling paradigm for simulating and optimising urban sustainability. Solar Energy, 81(9), 1196‐1211. Rowntree, R. A., Sanders, R. A., & Stevens, J. (1982). Evaluating urban forest structure for modifying microclimate: the Dayton Climate Project [Ohio, city and community vegetation, pollution]. Sani, S. (1987). Urbanization and the atmospheric environment in the low tropics: experiences from the Kelang valley region Malaysia. Schmidt, M. (2006). The contribution of rainwater harvesting against global warming. Technische Universität Berlin, IWA Publishing, London, UK. Shashua‐Bar, L., & Hoffman, M. (2000). Vegetation as a climatic component in the design of an urban street: An empirical model for predicting the cooling effect of urban green areas with trees. Energy and Buildings, 31(3), 221‐235. Shashua‐Bar, L., Potchter, O., Bitan, A., Boltansky, D., & Yaakov, Y. (2010). Microclimate modelling of street tree species effects within the varied urban morphology in the Mediterranean city of Tel Aviv, Israel. International Journal of Climatology, 30(1), 44‐57. Spangenberg, J., Shinzato, P., Johansson, E., & Duarte, D. (2008). Simulation of the influence of vegetation on microclimate and thermal comfort in the city of São Paulo. Revista SBAU, Piracicaba, 3(2), 1‐19. Spronken‐Smith, R., & Oke, T. (1999). Scale modelling of nocturnal cooling in urban parks. Boundary‐Layer Meteorology, 93(2), 287‐312. Sturman, A. P., & Tapper, N. J. (2006). Weather and climate of Australia and New Zealand. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 131 Thapar, H., & Yannas, S. (2007). Microclimate and urban form in Dubai. MSc Dissertation. Environment & Energy Studies Programme, Architectural Association School of Architecture, London. Torok, S. J., Morris, C. J., Skinner, C., & Plummer, N. (2001). Urban heat island features of southeast Australian towns. Australian Meteorological Magazine, 50(1), 1‐13. Upmanis, H., Eliasson, I., & Lindqvist, S. (1998). The influence of green areas on nocturnal temperatures in a high latitude city (Göteborg, Sweden). International Journal of Climatology, 18(6), 681‐700. Victorian Department of Human Services. ( 2010). Wong, N., Kardinal Jusuf, S., Aung La Win, A., Kyaw Thu, H., Syatia Negara, T., & Xuchao, W. (2007). Environmental study of the impact of greenery in an institutional campus in the tropics. Building and environment, 42(8), 2949‐2970. Yu, C., & Hien, W. N. (2006). Thermal benefits of city parks. Energy and Buildings, 38(2), 105‐120. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 132 Session 2A : Passive Design PLEA2014: Day 1, Tuesday, December 16 14:10 - 15:50, Auditorium - Knowledge Consortium of Gujarat Sustainable Habitat for Emerging Economies Futoshi MIYAOKAA Khuplianlam TUNGNUNGB Yuichiro KODAMAC [Nihon University] [Kobe Design University] [Kobe Design University] Email address of corresponding author: tungnung.khuplianlam@yahoo.com Arvind KRISHAND [Principal, CASA] ABSTRACT Affordance of thermal comfort, as a key to sustainable habitats in emerging economies, entails a socio-economically responsible response to the imperatives of local climate, low-energy, and lifestyle changes. Since passive house thermal comfort depends on occupant's lifestyle, low-energy architecture that integrates passive techniques and lifestyles becomes avant-garde. Thermal comfort is the absence of discomfort in the occupants mind because of the body’s interaction with environment parameters: temperature, humidity, and air speed, which are enhanced by ventilation. Since socio-cultural lifestyle changes due to globalization and developments resulted in new notions of comfort and adaptation to heat producing equipments, lifestyle is recognize as essential to low-energy paradigms during operations, and heat loss or gain through appropriate ventilations and storage of heat or cold in high thermal mass envelopes could be beneficial. The case study, Bidani Eco-house in Faridabad by Dr. Arvind Krishan is a haveli1 inspired plan form with appropriate open-spaces, orientation, or geometry, and envelope with low U-value local stones and glass facade, which are expected to reduce heating and cooling load if integrated with lifestyle. Research methods encompasses questionnaire with users, field survey, monitoring of hourly indoor temperature with data loggers, and a series of parametric simulations with Solar Designer ver. 6. The paper discusses passive techniques and lifestyles, through 1) indoor temperature fluctuation without air-conditioning to highlight the effects of ventilation modes, air change rates, and thermal mass on environmental comfort parameters, 2) comparative energy performance analysis of annual (2013) cooling, heating, and lighting load with GRIHA2 benchmark to validate the successful integration of lifestyle and passive design. The house low-energy EPI (Energy Performance Index) of 126MJ/m2/year shows that thermal comfort is affordable with relative low-energy in a rapidly changing cultural expectation of modern life. INTRODUCTION AND BACKGROUND This paper, essentially, reports "Sustainable Habitats for Emerging Economies" from the perspectives of "thermal comfort" afforded by the integration of passive design and lifestyle as a lowenergy solution during operations. Thermal comfort, as a subjective response or state of mind, is primarily influence by the body’s heat exchange with the environment parameters: temperature, humidity, air speed (Olesen & Brager, 2004), and corresponds to a temperature range of 20-30ºC DBT and 30-60% relative humidity in still air. (Govt. of India, Energy. n.d.). Personal parameters: clothing, activity, or metabolic rates are not covered in this report. Economic developments and socio-cultural lifestyle changes due to globalization resulted in new notions of comfort and adaptation to heat producing equipments, high energy use, and subsequent lost of the native habitats milieu. So, lifestyle is recognize as essential to low-energy operations. In India, lighting and household appliances such as: refrigerators, air conditions, water heaters, and ceiling fans accounts for 10% of electricity consumption, 1. Enclosed courtyard in private mansions in India and Pakistan. 2. An acronym for Green Ratings for Integrated Habitat Assessment, developed jointly by TERI and the Ministry of New and Renewable Energy, Government of India. Author A is a Dr. Design and Research Assist., Nihon University, Japan. Author B is a M. Design and Doctorate Candidate at Kobe Design University, Japan. Author C is a Dr. Eng. and Professor at Kobe Design University, Japan. Author D is a Professor and Former Dean, Head Dept. of Architecture, SPA Delhi. Currently Principal Architect, CASA (Center for Architectural Systems Alternatives). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 133 while housing and commercial sectors accounts for 29% of electricity consumption and rises at the rate of 8% annually. (Govt. of India Planning, 2011). In India, majority of the households are dependent on ceiling or table fans for cooling in summer, and air-conditioning use is relatively less. In view of the construction as the second largest economic activity (8%), and the projected urban population of about 600 million by 2030 (Govt. of India Planning, 2011), India's energy consumption during construction and operations is expected to increase, exponentially. This paper is limited to low-energy imperatives during operations, and doesn't cover full life cycle cost analysis for the construction period. Traditional architecture manifests the local climate, lifestyle, and materials. Faridabad, located at coordinates: 28.9°N, 77°E; and 216m above mean sea level is in ‘composite climate’, with extreme climate swings: maximum DBT of 45°C for about 2 and half months, followed by hot-humid monsoon, and minimum DBT of 3°C for a shorter winter heating period. The region, besides some haveli1 typology housing, is largely characterise by dense settlements, and compact planning with narrow pedestrian access that serves as socio-spaces and extension of work spaces to the semi-public ground floors adjacent to the access routes. In the poorer sections of the city, houses are often constructed next to each other with little or no setbacks. Traditional homes, in the area, are introverted spaces with a courtyard open to perimeter rooms and sky, high thermal mass local stone walls of 400-500mm thick, roof with 50mm stone slabs supported by wooden beams, small size openings with Jaalis3 and Chajjas4 for privacy, airflow, and shade, as shown in Figure 1. (Archinomy. n.d.). Shops and bathroom serves as buffer, and terrace can be used for outdoor sleeping. With the first floor reserved for women folks, these traditional introverted houses have small openings on the exterior walls, but larger openings to the internal courtyard. Bidani Eco-house is located in a medium density Faridabad residential neighbourhood, mostly with 1 or 2 storey detached houses on a plot size of 1000m2. b) Paper model with front facing north passage a) Section through private & public spaces Figure 1 (a) Section of a traditional dwelling at Khampur village, near west Patel Nagar, Delhi, and (b) Model of the dwelling with cusped mughal arch entrance, small openings with Jaalis3 and Chajjas4. Source: http://www.archinomy.com/case-studies/677/traditional-dwelling-in-delhi [02.06.2014]. Motivation and Objectives The past decades have witnessed unprecedented revolution of science and technology that brought great economic and socio-cultural benefits. However, it was also a nature-human dichotomy period, when vernacularism or passive design ideologies escaped our collective wisdom, subsequent adaptation to heat producing equipments, generic modernism, and destruction of the native habitats milieu. In an emerging society where majority of its population can't afford active heating and cooling systems, passive design techniques that envisaged reduction in artificial lighting, heating or cooling, and innovative use of locally sourced low embodied energy materials are key to low-energy paradigms. In Delhi's traditional architecture, mutually shading haveli1 heat sinks, high thermal mass local stone walls keep the inside cool due to time-lag in the day while its high emissivity allows rapid cooling of the surface at night. Within the framework of socio-cultural changes and new notions of lifestyle comforts, the paper aims to highlight Bidani Eco-house passive techniques, its integration with lifestyle, and subsequent low-energy paradigm when compared with GRIHA2 energy performance benchmark. HYPOTHESIS AND METHODOLOGY Re-interpretation of traditional passive techniques and pragmatic response to the adverse or advantageous climatic parameters, such as: temperature, solar radiation, humidity, air speed, etc presents sustainable habitat solutions, in contemporary emerging economies. Its subsequent manifestation in architecture is primarily defined by the availability of resources on the one hand, and lifestyle-praxis on 3. Perforated stone or lattice screens in Indo-Islamic architecture. 4. Sun-shading device for roof or windows in India, and usually supported on large carved brackets. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 134 the other. Unlike traditional compact plans, modern buildings produced much heat of their own and heat loss or gain through appropriate ventilations or air changes through open-spaces and storage of heat or cold in high thermal mass envelopes, through responsive lifestyles, could be beneficial. Through site observations and measurements with data loggers and a series of parametric simulations and analysis with Solar Designer ver. 6, the paper highlights 1) passive low-energy architecture techniques in the house, 2) effects of ventilation modes, air change rates, and thermal mass on thermal comfort parameters in various seasons, 3) energy performance analysis with GRIHA2 and simulations. Data loggers was logged from 7th - 10th January, 2014, in the living room and bedroom, and simulations were performed for a whole year covering 3 representative days of each of the 12 months, and best ventilation modes and air change rates were highlighted, in a manner close to how the house is operated. THE ARCHITECTURE Passive strategies differs base on a place climatic parameters: temperature, radiation, humidity, air speed, etc and their re-interpretation on the site, plan form and geometry, and envelope limits or allows the structures heat gain or loss. Given the composite climatic pre-requisites, Bidani Eco-house attempts to account for varied complex low-energy imperatives: minimizing summer heat gain during hot-dry season, maximization of passive ventilation during the hot-humid periods, passive solar heating in winter, and visual comforts. While the plan form and geometry's orientation aids or hinder solar radiation, the large volumetric composition of the living space with adjacent haveli1 heat sink is expected to enhanced porosity, thereby, ventilation and air-changes while thermal mass walls and mud-phaska5 roof attenuates heat gain in the day, it enhance heat loss at night, as shown in Figure 3 & 4. Additionally, mutual 1 shading with haveli voids and solid blocks or pergolas and trees, thermal buffer spaces, and evaporative cooling from water sprinkled on the front lawn are expected to reduce heat gain. Table 1. Passive Design Techniques in Bidani Eco-house Area/ Location Passive Techniques Functions Site Grass lawns and perimeter trees Evapotranspiration, bio-purifier Plan form & geometry Re-interpreted haveli1 typology with Heat sink, minimize heat gain, minimum east & west exposure visual comfort, privacy Buffer Toilets, stores, garage on S-west Reduce heat gain Courtyard Pergolas and north-east location Shade, heat sink Roof RCC with mud-phaska5 and stones Insulation, thermal mass Ceiling Concrete, white paint Diffused light, visual comfort Walls Low U-value local stones and concrete Thermal mass, low-energy Fenestrations Eaves/awnings, single glass, wire mesh Natural ventilation & safety Floor Stone with mortar on concrete Ground contact, thermal mass Verandah/ Porch Deep eaves and pergolas Shading Site Landscape The house creates its own landscape microclimate in responds to the adverse local composite climate and air pollution, within its inscribed territory. Bidani Eco-house with a built up area of 295 m² on a site area of 1000 m² has a width to depth ratio of 1:3, with the shorter side oriented towards the north and access road, which restricts design flexibility and limits generation of ideal plan form and geometry, or orientation. During the hot summer seasons, open lawns with >50% grass cover, and perimeter trees provides shade, acts as a bio-purifier to the hot-dusty air and cools the environment by evapotranspiration, and increase air speed due to narrow path of the hedges and open spaces, as shown in Figure 2. The site’s access road runs east west, and adjacent buildings and landscape trees provide shading from the harsh morning or afternoon sun from east and west. Plan Form and Geometry Buildings plan form's compactness or openness, and thereby, porosity to its surrounding landscape or geometric composition with respect to solar geometry, and envelope thermal mass can aid or hinder heat gain or lost, while its orientation can be crucial to control solar radiation and stabilizing extreme temperature swings. Temperature swings in the house is expected to be stabilized by shaded Northeast 5. A type of waterproofing on roof terrace in India. It consist of a mortar bed of mud soil compacted by pressing with a weighty instrument, laid to slope and tiles are laid on top. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 135 a) Plans b) View from north entry Figure 2 (a) Plan form with spaces layout around N-east oriented haveli1, and test rooms: 1. Atrium Living room, 2. Bedroom, and (b) High thermal mass walls with local stones, or concrete and greeneries. haveli1 heat sinks, high thermal mass local stones or concrete envelopes to the east and west, as shown in 1 Figure 2 & 3. The haveli heat sink opened towards northeast forms the central fulcrum, like traditional dwellings, with various spaces: bed rooms, dining room, large volumetric living space, etc around it, as shown in Figure 2(a). The oblique alignment of the plan form and geometry is expected to enhance passive cooling since only the narrowest elevations are exposed to the east-west low angle solar radiation while mud-phaska5 roof allows the building to consistently minimize high altitude mid-day solar radiations, but allows indirect natural light for visual comforts, as shown in Figure 3. Pergolas, louvers, eaves, and awnings provides shades to openings. The large atrium living-room with low sill windows wraps around the shaded N-East haveli1 heat sink, and the geometry of its ziggurat-like roof structure with glass and louvers on the vertical side allows for low altitude winter sun to penetrate while doubly functioning as summer time's hot air exhaust vent, as shown in Figure 3. These passive techniques, in combination with responsive lifestyle, such as: flexible cooling from various ventilation modes or air-changes as per seasonal conditions and evaporative cooling from vegetation are expected to maintain indoor "thermal comfort" in summer and winter, with low-energy. c) Winter heating a) Summer cooling Figure 3 (a) Summer cooling with geometry & pergola shading and heat sink to haveli1, and louvers, (b) Winter heating from south-east low altitude solar radiation. Building Envelope A building's envelope encompasses walls, floors, roof, windows, etc. The materials such as: high thermal mass local stones and concrete for the main envelope, beige granite stone floor (originally terrazzo), RCC slab roofing with mud-phaska5, etc, are all locally available within reasonable distances, and thereby, low embodied energy, as shown in Figure 4. Local stonewalls with 1.5Wm–2 low U value (Roaf. S, et al. 2001) and concrete provide thermal mass to attenuate diurnal thermal swings, and single glazed windows with safety grilles and awnings allows night ventilation in summer and southeast 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 136 b) Original view of the house in stones & concrete a) View of atrium living-room Figure 4 (a) Build envelope with high thermal mass local stone on east walls, and (b) North view of the original build envelop with high thermal mass stone and concrete. radiation ingress in winter. Buffer spaces, such as: toilets and stores, garage on the south-western perimeter, pergolas, eaves enhance thermal performance of the building by eliminating solar penetration to living spaces, and thereby, reducing cooling load in summer time, as shown in Figure 7(b). SIMULATIONS AND ANALYSIS In order to highlight the most representative way the house is used and the role of the building fabric, in various seasons of the year, a matrix combinations of ventilation modes, and air change rates were analyze through parametric simulations with Solar Designer ver. 6 (http://qcd.co.jp/). The outcomes are discussed in hourly indoor temperature fluctuation for various climate conditions of winter and summer, and monthly best for a whole year, as shown in Figure 5 & 6. Representative days of each month were selected based on the weather conditions: very sunny, cloudy, and sunny. These 3 days, after extensive parametric simulations, were found to be representative of the months' ambient temperature fluctuation pattern. The simulated atrium living room is 15m x 7m x 7m high to account for extra volumes of adjacent abutting smaller rooms, and openings of size 8mX5m on the S-East and 2mX2.4m on the Neast were incorporated, as shown in Figure 2 & 4. N-East and S-West walls are considered to be 30cm thermal mass local stones with low U-value of 1.5Wm–2, and others are 23cm concrete with plaster and no extra heating affect was considered from adjacent rooms. The floor is in earth contact, and adequate insulation added for mud-phaska5 roofing along with 20cm RCC. Deep eaves, both vertical and horizontal are incorporated considering the geometry, as shown in Figure 3 & 4. As internal heat sources, a constant 418.68kJ/h for refrigerators, 1.8MJ/h for laundry 2 hours/day, and 1.67MJ/h for 2 family members and 2 domestic help were set everyday for 8 hrs. Windows have curtain insulations. b) Temperature, radiation, & DI in May a) Temperature, radiation, & DI in January Figure 5 (a) Simulated effects of ventilation modes and air change rates in January, and (b) Simulated effects of ventilation modes and air changes rates in May, in the atrium living room. Extensive parametric simulations for various ventilation mode and air change rates for winter (January) and summer (May), on representative days, shows the best ventilation modes: night-ventilation 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 137 (30 ACH at night, 1 ACH in the day) in summer, and in winter air-tightness (1 ACH, both in the night and day), as shown in Figure 5. Next, monthly best ventilation modes and ACH for a year were selected and temperature fluctuation in the test room highlighted for each month, as shown in Figure 6. The maximum solar radiation on the South facade was about 2.657MJ/m2 in January and 1.2MJ/m2 in May, and the glass serves as the media for heat egress in summer and solar heat ingress in winter, as shown in Figure 3. The monthly average indoor temperatures are: January, 17.4ºC; February, 19.2ºC; March, 24ºC; April, 24.24ºC; May, 31.3ºC; June, 31.3ºC; July, 30.7ºC; August, 30.4ºC; September, 28.4ºC; October, 26.22ºC; November, 22.83ºC; December, 19.06 ºC, as shown in Figure 5 & 6. The maximum monthly temperatures for cooling period were: May, 36.86ºC; June, 35.98ºC; July, 35.04ºC; August, 34.4ºC; Sept, 31.6ºC; October, 30.1ºC, as shown in Figure 6. Based on this findings and the heating, cooling and lighting loads for the year 2013, we could surmise that attenuation of indoor temperatures swings in summer and lifestyle responsive to seasonal and daily temperature fluctuations have resulted in reduce energy consumption, as shown in Figure 7(b). Thermal performance of the atrium living room, "as-built orientation" and hypothetical "south orientation", was analyzed from the perspectives of energy performance, and, "as-built orientation" was about 4.9% more energy efficient under Flex Vent System, 18°C< AT <30°C, with 30ACH when AT (Ambient temperature) is 18-30°C, and 0.5ACH at other times for 8 hours occupancy per day. The envelope has sufficient number of operable doors and windows. Active cooling was required for parts of summer. However, night-ventilation allows the natural microclimate to prevail and the room air temperature dropped to almost the same level as the outdoor temperature, while closing the openings during the day allows the high thermal mass envelope to retain lower indoor temperature throughout the day in summer, and thereby energy savings. In hot-dry periods, evaporative cooling from water sprinkled grass lawns and ventilation airflow afforded by optimized open spaces and haveli1 attenuate heat gain. In winter, daytime ventilation or 'air-tightness' and green house effect from south-east facade glass could afford an average indoor temperature of about 17.4ºC and extra heating was required, as shown in Figure 5(a). Discomfort Index, DI=0.81Td+0.01H(0.99Td14.3)+46.3, where Td=Indoor Temperature(ºC), H=Relative Humidity (%), developed by the American Weather Bureau (US) in 1957, was used to calculate DI after finding the absolute humidity in g/kg of dry air, and relative humidity(%) on psychrometric chart. One percent of the population feels unpleasant if discomfort index exceeds 75, and all will become uncomfortable if it exceeds 80. The house, as-built, is uncomfortable with Discomfort Index above 75% in May, and parts of summer months, as shown in Figure 5(b). Figure 6 Simulated monthly best temperature (°C) fluctuation from January to December, in atrium living room (as-built), due to the effects of ventilation modes, air changes, shading, and thermal mass. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 138 ENERGY PERFORMANCE ANALYSIS AND SITE MEASUREMENTS The building envelopes high thermal mass helps in attenuating extreme temperature swings. But it also resulted in a stable low temperature, as shown in Figure 7(a). The atrium living-room recorded a low temperature of 15.5ºC average, with a high of 24.1ºC because of high thermal mass and painting of the top glass openings on south-east that blocks off solar radiation. The bedroom, on the other hand, has a comfortable indoor temperature of 19.2ºC average and a high of 24.1ºC since a heater was used and the bedroom had access to Southeast and Southwest solar radiation through glass windows and thermal mass walls retains heat, as shown in Figure 2(a). Through questionnaires, the authors ascertained thermal comfort, was afforded by Flex Vent system, where air-condition was on when the temperatures are not within comfort zone, say 18-30ºC. At other times, operable openings are opened and plenty of ventilation and air changes were allowed. The DI, Discomfort Index for site measurement was below 75% in both cases, as shown in Figure 7(a). a) Temperature & DI fluctuation on 7th to 10th January b) Energy Performance Index (EPI) Figure 7 (a) Temperature and DI (Discomfort Index) fluctuation due to lifestyle, ventilation modes, thermal mass, and shading in January, (b) Ecohouse Bidani’s annual (2013) heating and cooling load and energy performance index (EPI) as per GRIHA2 Version 3.0. The authors conducted site measurements on the 7th to 10th of January 2014, as shown in Figure 7(a). Bidani Eco-house is a residential building with a total built-up area of 295m², and a multigenerational residence. The grandparents live in the ground floor, approx. 195m2, while the son and his wife, and grand children used to live in the upper floor, approx. 100m2. The house occupancy is 24x7 for the grandparents, but the son and family do not continuously live in the house. Additionally, domestic helps encompassing a mother, father and 2 children also stays in the house sometimes. The possible sources of heat in the house are electronic equipments, such as: computers, TV, portable heat radiators, room electric heater, kitchen cooking stove, etc. The total heating, cooling, and lighting loads for 12 months, in 2013, was 37440MJ. According to GRIHA2 Version 3.0, “the annual energy consumption of energy systems in a residence (24×7 occupancy) should not exceed the benchmark limits of 360MJ/m2/year, as shown in Figure 7(b). Eco-house Bidani's Design EPI (energy performance index) of 126MJ/m2/year and 192MJ/m2/year for the total floor area and pro-rata occupied area respectively, as per occupancy validates the performance of passive design techniques, as well as the Bidani family's lifestyle responsiveness to passive ventilation cooling or passive solar heating, as shown in Figure 7(b). Simulations were done to analyze energy performance index for Eco-house Bidani, if the house was fully occupied all year round, and total heating and cooling load was 63265MJ for 295m2. So, the simulated energy performance index (EPI) for the whole building under Flex Vent System (18-30) was 214MJ/m2/year which is lower than GRIHA2 benchmark of 360MJ/m2/year, as shown in Figure 7(b). Energy performance for the house was calculated by applying the actual pro-rata area of 195m2 (occupied area), to the current annual (2013) consumption of 37440MJ and the result 192MJ/ m2/year. The actual EPI was lower than simulated heating and cooling load of 214MJ/m2/year, which further validates lifestyle responsiveness of the occupants. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 139 CONCLUSIONS In contrast to the conventional practices of closing doors and windows at night for security and to keep off unwanted bugs, appropriate ventilation modes or air changes through lifestyles responsive to diurnal and seasonal climate swings, resulted in a low-energy, efficient EPI (Energy performance index) of 126MJ/m2/year for total area. Bidani eco-house attenuated extreme climate swings of a composite climate through passive design techniques infused into the plan form, geometry, and orientation to minimize solar radiations in summer and enhance southeast winter sun or visual comfort. While local stones high thermal capacity afforded time lag, its high emissivity enhance heat loss at night. The historic usage of haveli1 or the high thermal mass of local stones, and evaporative cooling prevalent to Delhi's vernacular architecture are effectively re-interpreted towards "comfort affordance" through passive cooling or heating in summer or winter. Though 100% thermal comfort is not possible through passive cooling or heating, it is possible to reduce peak energy load. The build form and envelopes contiguous relationship with climatic parameters: temperature, radiation, and airflow directly affect the internal heat gain or loss. As learning, this paper highlighted various passive cooling and heating afforded by re-interpretation of traditional passive design techniques and subsequent integration with lifestyle towards low-energy paradigms for emerging economies. In view of Delhi's extreme climate and conduction of ambient heat, good solar glass could further attenuate heat gain or loss. Furthermore, the paper validates the possibility of using an interactive design tool, Solar Designer and Energy bills or GRIHA2 benchmark for energy performance as an effective method in assessing the design as well as lifestyle-praxis. To conclude, passive design techniques and responsive lifestyle praxis are both a necessity for low-energy sustainable habitats in emerging economies during operations. ACKNOWLEDGEMENTS The successful completion of these paper has been made possible through the encouragements, continue support, and guidance of teachers, project architects, house owners, family, friends, and in essence, all who directly or indirectly contributed to its progress and refinement. The opportunity to express our deepest gratitude has presented itself with the completion of this paper, and the authors are much indebted to, and acknowledge all concerned including but not limited to the house owners, Bidani family. REFERENCES Kodama.Y, et al. 2006. Influence of Ventilation on Mode Passive Cooling Effect - a proposal of Flex Vent System-. < http://www.unige.ch/cuepe/html/plea2006/Vol2/PLEA2006_PAPER240.pdf> Krishan, A. 2001. Shelter or Form. In: Krishan, A et al. 2007. Climate Responsive Architecture, A Design Handbook for Energy Efficient Buildings. Second reprint. New Delhi: Tata McGraw-Hill Publishing Company Limited, New Delhi, 2001. Print. p.32. Olesen, B.W. and Brager, G.S. 2004. A Better Way to Predict Comfort. ASHRAE Journal, August 2004. p. 21. <http://www.cbe.berkeley.edu/research/pdf_files/OlesenBrager2004_comfort.pdf> [12 March 2014]. Govt. of India, Ministry of New and Renewable Energy. (n.d.). CHAPTER – 2, CLIMATE AND BUILDINGS. Clause 2.4. p.19. <http://mnre.gov.in/solar-energy/ch2.pdf> [4th March 2014] Govt. of India, Planning Commission. The Final Report of the Expert Group on Low Carbon Strategies for Inclusive Growth, edited by Parikh, Kirit. New Delhi: May 2011. p. 47. < http://planningcommission.nic.in/reports/genrep/rep_carbon2005.pdf> [30 May 2014]. Archinomy. n.d. Traditional Dwelling in Delhi. <http://www.archinomy.com/casestudies/677/traditional-dwelling-in-delhi> [02 June 2014]. Roaf, S., Fuentes, M., Thomas, S., 2001.Bidani House. ECOHOUSE: A DESIGN GUIDE. p. 288-290. Oxford, England. Architecture Press, A division of Reed Educational and Professional Publishing Ltd. <http://www.unigaiabrasil.org/Cursos/Apresenta/PDFs/Marcelo/Ecohouse.A.Design.Guide.0750649046.pdf> [9th April 2014]. GRIHA1 Version 3.0. Clause 14.1.3. Avalable from: <http://planningcommission.nic.in/reports/genrep/Inter_Exp.pdf> [10 April 2014]. Olgyay, V. 1963. Design with climate, Princeton University press, Princeton, New Jersey, 1963. Print. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 140 A Comparative Study of Design Strategies for Energy Efficiency in 6 High-Rise Buildings in Two Different Climates Babak Raji1, MSc 1Faculty Martin J. Tenpierik1, PhD Andy van den Dobbelsteen1, PhD of Architecture and the Built Environment, Delft University of Technology, Delft, the Netherlands ABSTRACT Due to the ever growing trend of urbanization and population growth, the construction of high-rise buildings is inevitable and will also continue at an ever increasing pace. However, typical high-rise buildings (the traditional template of a rectilinear, air-conditioned box) are not energy efficient in many aspects of their design. In this research the impact of architectural design elements on building energy performance will be studied through a combined literature review and case study research on 6 high-rise buildings with different degree of sustainability and located in two climate types, sub-tropical and temperate. The exterior envelope, building form and orientation, service core placement, plan layout, and special design elements like atria and sky gardens are the subject of investigation. This study found that a double-skin façade with automated blinds and operable windows besides a narrow floor plan, the correct placement of core services in regards to solar heat gains, and the application of vertical shafts like atria, which bring daylight and natural ventilation deeper into the plan, are the strategies that effectively can provide energy savings for tall buildings. However, when the building has this potential to use energy efficient design strategies, the real performance depends on how the building is used by the occupants. Designers should therefore take user behavior into account during the design stage. INTRODUCTION Urbanization, insecurity of resources and climate change are key challenges toward the future of cities (Dobbelsteen, 2012). As cities become denser and buildings become taller, sustainability may be at stake. Tall buildings are source-intensive due to the excessive scale and complexity of design (Cook, Browning, & Garvin, 2013). A wrong design strategy can lead to more energy consumption. This paper addresses design strategies that help a high-rise building to be more energy efficient in both a temperate and a subtropical climate. In order to have high performance tall buildings, first there is a need to reduce the building’s demand for energy and the most straight forward approach is to design them in a way that reduces their appetite for consumption. METHODOLOGY 6 case studies with different degree of sustainability were selected from 2 climate types (temperate & sub-tropical). For each case, building-related energy performance data was collected through a literature review and contact with the energy consultants. This energy performance data of each group of buildings in one climate (3 cases) was compared to analyze the effectiveness of different design strategies for cooling, heating, ventilation and lighting in the specific climate type. Finally, energy-efficient design solutions were defined for both climates. The selection criteria for the cases were: - Considered by one of the rating systems or standards as a high-performance building - Availability of building-related energy performance data (metered or simulated) - Newly constructed office building that has been occupied for two years with at least 15 floors The difficulty with this kind of studies is that often the energy consumption data are neither measured nor 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 141 Sub-tropical Temperate made publicly available. Therefore, it is always difficult to normalize all of the conditions between different cases. Sometimes, simulated energy data are the only source or internal conditions such as occupancy rate, building function and office equipment may vary among cases. Climate variations during different years can also influence the energy consumption. Therefore all of these conditions ideally should be accounted for when making the comparison. In this paper, the presented energy figures are delivered energy in kWh/m2 of gross floor area, unless it is mentioned otherwise. For making the energy figures comparable, conversions were applied for some cases. The plan configurations and the energy performance data of the six cases are presented in Figure 1 & 2 respectively. Figure 1. Building orientation and plan configuration for the 6 buildings. Red color areas show the position of the service core and blue color areas present a vertical shaft like an atrium, a circulation void or an open void. Figure 2. Energy performance data of the six cases. (S)=Simulated; (M)=Metered; the electricity consumption is just for lighting, pumps and fans. 1The EUI for the Commerzbank (Goncalves & Bode, 2010) and the Post Tower (S Reuss 2014, pers. comm. 19 May) were originally calculated based on the net floor area. To convert the figures from net to gross floor area an efficiency factor (net area/gross area) of 61% & 57% is considered respectively for Commerzbank and Post Tower. In addition a very small amount of the cooling load is combined with the electricity usage in Commerzbank building that should be negligible. 2The energy consumption at 30 St Mary Axe (N Clark 2014, pers. comm. 12 May) is simulated on two scenarios: a fully air-conditioned design on levels 16-34 and a mixed-mode design on levels 2-15. 3The energy consumption of the Liberty Tower (Kato & Chikamoto, 2002) is converted from primary energy to delivered energy with an average efficiency factor around 45.4% for power plants in Japan. 41 Bligh Street (Yudelson & Meyer, 2013) building use a tri-generation system for combined cooling, heating and electricity generation. The 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 142 projected energy sources are gas and electricity. However, it is not clear how much is used to generate heat or lighting. 5Torre Cube (Wood & Salib, 2013) does not rely on an air-conditioning system for cooling, heating or ventilation. Therefore the energy consumption is zero in this building. The electricity consumption for lighting and equipment has not been published for this building. Therefore the predicted consumption is presented with a dashed line. TEMPERATE CLIMATE Cooling Among the three case studies in the temperate climate, the Post Tower has the lowest energy use (from the grid) for cooling by around zero. Cooling is provided through thermally active ceilings and a decentralized supplementary fan coil system. Cold water from the nearby Rhine River and a sunk well is used as a source. Furthermore, the building is oriented based on the sun path with the long axis almost along east-west. The Commerzbank’s energy consumption for cooling is around 29.5 kWh/m2. The building uses absorption chillers to generate cold water which is distributed through chilled ceilings. Natural ventilation throughout up to 80% of the year reduces the cooling need of this building. Both buildings have a double-skin façade with ventilated cavity and motorized blinds for solar control preventing excessive heat gains. Both buildings apply night-time ventilation and a BMS to control the operation of blinds and openings. The occupants can override this BMS to customize the climate to their desires. The Mary Axe building uses a decentralized air-conditioning system on each office floor. According to the simulation results, the cooling demand was lower when using natural ventilation in mixed mode zone compared to other one that was entirely air-conditioned. The total energy consumption of all of the 3 buildings is considerably less than of typical air-conditioned buildings. Heating Considering heating, the Mary Axe building has the lowest energy consumption. The air supply to the air handling units (AHUs) is provided by narrow slits between the glazing panels, then conditioned by the AHU and then distributed through adjusted ducts at ceiling level. Part of the exhaust air from the offices is used to ventilate the cavity inside the facade; therefore, in winter, the cavity will have a temperature similar to that of the indoor air, thereby reducing the heat loss through the envelope. Base on the simulation results of Mary Axe building, the heating demand is slightly higher when introducing natural ventilation into the building compared to a fully air-conditioned mode. The Post Tower can be ranked second best with an energy consumption for heating of around 30.8 kWh/m2. The energy source is waste heat from electricity production (district heating). Furthermore, the deep cavity (120-170 cm) within the double-skin façade acts as a thermal buffer between the outdoor and indoor air. On cold winter days, fresh air firstly is warmed up in the double-skin façade before it enters the perimeter fan coil units; thus reducing the need for heating. The energy consumption for heating of the Commerzbank is 42.5 kWh/m2, higher than of the Mary Axe and the Post Tower. The energy for heating is provided by the local district heating network and is distributed through thermostatically operated radiators. The double skin façade of this building has the narrowest cavity (20 cm) among the three buildings. However, the window-to-wall ratio of this building is lower (around 58%) than of the other cases which are fully covered with glass. Ventilation All of the three cases use a mixed-mode ventilation strategy (natural ventilation + mechanical ventilation). However, the duration of natural ventilation is different throughout the year. With the help of architectural elements (central atrium and the sky gardens) and special plan configuration of Commerzbank, internal-facing offices can be naturally ventilated throughout the entire year. The outwardfacing offices can also utilize natural ventilation up to 80% of the year. For the Post Tower, all of the working areas and communal spaces can be naturally ventilated with a combination of cross and stack ventilation. Only interior meeting rooms and conference halls are conditioned mechanically. The outer skin 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 143 of the façade is extended to create an aerodynamic form which increases the ventilation rate. In both projects, the double-skin façade is naturally ventilated and night-time ventilation is applied during summer. The office areas in the Mary Axe building are not ventilated directly through the façade. Fresh air first comes into 6 peripheral atria through small openings in the façade before this tempered air is distributed to the working stations. For the original design, it was predicted that the office areas could be naturally ventilated during 41-48% of the year. But with a change from owner occupation to multi-tenant occupation, most tenants rejected the energy-efficiency package with automated windows and choose for the year round air conditioning package instead (Wood & Salib, 2013). Because of the deep plan of this building, the central service core is mechanically ventilated. Besides, the cavity inside the facade is not ventilated with fresh air but with extracted air from the offices. Furthermore, the building does not use nigh-time ventilation. Lighting The Commerzbank has the highest electricity consumption (67.7 kWh/m2) among the case studies in the temperate climate. As it is not clear how much of this energy is used for lighting, it is difficult to determine the causes for this and might be derived from a prestigious design, more office equipment, higher number of occupants per square meter or architectural design features like window-to-wall ratio and plan depth. Considering the façade transparency, the Commerzbank has the lowest window-to-wall ratio of approximately 58%. This could mean that there is more need for artificial lighting. However, a full height central atrium and 9 spiral sky gardens bring a lot of natural light deep into the building interior. In the Post Tower around 85% of the working stations are located within 5 meters from the external façade. A considerable part of the office spaces therefore utilizes daylight reducing the energy demand for artificial lighting significantly. Furthermore, most of the meeting rooms and service spaces at the heart of the building are faced toward a central atrium and can therefore also be naturally lit. The office spaces can operate in stand-by mode when the rooms are empty. From the total electricity consumption, lighting is 6.2 kWh/m2. In the Mary Axe building, the distance between the core and perimeter ranges from 6.4 to 13.1m depending on the floor size. This building thus has a deeper plan compared to the other cases. However, the problem of a deep plan is solved here with the help of 6 triangular atria along the building perimeter. All of the rectangular office fingers can be naturally lit from three directions. The big central service core should always be artificially lit due to its central placement. The total electricity consumption for lighting are respectively 26.4 and 29.1 kWh/m2 for mixed-mode (levels 2-15) and fully air-conditioned (levels 16-34) zones. SUB-TROPICAL CLIMATE Cooling Among the cases in a sub-tropical climate, Torre Cube has the lowest energy consumption for both heating and cooling (0 kWh/m2) because it does not depend on an air-conditioning system. Due to the mild climate in Guadalajara, buildings in this city can be naturally ventilated throughout the entire year if designed well. Solar radiation intensity, however, is very high in this area making sun-shading an essential additional strategy for passive cooling. Adjustable external screens protect this building from excessive heat gain in summer. The 1 Bligh Street building in Sydney is equipped with a hybrid tri-generation system that simultaneously generates heat, cold and electrical power. 500 m2 of the roof of this building is covered with solar collectors that feed the absorption chiller to generate cold. Therefore, the building does not use electricity from the grid for cooling. Furthermore, the compact elliptical form has 12% less surface area than a rectilinear building of the same volume, thus reducing the heat gain/loss through the building envelope. In addition, a high-performance naturally ventilated double-skin façade with 60 cm cavity helps to reduce the heat gain through the envelope. However, there is some debate considering the land use and ecology of this building. The building’s orientation and configuration of plan are mainly derived from the urban grid and the desire to maximize the view, not from environmental concerns. While the service core 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 144 could have been used as solar buffer on the hot east and west side, it is placed on the south side (non-harbor side of the floor plate). The Liberty tower in Tokyo has an educational function, which because of high occupancy rates typically has a higher cooling demand than an office function. The building uses around 34 kWh/m2 for cooling which is higher than the other two sustainable buildings. However, the 1 Bligh Street building’s dependence on renewable energy (solar energy) for cooling does not mean that the cooling demand of this building is less than of Liberty Tower. This building does not seem to be oriented environmentally. The majority of lecture rooms are facing (south)east whereas the opposite (north)west contains the majority of service areas. Vertical and horizontal concrete fins on the façade protect the openings from high solar gains in summer. Heating As mentioned before, Torre Cube has zero energy use for heating due to the mild weather conditions of Guadalajara. During the cold months (December and January) daily mean temperature is around 17°C. As a result, the internal and passive solar heat gains are sufficient to warm up the small interior office spaces. Liberty Tower’s heating load is around 40 kWh/m2. The rectilinear shape of the building increases the surface area and therefore the heat gains/losses through the envelope. 1 Bligh Street building uses 73.7 kWh/m2 gas to feed a gas-fired tri-generation system which generates electricity and useful heat. It is up to 50% more efficient compared to conventional grid-connected systems. From the waste heat, ‘free’ cooling and hot water can be generated. The office spaces are fully air-conditioned and separated from the atrium by glass walls. Extracted conditioned air from the offices is used to temper the naturally ventilated atrium. However building’s energy use for heating has not been published. Ventilation 1 Bligh Street has two strategies for ventilation. The communal heart of the building is naturally ventilated but the working areas are fully mechanically ventilated. Natural fresh air is provided through an opening on the ground floor and a sky garden on the 15th floor and is distributed on all floors by stack ventilation in a full height atrium. The building is designed in a way that the perimeter cellular offices may potentially use single-sided natural ventilation if the interior glass panels are replaced with operable ones. But the deep floor plate does not allow for cross ventilation. With the help of natural ventilation, the annual cooling demand at Liberty tower was reduced by 17%. Two architectural elements that effectively have improved this natural ventilation strategy are the escalator void and a wind floor on the 18th floor on top of the circulation shaft. CFD analysis has shown that the wind floor increases the air flow rate by 30% (Kato & Chikamoto, 2002). As the escalator void is not segmented, there is a risk of extreme stack effect and draft inside the building. Furthermore, the introduction of fresh air directly into the working areas might provide discomfort especially for the occupants sitting near the air inlets. In the Liberty Tower cool fresh air comes in directly through the inlets below the fixed windows. The inability of the occupants to control their operation (fully controlled by a BMS) may limit their comfort and may result in user dissatisfaction (cold feet). The Torre Cube building uses different architectural elements to provide both cross and stack ventilation. Fan-shaped office wings help to funnel the air across the working spaces before it is exhausted through a central open void. Three open spiral sky gardens lead to a higher air circulation in the void. However, without a CFD analysis it is not clear if the sky gardens have a positive or a negative effect on buoyancy in the central void. Lighting 1 Bligh Street has a fully transparent façade. However, in 1 Bligh Street just 30% of permanent working stations are within 5 meters of this façade. Due to this deep plan (23.5 m from façade to central void), there are three working zones between the building perimeter and the atrium. A central atrium and transparent partitions are used to increase natural light penetration. Temporarily used spaces such as meeting rooms are placed in the mid-zone. The figures of electricity consumption for lighting, fans and ventilation has not been simulated but the total delivered electricity is around 32.1 kWh/m2. Torre Cube’s 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 145 electricity use for lighting has not been published. Because of a central void, the office wings in this building receive daylight from two sides, which allows for a deep office plan of about 9-12.5m. The electricity use for lighting and pumps of Liberty Tower is around 55.5 kWh/m2. Considering the façade transparency, the Liberty Tower and Torre Cube have compared lower window-to-wall ratio than 1 Bligh Street. EFFECTIVE DESIGN STRATEGIES FOR HIGH-RISES General design strategies for high-rise office buildings Concerning plan configuration, it is important to place the permanent work stations close to the envelope to reduce the need for artificial lighting. Dividing the internal zone into areas with different temperature is another important strategy that can reduce the cooling/heating load of high-rise buildings. Office workers expect a high degree of comfort in their work stations but tolerate a little bit of discomfort in lift lobbies and communal spaces. Plan form and building shape (or compactness) can influence the amount of heat gain/loss through the envelope. Circular and elliptical forms have an exposed surface area that is respectively 25% and 12% less than of a rectilinear building of the same volume. Furthermore, an aerodynamically curved form minimizes wind turbulence and downdraft at street level. Furthermore, the effectiveness of natural ventilation and daylight depends strongly on how the openings and solar shading devices are controlled. The absence of a central BMS might cause problems in attaining the right adjustments for providing indoor comfort conditions and may increase the energy consumption. Smart occupancy sensors cut down unnecessary consumption for lighting, and mechanical ventilation. In cellular offices, it is important that occupants can override the BMS to ensure their individual comfort. Psychologically, occupants with more control over their environment are more tolerant to high or low temperatures. However, the BMS should automatically switch off the air-conditioning system if occupants decide to open the window. Design strategies for high-rise office buildings in a temperate climate Façade transparency and plan depth are the two dominant factors with great influence on the electricity demand for lighting. A fully transparent façade is a common strategy in a temperate climate. However, it is important to provide a balance between the use of daylight, the solar heat gain in summer and the heat loss in winter. A double-skin façade with a deep cavity is an effective strategy for reducing the cooling and heating loads of high-rise buildings in temperate climates. A double-skin façade can act as a thermal buffer between the outdoor and indoor environment. Moreover, offices next to this façade can use natural ventilation for a longer period of time if fresh air first passes through the cavity in the double-skin façade before entering the offices. However, an effective ventilation strategy is highly needed inside this cavity especially during summer, otherwise the double-skin façade would act like a greenhouse and transfer a lot of heat into the building. Solar control devices within the cavity such as a motorized venetian blind allow for passive heating in winter, but prevent unpleasant glare and overheating in summer. A mixed-mode (natural and mechanical) ventilation strategy can reduce effectively the energy demand for cooling and mechanical ventilation. Some architectural elements that can help the air intake, circulation and exhaust are sky gardens and vertical shafts like atria and circulation voids. When using a full-height atrium, there is a risk of high temperature differences and extreme stack effects and drafts. For controlling this excessive stack effect, a full-height atrium is usually segmented into smaller zones with lower pressure difference. Design strategies for high-rise office buildings in a sub-tropical climate In a sub-tropical climate, solar radiation intensity is high. Therefore the most effective design strategies are those that reduce the solar heat gain. Such strategies include limited façade transparency on the east and west side of the building, the placement of service cores on the hot sides (double-sided core on 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 146 east and west) and the extensive use of shading devices. Glazing type is usually double-glazing with low-e coating. Due to high solar radiation angle in these areas, it is also important to shade the roof surface and utilize energy generation systems like solar collectors or PV panels. Placing the work stations along the north and south façade is a good strategy for reducing the electricity demand for artificial lighting. However, the size and position of openings should protect the occupants form direct solar radiation and glare. Using external shading and indoor blinds improves the quality of daylighting. As in a temperate climate, natural ventilation is also an effective solution for reducing the cooling demand in a sub-tropical climate. However, introducing humid outdoor air may reduce thermal comfort of the occupants as a result of which constant humidity control is an essential element of such a strategy. CONCLUSION Design strategies for tall office buildings were investigated through a comparative study of 6 highrises in a temperate and a sub-tropical climate. The total energy consumption of all of the 6 selected cases are considerably less than of typical air-conditioned buildings. This research explained the most effective design strategies that sustainable high-rises using them to reduce the energy consumption for cooling, heating, ventilation and lighting in both climates as the summery is presented in Table 1. It is found that a double-skin façade with automated blinds and operable windows besides a narrow floor plan, the correct placement of core services in regards to solar heat gains, and the application of vertical shafts like atria, which bring daylight and natural ventilation deeper into the plan, are the strategies that effectively can provide energy savings for tall buildings. Table 1. Comparison of the design strategies and the energy performance for the six cases. Temperate Design strategies Commerzbank 30 St Mary Axe Sub-tropical Post Tower Liberty Tower 1 Bligh Street Torre Cube Double-skin facade Deep cavity + + + NA + NA - + + NA + NA Ventilated cavity Natural ventilation Nigh-time ventilation Shading devices Narrow plan Energy recovery Energy absorption Environmental orientation Greenery systems Annual EUI (kWh/m2 gross floor area) + + + NA + NA + + + + + + + - + + - - + + + + + + + + + + + + + NA - - + - + - + + + - - + + - - - - - 139.7 (kWh/m2) 63 - 73.6 (kWh/m2) 42.8 (kWh/m2) 129.5 (kWh/m2) 105.8 (kWh/m2) -0Excluding electricity 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 147 ACKNOWLEDGMENTS We thanks the architects and engineers who responded with case-study information specifically Mr. Nigel Clark, the techical director of HilsonMoran Company by providing detailed operation data of the Mary Axe building. REFERENCES Cook, R., Browning, B., & Garvin, C. (2013). “Sustainability and energy considerations” in D, Parker & A, Wood (eds), The tall buildings reference book, New York: Routledge, pp. 145-155. Dobbelsteen, A. van den. (2012). “High-rise buildings: a contribution to sustainable construction in the city” in H, Meyer & D, Zandbelt (eds), High-rise and the Sustainable City, Amsterdam: Techne Press, pp. 120-147. Goncalves, J., & Bode, K. (2010). Up in the air. CIBSE Journal, December, pp. 32-34. Kato, S., & Chikamoto, T. (2002). Pilot study report: The Liberty Tower of Meiji University, pp. 25. Wood, A., & Salib, R. (2013). Natural ventilation in high-rise office buildings. New York: Routledge. Yudelson, J., & Meyer, U. (2013). The world's greenest buildings. New York: Routledge, pp. 213. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 148 The Potential for Natural Ventilation as a viable Passive Cooling Strategy in Hot Developing Countries Dima Albadra, MSc University of Bath d.albadra@bath.ac.uk Stephen Lo, PhD University of Bath ABSTR ACT Natural ventilation offers opportunities for reducing cooling energy-demand at low-cost in developing countries with limited resources. In this paper the natural day- and night-time ventilation potential for cooling in hot-arid and hot-humid Mediterranean climates is characterised against the key weather and building parameters affecting its performance. In particular, the study seeks to quantify the limits of outdoor environmental conditions under which natural ventilation is an effective strategy for achieving thermal comfort. Furthermore, the study explores the effects of certain building characteristics that enhance the performance of natural ventilation such as ventilation rates and thermal mass. This is achieved by short-term environmental monitoring and dynamic energy modelling of selected naturally ventilated domestic buildings in Lebanon and Jordan. The summer monitoring regime compared external and internal temperatures, relative humidity and air velocity in free-running ‘welldesigned’ buildings in order to identify the external environmental limits for effective day- and nighttime ventilation. Computer modelling of the monitored buildings was undertaken using IES VE to determine the design parameters affecting the performance of natural ventilation. Initial results show that computer modeling overestimate ventilation rates through windows with Venetian shutters. INTRODUCTION Natural ventilation is considered one of the simplest passive cooling strategies that allow the provision of a comfortable indoor environment at low operating costs. The two most beneficial natural ventilation regimes were considered: • Day time ventilation DTV (direct cooling); when buildings are ventilated during the day, and internal temperature and humidity are expected to follow closely external environmental conditions. DTV can enhance occupant comfort through higher indoor air velocities so DTV should only be applied when outdoor temperatures are within comfort limits and acceptable indoor air speed may be achieved, (Givoni, 1998). • Night time ventilation NTV (indirect cooling); where buildings are ventilated only at night can reduce the peak internal temperatures from external levels by pre-cooling the interior exposed thermal mass of a building. Many studies have been conducted to examine the potential of natural ventilation and the effect of different parameters on its performance. De Graca, Chen, Glicksman and Norford (2001), investigated DTV and NTV for an apartment building in Beijing and Shanghai. Shaviv, Yrzioro and Capeluto (2001), assessed the potential of NTV in terms of the reduction in maximum internal temperature as a function Dima Albadra is a PhD candidate at the University of Bath, Bath, UK. Stephen Lo is a senior lecturer in Sustainable Environmental Engineering at the Department of Architecture and civil engineering, University of Bath, Bath, UK. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 149 of the diurnal temperature swing. They concluded that in hot-humid Mediterranean climates a reduction of 3-6° C could be achieved by NTV depending on the level of thermal mass, air change rate and diurnal temperature swing. Artmann, Manz and Heiselberg (2008), conducted a parameter study on the performance of NTV for different moderate climates, building constructions, heat gains, air flow rates and heat transfer coefficients. The study found that even with moderate air change rates, overheating degree hours were considerably reduced by NTV. Yao, Li, Steemers and Short (2009) summarised current research on the potential of natural ventilation in two categories, the first focused on the calculation of ventilation driving forces (wind or stack effect) and the second on calculation of internal air temperature using computer thermal simulation or CFD programs. All the above studies used dynamic computer modelling tools in their assessment of natural ventilation. The vast majority of recent research on real naturally ventilated buildings was conducted to assess air quality or thermal comfort expectations and boundaries and were mostly related to studies on the adaptive thermal comfort theory. Fewer studies report on actual monitored cooling performance of naturally ventilated buildings. Examples of such studies are, Kolokotroni, Webb and Hayes (1998) on the applicability of NTV as a summer cooling strategy for office buildings by taking field temperature measurements. The results found that internal temperature was up to 4°C lower than the external temperature at the start of the following working day; however, due to internal gains the internal temperature had risen again by midday. Givoni (1998) examined fan assisted night ventilation for both shaded and unshaded windows in San Diego with three different levels of thermal mass levels, (light, medium and heavyweight). The study found that higher thermal mass lowered the maximum internal temperatures by 2°C and that they remained lower than external air temperatures throughout the day. However, buildings with low thermal mass failed to reduce internal temperatures below outside conditions and in some cases interior temperatures were higher. Based on this and several other studies on natural ventilation Givoni developed building bioclimatic charts BBCC with recommended boundaries for the external environmental conditions under which natural ventilation would be an effective strategy. It is well established that there are also key supporting factors for successful natural ventilation, namely adequate shading to limit unwanted excess solar heat gains, limiting internal heat gains from appliances, appropriate thermal mass and insulation. However, there is a lack of post-occupancy performance data from naturally ventilated buildings in hot climates where such strategies were applied. Most post-occupancy studies focus on the energy consumption of buildings rather than their passive performance. Computer simulation tools are widely used to predict the performance of natural ventilation without validation through real building monitoring. This paper reports on the monitoring of two naturally ventilated, best practice, new domestic buildings. Short-term environmental performance monitoring was conducted to establish the true natural ventilation potential of ‘well-designed’ buildings. The monitoring results were compared with the natural ventilation potential suggested by Givoni’s BBCC, and then the performance predicted by Dynamic building simulation tool IES. METHODS: The selected buildings: 1. Aqabba house AREE: a three story house, with a total floor area is 235 m2. The house was designed to be a prototype for low energy houses in the Aqabba region in hot-arid climates of Southern Jordan; it has 45 cm thick cavity walls of concrete blocks. 2. Casa Batroun: First and only BREEAM excellent awarded house in in the hot-humid climate of the eastern Mediterranean region of Lebanon, The house is separated into a ground floor flat and a 1st floor flat with two different constructions, masonry and timber, respectively. Both buildings were well shaded and had external Venetian shutters installed on the majority of windows, figure1. The monitored buildings were also unoccupied for the majority of the time except for night; however, kitchen appliances such as fridge and freezer remained on during the monitoring period and thus contibuting to internal gains. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 150 The monitoring methodology: Given that both buildings were classified as heavy weight, the potential benefits of thermal mass combined with day-time cross ventilation and night-time cross ventilation were examined. In order to measure the outdoor microclimate the following external environmental parameters were monitored: i) air temperature, ii) relative humidity, iii) wind speed and direction, and, iv) solar radiation on the horizontal plane. The internal measurements monitored were; relative humidity, air temperature and air velocity. Wireless sensors were placed on tripods at a height of 1.2m, (the normal seated height of occupants) in the middle of the rooms and transmitted to a data logger. The parameters were sampled at one minute intervals with averages recorded every 30 minutes. The monitoring period lasted for 10 to 14 days in each building depending on accessibility to the buildings, allowing 5 to 7 days for each of the two ventilation strategies. In DTV mode, windows were opened from 7am to 8pm and in NTV mode, from 10pm in AREE, and 8pm in Casa Batroun until 7am the following day. In order to establish the effectiveness of the two strategies the following performance indicators were examined; reduction in peak internal temperatures, and reduction in Kelvin hours (KhR) between external and internal temperatures over the period of ventilation, calculated in Kelvin hours (Kh). Figure 1 AREE (left), Casa Batroun (right). Computer modelling of the monitored buildings: Both monitored buildings were modelled for DTV and NTV performances using dynamic computer simulation. There are several programs and detailed thermal simulation tools able to model the energy performance of buildings. Computer software Virtual Environment IES was used to undertake the simulation as it not only is the approved energy software in the United Kingdom for energy analysis and Part L regulations (IES VE, 2010), but is also extensively used internationally (Altamimi & Fadzil, 2011; Rajagoplan & Luther, 2013; Blight & Coley 2013). The program comprises an integrated suite of applications, only four of which were used in this modelling work; ModelIT for the basic building geometry, SunCast for shading analysis, ApacheSim for thermal simulation, and Macroflo for building ventilation. A performance gap is frequently reported in literature between predicted and monitored energy use in buildings. Although the focus of this paper is on natural ventilation and not energy use, the following observations are still valid. One of the main reasons for the discrepancy between predicted and actual building energy performance is due to poor assumptions regarding occupants’ behaviour/patterns or issues with the built quality, (Menezes, Cripps, Douchlaghem & Buswell 2012). Other reasons include unreliability of weather files used, with various studies showing that using different weather files resulted in different simulation outputs. However, it is generally agreed that TMY2 weather file gave the closest output match to measured consumption (Crawley 1998; Michopoulos, Voulgari, Papakostas & Kyriakis 2012). To limit such possible discrepancies the following measures were taken: - Weather files used in the simulation were edited using a software ‘epw creator’. The new files corresponded to onsite measured weather variables; DBT, RH, Wind speed and direction and Global 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 151 solar radiation. Diffused and direct normal radiation were calculated from the measured data and included in the new weather files. - Occupancy profiles: Both monitored buildings were unoccupied for the majority of time, at durations were there was limited occupancy, detailed use of the building was recorded and reflected in the model. - Information on the buildings’ material specifications and constructions were provided by the architect and consultants (Karkoor, Visser 2013). However, issues with poor build quality and inaccuracies in the specifications of locally produced materials may still exist. RESULTS: DTV & NTV monitoring of AREE: Temperature °C Main observations, figure 2: 1) Up to 7K reduction in maximum temperature from external maxima with DTV and up to 9K with NTV. The lowest recorded temperatures were on the ground floor in the north-east room L0R3. The highest recorded internal temperatures were in room L1R4 on the 1st floor corridor, which had a south-west-west facing window and higher glazing ratio than other rooms. Additionally, this room had an external overhang while most other rooms had Venetian shutters. 2) Using NTV the day time internal temperature increased by only 2K compared to night time levels. To quantify and compare the overall performance, the Reductions in Kelvin degree hours (KhR) between internal and external temperatures were calculated and are shown in figure 3. Higher values of Kelvin degree hours indicate better ventilation/cooling performance. GFGF Figure 2 a) Internal peak temperatures for DTV (left), b) for NTV (right) at AREE. It is apparent from figures 2 & 3 that NTV has a better cooling potential than DTV for the hot arid climate of Aqabba, because of lower peak temperatures and more extended periods of reduction from outdoor temperatures. However, even with NTV internal temperatures remained above 30°C at all times during the monitoring period in peak summer. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 152 Kelvin degree hour reduction from external temperature levels, DTV, NTV average Figure 3 KhR=53.2Kh, 72Kh respectively, a 36.8% increase from DTV. DTV & NTV monitoring of Casa Batroun: Temperature °C Similar observations to AREE were made in terms of the reduction in internal temperature. However, more interestingly the performance of the two different floors differed significantly. Temperatures in the top flat were higher than the ground floor flat, (shown in figure 4), because; a) the ground floor was better shaded by nearby trees and adjacent buildings, b) the 1st floor had a higher exposed surface area resulting in additional solar gains through the roof, and walls, c) the effect of the ground floor’s higher thermal mass in storing coolth was more noticeable with NTV, where day time internal temperatures remained only 2K higher than night time levels. As for KhR an average KhR=44.5Kh, 61Kh for DTV and NTV, respectively, which is a 37% increase for NTV from DTV. Temperature °C Internal temperatures in the Ground floor (masonry construction) in red, 1st floor (timber Figure 4 construction) in green at Casa Batroun, a) DTV (right), b) NTV (left). Figure 5 a) internal peak temperatures for DTV (left); b) for NTV (right) at Casa Batroun. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 153 The monitored data of both buildings indicated that, the internal relative humidity followed external RH very closely. Recorded airspeed in the middle of the rooms that had Venetian shutters was minimal, while a maximum of 0.4m/s was recorded in a room that had an open terrace door with no shutters. Such low air velocities mean that relying on naturally driven air movement to provide cooling sensation for occupants is not possible. Building Bioclimatic charts BBCC: BBCC offer an easy method for initial assessment of the potential of a passive design strategy at early design stages. It suggests a comfort zone and the boundaries of climatic conditions within which DTV and NTV and other passive strategies can provide comfort. The boundaries are plotted on a conventional psychrometric chart. The best-known BBCCs are those developed by Givoni (lomas, Fiala, Cook & Cropper 2004). In this study, although internal temperatures remained above comfort levels for the majority of the time, the enhancement of internal conditions when adopting DTV strategy is much higher than that expected in conventional buildings, or of that suggested in Givoni’s BBCC; where it was limited to only 2K reduction. One possible explanation for this discrepancy between the findings of Givoni and those observed in this study, is that Givoni based his work on the monitoring of ‘thermally heavy’ buildings which had only 10cm thick concrete walls with insulation (Givoni, 1988), while AREE and Casa Batroun had about 45cm thick walls. On the other hand, for NTV the BBCC suggests up to 8K extension of the comfort zone which is consistent with the results reported in this paper. Therefore, based on the initial results of this study, high mass, well insulated and well shaded buildings, with limited internal heat gains would exhibit boundaries for DTV and NTV as shown in figure 6. Figure 6 Givoni’s BBCC for developing countries with new suggested boundaries for DTV. IES modeling results: The initial analysis of DTV performance has resulted in the model over predicting the rise in internal temperature in both models (figure 7a). In other words, the buildings performed better in reality than predicted by the simulation software. Similarly in NTV analysis (figure 7b), predicted temperatures dropped down below measured values during ventilation periods and increased above monitored levels during day-time. Clearly there was a discrepancy between measured and modelled results and the building seemed to perform better than predicted through simulation. The following sensitivity analysis was undertaken to identify the reasons behind such a gap. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 154 Temperature °C comparison between modelled and monitored results for one room in AREE, a) DTV (left), Figure 7 b) NTV (right). Sensitivity analysis: 1. Thermal mass: a range of thermal capacities were modelled, by increasing either the specific heat capacity of materials or their density, each material specification was increased to its highest realistic value as given in CIBSE guide A. Although such increases resulted in aligning the monitored and modelled temperatures closer when windows were closed, it had no significant impact when windows were opened. During ventilation hours, the model behaved in a similar way to the base case model. This indicated that the main issue could be in higher predicted ventilation rates. 2. Ventilation rates: most windows in both buildings had louvered Venetian shutters installed in addition to bug meshes, which will have an impact on the airflow rate through these windows. Macroflo allows the user to choose a window type for each window; louvered windows were chosen for windows with Venetian shutters. Data inputs required were the openable area and the discharge coefficient CD. A discharge coefficient relates the volume flow rate through an orifice to its area and the applied pressure difference (Karave, Stathopoulos & Athientis 2007). Several parameters affect the CD; these parameters are the opening area, wind speed, wind incident angle, and location of the opening in the façade (Karava, Stathopoulos & Athientis 2004). Due to the difficulty in determining the CD without testing, a range of discharge coefficients were considered. The discharge coefficient for a sharp edged orifice such as a window is usually taken as approximately 0.6 to 0.65 (ASHRAE fundamentals 2009), manufacturers’ data showed that louvered ventilators CD ranged from 0.3 to 0.1 (Renson 2009; Architectural louvers 2007). Therefore, a series of simulations with different CD was conducted. 5 cases were studied with CD equaling 0.4, 0.3, 0.2, 0.1 and 0.05. For every 0.1 reduction in CD, a reduction of approximately 25% was achieved in air change rates ach for the house. However, not even the lowest ventilation rates achieved (8 to 20ach, for CD = 0.05) brought the predicted internal temperatures significantly closer to the measured ones, figure 8 (a). Going below CD = 0.05 is realistically not possible as it results in an effective orifice area < 3% of window area. Figure 9 (a) shows the airflow in L0R5 for different CD, (b) shows airflow in different rooms for CD = 0.3. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 155 Figure 8 AREE DTV measured and modelled internal temperatures, a (left), b (right). Figure 9 AREE DTV a) airflows l/s in L0R5 (left), b) air flows l/s for CD = 0.3 (right). IES provides an option where Macroflo can be turned off and ventilation introduced through ApacheSim as natural or auxiliary ventilation. In order to determine the ventilation rate that would bring predicted temperatures down to the monitored levels; natural ventilation was introduced starting from 1ach to 10ach in addition to a continuous infiltration rate equal to 1ach. As could be seen in figure 8b, higher ventilation rates resulted in higher internal temperatures and the best match to monitored internal temperatures was achieved for 1ach. Additionally, low ventilation rates resulted in a mismatch in some observed ventilation patterns, such as the drop in internal temperature early in the morning when windows are first opened, as seen in Figure 8b. DISCUSSION: It was not possible to determine the actual ventilation rates in the monitored buildings as no such measurements were taken onsite. However, the recorded airspeeds in the middle of rooms indicated that there was a considerable air movement in some rooms which had no Venetian shutters, yet internal temperatures were much lower than predicted by the model. The model only achieved similar internal temperatures to the measured ones when there was very little to no ventilation. This indicates that whenever the modelled building is ventilated, internal temperatures closely follow external DBT regardless of the levels of thermal mass or ventilation rates. Additionally, it is clear from the sensitivity analysis that there is an issue in representing open windows in computer dynamic simulation tools such as IES because very high ventilation rates were predicted even for very small CD. A study by Coley (2008) on top hung window performance in IES found that the ventilation rate depends, to a great extent, on how the windows are represented, whether as a vertical ‘arrow slit’ hole in the wall or as a horizontal ‘letter box’ opening. False representation can result in up to four times higher airflows. Although IES provides an option for windows with louvers, their representation is not clear. Other parameters may be causing this performance gap and require further detailed investigation. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 156 CONCLUSIONS: The study presented investigated the free-running performance of two well-designed buildings in hot climates. The study found that up to 6K reduction could be expected for DTV and 9K for NTV. These findings were compared to previous studies and plotted on BBCC. Using computer modelling tools is a widely used approach for evaluating the performance of natural ventilation. Therefore, the monitored buildings were then modelled using computer software IES VE, in order to validate such an approach. The initial modelling of both buildings predicted higher internal temperatures than experienced in reality. A sensitivity analysis for AREE was conducted, thermal mass and ventilation rates were analysed. It was found that a ventilation rate of 1ach gave the closest match to monitored values. Low air-change rates however, were not possible to achieve in IES through opening windows in Macroflo, but rather it should be introduced in the thermal model ApacheSim as a fixed natural ventilation rate. Further investigation is required in order to fully understand how best to represent complex windows geometries in computer models, as this variable has a great impact on ventilation rates and consequently the internal temperatures. Further research is needed to identify other key factors responsible for the performance gap. REFERENCES Al-Tamimi, N & Fadzil, S. (2011). The potential of shading devices for temperature reduction in highrise residential buildings in the tropics. Procedia Engineering, 21:273–282 Architectural louvers., (2007). Louver Performance data. Retrived June 8, 2014 from: http://www.archlouvers.com/louvers/drawings/E2JS.pdf Artmann, A., Manz, H., & Heiselberg, P. (2008). Parameter study on performance of building cooling by night-time ventilation. Renewable energy,33: 2589-2598 ASHRAE. (2009). ASHRAE Handbook-Fundamentals. Atlanta: American Society of Heating Refrigeration and Air Conditioning Engineers, Inc. Blight, T & Coley, D. (2013). Sensitivity analysis of the effect of occupant behaviour on the energy consumption of passive house dwellings. Energy and Buildings, 66: 183-192 CIBSE, (2006). CIBSE Guide A – Environmental design. Norfolk: CIBSE Publications. Coley D. (2008). Representing top hung windows in thermal models. International journal of ventilation, 7: 151-158 Crawley, D. (1998). Which weather data should you use for energy simulations of commercial buildings? ASHRAE 1998 Transactions, 104 part 2 De Graca, G., Chen, Q., Glicksman, L., & Norford, L. (2001). Simulation of wind-driven ventilative cooling systems for an apartment building in Beijing and Shanghai. Energy and building , 1-11. Givoni, B. (1998). Climate considerations in building and urban design. New York: John Wiley & sons IES VE Integrated environmental solutions, (2010). Free IES VE-SBEM Approved for 2010 Regulations. Retrieved June 9, 2014 from: http://www.iesve.com/news/free-ies-ve-sbem-approved-for-2010regulations_1674_/corporate Karava, P., Stathopoulos, T & Athienitis, A. (2007). Wind-induced natural ventilation analysis. Solar Energy, 81: 20-30 Karava, P., Stathopoulos, T & Athienitis, A. (2004). Wind driven flow through openings: A review of discharge coefficients. International Journal of Ventilation, 3: 255-266 Karkour, M., (maya@ecoconsulting.net), September 02, 2013. Casa Batroun. Email to Albadra. D (dima.albadra@gmail.com). Kolokotroni M, Webb BC, Hayes SD. (1998). Summer cooling with night ventilation for office buildings in moderate climates. Energy buildings, 27: 231-237. Lomas, K., Fiala, D., Cook, M., Cropper, P. (2004). Building bioclimatic charts for non-domestic buildings and passive downdraught evaporative cooling. Building and environment, 39: 661-676. Menezes, A., Cripps, A., Douchlaghem, D., Buswell, R.. (2012). Predicted vs. actual energy performance of non-domestic buildings: Using post-occupancy evaluation data to reduce the performance gap. Applied energy, 97: 335-364. Michopoulos A, Voulgari V, Papakostas K, Kyriakis N. (2012). Evaluation of different weather files on energy analysis of Buildings. International journal of energy and environment, 3, pp.195-208. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 157 Rajagopalan P & Luther, M., (2013). Thermal and ventilation performance of a naturally ventilated sport hall within an aquatic centre. Energy and Buildings, 58: 111- 122 Renson., (2009). Louvre panels and grilles, retrived June 8, 2014 from: http://www.alunor.no/images/Marketing/kataloger/Renson/louvres_eng.pdf Shaviv, E., Yezioro, A., & Capeluto, I. (2001). Thermal mass and night ventilation as passive cooling design strategy. Renewable energy, 24: 445-452. Visser, F., (florentine_jordan@yahoo.com), January 07, 2013. AREE. Email to Albadra. D (dima.albadra@gmail.com). Yao, R., Li, B., Steemers, K., Short, A. (2009). Assessing the natural ventilation cooling potential of office buildings in different climatic zones in China. Renewable Energy, 34: 2697-2705 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 158 Shop Window Lighting: the Use of Sun to Improve Visual Appeal and Reduce Energy Demand Elena Anagnou, PhD Helena Coch, PhD Isabel Crespo, PhD [All from Architecture and Energy Group, School of Architecture of Barcelona– Polytechinic University of Catalonia (UPC), Barcelona, Spain] ABSTRACT The present study deals with the potential reduction of energy consumption for the lighting of shop window displays. Urban commerce has a very high impact on economics and at the same time, it is a highly energy-consuming sector. Light has the power of attracting people´s attention, which is one of the goals of the trading and selling activity and very high illuminance levels are usually recommended. In the Mediterranean areas, where daytime lasts for many hours, commercial activity takes place mainly under sunshine conditions and shop windows frequently fail to fulfil their main corporative goal, namely the unobstructed observation of the products exhibited. The necessary increase of artificial lighting illuminance levels to accent interior light conditions, due to extremely high-luminance urban surroundings, leads to an important increase of energy consumption, as a common solution. Nevertheless, the results are generally very poor, because reflections and other kinds of visual problems still defy solution and the final result is an economic and energy waste during daytime, especially in low latitude countries. The present study evaluates the visual and energetic benefits of an innovative passive design that obstructs solar rays and redirects them into the interior of the shop window scene. A scale model of this new design confirms the visual benefits produced by its use, via the different luminance maps tested. This new lighting passive system results in a very simple, effective and low cost solution that can be easily applied in economically emerging countries. The most important fact is that high illuminance levels are achieved and, simultaneously, there is important energy reduction, taking advantage of natural light instead of competing with sun power. INTRODUCTION Urban commerce and storefronts play a major role in the economy and the very image of a city, especially in climates that favour exterior human activities. Being a fundamental element of the commercial process, a proper visual presentation of shop windows is crucial. Lighting contributes to the creation of a corporate identity, ideally, it assists the production of the image desired by the trader and together with other components of the design it contributes to sales promotion (Rea, 2000). The present study analyzes the lighting requirements and visual inconveniences of window displays associated with the Mediterranean climate and how they are or should be dealt with, so that the aesthetic and commercial function of shop windows will not be inhibited. Elena Anagnou is a PhD holder in architecture. Helena Coch is a professor in the Department of Architectonical Construction I, Universitat Politècnica de Catalunya, Barcelona, Spain. Isabel Crespo is a professor in the Department of Graphical Expression in Architecture I, Universitat Politècnica de Catalunya, Barcelona, Spain. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 159 The complexity of light presentation of these everyday urban elements is linked to their exposure to varying light conditions in the outside environment. Being a part of the urban landscape, the illuminance levels in the interior of shop windows should be corresponding to the surrounding light, so that they would be able to stand up and yield good vision of the products on display. In Mediterranean climates, the temperature favours urban commerce, since duration of daily solar radiation is extensive and solar intensity is prevalent throughout the whole year. As a result of the very high level of exterior illuminance, though, the visual goals of window displays are often impeded. High energy consumption yet poor visual outcome The lighting design of window displays is required to perform multiple visual tasks, most important of which are to attract clients and to provide visual comfort and good vision of the products exhibited. In order to do so, two basic factors, among others, must be secured: adequate illuminance levels and luminance contrast between the display objects and their background. A recommended value for luminance ratio between the illuminated object and its surrounding area is 3:1, although this ratio can reach values from 10:1 to 30:1 (Carillo, et al., 2010). Regarding illuminance values, they can be as high as 5.000 lx during night-time and 10.000 lx during daytime (Freyssinier, Frering, Taylor, Narendran, & Rizzo, 2006). Moreover, certain studies prove that a higher illuminance level implies that more passersby are attracted by the shop windows (Carillo, et al., 2010). As a result, the consumption of energy for the lighting of window displays that remains on for at least 15 hours per day (Freyssinier, Frering, Taylor, Narendran, & Rizzo, 2006) is rather high. In Catalonia, window display lighting constitutes 11.2% of the total energy consumption for the lighting of a typical store (Institut Catalá d´Energia, 2008). Nevertheless, despite the high energy consumption, visual inconveniences insist; during the research discussed in this paper, a study of a representative sample of shop windows of Barcelona has been performed in its natural opposing lighting conditions, day and night, analyzing their visual results. The basic tools of this analysis have been the luminance maps and evaluation, via computer application, of the odds of glare and its possible sources. The study has confirmed and identified the main inconveniences that can impair the visual result and, therefore, the commercial purposes, during daytime hours. Specifically, the most frequent problems are: penetration of exterior light into the scene, which affects the balance of the luminances projected, failure to reach the appropriate contrasts for the projection of exhibits, as well as glare and annoying reflections. Among the analysed stores, as shown in Figure 1, the visual presentation of 65% of them failed to reach a contrast of 3:1 between all exhibits and their surroundings, as well as 65% of them failed to avoid annoying reflections. In addition, the conducted analysis confirms that the commercial storefronts mostly affected by such inconveniences, especially by the generation of glare, are those receiving direct solar incidence, even in cases that solar protection is provided. This effect can be generated not only by solar incidence in the actual presentation of the window, but also by the incidence in immediately adjacent areas, whether on the surrounding facade or the sidewalk that lies before it. In order to help face these visual disadvantages resulting from very high levels of exterior illuminance, a remedy often applied is to increase artificial illuminance projected during the day, in contrast to the one projected overnight. However, the intensity of sunlight in low latitude areas is so high that artificial lighting can never compete with the natural one. The inability of artificial lighting to compete with high solar illuminance makes it vain for any power to increase during the day, along with the corresponding energy additional charge this practice entails. This new study, instead, evaluates the benefits of the use of a passive design that takes advantage of the very solar ray incidence in order to increase interior illuminance effectively and at the same time reduce energy consumption. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 160 Figure 1 (a) Part of examined window displays that failed to reach contrast 3:1 between all exhibits and their surroundings, plus an example (b) Part of examined window displays with annoying reflections, plus an example. METHODOLOGY The present research evaluates the visual and energetic benefits of an innovative passive design that obstructs solar rays and redirects them into the interior of the shop window scene. In fact, the improvement of the quality of the observation of exhibits during daytime is assessed, along with the possible reduction in energy consumption. The research implementation is summarized in 4 main steps which are presented below and explained in the following subsections: 1. Design of a component that manages to obstruct solar radiation and re-direct it towards the interior of the scene. 2. Development of two scale models of the window analyzed, the one being independent of and the other one based on the proposed system. 3. Simulation of the unfavourable exterior light conditions. 4. Analysis of visual differentiation of the window display in the same lighting conditions, depending on the existence of the proposed system, or not. Development of the passive design component A typical window display of Barcelona has been selected among the sample observed, that is about four meters high and one meter wide. Based on the solar diagram of the city, several design tests of a system of reflective surfaces have been performed in order that, when the latter is incorporated in a typical window facing south, it could make the most of the solar incidence during the most unfavourable conditions; these are produced during summer solstice, as it has been confirmed after analysis in a simulation program that incorporates solar trajectories (Heliodon). The possible contribution of both flat 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 161 and curved reflective surfaces was examined in various positions, with the purpose of introducing into the showcase as much light as possible, especially during summer. The design finally proposed here, is based on two successive light reflections, as shown in Figure 2, where typical dimensions can also be appreciated. Rays that fall upon the curved surface will be reflected symmetrically around the axis joining the point of incidence and its centre of curvature (Peoglos, Raptis, & Christodoulides, 2004); subsequently, they will be cast on the plane surface and will be reflected towards the interior of the scene. Regarding the concave surface, in order to directly introduce as much light as possible, the use of a specular material is proposed. On the contrary, for avoiding intense reflections in the scene that, depending on sky conditions, may adversely affect the visual result, the use of a diffuse-reflection surface is proposed for the second reflection. Figure 2 (a) Geometry of the component proposed (b) Conduction of natural light in the interior of the scene. A vertical movable lamina should be placed in order to control the system and, depending on the solar height, to regulate the solar radiation introduced through the opening of the system. In addition, a fixed glass surface could help to avoid problems of water or dirt accumulation. In any case that this design is implemented, it will have to adjust to any special needs. Development of the scale models Two scale models of the window analyzed, the one being independent of and the other one based on the proposed system, have been developed. As in the present study the effect of chromatic contrast is not analyzed, all the components of the scene have been chosen to be of soft and slightly contrasting colours. The dimensions of the models have been adjusted to the artificial lighting projected, giving special care so that both the area of the surfaces and the luminous flux they receive, would be correctly scaled. In each one of the models two LED sources have been placed above two mannequins and 30cm of led stripe above the background of the scene, as shown in Figure 3. Because of their size and characteristics, the sources Prolight PM2B 3LxS-SD-3W Power LED have been chosen; in order to simulate the illuminance of the real window display, after measuring their luminous intensity and applying the basic laws of photometry, the appropriate scale of the model has been defined: (1) For wall washing lighting simulation, a strip led of 40W type 3528 has been used in each model. A potentiometer has also been connected to it, so that its intensity could be dimmed and, in this manner, possible reduction of energy for vertical illumination could be evaluated. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 162 Figure 3 Model based on the system proposed. Simulation of the desired exterior light conditions The study took place in Athens on December 10, during the time of zenith. In order to simulate the summer solstice in Barcelona (72° during the zenith) the table where each model was placed has been rotated through 43°, as shown in Figure 4, equal to the difference of solar altitude on December 10 in Athens (29o during the zenith). Furthermore, in order to obtain an overall assessment of the results for the entire year, a simulation of the equinox and winter solstice in Barcelona (49º and 26º correspondingly during the zenith) were simulated rotating the table through 20º, and -3º correspondingly. Figure 4 Simulation of the desired exterior lighting conditions by rotating through 43o the table where the model is placed. Evaluation of visual differentiation, dependeding on the existence of the proposed system The comparison of the visual outcome of the two different models has been realized via the observation of photographs and their equivalent luminance maps, in addition to the observation in situ. The comparison took into account the following visual inconveniences: 1. Annoying reflections. 2. Achievement of 3:1 contrast between products and their surrounding area. 3. Glare. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 163 In order to compare the possibility of glare in every situation, the Radiance based tool, called Evalglare, which evaluates that possibility through the daylight glare probability index, DGP (Wienold, 2009) has been used in addition to the luminance maps. RESULTS AND DISCUSSION Improvement of visual appeal The comparison of the visual result of the two scenes, the one being independent of and the other one based on the proposed system, has resulted in the expected improvement. Via the observation of the photographs and the luminance maps deriving from them, the positive effect of the redirection of the sunlight into the scene is appreciated, as shown in Figure 5. Specifically, when comparing the two models, the following parameters are observed: 1. The annoying reflections on the glass of the scene, after the incorporation of the proposed system, are significantly moderated. Looking at the model without the system (on the left) one can see the reflection of the photographer´s figure; however, looking at the model with the proposed design incorporated (on the right) this effect does no longer exist. 2. The desired contrast of 3:1 between the products and their surrounding area is achieved in the model where the proposed system has been incorporated, while simultaneously solar incidence upon products is avoided. 3. Large surfaces of high luminance that can dazzle there are in the model where the proposed design is not used; on the contrary, in the model based on the proposed system the area and luminance of these surfaces is moderated, thus glare is less probable. The comparison of the DGP index in the two cases is indicative of the improvement. Figure 5 (a) Photograph and luminance map (b) DGP index. Comparison of the two models, without (left) and with (right) the system proposed. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 164 Concerning the rest of the seasons, the design proposed has no negative effects. Instead, during the research discussed in this paper, the effect against the inconveniences of glare and annoying reflections has been examined and proven positive throughout the year. Energy demand reduction As predicted, the visual result of the model, following the incorporation of the system proposed is significantly improved. That alone signifies that the frequently applied technique of increasing the power of the lighting of window displays, is a remedy that can be abandoned along with the useless energy charge it entails; therefore, energy consumption for the lighting of display windows can be reduced. In addition, the possibility of further energy reduction has been verified with the help of the potentiometer connected to the vertical illumination of the model: The illuminance of the background lighting of the scene has been decreased, in the case of the design proposed, to the 1/3 of the illuminance projected before. Once again, when examined in the simulated lighting conditions, even with a drastic reduction of power applied the result has been positive, as shown in Figure 6. The comparison of the visual result of the two scenes, the one being independent of and the other one based on the proposed design with a power decrease to the 1/3 for vertical lighting, has resulted in improvement. Specifically, when comparing the two models, the following parameters are observed in the one based on the system proposed despite the power reduction: 1. The reflections of the glass of the scene are, once again, significantly moderated. 2. The desired contrast of 3:1 between the products and their surrounding area is achieved, while solar incidence upon them is avoided. 3. Fewer areas of high luminance that can dazzle are generated. The comparison of DGP index in the two cases is indicative of the improvement. Figure 6 (a) Photograph and luminance map (b) DGP index. Comparison of the two models, without (left) and with (right) the system proposed when wall wash lighting is reduced to 1/3. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 165 The effects observed above are the expected ones, since solar light intensity which is introduced in the scene, and not artificial lighting, is the one increasing significantly the illuminance in it. In that way, it is verified that under the most negative conditions for the visual result of shop windows, which arise during solar incidence, not only does the vertical lighting not have to be increased but, in contrast, it can be drastically reduced while, simultaneously, the visual outcome can be significantly improved. CONCLUSION This paper results in a passive design proposal that can offer very positive effects on the visual result of shop windows, improving the quality of their observation, thus fulfilling their commercial visual goals during the day and, in addition, reducing energy consumption. In low latitude countries, the inability of artificial lighting to compete with high solar illuminance makes it vain for any power to increase during daytime, along with the corresponding energy charge this practice entails. Instead, the increase of interior illuminance levels of shop windows by using the very illuminance of the sun may improve the visual outcome and reduce energy consumption for the lighting of the scene. The positioning of the proposed design of sunlight redirection, with the dual effect of both obstruction of its incidence and the benefit of its intensity to increase illuminance in the interior of the shop window, has very positive effects on its visual presentation, especially in terms of reducing annoying reflections and the possibility of glare. Moreover, these effects are valid even when reducing the installed power for the lighting of the background of the scene, thus making possible, in combination with the use of sensors and resistors, further reduction of energy consumption for the lighting of the display window. This new lighting passive system results in a very simple, effective and low cost solution that can be easily applied in economically emerging countries, with considerable environmental and economic benefits when incorporated in the highly energy-consuming display windows. Therefore, it is deduced that, instead of trying to compete with the sun, it is better for one to ally with it. ACKNOWLEDGMENTS The authors would like to thank Axel Jacobs for providing the possibility of luminance maps production using their software online (http://www.jaloxa.eu/webhdr/) and Benoit Beckers for providing the Heliodon2 software. This work has been conducted with the support of the Diputació de Barcelona. This paper is supported by the Spanish MEC under project BIA2013-45597-R. NOMENCLATURE dM = dR = DGP dimensions of the model dimensions of the real window display = Daylight Glare Probability REFERENCES Carillo, C., Diaz-Dorado, E., Cidrás, J., Bouza-Pregal, A., Falcón, P., Fernández, A., & ÁlvarezSánchez, A. 2013. Lighting control system based on digital camera for energy saving in shop windows. Energy and Buildings 59: 143–151 Freyssinier, J. P., Frering, D., Taylor, J., Narendran, N., & Rizzo, P. 2006. Reducing light energy use in retail display windows. Sixth International Conference on Solid State Lighting. San Diego: Proceeding of SPIE. Institut Catalá d´Energia. 2008. Caracterització del consum energétic en els establiments del sector de pimes de comerç i serveis de Catalunya. Barcelona. Peoglos, B., Raptis, I., & Christodoulides, C. 2004. Optical Instruments. In B. Peoglos, I. Raptis, & C. Christodoulides, Experimental Physics Techniques. Athens: National Technical University of Athens. 3: 39-82. Rea, M. 2000. IESNA Lighting Handbook: Reference & application. 9th Edition. New York: Illumination Engineering Society of North America. Wienold, J. 2009. Dynamic daylight glare evaluation. IBPSA Conference. Glasgow: Building Simulation. pp. 944-951 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 166 Developing Bioclimatic Zones and Passive Solar Design Strategies for Nepal Susanne Bodach, M.Sc. Technische Universität München, Germany Susanne.bodach@gmail.com ABSTRACT Nepal displays a highly varying topography which is leading to a variety of climatic conditions. With the introduction of modern construction technologies in the country, the building sector has adopted uniform design and building techniques that often neglects local climate and rely on energy-intensive mechanical means to provide thermal indoor comfort. The definition of a climate classifications for building design can be an important decision making tool towards climate-responsive and energyefficient architecture. This paper represents the groundwork for developing bioclimatic zones for building design in Nepal. Based on climatic maps areas of similar climatic conditions were identified. Climate data of various locations within these zone were collected and analysed. A bioclimatic approach was adopted using the psychrometric chart in order to identify passive design strategies for each locations. Finally, an overview of appropriated design strategies for summer and winter for each zones is developed. INTRODUCTION Climate-responsive design is considered to be one of the major requirements to drive the building sector towards sustainable development (Szokolay, 2008). However, architects and building planners are still guided by universal design style that is rather focusing on form language and neglecting the local climate conditions (Liedl, Hausladen, & Saldanha, 2012). Climate classification for sustainable building design can fill the gap and guide building professionals which design strategies are suitable in a certain climate context. Many countries that have a variety of climates within their territory have developed a climatic classification for building design - also called bioclimatic zoning. Climate maps for Nepal have been developed based on physiological features and vegetation. However, no building design specific climate zoning is available for the country. Few authors identified the climate-responsive design strategies for specific locations in the country (Upadhyay, Yoshida, & Rijal, 2006). This research is the first comprehensive study aiming to provide the groundwork for developing a bioclimatic zoning and the appropriate design strategies for the whole territory of Nepal. There are several approaches to define a climate classification for building design. Givoni (1969) distinguishes between four main climate classes, namely hot, warm-temperate, cool-temperate and cold climates; using sub classification he elaborated a total number of eleven climate types for the whole planet. Koenigsberger (1974) developed six climate zones for building design in the tropics based on the two climate factors, temperature and humidity; these factors dominantly influences thermal comfort. Many countries with high climatic variations have developed their own climate zones for building design 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 167 which is often used for defining thermal performance standards for buildings. There is no universal approach for the definition of such a zoning. Most classifications use climate variables (such as temperature, humidity, precipitation, solar radiation, wind conditions) as main criteria (Table 1). Building design factors, e.g. heating and cooling degree day, effective temperature, temperature swing or passive design strategies, are often used as secondary criteria or in combination with climate variables. There are two countries (Argentina, Brazil) that have used only passive design criteria to define the bioclimatic zoning. In some classifications topographical criteria like latitude, longitude, altitude and distance to the coast are added to differentiate bioclimatic zones. Most classifications use the bio-climate chart to identify passive solar design strategies for the different zones. The way of defining climate classification is from country to country different and few interesting examples are described in detail in the following. USA’s climate zones are developed based on the need for heating and cooling using the amount of heating degree days (HDD) and cooling degree days (CDD) (ASHRAE, 2007). China uses the mean temperature in the hottest and coldest month as main criteria. Complementary criteria is the number of days that average temperature is below 5 °C or above 25 °C (Lam, Yang, & Liu, 2006). For the development of the Indian zoning monthly climate data of mean temperature, relative humidity, precipitation and number of clear days were analysed from 233 meteorological stations of the country (Bansal & Minke, 1988). Defined climate conditions must prevail for more than six month; otherwise the location is classified as composite climate. Brazil, being the fifth largest country on the globe with a range in latitude of about 40°, has developed bioclimatic zones adopting the bioclimatic chart from Givoni for hot developing countries. Climate data from 330 locations was plotted on the chart to identify passive design strategies for each location. According to the predominant design recommendations these locations were grouped into different climate classes resulting into eight bioclimatic zones (ABNT, 2003). Argentine has used three indicators to define the zoning: 1. Heating degree days (HDD); 2. Effective temperature (ET) on a typical summer day; 3. Average daily thermal swing for the relevance of thermal mass. The HDD is a key indicator for the heating demand in winter dividing the country in the six main zones. The temperature swing being an indicator for the incorporation of thermal mass is used to classify the four warmer zones into 12 subzones. The two colder zones are sub-classified indicating the potential for passive solar heating Table 1. Climate Classifications for Building Design and Used Criteria Target region World Egypt India USA China California Tropics North-east India Chile World Peru Venezuela Brazil Argentine Climate X X X X X X X X X X X X Used criteria Passive Topography design X X X X X X X X X X X X X X Source (Givoni, 1969) (Mahmoud, 2011) (Bansal & Minke, 1988) (ASHRAE, 2007) (Lam et al., 2006) (The Pacific Energy Centre, 2006) (Koenigsberger, 1974) (Kumar, Mahapatra, & Atreya, 2007) (INN, 1977) (Liedl, 2011) (Chang Escobedo, 2008) (Rosales, 2007) (ABNT, 2003) (IRAM, 1996) METHODS Research region Nepal expands from the Gangetric plain at an elevation of 60 m up to the high Himalaya Mountains with the highest peak in the world the Mt. Everest at an elevation of 8,848 m. The highly diversified 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 168 geography leads to large variation in climate. The climatic diversity has been also reflected in the traditional architectures. While traditional houses in the upper Himalayan have a compact building typology and are attached to each other aiming to reduce heat loss, houses in the subtropical plain have a more elongated floor plan and are distributes in loose settlement pattern allowing air penetration (Bodach, Lang, & Hamhaber, 2014). With the modernisation of the constructions sector in Nepal, traditional building techniques are replaced by universal design, modern construction technologies and materials. New buildings in urban centres of Nepal are built using column-beam structure of reinforced concrete combined with brick filling walls and flat roofing. Facades with large unshaded glazing area and aluminium panel cladding are typical design options for commercial buildings. Due to centralisation and issues of prestige, modern building practises from the capital Kathmandu are spreading out in other parts of the country where the climatic conditions are very different. Architects and engineers are trained in the capital or in India bringing design ideas and construction techniques from these places that are often inappropriate in the climate of the place. In contrast to other Asian countries, Nepal has not developed any standards or regulation for a more sustainable building design. The national building code is concerned about structural safety and does not contain any standards on energy efficiency. Currently, the Department of Urban Development and Building Construction (DUDBC) which is the government organisation responsible for drafting building regulation, has started the process to develop green building technology guidelines. However, the lack of a proper bioclimatic zoning makes it difficult to define standards for climate-responsive building design or envelope insulation. A climate classification for building design will be a useful tool for regulators as well as building professionals to enhance climate-adapted design practices and the step forwards towards a more sustainable development of the Nepalese building sector. Bioclimatic approach and thermal comfort The bioclimatic approach explores the opportunities to design according to the local climate conditions. Olgyay (1963) developed the first bioclimatic chart based on outdoor climate conditions aiming to identify mitigation measures like solar radiation, air movement or shading to achieve a comfortable indoor climate. Givoni (1963) developed a bioclimatic chart based on indoor conditions using the standard psychrometric chart. His chart has been widely used to identify passive design strategies for different bioclimatic zones (Lam et al., 2006; Rakoto-Joseph, Garde, David, Adelard, & Randriamanantany, 2009; Singh, Mahapatra, & Atreya, 2007). Some countries have used solely his chart to define the climate classification for building design (ABNT, 2003). The main challenge for developing a bioclimatic chart is the definition of the thermal comfort zone. Thermal comfort is defined as a subjective response of a person in regard to satisfaction with the thermal environment (ASHRAE, 2010). It is influenced by environmental factors, such as air temperature, air movement, humidity, radiation, and personal factors like metabolic rate, clothing, state of health and acclimatization (Szokolay, 2008). For naturally ventilated buildings ASHRAE Standard 55 proposes the adaptive thermal comfort approach and defines a range of acceptable indoor temperature of 2.5 K above and below optimum comfort temperature. Thereby, the comfort temperature is calculated by the outdoor temperature using the equation (1). (1) (de Dear & Brager, 2002) comfort temperature and Tout is the mean outdoor temperature where Tc is the optimum However, some studies question the applicability of the adaptive model, particularly, in warm and humid climates (Harimi, Ming, & Kumaresan, 2012). Two studies on thermal comfort in Nepal have found that people feel comfortable at temperatures far below and above international comfort standards (Rijal, Yoshida, & Umemiya, 2010). Comparing the optimum comfort temperature using equation (1) and the actual comfort temperature found in the field, temperature differences between 0.2 and 9 K are 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 169 recorded depending on the region (Table 2). That means the adaptive thermal comfort model of ASHRAE 55 might be not applicable to Nepal. Table 2. Comparison of predicted comfort temperature and actual found in the field Summer Location Altitude Tmean out Tc pred Tc field Tmean out Banke 150 m 31.4 27.5 30.0 15.2 Bhaktapur 1,350 m 22.2 24.7 25.6 10.6 Dhading 1,500 m 25.4 25.7 29.1 13.3 Kaski 1,700 m 18.8 23.6 23.4 8.9 Solukhumbu 2,600 m 13.1 21.9 21.1 4.0 Mustang 3,705 m n.s. n.s. n.s. 6.0 Tmean out Mean outdoor temperature (Rijal et al., 2010) Tc pred Predited comfort indoor temperature (de Dear & Brager, 2002) Tc field Comfort temperature according to field study (Rijal et al., 2010) Winter Tc pred 22.5 21.1 21.9 20.6 19.0 19.7 Tc field 16.2 15.2 24.2 18.0 13.4 10.7 Updating his original research work, Givoni proposed an extended comfort zone for hot developing countries that considers the acclimatization resulting from living in naturally ventilated buildings (Givoni, 1992). It defines temperatures between 18°C and 29°C and humidity levels from 4g/kg up to 17g/kg as comfortable. Givoni's extended comfort zones is used in this study because it is evaluated the most appropriate approach for the Nepalese context. Givoni did not define zones for passive design strategies in his updated chart for hot developing countries. Therefore, this study uses the boundaries defined by Gonzalez et al (1986) for warm and humid climates in developing countries. The upper limit for ventilation is set to absolute humidity of 20.5 g/kg. The ventilation zone was extended to 100% of relative humidity taking reference to several studies conducted in hot and humid climates that found out that local people can cope with higher humidity by increasing ventilation (Gonzalez et al., 1986; Shastry, Mani, & Tenorio, 2012). Solar passive heating zone is defined between 10.5°C and 20.0°C (outside comfort zone). Mechanical heating is needed up to a temperature of 10.5°C. The upper boundary for evaporative cooling is set at the wet bulb temperature line of 24°C. Humidification is needed below wet bulb temperature of 10.6°C. Nepal’s climatic diversity is mainly caused by the high variation in altitude. Therefore, elevation was chosen as the main criteria for developing bioclimatic zones. Meteorological data (temperature, precipitation) from 26 weather stations were collected either directly from Department of Hydrology and Meteorology or derived from United Nation's Food and Agriculture Organisation (FAO) climate database (FAO, 2014). The station data was used to generate a typical meteorological year (TMY) for each location using the recognized software tool METEONORM (Meteotest, 2014). The TMY was then analysed using the bioclimatic approach. RESULTS The plotting of the climate data of 26 locations on the psychrometric chart shows clearly that Nepal has a composite climate that is strongly influenced by the monsoon. Composite means there is no dominating climate for six following months. Instead there are four different seasons that are leading to different design strategies: 1. Winter season (December to February); 2. Pre-Monsoon (March to May); 3. Monsoon or summer season (June to September); 4. Post-monsoon (October to November). The analysis of the bioclimatic chart of 26 locations led to four different bioclimatic zones (Figure 1): 1.Warm Temperate (below 500 m); 2. Temperate (500-1500 m); 3. Cool temperate (15012500 m); 4. Cold (above 2500 m). Table 4 gives an overview about the climatic conditions in each zone. The climate and design strategies for each zone are discussed in the following. In the warm temperate climate daily temperature rises in pre-monsoon and monsoon season well above the comfort zone reaching up to 35°C. While relative humidity is below 60 % in the pre-monsoon, it increases up to 90 % in monsoon season. The winter month are warm with average temperatures above 10°C. Figure 2 shows that the main design strategies for warm temperate climate zone is natural 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 170 ventilation in summer and passive solar heating in winter. For the pre-monsoon time thermal mass is recommendable. Table 4. Characteristics of different bioclimatic zones of Nepal Bioclimatic zones Summer temperature Mean maximum Mean minimum Winter temperature Mean maximum Mean minimum Relative Humidity Warm temperate Temperate Cool Temperate Cold 29 – 35°C 22 – 26°C 25–35°C 18–25°C 22 – 26°C 14 – 18°C 16 – 22°C 7 – 12°C 21 – 26°C 9 – 15°C 25 – 90% 17 – 25°C 5 – 10°C 20 – 90% 11 – 20°C -2 – 5°C 30 – 90% below 10°C below -2°C 10 – 90% Figure 1 Bioclimatic zoning for Nepal Figure 2 Bioclimatic chart for warm temperate (left) and temperate climate zone (right) In the temperate climate zone average summer temperatures are more moderate, hardly exceeding the comfort zone. During pre-monsoon mean temperatures and humidity is very much comfortable. However, in some places day temperature can rise up to 35°C. In the monsoon season relative humidity might increase above 80% in few locations. In winter temperatures fall below the lower comfort limit; 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 171 night temperature can drop down to 5°C. However, day temperature might fall within the comfort zone around 20°C. The most important design strategy for the temperate climate zone is passive solar heating combined with thermal mass (Figure 2). This strategy can balance the high temperature swing during the colder months of the year. Few mechanical heating might be necessary in winter. During monsoon season ventilation is required. Thermal mass can bring relieve and absorb excessive heat in pre-monsoon season. Figure 3 Bio-climatic chart for cool temperate climate zone In the cool temperate climate day temperature in pre-monsoon and monsoon season are within the comfort zone. During monsoon time temperatures are between 15-20°C and relative humidity rarely rises above 80%. In winter average temperatures are clearly below comfort. Night temperatures can down up to the freezing point. Passive solar heating is the most essential strategy used all over the year (Figure 3). In summer thermal mass that store solar heat gain during the day might compensate night temperatures that are often below the lower comfort limit. In winter solar heat gains might contribute to reduce the heating demand by mechanical means. However, mechanical heating is necessary from October to March. Figure 4 Bio-climatic chart for cold climate zone In the cold climate temperature hardly reach the comfort zone (Figure 4). During summer, day time temperature rarely rises above 18°C. During winter average temperature are around the freezing point. In the cold climate of Nepal passive solar heating is the only design strategy that can be applied. It will reduce the heating demand during the summer month. However, mechanical heating is required all over the year. DISCUSSION The bioclimatic chart for the warm temperate climate indicates passive solar heating as main 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 172 climate-responsive design strategy in winter. By capturing the solar radiation during the day and storing the heat in the thermal mass of the building, lower night temperature can be compensated. By this way mechanical heating is not required. Thermal mass is also desirable during the pre-monsoon for cooling purpose. In contrast the warm and humid summer claims for light building materials like applied in the traditional architecture of the region. The solution to this conflicting design strategies might be the application of high thermal mass in the interior of the building, e.g. for interior walls, floors and ceilings. A suitable construction technique for the exterior walls could be the reverse brick veneer wall. High thermal mass of the northern outside wall without solar exposure is also possible. In any case shading of the openings and the construction elements of high thermal mass has to be provided in summer to avoid overheating. Furthermore, building design should enhance air movements within the building through cross or stack ventilation. The temperate climate zone is the most comfortable bioclimatic zone of Nepal. Passive solar heating combined with the minimisation of air filtration and good insulation of the building envelope can fulfil most of the heating demand in winter. High thermal building mass is desirable for passive heating as well as passive cooling due to the high daily temperature swing. Enhancing natural air movement through cross or stack ventilation is required during the warm and humid monsoon season. In cool temperate climate of Nepal passive solar heating strategies is required all over the year. Building layout should be compact and of high thermal mass. Optimising the design for passive solar heating can reduce the amount of mechanical heating. High solar radiation available in winter can also be used for active solar heating by using solar thermal collectors of solar air heating. The only bioclimatic design strategy in cold climate of Nepal is passive solar heating. However, active heating is needed all over the year. Compact building layout, reduction of air infiltration and good insulation of roof, walls and windows are the imperative to protect from the cold in this harsh mountain climate. The application of active solar heating to support a conventional heating system is recommended. CONCLUSION This study developed the first bioclimatic zoning for Nepal. The main passive solar design strategies for the four different bioclimatic zones were identified using the bioclimatic chart. This new climate classification can help planners and architects to make general decisions at early design stage to develop more climate-responsive and energy-efficient buildings. Furthermore, it might be useful for the development of appropriate building energy regulations. However, the qualitative approach of the bioclimatic chart has its limitations due to the fact that it only considers two climate factors: temperature and humidity. The micro-climatic conditions can vary and a detailed analysis of the site might be necessary to come up with site-specific solutions. Due to the fact that the climate in Nepal is of composite character design strategies might conflicting each other. Therefore, further research is needed to quantify the effectiveness of the passive design strategies in each climate zones. REFERENCES ABNT. (2003). Desempenho térmico de edificações Parte 3: Zoneamento bioclimático brasileiro e diretrizes construtivas para habitações unifamiliares de interesse social. Rio de Janeiro: Associação Brasileira de Normas Técnicas (ABNT). ASHRAE. (2007). Standard 90.1-2007 Normative Appendix B – Building Envelope Climate Criteria. ASHRAE. (2010). ANSI/ASHRAE Standard 55-2010 Thermal Environmental Conditions for Human Occupancy. Atlanta. Bansal, N. K., & Minke, G. (1988). Climatic Zones and Rural Housing in India, Part 1 of Indo-German Project on PASSIVE SPACE CONDITIONING (Scientific.). Jülich: Kernforschungsanlage Jülich GmbH. Bodach, S., Lang, W., & Hamhaber, J. (2014). Climate responsive building design strategies of vernacular architecture in Nepal. Energy and Buildings, 81, 227–242. doi:10.1016/j.enbuild.2014.06.022 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 173 Chang Escobedo, J.-A. (2008). Guía de aplicación de arquitectura bioclimática en locales educativos. Lima: Ministerio de Educaion, Oficina de Infraestructura Educativa. De Dear, R. J., & Brager, G. S. (2002). Thermal comfort in naturally ventilated buildings: revisions to ASHRAE Standard 55. Energy and Buildings, 34(6), 549–561. doi:10.1016/S0378-7788(02)000051 FAO. (2014). CLIMWAT 2.0 Database. 2013. Retrieved April 10, 2014, from http://www.fao.org/nr/water/infores_databases_climwat.html Givoni, B. (1969). Man, Climate and Architecture. (H. J. Cowan, Ed.)Elsevier;(). Amsterdam - London New York: Elsevier Publishing Company Limited. Givoni, B. (1992). Comfort, climate analysis and building design guidelines. Energy and Buildings, 18(1), 11–23. Gonzalez, E., Hinz, E., Oteiza, P. de, & Quiros, C. (1986). Proyecto clima y arquitectura, Volumen 1. Mexico: Ediciones G. Gili. Harimi, D., Ming, C. C., & Kumaresan, S. (2012). A Conceptual Review on Residential Thermal Comfort in the Humid Tropics. International Journal of Engineering Innovation & Research, 1(6). INN. (1977). Arquitectura y construcción - Zonificación climático habitacional para Chile y recomendaciones para el diseño arquitectónico NCH1079.Of77. Chile: Instituto Nacional de Normalizacion (INN). Retrieved from http://seigrapa.weebly.com/uploads/1/1/8/2/11828201/nch1079-1977.pdf IRAM. (1996). Clasificación bioambiental de la República Argentina IRAM 11603:1996. Instituto de Normalizacion y Certificacion (IRAM). Retrieved from http://www.scribd.com/doc/58167249/IRAM-11603 Koenigsberger, O. H. (1974). Manual of Tropical Housing and Building: Climatic design (University.). Longman. Kumar, M., Mahapatra, S., & Atreya, S. K. (2007). Development of bio-climatic zones in north-east India. Energy and Buildings, 39, 1250–1257. doi:10.1016/j.enbuild.2007.01.015 Lam, J. C., Yang, L., & Liu, J. (2006). Development of passive design zones in China using bioclimatic approach. Energy Conversion and Management, 47(6), 746–762. doi:10.1016/j.enconman.2005.05.025 Liedl, P. (2011). Interaktion Klima-Mensch-Gebäude Planungsswerkzeuge für die Konzeptphase von Verwaltungsgebäuden in unterschiedlichen Klimaregionen. Technische Universität München. Liedl, P., Hausladen, G., & Saldanha, M. (2012). Building to Suit the Climate: A Handbook (Vol. 2012, p. 176). Walter de Gruyter. Mahmoud, A. H. a. (2011). An analysis of bioclimatic zones and implications for design of outdoor built environments in Egypt. Building and Environment, 46(3), 605–620. doi:10.1016/j.buildenv.2010.09.007 Meteotest. (2014). Meteonorm Software. Bern: Meteotest Company. Olgyay, V. (1963). Design with the Climate (p. 190). Princeton, New Jersey: Princeton University Press. Rakoto-Joseph, O., Garde, F., David, M., Adelard, L., & Randriamanantany, Z. a. (2009). Development of climatic zones and passive solar design in Madagascar. Energy Conversion and Management, 50(4), 1004–1010. doi:10.1016/j.enconman.2008.12.011 Rijal, H. B., Yoshida, H., & Umemiya, N. (2010). Seasonal and regional differences in neutral temperatures in Nepalese traditional vernacular houses. Building and Environment, 45(12), 2743– 2753. doi:10.1016/j.buildenv.2010.06.002 Rosales, L. (2007). Zonas climáticas para el diseño de edificaciones y diagramas bioclimáticos para Venezuela. Tecnología Y Construcción, 23(1), 45–60. Retrieved from http://www2.scielo.org.ve/pdf/tyc/v23n1/art05.pdf Shastry, V., Mani, M., & Tenorio, R. (2012). Impacts of Modern Transitions on Thermal Comfort in Vernacular Dwellings in Warm-Humid Climate of Sugganahalli (India). Indoor and Built Environment. doi:10.1177/1420326X12461801 Singh, M. K., Mahapatra, S., & Atreya, S. K. (2007). Bio-climatic Chart for Different Climatic Zones of Northeast India. In Proceedings of 3rd International Conference on Solar Radiation and Day Lighting (SOLARIS 2007) (pp. 2–7). New Delhi: Anamaya Publishers. Szokolay, S. V. (2008). Introduction to architectural science: the basis of sustainable design. Journal of the American College of Radiology : JACR (Vol. 8, pp. 259–64). doi:10.1016/j.jacr.2010.08.020 The Pacific Energy Centre. (2006). California Climate Zones and Bioclimatic Design. The Pacific Energy Centre. Retrieved from http://www.energy.ca.gov/ Upadhyay, A. K., Yoshida, H., & Rijal, H. B. (2006). Climate Responsive Building Design in the Kathmandu Valley. Journal of Asian Architecture and Building Engineering, (May), 169–176. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 174 Session 2B : Low carbon cities and neighborhood development PLEA2014: Day 1, Tuesday, December 16 14:10 - 15:50, Compassion - Knowledge Consortium of Gujarat Improving Pedestrian Thermal Comfort by Pavement-Watering during Intense Heat Events M. Hendel M. Colombert, PhD Y. Diab, Pr L. Royon, PhD Paris City Hall martin.hendel@paris.fr Université Paris Est, EIVP Université Paris Est, EIVP Univ Paris Diderot, MSC ABSTRACT From the late 19th until the mid-20th Century, pavement-watering was used to prevent dust cloud from forming. This practise has since been lost, but is now stirring new interest as a tool for urban heat island mitigation, climate change adaptation and pedestrian thermal stress reduction. To evaluate the potential of pavement-watering, two daytime watering methods were tested over the summer of 2013 in Paris, France: the pavement and sidewalk of a N-S street and the pavement of an E-W street. The effectiveness of the method was measured according to mean radiant temperature (MRT) and Universal Thermal Climate Index (UTCI) equivalent temperature reductions, determined by a statistical analysis. MRT and UTCI reductions were observed at both sites. While daily effects were highest at the N-S site, highest maximum hourly cooling was observed at the E-W site, reaching 2.9°C for MRT and 1.2°C for UTCI. Overall, hourly cooling was most often statistically significant at night and at the N-S site. INTRODUCTION Paris’ strong hygienist movement during the 19th Century led to the development of its dual water supply. Street cleaning has since relied on the use of non-potable water. Until the mid-20th Century, streets could be watered up to five times a day on hot summer days to prevent dust clouds from forming (Girard, 1923). According to reports by urban managers at the time, many inhabitants also it had a cooling effect. As mechanized cleaning was generalized, these practises were lost and nearly forgotten. Urban areas, through a combination of radiation trapping, wind obstruction, and low surface humidity, create a localized warming effect know as the urban heat island effect (Grimmond, 2007; Oke, 1973). Today, climate change is expected to increase the frequency and intensity of heat-waves on all continents, including in the Paris region (Lemonsu, Kounkou-Arnaud, Desplat, Salagnac, & Masson, 2012). Unfortunately for urban dwellers, heat-waves have been found to interact with urban heat islands and increase their intensity (Li & Bou-Zeid, 2013). This mechanism helps explain why heat-waves are more devastating in densely populated cities than in rural areas, such as was the case in Paris during the August 2003 heat-wave (Robine et al., 2008). Adaptation to more frequent and more intense heat-waves is therefore crucial in dense cities. In this context, pavement-watering may once more have a role to play in cities as an emergency counter-measure against heat-waves. By artificially reintroducing the evaporative mechanism at work in rural soils, pavement-watering is expected to positively impact pedestrian thermal comfort by reducing surface and air temperatures, while only marginally increasing air humidity. In Japan, field and numerical studies of pavement-watering have been carried out over the last twenty years or so (Kinouchi & Kanda, 1997, 1998; Nakayama & Fujita, 2010; Nakayama & Hashimoto, 2011; Takahashi, Asakura, Koike, Himeno, & Fujita, 2010; Yamagata, Nasu, Yoshizawa, Martin Hendel is a PhD Student at the Water and Sanitation Department, Paris City Hall, Paris, France. Morgane Colombert is a Lecturer at Université Paris Est - EIVP, Paris, France. Youssef Diab is a Professor at Université Paris Est - EIVP, Paris, France. Laurent Royon is an Assistant Professor at Univ Paris Diderot, Paris, France. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 175 Miyamoto, & Minamiyama, 2008). In Paris, computer simulations of the method at the city-scale (Météo-France & CSTB, 2012) have been conducted in recent years as well as pavement-watering field experiments (Bouvier, Brunner, & Aimé, 2013). Review of previous field work reveals strong variability in the micro-climatic effects of pavementwatering as well as in their measurement methodology. Reported air cooling ranges from 0.4°C to 4°C, while measurement heights vary from 0.5 m to 2 m (Bouvier et al., 2013; Takahashi et al., 2010; Yamagata et al., 2008). Furthermore, only a few studies study the effect on pedestrian thermal comfort. This variability in methods only highlights the need for standardization in urban micro-climatic measurements as was outlined by Johansson, Thorsson, Emmanuel, & Krüger (2014). To better understand the potential of pavement-watering to improve pedestrian comfort in Paris, two watering methods were tested over the summer of 2013 at two locations: rue du Louvre in the 1st and 2nd Arrondissements and Belleville in the 20th Arrondissement. This article will present the method used to analyse field measurements and the effects of pavement-watering on pedestrian thermal comfort as estimated by mean radiant temperature (MRT) and the Universal Thermal Climate Index (UTCI). MATERIALS AND METHOD Micro-climatic parameters were investigated at two sites in Paris, France over the summer of 2013, hereafter referred to as Louvre and Belleville. For the former, measurements were conducted on rue du Louvre, near Les Halles in the 1st and 2nd Arrondissements, while at Belleville they took place on rue Lesage and rue Ramponeau in the 20th Arrondissement. Watered and control weather station positions are illustrated in Figure 1. Two twin weather stations were positioned for each site, each pair measuring identical parameters. Each position was chosen to ensure that the urban environment of each station was as identical as possible (traffic, materials, urban morphology, sky view factor, …). On rue du Louvre, watering took place on the sidewalk and pavement, each paved with asphalt concrete. Both watered and dry portions of the street are approximately 180 m long and 20 m wide. The street canyon has an aspect ratio approximately equal to one (H/W=1) and has a roughly N-S orientation. At Belleville, watering was limited to the cobblestone pavement only. The watered portion was located on rue Lesage while the dry portion was on rue Ramponeau, a parallel street nearby. The watered portion was approximately 40 m long and 4 m wide. Both canyons have an aspect ratio approximately equal to one and have a roughly E-W orientation. N N N N 500 m Weather stations Watered portion Weather stations Dry portion 10 m Watered portion Dry portion 50 m Figure 1: Map of weather station positions at the Louvre (left) and Belleville (right) test sites. Instruments Weather station design is presented in Figure 2. Instruments within pedestrian reach were protected behind a 2-m cylindrical white-painted steel cage. All parameters were recorded every minute and smoothed with a one-hour moving average. The final series was obtained by keeping four data points per 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 176 hour from the smoothed series. All data is presented in local daylight savings time (UTC +2). Table 1 lists the instruments used for this analysis as well as their height and accuracy. Street sign post Wind, Temperature, Humidity, Presence of rain, Solar Irradiance Measurement cage 4m Temperature, Humidity, Globe Temperature (1,5 m height) 2m Temperature, Heat Flux (5 cm depth) 0,30 m 1,60 m minimum 1m 1,60 m Figure 2: Weather station design and instrumentation. The temperature and heat flux sensor was only installed at the Louvre site. Table 1: Instrument type, measurement height and accuracy Height Parameter Instrument Air temperature Relative humidity Black globe temperature Wind speed Sheltered Pt100 1/3 DIN B Sheltered capacitive hygrometer Pt100 1/2 DIN A - ISO 7726 2D ultrasonic anemometer 1.5 m 1.5 m 1.5 m 4m Accuracy 0.1°C 1.5% RH 0.15°C 2% Thermal comfort evaluation The effects of pavement-watering were quantified by MRT and UTCI. Since we are studying the potential for pavement-watering to reduce the health impacts of intense heat-waves, one of the indexes should be able to properly assess heat-related health impacts. Thorsson et al. (2014) found that MRT was a better predictor for heat-related mortality than air temperature, which is more commonly used. MRT was therefore chosen for these reasons. However, although it may be effective at predicting heat-related mortality, MRT is a relatively new index for heat stress and does not currently allow us to evaluate intermediate levels of heat stress. Furthermore, it ignores the effects of wind speed, air temperature or humidity on thermal comfort. We therefore need to include an index that takes all relevant climatic aspects into account. One of the more recently developed thermal comfort indexes is the Universal Thermal Climate Index (UTCI) (Blazejczyk et al., 2010). UTCI was developed by international experts from Commission 6 of the International Society of Biometeorology (ISB) and European COST Action 730 from the year 2000 to 2009. It is based on a special version of the multi-node Fiala thermophysiological model. Air temperature, humidity, wind speed and MRT are used as well as assumptions on the metabolic activity and clothing of pedestrians to calculate an equivalent air temperature for reference conditions. To calculate MRT, black globe temperature (Tg), air temperature (Ta) and wind speed (v) measurements from the weather station were used according to the method described by ASHRAE (2001). Air temperature, relative humidity, MRT and wind speed measurements were then used to obtain UTCI equivalent temperature, which was fast-calculated with the FORTRAN code written by Peter Bröde in 2009, adapted for use with the R software environment. The source code is freely available at http://www.utci.org/utci_doku.php. Inaccuracies are introduced into both MRT and UTCI by the use of 4-m wind speed rather than 1.5m and 10-m wind speed, respectively, as well by globe temperature measured inside the cylindrical cage. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 177 Watering method Watering was started if certain weather conditions were met based on Météo-France’s three-day forecast. These as well as those for heat-wave warnings in Paris are presented in Table 2. BMIMax and BMIMin refer to the 3-day mean of maximum (Tx) and minimum (Tn) air temperature. At the Louvre site, cleaning trucks were used to sprinkle sidewalk and pavement every hour in the morning (6:30 am to 11:30 am) and every 30 minutes in the afternoon (2 pm to 6:30 pm). At the Belleville site, a removable 40-m watering pipe was laid along the gutter to water the pavement on rue Lesage continuously from 7 am to 7 pm. In terms of the watered surface ratio, rue du Louvre was 100% watered, while rue Lesage was approximately 33% watered. Water used for this experiment was supplied by the city’s 1,600-km non-potable water network, principally sourced from the Ourcq Canal. Table 2: Weather conditions required for pavement-watering and heat-wave warnings Parameter Pavement-watering Heat-wave warning level BMIMin BMIMax Wind speed Sky conditions ¯ 16°C ¯ 25°C Κ 10 km/h Sunny (less than 2 oktas cloud cover) ¯ 21°C ¯ 31°C - Data selection and interpretation method Because many weather conditions were encountered over the duration of the experiment, only days of Pasquill atmospheric stability class A-B or more were retained for the upcoming analyses (Pasquill, 1961). This provision limits selected days to those with clear skies (less than 3 oktas) and low wind speeds (less than 3 m/s). To interpret the effect of pavement-watering on MRT or UTCI, the difference between the watered and control stations is analysed, calculated as: ydifference=ywatered – ycontrol. Negative values indicate that the watered station parameter is lower than that of the control station, and vice versa. However, even with these provisions in mind, it is not possible to determine the effect of pavementwatering by analyzing the raw data from the weather stations. To eliminate the high natural variability in the data, a statistical representation of the daily profile of the difference is used instead, calculated for dry (control) days and case (watered) days. Finally, because no watering occurs between midnight and 6 am, days will be divided into 24-hour periods beginning at 6 am and ending at 5:59 am the next day. Thus, when we refer to data from July 8th for example, this means from July 8th at 6 am to July 9th at 5:59 am. RESULTS Weather stations recorded continuously from July 2nd until September 10th, 2013. Over this period, ten days met the conditions set for pavement-watering. Of the ten watered days, July 8th, 9th, 10th and 16th were the coolest (TxΚ30°C), with July 22nd, 23rd, August 1st and 2nd being the warmest (Tx¯35°C, Tn¯20°C). August 23rd and September 5th were also watered and had intermediate temperatures (35°C¯Tx¯30°C). Several measurement interruptions occurred over this period. Rue du Louvre was most affected with its control station unoperational from July 19th until August 19th and from September 4th until September 10th. At Belleville, only one interruption occurred from the 22nd to the 25th of July. These events were poorly timed, resulting in the absence of control measurements on July 22nd and 23rd at either site and on August 1st and 2nd at the Louvre site. It should therefore be kept in mind that results from rue du Louvre only include the coldest watered days, while July 22nd and 23rd are missing from the Belleville data. Effects on mean radiant temperature Figure 3 illustrates the difference between watered and control stations at the Louvre (top) and Belleville (bottom) sites for MRT. Deviations between the blue and red curves are statistically 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 178 6 significant only if the blue curve is not between the dotted red lines. As can be seen, both sites behave quite differently, with signal amplitudes ranging from [-2°C;4°C] at the Louvre site to [-6°C;10°C] at the Belleville site. 24-hour mean and maximum cooling effects are summarized in Table 3. 24-hour average and daily maximum effects are quite different between sites. Although a net cooling effect of 0.36°C is seen at the Louvre site, a warming of 0.09°C is seen at Belleville. However, the latter result is not statistically significant. No statistically significant 24-hour effect is therefore detected for Belleville, while a statiscally significant average cooling of 0.36°C is seen for Louvre. When the hourly curves in Figure 3 are considered, the MRT difference curve is always lower on watered days than on dry days at the Louvre site. This is not the case at the Belleville site. However, the deviations between the control and watered day curves are not always statistically significant at either site. Overall, the effects are most often statistically significant at night and at the Louvre site. The maximum effects reported in Table 3 are statistically significant at both sites. 2 0 -2 MRT Difference (in °C) 4 Control days Watered days 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 00:00 02:00 04:00 10 5 0 -10 -5 MRT Difference (in °C) 06:00 Control days Watered days 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 00:00 02:00 04:00 06:00 Time Figure 3: Difference in mean radiant temperature between twin stations at the Louvre (top) and Belleville (bottom) sites. Solid lines indicate the mean value for control (red) and watered (blue) days, dotted red lines indicate the 95% confidence interval of the difference between the control and watered day mean curves. Table 3: Mean and maximum cooling effects of pavement-watering on MRT at Louvre or Belleville Maximum effect Site Mean (24-hour) effect Louvre Belleville 0.36°C -0.09°C 1.79°C 2.94°C Effects on pedestrian thermal comfort Figure 4 illustrates the hourly difference between watered and control stations at the Louvre (top) and Belleville (bottom) sites for UTCI. As was the case for MRT, both sites behave quite differently, despite reduced signal amplitudes compared to that of MRT with [-0.5°C;1.75°C] at Louvre and [-2°C;3.5°C] at Belleville. 24-hour mean and maximum cooling effects are summarized in Table 4. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 179 2.5 Unlike for MRT, the effects on UTCI are relatively similar between sites. UTCI is reduced by an average of 0.21°C at the Louvre site and by 0.12°C at the Belleville site. However, only the effect on rue du Louvre is statistically significant. No statistically significant 24-hour effect on UTCI is therefore visible at the Belleville site. When the hourly curves in Figure 4 are considered, statistically significant effects exist, although they are less numerous than for MRT. Maximum effects are significant and in the same order of magnitude, between 0.98°C and 1.20°C. As was the case for MRT, the hourly effects are most often significant at night and at the Louvre site. 1.5 1.0 0.5 0.0 -1.0 -0.5 UTCI Difference (in °C) 2.0 Control days Watered days 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 00:00 02:00 04:00 4 06:00 06:00 2 1 0 -1 -3 -2 UTCI Difference (in °C) 3 Control days Watered days 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 00:00 02:00 04:00 06:00 Time Figure 4: Difference in UTCI between twin stations at the Louvre site (top) and at the Belleville site (bottom). The dashed lines indicate the mean value for control (short red dashes) and watered (long blue dashes) days, while red dotted lines indicate the 95% confidence interval of the difference between the control and watered day mean curves. Table 4: Mean and maximum cooling effects of pavement-watering on UTCI at Louvre or Belleville Maximum effect Site Mean (24-hour) effect Louvre Belleville 0.21°C 0.12°C 0.98°C 1.20°C DISCUSSION For both MRT and UTCI, 24-hour average effects are always greater at the Louvre site, while maximum effects are higher at the Belleville site. Furthermore, the 24-hour average effects site are only statistically significant at the Louvre site, while maximum effects are always significant regardless of the site considered. Finally, the hourly effects are most often significant at night and at the Louvre site. Maximum effects for MRT were in the order of a few degrees Centigrade while they were around one degree for UTCI. Overall, greater, more significant results were found for MRT than for UTCI. On the one hand, the more significant and higher 24-hour results obtained at rue du Louvre are most likely explained by the proportion of street that is watered. While rue du Louvre was watered over 100% of its width, only a third of rue Lesage was watered. On the other hand, the highest maximum effects were reached at the Belleville site. This may be 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 180 linked to the absence of data on the hottest watered days for the Louvre site, when the effects are expected to be highest. However, it could also be linked to the Belleville site’s orientation, since both site canyons have the same aspect ratio. For an aspect ratio of one, N-S streets experience similar conditions on either side of the street, while E-W streets have very different conditions between the North and South sidewalks, due to predominant daytime sunlight or shade, respectively. Maximum heat stress conditions are therefore much higher on the North side of an E-W street than on its South side or in a N-S street during the summer. These links between aspect ratio, orientation and pedestrian thermal comfort were studied by Ali-Toudert & Mayer (2006). Since it is expected that the effects are greatest in the hottest conditions, cooling may be increased by watering the North sidewalk rather than the pavement on rue Lesage. However, because of the missing data at the Louvre site, this cannot be confirmed. Finally, the difference in paving materials may also explain some of these aspects. We now look to evaluate the health or comfort impacts of these effects. No universal relation with mortality risk has been established for MRT at this time to our knowledge, while UTCI is scaled according to five heat stress levels for hot environments. To evaluate the effect of pavement-watering on pedestrian thermal comfort through UTCI, we must therefore compare the effect with the span of the UTCI heat stress categories, i.e. between 6° and 8°C. With a maximum cooling effect of 1.2°C, pavement-watering has a limited effect on pedestrian thermal comfort, only rarely causing a downwards shift in heat stress category. However, it should be noted that pavement-watering performs well in comparison to other urban surface cooling methods, such as high-albedo pavement materials (Erell, Pearlmutter, Boneh, & Kutiel, 2013). CONCLUSIONS The effects of pavement-watering on pedestrian thermal comfort were evaluated via MRT and UTCI, using data collected during a field experiment of pavement-watering conducted at two sites in Paris, France over the summer of 2013. The N-S site was entirely watered, while only the pavement, corresponding to a third of total width, of the E-W site was watered. For both MRT and UTCI, average cooling effects were highest at the N-S site, reaching 0.36°C for MRT and 0.21°C for UTCI, while maximum cooling was highest at the E-W site, with up to 2.9°C for MRT and 1.2°C for UTCI. Average cooling at the E-W site was not statistically significant, while maximum cooling was significant at both sites. The higher effect reached at the Belleville site may be closely tied to the street’s orientation, but further measurements would be necessary to confirm this. If this is the case, it may be more efficient to water the North sidewalk of E-W streets rather than their pavement. To our knowledge, no tools currently exist to evaluate the health impact of observed MRT cooling. It is therefore not yet possible to estimate the health impacts of pavement-watering with MRT. In terms of UTCI, observed cooling was found to be limited in comparison to the level necessary to obtain a downwards shift in heat stress category. Regardless, the method performs better than others such as high-albedo pavement materials. In order to continue to evaluate pavement-watering against other methods such as urban greening or artificial shading, further studies such as that conducted by Shashua-Bar, Pearlmutter, & Erell (2011) should be conducted in the Parisian urban environment. Trials should take all relevant decision-making factors into account, including the cost, water consumption and feasibility of each method. NOMENCLATURE BMI MRT Tn Tx UTCI = = = = = biometeorological index [°C] mean radiant temperature [°C] minimum daily temperature [°C] maximum daily temperature [°C] universal thermal climate index [°C] 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 181 REFERENCES Ali-Toudert, F., & Mayer, H. (2006). Numerical study on the effects of aspect ratio and orientation of an urban street canyon on outdoor thermal comfort in hot and dry climate. Building and Environment, 41(2), 94–108. doi:10.1016/j.buildenv.2005.01.013 ASHRAE. (2001). ASHRAE Fundamentals Handbook 2001 (SI Edition.). American Society of Heating, Refrigerating, and Air-Conditioning Engineers. Blazejczyk, K., Bröde, P., Fiala, D., Havenith, G., Holmér, I., Jendritzky, G., … Kunert, A. (2010). Principles of the new Universal Thermal Climate Index (UTCI) and its Application to Bioclimatic Research in European Scale. Miscellanea Geographica, 14, 91–102. Bouvier, M., Brunner, A., & Aimé, F. (2013). Nighttime watering streets and induced effects on the surrounding refreshment in case of hot weather. The city of Paris experimentations. Techniques Sciences Méthodes, (12), 43–55 (In French). Erell, E., Pearlmutter, D., Boneh, D., & Kutiel, P. B. (2013). Effect of high-albedo materials on pedestrian heat stress in urban street canyons. Urban Climate. doi:10.1016/j.uclim.2013.10.005 Girard, L. (1923). Lutte contre la poussière. In Le nettoiement de Paris (pp. 109–120). Paris, France: Léon Eyrolles. Grimmond, S. (2007). Urbanization and global local effects of environmental change۷ : urban. The Geographical Journal, 173(1), 83–88. Retrieved from http://www.jstor.org/stable/30113496 Johansson, E., Thorsson, S., Emmanuel, R., & Krüger, E. (2014). Instruments and methods in outdoor thermal comfort studies – The need for standardization. Urban Climate. doi:10.1016/j.uclim.2013.12.002 Kinouchi, T., & Kanda, M. (1997). An Observation on the Climatic Effect of Watering on Paved Roads. Journal of Hydroscience and Hydraulic Engineering, 15(1), 55–64. Kinouchi, T., & Kanda, M. (1998). Cooling Effect of Watering on Paved Road and Retention in Porous Pavement. In Second Symposium on Urban Environment (pp. 255–258). Albuquerque, NM. Retrieved from https://ams.confex.com/ams/nov98/abstracts/77.htm Lemonsu, A., Kounkou-Arnaud, R., Desplat, J., Salagnac, J.-L., & Masson, V. (2012). Evolution of the Parisian urban climate under a global changing climate. Climatic Change, 116(3-4), 679–692. doi:10.1007/s10584-012-0521-6 Li, D., & Bou-Zeid, E. (2013). Synergistic Interactions between Urban Heat Islands and Heat Waves: The Impact in Cities Is Larger than the Sum of Its Parts. Journal of Applied Meteorology and Climatology, 52(9), 2051–2064. doi:10.1175/JAMC-D-13-02.1 Météo-France, & CSTB. (2012). EPICEA - Rapport final (p. 31 (In French)). Nakayama, T., & Fujita, T. (2010). Cooling effect of water-holding pavements made of new materials on water and heat budgets in urban areas. Landscape and Urban Planning, 96(2), 57–67. doi:10.1016/j.landurbplan.2010.02.003 Nakayama, T., & Hashimoto, S. (2011). Analysis of the ability of water resources to reduce the urban heat island in the Tokyo megalopolis. Environmental Pollution (Barking, Essex : 1987), 159(8-9), 2164– 73. doi:10.1016/j.envpol.2010.11.016 Oke. (1973). City size and the urban heat island. Atmospheric Environment, 7(8), 769–779. Pasquill, F. (1961). The estimation of the dispersion of windborne material. The Meteorological Magazine, 90(1063), 33–49. Robine, J.-M., Cheung, S. L. K., Le Roy, S., Van Oyen, H., Griffiths, C., Michel, J.-P., & Herrmann, F. R. (2008). Death toll exceeded 70,000 in Europe during the summer of 2003. Comptes Rendus Biologies, 331(2), 171–178. doi:10.1016/j.crvi.2007.12.001 Shashua-Bar, L., Pearlmutter, D., & Erell, E. (2011). The influence of trees and grass on outdoor thermal comfort in a hot-arid environment. International Journal of Climatology, 31(10), 1498–1506. doi:10.1002/joc.2177 Takahashi, R., Asakura, A., Koike, K., Himeno, S., & Fujita, S. (2010). Using Snow Melting Pipes to Verify the Water Sprinkling’s Effect over a Wide Area. In NOVATECH 2010 (p. 10). Thorsson, S., Rocklöv, J., Konarska, J., Lindberg, F., Holmer, B., Dousset, B., & Rayner, D. (2014). Mean radiant temperature – A predictor of heat related mortality. Urban Climate, 1–14. doi:10.1016/j.uclim.2014.01.004 Yamagata, H., Nasu, M., Yoshizawa, M., Miyamoto, A., & Minamiyama, M. (2008). Heat island mitigation using water retentive pavement sprinkled with reclaimed wastewater. Water Science and Technology: A Journal of the International Association on Water Pollution Research, 57(5), 763–771. doi:10.2166/wst.2008.187 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 182 Sunlight Availability for Food and Energy Harvesting in Tropical Generic Residential Districts Abel Tablada, PhD Zhao Xi, M.S. [National University of Singapore] abel@nus.edu.sg [National University of Singapore] ABSTRACT Increasing food and energy self-sufficiency in residential areas is one of the key measures to reduce greenhouse gases emissions as well as to mitigate and adapt to climate change. The objective of the study is to verify the impact of building typologies and urban forms with relative high density on sunlight availability. Computational tools are employed to obtain quantifiable indicators of the potential of each variant for energy and food harvesting. Three typical residential typologies, namely, point block, slab block, and contemporary block, in Singapore were identified. Point block typology was assessed in this paper. Twenty five point block cases were assessed in terms of solar access by using three density and geometry parameters: plot ratio, site coverage and building height. Each case was considered on a plot of 520x520m2. Singapore’s weather and sky conditions (1.3°N) were used for the analysis. Out of the 25 cases, six with the lowest plot ratio between 0.8 and 1.9 achieved food self-sufficiency when a hybrid (conventional and vertical) farming method was applied. The cases with the lowest building height (<42 m, <14 storeys) achieve energy self-sufficiency due to the maximum exposed area with PV per amount of residents. The indicators having the higher impact on the food and energy selfsufficiency were plot ratio and building height respectively. The study provided the basis for future environmental and energy assessments including the potential of each variant for natural ventilation and carbon footprint reduction. INTRODUCTION The rapid urbanisation process occurring along the tropical belt promotes the construction of new residential districts, especially in south-east Asia (SEA) countries. In Singapore, the construction of new dwellings is an important part of the building sector. The need to build higher density residential districts to accommodate the growing population in the land-scarce Estate-Island oblige the demolition of relatively old housing estates and build new ones. The land occupied for building new residential districts is, most of the time, located in the peri-urban area in which agricultural activities are foremost. There is a twofold negative effect: increasing food demand by growing population and decreasing farming areas around the cities by the construction of new residential districts. This situation increases the food dependence of Singapore from neighbouring countries. Therefore, increasing local selfsufficiency in terms of food and energy in residential areas is one of the key measures to reduce greenhouse gases emissions as well as to mitigate and adapt to climate change. Urban farming has proved to considerably reduce the food supply chain as well as to socially and Author A is an assistant professor at the Department of Architecture, National University of Singapore. Author B is a research fellow at the Department of Architecture, National University of Singapore. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 183 economically benefit the local population. In addition, the installation of solar panels (photovoltaic and solar collectors) on building envelopes will be the norm if the current trend of more efficient and less expensive panels is prolonged. Therefore, the integration of farming areas and the potential installation of solar panels should be considered as some of the design parameters for new residential districts in Singapore and SEA. Numerous studies have tackled the problem of the insertion of productive gardens in urban areas from the point of view of land use planning (Indraprahasta, 2012), of the use of roof top (Astee & Kishnani, 2010) and of vertical farming and high-tech methods (Despommier, 2013). In addition, several studies have been conducted on the potential of the urban form to harvest solar energy (Kanters & Horvat, 2012) and on the calculation methods of solar distribution on urban environments and solar access (Ibarra & Reinhart, 2011). In the context of Singapore, several studies have dealt with the impact of urban form on several environmental indicators like daylight and natural ventilation (Zhang el al. 2012; Lee et al. 2013) However, no study was found in literature about the impact of urban form on sunlight availability for food production. There is also a need to conduct a comprehensive study dealing with both food and energy harvesting in tropical high density areas. Therefore, the objective of this study is to verify, through the use of computational tools, the impact of a series of densities and urban forms on sunlight availability for one of the three identified typical public housing typologies in Singapore: point block. This is translated into coefficients of self-sufficiency in terms of food (vegetables and fruits) and energy. The study provides the basis for future environmental and energy assessments including the potential of each variant for natural ventilation and carbon footprint reduction. METHOD The study is divided into three main stages: (1) calculation of solar availability on the point-block cases, (2) calculation of the potential of food harvesting on ground and facades and, (3) calculation of the potential of energy harvesting by Photovoltaic (PV) panels on facade and roof surfaces. Building typology and simulation cases Three typical public housing typologies from Singapore (Housing Development Board (HDB)) and, to some extent, from other SEA new urban areas will be assessed: point block, slab block and contemporary block. However, in this paper only the analysis of the point block typology is presented. Table 1 shows a summary of the twenty five point block cases that are assessed in terms of solar access by using density and geometry parameters: plot ratio (PR), site coverage (Cs) and building height (Hb). PR is defined as the ratio of the gross floor area of all buildings to the area of the analysed plot where all buildings are located. Cs is defined as the ratio of the ground floor area of all buildings to the area of the analysed plot. Hb may vary from case to case but all buildings in the same case has equal Hb. The plot area is the same for every case: 520 x 520m2 (27 ha). This represents the equivalent area of a typical large precinct in Singapore or several small ones including the area for car circulation. A similar plot area was used in a natural ventilation study for typical residential district in Singapore (Lee et al. 2013). Figure 1 shows the different Cs and building arrangements for the point-block typology. The different densities were defined departing from the typical PR of the most recent HDB developments in Singapore (PR = 3.0). Then a matrix was developed considering PR, Cs and building height. The height varies in order to keep the same -or very similar- PR on the cases coinciding with the diagonal of the matrix as shown in Figure 1. That means a case (2-4) having site coverage 13% with 27 floors has the same PR = 3.0 than a case (4-2) with Cs of 21% and 17 floors. For the sunlight availability simulations, every case is partially replicated in order to take into account the effect of neighbouring buildings. Figure 2 shows the extended models for three Cs: 10%, 16% and 27%. Singapore’s weather conditions (1.3°N) are considered for the analysis. The sky conditions are considered as intermediate sky which is typical for Singapore along the year. However, the actual values of solar radiation are taken into account from the ASHRAE International Weather for Energy Calculations (IWEC) Data (US Department of Energy, 2013). The sunlight availability of three cases is contrasted with the conditions of Hanoi, Vietnam (21°N) to assess the influence of latitude. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 184 Table 1. Summary of 25 cases and building indicators corresponding to Point Block typology. Cases Nr floors Nr Buildings 1‐1 to 1‐5 2‐1 to 2‐5 3‐1 to 3‐5 4‐1 to 4‐5 5‐1 to 5‐5 10 to 36 12 to 32 14 to 30 14 to 28 13 to 25 36 48 60 76 100 Site coverage (Cs) 10% 13% 16% 21% 27% Plot Ratio (PR) 0.8 to 3.0 1.3 to 3.6 1.9 to 4.2 2.5 to 4.9 3.0 to 5.8 520 m Schematic of the 25 cases corresponding to the point block. The analysed cases are Figure 1 organized according to a matrix of site coverage (X-axis), building height (Y-axis) and plot ratio (diagonal). Case 1-1 is at bottom left, case 1-5 is at top left and case 5-5 is at top right. 520 m Schematic of the simulation model for three of the 5 different building dispositions Figure 2 according to site coverage of 10%, 16% and 27%. The framed area is the analysed area. Settings and calculation method Sunlight availability. The software Daysim (version 3.1b) (Reinhart and Walkenhorst, 2001) was used to perform the calculations of the solar availability (illuminance levels) on the ground, facade and roof surfaces. Daysim is a Radiance-based program (Ward and Shakespeare, 1998) able to simulate time 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 185 series of solar irradiances and illuminance levels. This version of Daysim can be integrated into Autodesk Ecotect which facilitates the modelling of the geometries while allows more accurate calculations in comparison with the Ecotect Solar access model (Ibarra & Reinhart, 2011). Radiance simulation parameters were defined according to ‘scene complexity 1’. This means 5 ambient bounces, 1000 ambient divisions and 300 ambient accuracy among other parameters. The simulation accuracy is increased by using the DDS model which more precisely accounts for the effect of obstructions (neighbouring buildings) on the incident radiation and illuminance values. Daysim calculates both the Daylight Autonomy (DA) and the average illuminance levels per point for the analysed period (whole year). The DA is a climatic-based index which denotes the percentage in which a minimum –defined by user- illuminance level is achieved by daylight alone for a specific time interval (Reinhart and Walkenhorst, 2001). In this study, DA is used to determine the percentage along the whole year in which each analysed point receive more than 10 000 lux from 8:00 till 18:00. The optimal illuminance level for certain vegetables and fruits is considered to be 10 000 lux for about 8 hours (Conover and Flohr, 1996). When the DA is below 80% (less than 8 hours with 10 000 lux), a reduction coefficient is applied for the calculation of the annual yield. A grid of points (lighting sensors) was generated in every case to obtain the illuminance levels. For the ground a grid of 5 m by 5 m was generated among the buildings. For the facades grids of 5 m (X or Y axes according to orientation) by 12 m (Z axis) were generated adapted to the building height. In the case of the roof, a single point was defined due to the lack of obstructions from other buildings. No distinction is made about the different solar availability per facade orientation and height. The average of all facade points will be considered for the calculation of both the farming and solar energy potential. The effect of shading devices (30 cm overhang along all facades on every floor) on ground and facade illuminance levels (lux) is considerable: 4% lees sunlight on ground surface and 12% less on facades. However, modelling all shading devices increase the calculation time several times, therefore, an equivalent reflectance coefficient (-20%) on the facade was applied to account for the ground illuminance reduction due to the shading devices. But since the facade reflectance values have little impact on the facade illuminance levels, another coefficient is applied directly on the final illuminance levels to account for the presence of horizontal shading devices. For overhangs of 30cm, a coefficient of 0.9 will be applied accounting for a 10% reduction. Population and area for farming and PV panels. The amount of population per case is calculated considering that 70% of GFA is residential, 20% institutions and 10% commercial. The total amount of residents per case were calculated according to the average area per capita in the HDB of the last decade 2 equal to 25m and considering a floor plan efficiency (rental flat area out of GFA of residential building) equal to 85%. For the farming activity we consider part of the ground area and part of the facade while for the solar energy harvesting we consider part of the roof surface and part of the facade. The farming area on the ground was derived from the actual land use at Punggol New Town in the northeast of Singapore. From the total plot area, 15% is considered to be for roads and 35% for open space and recreation. The 50% remaining area is distributed between buildings and farming areas according to the different Cs. The area for car parks is considered to be underground or above ground. In the latter case, the roof of the car park is considered to be covered by playgrounds, circulation, green and farming areas. The farming area on the facade considers 50 cm of planters along 30% of the perimeter of the facade. Three fifths of the planter’s thickness (30cm) is projected outside of the building facade acting as a shading device. The remaining is considered to be inside the facade perimeter as part of a balcony or external common corridor. The other 70% of the facade perimeter is considered to be used for the installation of Building Integrated Photovoltaic (BIPV) panels. Eighty percent of the roof area is considered to be covered by solar panels. Selection of crops and food self-sufficiency. A selection of crops was done in order to calculate the yield potential of the farming areas and the potential of food (vegetables and fruits) self-sufficiency 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 186 of the total population in each case. The criteria to select the type of crops were (1) suitability for local context, (2) preference among local residents and, (3) productivity. A reduced variety of vegetables and fruits were chosen with different productivity indices. The vegetables are Kang Kong (30%), Water Mimosa (20%), Chinese Celery (20%), Water Cress (10%) and Pumpkin (20%); the fruits are Dragon Fruit (80%) and Banana (20%). The yield from the crops ranges from 5 to 30 tonnes per hectare for the Pumpkin and the Dragon Fruit respectively. The food cycle per year is from 1 to 18 for Banana and Kang Kong respectively. Two scenarios are considered regarding the technology used for farming. The first one, termed as ‘conventional’, refers to urban ground farming methods. This method considers both the traditional ground soil gardening and the use of soil planters or containers. The second scenario is termed as ‘hybrid’ and it is a combination of the ‘conventional’ and the ‘vertical’ methods (50% ground surface each). The ‘vertical’ method refers to hydroponics, aeroponics and vertical soil-based structures like the A-shaped SkyGreen system introduced in Singapore. The vertical methods are considered to be around 4 times more productive than the conventional ones (Mugundhan, 2011). Only vegetables are considered to grow using the ‘vertical’ method. Table 2 shows the area needed for vegetables and fruits in order to achieve self-sufficiency per capita for the two scenarios. Table 2. Area needed in order to achieve food self-sufficiency per capita (2 scenarios) Yield needed Area needed per year per Area needed per year per per year per capita capita capita (t) (hybrid) (m2) (conventional) (m2) Vegetables 109 9.4 3.7 Fruits 55 14.8 14.8 Total 164 24.2 18.5 PV panels and energy self-sufficiency. Polycrystalline Silicon (pc-Si) and Thin-film Amorphous Silicon Copper Indium Selenide (a-Si CIS) on a horizontal position were considered for the roof and facade respectively. Typical efficiencies and temperature factors were considered: 13% and 8% for pc-Si and a-Si CIS respectively. The energy use per capita of 1287 kWh from April 2013 till March 2014 is considered for the calculation of energy self-sufficiency corresponding to a typical HDB apartment (Singapore Power Group, 2014). Solar collectors for water heating are not considered at this stage. RESULTS The results corresponding to the solar availability analysis and the potential of food and energy harvesting for 25 cases of point block typology are presented in this paper. Sunlight availability The results of the illuminance levels (luxh) per point are averaged for the ground and facade respectively. Figure 3a shows the average illuminance levels for the 25 cases on ground and facade points. As expected, when density (PR) increases, sunlight availability (illuminance levels) decreases. The decrease of illuminance levels is less evident on the facade points (15%) than on the ground (45%) because all facades, disregarding its orientation and the plot density, have a limited (≤ 0.5) sky view factor in comparison with less obstructed horizontal surfaces (≤ 1.0). Figure 3b shows the average percentage of time in which illuminance levels are above 10 000 lux. Here the differences between the lowest and highest densities are lower both for the ground points (around 12%) and facade points (6%). Farming potential and food self-sufficiency Figure 4a shows the ratio of food self-sufficiency for the conventional and hybrid cultivation methods. The differences between the two methods are more evident on lower densities in which larger ground area is available for the installation of the vertical farming systems. The impact of the different urban density indicators was analysed. The Hb alone has a minor 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 187 influence on the ratio of self-sufficiency (R2=0.2). The increase of Hb results on a counterbalance effect: first it allows larger building facades with a higher potential for the installation of planters, but it also means a larger amount of residents. The second factor is more influential and make that, in general, the higher the building the less potential for self-sufficiency. Cs has a higher impact on food self-sufficiency because the amount of land available for farming is directly proportional to Cs. Therefore, there is a stronger correlation between food self-sufficiency and Cs (R2= 0.65). PR, shown in Figure 4b, and population density have the strongest correlation: R2= 0.96 and R2= 0.95 respectively. This is expected because the building and population density take into account both the horizontal (Cs) and vertical (Hb) densities factors. Average illuminance per hour (klx) Percentage illuminance >10 klx (%) 45 90 40 85 35 80 75 30 70 25 65 20 60 15 55 10 50 1‐1 2‐1 3‐1 Ground a) 4‐1 5‐1 1‐1 Facade 2‐1 3‐1 4‐1 Ground b) 5‐1 Facade (a) Average illuminance levels per hour for the ground and facade points and (b) Figure 3 average percentage of time in which illuminance levels are higher than 10 000 lux. Ratio of food self‐sufficiency vs plot ratio Ratio of food self‐sufficiency 3 3 2.5 2.5 2 2 1.5 1.5 1 1 0.5 0.5 0 0 1‐1 a) y = 2.0711x‐1.143 R² = 0.9605 2‐1 3‐1 Conventional 4‐1 0 5‐1 Hybrid b) 1 2 Plot Ratio 3 4 5 6 7 Power (Plot Ratio) (a) Ratio of food self-sufficiency for the conventional and hybrid cultivation methods Figure 4 (ratio ≥ 1 self-sufficient) and (b) Correlation of the ratio of food self-sufficiency to plot ratio. Solar energy potential and energy self-sufficiency Based on the assumption of having 80% of the roof surface covered by PV panels and by considering BIPV as shading devices (30cm) on 70% of the facade perimeter on each floor the following results shown in Figure 5 were obtained in terms of energy self-sufficiency. As expected, the highest 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 188 energy output was obtained on the densest case (5-5 with PR = 5.8) due to the larger total roof surface and facade perimeter (more and higher buildings). However, when calculating the energy output relative to the amount of residents the cases with the lowest Hb (<42 m, cases 1-1, 2-1, 3-1, 4-1 and 5-1) are the only ones achieving energy self-sufficiency (> 98%) as shown in Figure 5a. This is also evident in 2 Figure 5b which shows a strong negative correlation (R =0.98) between energy self-sufficiency and Hb. The taller the building the lower the energy self-sufficiency due to the fact that the increase of PV panels does not counteract the effect of the larger amount of population on energy demand. Different from food self-sufficiency, PR has a much lower correlation with the energy self-sufficiency (R2=0.43). This may be explained by the fact that even with the same plot ratio, two cases may have different roof area for PV panels. I.e. cases 1-5 and 5-1 (PR = 3) have 37% and 77% of energy self-sufficiency respectively. PV output (MWh) and energy self‐sufficiency 20 18 16 14 12 10 8 6 4 2 0 Ratio of energy self‐sufficiency vs height (m) 1.4 1.4 1.2 1.2 1 0.8 0.6 0.4 1‐1 2‐1 3‐1 PV output a) 4‐1 1.0 y = 20.351x‐0.825 R² = 0.9863 0.8 0.6 0.2 0.4 0 0.2 20 5‐1 Self‐sufficiency b) 40 60 Building height 80 100 120 Power (Building height) (a) Energy output from PV panels (roof + facade) and energy self-sufficiency and (b) Figure 5 correlation of the ratio of energy self-sufficiency to building height. Sunlight availability on higher latitudes A comparison was made between Singapore and Hanoi in terms of DA (%) as shown in Table 3. The reduction of the DA in Hanoi is significant for the ground if PR is higher than 3. For the facade sunlight availability, this reduction is significant for all PR. Therefore, the impact of higher densities on the reduction of sunlight for farming and energy harvesting becomes larger with higher latitudes. Table 3. Comparison between Singapore and Hanoi, DA (%) DA [%] Ground average Cases 1-1 3-3 5-5 Plot Ratio 0.8 3 5.7 Singapore Hanoi 87 83 75 81 73 55 Ratio [-] Hanoi / Sing 0.93 0.88 0.73 DA [%] Facade average Singapore Hanoi 73 70 66 52 46 42 Ratio [-] Hanoi / Sing 0.71 0.66 0.64 CONCLUSION This paper describes the process and results of the first stage of a study on solar availability on three typical public housing typologies. Twenty five cases corresponding to the point block typology were analysed in this paper. Sunlight availability was calculated in order to predict the potential of food (fruits and vegetables) and energy harvesting and the degree of self-sufficiency on each of the 25 cases. Ground and facade surfaces were considered for the farming activities while facade and roof surfaces were considered for the installation of PV panels. The results show that food self-sufficiency is achieved in 6 of the 25 cases corresponding to the cases with the lowest PR (0.8 ≤ PR ≤ 1.9) if a hybrid farming 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 189 method is applied (conventional + vertical). If conventional method of ground-based farming is used, only two cases achieve self-sufficiency. Regarding energy harvesting, the cases with the lowest building height (< 42 m, < 14 storeys) achieve energy self-sufficiency due to the maximum exposed area with PV per amount of residents. Therefore, the indicators having the highest impact on the food and energy selfsufficiency are the plot ratio and building height respectively. However, this may not be fully applicable on other typologies and urban forms. Site coverage is still a crucial factor in providing food autonomy in urban areas due to the higher importance of ground than facade surfaces for the total food production. From this study we can conclude that in tropical regions the reduction of food and energy selfsufficiency due to denser urban environments is more a consequence of the reduction of the farming and PV area in relation to the total population than to the reduction of the sunlight availability. However, as shown in the case of Hanoi, with higher latitudes and a lower frequency of the sun near the zenith, the impact of the surrounding obstructions on reducing the sunlight availability increases. The other two typical housing typologies in Singapore, ‘slab’ and ‘contemporary’, will be the continuation of this study. In addition, the influence of the facade and plot orientations and of the sunlight availability at different facade heights will also be analysed together with the integration of other types of energy harvesting and conservation technologies like solar thermal and algae bioreactors. These studies will provide the basis for further environmental and energy assessments as well as a framework for a more comprehensive discussion about the impact of food and energy self-sufficiency strategies on several urban indicators in pursuit of a drastic carbon footprint reduction. ACKNOWLEDGMENTS This research is funded by the Academic Research Fund (AcRF) Grant from the Ministry of Education (MOE) and the National University of Singapore. REFERENCES Astee, L.Y., and Kishnani, N. T. 2010. Building integrated agriculture: Utilising rooftops for sustainable food crop cultivation in Singapore. Journal of Green Building, 5 (2): 105-113. Cheng, V., Steemers, K., Montavon, M., and Compagnon, R. 2006. Urban form, density and solar potential. The 23rd Conference on Passive and Low Energy Architecture, Geneva, Switzerland. Conover, C. and Flohr, R. 1996. Light, fertilizer and cultivar selection affect growth and yield of containerized patio tomatoes. Commercial Foliage Research Reports, 96 (1). Despommier, D. 2013. Farming up the city: The rise of urban vertical farms. Trends in Biotechnology, 31 (7): 388-389. Ibarra, D., and Reinhart, C.F. 2011. Solar availability: A comparison study of six irradiation distribution methods, Proceedings of Building Simulation 2011, Sydney: 2627-2634. Indraprahasta, G.S. 2013. The potential of urban agriculture development in Jakarta. Procedia Environmental Sciences, 17: 11-19. Kanters, J., and Horvat, M. 2012. Solar energy as a design parameter in urban planning. Energy Procedia 30: 1143 – 1152. Levinson, R., Akbari, H., Pomerantz, M., and Gupta, S. 2009. Solar access of residential rooftops in four California cities. Solar Energy, 83 (12): 2120-2135. Reinhart, C.F., and Walkenhorst, O. 2001. Validation of dynamic radiance-based daylight simulations for a full-scale test office with outer venetian blinds. Energy & Buildings, 33 (7):683-697. Lee, R.X., Jusuf, S.K., and Wong, N.H. 2013. The study of height variation on outdoor ventilation for Singapore’s high-rise residential housing estates. International Journal of Low-Carbon Technologies, 0:1-19. Singapore Power Group Ltd. 2014. http://www.singaporepower.com.sg/ US Department of Energy. EnergyPlus weather data. (accessed on September 2013) http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data3.cfm/region=5_southwest_paci fic_wmo_region_5/country=SGP/cname=Singapore Ward, G., and Shakespeare, R. 1998. Rendering with radiance. The Art and Science of Lighting Visualization. Morgan Kaufmann Publishers. Zhang, J., Heng, C.K., Malone-Lee, L.C., Jun, D.C.H., Janssen, P., Kam, S.L., and Tan, B.K. 2012. Evaluating environmental implications of density: A comparative case study on the relationship between density, urban block typology and sky exposure. Automation in Construction, 22: 90–101. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 190 The Cooling Effect of Green Strategies Proposed in the Hanoi Master Plan for Mitigation of Urban Heat Island Andhang Rakhmat Trihamdani, M.Eng. Han Soo Lee, Dr. Tetsu Kubota, Dr. Takahiro Tanaka, Dr. [Grad. School for IDEC, Hiroshima Univ., Japan] andhang.rt@gmail.com [Grad. School for IDEC, Hiroshima Univ., Japan] [Grad. School of Science and Engineering, Saitama Univ., Japan] [Grad. School of Engineering, Hiroshima Univ., Japan] Tran Thi Thu Phuong, M.Eng. [Vietnam Institute Urban and Rural Planning, Vietnam] Kaoru Matsuo, M.Eng. [Grad. School of Engineering, Hiroshima Univ., Japan] ABSTRACT This study aims to assess the impacts of land use changes brought by the Hanoi Master Plan on its urban climate using the Weather Research and Forecasting (WRF), focusing on the cooling effect of the green strategies that were proposed in the master plan. The results show that even after implementing the master plan, the peak air temperature in the urban areas still remains at the same level of 41°C. However, the expansion of built-up areas largely increases the UHI intensity and raises the nocturnal air temperatures in the built-up areas. The centralized green spaces proposed in the master plan is seen to be insufficient to mitigate UHIs compared to the equally distributed green spaces. The urban air temperature in Hanoi is increased when the westerly or south-westerly Foehn winds flow over the city during the daytime. In contrast, relatively strong and cool southerly winds prevail during the night-time and contribute to reduction in the nocturnal air temperature in the city. INTRODUCTION Emerging economies in Southeast Asia are likely to see serious energy shortages, especially in terms of electricity, due to the recent rapid economic growth. Most of the cities in this region have a hot and humid climate during the summer months, and the growing energy consumption caused by airconditioning in buildings is, therefore, a major concern. Meanwhile, these cities are suspected of already experiencing urban heat islands (UHIs) as a result of the rapid urbanization. The further rise of urban temperature would lead to a significant increase in energy demand for cooling. Currently, these Southeast Asian cities tend to propose large-scale master plans and increase their urban population. This would result in a dramatic change in land use and therefore the urban climate. In Hanoi, a long-term urban development plan, namely the Hanoi Master Plan 2030, was implemented in 2011 with the aim of developing the city into a more sustainable capital (VIUP, 2011). One of the key concepts of the master plan is to maintain abundant green coverage in the city through a systematic green network, including green buffers and green belts, for example. Although the master plan had considered various environmental issues such as air pollution, water quality and eco-system, this assessment did not take into account the impact of this development on UHIs in the city. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 191 Therefore, the objective of this study is to investigate the UHI effects in Hanoi under the present land use conditions as well as under those conditions proposed by the master plan, focusing especially on the cooling effects of the green strategies. Numerical simulations, specifically meso-scale urban climate modelling using Weather Research and Forecasting (WRF) are performed for this purpose. HANOI MASTER PLAN 2030 As the capital city of Vietnam, Hanoi is the second largest city of the country, which make up an area of about 3,300 km2. The city center is located in the delta area along the Red River, which is about 90km inland from the coastal line. Hanoi experiences a typical tropical monsoon climate, comprising a hot-humid season (April to October) and a cool and relatively dry season (November to March). The southeast monsoon wind prevails during the hot season. The maximum monthly average air temperature is observed in June, which is about 30°C, while the minimum average value of about 16.5°C in January. Figure 1 Land use and land cover for (a) the current status and (b) the Hanoi Master Plan 2030. Source: VIUP, 2011 The Vietnam government officially implemented the Hanoi Master Plan 2030 in July 2011. In the master plan, the population of Hanoi is projected to reach 9.2 million by 2030. The main target of the master plan is to develop Hanoi as a green-cultured and civilized-modern city. In order to achieve that target, the master plan proposes a series of spatial development strategies for the capital city. One of them is, as described before, the green network consisting of two major green strategies which are the green belts and the green buffers. As a result, the green coverage in the master plan would account for about 52% of the total land of the city. Figure 1 shows the land use changes before and after the implementation of the master plan. To meet the demand of expanding urban development, 28% of the city’s natural land will be allocated for urban construction land by 2030. In total, the constructed land will rise sharply by almost three times, from 46,340 ha (14%) to more than 129,500 ha (39%). DATA AND METHODOLOGY Weather Research and Forecasting Meteorological modelling is performed to obtain basic weather elements such as air temperature, humidity, and surface wind, using the Advanced Research Weather Research and Forecasting (WRFARW) model (version 3.5) (Skamarock et al., 2008). WRF is a three dimensional non-hydrostatic mesoscale meteorological model developed at the National Center for Atmospheric Research (NCAR) based on the non-hydrostatic compressible form of the governing equations in spherical and sigma coordinates with physical processes, such as cumulus scheme, microphysics, planetary boundary layer (PBL) processes and atmospheric radiation processes, incorporated into a number of physics parameterizations. This model has been widely used in atmospheric research and operational forecasting needs. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 192 (b) (a) Figure 2 (a) Computational domains for the WRF simulation. (b) Domain 4 covers all of Hanoi City (100x100 grid points), with 1 km resolution. The green strategy areas are indicated by colors. Green represents the green belts and red is for the green buffers. The WRF simulations in this study adopt an interactive grid nesting with four domains that have horizontal resolutions of 27, 9, 3 and 1 km for domains 1, 2, 3 and 4, respectively (Figure 2a). The domain 4 covers all of the Hanoi City (Figure 2b). The 30 sigma levels are set up vertically. The initial and lateral boundary conditions are imposed every 6 hours using the NCEP FNL Operational Global Analysis data with 1°x1° latitude-longitude resolution (http://rda.ucar.edu/datasets/ds083.2/). Simulation scenarios As shown in Figure 2b, the green belts are large green spaces located inside the urban development area with the aim of improving the micro-climate conditions, while the green buffers are the boundary space between the existing urban areas and expanded urban clusters. To study the effect of these strategies on the urban climate in the future, a comparison is performed between the current condition (hereafter referred as U_CUR) and the master plan scenario (hereafter referred as U_HMP). Both of the green strategies are implemented in the U_HMP. In order to evaluate the effectiveness of the green strategies, two scenarios are designed which are U_NOGREEN and U_GREEN. In U_NOGREEN, both of the green strategies are not taken into account and converted into the built-up area (Figure 3c). Meanwhile in U_GREEN, the same amount of strategic green areas in the master plan are redistributed to new locations, resulting in smaller green spaces but equally distributed in all of the city (Figure 3d). (a) (b) (c) (d) Figure 3 LULC of domain 4 for (a) U_CUR, (b) U_HMP, (c) U_NOGREEN, and (d) U_GREEN. Model validation The simulation was conducted for one month from 00:00 UTC 1 to 00:00 UTC 30 June in 2010, which was the hottest period over the period of 2000-2012. Subsequently, simulation results on 17 June of a typical hot sunny day is mainly analyzed in the following sections. Figure 4 presents the comparison of air temperature and wind speed between the simulated and the 3-hourly observation data at Lang station (21.02°N 105.8°E) located in the Hanoi city center (see Figure 1a). Both of the simulated air temperature (at 2 m above ground surface) and wind speed (at 10 m above ground surface) show good 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 193 agreement with the observed values with a coefficient of determinant of 0.92 and 0.13, respectively. The simulation conditions are shown in Table 1. (a) (b) Figure 4 Comparison between the observed and simulated value for (a) air temperature and (b) wind speed at Lang station located in Hanoi city center. Items Simulation period Vertical grid Horizontal grid Meteorological data Table 1. WRF Simulation Conditions Conditions Land use/land cover (LULC) data Microphysics Long-wave radiation Short-wave radiation PBL scheme Cumulus scheme Surface scheme Surface layer 00:00 UTC 1 to 00:00 UTC 30 June in 2010 30 layers 100x100 grids NCEP FNL Domain 1 and 2: USGS (default); Domain 3:GLCNMO; Domain 4: ALOS ANVIR-2 and National Digital LULC data WSM 3-class RRTM long-wave scheme Dudhia short-wave scheme YSU Scheme Kain-Fritsch scheme NOAH-LSM Monin-Obukhov scheme RESULTS AND DISCUSSION Urban climate in current status and master plan condition Figures 5 and 6 present the spatial patterns of the simulated air temperature at 2 m above the ground surface and the winds at 10 m above the surface for U_CUR, U_HMP, U_NOGREEN and U_GREEN at 1:00 and 16:00 LST on 17 June, respectively. This section compares U_CUR and U_HMP. At night, the highest air temperature is observed not in the city center but in the lee of western mountainous region in both scenarios (Figure 5). This result is partly due to the effect of Foehn wind from Laos (Nguyen & Reiter, 2014). As shown in Figures 5 and 6, westerly or south-westerly winds prevail most of the day in the western mountainous region. While these westerly winds pass over the mountain range situated near the border of the Hanoi region, the air gets drier and the temperature increases rapidly. These westerly winds flow over most parts of the simulated area during the daytime (9:00-17:00) and bring hot and dry air to the city. In contrast, relatively cool south-easterly or southerly winds prevail over plain areas in the east of 105.65°E during the night-time (18:00-8:00) (Figure 5). In U_CUR, the minimum nocturnal air temperatures in urban and suburban areas are recorded as approximately 31-32°C, while the air temperature is approximately 32-33°C in the lee of mountainous areas. The daily peak air temperatures were obtained in most parts of the city around 16:00 in both conditions (Figures 6a and 6b). Although the peak air temperatures remain almost the same level even after implementing the master plan, the high air temperature areas of 40-41°C expand widely over the planned built-up areas in U_HMP. At 1:00, the air temperature difference between U_CUR and U_HMP increases by up to 2-3°C over the expanded built-up areas under the master plan condition. During the daytime, the increase in air temperature over the expanded built-up areas is not significant, which is up to 1°C, compared to that in the night-time. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 194 Figure 5 Air temperature and wind distribution at 1:00 on 17 June for (a) U_CUR, (b) U_HMP, (c) U_NOGREEN, and (d) U_GREEN. Figure 6 Air temperature and wind distribution at 16:00 on 17 June for (a) U_CUR, (b) U_HMP, (c) U_NOGREEN and (d) U_GREEN. Master plan condition and that without green network In U_NOGREEN, the planned green spaces are turned into the built-up areas. As expected, the high air temperature areas of 33-34°C are enlarged over the expanded built-up areas at 1:00 (Figure 5c), while the high air temperature areas of 40-41°C are found over the same expanded built-up areas at 16:00 (Figure 6c). The increase in air temperature in the built-up areas becomes noticeable during the nighttime (1:00), with the maximum increase of 1°C from U_HMP to U_NOGREEN. The incremental zones of the air temperature from U_HMP to U_NOGREEN correspond to the expanded built-up areas and those transformed from the planned green spaces. The diurnal average air temperatures of seven days (13-20 June) in Lang and Tu Liem for three scenarios are shown in Figure 7. These two locations represent the existing city center (Lang) and the green areas located in the green belt (Tu Liem), respectively (see Figure 1a). As shown, the average air temperatures for U_CUR and U_HMP at Tu Liem are lower than the corresponding air temperatures at Lang by up to 1°C during the night-time and by up to 0.5°C during the daytime, respectively. This is mainly due to the difference of land cover between the two places. Tu Liem still maintains the natural land cover (i.e. mixed shrubland or grassland) even after the implementation of the master plan. However, if all the green areas are converted to built-up areas, the air temperatures at Tu Liem are increased significantly by up to 1.5°C at night, reaching the similar values to those at Lang. Meanwhile, the removal of green strategies results in a slight increase in the average air temperatures in the existing city center (i.e. Lang). As shown in Figure 7a, the nocturnal average air temperatures in U_NOGREEN are approximately 0.3°C higher than those in U_HMP. As shown in Figure 7, in both Lang and Tu Liem, the average air temperatures in U_CUR increase slightly faster than those in the other scenarios during the morning hours from 8:00 to 10:00. This occurs, although not prolonged, probably due to the difference of heat capacity of the whole urban fabric between U_CUR and the other scenarios. Thermal capacity of the whole urban surface is largely increased by the expansion of the city, which in turn, slows the increase in air temperature during the morning in the cases of U_HMP and U_NOGREEN than that in U_CUR. On the other hand, the green areas induce a relatively faster heat release as illustrated in Figure 7b. As shown, except for the above mentioned morning hours, the average air temperatures in U_CUR and U_HMP in Tu Liem are significantly lower than those without the green areas (U_NOGREEN). Nevertheless, the proposed green areas do not result in the large reduction in air temperature at Lang because the city center is located 4 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 195 km away from the green belt. (a) Figure 7 (b) Diurnal variation of average air temperature over seven days (13th-20th June) at (a) Lang and (b) Tu Liem. Impacts of the distribution of green spaces Figure 8 shows the air temperature distribution over the built-up areas in U_HMP and U_GREEN at 1:00 and 16:00 on 17th June 2010, respectively. As shown in Figure 8a, the nocturnal air temperatures in the built-up areas range from 31-34.5°C, with the proportion of the area at specific temperature peaking at 33.5°C. In U_GREEN, the proportions of the area with the air temperature of 32.5-34°C are reduced, while the proportions of the area with the lower air temperatures of 31-32°C are increased. This air temperature shift indicates the reduction in UHI intensity after the green spaces are equally distributed within the city. In the daytime, the amount of areas with the air temperature of 39.5°C is reduced and shifted down to the ranges of 38.5-39°C, indicating the temperature reduction by 0.5-1°C (Figure 8b). The impact of the relocation of the green spaces on the reduction of UHI intensity is greater in the night-time than that in daytime. (a) (b) Figure 8 Air temperature distribution over the built-up areas on 17th June at (a) 1:00 and (b) 16:00. Effects of prevailing winds Figure 9 depicts wind roses at the city center (Lang) over the one month simulation period (June 2010) analyzed by the corresponding air temperatures (a, b) and the wind speeds (c, d) at 10 m above the ground surface during the daytime (09:00-17:00) (a, c) and the night-time (18:00-08:00) (b, d), respectively. It is noted that the three scenarios did not show any remarkable differences in terms of ground surface wind conditions (see Figures 5 and 6). Therefore, the data from U_CUR are analyzed in this section. As seen in the previous section, the southerly and south-easterly winds with relatively high wind speeds of 4-7 m/s prevail during the night-time over the whole month (Figure 9bd). In contrast, the prevailing wind direction differs during the daytime as shown in Figure 9ac. The westerly or southwesterly winds prevail over approximately 38% of the month (Figure 9ac). Nevertheless, the northeasterly and south-easterly winds also prevail during the same period. It is interesting to note that when the westerly or south-westerly winds blow over the city center, the air temperatures are relatively higher (38-41°C) than the air temperatures under the conditions of north-easterly or south-easterly winds (2840°C), though the corresponding wind speeds do not differ significantly between the different wind directions (Figure 9ac). This result supports the assumption that urban air temperature in Hanoi is increased when the westerly or south-westerly Foehn winds flow over the city during the daytime. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 196 (a) (b) (c) (d) Figure 9 Wind roses at Lang station in June 2010, analyzed by (a) air temperature (daytime), (b) air temperature (night-time), (c) wind speed (daytime), and (d) wind speed (night-time). Figure 10 Variations in the simulated air temperature and wind speed from U_HMP and U_NOGREEN along the west-east cross-section of the meridional line at 21.04°N. Figure 11 Variations in the simulated air temperature and wind speed from U_HMP and U_NOGREEN along the south-north cross-section of the meridional line at 105.81°E. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 197 Figures 10 and 11 show the spatial variations in the air temperature and wind speed along the cross sections (see Figure 2b) of domain 4 from U_HMP and U_NOGREEN on 17 June. The horizontal indicator for the LULC category in each panel is based on the master plan. Therefore, from Figures 10 and 11, the effects of the green spaces and surface winds on the air temperature can be discussed. This cross-section traverses two mountains at heights of approximately 260 m and 280 m in the western region at 105.22°E and 105.40°E, respectively. The discrepancies between the air temperatures in the strategic green areas under U_HMP and the built-up transformed areas in U_NOGREEN are large, up to 2-3°C, when the relatively strong and cool southerly winds prevail at 1:00 (Figure 10a). In this circumstance, the reduction in the air temperature due to the effect of the green strategy can be observed in some parts of surrounding built-up areas, up to 1°C. These discrepancies are reduced in the daytime as the wind direction changes from the south to the west (Figure 10b). The surface wind conditions largely affect the air temperature distribution. As shown in Figure 10a, the leeward sides of both mountains receive relatively higher westerly winds and the air temperature increase while flowing down the slopes of the mountain at 105.4°E (i.e. the Foehn effect). The southerly winds flow from the coastal area and pass through the irrigated croplands in the south at 20.85°N and bring relatively lower air temperature before entering the urban areas (Figure 11a). From the above point, the air temperature in U_NOGREEN gradually increases towards the north and then rapidly decreases. Then, the air temperature simulated in U_NOGREEN depicts a direct and linear response to the corresponding winds (Figure 11a). The cooling effect of the surface wind are clearly seen when the air temperature in built-up areas over the eastern half from 105.65°E drops by nearly 1°C particularly when the southerly winds prevail at night (see Figure 5ab and 10a) CONCLUSIONS Based on the results of the numerical experiments, the main findings are summarized as follows: 1. In general, the daytime peak air temperature rises up to 40-41°C over the built-up areas in the city center in the current condition. 2. If the LULC is changed according to the master plan, the daytime peak air temperature is predicted to remain at almost the same level as the current condition. However, the new hotspots would expand widely over the planned built-up areas. 3. The strategic green spaces would not sufficiently mitigate UHIs in the city because they are located far away from the city center. On the other hand, the equally distributed green areas show a better performance in the reduction of UHI intensity, especially at night. 4. The south-westerly or westerly winds are dominant in the daytime and bring hot and dry air to the city, likely due to the Foehn wind. The air temperature in Hanoi City is increased when these Foehn winds flow over the city during the daytime. The occurrence of this phenomenon was found to be approximately 38% in June 2010. In contrast, relatively strong and cool southerly winds prevail during the night-time and contribute to reduce the nocturnal air temperature in the city. Due to the discrepancy in term of wind condition, heat islands appearing in the western built-up region of Hanoi are found to be more intense than in the city center at night. ACKNOWLEDGMENTS We sincerely appreciate the generous supports given by the Vietnam Institute of Urban and Rural Planning. This research was supported by a grant from Mitsui & Co., Ltd. Environment Fund. REFERENCES Nguyen, A.-T., & Reiter, S. (2014). A climate analysis tool for passive heating and cooling strategies in hot humid climate based on Typical Meteorological Year data sets. Energy and Buildings, 68, 756– 763. Skamarock, W. C., Klemp, J. B., Gill, D. O., Barker, D. M., Duda, M. G., Wang, W., & Powers, J. G. (2008). A Description of the Advanced Research WRF Version 3. VIUP. (2011). The Hanoi Construction Master Plan 2030. Hanoi: VIUP. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 198 Tall Buildings and the Urban Microclimate in the City of London Sujal V. Pandya, B. Arch., M. Sc, COA Luisa Brotas, Arch, PhD, ARB, RIBA Low Energy Architecture Research Unit London Metropolitan University, United Kingdom Low Energy Architecture Research Unit, Sir John Cass, Faculty of Art, Architecture and Design, London Metropolitan University, United Kingdom l.brotas@londonmet.ac.uk ABSTRACT The design of tall buildings can be quite complex not only in architectural, structural or façade system terms but also in their interrelations with the environment. Recent design developments have shown a growing consideration for this micro-climate interface as a part of a sustainable design strategy for tall buildings towards bioclimatic urban fabrics. Architects and urban planners recognize that the importance of activities at street level is a constant reality. Many efforts are made to develop sustainable skyscrapers where the co-relation with the surrounding environment influences the architecture, land use pattern, public realm and street activities. The aim of this study is to understand the environmental or bioclimatic factors, which take part in the creation of microclimate relations between tall buildings and the urban city fabric. The research includes a brief understanding of bioclimatic design of tall building focusing in its relation to surrounding neighbourhoods. The impact of environmental factors namely wind flow patterns around buildings, daylight availability and overshadowing, air temperature and humidity, as well as the urban heat island phenomena, landscape and albedo of materials, vehicular & pedestrian movements and activity patterns at street levels in tall building clusters is addressed. A series of environmental variables measured during summer and simulated results of Canary Wharf and Liverpool Street clusters are compared to understand how the urban microclimates are affected within two different urban fabric typologies in London. Outcomes of literature review and these two case studies give useful guidelines to be utilized in the process of Sustainable urban design. INTRODUCTION Cities all around the world are recognized by their skylines (Figure 1) and there are many drivers to its growth. Land scarcity and high real estate value, commercial opportunities and corporate demand, as well as the attraction to the cluster are often allied with a growing city population and its social needs. . Figure 1 Canary Wharf skyline, London. Sujal V. Pandya is a Visiting Researcher at the Low Energy Architecture Research Unit at London Metropolitan University. Dr Luisa Brotas is Course Leader of the MSc Architecture Energy and Sustainability and co-Director of the Low Energy Architecture Research Unit at the Sir John Cass Faculty of Art, Architecture and Design at London Metropolitan University, London, United Kingdom. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 199 The City of London is a world financial centre and needs to project an image of power and efficiency. The scarcity of land within the neighbourhood is further restricted by a surrounding mesh of residential buildings, on average with less than 3 floors (London has around 52% dwelling of the type terraced, semi detached and detached). High rise buildings being structures with 12 floors minimum contrast significantly with the low height London residential stock. According to the England housing survey, London only has 20% of dwellings that are 3 or more storeys in height. (EHS, 2012) This contrast with high-rise clusters developed within the city and in particular in the city of London neighbourhood (Canary Wharf Tower is 235m height). Tall buildings in a cluster are a growing tendency as a solution to the need of office space within the limited land available. The Greater London Authority promotes this concept as an efficient way to build offices that need to be energy efficient and with good public transport access and capacity, in line with sustainable cities and the Agenda 21. (GLA, 2014) This creates an opportunity for sustainable development in a holistic approach that goes beyond the architecture, structural or façade system to include the interface between the street environment and the building. Recent design developments have shown major consideration over this microclimate interface as a part of a sustainable design strategy for tall buildings in the direction of making bioclimatic urban fabrics. Likewise local climate variations influence the urban boundary and should be given much thought. The amount of solar radiation can be significantly reduced in urban areas as a result of overshadowing, air pollutants and wind flow patterns adding to the greenhouse effect. The Urban Heat Island phenomenon can be aggravated by high-density materials, surfaces with low albedo, reduced green spaces (evapotranspiration) and anthropogenic heat. This impacts the surface and air temperatures near the ground. Conversely high-rise buildings require special attention to the effect of strong winds at high levels. At street level wind may be much reduced overall (though it may increase locally due to a ‘wind tunnel’ effect) and buildings casting shadows may create dark and cold spots for pedestrians. The character of the urban cluster / fabric play a big role in such microclimates. (Littlefair, 2000; Erell, 2011) BIOCLIMATIC DESIGN CONSIDERATIONS FOR URBAN FABRICS WITH TALL BUILDINGS The effect of climate on the activities of occupants and its health is the base of Bioclimatology. The bioclimatic design approach is not limited to the building but extends beyond and involves external environmental conditions, site location, geography, surrounding buildings and their aesthetic expressions, land use pattern, the neighbourhood cluster, public accessibility and activities. (Clair, 2010) The major bioclimatic design considerations for high-rise clusters are: a. Urban geometry (Figure 2): variation in heights and in-between distances of buildings (street canyon); b. Local climate: external air temperature, relative humidity and prevailing wind direction and speed; c. Orientation and overshadowing (Figure 3): position of tall buildings with respect to south (northern hemisphere) and its exposure to solar radiation and overshadowing over low-rise buildings within the cluster; d. Solar access, solar radiation and albedo (Figure 4); e. Urban canyons (Figure 5). Figure 2 Urban Geometry. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Figure 3 Orientation and Overshadowing. 200 SVF Reflection L Absorption . H Absorption Canyon Axis W Figure 4 Solar radiation and albedo. Figure 5 Urban canyon. CASE STUDIES This study assessed two different high-rise urban typologies in London to understand the interaction/impact of buildings at the urban scale. Results from simulations as well as real data recorded in a summer period are presented next. Cluster 1: Canary Wharf, London Coordinates: 51.5036° N 0.0183°W for 1 Canada Square, Figure 6 Solar access and solar radiation: The Central garden is the biggest open space within the Canary Wharf cluster. It works as a ‘buffer space’ as a result of the microclimate conditions originated within these high-rise buildings of the square. The average direct solar radiation ranges from 120Wh to more than 1200Wh during the summer months from June to August. As seen in Figure 7, the road along the edge of the south buildings receives very low radiation during day time (9 am to 5 pm), around 120Wh, while the open space in front of plaza receives much higher levels of solar radiation as facing buildings are distant enough to avoid significantly overshadowing. This correlates well with the monitored air st temperature made at the plaza, the garden and the south-west corner low rise block during July 1 to July th 13 2011 as seen in Table 1 and elsewhere (Pandya, 2011). 0 200m 550 m Figure 6 Canary Wharf cluster. Central Garden Road 550 m N Plaza Figure 7 Direct solar radiation – June, July, August 9am to 5pm. The simulation results of direct solar access on vertical surfaces, as shown in Figure 8, reveal that the ‘unobstructed’ south edge buildings receive almost 6000Wh/m2 during Summer solstice. Buildings on the north side of the central garden have much lesser solar access. This amounts to around 2000Wh at ground level, increasing to around 4000Wh on top of the façade, as surrounding buildings no longer significantly obstruct high storeys. This has an impact on temperature and daylight availability to the spaces. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 201 Figure 8 Solar access on Summer solstice. Urban Air Flow: For both clusters, CFD Simulations with ENVI-MET software were made. st Results are presented for July 1 at 9am. The average wind speed is 3m/s and the prevailing wind direction in July is southwest (as per meteorological data for London). Figure 9, location A highlights that within the narrow canyons the wind flow vectors are longer which means high speed, while at the open spaces such as Reuters’ plaza or at the central garden, the wind speed is less intense and there are no significant deviations in the direction as a result of lack of obstructing tall buildings. A 1 Reuters’ Plaza Central Garden B Reuters’ Plaza Figure 9 Wind flow simulation plan on July 1st, 9am. Figure 10 Wind flow simulation - Section 1. Figure 10 shows that when buildings are fairly distant, downward wind loses its speed when it enters the larger open space. As it reaches the opposite high-rise building, the speed increases gradually. At the Canary Wharf cluster the corners of the buildings or edge conditions are most affected by incoming wind with higher speeds and get deviated due to obstructing edges as seen in Figure 9. This generates wind turbulences with high wind speed (see Figure 11). As both north and south edges are blocked by highrise buildings (Figure 9, location: B) wind turbulence and speed increases. Reuters’ Plaza Figure 11 Wind flow contour lines for wind speeds - Section 1. Cluster 2: Liverpool street, Bishopsgate, London Coordinates: 51.31° N, 0.4° W, Figure 12 Solar access and solar radiation: Liverpool Street is surrounded along their edges with continuous building facades. These urban canyons are mainly oriented 20° to 26° from north. Figure 13 shows that narrow streets, having smaller canyon ratios (width to height) receive a smaller amount of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 202 direct solar radiation (street in front of Tower 42). The plaza, an important public space in the neighbourhood, receives more than 2500Wh. 0 200m N Garden Swiss Re Tower 42 Plaza Figure 12 Liverpool street, Bishopsgate cluster Figure 13 Direct solar radiation June, July, August. 9am to 5 pm Figure 14 Solar Access on Summer solstice In case of the Swiss Re building, the rear open space is used as a café and restaurant esplanade during summer, as it benefits form direct sun from southeast and south. This is not the situation at the front entrance as it is blocked by continuous buildings, and solar radiation is reduced to 500Wh. Figure 14 shows that the orientation of canyons/streets and buildings heights along the edge of them, can significantly affect the solar access at ground level, from 400 to 1500Wh/m2. Solar access on the façade at higher floors can reach 7500Wh/m2. At this cluster most of the high-rise buildings are fully glazed with anti-reflective glass, other surfaces also have reduced albedo. This prevents a high amount of solar radiation to be reflected towards the surrounding cluster. Conversely it may minimize problems of reflected glare. However, it is likely to increase the cooling loads of the spaces as well as aggravate thermal stress at the façade. Urban Air Flow: The irregular urban geometry of the Liverpool street cluster, with varying canyon orientations and monotonous flat building fabrics, shows dramatic changes in wind flows. Figure 15, Location A – Bishopsgate road is a narrow canyon with building heights around 25 to 40m. The length of the canyon is about 65m and its orientation is southwest to northeast. This geometry permits the southwest wind with high speed entering the canyon. Figure 16, Section 1 shows airflow vectors going in the negative direction, upward, as a result of the close vertical surfaces on both sides of the in-between space. As the aspect ratio (H/W) is quite high, the incoming wind speed is higher than the one flowing above the roof of the building. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 203 A 1 st Figure 15 Wind flow simulation on 1 July at 9am. Figure 16 Wind flow simulation - Section 1. Figure 17 Wind flow contours for wind speed - Section 1. The irregular street pattern and building edges create negative turbulences (Figure 17) that affect the wind flow and speed. This phenomenon is stronger around tall buildings. COMPARATIVE ANALYSIS Table 1 Comparative analysis of Canary Wharf & Liverpool Street clusters, London. Clusters Total Built up area (m²) Canary Wharf 302,500 (Cluster coverage) 67,624 (built---up) 302,500 (Cluster coverage) Liverpool Street 302500 m² High Rise building foot print (m²) Tall building Height range (m) 302,500 (Cluster coverage) 35,769 (built---up) 105 235 153 200 140 302,500 (Cluster coverage) 125 158,114 Urban Canyon profiles Canyon Orientations N 235 m Canary Wharf S 242 100 93 16108 Urban Geometry in Plan Clusters 180 8° 98° N N S N Liverpool Street 26° 34° 116° 242 m S N N S The urban geometry of the Canary Wharf cluster has a regular organization of buildings along the road edges with open spaces within. The Liverpool Street cluster presents an irregular street pattern and the building footprint follow the available space. As seen in Table 1, the cross sectional profiles of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 204 canyons within Liverpool Street cluster are narrower on the north-south axis and the scattered high-rise buildings pop up in the low-rise settlement. This kind of geometry shows uneven height variations in the cross sectional profile of the canyons. Table 2 Comparative Analysis of Canary Wharf & Liverpool Street clusters, London. Street Pattern Clusters Sky View Factor Canary Wharf Direct Solar Radiation Map 0.35 0.37 street plaza garden N 0.13 0.18 0.23 Liverpool Street garden street plaza 0.252 AVG. 0.49 0.22 0.26 0.26 0.2 0.286 AVG. 1200 Wh max. avg. for summer months 2280 Wh max. avg. for summer months N Clusters Canary Wharf Green Cover (m²) 302,500 (Cluster coverage) Air Temperature (°c) 01/07/2011 25.5 25.1 24 13,244 (green cover) Plaza Liverpool Street Street 302,500 (Cluster coverage) 2,344 (green cover) Relative Humidity (%) 01/07/2011 22.9 23.2 Plaza Street 24.7 23.3 24.2 Garden Plaza Street 26.4 25.8 25.7 Garden 22.9 Garden Plaza Street Garden The garden area of Liverpool Street shows higher air temperature and lower humidity than the street and plaza places (lowest temperature and high relative humidity). Conversely the Canary Wharf garden has lower air temperature and higher humidity than other parts of the cluster. Similarly, the open space like the plaza has higher wind speeds and the garden has lower wind speed due to the trees blocking the landscape. The variations in wind speeds at different places is also dependent on the prevailing wind direction and the urban fabric characteristics such as building heights, vegetation, water body and so on. The comparison of spot measurements of air temperatures at the three locations: plaza, streets and garden areas, revealed that the plaza at Canary Wharf has high air temperature during afternoon hours. See Table 2. This is a result of the hard paved surfaces with high reflectance and better solar exposure versus a more shadowed garden with high evapotranspiration. The Liverpool Street plaza appears to be more affected by the proximity of the surrounding buildings, their irregular geometries (ie more surface area) with low surface reflectance and higher thermal mass, hence absorbing more direct solar radiation. Heat is released to the air later at night when temperatures are lower. The open space is also more in the shadow. Therefore during day the air temperature remains low. Taking this study in consideration the most prominent bioclimatic factors affecting the urban microclimate are: urban geometry, urban canyons, sunlight availability and solar radiation, urban air flow, urban landscape and urban air quality and surface materials. It is understood that all are interdependent. However, the geometry, orientation and fabric of the buildings strongly impact the amount of solar radiation being absorbed/reflected by building facades and streets. The climatic conditions and building geometry play a major role at promoting sunlight availability, ameliorating air temperature and promoting air flow/ventilation to spaces. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 205 CONCLUSIONS Environmental factors such as air temperature and humidity, wind speed and direction, solar radiation and daylight conditions, noise and air pollution at ground level can make open areas in cities an appealing and comfortable place for pedestrians. Results presented above discuss the influence of these factors within two high-rise clusters in London. Furthermore, they strongly impact the energy performance of surrounding buildings and its comfort conditions and should be given much thought towards sustainable urbanism within existing and future cities. Canary Wharf has a rich landscape planning in terms of public realm, environmental sustainability, ecology and usage. The central garden permits solar access, a pleasant airflow and acts as a sound barrier from busy roads nearby. Road includes coniferous trees, which helps disperse the pollution particles within the canopy layer. Deciduous trees at the street side provide protection from direct sun and wind turbulence whilst still allows air movement between the trees and the adjacent buildings for ventilation. Liverpool Street has a compact and dense urban fabric with reduced open spaces, which can be used for urban landscape. Trees in internal courtyards of buildings may be beneficial to the building and its users but not at the urban scale. In hot periods, when the temperature rises above comfort level, direct and the reflected solar radiation from building surfaces increase the Heat Island effect. To minimize this, soft paved surfaces such as sand gravels or green patches of lawn and tree cover are necessary. Climate resilience and demands for sustainable living are not limited to environmental friendly building design, but requires much more attention to bioclimatic factors affecting the urban fabric and its microclimate. Bioclimatic factors can be man made or organic but ultimately affect human living conditions and the environment. It is a necessity to tackle at both building and urban scale. ACKNOWLEDGMENTS The first author acknowledges the following individuals and institutions: his parents and family members, CEPT University in Ahmedabad, India; Dr Luisa Brotas, Course Leader of MSc Architecture Energy & Sustainability at London Metropolitan University, Dr Axel Jacobs and Professor Fergus Nicol from London Metropolitan University as well as London Metropolitan University for the Scholarship Award that made this study possible; The Architectural Association, School of Architecture in London, the London underground and Bus transport; and last but not least ‘The Almighty God’ for giving directions and courage to overcome all the obstacles and paths to success. NOMENCLATURE θ = the line running north-south and angle between canyon axis which determines the orientation of urban canyon W = width of street H = average height of buildings on both sides L = length of canyon α = angle between building top edge and street plane SVF = sky view factor REFERENCES Clair, P.S, 2010. The Climate of Tall buildings- An investigation of building height and bio-climatic design. Available at: http://www.peterstclair.com/pdf/The-Climate-of-Tall-Buildings-ScienceReview_LR.pdf [accessed: May 17, 2011] EHS, 2012. English Housing Survey: Homes report 2010, Department for Communities and Local Government, National Statistics, Crown copyright Erell, E. et. al, 2011. Urban Microclimate: Designing the spaces between buildings. London: Earthscan. GLA, 2014. Great London Authority. https://www.london.gov.uk/ [accessed: September 10, 2014] Littlefair, P.J, et. al, 2000. Environmental site layout and planning, London: BRE Publications Pandya S., 2011. Tall Buildings and Urban Microclimate: A Bioclimatic study of urban fabric, MSc Thesis, London Metropolitan University 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 206 Baseline Scenario of Energy Consumption of Urban Multi- Storey Residential Buildings in India Kanagaraj Ganesan 1 Bhanware, Prashant 1 1 Cusack, Kira2 Greentech Knowledge Solutions Pvt. Ltd. kanagaraj@gkspl.in Chetia, Saswati1 2 Jaboyedoff, Pierre2 Maithel, Sameer1 Effin’Art Sarl KEY WORDS: Urban residential buildings, Energy monitoring, Energy Performance Index ABSTRACT: This paper presents results of monitoring of energy consumption in sample urban residential buildings in India. The work was carried out under the Indo-Swiss Building Energy Efficiency Project (BEEP) as background research leading to the development of energy efficiency guidelines for the design of new residential buildings in the composite and warm-humid climatic regions of India. The work involved: a) Collection of monthly energy consumption data for a period of 1-year for 732 households in DelhiNCR (composite climate) and 426 households in Chennai (warm-humid climate). b) Detailed monitoring of four residential flats (two each in Delhi and Chennai) for one year duration. The monitoring included: discrete logging of hygrothermal properties of individual rooms & ambient conditions; electricity consumption of comfort conditioning equipment like fans, desert coolers and airconditioners 1 c) Analysis of the collected data to calculate Energy Performance Index (EPI), monthly energy consumption profiles, share of energy consumption for comfort conditioning etc. The information from the analysis of monthly energy consumption data and monitoring campaign is intended to be used to define inputs and validate outputs of energy simulation models for typical residential flats in the two climatic regions. The energy simulation models will be further used to evaluate the potential of passive and active strategies for reducing energy consumption and improving thermal comfort in the residential buildings. 1 Energy Performance Index (EPI) for the analysis in the paper is defined in terms of annual purchased electricity (in kWh) divided by built-up area (in m2) of the flat. The built-up area includes covered area of the flat and does not include balcony areas, semi-covered areas and common areas like lifts and lobbies etc. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 207 INTRODUCTION India is experiencing an unprecedented urbanization due to the cities transforming into economic hubs. According to 2011 census data, about 31% of the India’s population was residing in the urban centers, and this percentage is expected to increase to 40% by 2030. It is estimated that the total constructed built-up area would increase from 8 billion square meters in 2005 to 41 billion square meters in 2030 (about 5 fold increase) (Mckinsey & Company, 2009). This situation is significantly different from the developed countries, where bulks of the buildings are already constructed. This provides both challenges and opportunities to building sector stakeholders to develop this building stock appropriately. As per CEA report (CEA 2005), residential sector consumes 21% of the total electricity generated in India, which is about 3 times more than that of commercial buildings. One of the reasons for this is that the built-up area of residential buildings is about 7 folds more than that of commercial buildings (Mckinsey & Company, 2009). The energy use intensity of the residential buildings is expected to grow because of increase in air conditioned area, better access to electricity and increase in ownership and usage of appliances. There is an inevitable rise in the density of residential urban development due to scarcity of land and the desire to curtail suburban sprawl. It is now common for city planning authorities to encourage Floor Space Index (FSI) of up to 4; FSI of 1.5 to 2 is becoming commonplace. As per census data, urban residential household will increase by ~2 folds from 2014 to 2032; Greentech Knowledge Solutions Pvt. Ltd. based on EMPORIS (EMPORIS data, 2014) data projected that during this period the share of highrise building will increase by ~ 5 folds. CONTEXT AND METHODOLOGY FOR DEVELOPING THE GUIDELINES The guidelines for residential buildings design will focus on the recommendations for the reduction of operational energy, reduction of embodied energy and improvement of thermal comfort of the residents. The flowchart showing the complete methodology for development of design guidelines for residential buildings is shown in Figure 1. The work presented in this paper concerns only with the monitoring and analysis of monthly energy consumption and is highlighted in grey boxes in Figure 1. Monthly energy consumption data for one year duration was collected from 732 residential units in Delhi - NCR (from 4 residential complexes) and 426 residential units in Chennai* (from 6 residential complexes) to understand the baseline scenario of energy consumption in the composite climate 2 (represented by Delhi-NCR) and warm-humid (represented by Chennai). The composite and warm3 humid climatic regions of India. This constitutes almost two third of the total geographical area of India (see Figure 2). 2 As per National Building Code 2005, “Climatic zone that does not have any season for more than six months may be called as composite zone”. In India, composite climate is characterized by hot summer season and moderate winter season. 3 As per National Building Code 2005, “Warm-humid climate are characterized by two factors: monthly mean maximum temperature (MMMT) and mean monthly relative humidity (MMRH) percentage. e.g. MMMT above 30 ⁰C with MMRH of above 55% and MMMT above 25 ⁰C with MMRH of above 75% 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 208 Figure 1. Flowchart showing approach for development of design guidelines for residential buildings under Indo-Swiss Building Energy Efficiency Project (BEEP) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 209 Figure 2 Climate classification map of India as per NBC 2005 The selected residential complexes were multi-storey apartment buildings of 3 to 15 storeys. The built-up area of residential units ranged from 80 m2 to130 m2. The residential units were having 2 or 3 bedrooms and a drawing/dining room. The residents of these houses primarily represents middle and middle upper income group of India. A mathematical model was used to filter monthly electricity consumption values, which were either too high or too low. After data filtering, 89% of the data from the residential units in the composite climate and 90% of the data from the warm-humid climate was considered for further analysis. The filtered data was statistically analyzed in terms of Energy Performance Index (EPI) distribution, monthly energy consumption profiles, share of electricity for space comfort conditioning. Time series data (for 2 years) from one of the residential complexes in Delhi was used to infer the trend of energy consumption. Subsequent to the collection of monthly electricity data, a monitoring campaign was carried out in the four selected residential units (two each in Delhi and Chennai). The monitoring included discrete logging of space and ambient hygrothermal conditions and energy consumption monitoring of space comfort conditioning equipment (Figure 3). Figure 3 (a) Temperature humidity logger assembly for monitoring ambient conditions 3(b) Globe temperature logger assembly for measuring mean radiant temperature in the room 3(c) Energy loggers for monitoring energy consumption of comfort conditioning equipment Information from the monthly energy consumption data analysis and monitored data will be used to define inputs and validate outputs of the baseline energy simulation models for both the climatic 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 210 zones. Potential of individual and group of strategies will be evaluated by conducting parametric runs on baseline simulation models. Inferences drawn from the simulation analysis and recommendations for reducing the energy requirements for common facilities will be used to formulate climate specific design guidelines for reducing operational energy and improving thermal comfort in the residential buildings. This paper discusses the baseline scenario of energy consumption of urban residential buildings in both composite and warm-humid climatic regions of India. RESULTS AND ANALYSIS A. Results of analysis of monthly electricity consumption data for sample flats in composite and warmhumid climates Figure 4(a) and 4(b) shows EPI distribution graphs for residential units belonging to composite and warm-humid climates. The mean EPIs for residential flats in composite and warm-humid climate are calculated as 48 kWh/m2.year and 43 kWh/m2.year respectively. Figure 4(a) EPI distribution of residential units in the composite climatic region 4(b) EPI distribution of residential units in the warm-humid climatic region Figure 5 shows average monthly energy consumption profile of two residential complexes, one each in composite (78 residential units) and warm-humid climates (243 residential units). The monthly electricity consumption for the residential complex in composite climate shows steep increase in energy consumption during the summer and monsoon months (May-August). This is attributed to the operation 4 of comfort conditioning equipment. November can be considered as base month , when the need for comfort cooling or heating is minimum. Monthly electricity consumption for the residential complex in warm-humid climate (Figure 5) shows a flatter profile during April to November, with peak appearing during the June and July months. This is due to extended warm and humid seasons, when comfort conditioning is required. December and January can be considered as base month when the requirement for comfort cooling or heating is minimum. EPI range distribution for both composite and warm-humid climate (Figure 6(a) & 6(b)) shows that 16% of the residential units in the composite climate and 22% in the warm-humid climate have a high EPI of more than 70 kWh/m2.year. 4 “Base month” is the month during which there is almost no comfort cooling or heating requirements in the climatic region. During this time the energy consumption will be from refrigerator, lighting, washing machine, electric geysers, kitchen appliances, TV, computers, etc. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 211 Figure 5 Average monthly energy consumption profiles of two residential complexes in composite and warm-humid climates Figure 6(a) EPI range distribution for residential units in composite climate Figure 6(b) EPI range distribution for residential units in warm-humid climate Time series energy consumption data for year 2007 and 2009 was collected for one of the residential complexes (78 residential units) in Delhi. Figure 7 shows that there is an increase in the average EPI of the complex by 16% in the year 2009 compared to 2007. This increase in average EPI is primarily attributed to the increase in energy consumption (by 20%) during the summer and monsoon periods (April to September). This reflects higher ownership and use of air-conditioning equipments with the increase in disposable income and availability of easy financing options and an aspiration for higher comfort. Figure 7 Time series plot of a residential complex in Delhi 7(a) Mean EPI 7(b) Cumulative energy consumption per flat for summer and monsoon months (April- September) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 212 Figure 8(a) & 9(b) show that there is an increase in average EPI of residential units with the increase in ownership of number of air-conditioners in both composite and warm-humid climates respectively. There is an overall trend of increase in number of residential units with higher EPI. Figure 8(a) Distribution of average EPI with respect to air conditioners ownership for residential units in composite climate 8(b) Distribution of average EPI with respect to air conditioners ownership for residential units in warm-humid climate B. Results of detailed energy monitoring in three flats of the composite climate Analysis of energy consumption in 3 flats in the composite climate: a) Flat A having a below average EPI (in the range of 30 to 40) b) Flat B having an above average EPI (in the range of 60 to70) , and c) Flat C having high EPI (> 70EPI) is presented below: Figure 9 shows monthly energy consumption and monitored energy consumption for comfort conditioning equipment for Flat A in Delhi. This flat used convective ceiling fans from mid-March to mid-October, 2 evaporative desert coolers were used from April to June and 2 air conditioners were used from June to July. The EPI of this flat was 35 kWh/m2.year. Monitored data shows that almost 33% of the annual energy consumption is attributed to operation of comfort space conditioning equipment. Balance, 67% of the electricity (referred in this paper as base energy consumption i.e. energy used for purposes other than comfort cooling) is used for refrigerator, lighting, washing machine, electric geysers, kitchen appliances, TV, computers, etc. Figure 9 Monthly energy consumption profiles for below average EPI in composite climate Figure 10 shows monthly energy consumption of two more residential units in the same residential complex. Residence-B uses 2 air-conditioners predominantly for comfort cooling and have an EPI of 65 kWh/m2.year (~1.8 times compared to Residence-A). Residence-C with 4 air conditioners have an EPI of 117 kWh/m2.year (~3.3 times compared to Residence-A). If energy consumption during February is 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 213 5 considered as base energy consumption for Residence B and Residence C, then the contribution of energy consumption for comfort space conditioning can be as high as 38% for Residence B and 65% for Residence-C. Figure 10 Monthly energy consumption profiles of three residential units in composite climate CONCLUSIONS a) The mean EPIs for sample residential flats of 2-3 bedrooms in composite (732 flats) and warmhumid climate (426 flats) for the year 2009 are calculated as 48 kWh/m2.year and 43 kWh/m2.year respectively b) Energy consumption for comfort cooling is a significant part of the electricity consumption. Detailed analysis of energy consumption in three sample flats shows that the contribution of energy consumption for comfort space conditioning, increases with the increase in EPI (and increased usage of air-conditioners) and for the three flats was estimated to vary between 33% to 65% of the total energy consumption. c) Analysis of time-series data for one residential complex for 2007 and 2009 shows 16% increase in average EPI, which indicates towards the trend of increase in energy consumption in the urban residential buildings d) Detailed energy consumption monitoring of a flat, which utilizes a combination of fans, evaporative coolers and ACs for cooling, shows potential of large energy savings by appropriate and energy-efficient use of comfort cooling appliances. With bulk of the construction in building sector bound to happen in housing sector in the next two decades, there is an urgent necessity for guidelines for designers to effectively integrate the potential strategies for reducing energy consumption and for augmenting thermal comfort as well as guidelines for residents to use energy efficiently for space cooling. ACKNOWLEDGEMENT The authors would like to acknowledge the support of the Swiss Agency for Development and Cooperation and Bureau of Energy Efficiency, the two implementing agencies of BEEP, for support and guidance in the work on the development of energy efficient residential building design. The authors would also like to acknowledge the contribution of Tara Nirman Kendra of Development Alternatives and Conserve Consultants Pvt. Ltd. for assisting BEEP in the collection of monthly energy consumption data and monitoring of residential units in Delhi-NCR and Chennai respectively. REFERENCES • • • McKinsey & Company. (2009). Environmental and energy sustainability: An approach for India Central Electricity Authority (CEA). (2008). All India Electricity Statistics 2008: General Review 2008 EMPORIS data. (2014). Retrieved June, 2014, from http://www.emporis.com/country/india 5 Base energy consumption is the energy consumption excluding the comfort space conditioning. This also excludes energy consumption for space heating. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 214 Session 2C : User behavior, thermal comfort & energy performance PLEA2014: Day 1, Tuesday, December 16 14:10 - 15:50, Grace - Knowledge Consortium of Gujarat Cool spots in hot climates: a means to achieve pedestrian comfort in hot climates Priji Balakrishnan, MArch Architectural Association School of Architecture, London prijibalakrishnan@gmail.com ABSTRACT With the advent of automobiles, an increasing number of emerging cities are being planned for motorists. Ironically, motorists are least affected by harsh climate or distance. Planning is essentially required for pedestrians to traverse in the shortest possible routes with a pleasurable thermal comfort experience. Defining pedestrian comfort is complex, as it depends directly on dynamic climatic factors and also on the physiological and psychological factors. The current standards for outdoor comfort are based on static models which do not take into account the variability of climatic factors in the urban scenario and its impact on physiological factors with time. This paper focuses on understanding thermal comfort of pedestrians through literature and fieldwork to draw comfort limits and find the most influential factors that affect pedestrian comfort in hot climates. The research is carried out in the city of Sharjah, in UAE which experiences hot desert climate. The three factors that were identified to be the most influential for pedestrian comfort in Sharjah were – providing shade, enhancing wind movement and reducing mean radiant temperature. This paper also explores a design solution – ‘cool spots’ incorporating these factors in a way that best suits the urban context of Sharjah. INTRODUCTION UAE is a country that overturned the sands of the desert to become one of the most booming economies of West Asia. This resulted in rapid urbanization, and despite harsh climate, the urban centers developed without much consideration to these factors. Wide glass-faced urban canyons resulted in an urban fabric hostile to pedestrians. The problem faced by the pedestrians in UAE is that there is too much sun and too little shade. Designing for pedestrian comfort in hot climates like UAE requires one to understand the climate, comfort limits and thermal adaptation of a pedestrian in the specific context. There are a number of thermal comfort assessment methods that have been developed. Some of the commonly used ones for outdoors are steady-state models like Predicted Mean Vote Index (PMV) or Predicted Percentage Dissatisfied Index (PPD) (Fanger,1982), Index of Thermal Stress (ITS) (Givoni, 1976) and Physiological Equivalent Temperature (PET) (Mayer & Hoppe,1987). The problem with these steady-state models is that they cannot account for the dynamic thermal adaptation of humans (Chen & Ng 2011). Hence, in this paper, comfort limits are identified through literature studies and a series of experimental fieldwork in Sharjah, which helps confirm these limits. The cool spot is designed using these limits. For the purpose of evaluation of the conditions in the cool spot, PET is used, which gives an understanding of the difference in the felt temperature inside and outside the cool spot. Priji Balakrishnan is a PhD student in the Architecture and Sustainable Design pillar, Singapore University of Technology and Design, Singapore 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 215 FACTORS AFFECTING PEDESTRIAN COMFORT The three factors that influence pedestrian comfort are environmental, physiological and psychological. The environmental factors depend on parameters of air temperature; air movement; radiation and humidity. Physiological factors depend on heat balance of the body, which is largely influenced by metabolic rates, physical activity and the type of clothing. Psychological factors depend on nature of space and usage, seasonal expectation and cultural and regional expectations. This paper deals with pedestrian comfort in an urban context during hot periods. Hence, it addresses factors and limits in the light of reducing heat gains or enhancing cooling. CLIMATE OF SHARJAH Figure 1 shows a graphical representation of the monthly average temperatures in Sharjah (25.330N and 55.430E). The weather data is obtained from the Meteonorm global meteorological database (version 6.1) which represents a ten year average of the data files. Based on the monthly variations of Dry Bulb Temperature (DBT) of Sharjah, the annual cycle can be divided into three distinct periods – a four month period of mild weather (December to March inclusive), a warm period (November and April) and a hot period (May to October inclusive ) (Yannas 2008). Figure 1 Graphical representation of the monthly average temperatures. COMFORT LIMITS In outdoor spaces, the aim is to provide tolerable thermal conditions to prolong the exposure time of the pedestrians (Tahbaz, 2010) rather than trying to achieve ideal thermal comfort. Hence it is significant to study the varying comfort limits for pedestrians in a particular climate. Comfort Limits based on Environmental Factors Figure 2 (a) shows the bioclimatic chart plotted for a clo value of 0.4 and 1.3 Met which indicates summer clothing and sedentary activity (like slow walking) respectively. The cluster of dots in the chart represents the DBT and Relative Humidity (RH) (source Meteonorm 6.1) values for all the days throughout the year. The chart is based on the limits described by Arens, Gonzalez and Berglund (1986) and assumes that air temperature is equal to the mean radiant temperature. The measure of comfort used to determine boundaries of the chart is skin wettedness (fraction of skin covered by sweat). 0 Comfort Limits of Air Temperature: Up to DBT 25 C in Figure 2(a) the pedestrians are 0 0 comfortable without an external shade. From DBT 25 C to 32 C pedestrians are comfortable underneath a shade and still conditions of air. The boundaries of comfort limit can be extended up to 390C DBTunderneath shade and 2m/s wind speed. This limit conforms to the findings of the fieldwork survey conducted by Thappar and Yannas (2008) in Dubai (25.140N, 55.170E) (a neighbor city that shares similar climate with Sharjah) where ambient temperatures close to 400C in shade, with wind velocities of 2m/s were perceived as acceptable conditions. Comfort Limits of Air Velocity: The comfort boundary can be stretched to 420C with an air velocity of 4 m/s and further to 440C up to a mean maximum of 6m/s (Arens et al. 1986), provided the turbulence is low. Above 6m/s, the mechanical effects of wind counterbalance any of its positive effects on comfort. Air velocity at high temperatures has opposite influences on comfort. It increases the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 216 evaporative capacity of the air, and hence the cooling impact on the skin, but also causes higher convective exchange that warms the body Givoni (1976). Hence, there is an optimum velocity of air movement that produces the highest cooling. This optimum velocity depends on temperature, humidity, metabolic activity and clothing. Based on Givoni’s (1976, p 66 - 67) method, at a temperature of 400C and vapour pressure 30mmHg, for a pedestrian whose metabolic activity is 2.4 – 3 Met (walking at 4km/hr), the optimum velocity averages between 2.9 to 4.0m/s. For metabolic activities at 1.2 – 1.4 Met under the same conditions, the optimum velocities average between 1.2 -2.0m/s Figure 2 (a) Bioclimatic chart based on environmental factors. (b) Bioclimatic chart based on physiological factors. Comfort Limits of Humidity: Vapour pressure below 5mmHg is likely to cause respiratory discomfort (Erell, Pearlmutter&Williamson, 2011). Vapour pressure of air above 37mmHg corresponds to the vapour pressure of skin and hence indicates the highest level of discomfort. Vapour pressure of 22mmHg corresponds to a wet bulb temperature of 240C (at ambient temperature of 340C), which Givoni (1998) suggests as the upper limit for the application of direct evaporative cooling of DBT. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 217 Comfort Limits: Bioclimatic Chart based on Physiological Factors Figure 2(b) defines the boundaries of extended comfort achieved by the human body through physiological responses to cooling. Up to about 250C, the heat balance of the body is maintained by the minimum thermoregulatory processes and heat exchanges required. This takes place mainly through convection and radiation. At 280C, the body starts perspiring and above 280C, true sweating occurs wherein water is exuded to the skin (Oke 1987). At about 320C – 340C, convective losses turn into convective gains as the gradient between skin and ambient temperature becomes nil. At an ambient temperature of 350C and above, only evaporative losses from skin help in heat dissipation. The chart clearly indicates that during a large span of the hot period, evaporative loss is the only means of heat regulation for the body. The rate of evaporative losses depends on body to air, vapour pressure gradient and air velocity (Givoni 1976), hence making both these limits influential for pedestrian comfort. FIELDWORK The fieldwork was carried out to validate the theoretical understanding of the comfort limits described above and to identify the most influential factors of comfort in this context. Outdoor comfort studies are usually conducted by questionnaire surveys, where subjects are interviewed. The limitation of this method as described by Ng E et al (2012) is that the thermal sensation of the subject is captured under relatively static climatic conditions. This fieldwork study introduces an experimental methodology - measuring the skin temperature of a pedestrian in a series of three experiments while in motion, hence giving an understanding of thermal stress experienced over time. The fieldwork was conducted in Sharjah during the two hottest months – August to September, during the day to measure thermal stress in the worst case scenario. The thermal stress was weighed based on skin temperature relative to air temperature, as it is determined by the local equilibrium conditions of heat flow from the body core to the skin and the heat loss from the skin to the environment (Givoni 1976). Observation of skin wettedness and sweating were also noted at regular intervals. Instruments used for fieldwork were an anemometer for measuring wind speed, two hand held data loggers – one for measuring globe temperature and one for logging DBT and RH, and a temperature logger with a thermistor surface temperature probe to measure skin temperature. Skin temperature measurements are usually taken at 16 points and their mean weighted average is the resultant, but in cases where the DBT is at 320C and above, the variation of temperature over the whole skin is less than 20C (Givoni 1976). For this reason, the skin temperature was approximated using temperature measurements on just one point on the body, that is, on the wrist of the subject. Measurements of skin temperature, ambient air temperature and relative humidity were recorded simultaneously and spot measurements of wind speeds were taken. Notes based on observation and questioning were recorded to mark levels of discomfort through sweating. The recorded skin temperature corresponds to conditions stated by Givoni (1976) – skin temperature in comfortable conditions is 330C; at moderate heat conditions, 350C and during severe heat, it reaches 370C. The test subject was a male, 58 years old with dark complexion and is a representative sample of the labour workforce in UAE, who are most likely to use outdoor spaces during the hottest periods of day. Experiment 01 – Alternate walks in sun and shade The walk was conducted during day time and the average temperatures recorded were 420C. Given the temperature and clear skies, the walk is not a leisure walk, rather a deterministic walk i.e. a walk for a necessary activity. In case of leisure walk, the discomfort expressed could be higher and at more frequent intervals than for a deterministic walk. The walk spanned about an hour with a routine - 10 minutes walk in the sun (without any overhead gear) and 5 minutes rest in an air-conditioned/shaded zone. The candidate was given the choice of reducing or increasing either the exposure or the rest time based on the comfort levels. Figure 3 shows the graph with the plotted skin temperature, ambient air temperature and RH measured during the walk. The shaded regions in the graph indicate the time span resting in shaded/air 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 218 conditioned zone and the lighter regions indicate the time span walking while exposed to direct sunlight. The circled numbers mark the descriptive note of the conditions and the observations made. Few of the significant ones are shown in Table 5. Table 5. Descriptive Note of Conditions and Observations Made During Experiment 01 1 Description The walk started from an air-conditioned indoors maintained at comfortable temperatures between 260C – 270C. The skin temperature measured was 33.3 0C which is consistent with Givoni’s (1976) conditions of skin temperature. 3 Walking at medium pace (4km/hr) in the sun, air temperature recorded was 400C 420C. The skin temperature increased from 350C to 38.60C in 9 minutes. Within 5 minutes, the skin was clammy and at the end of 9 minutes, the forehead was observed to be wet and the candidate expressed the need to be in a shaded/airconditioned zone. 4 Stepping into an air-conditioned mall with an indoor temperature between 250C 260C, the skin temperature dropped to 34.50C in 5 minutes. The forehead and body were dry within this time. 5 Resuming the experiment of walking in the sun, recorded air temperature is at 40 – 410C. The skin temperature increased to 380C in 4 minutes. The rate of sweating increased and the candidate expressed his forehead and back to be wet. 6 Walking in an enclosed, non air-conditioned market with ceiling fans (at 4m high); the air temperature and wind speed recorded was 350C, and 0.0 - 0.5m/s respectively. Though the skin temperature dropped to 360C in 3 minutes, the candidate expressed discomfort and the need to be outside the market at the end of 3 minutes. The ambient wind speed before walking into the market was measured at 1.1 -1.5 m/s which dropped inside the market. This could be the possible explanation for the candidate’s discomfort. 7 Continuing the walk in the sun, at an air temperature of 41 – 420C., the skin temperature increased from 360C to 390C in 2 minutes and to 39.80C in the next 2 minutes. The candidate expressed high level of discomfort and it was observed that the clothing was completely wet with sweat. 8 Resuming a slow walk in the air-conditioned mall, air temperature was recorded to be 25-260C and the skin temperature dropped to 340C in 3 minutes. 9 10 Walking with the back facing the sun was claimed to be more comfortable than the other way round; recorded air temperature was 41-430C. The skin temperature reached 390C in 4 minutes with profuse sweating. While seated and taking rest in the air-conditioned mall, recorded temperature was 250C.The skin temperature reached 330C in 9mins. Observations and Conclusions: Starting from comfortable indoor conditions and walking at a pace of 4km/hr, a 10 minute exposure to direct sunlight can be considered the maximum acceptable before an air-conditioned resting zone is required to bring down the skin temperature to comfortable conditions. This distance, when calculated based on the above study is 667m. If these resting zones were to be replaced by shaded zones (or cool spots), the minimum distance of exposure to avoid thermal stress can be reduced to half, at approximately 300m for urban design considerations. The time taken for the rise in the candidate’s skin temperature is inversely co-related to his exposure in the sun. Hence, if a particular path has to designed with air-conditioned nodes, the distance between the nodes have to decrease incrementally to ensure that the skin temperature remains in the comfortable range. This experiment also shows that in hot climates like Sharjah, skin wettedness influences thermal 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 219 comfort in a shaded zone more than a decrease in 50C-60C in DBT. Hence, an increase in wind speed would be more beneficial in design application rather than a decrease in a few degrees of temperature. The rise in skin temperature reduces considerably after it reaches 390C. Figure 3 Graph showing measured skin temperature, ambient DBT and RH during Experiment 01. Experiment 02 – Walk in Shade In experiment 02, the candidate was asked to walk for 10 minutes in a shaded zone without any exposure to the sun. Figure 4 shows the graph plotted with the measured skin temperature, ambient air temperature and relative humidity. The shaded region in the graph indicates the measurements taken when the candidate walked under the building shade for 10 minutes. The lighter region shows the measurements taken indoors. Figure 4 Graph showing measured skin temperature, ambient DBT and RH during Experiment 02. Observations and Conclusions: As seen from the graph, though the ambient temperature increased from indoors to outdoors from 260C to about 36.40C, the skin temperature increased only slightly, from 33.20C to 33.60C. The measured wind speed was 0.8 -1.3 m/s and hence, the skin remained relatively dry. Experiment 03 – Standing in sun In continuation with experiment 01, the candidate was asked to stand in the sun for 10 minutes without any head gear. The wind speed was measured to be 2–2.5 m/s. This experiment was conducted 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 220 to understand the effect of wind as a standalone factor to thermal stress. Figure 5 shows the graph with measurements plotted during the experiment. Observations and Conclusions: At ambient temperature of 400C - 430C, the skin temperature remained at an average 37.20C and reached a maximum of 37.60C unlike in experiment 01, where it rose to 390C. This could be due the effect of wind speed at 2-2.5 m/s that continuously enhanced the evaporation of the sweat, hence cooling the skin. Figure 5 Graph showing measured skin temperature, ambient DBT and RH during Experiment 03. DESIGN OF COOL SPOTS The three factors that were identified through research and fieldwork to be the most influential in Sharjah for improving pedestrian comfort were providing shade – cut off all the solar radiation, enhancing wind movement – providing a minimum air movement of 2m/s and a maximum of 4m/s at pedestrian level and reducing mean radiant temperatures – maintaining immediate surrounding surfaces of a pedestrian close to air temperature. The design and concept of the cool spot is a direct synthesis of the three factors mentioned. Among the five categories of fieldwork (1.Measuring surface temperature of common surfaces in the urban fabric, 2.Measuring conditions underneath different kind of shades, 3.Datalogging conditions underneath the Masdar cooling tower in Masdar city, Abu Dhabi, 4.Measuring thermal stress of a pedestrian and 5.Observing pedestrian activity in Sharjah) that contributed to the design of the cool spot, only one category – measuring thermal stress of a pedestrian has been explained in this paper. The geometry of the cool spot is designed to allow for the wind to pass through it from any direction. Fans are integrated in the centre in case of no wind condition. The cool spot is designed so as to ensure the central space of diameter 3.2m remains shaded throughout. Materials used for cool spot do not heat up above the air temperature as shown in Figure 6(a). Figure 6 (a) Elevation of Cool Spot showing surface temperature of the various materials at 400C ambient DBT. (b) Map showing placement of cool spots. (c) Graph showing PET values. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 221 The Physiologically Equivalent Temperature (PET) was calculated using RayMan version 1.2 inside and outside the cool spot for a typical day in August, shown in Figure 6(c). The results show that during the hottest time of the day, the felt temperature (PET) in the cool spot is 100C lesser than the outside temperature. Figure 6(b) shows a map of a pedestrian route in Sharjah connecting a residential/commercial area to an intercity bus stand. This stretch of 1.6 km, with the introduction of cool spots at a minimum distance of 60m (to prevent the interference of the wind flow pattern between two cool spots) and maximum distance of 300m ensures a pedestrian experience with reduced thermal stress. CONCLUSION Designing for pedestrian comfort requires understanding the climate and the comfort limits. Initial stages of this research (not mentioned in this paper) looked at design strategies to reduce air temperature. This was overruled as fieldwork in Sharjah revealed that air movement had more influence on comfort than a few degrees decrease in temperature. Through literature review and fieldwork it was understood that, during the hot period in Sharjah temperatures of 400C – 420C were considered acceptable, provided the pedestrian is in complete shade and there is a minimum wind speed of 2m/s and maximum 4m/s. Although the fieldwork conducted in this study used only one subject, it can be extended to larger group to arrive at a more comprehensive results. Such a study could be conducted using a similar methodology as outlined in this work. This paper also proposes a design solution - Cool Spots - to achieve pedestrian comfort in Sharjah. The design of the cool spots is based on the three factors – provide shade, enhance wind movement and reduce mean radiant temperatures close to air temperature. The concept of a cool spot as an urban furniture is specific to Sharjah/UAE as it is a response to both the cultural as well as the climatic expectations of the pedestrians there. It was observed that people felt comfortable underneath a tree canopy, especially under the dense and wide one of a banyan tree. Given the desert climate in Sharjah, trees like the banyan are not a common sight. Cool spots are intended to serve the function of such a canopy, but unlike trees that reduce the wind speed underneath them, cool spots enhance it. The felt temperature (PET) within the cool spot was analysed (using RayMan 1.2) to be 100C lesser than the outside temperature at peak conditions (noon) during one of the hottest month (August) in Sharjah. REFERENCES Arens, E,R. Gonzalez, & L. Berglund. (1986). Thermal comfort under an extended range of environmental comfort. ASHRAE Transactions. Vol 92 . Part 1B. Chen, L & Ng, E. (2011). Outdoor thermal comfort and outdoor activities: A review of research in the past decade. DOI: 10.1016/j.cities.2011.08.006 Erell,E., D. Pearlmutter, & T. Williamson. (2011). Urban Microclimate: Designing Spaces Between Buildings. Earthscan. Fanger, P.O (1982). Thermal comfort. Analysis and application in environment engineering. McGraw Hill Book Company. New York. Givoni, B. (1998). Climate Considerations in Building and Urban Design. Van Nostrand Reinhold. Givoni, B. (1976). Man, climate and architecture. Applied Science Publishers. London. Mayer, H., & Höppe, P. (1987). Thermal comfort of man in different urban environments. Theoretical and Applied Climatology. Oke, T.R. (1987). Boundary Layer Climates. Meuthen & Co.London Penwarden, AD & AFE Wise. (1975). Wind Environment Around Buildings. BRE, Building Research Establishment. Tahbaz, M. (2010). Toward a New Chart for Outdoor Thermal Analysis. Network for Comfort and Energy Use in Buildings (nceub). Proceedings of Conference: Adapting to Change: New Thinking on Comfort. London. Thapar,H & S.Yannas. (2008). Microclimate and Urban Form in Dubai.Proc PLEA. Dublin Yannas,S. (2008). Challenging the Supremacy of Airconditioning. 2A Architecture & Art, Issue 7, pp472 – 477. Ng, E. & Cheng,V. (2012). Urban human thermal comfort in hot and humid Hong Kong. DOI:10.1016/j.enbuild.2011.09.025 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 222 The Effect of Natural Ventilation and Daylighting on Occupants’ Health in Malaysian Urban Housing Yaik-Wah Lim, PhD Yong-Long Lim, PhD Department of Architecture Faculty of Built Environment Universiti Teknologi Malaysia limyaikwah@gmail.com Department of Architecture Faculty of Built Environment Universiti Teknologi Malaysia ABSTRACT Terraced houses have been rapidly constructed in Malaysia since 1960’s and account for 44% of the existing urban housings. The spatial characteristics of the houses have been remained the same for decades although these houses have very constrained use of natural ventilation and daylighting due to deep planning. This kind of design causes indoor thermal and visual discomforts due to gloomy indoor spaces, low air change rate and poor indoor air quality. As studies proved that indoor environmental stressor can produce negative stress on occupants’ health, the effect of natural ventilation and daylighting on occupants’ comfort and health in the terraced houses was investigated. Case study of 80 terraced houses in Johor Bahru, Malaysia was conducted to identify the critical comfort and health issues due to natural ventilation and daylighting. The relationships between occupants’ comfort, behavior and health were studied through questionnaire survey. The findings demonstrated significant linear relationships between indoor comfort and health. However, occupants’ behavior did not give significant impact on comfort and health. Besides, the effects of natural ventilation and daylighting performances on specific health issues were also studied. The findings concluded that the by-law requirement of 5% window-to-floor ratio for natural ventilation is inadequate for occupants’ comfort and health, thus further review is needed. Proper consideration of natural ventilation and daylighting design strategies in terraced house is essential as it determines how the occupants can manage the indoor environment to achieve comfortable and healthy living environment. INTRODUCTION Since 50 years ago terraced houses have been rapidly constructed in Malaysia due to the increasing demands for urban housing. This housing typology accounts for more than 40% of the existing housing stocks in the urban areas (Malaysia Department of Statistic, 2000). The origin of this housing typology is adopted from the British terraced house design which is also known as “row house”. This type of house has relatively narrow and deep plan with limited fenestration at the front and rear facades. The typical width of Malaysian terraced house ranges from 18 ft. (5400 mm) to 25 ft. (7500 mm); while the length is 65 ft. (19500 mm) to 80 ft. (24000 mm). The housing layout is usually planned repetitively and monotonously in rows of rectangular lots. The boundaries of the houses are defined by perimeter chain- 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 223 linked fences or bricks walls (Hashim and Rahim, 2008; Omar et al., 2010). The typical spatial characteristics of the terraced houses in Malaysia have been remained the same for decades (Omar et al., 2010). Most terraced houses have combined living and dining hall, minimum three bedrooms, two or three bathrooms, a kitchen at the back and a car porch at the front. For doublestorey terraced house, the space under the staircase is commonly used as storage area. On the first floor, master bedroom with an attached bathroom is usually located at the front; while two other bedrooms which share a bathroom sitauted at the back. Besides, some of the terraced houses have a small family hall placed in the center to connect all the bedrooms with the staircase. Examples of typical Malaysian terraced house design and layout are shown in Figure 1. Figure 1 Example of typical terraced houses in Malaysia: (a) Frontage, (b) Layout plans The roofs are the major building envelopes of terraced houses that are exposed directly to solar radiation. Hence, proper insulations are needed to reduce the heat conduction from the roof into the indoor spaces. Despite global illuminance as high as 130 klx, the openings on the front and back façades and the roofs are the very limited sources for daylighting. The openings also allow unwanted solar radiation heat gain. Besides, natural ventilation in terraced houses is constrained by the small windowto-floor ratio (WFR). Moreover, tropical climate has high air temperatures, high relative humidity and very low wind speeds (Agung and Mohd Hamdan, 2006; Lim et al., 2012; Lim, 2013). In Malaysia, some research works had been conducted in the terraced houses in tropical climate. Sadafi et al. (2011) investigated the thermal effects of internal courtyard in a tropical terraced house in Malaysia. The findings showed that internal courtyard allows better natural ventilation but increases the radiation heat gain. Hence, efficient openings and shading devices are needed in order to improve the thermal conditions of the courtyard’s surrounding spaces. Kubota et al. (2009) examined the effects of night ventilation technique on indoor thermal environment for terrace houses in Malaysia. The findings concluded that the indoor humidity control during the daytime such as by dehumidification would be needed when the night ventilation technique is applied to Malaysian terraced houses. Otherwise, full-day ventilation would be a better option compared with night ventilation. Zakaria (2007) studied sustainable housing for residential-industrial neighborhoods in Malaysia by looking into several indoor environmental quality (IEQ) aspects. Questionnaire surveys, physical measurements and interviews were conducted for housing area in Pasir Gudang, Johor, Malaysia. Nevertheless, the focus of the study was on the IEQ especially air quality due to pollution from industries. The researcher did not emphasize the impact of other IEQ aspects on occupants’ health. Most of the previous research on terraced houses in Malaysia was limited to indoor environmental performance. However, studies have shown that indoor environments, including work and living spaces, have major impact on occupants’ well-being (Bluyssen et al. 2011; Choi et al., 2012; Todorovic and Kim, 2012). Environmental stressor such as discomfort air temperature, poor air quality and inadequate 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 224 lighting can produce negative stress. Thereby, this paper studies the impacts of natural ventilation and daylighting on occupants’ comfort and health in existing Malaysian typical terraced houses. METHOD Case study of indoor environment, occupants’ perceived comfort and health in Malaysian terraced houses was conducted. Questionnaire method was employed to conduct survey in 10 different terraced housing estates in Johor Bahru, Malaysia during the months of July to September in year 2013. All the selected terraced housing estates were located within the neighbourhood of Skudai, Johor Bahru, Malaysia. Random purposeful sampling method was employed in order to cover various types of terraced houses including intermediate, corner and end lot units. The total number of study cases was 80 houses. The summaries of respondents according to house types are as shown in Table 1. Type Table 1. Summary of study cases according to housing type No. of Storey No. of House(s) Percentage (%) Intermediate Unit Corner Unit End Unit 1 2 3 1 2 1 2 Total 29 35 1 8 3 3 1 80 36.3 43.7 1.3 10.0 3.7 3.7 1.3 100.0 81.3 13.7 5.0 100.0 The design of the questionnaire was divided into 4 major sections. Section 1 was intended to evaluate the effect of natural ventilation on indoor temperature and air quality. Section 2 was to obtain feedbacks from occupants regarding to the use of daylighting for task performance and visual comfort. Section 3 investigated the behavior of the occupants to use and control the natural ventilation and daylighting. Section 4 aimed to evaluate the occupants’ psychological and physical health. All the questions were using 1 (lowest) to 5 (highest) scales. The data collected were analyzed using statistic methods. First of all, the means of each question in Section 1, 2 and 4 were computed in order to identify the most critical indoor comfort and health issues. Then, the correlations between Natural Ventilation (Section 1) and Health (Section 4), Daylighting (Section 2) and Health (Section 4), Natural Ventilation (Section 1) and Behavior (Section 3), Daylighting (Section 2) and Behavior (Section 3) as well as Behavior (Section 3) and Health (Section 4) were analyzed. Spearman’s rho correlation was employed for this analysis since the data was qualitative in nature thus fall under the category of non-parametric test. This correlation tests were important to understand the significant relationships among the variables. Finally, more detailed analysis was conducted by investigating the correlation for specific questions from Section 1 and 2 against specific questions from Section 3 and 4. These questions were selected for analysis due to the relevancy such as indoor temperature comfort against behavior of controlling airconditioning and natural ventilation. Spearman’s rho correlation was employed for this analysis to look into the factors of indoor comfort which influenced the occupants’ behavior and health. RESULT AND ANALYSIS Figure 2 shows the 5 most critical comfort issues and the 5 most critical health issues accoding to the respondents’ opinions. The lowest score for comfort issues was discomfort indoor air temperature during noon time with mean 2.69. Meanwhile, the second lowest score was air movement in other bedrooms with mean 2.78. Indoor temperature and air movement were the major indoor comfort issues. Only 1 of the issues was related to daylighting due to insufficient brightness in toilet. On the other hand, the mean scores for health issues ranged higher than the mean scores for indoor environmental issues. Among the 5 lowest scores, 2 of them were related to psychological health that 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 225 influenced their minds and emotions. Among all the physical health issues, lethargy or tiredness obtained the lowest score. This issue can be related to poor ventilation and air quality as well as insufficient daylighting. Figure 2 Critical comfort and health issues with the lowest mean scores Spearman’s rho correlation tests were used to identify the strength of the linear relationship between 2 variables. Table 2 shows the correlation among Sections 1 to 4. The analysis showed positive linear relationship between Natural Ventilation (Section 1) and Occupants’ Health (Section 4), Daylighting (Section 2) and Occupants’ Health (Section 4), as well as Daylighting (Section 2) and Occupants’ Behaviour (Sectoion 3). The value of 'sig. (2-tailed)' (0.000) was less than the predetermined alpha value (0.01/2 = 0.025), thus the stated null hypothesis was rejected. There existed adequate evidence to show that there was significant positives linear relationship between these variables. This conclusion was made at the significance level of 0.01. Table 2. Correlations among Section 1 to 4 SECTION 3 (Occupants’ Behaviour) Spearman's rho SECTION 4 (Occupants’ Health) SECTION 1 (Natural Ventilation) Correlation Coefficient 0.198 0.549** Sig. (2-tailed) 0.079 0.000 SECTION 2 (Daylighting) Correlation Coefficient 0.474** 0.506** Sig. (2-tailed) 0.000 0.000 SECTION 3 (Occupants’ Behavior) Correlation Coefficient - 0.199 Sig. (2-tailed) - 0.077 **. Correlation is significant at the 0.01 level (2-tailed). There was a positive linear relationship between Natural Ventilation (Section 1) and Occupants’ Behavior (Section 3). The value of 'sig. (2-tailed)' (0.079) was more than the predetermined alpha value (0.05/2 = 0.025), thus the stated null hypothesis was accepted. There existed not adequate evidence to show that there was significant positives linear relationship between these 2 variables. For the correlation between Occupants’ Behavior (Section 3) and Occupants’ Health (Section 4), there was a positive linear relationship between these 2 variables. The value of 'sig. (2-tailed)' (0.077) was more than the predetermined alpha value (0.05/2 = 0.025), thus the stated null hypothesis was accepted. There 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 226 existed not adequate evidence to show that there was significant positives linear relationship. g) Lethargy / Tiredness f) Headache d) Chest tightness 0.004 ** e) Dry / irritated skin c) Dry / irritated throat 0.063 Q3. Able to concentrate well b) Blocked / runny nose 0.013 * 0.157 0.001 ** 0.000 ** 0.123 0.078 0.000 ** 0.022 * 0.215 0.000 ** 0.006 ** 0.000 ** 0.008 ** 0.879 0.009 ** 0.010 ** 0.027 * 0.000 ** 0.016 * 0.030 * 0.049 * 0.121 0.549 0.583 0.146 0.378 0.104 0.014 ** 0.000 ** 0.001 ** 0.029 * 0.355 0.247 Q4. Bad smell 0.051 Q3. Stuffy air 0.077 f. Toilet 0.000 ** 0.446 e. Other Bedroom 0.484 d. Master Bedroom 0.995 c. Kitchen Q5. Physical Health / Symptoms 0.001 ** Q2. Able to sleep well Q1. Feel healthy Q3. Open windows for ventilation during night time Q2. Open windows for ventilation during day time 0.010 ** 0.589 b. Dining Hall Q2. Air Moveme nt in: Psychological Health 0.059 a. Living Hall 0.690 0.260 d. Night 0.000 ** c. Afternoon 0.559 b. Noon 0.912 Section 1 – Natural Ventilation Q1. Indoor Tempera -ture during: 0.879 a. Morning 0.473 Q1. Switch on A/C Table 3. Correlation Sig. (2-tailed) for selected questions in Section 1 (Natural Ventilation) Section 3 – Occupants’ Section 4 – Occupants’ Health Behaviour **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 227 g. Lethargy / Tiredness 0.000 ** f. Headache a. Dry eye / Watering eyes (Epiphora) 0.012 * 0.067 Q4. Do not feel depression / anxiety 0.138 Q3. Concentrate well Q2. Sleep well Q1. Feel healthy 0.000 ** 0.001 ** 0.077 0.000 ** 0.000 ** 0.002 ** 0.442 0.023 * 0.001 ** 0.001 ** 0.006 ** 0.052 0.667 0.000 ** 0.000 ** 0.022 * 0.000 ** e. Cooking 0.001 ** 0.000* * 0.000 ** 0.000 ** d. Leisure 0.002 ** c. Computer work 0.385 Q2. Glare / contrast when windows are unshaded Q3. Using daylight to do: Q5. Physical Health / Symptoms 0.049 * f. Toilet 0.299 0.559 0.021 * e. Other Bedroom b. Writing Q6. Enjoy view through windows 0.224 d. Master Bedroom a. Reading Psychological Health 0.000 ** c. Kitchen 0.000* * Section 2 - Daylighting Q1. Sufficient daylight brightness in: 0.000 ** a. Living Hall Q5. Open window curtain / blinds for daylight Q4. Switch on electric lighting during day time Table 4. Correlation Sig. (2-tailed) for selected questions in Section 2 (Daylighting) Section 3 – Occupants’ Section 4 – Occupants’ Health Behaviour **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Table 3 presents Spearman’s rho correlation analysis results for the selected questions regarding to natural ventilation, occupants’ behavior and occupants’ health. The results showed that indoor temperature did not give significant impact on occupants’ behavior except “indoor temperature in the morning” against “opening windows for ventilation during daytime”. However, indoor temperature yielded significant linear relationship with occupants’ psychological health which includes “feeling healthy”, “able to sleep well” and “able to concentrate well”. The value of 'sig. (2-tailed)' was less than the predetermined alpha value (0.01/2 or 0.05/2) except “indoor temperature during noon time” against 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 228 “able to concentrate well”. From the correlation analysis, the results demonstrated that air movement gave substantial impact on occupants’ physical health. For instance, “air movement in living hall” yielded significant positive linear relationship with health symptoms like “blocked / runny nose”, “chest tightness”, “headache” and “lethargy / tiredness”. On the other hand, there were significant positive linear relationship between “stuffy air” and health symptoms such as “chest tightness”, “headache” and “lethargy / tiredness”. The values of 'sig. (2-tailed)' for these correlations were less than the predetermined alpha value (0.01/2 = 0.025), thus the stated null hypothesis was rejected. This conclusion was made at the significance level of 0.01. Spearman’s rho correlation analysis results for the selected questions related to daylighting, occupants’ behavior and occupants’ health are as stated in Table 4. The results indicated that “sufficient daylight brightness” had significant positive linear relationship with “feeling healthy”. Besides, there were also significant relationship between “sufficient daylight brightness” and physical health symptoms such as “epiphora”, “headache” and “lethargy / tiredness”. Glare or contrast from unshaded windows did not give significant relationship to occupants’ behavior and occupants’ health. On the contrary, sufficient daylight to perform tasks such as reading, writing, computer work, leisure and cooking directly influenced occupants’ behavior to control electric lighting during day time. Apart from that, “using daylight” for task performance yielded significant positive lieanr relationship with the ability to “concentrate well”. The values of 'sig. (2-tailed)' for these correlations were less than the predetermined alpha value (0.01/2 = 0.025). DISCUSSION According to Malaysian Uniform Building by-law (UBBL), every room for residential purposes shall be provided with daylighting and natural ventilation by windows having a total area of not less than 10% of the clear floor area of such room and 5% of them shall be open able. Besides, every bathroom or toilet shall be provided with daylighting and ventilation by openings having a total area of not less than 0.2 m2 and such openings shall be open able (Lembaga Penyelidikan Undang-undang, 2013). Although all the terraced houses complied with UBBL, the findings reflected that the current requirement of 5% WFR for natural ventilation is inadequate for comfortable and healthy living as many occupants felt that there was insufficient indoor air movement including in the toilets (refer to Figure 2). The findings indicated that natural ventilation issues were more serious than daylighting issues in the terraced houses. One of the main reasons was the activities or tasks performed in terraced houses do not require high level of brightness, whereas most of the occupants in these houses relied on natural ventilation. Furthermore, the outdoor temperature is high and air velocity is low in Malaysia. Thus, more considerations shall be given to improve the indoor thermal comfort through proper passive ventilation strategies. The mean scores for health problems were relatively high (above 3.00 – average) in spite of the low mean scores for natural ventilation and daylighting. The analysis of health issues demonstrated that indoor living environment in terraced houses affected both the physical and physiological health. Occupants may feel unhealthy due to dissatisfaction with the indoor environment. Psycho-social effects, which relate to emotional and behavioral responses, influence how occupants perceive and behave in certain environment (Loewenstein, 2001; Bluyssen, 2010). For example, discomfort indoor temperature and insufficient daylight brightness caused psychological health isssues (refer to Table 3 and 4). The correlation tests evidenced significant linear relationship between indoor environment and health. The better the natural ventilation and daylighting performances in the terraced houses, the healthier the occupants perceived. On the contrary, the correlation tests proved that the relationships between behavior and natural ventilation as well as behavior and health were less significant. Only the relationship between behavior and daylighting was significant. For instance, the use of daylight for task performance gave significant effect on the occupants’ behavior to control electric lighting. The findings of this study suggest to review current terraced house design in relation to natural 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 229 ventilation and daylighting. The existing intermediate terraced house unit with deep planning design is not effective for passive strategies. Lager openings (higher WFR) are needed to allow sufficient natural ventilation and air movement to avoid symptoms such as chest tightness and lethargy. However, proper shading of the openings is necessary to allow sufficient daylight while eliminating solar heat gain. CONCLUSION The study concludes that there is significant positive linear relationship between indoor environments and occupants’ health. The current by-law requirement of 5% window-to-floor ratio for natural ventilation is inadequate for occupants’ comfort and health, thus further review is needed. Proper consideration of natural ventilation and daylighting design strategies in terraced house is essential as it determines how the occupants can manage the indoor environment to achieve comfortable and healthy living environment. ACKNOWLEDGEMENT The author would like to acknowledge the research funding by Universiti Teknologi Malaysia (UTM), Ministry of Education (MOE) through Research University Grant (GUP), Vote 07H36, titled “Dynamic Shading as Daylight and Solar Control for High-rise Office in Tropical Climate”. Besides, the author is the recipient of MT-DYNAC-UTMISI Partnerships in Research and Development Grant. REFERENCES Agung, N. M. and Mohd Hamdan, A. (2006) The development of solar chimney geometry for stack ventilation in Malaysia's single storey terraced house. Architecture & Environment. 5(2), October, pp. 77-96. Bluyssen, P. M. (2010) Towards new methods and ways to create healthy and comfortable buildings. Building and Environment, 45, pp. 808-818. Bluyssen, P. M., Aries, M., Dommelen, P. V. (2011) Comfort of workers in office buildings: The European HOPE project. Building and Environment, 46, pp. 280-288. Choi, J.-H., Beltran, L. O., Kim, H. -S. (2012) Impacts of indoor daylight environments on patient average length of stay (ALOS) in a healthcare facility. Building and Environment, 50, pp. 65-75. Hashim, A. H., & Rahim, Z. A. (2008) The Influence of Privacy Regulation on Urban Malay Families Living in Terrace Housing. International Journal of Architectural Research, 2(2), pp. 94-102. Kubota, T., Toe, D. H. C. and Ahmad, S. (2009) The effects of night ventilation technique on indoor thermal environment for residential buildings in hot-humid climate of Malaysia. Energy and Buildings, 41, pp. 829-839. Lembaga Penyelidikan Undang-undang (2013) Building (Federal Territory of Kuala Lumpur) By-laws 1985. Petaling Jaya, Malaysia: International Law Book Services, ISBN 9789678923255. Lim, Y. W. (2013) Indoor Environmental Comfort in Malaysian Urban Housing. American Journal of Environmental Science. 9(5), 431-438. Lim, Y. W., Mohd Zin, K., Mohd Hamdan, A., Ossen, D. R. and Aminatuzuhariah, M. A. (2012) Building Façade Design for Daylighting Quality in Typical Government Office Building. Building and Environment. 57, pp. 194-204. Loewenstein GF, Weber EU, Hsee CHK, Welch N. (2001) ‘Risk as feelings’. Psychologic Bull 2001;127:267–86. Malaysia Department of Statistic (2000) General Report of the Population and Housing Census, Department of Statistics Malaysia. Omar, E. O., Endut, E. and Saruwono, M. (2010) Adapting by altering: spatial modifications of terraced houses in the klang valley area. Asian Journal of Environment-Behaviour Studies, 1(3), pp. 1-10. Sadafi, N., Salleh, E., Haw, L. C. and Jaafar, Z. (2011) Evaluating thermal effects of internal courtyard in a tropical terrace house by computational simulation. Energy and Building, 43, pp. 887-893. Todorovic, M. S. and Kim, J. T. (2012) Buildings energy sustainability and health research via interdisciplinarity and harmony. Energy and Buildings, 47, pp. 12-18. Zakaria, R. (2007) Sustainable housing for residential-industrial neighbourhoods in Malaysia – a study on the elements of indoor enivironmental quality improvement. Doctor of Philosophy, Faculty of Built Environmental and Engineering, Queensland University of Technology. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 230 Thermal Comfort in Offices in India: Behavioral Adaptation and the Effect of Age and Gender Madhavi Indraganti, PhD Ryozo Ooka, Ph.D. [Prince Sultan Univ., Riyadh, Saudi Arabia] [The Univ. of Tokyo, Japan] mindraganti@pscw.psu.edu.sa; Madhavi.indraganti@fulbrightmail.org Hom B Rijal, Ph.D. [Tokyo City Univ., Japan] ABSTR ACT Reports on occupant’s behavioral adaptation in India are limited in the literature. We analyzed the data from a recent thermal comfort field study of office buildings in two capital cities of Chennai and Hyderabad in India. Behavioral adaptation formed a key mechanism contributing to the subject’s thermal comfort and user satisfaction in buildings. In mixed mode (MM) buildings, use of AC and/ or fan during temperature excursions proved to be an important and power saving adaptation. We present the logistic algorithms to predict the use of ACs and fans in MM buildings. Females, young subjects, and thin people had statistically significant and higher comfort temperature than males, older people, and obese occupants respectively. Females accepted the environments better. These findings might determine the design direction of future indoor environments. The occupants have undertaken several behavioral control actions throughout the year without many seasonal differences. Staying in airy place, drinking beverages, changing posture, and avoiding direct sunlight were the most prominent actions. INTRODUCTION Human existence hinges on thermal adaptation. In order to maintain the deep body temperature at 37 °C at all times, human beings adapt continuously. The adaptations are mainly physiological, psychological, environmental, and behavioral. Controlled by hypothalamus, physiological vasomotor regulation happens almost instantly. Human-envelop interaction greatly influences environmental and behavioral adaptation in buildings. This in turn also affects thermal satisfaction and energy consumption in them (Nicol & Humphreys, 2004; Brager, et al., 2004). Environmental adaptation by using various controls results in energy saving (Brager & Baker, 2009). India’s building energy consumption increased by about 3% per annum. It was 196.04 Million ton of oil equivalent in 2011, of which lighting, heating, ventilation, and air-conditioning constituted a major portion (IEA, 2011). India has an ever-widening energy supply-demand gap (Central Electricity Planning Authority, India, 2012). South India faces the maximum energy deficit of close to 30 %. Moreover, India is yet to have custom made adaptive thermal comfort standards. Recent research in Indian buildings proved that occupants in Indian buildings expressed comfort at much higher temperatures than expected (Indraganti, et al., 2014; Indraganti, 2010; Deb & Ramachadraiah, 2010; Dhaka, et al., 2013; Honnekiri, et al., 2014). Researchers attributed this to a wide range of adaptations in some of the studies. Understandably, mixed mode buildings with adaptive use of air-conditioners consume much lower energy than compared to the buildings air-conditioned throughout their active life. Madhavi Indraganti is an Associate Professor + Fulbright Scholar at the Architecture & Interior Design Dept., Prince Sultan University, Riyadh, Saudi Arabia. Ryozo Ooka is a Professor at the Institute of Industrial Science, The University of Tokyo, Japan. Hom B Rijal is an Associate Professor at the Department of Environmental & Information Studies, Tokyo City University, Yokohama, Japan 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 231 Indraganti (Indraganti, 2010a) reported the behavioral and occupant adaptation in apartments in India. Occupant adaptation with fans and windows and obstacles to adaptation in Indian offices were presented in Indraganti et al., (Indraganti, et al., 2014a). Others studied the mixed mode buildings for adaptation in summer (Honnekiri, et al., 2014). However, occupants undertake several behavioral control actions besides the operation of environmental controls in offices. They need investigation. Sixty five percent of India’s population is below 65 yrs. This gives it a rich demographic dividend (Basu, 2007). It particularly means that, India would have young population in the work environments and more women than is it now. Thermal necessities of this young group would be major drivers for design decisions in the future. Indian offices have about 25% female occupants now (Indraganti, et al., 2014). Researchers noted that occupant’s age, body constitution, and gender influenced their comfort perceptions in both homes and offices (Indraganti & Rao, 2010; Karyono, 2000; Fanger, 1970). For improving thermal satisfaction in buildings, we need to understand them. Therefore, this paper aims to explain the effect of age, body constitution, and gender on thermal comfort and highlights the user behavior in undertaking various adaptive control actions in offices in India. We also aim to develop algorithms to predict the use of air conditioners and fans in mixed mode offices. For this study, we use the long-term thermal comfort field study data obtained from offices in India (Indraganti, et al., 2014). METHODS AND FIELD SURVEY We conducted a thermal comfort field survey in 28 office buildings from 01-2012 to 02-2013. It was in two State Capitals: Chennai (N13°04’ and E80° 17’) and Hyderabad (N17°27’ and E78° 28’) with warm humid wet land coastal climate and composite climates respectively. These have four distinct seasons: summer, Southwest monsoon (SWM), Northeast monsoon (NEM) and winter. The surveys were paper based surveys. About 2787 occupants gave 6042 datasets. In all the offices close to the subjects at 1.1m level from the ground, we measured the indoor air temperature (Ta), globe temperature (Tg), air velocity (Va), and relative humidity, while they filled in the questionnaires (Fig.1). High precision digital instruments and standard protocols were used with accuracies: thermometers: ±0.5 °C, hygrometer: ±5%, anemometer: ± 0.01 m/s. We spaced the surveys at four to six weeks. (1) The instrument setup, (A) Thermo-hygro meter (TR 76Ui), (B) Hot-wire Figure 1 anemometer (Testo 405) (C) Globe thermometer (TR 52i), (2, 3) Typical survey environments The Survey Questionnaire The questionnaire had three sections: (1) personal identifiers, (2) thermal responses and (3) Behavioral control actions undertaken (McCartney & Nicol, 2002). While the survey was going on, the interviewers noted down their clothing (Icl_tot), activity (Met) and the personal environmental controls in use in that space. Thermal responses included standard questions on thermal sensation (TS); preference (TP); acceptability (TA); and sensation and preference for other environmental parameters. Indraganti et al. elaborated the methods and questionnaires (Indraganti, et al., 2014). We measured TS with ASHRAE’s seven point scale having: cold (-3); cool (-2); slightly cool (-1); neutral (0); slightly warm (1); warm (2); and hot (3) and TA through a direct question with 0: acceptable; and 1: Unacceptable. THE SUBJECT AND BUILDING SAMPLE The occupants were in the age group of 18- 70 years and were associated with the environments for 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 232 longer than three months. The age and gender profile of the subjects was similar in both the cities as shown in Fig. 2. Women constituted about 21 – 25 % of the sample. Majority of the subjects were in 2535 years age group. Figure 2 Age and gender profile of the subjects in the survey (M: Male; F: Female) Building Types and Modes of Operation We surveyed fourteen buildings in each city. These are of three types: (1) fully naturally ventilated, (2) mixed mode (MM) and (3) air-conditioned throughout (ACall). We had thirteen MM buildings, fourteen ACall buildings and one NV building. Of the total 6048 sets of data, 3804 came from the ACall and 2212 were from the MM buildings. We collected thermal responses from the subjects when the buildings were operated in naturally ventilated (NV) mode and air-conditioned (AC) mode in all the buildings. However, due to frequent power outages, in some ACall buildings, AC was switched-off during an outage and the building was not run NV mode either. Data collected in ACall buildings with AC switched off is termed as ACoff. About 10 % of the data was collected in ACoff mode. We used the data collected in NV and AC modes here. Table 1. Descriptive statistics of outdoor and indoor environmental variables NV C H All Mean SD Mean SD Mean SD To ( °C) 26.9 2.7 25.4 3.0 25.5 3.0 Ta ( °C) 29.7 2.2 29.0 1.9 29.1 1.9 Tg ( °C) 29.5 2.1 28.7 2.0 28.8 2.0 Va (m/s) 0.46 0.33 0.13 0.21 0.17 0.25 RH (%) 59.7 5.5 43.1 11.1 44.7 11.7 Icl_tot (clo) 0.69 0.08 0.71 0.08 0.70 0.08 Met (met) 1.0 0.1 1.0 0.1 1.0 0.1 C: Chennai; H: Hyderabad; SD: Standard deviation C Mean 28.9 26.5 26.7 0.15 50.1 0.71 1.0 SD 2.8 1.5 1.4 0.20 7.5 0.09 0.1 AC H Mean SD 27.5 3.9 26.1 1.6 25.6 1.7 0.05 0.08 45.6 10.8 0.69 0.06 1.0 0.0 All Mean 28.4 26.3 26.2 0.11 48.2 0.70 1.0 SD 3.4 1.6 1.6 0.17 9.3 0.08 0.1 RESULTS AND DISCUSSION OUTDOOR AND INDOOR ENVIRONMENTS (a) Field survey data superimposed on the psychrometric chart; (b) Intercity variation in Figure 3 TA; (c) Modal variation in mean Tg when subjects voted on TA scale. Error bars indicate 95% CI. Outdoors were hot in summer and warm through the rest of the survey. Being a coastal city 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 233 Chennai was more humid throughout, and Hyderabad was humid in monsoon seasons. We obtained outdoor daily mean temperature (To) data from meteorological records. Chennai indoors were significantly warmer and more humid than Hyderabad at 95% confidence interval (CI)(All declarations are at 95% CI unless and otherwise specified explicitly). Interestingly, we recorded 80% of the data in NV and AC modes when Tg was less than 29.8 and 27.5 °C and Va was around 0.25 m/s and 0.15 m/s respectively. About 70% and 77% environments in NV and AC modes had humidity ratio (Wv) less than 12 ga/kgda, the upper limit suggested in ASHRAE Std-55 (Fig.3, Table 1) (ASHRAE, 2010; Indraganti, et al., 2014). Occupants achieved indoor air movement primarily by using the common fans and through the operation of openings in addition, in NV mode. A few ACall offices in Hyderabad did not have fans however. Women had slightly but significantly higher Ict_tot than men did (N = 6048, p < 0.001). SUBJECTIVE THERMAL RESPONSES Fig.4 shows the probit lines of proportion voting on a given TS scale point (X) or lower against Tg. It also shows the probits for the proportion voting comfortable (-1≤ TS ≤ 1) and juxtaposed with the actual proportion comfortable. From these we can observe that in NV mode, 80% subjects voted in the central three categories on TS when 25.6 ≤ Tg ≤ 28.1 ºC. Probit lines indicating the percentage voting at a given TS scale point (X) and lower on Figure 4 Tg in NV and AC modes in India. Also shown are the probits for the proportion voting comfortable and the actual proportion voting comfortable (-1≤ TS ≤ 1) at each 1 K bin of Tg. X: Scale value of TS Thermal acceptability remained the same at around 72 - 71% in both NV and AC modes. Interestingly occupants accepted the NV environments at 28.5 ºC (Tg-mean) and in AC at 26.2 ºC. (Fig. 3b,c) It may be possible that in AC environments, subjects’ acclimatization to the narrower thermal regime perhaps had influenced the TA outcome. Brager and de Dear (Brager & de Dear, 2000) demonstrated that people who were exposed to a small range of temperatures (mostly through HVAC systems) developed high expectations for homogeneity and cool temperatures, and were soon critical of the subsequent thermal migrations indoors. EFFECT OF GENDER, AGE, BODY FAT ON THERMAL ACCEPTABILITY AND COMFORT TEMPERATURE Figure 5 (a) Mean Tcomf significantly varying with gender, body constitution, and age group; The 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 234 error bars indicate 95% CI. (b) Logistic regression Tg and TA for both the genders (p<0.001). Actual data are superimposed. More men were critical of an environment than women were. We estimated the comfort temperature (Tcomf) using the Griffith’s method taking 0.5 as the coefficient, similar to others (Griffiths, 1990; Humphreys, et al., 2013). The comfort temperature varied with the outdoor temperature. Mean Tcomf in NV mode was 28 °C and in AC was 26.4 °C. Notably, women had higher comfort temperature than men (Fig. 5). Similarly, subjects younger than 25 years old had higher Tcomf, than the older group. At normal activity (1.0 met) and at common indoor clothing (Ict_tot ≤ 0.7 clo), we noted women having 0.6 K higher Tcomf than men did. It equaled a sensation scale value of 0.22, (N = 3239, p<0.001). Conversely, Parsons (Parsons, 2002) recorded no gender differences in thermal comfort for the same Met and Icl_tot. In this context, it is important to note that females in India are mostly (99%) dressed in loose fitting Indian attires, with much better scope for thermal adaptation (Indraganti, et al., 2014). This in part explains higher Tcomf of women. A Finnish experiment, found TA in females significantly lesser than males. He attributed this in part to the unawareness about the thermostat and HVAC systems (Karjalainen, 2007). The mean temperature where younger subjects accepted the environment was higher than that of their older counterparts. Women also accepted the environment better. For example, at 32 ºC the acceptability among women was about 10.6 % higher as seen in Fig. 5b. We noted a similar trend between the two age groups. Researchers in Indonesia found men feeling warmer than women in offices (Karyono, 2000), unlike the Fanger’s experiment on American and Danish subjects where there were no significant gender differences in comfort sensation (Fanger, 1970). This is despite the fact that women (mean 0.78 clo) had significantly higher clothing insulation than men (mean 0.68 clo) in India (N = 6048, p<0.001). Thin subjects (body mass index (BMI) < 18.5 kg/m2) (WHO, 2004) recorded comfort at 27.1°C while fat people (BMI > 25 kg/ m2) expressed comfort at 0.7 K lesser (Fig 5a). Karyono also noted the same 0.7 K difference (N = 3865, p < 0.001), while Fanger recorded 0.26 K difference between thin and fat college age subjects (Karyono, 2000; Fanger, 1970). However, we found no statistically significant differences in acceptability among thin and fat people. ALGORITHM TO PREDITCT THE USE OF CONTROLS IN MM BUILDINGS: AC AND FAN In the era of power outages and expensive energy tariffs, mixed mode buildings with AC usage limited to the overheated periods come as a welcome respite. As the outdoor conditions became warmer, occupants adopted through ACs in mixed mode offices: for ex. during the mid-day and in summer and monsoon seasons. It formed an important adaptation strategy in addition to the use of fans. Use of AC in Chennai MM buildings was 82% (mean) while in Hyderabad MM, it was 20%. Chennai has very hot and humid climate, which could have triggered higher AC use. On the other hand, fan usage was much higher in Hyderabad MM. It was 78%, as against 43% in Chennai MM. Interestingly; in Chennai ACall buildings, we noted 20% fan usage, while ACall in Hyderabad had very few fans available. Logistic regression with outdoor daily mean temperature showing the proportion of (a) Figure 6 AC use in MM buildings in Chennai and Hyderabad; (b) fan use in ACall and MM buildings in Chennai 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 235 (C) and Hyderabad (H). Also shown juxtaposed are the actual proportions in 1K bins of To. For all the equations p<0.001, and slopes are significantly different at 95% CI. The surveys provided binary data on the use of various environmental controls. We then applied the logistic regression on the ‘control usage’ against its stimulus (i.e. temperature) to develop the algorithm (Nicol, 2001; Rijal, et al., 2008). It yielded the following equations as shown in Table 2 and Fig. 6, where p is the probability of a control in use and To is the outdoor daily mean temperature. Table 2. Logistic regression of AC and fan usage in Chennai (C) and Hyderabad (H) Control Case Equation Sample size Negelekerke R2 Standard Error AC C: MM logit(p) = 0.30 To - 6.86 723 0.134 0.043 H: MM logit(p) = 0.35 To - 10.93 1489 0.394 0.02 Fan C: ACall logit(p) = 0.59 To - 17.62 1389 0.264 0.04 C: MM logit(p) = 0.75 To - 20.89 672 0.300 0.07 H: MM logit(p) = 0.51 To - 13.07 1356 0.143 0.05 Logit(p): Probability of a control being in use From these equations, we can estimate that 89.4% and 28.5% ACs would be in operation when the To is at 29 °C for MM buildings of Chennai and Hyderabad respectively. Higher AC use in Chennai is perhaps due to its warmer thermo-hygro regime. Similarly, at To of 29 °C, nearly 40%, 68% and 85% fans would be on in Chennai-ACall, Chennai-MM, and Hyderabad MM buildings. Rijal et al. noted 81% fans on at 29 °C in Pakistan (Rijal, et al., 2008). It is important to note that in mixed mode buildings of Hyderabad, subjects have made use of the fans more than the ACs, making great energy dividends. ADAPTATION THROUGH BEHAVIORUAL CONTROL ACTIONS Occupants in real environments continuously adapt through various behavioral control actions. These vary with the thermal stimuli and happen in immediate response to the stimuli, throughout the day and year. These can vary with season. Therefore, we included a multitude of behavioral actions listed in Fig. 7 in the questionnaire. In all the surveys, the subjects chose from the fourteen possible behavioral control actions they must have undertaken during or fifteen minutes prior to the survey (Rijal, et al., 2010). These were noted down as binary data, an action in use: 1 and, not in use: 0. On analysis of the responses, we noted very few seasonal changes in the pattern of behavioral adaptation. Nevertheless, there were slight differences in NV and AC modes, overall. Proportion using various behavioral control actions round the year in both NV and AC Figure 7 modes. Staying in airy place is an important behavioral action. Error bars indicate 95% CI of mean. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 236 Among all the adaptive actions, ‘staying in airy place, drinking cold/ hot beverages, changing posture’ were most prominent in both NV and AC modes. In addition, in NV mode, occupants also ‘avoided direct sun light, rested, rinsed face and hands and stayed away from heat sources.’ A higher percentage of subjects adapted through behavioral actions in NV mode than in AC mode, possibly due to the warmer conditions in NV. Similarly, subjects in Swiss offices consumed significantly more cool drinks as the temperature went up (Haldi & Robinson, 2008). As the conditions in the offices continued to be slightly warm throughout the year, the adaptive actions generally used in winter conditions like, ‘stay in a warmer place, move close to direct sunlight, move away from airy places, were not used much. The seasonal differences were also not significant. ‘Staying in airy place’ was the most frequent behavioral action in summer (used by about 30 – 21 % in NV and AC modes). More importantly, subjects adopting this action encountered 0.9 K warmer indoor (Ta) environments and lower air speeds than otherwise (N = 1487, p < 0.001). Similarly, mean Ta in AC when people ‘avoided direct sun’ was ⅓ K higher than otherwise, (N = 4310, p = 0.001). As we can see, subjects have adopted through these actions and responded to their immediate thermal environment. Actions like drinking cold and hot beverages and adding removing clothing /slippers were some of the actions that subjects adopted in both summer and winter over a wide temperature regime. Female occupants predominantly preferred to use extra layers (sweaters or Dupatta/ shawls) adaptively when challenged by cold drafts. Indian ensembles like sari and Salwar-Kamiz were much more tenable for adaptation unlike the western outfits of men (Indraganti, et al., 2014). In addition, men had dress code while women had none. Liu et al. also observed season specific clothing adaptation in Chinese workplaces (Liu, et al., 2012). In addition to these, Rijal et al. found subjects taking extra showers and resting in summer in residential areas (Rijal, et al., 2010). CONCLUSIONS This paper discussed the behavioral adaptation and the effect of age, gender and body constitution on thermal comfort in Indian offices, relying on a recent field study data. The occupants undertook several behavioral control actions throughout the year. In mixed mode (MM) buildings, use of AC during temperature excursions proved to be an important and power saving adaptation. We presented algorithms to predict the AC and fan usage in always air-conditioned (ACall) and MM buildings in Chennai and Hyderabad. Females, thin people and the subjects under 25yrs age group had higher comfort temperature than males, ‘over 25 yrs’ age group and obese people. Women accepted the environments better. All these differences are statistically significant. The engineering significance of these differences may be limited, but these would determine the design direction of indoor environments of the future. ACKNOWLEDGMENTS The Japan Society for Promotion of Science and The University of Tokyo, Japan funded this research. Under the supervision of Gail Brager of the Centre for the Built Environment, University of California Berkeley, the Fulbright Grant provided for part of the data analysis. We acknowledge their support. We also thank Mukta Ramola and Prakash K for the field surveys, Government of India and Andhra Pradesh, all the heads of the offices, and the subjects for their co-operation and participation. REFERENCES ASHRAE, 2010. ANSI/ ASHRAE Standard 55-2010, Thermal environmental conditions for human occupancy, Atlanta: s.n. Basu, K., 2007. India's demographic dividend. [Online] Available at: http://news.bbc.co.uk/2/hi/south_asia/6911544.stm Brager, G. & Baker, L., 2009. Occupant Satisfaction in Mixed-Mode Buildings. Building Research & Information, 37(4), pp. 369-380. Brager, G. & de Dear, R., 2000. A standard for natural ventilation. ASHRAE Journal, 42(10 O), pp. 21-28. Brager, G., Paliaga, G. & de Dear, R., 2004. Operable Windows, Personal Control. ASHRAE transaction, Volume 110-2, pp. 17-35. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 237 Central Electricity Planning Authority, India, 2012. Power Scenario at a Glance. [Online] Available at: http://www.cea.nic.in/reports/planning/power_scenario.pdf Deb, C. & Ramachadraiah, A., 2010. Evaluation of thermal comfort in a rail terminal location in India. Building and Environment, Volume 45, pp. 2571-2580. Dhaka, S. et al., 2013. Evaluation of thermal environmental conditions and thermal perception at naturally ventilated hostels of undergraduate students in composite climate. Building and Environment, Volume 66, pp. 42-53. Fanger, P. O., 1970. Thermal Comfort, Analysis und Applications in Environmental Engineering. Copenhagen: Danish Technical Press. Griffiths, I. D., 1990. Thermal Comfort in Buildings with Passive Solar Features: Field Studies , s.l.: s.n. Haldi, F. & Robinson, D., 2008. On the behaviour and adaptation of office occupants. Building and Environment, Volume 43, pp. 2163-2177. Honnekiri, A., Brager, G., Dhaka, S. & Mathur, J., 2014. Comfort and adaptation in mixed-mode building in a hot-dry climate. Cumberland Lodge, Windsor, UK 10-13 April , Network for Comfort and Energy Use in Buildings, http://nceub.org.uk. Humphreys, M., Rijal, H. & Nicol, J., 2013. Updating the adaptive relation between climate and comfort indoors; new insights and an extended database. Building and Environment , Volume 63, pp. 40-45. IEA, 2011. Energy balances of non-OECD countries, s.l.: www.iea.org. Indraganti, M., 2010a. Behavioural adaptation and the use of environmental controls in summer for thermal comfort in apartments in India. Energy and Buildings, pp. 42 (1019-1025). Indraganti, M., 2010. Thermal comfort in naturally ventilated apartments in summer: findings from a field study in Hyderabad, India. Applied Energy, Volume 87, pp. 87 (866-883). Indraganti, M., Lee, J., Zhang, H. & Arens, E., 2014. Versatile Indian sari: clothing insulation with different drapes of typical sari ensembles. Windsor Great Park, London, Network for Comfort and Energy Use in Buildings, http://nceub.org.uk. Indraganti, M., Ooka, R., Rijal, H. B. & Brager, G. S., 2014. Adaptive model of thermal comfort for offices in hot and humid climates of India. Building and Environment, April, 74(4), pp. 39-53. Indraganti, M., Ooka, R., Rijal, H. B. & Brager, G. S., 2014. Adaptive model of thermal comfort for offices in hot and humid climates of India. Building and Environment, April, 74(4), pp. 39-53. Indraganti, M., Ooka, R., Rijal, H. & Brager, G., 2014a. Occupant behaviour and obstacles in operating the openings in offices in India. Cumberland Lodge, Windsor, London, UK, Network for Comfort and Energy Use in Buildings, http://nceub.org.uk. Indraganti, M. & Rao, K. D., 2010. Effect of age, gender, economic group and tenure on thermal comfort: A field study in residential buildings in hot and dry climate with seasonal variations. Energy and Buildings, Volume 42, pp. 273 - 281. Karjalainen, S., 2007. Gender differences in thermal comfort and use of thermostats in everyday thermal environments. Building and Environment, Volume 42, pp. 1594- 1603. Karyono, T. H., 2000. Report on thermal comfort and building energy studies in Jakarta, Indonesia. Building and Environment, Volume 35, pp. 77-90. Liu, J., yao, R., Wang, J. & Li, B., 2012. Occupants’ behavioural adaptation in workplaces with non-central heating and cooling systems. Applied Thermal Engineering , Volume 35, pp. 40- 54. McCartney, K. J. & Nicol, J. F., 2002. Developing an adaptive control algorithm for Europe. Energy and Buildings, Volume 32, p. 623–635. Nicol, F. J., 2001. Characterising occupant behaviour in buildings: towards a stochastic model of occupant use of windows, lights, blinds, heaters and fans. Rio, Brazil, Proceedings of the seventh international IBPSA conference. Nicol, J. & Humphreys, M., 2004. A stochastic approach to thermal comfort—occupant behaviour and energy use in buildings. ASHRAE Transactions, Symposia, pp. 554 - 568. Parsons, K. C., 2002. The effects of gender, acclimation state, the opportunity to adjust clothing and physical disability on requirements for thermal comfort. Energy and Buildings, Volume 34, p. 593–599. Rijal, H. B. et al., 2008. Development of adaptive algorithms for the operation of windows, fans, and doors to predict thermal comfort and energy use in Pakistani buildings. ASHRAE Transactions, SL-08(056), pp. 555-573. Rijal, H., Yoshida, H. & Umemiya, N., 2010. Seasonal and regional differences in neutral temperatures in Nepalese traditional vernacular houses. Building and Environment, 45(12), pp. 2743-2753. WHO, 2004. BMI Classification. [Online] Available at: http://apps.who.int/bmi/index.jsp?introPage=intro_3.html [Accessed 26 5 2014]. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 238 Factors Influencing Window Opening Behavior in Apartments of Indonesia Meita T Arethusa Tetsu Kubota, Dr Agung Murti N, Dr Hiroshima University Hiroshima University University of Brawijaya Sri Nastiti E, Dr I Gusti N Antaryama, Dr Tomoko Uno, Dr Sepuluh November Institute of Technology Sepuluh November Institute of Technology Mukogawa Women’s University ABSTRACT This study aims to investigate the factors influencing occupants’ window-opening behavior in apartments in the major cities of Indonesia. For this purpose, a field survey was carried out for 347 respondents in the city of Surabaya, covering detailed household and building profile, cooling energy consumption, thermal conditions, satisfactions and preferences, and duration and reasons for opening doors or windows. The results showed that the majority of respondents in the naturally-ventilated public apartments open at least one of the doors and windows during daytime (60-80%) and tend to close them during nighttime. In contrast, very few respondents in the air-conditioned private apartments utilize their openings. The respondents in public apartments tend to open either door or window on one side of living room, but not both. Privacy and security were found to be the main reasons affecting this behavior. The average duration of opening doors/ windows was 16-17 hours/day in the public apartments and less than 5 hours/day in the private apartments. Multiple regression analyses were carried out to further investigate the factors influencing the duration for opening doors/ windows. It was found that the opening behavior is highly influenced by not only thermal conditions, but also other factors such as size of balcony, size of corridor space, usage of cooling appliances, and background noise. INTRODUCTION Indonesia has been experiencing high economic growth and the middle class is now on the rise. The total population of Indonesia increased by more than double over the last four decades whereas the nationwide final energy demand rose by 14 times over the same period. The growing housing demand for the above emerging middle class is expected to require more apartments in the near future. In locations with high thermal stress such as Surabaya, the ownership of air conditioners is becoming less luxurious even in residential buildings (Ekasiwi, 2013). Therefore, it is important to determine energysaving strategies for the future middle-class apartments. Window-opening is considered to be one of the major adaptive behaviors of occupants to achieve their thermal comfort particularly in naturally ventilated houses in hot-humid climate. Studies by Rijal et al. in UK and Japan (2007, 2013) showed that the highest usage of window-opening was found in summer and the lowest was in winter. Since the number and size of openings in apartments are limited, Meita T Arethusa is a master student in the Graduate School for International Development and Cooperation (IDEC), Hiroshima University, Hiroshima, Japan. Tetsu Kubota is an associate professor in the Graduate School for IDEC, Hiroshima University, Hiroshima, Japan. Agung Murti N is a lecturer in the Department of Architecture, University of Brawijaya, Malang, Indonesia. Sri Nastiti E and I Gusti N Antaryama are lecturers in the Department of Architecture, Sepuluh November Institute of Technology, Surabaya, Indonesia. Tomoko Uno is a lecturer in the Department of Architecture, Mukogawa Women’s University, Hyogo, Japan. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 239 window-opening/closing behavior may affect their thermal comfort more significantly compared to that in landed houses. The behavior of opening windows may also have direct effects on energy consumption for cooling by changing air-flow rate inside the buildings (Fabi et al., 2012). Therefore, the factors affecting people’s behavior of opening or closing their windows are continuously studied. Finding these factors may help architects to create designs which can encourage people to actively open their windows. To date, most of the relevant studies tried to examine the relation between window-opening behavior and thermal conditions. This is because the action of opening windows are assumed to be stimulated by people’s reactions to thermal discomfort. The frequency and length of opening windows were found to be much less in air-conditioned buildings rather than those in naturally ventilated buildings (Rijal et al., 2007, and Rijal et al., 2013). Nevertheless, there are various other factors that may affect the behavior of opening windows, including physical environmental, contextual, psychological, physiological, and social factors (Fabi et al., 2012). Nonetheless, studies involving social and psychological factors are mostly conducted not in the field of building science, but in psychology. This paper aims to investigate various factors influencing window-opening behavior in apartments in hothumid climate of Indonesia. METHODS Surabaya is located on 7°9’21” South Latitude and 112°36’-57” East Longitude. It is the capital city of East Java Province and the second biggest city in Indonesia. The city has a hot-humid climate, with monthly average temperature ranging from 27.2-29.0°C, monthly average relative humidity of 65.980.9%, and monthly average wind speeds of 2.12-3.10m/s (NCDC, 2014). Both private and public apartments in Surabaya were addressed in this study (Fig. 1a). Private apartments are normally high-rise buildings and contain 14-33 floor heights. Most of the houses in private apartments are equipped with air-conditioners. On the other hand, public apartments contain 3-5 floor heights and almost all the houses are naturally ventilated. Public apartments were initially built to resettle slum squatters for providing better living environments (hereafter, ‘old public apartment’). However, since 2008, the target was extended into low to middle income classes in response to housing demand (hereafter, ‘new public apartment’). In this study, 8 public and 8 private estates from a total of 48 housing estates were selected through proportional stratified samplings (Table 1). The field survey was conducted from September to October 2013, consisting of face-to-face interviews and one-week thermal measurements. The interviews covered 347 respondents, comprising the following detailed information: (1) household and building profile, (2) energy consumption and usage of cooling appliances, (3) thermal satisfactions and preferences, and (4) duration and reasons for Ping-pong ball Sensor Data logger Old public Figure 1 Old public Figure 2 New public (a) Views of apartments Private Indoor Outdoor (b) Thermal measurements (a) Views of apartments from each category and (b) one week thermal measurements New public Typical room arrangement in apartments 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Private 240 opening doors or windows. One-week thermal measurements were conducted in 30 apartment houses to measure globe temperature, indoor and outdoor air temperature, and indoor and outdoor relative humidity. Globe temperature was measured using TR-52i (T&D Corporation) by inserting the sensor into a black-painted-Ping-Pong ball. The Ping-Pong ball was then positioned at 110 cm height above floor. Indoor air temperature and relative humidity were measured using TR-72ui (T&D Corporation) which was placed slightly below the Ping-Pong ball. The equipment was installed in the living room and set to avoid direct sunlight. Outdoor measurement instruments (TR-73ui; T&D Corporation) were placed in open spaces at a certain height under the shade to prevent the effect of solar radiation (Fig. 1b). RESULTS AND DISCUSSION Profile of Respondents As shown in the Table 1, the average household sizes are 3.5 to 3.6 for the public apartments and 1.9 for the private apartments. Age of respondents ranges from 20 to more than 60. The monthly average household income is the highest in the private apartments. More respondents in the new public apartments have a higher income than those in the old public apartments. Typical unit in public apartments (both old and new) consist of a room, a balcony, and a bathroom; with floor areas of 18-21m² (Fig. 2). However, the newest type in the new public apartments has slightly larger floor areas (2432m²) and consists a bedroom, a living room, a kitchen, a bathroom, and a balcony. On the other hand, private apartments have two types of room arrangements. The first is single room which contains only one room (for bedroom and kitchen) and a bathroom. The second is family room which has two bedrooms, one living room (also functioning as kitchen and dining room), and a bathroom. Daily Usage Pattern of Window Opening Behavior The following four typical openings in apartments are considered in this paper: front door, front window, back door, and back window. In most of the apartments, the front door (facing corridor space) is directly placed on one side of the living room, whereas the back door is normally placed on the rear side of the living room adjacent to the balcony. The size of the doors per floor area ranges from 6% to 10%, whereas the size of windows per floor area ranges from 2% - 8%. The size of balcony per floor area ranges from 9% - 22% and the size of corridor space per floor area ranges from 15% - 28%. As shown in Table 2, the respondents use their openings for 14.8 hours on average. The average opening duration during daytime (6:00-18:00) (9.7 hours) is almost two times longer than that of nighttime (5.1 hours). This tendency is seen similarly for all the categories. In the private apartments, the average duration of opening doors or windows is very short, even in daytime (less than 5 hours). The respondents in public apartments tend to open their doors (12.8 hours) for a longer period than to open Table 1 Profile of respondents Sample size Household size (persons) Age (%) 20-30 (years old) 31-40 41-50 51-60 >60 Monthly income (%) <100 (US$)* 100-200 200-300 300-400 >400 No. of apartments Built before 2008 Built after 2008 Floor area (m²) Whole sample 347 3.5 Old public 208 3.6 New public 101 3.5 Private 38 1.9 22.0 34.6 25.2 11.3 7.0 13.8 38.3 30.0 11.7 6.6 25.3 32.6 22.1 12.6 7.4 70.4 14.8 3.7 3.7 7.4 7.5 36.9 21.0 12.4 10.4 8.7 49.0 21.6 11.5 6.3 7.9 23.8 26.7 15.8 18.8 5.3 2.6 7.9 10.5 7 9 18-38 5 18-21 3 24-32 2 6 18-38 *exchange rate: 1 IDR=9,387 USD (Average per 2012, The World Bank) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 241 Table 2. Average hours for opening doors or windows Whole sample 14.8 9.7 5.1 11.5 8.1 5.8 6.2 8.5 2.9 100% 80% 60% 40% 20% 0% 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 am Open both pm Open door only Open window only Open window only Front openings Open both Open door only 100% 80% 60% 40% 20% 0% am Open window only Close both Rear openings 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 Open both pm Open door only am Open window only Close both Rear openings (c) Private apartments Daily usage patterns for opening doors and windows Reasons Whole for opening Old public doors or windows To get fresh air To let wind enter To cool the room To remove odor View Others New public To get fresh air To let wind enter To cool the room To remove odor View Others 0 20 40 60 80 20 40 60 80 Old public apartments 20 40 60 Frequency (%) Whole sample 80 40 60 80 20 40 60 Frequency (%) Old public apartments 80 20 40 60 80 Private apartments New public Private Privacy Security Insects Others Dust/pollution Rain Noise AC Don’t know 0 0 Frequency (%) New public apartments Insects Privacy Security Others Dust/pollution Rain Noise Don’t know AC 0 20 Frequency (%) Whole Old public (b) Reasons for not opening doors or windows Privacy Insects Security Dust/pollution Others Rain Noise AC Don’t know To get fresh air To let wind enter To remove odor View To cool the room Others 0 Frequency (%) Whole sample Private To get fresh air To let wind enter To cool the room To remove odor Others View 0 Frequency (%) Figure 4 pm Open door only am Close both Front openings Figure 3 (a) Open both am Open window only Close both 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 (b) New public apartments pm am Open window only Rear openings am Close both 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 am 100% 80% 60% 40% 20% 0% am Frequency 100% 80% 60% 40% 20% 0% Open door only pm Open door only (a) Old public apartments pm Private 4.8 3.1 1.7 1.4 3.4 0.7 1.0 3.4 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 Open both Frequency Frequency Frequency Open both 100% 80% 60% 40% 20% 0% am Close both 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 am New Public 17.0 10.4 6.6 12.8 10.6 4.6 8.6 10.2 3.5 am Front openings 100% 80% 60% 40% 20% 0% Old Public 15.5 10.5 4.9 12.8 7.6 7.5 5.6 9.1 2.8 Frequency Frequency All openings Day Night Doors Windows Front door Front window Back door Back window AC Dust/pollution Insects Security Privacy Noise Rain Don’t know Others 0 20 40 60 Frequency (%) New public apartments 80 0 20 40 60 80 Frequency (%) Private apartments Reasons for: (a) opening doors/windows; and (b) closing doors/windows windows (7.6 and 10.6 hours for old and new public apartments respectively). In contrast, the respondents in private apartments tend to open their windows (3.4 hours) longer than doors (1.4 hours). The back door is used the longest (8.5 hours), whereas the back window is least used (2.9 hours). This tendency is found similarly among the respondents in public apartments. In contrast, the respondents in private apartments tend to open their back window longer than other openings (3.4 hours). The respondents in the old public apartments tend to open their front door longer by 1.9 hours than the front window, while those in new public apartments tend to open the front window longer by 4.0 hours than the front door. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 242 The daily usage patterns of doors and windows in each category are showed in Fig. 3. The figure clearly shows that the occupants in the public apartments (both old and new) open at least one door or window in both front and rear side of living room particularly during daytime (60-80%), while those in the private apartments rarely open doors or windows (0-30%). In general, most of the respondents in public apartments tend to open only one opening for one side of unit at one time. Less than 20% of the respondents open both windows and doors even during daytime. In the case of old public apartments, the usage of front door (40-60%) was higher compared to the front window (30-40%). In contrast, the respondents in the new public apartments tend to open the front window (50-60%) more than the front door (20-40%). About 40-50% of the respondents of both public apartments open only the back door and tend to close the back window. In the nighttime, the respondents in both apartments tend to close their front window and door, while 20-40% of them still open the back door. Reasons for Opening and Closing Windows Fig. 4 illustrates major reasons for respondents to open or close their doors or windows. The highest reasons were found to be: ‘obtaining fresh air’ (74.3%), ‘letting wind to enter’ (66.2%), and ‘to provide cooling’ (45.4%) for all the categories. On the other hand, the top reasons for not opening doors or windows were different for each category: ‘AC usage’ (52.6%) for the private apartments, ‘insects’ for the old public apartments (52.9%), and ‘privacy’ (57.0%) for the new public apartments. ‘Security’ reason may especially affect the closing behavior during nighttime, while ‘privacy’ may be the main reason for closing behavior during daytime. As previously discussed, the pattern and average duration of opening front door and window were different between old and new public apartments, unlike the back door and window (see Fig. 3). Furthermore, the respondents were found to open either door or window on one side of living room, but not both. Fig. 4 implies that the occupants in new public apartments are more concerned about privacy and security than those in old public apartments. In the old public apartments, these concerns should be less because most of the residents were relocated from the same areas. This may be one of the reasons for opening front door more in the old public apartments than those in the new public apartments. Thermal Conditions Fig. 5 shows the results of one-week thermal measurements. As indicated, the outdoor air temperature ranges from 25.6-36.7°C whereas the relative humidity ranges from 27-77% during the measurement periods. The indoor air temperature ranges from 29.1-33.0°C for naturally ventilated apartments (i.e. public apartments) and 25.8-31.6°C for air-conditioned apartments (i.e. private apartments). Even in the naturally ventilated apartments, the diurnal air temperature ranges are smaller than that of the outdoor temperature, though the mean indoor air temperatures in both old and new public apartments are slightly higher than the mean outdoor air temperature (30.3°C). The previous results showed that most of the occupants open doors or windows particularly during daytime (see Fig. 3). This implies that the ventilation rates in these public apartments were not necessarily sufficient to change the indoor air even when the doors or windows were opened during daytime. Thermal Sensations and Preferences Fig. 6 shows the sensation and preference of respondents for thermal condition, air flow, and humidity in the living room during the day. The sensations were measured in a 7-point scale while preferences were measured in a 5-point scale. More than 57% of the respondents regard the thermal condition in their living room as ‘warm’ to ‘hot’ even for those in the private apartments. Accordingly, the preferences for cooler environments are evident (more than 68%). Despite the use of airconditioners, more than 97% of respondents in the private apartments prefer cooler indoor conditions. Meanwhile, more than 40% of the respondents found humidity in the living room to be ‘slightly humid’ to ‘very humid’, thus they prefer ‘less humid’ conditions (38.5%). As for the indoor air flow, more than 70% of respondents consider it to be ‘slightly high’ to ‘very high’. However, more than half of the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 243 40 36 32 28 24 Whole Old samples Public 80 36 Relative humidity (%) Operative temperature (°C) Air temperature (°C) 40 32 28 24 New Private Outdoor Public Whole Old samples Public 60 40 20 New Private Outdoor Public Whole Old samples Public New Private Outdoor Public Statistical summary of air temperature, operative temperature, and relative humidity Figure 5 Air flow during the day - Whole sample Humidity during thedeviation) day - Whole sample during the day mean and ± one (5th andThermal 95thcondition percentiles, standard neutral Sensation Preference no change 0% 20% Sensation hot 40% Preference much cooler cooler 60% 80% humid 100% slightly humid 0% 20% 40% Thermal condition during the day 80% 100% 0% 20% high no change 40% 60% 80% 100% Frequency Humidity during the day Air flow during the day Thermal sensations and preferences for thermal condition, air flow, and humidity (c) Usage of air conditioning (b) Usage of fan (a) Ownership of cooling appliances Whole sample Whole Whole Old Public Old Public Old Public Exhaust New Public AC Private Whole day New Public Fan 0% 20% 40% 60% 80% 100% Ownership level (%) Figure 7 60% more wind Preference Frequency Frequency Figure 6 no change less humid neutral slightly high Sensation dry neutral Night Private Day 0 4 8 12 16 Whole day New Public Night Private 20 Usage time (hours/ day) Day 0 2 4 6 8 10 Usage time (hours/ day) Ownership and usage of cooling appliances respondents do not prefer to change the current air flow conditions, while about 30% prefer even higher air flow. This result clearly indicates their high preference for a better air flow condition in their apartments. Usage of Fan and Air Conditioner Fig. 7 shows the ownership and usage of cooling appliances in respective apartments. More than 90% of houses in public apartments (both old and new) owned one or more fans, whereas more than 90% of houses in private apartments were equipped with air conditioner. The average daily usage time of fan is 16.7 hours for old public apartments, 13.1 hours for new public apartments, and 1.4 hours for private apartments. The occupants in public apartments use fan longer during nighttime (10.2 hours and 8.8 hours) than daytime (6.4 and 4.1 hours), in contrast with the usage of doors/windows (see Table. 2). On the other hand, the usage of air conditioner was the highest in private apartments (5.1 hours during nighttime and 1.9 hours during daytime). Factors Influencing Window Opening Behavior: Multiple Regression Analyses Multiple regression analyses were carried out for five target variables: (a) daily usage of doors/windows, (b) daytime usage of doors/windows, (c) nighttime usage of doors/windows, (d) daily usage of doors, and (e) daily usage of windows. In this analysis, thermal conditions are represented by thermal sensations and preferences. A total of 44 variables were considered. Step-wise method with pairwise deletion was utilized to determine the factors which best explain the window opening behavior of apartments’ users in Surabaya. The result for daily usage of doors/windows (Table 3a) shows that high correlations are presented by ‘size of balcony’ (0.24***) and ‘size of corridor space’ (0.22***). As the size of these two transition space larger, the duration of opening doors/windows tend to be longer. ‘Energy for cooling’ decrease 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 244 when the duration of opening doors/windows increase (-0.18**), in contrast with ‘usage of fan’ during daytime (0.18**). On the other hand, the daily usage of doors/windows tend to be longer when the respondents prefer noisier condition (0.12*). During daytime (Table 3b), ‘size of balcony’ and ‘usage of fan’ remain to have high influences (0.19***), followed by ‘monthly income’ (-0.18***) and ‘size of household’ (0.16**). This means that respondents with higher income tend to open their door/windows shorter during daytime. Further, space limitation can be one of the reasons for the respondents with higher household size to open their doors/windows longer during daytime. ‘Preferences for thermal condition’ shows relation of 0.12* which implies that daytime usage of doors/windows is longer when the respondents has adjusted with hotter thermal condition. During nighttime (Table 3c), influences are showed by ‘preferences for background noise’ (0.14*) and ‘energy for cooling’ (-0.13*). Table 3d shows that the duration of opening doors tend to be longer when the ‘size of balcony’ and ‘size of front openings’ increase. ‘Preferences for background noise’ also give influence of 0.19***, followed by ‘sensation for air quality’ (0.15**), ‘floor level’ (-0.13*), and ‘quality of life’ (-0.12*). In most of apartments in Indonesia, first floor is used for public spaces instead of living spaces. Thus, the negative correlation from ‘floor level’ might be caused by other reason such as ‘age of respondents’. In fact, the relation of ‘age’ and floor level is -0.29***, which indicates that older people tend to live in the lower floor. It is predicted that people in older age who mostly live in those apartments longer tend to open their door longer than younger respondents. On the other hand, ‘size of corridor space’ has positive correlation with duration of opening windows (0.48***, Table 3e), in contrast with ‘size of front openings’ (-0.15**). This perhaps indicates that rather than the ‘size of front openings’, size of corridor space play more important role in encouraging the respondents to open their windows longer. ‘Preferences for natural lighting during the day’ and ‘satisfaction for life’ also show influences to the daily usage of windows with correlation of 0.13* and 0.11*, respectively. Table 3 Results of multiple regression analysis, explaining window opening behavior in apartments of Surabaya (a) Daily usage of doors/windows Independent variable Size of balcony per floor area Size of corridor space per floor area Energy for cooling Usage of fan during daytime Preferences for background noise R² Adj. R² n β 0.24 0.22 -0.18 0.18 0.12 0.19 0.17 331 r *** *** ** ** * * * 0.24 0.15 -0.16 0.16 0.19 *** ** ** ** *** (b) Daytime usage of doors/windows Independent variable (c) Nighttime usage of doors/windows β Independent variable r β Size of balcony per floor area Usage of fan during daytime Monthly income 0.19 0.19 -0.18 *** *** *** 0.29 0.26 -0.25 *** *** *** Preferences for background noise Energy for cooling Size of household Energy for cooling Preferences for thermal condition during the day R² Adj. R² n 0.16 -0.13 ** * 0.25 -0.11 *** * R² Adj. R² 0.04 0.03 0.12 * 0.11 * n 331 0.23 0.21 331 * * (d) Daily usage of doors Independent variable r 0.14 -0.13 * 0.15 ** * -0.15 ** * * (e) Daily usage of windows β Independent variable r Size of balcony per floor area 0.37 *** 0.39 *** Size of front openings per floor area * 0.20 *** 0.12 Preferences for background noise Sensation for air quality Floor level 0.19 0.15 -0.13 *** ** * 0.19 0.14 -0.24 *** ** *** Quality of life -0.12 * -0.17 *** R² Adj. R² 0.29 0.26 * * n 331 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Size of corridor space per floor area Size of front openings per floor area Preferences for natural lighting during the day β r 0.48 *** 0.31 *** -0.15 ** -0.15 ** 0.13 * 0.20 *** Satisfaction for life 0.11 * 0.12 * R² 0.27 * Adj. R² n 0.25 331 * 245 CONCLUSIONS (1) The respondents in the naturally ventilated public apartments showed relatively high utilization of openings on both front and rear side of unit during the daytime (60-80%), whereas very few respondents in the air-conditioned private apartments open their doors/ windows (0-30%). In the public apartments, the occupants tend to open either window or door on each side of unit, but not both. The respondents in the old public apartments open their front door more than the front window during daytime, in contrast with the respondents in the new public apartments. In the case of rear opening, the respondents in both old and new public apartments tend to open the back door rather than back window. The average duration of opening doors/windows was 16-17 hours/day in the public apartments and less than 5 hours/day in the private apartments. Obtaining fresh air, letting wind enter, and obtaining cooling were the main reasons for respondents to open doors/windows, whereas concerns of privacy and security may be the reasons for them to close their doors/windows. (2) The diurnal indoor air temperature ranges in the selected apartments were smaller than that of the outdoor temperature, even for naturally ventilated apartments. It indicates that the ventilation rates in these apartments were not necessarily sufficient to change the indoor air even when the openings were utilized. (3) The occupants tend to perceive their thermal condition as ‘hot’, thus the preference for cooler indoor condition was evident. To improve their thermal comfort, the occupants in naturally ventilated apartments tend to increase indoor air speed to enhance skin evaporation by opening doors/windows. Fans were also utilized to further increase the air speed, especially during the nighttime. Correspondingly, most of the respondents regarded their sensation for air flow as ‘high’ and prefered ‘no change’ or ‘more air flow’ for the current air flow condition. (4) The results of multiple regression analyses implied that among the selected factors, the behavior of opening and closing doors/windows is affected not only by thermal sensations and preferences, but also building profiles and usage of cooling appliances. Major factors included ‘size of balcony’, ‘size of corridor space’, ‘usage of fan’, ‘energy for cooling’ and ‘preferences for background noise’. In designing apartments, more attention should be drawn to the size of balcony and corridor space to encourage the occupants to more actively open their doors/windows. (5) In this study, air temperature, relative humidity, and globe temperature were measured during one week period. In the future study, the resulting indoor air flow and air change rates under the open/closed window conditions should be analyzed to further understand the occupants’ adaptive behavior by opening windows. REFERENCES Ekasiwi, S., Majid, N., Hokoi, S., Oka, D., Takagi, N., and Uno, T. (2013). Field survey of air conditioner temperature settings in hot, humid climates, part1: questionnaire results on use of air conditioners in houses during sleep. Journal of Asian Architecture and Building Engineering, 12(1), 141-148. Fabi, V, Andersen, R., Corgnati, S., and Olesen, B. (2012). Occupants’ window opening behavior: A literature review of factors influencing occupant behavior and models. Building and Environment (58), 188-198. NCDC. Climate Data Online. Global Surface Summary of Day Data < https://gis.ncdc.noaa.gov/map/viewer/#app=cdo> (accessed 18 May 2014). Rijal, H., Tuohy, P., Humphreys, M., Nicol, J., Samuel, A., and Clarke, J. (2007). Using results from field surveys to predict the effect of open windows on thermal comfort and energy use in buildings. Energy and Buildings (39), 823-836. Rijal, H., Honjo, M., Kobayashi, R., Nakaya, T. (2013). Investigation of comfort temperature, adaptive model and the window-opening behavior in Japanese houses. Architectural Science Review, 56(1), 54-69. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 246 Session 2D : Tools and methods/ framework PLEA2014: Day 1, Tuesday, December 16 14:10 - 15:50, Trust - Knowledge Consortium of Gujarat An Operational Indicator System for the Integration of Sustainability into the Design Process of Urban Wasteland Regeneration Projects Martine Laprise, MArch Sophie Lufkin, PhD Emmanuel Rey, Prof. PhD Laboratory of Architecture and Sustainable Technologies (LAST), Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland. Laboratory of Architecture and Sustainable Technologies (LAST), Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland. Laboratory of Architecture and Sustainable Technologies (LAST), Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland. ABSTRACT In the context of sustainable development of European post-industrial cities, urban wastelands offer an important potential of surfaces to recapture. The regeneration of these sites is indeed an opportunity to simultaneously create density within the existing built fabric and revitalize some portions of cities and metropolitan areas. Although the launching of several initiatives of this type can be observed, their implication toward sustainable development is in most cases implicit and superficial. In point of fact, integration of sustainability into urban wasteland regeneration projects cannot be summarized by a mere density issue. It requires a proactive search for global quality, implemented in a participative way into the project dynamics, and a continuous monitoring of environmental, social and economic dimensions adapted to such projects. Specifically addressing these considerations, this paper introduces the development of an operational indicator system for the integration of sustainability into the design process of urban wasteland regeneration projects. It aims to provide a tool for structured and continuous evaluation, hinged on their specific characteristics, and to give useful basis to stakeholders involved in their management. Subsequently, the paper presents a first test application performed on a project underway in Switzerland, which validates its usability. Further work suggests the integration of the system into a digital monitoring tool in order to make it applicable to a variety of projects of this type. INTRODUCTION Although Urban Wasteland Regeneration Projects (UWRP) embody a strategic potential to revitalize and densify existing urban fabrics, they are often not as sustainable as they may seem (Rey, 2007). Indeed, this strategy - as part of the compact and polycentric city model (Jenks, 1998; Rogers & Gumuchdjian, 1998) - is no guaranty of inclusion of the three pillars of sustainability: the economic, the ecological and the sociocultural (Andres & Bochet, 2010). Integration of sustainability into UWRP goes through the pursuit of a global quality and a constant follow-up of environmental, sociocultural and economic dimensions adapted to projects on those sites (Rey, 2012). In this sense, existing evaluation Emmanuel REY is professor at the Ecole Polytechnique Fédérale de Lausanne (EPFL) and head of the Laboratory of architecture and sustainable technologies (LAST), Switzerland. Dr. Sophie Lufkin is a scientist and Martine Laprise is a doctoral assistant at the same laboratory. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 247 systems for large scale developments (LEED ND, BREEAM communities, HQE aménagement, DGNB New urban districts, etc.) do not address the specificities of UWRP. Therefore, few studies have developed methodologies for sustainability assessment adapted to the characteristics of the projects located on these sites. However, our analysis reveals that they are dissociated from the design process and do not totally address the specificities of UWRP. As a result, none of them are currently operational. Hence, the purpose of this paper is to introduce SIPRIUS (Rey, 2012), an operational indicator system for the integration of specific sustainability issues into the design process of such projects, as well as the methodology that precedes its development. Afterwards, a test application on a case study in Neuchâtel (Switzerland) demonstrates the relevance and applicability of the indicator system to UWRP. SPECIFICITIES OF UWRP Wastelands have a unique identity, whether positive or negative (cultural symbol, economic and social stigma, sense of insecurity, risk of contamination, etc.). Projects on these sites are not limited to a single building. Quite the contrary, their scale ranges from urban planning to architectural design. Hence, neighborhood scale seems the most appropriate to encompass the full implications of these projects (CABERNET, 2006). But unlike new neighborhood developments, urban wastelands are already transformed and yet abandoned. Economic and ecological potential of existing buildings – and consequently architectural heritage – implies making decisions on the level of conservation (OFEN, 2013). Moreover, because they are disconnected from their urban context and emptied from permanent population, projects on wastelands neither can be considered as neighborhood renewals. Urban wastelands are not irreversible but their regeneration is highly complex. This is due in part to the long duration of the regeneration process, which involves a variation of several elements (conditions, needs, modification of general terms, changes in project leaders and actors, etc.). Moreover, the implication of a multitude of stakeholders with varying degrees of influence and interest tends to complicate the process (Doak & Karadimitriou, 2007). REQUIREMENTS FOR THE OPERATIONAL INDICATOR SYSTEM Given their specificities, having a clear idea of where UWRP are heading in terms of sustainability is crucial to build a solid foundation for their future (Hollander, Kirkwood, & Gold, 2010); particularly since most UWRP refer partially to sustainable development, generally in favor of environmental aspects (Franz, Pahlen, Nathanail, Okuniek, & Koj, 2006). These considerations call for the development of an operational indicator system tailored to the needs of UWRP in order to integrate sustainability into their design process. This objective is reflected in the following specifications: 1. Search for a global quality: The indicator system covers a relatively wide range of parameters to address the environmental, social and economic sustainability, equally and concurrently; 2. Appropriateness to UWRP: The indicator system meets the specificity inherent to UWRP. In particular, adaptation to the scale and complexity of the project and consideration of an already built-up site; 3. Inclusion of the principles of monitoring: The indicator system ensures an operational assessment, i.e. visualization of the various phases of the project and establishment of reference values in order to follow and act on performance trends. METHODOLOGY The methodology used to establish the indicator system is based on three main steps, namely the determination of criteria, indicators and then reference values for each indicator. It is worth nothing that a test application was done in parallel with the construction of the indicator system, which helped to perform various practical settings and iterative improvements. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 248 Identification of criteria The first step is to determine a list of criteria that portrays the multidimensional aspects of UWRP in the context of sustainable development. In this sense, the list of criteria endeavors to give an equivalent importance to the three pillars of sustainability. It is based on the definition of fundamental objectives of sustainability and on operational considerations coming from practical experience of the test application. Given the requirements related to the project’s scale, the identification of criteria seeks to distinguish those that refer to the context of those that refer to the project. 1. Context criteria: concern aspects which are clearly beyond the physical boundaries of the site. Either the project has an impact in a wider sphere than that defined by the wasteland or external factors interact with the project; 2. Project criteria: involve aspects whose issues are within the boundaries of the site. These criteria relate to built-up and unbuilt areas. Identification of indicators To assess the selected criteria, it is necessary to determine one (or more) indicator(s) that is a “value” that can be “measured” to indicate the degree of satisfaction with each criterion. It is essential to note that the notion of value is to be understood in its broadest sense: it can be both qualitative and quantitative, provided that it gives an explicit indication on the project. To ensure the legitimacy of the system, the selection of indicators is subject to a number of methodological rules and fundamental principles (Bossel, 1999). They stress that the indicator should be: 1. Exhaustive: together, represent proportionally and holistically the three dimensions inherent to the concept of sustainable development; 2. Relevant: synthetically reflect the performance of the project in relation to a given criterion; 3. Sensitive: respond significantly to variations of the parameter that is evaluated for both quantitative and qualitative indicators; 4. Objective: eliminate ambiguity. Requires a precise definition of the indicator and its evaluation method; 5. Accessible: depend on known values or known quantities and reflects the reality of the usual practice. Quantitative indicators must be easily calculated, qualitative indicators depend on a clear description; 6. Readable: ensure simplicity of interpretation, as it is intended to contribute to decision-making and to communication of the results to multidisciplinary stakeholders. Identification of reference values Finally, reference values are defined for each indicator. These values may correspond to quantitative data, from overall performance encountered in professional practice, or qualitative characteristics, defined by a description of specific issues or concrete elements that are related. Aiming to include monitoring principles, a set of determined values is used for the measurement and follow-up on project performance. 1. 2. 3. 4. Limit Value (VL): Minimum value required for any project (or veto value); Average Value (VA): Value corresponding to the usual practice, no particular performance; Target Value (VT): Value to target in order to achieve a greater performance; Best Practice Value (VB): Value corresponding to a particularly high performance. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 249 RESULTS Operational indicator system - SIPRIUS The selection and identification of criteria and indicators, followed by the definition of reference values, led to the creation of an operational indicator system. Entitled SIPRIUS, it is a catalog from which planners can choose indicators considered as significant for the monitoring of their UWRP. Depending on the specific characteristics of some projects, it is possible to add indicators to those already provided. SIPRIUS is composed of 9 criteria and 21 indicators relating to the context presented in Table 1 as well as 12 criteria and 21 indicators relating to the project presented in Table 2. Table 1. Summary List of Criteria and Indicators Related to Context Sustainability Criterion Criterion Indicator Indicator Pillar Code Title Code Title Mobility Environment C1 C1a Quality of service in public transport C1b Number of parking spaces C1c Tying status with soft mobility networks C2 Pollution C2a Average annual emission of NO2 C2b Acidification Potential (AP) C2c Global Warming Potential (GWP) Noise C3 C3a Average emission of noise - day C3b Average emission of noise - night Proximity of school Sociocultural C4 C4a Average distance to a nursery facilities C4b Average distance to kindergarten C4c Average distance to an elementary school C4d Average distance to a junior school C4e Average distance to a high school C5 Proximity of C5a Average distance to a commercial zone commercial facilities Proximity of C6 C6a Average distance to a public park recreational facilities C6b Average distance to a recreational greenspace/natural area C6c Average distance to a cultural center C6d Average distance to a sport center Population Economy C7 C7a Net population density Job C8 C8a Net employment density C9 Local Economy C9a Proportion of work by local companies Table 2. Summary List of Criteria and Indicators Related to Project Criterion Indicator Indicator Sustainability Criterion Pillar Code Title Code Title Land P1a Land use coefficient Environment P1 P2 Energy P2a Non-renewable primary energy for construction, renovation and demolition of buildings P2b Non-renewable energy for buildings in operations P3 Water P3a Infiltration surfaces and stormwater management Biodiversity P4a Green surfaces P4 Well-being P5a Annual hours of overheating Sociocultural P5 P5b Interior noise level P5c Average daylight factor P5d Degree of electrosmog P5e Degree of individualization of housing 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 250 Economy P6 Security P7 P8 Heritage Diversity P9 Direct costs P10 P11 P12 Indirect costs External costs Flexibility P5f P6a Quality of outdoor spaces Degree of security P7a P8a P8b P8c P9a P9b P10a P11a P12a Degree of enhancement of heritage Degree of functional mix Potential of social diversity Degree of universal access Investment costs Gross rental yield Annual operating costs External costs Degree of flexibility of buildings Each indicator is developed in a synthetic datasheet that includes all necessary informations to perform an assessment. As examples, the datasheet of two indicators assessing the environmental dimension of UWRP are illustrated in Table 3 and Table 4, respectively C1c - "Tying status with soft mobility network" and P1a - "Land use coefficient". In an attempt to be representative of the variety of assessment methods, one addresses the context and uses qualitative values while the second refers to the project with quantitative measurements. The reference values are assigned in this case relatively to the Swiss context since the test application was carried out on an ongoing project in this country. Data source Table 3. Datasheet of the indicator C1c C1c Tying status with soft mobility networks Intensity of connection to different networks for pedestrians and bicycles Analysis of the number and quality of the various links Qualitative scale (from 0 to 5) Level 2 (in 5) The project is characterized by a relatively weak consideration of soft mobility and has connections only in two distinct directions Level 3 (in 5) The project is characterized by a moderate consideration of soft mobility and has connections in three distinct directions Level 4 (in 5) The project provides an important consideration of soft mobility and has connections in three distinct directions Level 5 (in 5) The project includes a systematic consideration of soft mobility (many specific devices) and has connections in four distinct directions Plan and project guidelines Indicator Definition Evaluation mode Measurement unit VL (Limit Value) VA (Average Value) VT (Target Value) VB (Best Practice Value) Data source Table 4. Datasheet of the indicator P1a P1a Land use coefficient Ratio of gross floor area and the area of land Measurement of surfaces considered from the project plans. Quantitative scale 0.5 (Limit to ensure public transport offer) 1.0 1.5 2.0 (Density of central areas – Switzerland) Project plan and cadastral data of the land Indicator Definition Evaluation mode Measurement unit VL (Limit Value) VA (Average Value) VT (Target Value) VB (Best Practice Value) Test application The Ecoparc neighborhood consists in the regeneration of an urban wasteland of about 4 hectares located in Neuchâtel, Switzerland. The triggering of the regeneration process dates back to 1989; since then, Bauart Architects and Planners Ltd is in charge of developing the project. The latter involves the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 251 creation of a new dense urban center based on a mix of functions (housing, offices and educational buildings), combining new constructions with transformation of old industrial buildings. Ecoparc is committed to a sustainable development approach. The project represents the inherent complexity of this type of operation (multiple stages of development, many actors involved, etc.) (Rey, 2002). The status of Ecoparc at the time of the assessment is an intermediate stage typical of UWRP (Figure 1). Figure 1. Site plan of Ecoparc at assessment. In black, buildings in operation. In yellow, buildings in execution phase. In orange, buildings in design phase. (Bauart document, 2005) For the sake of complete verification, all indicators with available reference values were evaluated during the test application, including those that were not subject to particular attention. To integrate the results in the project dynamics - in accordance with the principles of monitoring - a bar chart histogram that follows the entire design process is used for each type of criteria. This paper exposes in Figure 2 and Figure 3 the results of the two indicators presented in Table 3 and 4. The project of Ecoparc includes a connection to pedestrian/bicycle networks in the four cardinal directions. In this sense, it tends to tie links that did not previously exist and increases the space reserved for pedestrian / bicycle along the main street. Expected final value corresponds to the target value VT. Figure 2. Test-application: evaluation of indicator C1c (Tying status with soft mobility network). Ecoparc is characterized by an “optimal density” that simultaneously combines a quantitative contribution to urban regeneration and creates a neighborhood that takes into account the qualitative characteristics of the site. The expected final situation is 2.11 slightly above the Best Practice value VB. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 252 Figure 3. Test-application: evaluation of indicator P1a (Degree of flexibility of buildings). VALIDATION OF THE INDICATOR SYSTEM The test application has resulted in a real and iterative verification of the practical relevance of the methodology developed by SIPRIUS. It has proven that the indicator system is operational and can be used to assess both types of indicators. Table 5 shows the results for the expected final situation of Ecoparc. In general, the evaluation has confirmed that Ecoparc falls significantly in a sustainable development approach. Indeed, the vast majority of indicators provides an expected final situation that meets the objectives as shown in Table 5. In that sense, the indicator system has contributed to raise awareness about various aspects of sustainability within the project. Table 5. Test application: distributions of indicators based on the values obtained for the expected final situation in respect of the initial situation and the reference values. Status Context Project 19 19 Number of indicators evaluated Expected final situation greater than or equal to the 16 (84%) 19 (100%) objectives Expected final situation greater than or equal to VL 19 (100%) 19 (100%) Expected final situation greater than or equal to VA 17 (89%) 18 (95%) 13 (69%) 14 (74%) Expected final situation greater than or equal to VT 8 (42%) 2 (11%) Expected final situation greater than or equal to VB In addition, the test application showed that SIPRIUS takes into account the requirements for a holistic evaluation of UWRP: it includes global quality and is adapted to the specificities of urban wastelands. Moreover, the graphical representation of the results helps to visualize the multiple phases of the project and the determination of reference values as “level of performance”. In this sense, it sets basis to project monitoring by aiming at a greater sustainability. These complementary aspects validate that a relevant operational indicator system can be developed to suit the needs of UWRP for an integration of sustainability issues in the design process. Toward an operational monitoring tool To concretize sustainability targets, their integration into the project dynamics of urban wasteland regeneration and their continuous follow-up is an essential condition. The indicator system SIPRIUS contributes to this objective. It highlights the strengths and weaknesses of the project and feeds interactions amongst planners and decision makers. It also contributes to the transmission of results to audiences from various perspectives. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 253 Nevertheless, its use depends primarily on the involvement and motivation of the stakeholders. Thus, the adaptation and transposition into a digital device in order to make a fully operational monitoring tool applicable to a multitude of regeneration projects would concretely facilitate the integration of sustainability in UWRP. Further work will be carried out in order to reach this objective, and will be the subject of future publications. CONCLUSION Three successive stages - identification of criteria, indicators and reference values - led to the creation of the operational indicator system SIPRIUS. The creation of the indicator system was done in parallel to the completion of a comprehensive test application. SIPRIUS meets the requirements of a search of global quality, is adapted to the specificities of UWRP and includes monitoring principles. The test application demonstrates that the indicator system is operational and contributes to integrate sustainability into the design process of the project. A transposition toward a digital monitoring tool in order to facilitate evaluation of diverse UWRP is suggested. Research is moving in this direction. ACKNOWLEDGMENTS We would like to acknowledge financial support from Swiss National Science Foundation within the framework of Project No 100013_143376. REFERENCES Andres, L., and Bochet, B. 2010. Regenerating brownfields and promoting sustainable development in France and in Switzerland: what convergences? Revue d’Economie Régionale & Urbaine, (4): 729– 746. Bossel, H. 1999. Indicators for Sustainable Development: Theory, Method, Applications (A report to the Balaton Group). Winnipeg, Canada: IISD International Institute for Sustainable Development. Retrieved from http://www.ulb.ac.be/ceese/STAFF/Tom/bossel.pdf CABERNET. 2006. Sustainable Brownfield Regeneration: CABERNET Network Report. University of Nottingham. Doak, J., and Karadimitriou, N. 2007. Actor Networks: The Brownfield Merry-Go-Round. In T. Dixon, M. Raco, P. Catney, & D. N. Lerner (Eds.), Sustainable brownfield regeneration : liveable places from problem spaces. Oxford: Blackwell: pp. 67–88 Franz, M., Pahlen, G., Nathanail, P., Okuniek, N., and Koj, A. 2006. Sustainable development and brownfield regeneration. What defines the quality of derelict land recycling? Environmental Sciences, 3(2): 135–151. Hollander, J. B., Kirkwood, N., and Gold, J. L. 2010. Principles of brownfield regeneration : cleanup, design, and reuse of derelict land. Washington: Island Press. Jenks, M. 1998. The compact city: a sustainable urban form? [Reprinted]. London etc: Spon. OFEN, O. fédéral de l’énergie. 2013. Réhabiliter des friches industrielles pour réaliser la société à 2000 Watts. Pediaditi, K. E., Wehrmeyer, W., and Chenoweth, J. 2005. Monitoring the Sustainability of Brownfield Redevelopment Projects The Redevelopment Assessment Framework (RAF). Contaminated Land & Reclamation, 13(2): 173–183. Rey, E. 2002. The Ecoparc project in Neuchâtel : sustainable regeneration of an urban wasteland (pp. 963–966). Presented at the Proceedings of PLEA 2002, Toulouse: PLEA - Passive and Low Energy Architecture. Rey, E. 2007. Des friches urbaines aux quartiers durables. Tracés, (12): 13–15. Rey, E. 2012. Régénération des friches urbaines et développement durable : vers une évaluation intégrée à la dynamique du projet. Louvain-La-Neuve: Presses Universitaires de Louvain. Rogers, R. G., and Gumuchdjian, P. 1998. Cities for a small planet (Icon Editions.). Boulder, Colo.: Westview. Wedding, G. C., and Crawford-Brown, D. 2007. Measuring site-level success in brownfield redevelopments: A focus on sustainability and green building. Journal of Environmental Management, 85(2): 483–495. Williams, K., and Dair, C. 2007. A framework for assessing the sustainability of brownfield developments. Journal of Environmental Planning and Management, 50(1): 23–40. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 254 A simplified Approach to integrate Energy Calculations in the Life Cycle Assessment of Neighbourhoods Damien Trigaux, PhD Karen Allacker, Prof. Frank de Troyer, Prof. [KU Leuven] damien.trigaux@asro.kuleuven.be [KU Leuven] [KU Leuven] ABSTRACT Life Cycle Assessment (LCA) is a method which can be used to effectively evaluate and optimize the environmental impact of the built environment. However, when carrying out an LCA on the neighbourhood scale, estimating the energy consumption in buildings is problematic because most energy simulation tools require a lot of input data, which are not available in the master planning stage. This paper proposes a simplified approach to evaluate the heating energy consumption in neighbourhoods, taking into account the neighbourhood layout and shading caused by interacting buildings. The proposed approach, which is implemented for the Belgian context, is a refinement of the existing Equivalent Degree Day (EDD) method, by including results from both semi-dynamic and dynamic solar gain calculations. To illustrate this new approach, a parametric neighbourhood model is developed, linked to the energy simulation software EnergyPlus and LCA calculations. Simulations of a medium-density urban block reveal substantial differences in heating energy consumption, depending on shading patterns, confirming the importance of integrating simple but reliable energy calculations in neighbourhood LCA. 1. INTRODUCTION AND OBJECTIVES In order to move towards a more sustainable built environment, new urban developments need to be planned and organized differently. As shown in previous studies (Trigaux, Allacker, & De Troyer, 2014) (Herfray, 2012), life cycle assessment (LCA) is a method which can be used to effectively evaluate and optimize the environmental impact of buildings and neighbourhoods. However, when carrying out an LCA on the neighbourhood scale, estimating the energy consumption in buildings is problematic, especially for passive and low energy design. To date, most building energy simulation tools require a large amout of data, which is unavailable in the early stage of the design process. Although technical aspects are often not considered during the master planning stage, decisions related to the neighbourhood layout, building compactness and solar shading can also affect the heating energy demand importantly. Therefore a fast but reliable energy calculation method is needed, taking those aspects into account. This paper proposes a simplified approach to evaluate the heating energy consumption in neighbourhoods during the master planning stage; which can be integrated in an LCA study. The proposed approach is based on the Equivalent Degree Day (EDD) method (Diensten voor de programmatie van het wetenschapsbeleid, 1984), giving a first estimation of the heating energy demand. In order to accurately consider the impact of solar gains and shading caused by interacting buildings, the original method is refined, using results from both semi-dynamic and dynamic solar gain calculations. Damien Trigaux is a PhD researcher in the Department of Architecture, KU Leuven, Leuven, Belgium. Karen Allacker is a professor in the Department of Architecture, KU Leuven, Leuven, Belgium. Frank De Troyer is a professor in the Department of Architecture, KU Leuven, Leuven, Belgium. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 255 This new method, further referred to as dynamic Equivalent Degree Day method, is illustrated based on a parametric neighbourhood model linked to the energy simulation software EnergyPlus (U.S Department of Energy, n.d.). The energy calculations are moreover integrated in a broader LCA study. In the subsequent section the methodological aspects are described, focussing on the Dynamic EDD method and the LCA. In section 3 the parametric model is described and used to analyse the energy consumption and life cycle environmental impact of a medium-density urban block. Conclusions and recommendations are drawn in the final section. 2. METHODOLOGY From the Degree Day method to the dynamic Equivalent Degree Day method The Degree Day (DD) method is an existing method to estimate heating requirements in buildings, when no layout decisions are taken. The basic assumption is that the yearly heating demand at a specific location is proportional to the number of DD (°d) at that location (Diensten voor de programmatie van het wetenschapsbeleid, 1984a). For each day of the heating season, the difference between the average indoor temperature (Ti) and average outdoor temperature (Te) is calculated. The sum of all these daily temperature differences over the whole heating season results in the number of DD. This is illustrated in Figure 1 with the DD 15/18 for the temperate Belgian climate. In this specific case it is assumed that no heating is required when the daily average outdoor temperature is higher than 15°. Furthermore, a fixed average indoor temperature of 18°C is considered. In Figure 1, the number of DD is represented by the surface enclosed by the indoor and outdoor temperature curves and the lines delimiting the heating season. For practical reasons, this surface is approximated using monthly average outdoor temperatures, as illustrated by the hatched surface in Figure 1. In this paper, 2738°d are calculated as representative for the Belgian climate, based on the EnergyPlus test reference year for Brussels. DD 15/18 for the temperate Belgian climate (Diensten voor de programmatie van het Figure 1 wetenschapsbeleid, 1984a, p.33). Based on the number of DD, the heating energy demand (Qj) is estimated using Formula 1 (Diensten voor de programmatie van het wetenschapsbeleid, 1984b, p.39), which includes the impact of heat transmission losses through the building skin and heat ventilation losses: Q j = (U m * S + V * n * 0.36) * 3600 * 24 * °d 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad (1) 256 With: • • • • • Um = average heat transfer coefficient (W/m²K) S = heat loss surface (m²) V = inside building volume (m³) n = air change per hour (1/h) °d = number of Degree Days The Equivalent Degree Day (EDD) method is a refinement of the DD method, as the latter often leads to an overestimation of the heating demand (Diensten voor de programmatie van het wetenschapsbeleid, 1984b). Internal heat gains (resulting from people, electric devices and artificial lighting) and solar gains, which are not considered in the DD method, often result in a reduction of the heating demand, especially in well-insulated buildings. For this reason, the DD method was refined by defining EDD (eq d°). EDD (Figure 2) are calculated based on two temperature curves: the temperature curve of no more heating (TNH) and the temperature curve without heating (TWH). The first one (TNH) is defined as the indoor temperature above which no heating is required. This TNH is lower than the original indoor temperature of 18°C, since the internal gains will be sufficient to compensate the heat losses. The second temperature curve (TWH) is the increased indoor temperature, resulting from solar gains, when the building is not heated and not occupied. Equivalent Degree Days 15/18 for the temperate Belgian climate (Diensten voor de Figure 2 programmatie van het wetenschapsbeleid, 1984a, p.36)). The TWH is hence calculated, based on the useful solar gains in a building. These solar gains can be estimated by using several approaches, ranging from static to dynamic simulations. In the original EDD method, a static approach was followed, based on average solar radiation data for two characteristic months of the year (respectively March and December) (Diensten voor de programmatie van het wetenschapsbeleid, 1984b). In previous research, Allacker (Allacker, 2010) determined an average of 1200 eq d° for residential buildings in the Belgian context, based on an analysis of two dwelling types, and for several insulation levels. The calculation of this average EDD was based on the Flemish Energy Performance of Buildings (EPB) regulation (Flemish Government, 2005). This estimation is used in the analysis (section 3) as a reference base for the more dynamic calculations (see next paragraph). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 257 This paper proposes a new method, the dynamic Equivalent Degree Day method, based on two more dynamic solar gain calculations. Firstly, a semi-dynamic calculation was made based on the Flemish EPB regulation (Flemish Government, 2005). In this first approach, a characteristic day of each month is considered and shading patterns, resulting from neighbouring buildings, trees, sheds or side walls, are approximated by defining a set of obstruction and overhang angles per window. For each window those angles are then projected on the visible part of the sky dome to calculate the reduction in direct solar radiation, compared to unshaded conditions. As illustrated in Figure 3 for a dwelling in a rectangular urban block, this approximation can lead to an overestimation of shading patterns and thus lower solar gains than in reality. Secondly, a dynamic energy caculation, using the the software EnergyPlus, was made to calculate the indoor temperature without heating (TWH). In this second approach, solar gains are simulated, based on detailed reflection algorithms, for all days of a test reference year. The results of both approaches were compared and are discussed in the subsequent section. Stereographic projection of shading obstructions for a window in a rectangular urban Figure 3 block. Real obstructions (left) are compared with the EPB approximation (right). LCA method The environmental impact assessment in this paper is based on an existing LCA method developed within the MMG (“Milieugerelateerde Materiaalprestatie van Gebouwelementen”) research project, commissioned by the Public Waste Agency of Flanders (OVAM) (Allacker et al., 2013). This method, specific for the Belgian context, evaluates the environmental performance of building elements. Besides individual impact indicators, the MMG method allows to assess the environmental impact based 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 258 on an aggregated indicator, expressed in environmental costs (i.e. external costs caused by environmental impacts). Based on the MMG database of building elements, we developed a simple tool to assess the environmental impact of buildings and neighbourhoods. Using a limited number of input data, building elements can be combined to buildings, which in turn can be clustered to a neighbourhood model. In this paper, this tool is applied for the LCA calculations. 3. SIMULATION RESULTS Parametric model To illustrate the methodology, we defined a parametric neighbourhood model of rectangular urban blocks (Figure 4), linked to the EnergyPlus software. Although many geometric variants are possible, we focus on a medium-density urban block in order to evaluate the impact of shading interactions. This block consists of 15 m high buildings around a courtyard of 50m by 20m and is separated from other blocks by 10m wide streets. In this paper, only one side of the urban block with a north-south orientation is analysed (Figure 4). However, the other sides could be evaluated in a similar way. For the simulations, the building is subdivided in a grid of 25 dwellings of 100m². The glazed surfaces, which are assumed to be 25% of the façades, are approximated by two big windows, oriented respectively to the street and courtyard. Futhermore, building elements, fulfilling the low energy standard, are defined, using elements from the MMG database. Figure 4 Parametric neighbourhood model. Dynamic EDD calculations For the 25 dwellings the dynamic EDD are calculated using both the EnergyPlus and EPB approach. In order to analyse the impact of shading, each dwelling is simulated both in shaded and unshaded conditions. The results are shown in Figure 5 and expressed in percentage compared to a reference dwelling in unshaded conditions. When looking at the results for the unshaded conditions, the impact of internal and solar gains is clearly noticeable in the number of EDD. As an example, the reference dwelling is characterized by 1176 eq°d based on the EnergyPlus approach (1173 eq°d based on the EPB approach), which means a reduction of about 60% of the estimated heating requirements, compared to the standard DD method (2738 °d). Furthermore, higher EDD are calculated for the dwellings located under the roof and on the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 259 ground floor. This is a consequence of the higher heat transmission losses through the building skin, resulting in lower heat gain utilization. Regarding the comparison between the semi-dynamic and dynamic approach, similar results were found, except for the dwellings on the ground floor. In this case the simplified ground heat transfer calculation in EPB seems to overestimate the heat losses via the ground. Concerning the results in shaded conditions, an important increase of the EDD is noticed, compared to the unshaded conditions. For the EnergyPlus approach, this increase ranges from about 5% for the dwellings under the roof to about 35% for the dwellings on the ground floor. Similar results were found for the EPB approach but with bigger differences between the shaded and unshaded conditions. This is a direct consequence of the EPB approximation based on obstruction and overhang angles (Figure 3). EDD of the analysed dwellings in shaded and unshaded conditions, based on the Figure 5 EnergyPlus and EPB approach. The results are projected on the courtyard façade and expressed in percentage, compared to a reference dwelling in unshaded conditions (indicated by the black frame). Environmental impact calculations Based on the calculated EDD, the heating energy consumption of the 25 dwellings can be estimated and summed up over the whole building. This total heating energy consumption can then be used as input in the LCA calculation tool. The results of the environmental impact assessment are shown in Figure 6, with a distinction between the impact of building materials and heating energy use. Six models to estimate the energy consumption are compared, including the standard DD method (2738°d), the static EDD using the average of 1200 eq°d and the dynamic EDD based on EnergyPlus and EPB (both for unshaded and shaded conditions). Firstly, the results show a significant overestimation of the building environmental impact, when applying the standard DD method: the life cycle environmental costs are about 30% higher, compared to the EnergyPlus model for shaded conditions. Secondly, an average of 1200 eq°d seems a good approximation for the unshaded conditions. However, a difference of about 5% in life cycle environmental costs and about 15% in heating environmental cost was noticed between the shaded and unshaded conditions. It is expected that this difference could even be bigger, when using high-insulated passive building elements, increasing the internal utilization of solar gains. Therefore, a dynamic EDD calculation based on EPB or EnergyPlus is recommended, especially in dense neighbourhoods. Finally, only small life cycle impact differences are found between the EPB and EnergyPlus approach (from about 1% to 2% for respectively the shaded and unshaded conditions). The semi-dynamic calculation hence seems (for this case study) a good approximation for the more complex dynamic calculation. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 260 Building life cycle environmental cost calculated using 6 heating energy calculation Figure 6 methods: standard DD (2738°d), static EDD (1200 eq°d) and dynamic EDD based on EnergyPlus and EPB (for unshaded and shaded conditions). 4. CONCLUSIONS In this paper, a simplified approach is developed to estimate the heating energy consumption in the context of neighbourhood LCA. The existing EDD method is refined by including results from both semi-dynamic and dynamic solar gain calculations. Simulations of a medium-density urban block reveal substantial heating energy demand differences between shaded and unshaded conditions, stressing the importance of more dynamic solar gain calculations, especially for dense urban developments. Nevertheless, the static EDD (1200 eq d°) seems to be a good approximation, if supplemented with dynamic EDD for the most critical housing units. Furthermore, because of limited life cycle impact differences, compared to a full dynamic simulation, the semi-dynamic EDD, based on EPB, seems to be a valuable method for integration in an LCA tool. However, to avoid an overestimation of shading patterns, it is recommended to refine the EPB approximation by using a variable obstruction angle for different orientations. Concerning further research, we recommend validating the above conclusions by simulating more case studies and variations in urban block geometry. As the number of Equivalent Degree Days depends on the building insulation level, the influence of this parameter should be analysed in detail, especially for high-insulated passive buidlings. Finally, as the subdivision of each building block in constituting dwellings increases the simulation time, we recommend investigating whether simulations can be limited to a set of representative dwellings that could be used for interpolations. REFERENCES Allacker, K. (2010). Sustainable building, The development of an evaluation method (PhD dissertation). KU Leuven, Heverlee. Allacker, K., Debacker, W., Delem, L., De Nocker, L., De Troyer, F., Janssen, A., … Van Dessel, J. (2013). Environmental profile of building elements. Mechelen: OVAM. Retrieved from www.ovam.be/jahia/Jahia/pid/2594?lang=null 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 261 Diensten voor de programmatie van het wetenschapsbeleid. (1984a). Ontwerp en thermische uitrusting van gebouwen - deel 1 (translated title: Design and technical equipment of buildings - part 1). Brussels. Diensten voor de programmatie van het wetenschapsbeleid. (1984b). Ontwerp en thermische uitrusting van gebouwen - deel 2 (translated title: Design and technical equipment of buildings - part 2). Brussels. Flemish Government. Besluit van de Vlaamse Regering van 11 maart 2005 tot vaststelling van de eisen op het vlak van de energieprestaties en het binneklimaat van gebouwen - bijlage I (2005). Herfray, G. (2012). Contribution à l’évaluation des impacts environnementaux des quartiers (PhD dissertation). Mines ParisTech, Paris. Retrieved from http://tel.archivesouvertes.fr/docs/00/65/82/20/PDF/Herfray.pdf Trigaux, D., Allacker, K., & De Troyer, F. (2014). Model for the environmental impact assessment of neighbourhoods (pp. 103–114). doi:10.2495/EID140091 U.S Department of Energy. (n.d.). EnergyPlus Energy Simulation Software. Retrieved January 9, 2014, from http://apps1.eere.energy.gov/buildings/energyplus/ 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 262 Session 3A : Lessons from vernacular architecture PLEA2014: Day 1, Tuesday, December 16 16:05 - 17:45, Auditorium - Knowledge Consortium of Gujarat The ‘Teatinas’ of Lima: Energy Analysis and Possibilities of Contemporary Use Martin Wieser, PhD Pontificia Universidad Católica del Perú, PUCP ABSTRACT “Teatinas”, roof openings for zenithal ventilation and daylighting, were systematically used in buildings in the city of Lima and in most of the Peruvian coast from the mid-18th century to the end of the 19th century. The purpose of the study was to evaluate the thermal and lighting performance of the rooms where the “teatinas” were used and to assess an eventual use of similar resources in contemporary architecture. After defining the climate of the city, the buildings where they were installed and the “teatinas” themselves, the thermal and lighting conditions resulting from their use were calculated based on comparative measurements and simulations: air temperature and relative humidity, ventilation, lighting levels and glare. The results showed that the presence of a “teatina” in a room provides comfortable hygrothermal conditions, good air intake and circulation inside a room. As compared to conventional windows, the “teatina” allows for a more even distribution of daylight inside the space and more possibilities of avoiding glare. Finally, it is concluded that “teatinas”, consistent with the climate and daylighting conditions of the city of Lima, did fulfill the comfort requirements of homeowners of that period of time and are a valid reference and concrete alternative in the current search for comfortable spaces in energy-efficient buildings. INTRODUCTION The addition of teatinas on the roofs of Lima buildings became widespread after the 1746 earthquake. Despite the temperate climate of the Peruvian coast, its recurrent use was due to the need of providing natural ventilation and light to spaces which, due to the density and homogeneity of the urban grid, had little or no access to outside breezes and daylight. Photo 1: View of the Church of San Francisco from the Cathedral of Lima (circa 1860) with dozens of “teatinas” on the roofs. http://www.corbisimages.com/stock-photo/rightsmanaged/IH074568/church-of-san-francisco-lima?popup=1. Author is a professor in the Department of Architecture, Pontificia Universidad Católica del Perú PUCP, Lima, Perú. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 263 Photo 2: “Teatina” on a building located at the intersection between Camaná and Callao Streets, showing typical proportions and characteristics. Teatinas were systematically incorporated in buildings until the end of the 19th century. Afterwards, Neoclassical architecture, and later modern architecture, excluded them completely. The strategy of catching wind and capturing natural light through the roof has not been maintained or resumed in the Peruvian coast. This study examines the objective thermal and lighting conditions achieved by the presence of a teatina in a room, and its ability to provide thermal and lighting comfort in the spaces where it is present. This study has been more fully developed and submitted in a doctoral dissertation [1]. METHODOLOGY Given the inexistence of similar research studies in Peru, it was necessary to address previously certain aspects in order to be able to choose appropriate indicators for the energy assessment to be developed within this study. In addition to determining the definition, the variables and the evaluation of thermal and lighting comfort, it was necessary to identify the energy assessment tools and the most common environmental control strategies in architecture. The geographic, climatic and lighting characteristics of the city of Lima were identified and a prior study was made to relate the teatina to its historical context. Ten buildings containing a total of ninety-seven teatinas were surveyed and measured in order to identify typical construction characteristics and components –as related to the particular features of the urban grid– of the buildings and rooms where they were found. Afterwards, a ‘model room’ containing a ‘model teatina’ was defined. Jointly, both have typical characteristics insofar as orientation, layout, dimensions, form and finishes. These models were used in the simulations and some assessments were performed in settings with similar characteristics, since they had to be made on-site. The ‘model’ room is a quadrangular space, 5 meters long by 5 meters wide, with a height of 4.20 meters. For the lighting assessment, a ‘contemporary’ height of 3.00 meters was considered as an additional variant. The dimensions of the ‘model’ teatina were defined based on the sampling average, as well as the typical orientation of the opening (SSW). Figure 1. Dimensions and characteristics of the model “teatina”. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 264 The purpose of the energy assessment was to identify the thermal and lighting behavior of the ‘room/teatina’ combination as compared to its objective capacities of providing comfort to its occupants. Specifically, the following aspects were evaluated: Aspects to be Evaluated 1. Thermal Assessment Solar Radiation Ventilation 2. Lighting Assessment Lighting Level Glare Table 1. Energy Assessment Aspects Method Resource or Instrument Computer Simulation On Site Nomogram/AutoCAD Thermo-Hygrometer/Anemometer/Smoke Models Calculation Lux Meter/ Model Glare Index RESULTS Thermal Assessment In order to calculate the effect of solar radiation on the teatina, both that which passes directly through the opening and that which falls on its opaque structure were taken into account. In the first case, the equidistant solar projection was used to identify the months and hours when direct solar radiation entered through the opening, and a nomogram was superimposed on the former in order to obtain the incidental energy. Using AutoCAD, a graphic image of the entrance of direct solar radiation into the room was obtained. Figure 2: AutoCAD simulated rendering on the month when most direct solar radiation enters through the “teatina”: December The result was that the cumulative daily average of solar energy entering the room in the months of January and February ranges from 2.5 to 3.5 kWh, being higher at around 4:00 pm (approximately 0.9 kWh), when the eave presents the least obstruction and the sunrays are most perpendicular with respect to the opening. Figure 3: Nomogram superimposed on solar projection. Based on times of day, the amount of direct and diffuse solar energy through the “teatina” opening facing SSW is inferred. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 265 To predict the energy passing through the opaque surfaces of the teatina, thermal transmittance was calculated taking into account the ‘sun-air’ temperature (tsa). Given that the air temperature entering the teatina is equal to the external temperature, the following formula (Evans, 1980) was applied: Q = U • A • α • R • ro, [2] According to this calculation, the approximate amount of energy (Q) passing through the opaque surfaces of the teatina is close to 0.17 kWh. This value proved to be rather low, since the transmittance of the element, consisting of a packed mud layer about 10 cm thick, which also shares the characteristics of the rest of the covering, is equally low. From the values obtained, it can be stated that both the direct and diffuse solar radiation entering through the teatina opening and the heat passing through the opaque material are not enough to significantly raise the interior temperature of the room. The relatively low energy values and the ventilation provided by the component itself ensure continuous outflow of excess heat. To evaluate the ventilation in a room having a teatina, the energy conditions of the air and the air movement within the space were assessed. On-site testing in the Casa de la Riva during 72 consecutive hours comprised temperature and relative humidity measurements in a single room under two different conditions: with the teatina open and with the teatina closed. The results are depicted in a psychrometric chart showing summer comfort zone limits and comfort zone corrected with ventilation, as per Coch and Serra (1994) [3]. Figure 4. Psychrometric chart showing temperature and relative humidity data, considering vertical opening of the open (left) and closed “teatina” (right). Outdoor stations (in blue) and indoor stations (at human occupant level, in red) are represented. Testing was carried out during particularly hot and sunny summer days. Although on both days the external temperature was close to 29 °C at the warmest moments, it was observed that, with the teatina closed, indoor temperature reached almost 28 °C, while with the teatina open, indoor temperature stayed at around 26° C. These results are due to the presence of an indoor thermal mass that was cooled by the ventilation itself in the preceding hours, and to the ability of constantly expelling the air being heated inside. As to the air movement pattern within a room equipped with a teatina, site measurements were made in the Casa de la Riva and the Casona de San Marcos to determine speeds and direction, using in this case an anemometer and a smoke machine. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 266 Photo 3: Smoke tests to detect indoor air movements created by an open “teatina”. Casona de San Marcos, Lima. Figure 5: Typical wind movement patterns inside a room with a “teatina” and a door (left) or a window (right) facing the “teatina”. Wind directions are rather particular since, regardless of the location and the size of the exit opening, the incoming air direction is markedly vertical. Once it reaches the floor, it spreads horizontally in all directions, mainly away from the inflow direction. Incoming air mixes with existing air and finally leaves through the opening on the opposite side (Figure 5). Wind speed inside the room, at the lower section of the teatina, is between 30% and 60% of outdoor wind speed. Outdoor air descends rapidly once inside the room, though it is warmer than existing indoor air, because of the pressure differences created by the wind itself, since both, outdoor and indoor air, quickly mix due to the existence of continuous convective movements within the space. The resulting temperatures inside a room equipped with a teatina, in addition to wind presence and speed, ensure thermal comfort for occupants under typical hot summertime conditions in the city of Lima. Lighting Assessment Illuminance measurements were performed using scale models, considering a work plane at 80 cm from the floor, overcast sky and ‘model’ room and teatina characteristics. Table 2 shows average DF results (%, Daylight Factor) from various situations, followed by iso-DF curves (Figure 6). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 267 Photos 4 and 5. View of the Room A.01 model and detail of the roof and “teatina”. Room Model A.01 B.01 B.02 C.01 C.02 Table 2. Characteristics of Openings and the Room to Measure Lighting Levels Description DF (%) Average Teatina / roof covering with joists / ‘traditional’ roof height (4.20 m) 1.13% Teatina / smooth covering / ‘traditional’ roof height 1.61% Teatina / smooth covering / ‘conventional’ roof height (3.00 m) 1.94% High window / smooth covering / ‘traditional’ roof height 1.98% High window / smooth covering / ‘conventional’ roof height 2.37% Figure 6. Light distribution at the work plane in the various rooms. Considering that minimum values suggested for home environments range from 0.5% to 0.6% (Baker & Steemers, 2002) [4], it can be stated that the results obtained under any of these scenarios far exceed those requirements. To the extent that the lighting level values obtained from Europe overcast sky and in the north of the United States of America represent approximately one third of those reached in our tropical latitudes, it can be stated that the illuminance obtained from all of the results are suitable even for tasks requiring greater precision (classrooms, reading areas, etc.). Light distribution patterns show that in a room with a teatina, lighting is distributed more evenly as compared to a room with a conventional high window with the same dimensions. In addition, the greater height of the space is confirmed to provide more homogeneous illuminances, with higher values in the area facing the opening. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 268 To determine the glare, the Unified Glare Ratio (UGR) was applied, using the following formula (CIE, 1995, p. 117): UGR = 8 log [(0,25/Lb) • Σ (L2•ω/p2)] [5] Possible glare was determined by considering the ‘model’ room with teatina, a sight line at a height of 1.20 meters, five different points inside the room and four typical scenarios of sky brightness and lighting level (Figure 7). Projections were made considering both the ‘traditional’ and the ‘conventional’ (contemporary) height of the room. The results of the various scenarios are shown in Figures 11, 12, 13 and 14. Figure 7. Model Rooms and Measurement Points for Assessing Glare Figures 8 and 9. Results in a room with a ‘traditional’ height (4.20 m, B.01) and a conventional height (3.00 m, B.02), with “teatina”. Figures 10 and 11: Results in a room with a ‘traditional’ height (4.20 m, C.01) and a conventional height (3.00 m, C.02), with high window. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 269 Despite the relativity of the results given that both the height and the sight line direction are usually variable, it is confirmed that the overhead lighting – in this case provided by the teatina – performs better in preventing the visual discomfort associated with the glare phenomenon, as compared to conventional lateral natural light. CONCLUSIONS Teatinas, consistent with the climatic and light characteristics of the city of Lima, did meet the comfort requirements of the city inhabitants of the time, who incorporated them systematically in buildings for more than one hundred years since the mid-18th century. In addition to good wind uptake and distribution inside the rooms, their lighting performance is better than that of a conventional window, because teatinas distribute light more evenly within a space and are more likely to prevent glare. The thermal and lighting conditions achieved in rooms having teatinas, show that these elements continue to represent a valid design alternative to be considered. Based on the identification and comprehension of the phenomena associated with the energy performance of teatinas, alternatives and specific details may be proposed in order to improve their efficiency and adapt them to various contemporary applications. REFERENCES 1. Wieser, M. (2007). Las teatinas de Lima. Análisis energético-ambiental y perspectivas de uso contemporáneo. Doctoral Dissertation. Barcelona: Universidad Politécnica de Cataluña, Programa de Ámbitos de investigación en energías y medio ambiente en arquitectura. 2. Evans, M. (1980). Housing, Climate and Comfort. London: The Architectural Press. 3. Coch, H. & Serra, R. (1994). El disseny energètic a l’arquitectura. Barcelona: Edicions UPC. 4. Baker, N. & Steemers, K. (2002). Daylight Design of Buildings. London: James & James. 5. CIE. Comission Internationale de L'Eclairage. (1995). Technical Report. Discomfort Glare in Interior Lighting. CIE. 117. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 270 Influence of Greenery in Cooling the Urban Atmosphere and Surfaces in Compact Old Residential Building Blocks: A Building Morphology Approach Zheng Tan, MSc Edward Ng, PhD [The Chinese University of Hong Kong] tanya@cuhk.edu.hk [The Chinese University of Hong Kong] ABSTRACT Outdoor thermal environment at great extent determines the possibility and vitality of outdoor social activities and communication. The hot air and heated surfaces in tropical/subtropical areas during summer period would compromise the outdoor thermal environmentt in the city. This study aims to investigate the effect of street trees in cooling the air and surfaces in traditional high-dense residential area in Hong Kong. Sham Shui Po is a traditional residential district in Hong Kong with relatively regular building arrays oriented in different orientations. With simplified morphology, this paper aims to investigate the influence of building morphology and tree configuration on greenery cooling. In the simulation, the testing parameters including sky view factor (SVF) and orientation of street canyon, and the leaf area density (LAD) of the tree crowns. Based on the findings from parametric study, case study adopting the existing building morphology was also conducted to offer further understanding in greenery cooling. This study is expected to offer site-specific greening design suggestions for tropical/subtropical cities with compact building layouts. The simulation result shows that cooling efficiency varies with tree crown density. With the average of the LAD profile being 0.5 or above, substantial cooling in urban atmosphere and surfaces can be achieved in compact residential neighborhoods. And with denser tree crowns of LAD averaged at 1.0 or above, the mean radiant temperature would be reduced to a comfortable level of 33 degree in hot summer.The study also revealed with the high summer solar angle at noon, the cooling outcome under different orientations would be more homogenous with higher SVF; and for lower SVF condition, the cooling effect differs along streets sections and varies with street orientation. For subtropical cities, it is an effective measure to shade the street-crossing areas in high-dense neighborhoods to avoid overheat problem. INTRODUCTION As one of the world’s highest-density cities, Hong Kong suffers serious urban heat island (UHI) with a maximum intensity of 4 °C. To mitigate the UHI effect, several measures have been proposed, including greenery. and study by Ng et al. (2012) revealed that greening coverage larger than 33% would reduce the local thermal load, one of the two most important urban climatic evaluation factors, by 1 class in the Urban Climatic Map of Hong Kong. Nevertheless, the greening coverage ratio is much lower than 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 271 the recommended value in many densely built urban areas. The situation is particularly serious in aging residential districts with compact building layouts and large population density. Greening design should be applied in residential neighborhoods to provide better living condition (Givoni, 1998). As a planning reference, site-specific greenery design strategies should be established upon clarifying the local thermal condition in the neighborhoods. Thus, the current study investigates the impact of urban tree design on the enhancement ofoutdoor thermal enironment in compact residential neighborhoods. From previous studies, it has been pointed out that sky view factor (SVF) is a key parameter to be considered in morphology-approach thermal study. Study of Mills (1996) demonstrated that minimizing the relative exposure and maximizing the SVF improve thermal performance of building group design in tropical and subtropical areas. Giridharan et al. (2004) studied the UHI intensity and mitigation in Hong Kong, and it was found that 1% reduction on sky view factor (SVF) reduced daytime UHI by 1-4%. Correlation between intra-urban temperature difference and areal means of SVF has been studied by Unger (2009). It was summarized that ground level SVF was recommended for individual point/site investigation, while areal averaged SVF at pedestrian height obtained a better result in predicting the temperature deviation between sites. An inverse proportion was found between the areal averaged SVF and air temperature elevation in urban areas of Hong Kong (Chen et al., 2012), and in high-density city, the influence of SVF as a quantifier was different in various density. Giridharan et al. (2008) analyzed the vegetation effect on UHI mitigation under different SVF settings, but the study only covered relatively small range of low SVF values. To investigate the vegetation effect in the context of building morphology, urban canyon as the basic morphology unit has been selected as setting in previous studies. Study by Shashua-Bar et al. (2006) revealed that with the tree coverage being 64% and the aspect ratio ranging from 0.2 to 0.6, the cooling effect of the tree achieved 2.3K. Ali-Toudert et al. (2007) studied the effect of configuration and vegetation cooling in urban canyon by simulation, and the results showed that compared to air temperature, radiant environment was more sensitive to canyon geometry. The vegetation created a decrease on PET value of 22K within the tree canopy in canyon with aspect ratio of 2, and the value was even higher as 24K when the aspect ratio was 1. Shashua-Bar et al. (2012) studied the passive cooling efficient of green urban canyon with numerical modeling, and the result showed the greenery cooling effect was highly related to the canopy coverage. Sham Shui Po in Hong Kong is a traditional high-density residential district. This district is also a climate sensitive waterfront area with summer prevailing wind from the sea. In the study, the building morphology of SSP will be adopted in parametric studies with ENVI-met simulation to investigate the influence of tree crown configuration on the cooling efficiency under the context of different sky view factor and street orientations. Based on the finding from parametric study, the greenery design scheme will be tested with the existing morphology of the district as case study. Data from small scale site survey will be used to validate the microclimate and vegetation model in ENVI-met with Hong Kong’s condition. Figure 1 Locations of the selected sites showing on the Urban Climatic Map of Hong Kong (Ng et al., 2009) with the building layout (left) and street views (right). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 272 METHODOLOGY Site Survey and Model validation This study considered approximately 1350000 m2 traditional residential areas with compact building block arrays in the SSP district. In the selected site, the total building volume was 16,722,000 m3 and the average building height was 32 m with a standard deviation of 15. The site is located in the waterfront areas, and wind from the sea dominates the district in the summer. Several building blocks are parallel to the summer sea breeze, whereas the others are tilted at approximately 45° (Fig.1). Figure 2 Building layout of the survey site and locations of the testing trees. The fisheye image of tree13 and the estimated LAD are also presented. A small-scale site survey was conducted to determine the influence of tree crown configuration on the microclimate under tree canopies. The measuring site was a residential area with relatively low traffic; and the measuring period was between 13:00–14:00 on a clear sunny day in early July. The measured data were also used to validate the vegetation model of the ENVI-met program. Fig. 2 shows the building layout of the survey site and the locations of three testing trees. The profile of leaf area density (LAD) of each testing tree was estimated based on fisheye images and the vertical configuration of the tree crown; and was inputted to the plant database in ENVI-met model for calculation. Thermal indicators were measured under the canopy of testing trees and a nearby (within 20 m distance) reference point exposed to sunlight using a TESTO measuring instrument, FLIR thermography camera and black globe thermometer. Table 1 compares the measured and simulated values on air temperature, surface temperature, and means radiant temperatures (Tmrt). Table 1. Measuring result from the pilot study Air Temperature 1. Ficus microcarpa in shade exposed point 32.8 / 33.6 33.3 / 33.8 2. Melaleuca cajuputi in shade exposed point 32.9 / 32.9 33.5 / 33.2 3. Spathodea campanulata in shade exposed point 33.1 / 32.9 33.5 / 33.0 Surf Temperature 33.4 / 33.8 62.5 / 58.5 36.4 / 36.9 61.0 / 57.8 43.9 / 46.0 61.0 / 57.3 Tmrt 35.3 / 37.4 68.1 / 70.9 37.5 / 38.9 66.4 / 70.7 48.0 / 48.7 66.5 / 70.7 (Figures in black are the measured data and figures in grey are the calculated results. Tmrt calculated based on the measured data with the method of Thorsson et al. (2007), and by the ENVI-met model, respectively.) Table 1 indicates that the model successfully reproduced the microclimate under tree canopies with various LAD and the thermal condition of reference points under the effect of urban morphology. The measured ground-surface temperatures were approximately 4° higher than the calculated results in the exposed points; this result can be attributed to the thermal property of the ground material. The surface material of the survey site is dark asphalt which would present relatively higher surface temperatur. Based on the data, it can be concluded that the ENVI-met model is a reliable tool for green design in urban areas and microclimate study under noontime high solar angle during summer in Hong Kong. PARAMETRIC STUDY OF GREENING DESIGN Based on the original building layout, the building morphology of the SSP site was simplified into 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 273 aligned arrays and tilted blocks with uniform building height for the parametric study (fig.3). Urban trees that are 20 m high with averagely dense (average LAD = 0.15 m2/ m3 within the tree crown) and dense crowns (average LAD = 0.95 m2/ m3 within the tree crown) were planted in the aligned and tilted streets as shown in Fig. 3. The aligned streets were oriented at north–south and east–west to test the influence of orientations, whereas the building heights were set at 30 and 60 m to simulate low and high SVF conditions, respectively. These two values were selected because 30 m is the average building height in the SSP site and 60 m is the weighted average building height in Hong Kong as a whole. Therefore, greening design schemes are studied under SVFs ranging from 0.2 to 0.3 and 0.4 to ¬0.5. Cubes were set with uniform height at the model boundaries to simulate the effect of the surrounding morphology. For simulation setting, 250*250*30 grids version was selected for model domain in the case study with grid sizes being 6m. The study focused on thermal comfort at the pedestrian level; hence, dense (four layers) vertical telescoping grids were set in the first 2 m to improve the accuracy of the results. Input wind data including wind directions and pedestrian wind velocity can be extrapolated down from the 500m height wind data of the SSP district and the wind profile power law expression (1). u(z)/uref = (z/zref)α (1) u(z) and uref represent the mean wind speed at a certain height z and at a suitable reference height and α is the terrain roughness at 0.15. The input wind speed was adjusted, and the wind speed at the aligned street was maintained at 2.7 m/s–3.0 m/s at a height of 2 m. The other meteorological input data are listed in Table 2, and these data were based on the averages obtained during hot summer days in August 2012 (data source: Hong Kong Observatory, SSP station). The total simulation duration was 8 h (from 6:00 to 14:00). Figure 3 Simplified building morphology applied and the greening design scheme and two ranges of SVF values would be tested in the parametrict stuy. Table2. Input data used in the simulation Initial Temperature atmosphere [K] Relative Humidity 2m [%] 303 Heat Transmission Walls [W/m2K] Heat Transmission roof [W/m2K] 2 Albedo Walls Albedo Roofs 0.2 70 2 0.3 Data Analysis. This study investigates the effect of urban trees on the enhancement of outdoor thermal comfort under different morphologies. hence, the most critical daytime period in summer in subtropical regions was selected for analysis, which would be noontime and early afternoon when the solar radiation is intense and the surfaces have been heated up. In this study, simulation result in 13:00 (the early afternoon) is chosen. The efficiency of urban trees in differnet morphology settings is evaluated on the aspects of atmosphere and surfaces cooling. As a more comprehensive indicator for environmental impact for human comfort, the mean radiant temperature will also be investigated.  0.4-0.5 SVF (building height 30m) In the first scheme, building height was set to 30 m. The cooling effect on the air and surface temperatures were also examined under SVFs ranging from 0.4 to 0.5. Fig. 4 depicts the simulation results for the two oriented base cases without greenery. Fig. 5 displays the reduction in the air and surface temperatures of trees with the variation in tree crown density when the aligned streets were 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 274 oriented at east–west. The maximum magnitude of air cooling was 1.8 K in the middle section of the shaded area at a height of 1.5 m under dense tree crowns. The edge of the shaded section was cooled by 1.2 K and was reduced to 0.8 K approximately one block away in the downwind area. The greenery effect was slight on areas with tilted streets (0.5K) because wind speed is reduced. Given cases with averagely dense tree crowns, the cooling effect was roughly halved. With respect to surface cooling magnitude, the shadowed areas on the ground were limited to the edge because the buildings were relatively short and because SVFs were high. Therefore, surface temperature distribution is quite homogeneous (40 and 45 K for aligned and tilted streets, respectively). The surface temperature was lowered by 5 K and 9 K in aligned and 9K tilted streets, respectively, given averagely dense tree crowns. Furthermore, the cooling magnitudes by denser tree crowns were 11 K and 17 K, respectively (Fig. 5). With the average dense tree crowns the mean radiant temperature under the tree canopy was about 49K and with dense tree crown the value was cooled down to 33K. The efficiencies of the dense tree crowns (average LAD 1.0) were basically twice that of the averagely dense tree crowns (average LAD 0.2) in terms of cooling the air and surface and reducing the mean radiant temperatures. When the aligned streets were oriented to north–south, the reduction in air temperature by the street trees was basically similar to that in the east–west orientation. Moreover, the decrease in surface temperature was approximately 1 K lower in the east–west orientation for both aligned and tilted streets. Figure 4 simulation results on on air temperature, surface temperature, and Tmrt for basecases of two orientations under high SVF (ranging at 0.4-0.5) Figure 5 cooling magnitude of street trees compared to base case in 13:00 for high SVF scenario (aligned streets E-W oriented). (1) air cooling by averaged and dense tree crowns; (2) surface cooling by averaged and dense tree crowns; (3) Tmrt reduction by averaged and dense tree crowns 0.2-0.3 SVF (building height 60m) In the second scheme, building height was set at 60 m. The SVF values of the streets at ground level ranged from 0.4 to 0.5. Fig. 6 shows the simulation result for the base case. The air temperature of the street at pedestrian level was lower and the shaded area of ground surface was larger than those in the 30 m scheme. When the aligned streets were set to be east-west oriented, it can be seen that the cooling magnitude on air temperature was smaller than the larger SVF case, as depicted in Fig. 7, and with the shading of trees, the air temperature would be cooled down to similar level in the streets with SVF of 0.2-0.3 and 0.4-0.5. On the other hand, for surface temperature and Tmrt the cooling magnitudes were almost the same for high and low SVF scenarios; the cooling effect presented more various levels in streets with  30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 275 low SVF. The influence of crown configuration on cooling efficiency showed similar pattern, i.e. the denser tree crowns were more effective and for tilted streets, the thermal effect of trees was minor in air cooling but significant in cutting down the surface temperature and Tmrt. Figure 6 simulation results on air temperature, surface temperature, and Tmrt for basecases of two orientations under low SVF (ranging at 0.4-0.5) Figure 7 cooling magnitude of street trees compared to base case in 13:00 for low SVF scenario (aligned streets E-W oriented) (1) air cooling by averaged and dense tree crowns; (2) surface cooling by averaged and dense tree crowns; (3) Tmrt reduction by averaged and dense tree crowns Figure 8 cooling magnitude of street trees compared to base case in 13:00 for low SVF scenario. (aligned streets N-S oriented) (1) air cooling by averaged and dense tree crowns; (2) surface cooling by averaged and dense tree crowns; (3) Tmrt reduction by averaged and dense tree crowns Hong Kong is located south of the Tropic of Cancer. The sun casts shadows in two directions on aligned building arrays in the early afternoon during summer before, during, and after the solstice. The shadow on the ground is enlarged along with diversity in surface temperature and the radiative environment under low SVF. The shadowed area of tilted blocks was larger in the northwest–southeast streets than in the northeast–southwest streets. Trees in streets with low SVF will shade the ground surface and homogenize the temperature. The simulation results indicated that surface of the aligned streets oriented at north–south was less heated by 5 K at the center of the road than that oriented east– west. The difference was reduced to less than 3 K in streets with averagely dense tree. The surface temperature of aligned streets with dense trees was cooled to 30 K, and the surface temperature of the tilted streets was 28 K in either orientation. In addition, greenery cools the several “hot spots” generated at street crossing areas, as a result of building geometry and solar angle in the summer. The tree shading is most needed in these hot spots (fig. 7 and fig.8). For the mean radiant temperature, the cooling magnitude in the vegetated area was relatively more diverse under low SVF scenario, especially for north-south oriented streets. Based on the study of Forwood et al. (2000) in Sydney, it was revealed that the comfortable range for Tmrt was 24 to 30 degree. The simulation result showed with the averaged 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 276 LAD at 1.0 level, the Tmrt value under canopy was cut down to 33 degree for both high and low SVF environment, greatly improve the outdoor thermal comfort in the compact residential neighborhoods in subtropical Hong Kong. CASE STUDY OF GREENING DESIGN The green design scheme was tested using the existing building layout in the SSP area based on the parametric study results to investigate the thermal influence of trees on the urban environment further. The “aligned streets” in the site were oriented at northeast–southwest and were parallel to the direction of the local wind, whereas the “tilted streets” were oriented at east–west. Trees with relatively dense crowns (average LAD = 0.5) were planted in these streets (fig. 9). The simulation settings remained similar to those of the parametric study. However, the inputted wind speed data at a height of 10 m was adjusted to ensure that the wind velocity at pedestrian level was consistent with that in the parametric study. Fig. 9 shows the building layout in the SSP site under the greenery design scheme, as well as the SVF and wind speed distribution in the simulation of the base case. Fig. 10 presents the magnitude by which air and surface temperatures are lowered by the street trees. The distribution of SVF and wind speed in the base case was complicated by the irregular building morphology in actual urban environments; and two two parameters are closely related to the air and surface temperature in the area. The case study further confirms the findings from the parametric studies. The applied LAD in the case study was moderate compared with that adopted in the parametric study; therefore, the cooling magnitude was also at moderate level. With the LAD at about 0.5 levels, the trees reduced the Tmrt by about 28-30 degree and the value under tree canopy was about 34 degreee in the SSP neighborhoods. Figure 9 (1) greenery test scheme in case study with existing morphology in SSP site; (2) SVF distribution in the basecase; (3) wind speed distribution in the basecase Figure 10 simulation results for case study of SSP site: (1) basecase air temperature distribution; (2) magnitude of air cooling by trees; (3) basecase surface temperature distribution; (4) magnitude of surface cooling by trees; (5) reduction of Tmrt by trees SUMMARY To address the problem of greening design for improvement of the outdoor thermal environment in high-density cities, parametric study and case study were conducted on compact residential building blocks in Hong Kong. The effects of tree crown configuration under varying SVFs and street orientations were tested by simulation with the three-dimensional microclimate model and ENVI-met. In this study, the cooling efficiency of urban trees in residential neighborhoods was evaluated based on 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 277 cooling the air and ground surfaces, as well as mean radiant temperature reduction. Air cooling is dependent on the air velocity in the street, whereas the latter two indicators are morphology-related. Findings of the study can be summarized as below: 1. Influence on urban thermal environment varies with the crown density of urban trees. Tree crowns with an average LAD of above 0.5 are recommended for improving the outdoor thermal environment of high-density cities. In compact residential neighborhoods with street trees of this dense, cooling effect in air temperature and surface temperature would achieve 1K and 10K, respectively. With dense tree crowns of LAD averaged at about 1.0 level, the mean radiant temperature would be cut down to 33 degree under the canopy; given the hot summer in subtropical Hong Kong, it can be considered the urban trees provide a comfort microclimate for residents and greatly improve the outdoor thermal environment. 2. With respect to the influence of building geometry, the absolute cooling magnitudes of air and ground-surface temperatures were smaller in low SVF (0.2–0.3) than in high SVF (0.4–0.5) scenarios. The ground surface is exposed at varying levels in subtropical cities because of the high solar angle at noontime in the summer. Moreover, diversity in radiative environment would be enlarged as a result of builfing geometry. Street trees with certain LAD level could cool down the ground surface and even the thermal differences. And it is more critical to shade the street-crossing areas with tall dense trees to avoid overheat in high-dense neighborhoods with low SVF. Nonetheless, this study has several limitations. First, additional environmental indicators should be included to evaluate the thermal effect of urban trees comprehensively. Second, the relationship between greening design and the thermal comfort of residents in the compact urban environment of tropical/subtropical cities should be further studied. REFERENCES Ali-Toudert, F., and H. Mayer. 2007. Thermal comfort in an east–west oriented street canyon in Freiburg (Germany) under hot summer conditions. Theoretical and Applied Climatology, 87(1-4), 223-237. Chen, L., Ng, E., An, X., Ren, C., Lee, M., Wang, U., and Z. He. 2012. Sky View Factor Analysis of Street Canyons and Its Implications for Daytime Intra-Urban Air Temperature Differentials in HighRise, High-Density Urban Areas of Hong Kong: a GIS-Based Simulation Approach. International Journal of Climatology, 32(1): 121-136. Forwood, B., Hayman, S., and S. Tadepalli. 2000. Thermal comfort in urban open spaces. In: de Dear RJ, Kalma JD, Oke TR, Auliciems A, editors. Biometeorology and Urban Climatology at the turn of the millennium: A selection of papers from the International Conference on Urban Climatology and the International Congress on Biometeorology (ICB-ICUC’99), vol. WCASP-50, WMO/TD-No.1026, Geneva: World Meteorological Organization, 2001. Givoni, B. 1998. Climate considerations in building and urban design. John Wiley & Sons. Giridharan, R., Ganesan, S., and S. S. Y. Lau. 2004. Daytime Urban Heat Island Effect in High-Rise and High-Density Residential Developments in Hong Kong. Energy and Buildings, 36(6): 525-534. Giridharan, R., Lau, S. S. Y., Ganesan, S., and B. Givoni. 2008. Lowering the Outdoor Temperature in High-Rise High-Density Residential Developments of Coastal Hong Kong: The Vegetation Influence. Building and Environment, 43(10): 1583-1595. Mills, G. 1997. The Radiative Effects of Building Groups on Single Structures. Energy and Buildings, 25(1): 51-61. Ng, E., Chen, L., Wang, Y., and C. Yuan. 2012. A study on the cooling effects of greening in a highdensity city: an experience from Hong Kong. Building and Environment, 47: 256-271. Unger, J. 2009. Connection between urban heat island and sky view factor approximated by a software tool on a 3D urban database. International Journal of Environment and Pollution, 36(1): 59-80. Shashua-Bar, L., Tsiros, I. X., and M. Hoffman. 2012. Passive cooling design options to ameliorate thermal comfort in urban streets of a Mediterranean climate (Athens) under hot summer conditions. Building and Environment, 57, 110-119. Shashua-Bar, L., Hoffman, M. E., and Y. Tzamir. 2006. Integrated thermal effects of generic built forms and vegetation on the UCL microclimate. Building and Environment, 41(3), 343-354. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 278 Design Optimization of Vernacular Building in Warm and Humid Climate of North-East India Manoj Kumar Singh, PhD Sadhan Mahapatra, M.Tech. Jacqu es Teller, PhD Local Environment Management and Analysis (LEMA), Université de Liège, Chemin des Chevreuils, 1 - 4000 Liège, Belgium; Integrated Research and Action for Development (IRADe), C-80, Shivalik, Malviya nagar, New Delhi 110017, India Department of Energy, Tezpur University, Tezpur, 784028, Assam, India Faculté des Sciences Appliquées, Department ArGEnCo, Local Environment Management and Analysis (LEMA), Université de Liège, Chemin des Chevreuils, 1 - 4000 Liège, Belgium ABSTRACT Vernacular buildings are evolved through trial and error method over the period of time. These buildings are constructed more on ‘design-based approach’ suited to a particular climatic condition and socio-cultural setup rather than emphasizing technological solutions or prescriptive requirements. However, in recent times, due to quest for better thermal comfort, energy consumption is increasing in these naturally ventilated buildings. So, it is an urgent need to analyse the present level of thermal comfort and the occupant’s expectation in these buildings. In case of design based approach, passive solar design, ventilation, insulation on the building envelope, shading and glazing area, proper orientation of buildings etc. are the key parameters for optimization process. In this study, a vernacular building of warm and humid climatic zone of North-East India is considered. Thermal performance study has been done by carrying out year long measurements of environmental parameters both at indoor and outdoor of the building along with thermal comfort survey and interaction with the occupants. The comfort and neutral temperature for different seasons of the year have been evaluated in the study. Solar energy modular simulation tool TRNSYS 17 is used to carry out simulations of the building. Building 3D model is generated in TRNSYS and design optimization has been done by carrying out parametric simulations for different scenario such as wall thermo-physical properties and thickness, window to wall ratio, glazing type, orientation, shading, infiltration, ventilation and internal load. The objective of the simulations is to improve the indoor thermal environment close to the comfort temperatures obtained during comfort survey. Indoor temperature profile of the optimized building shows significant reduction in number of discomfort hours compared to the base case. INTRODUCTION Vernacular buildings are the structures that use the bioclimatic concepts and locally available building material to a large extent (Singh et al., 2011b). This provides an edge to vernacular buildings to withstand with the local climate constraints through adaptation. However, vernacular buildings are mainly constructed on design based approach and evolve over the period of time through trial and error method (Ruiz and Romero, 2011; Singh et al., 2010a; Singh et al., 2010b; Singh et al., 2011b). These buildings attract attention of researchers because these structures represent an excellent harmony between environment, available building material and resources, socio-economic status and sociocultural need of occupants, climate pattern and comfort, thus putting forth a unique example of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 279 sustainability (Kulkarni et al., 2011; Orehounig and Mahdavi, 2011; Singh et al., 2011b). In modern times, with the changing lifestyles, comfort standards and energy needs are increasing. Hence, it is important to look at the energy saving potential and sustainability presented by bioclimatic aspects of vernacular buildings (Singh et al., 2010a; Singh et al., 2010b). Vernacular buildings of North-East India are naturally ventilated and constructed using locally available building materials. Shape and form of these buildings are evolved over time to meet the socio-cultural and day to day requirements (Singh et al., 2010a; Singh et al., 2011b). These buildings are still favoured by people of the region and are still being widely constructed (Singh et al., 2010a; Singh et al., 2011b). However, in the present context of increasing comfort requirement and energy efficiency regulation and guidelines, it is an urgent necessity to carry out the thermal performance study of these vernacular buildings (Auliciems, 1981; Brager and Dear de, 1998). In this study, a vernacular house located in warm and humid climate (Tezpur, India) is considered. Selected house is modelled in TRNSYS 17 (Transient System Simulation), most widely used solar energy modular program to carry out the thermal performance study of buildings (Bansal and Bhandari, 1996; Beckman, 1994; Datta et al., 2001). It is a very powerful solar modeling and simulation tool (Bansal and Bhandari, 1996; Orehouning and Mahdavi, 2011). In this study, multi-zone building is integrated to simulation studio by Type 56. Number of studies has been carried out in different parts of the world on the thermal performance of modern buildings by using TRNSYS. However, no study has been done on the design aspect of the vernacular houses of North-East India. Thermal simulations are carried out to see the effect of different design features on indoor temperature. Based on the analysis of simulation data suggestions are made to improve the indoor temperature variation inside the house over the year. Table1 Properties of the selected vernacular house House details Building type Climatic zone Build up area (m2) Building material Ventilation type Temperature range Layout and orientation Relative humidity (%) Altitude (m) Elevation of building Properties Vernacular house (Local common name : Assam type) Warm and humid 94 Brick, cement, sand, plywood, asbestos sheet/wood, galvanized tin sheet Naturally ventilated Summer temperature : Maximum : 30 – 35 0C; Minimum: 22 – 27 0C Winter temperature : Maximum : 25 – 30 0C; Minimum: 20 – 15 0C Open layout, NW-SE 75 - 90 48 4.8 m (floor to eaves 3.8m and ceiling to roof top 1m) North-East India is classified into three climatic zones (warm and humid, cool and humid and cold and cloudy) and vernacular houses in each climatic zone possess distinct climatic responsive features (Singh et al., 2007). Table 1 and 2 present the specific details and building materials of the selected vernacular house in warm and humid climatic zone. Figure 1 presents the layout of the selected vernacular house in warm and humid climatic zone (numbers in the Figure 1 represent the zone number). It can be observed from Figure 1 that openings (windows and ventilators) are evenly distributed on the facade of the house. It is found that windows of the zone 2 and 3 are made up of wood with single glazing (30% of total window area). Ventilators are made up of wood with single glazing (35% of total ventilator area). It is found from the thermal performance study of the selected house that the maximum indoor temperature swing is 10 0C (Singh et al., 2010a). It is also found from thermal performance analysis that the house is more comfortable in pre-summer and summer season compared to pre-winter and winter season (Singh et al., 2009; Singh et al., 2010a). Figure 2 represent the 3D drawing created in Trnsys3D and Google SketchUp. In 3D model, window on exterior façade are constructed by adding all the windows on exterior wall of same zone (keeping the area same) to reduce the complexity of the model. Since the selected vernacular house is naturally ventilated so auxiliary heating, cooling and mechanical ventilation are kept off for all simulations. In case of naturally ventilated building indoor air temperature variation is the most important parameter so entire study is focussed on analysis of indoor temperature variation in different zones of the house. In this house zone 2 and zone 3 are occupied for maximum duration of time so due consideration is given to the temperature profile of these two zones. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 280 Indoor temperature variation of base case is also compared with the data collected during thermal monitoring work carried out in 2008 to judge the accuracy of the developed model. Figure 1 Layout of the house with different zones Figure 2 3D Drawing of the vernacular house Figure 3 Methodology of the study METHODOLOGY Vernacular house in warm and humid climate is locally called Assam type in this region (Singh et al., 2011b). This build form is very popular and it has wide acceptance because it fits well into socialcultural setup, economical to construct, easy to maintain and above all meets the climatic constraints (Singh et al., 2011b). In modern times, with changing life style, demand for better comfort and energy use regulations is forcing occupants to explore different options that modify the indoor environment. Thus, it becomes necessary to study the design aspects of vernacular house for its energy efficiency. This enables us to understand the thermal behavior of the vernacular house with respect to design modification required in the building design. Figure 3 represents the methodology followed to carry out the present study. Parametric simulation studies are carried out by using TRNSYS and MATLAB simulation tool is also used to process the simulation data. Three different 3D models like without false ceiling, false ceiling with common attic space and false ceiling with individual attic space are made in 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 281 Google SketchUp. All these models are used to carry out the simulation for all eight possible building orientations. It has been tried to find out the best model by analyzing the simulation data which provides close result to thermal comfort survey data (Singh et al., 2010a; Singh et al., 2011a). This house is being selected as ‘base case’ for further simulations with design modifications. Figure 3 also shows the seven cases for which the base case model simulation is carried out with design modifications. Table 3 presents the specific details of seven cases for which the simulations are carried out. The different scenario for the simulation are (i) applying different insulation to the walls of base case house (ii) replacing windows and ventilators with double glazing of base case (iii) increase and decrease the windows and ventilators area to base case (Table 2) (iv) replace the increased windows and ventilators area with double glazing to base case. The house considered for this study is naturally ventilated so the zone temperature is considered as the main output parameter along with zone heat gain due to infiltration. The infiltration is kept at 3 ACH (air changes per hour) for this naturally ventilated house for all simulation cases. Table 2 Input parameters for base case building Thermal conductivity (W/m-K) 0.721 0.811 61.06 0.245 0.17 0.144 0.33 0.16 Building materials Plaster Brick Tin sheet Asbestos sheet Wood(window, doors and ventilators) Foam insulation (10cm) U1 Wooden wool (10 cm) U2 Mineral wool (10 cm) U3 Density (kg/m3) 1762 1820 7520 1520 900 1.4 0.025 0.90 Specific heat (kJ/kg-K) 0.84 0.88 0.50 0.84 1.7 10 400 80 Table 3 Wall construction and thermo-physical properties of materials with thickness Case Wall configuration External wall Base case+single glazing window with wooden frame Base case + double glazing window Base case + wall with insulation 1 Base case + wall with insulation 2 Base case + wall with insulation 3 Base case + decreased window area Base case + increased window area Base case + increased window area + double glazing Internal wall Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) Plaster (1.5cm) + brick (11cm) + plaster (1.5) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) + insulation U1 (10cm) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) + insulation U2 (10cm) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) + insulation U3 (10cm) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) Plaster (1.5cm) + brick (23cm) + plaster (1.5cm) Plaster (1.5cm) + brick (11cm) + plaster(1.5) Plaster (1.5cm) + brick (11cm) + plaster (1.5) + insulation U1 (10cm) Plaster (1.5cm) + brick (11cm) + plaster (1.5) + insulation U2 (10cm) Plaster (1.5cm) + brick (11cm) + plaster (1.5) + insulation U3 (10cm) Plaster (1.5cm) + brick (11cm)+ plaster (1.5) Plaster (1.5cm) + brick (11cm) + plaster (1.5) Plaster (1.5cm) +brick (23cm) + plaster (1.5cm) Plaster (1.5cm) + brick (11cm) + plaster(1.5) Over all heat transfer coefficient (W/m2 K) External Internal wall wall 2.103 3.056 2.103 3.056 0.316 0.331 0.568 0.621 0.343 0.304 2.103 3.056 2.103 3.056 2.103 3.056 BUILDING MODEL GENERATION AND SIMULATION TRNSYS 17 simulation tool is used to model the selected vernacular house to study the energy flow in the house as well as in between the zones of the house. TRNSYS is a quasi-state simulation tool (Bansal and Bhandari, 1996; Beckman, 1994). Its modular structure provides a tremendous flexibility and facility to users to customise the generated model (Singh et al., 2009; Beckman et al., 1994). It runs through hourly values but user can reduce the time step according to the system requirement (Beckman et al., 1994; Singh et al., 2011b). A systematic approach has been adopted to develop a multi-zone model 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 282 of the selected vernacular house in TRNSYS using TYPE 56. Type 56 in the simulation tool also provides a provision to give 3D geometric surface information as input for detailed radiation calculation. This increases the accuracy of the calculations. Using Trnsys3D and Google SketchUp 3D model, vernacular house specifying zones is created. Once the building is defined properly the building variable needs to be updated and linked to TYPE 56. RESULTS AND DISCUSSION The vernacular house in warm and humid zone is generally constructed in three different pattern such as (a) without false ceiling (room air is in direct contact with roof), (b) with ceiling and attic space is common (most common type of construction) and (c) with ceiling but individual attic space above each room (walls of the houses are load bearing). 3D models for above three types of construction of vernacular house is generated in Google SketchUp and TRNBuild and imported to simulation studio as Type 56 (multi zone building). The simulations are carried out for all 8 possible orientations (00 due North, 450, 900, 1350, 1800, 2250, 2700, 3150). The build up area of the selected vernacular house is 94 m2. Figure 1 presents the configuration (numbers in the Figure 1 represent the zone number) of the house viz. zone 1: veranda, zone 2: living room/bed room, zone 3: bed room, zone 4: kitchen and zone 5: store room. Based on the functionality and specific requirements of rooms in a vernacular house, zone numbering is done in the selected vernacular house. Subsequently analysis of the simulation data has been carried out keeping in mind the requirement of the zones. In this study, due consideration has given to the temperature profile of zone 2 (living room/bed room) and zone 3 (bed room). Detailed thermal comfort model is applied by defining the geo-position. Thermal comfort, operative temperature and mean radiant temperature are calculated at the centre of each zone to analyse the thermal comfort. Since the selected house is naturally ventilated, heat gain/loss due to infiltration over 24 hours for entire year is also studied. Simulation data of zone 2 and 3 of the house with no ceiling, ceiling with common attic and ceiling with individual attic is analysed. It is found that the house with no ceiling show higher daily indoor temperature swing. Also roof of all the house is made up of galvanised tin sheet, so it gain and loose heat quickly. This happens because indoor air in this house is in direct contact with roof and in day time it gains heat inside quickly and loose heat quickly in the night time. It is also observed that temperature fluctuation remains high for most part of the year except for the last three months when daily temperature swing is less. This can be explained by observing the local wind velocity profile. Low wind velocity greatly affects the infiltration and natural ventilation. Again looking at the indoor temperature swing of house having ceiling with common attic and ceiling with individual attic, it can be found that difference is very less. Hence, it can be concluded that the house with individual ceiling shows slightly better thermal performance but considering the complexity of construction and safety this can be neglected (Assam lies in seismic zone V). Hence the house having ceiling with common attic has been considered for further study. Table 4 Indoor temperature range from field measurements and simulation of zone 2 Climatic zone (place) Warm and humid (Tezpur) Month January April July October Range of indoor temperature (°C) Field measurement Simulation and comfort survey 13 - 23 13 - 22 22 - 28 22 - 29 27 - 34 27 - 33 22 - 28 23 - 29 Figure 4 and 5 presents the daily maximum and minimum temperature variation for zone 2 of the simulated vernacular house for all orientations. Similar kind profiles are also obtained for zone 3 of the selected house. It is observed from these figures that in pre-summer and summer the orientations has no effect on the indoor maximum and minimum temperature profile of the selected vernacular house. This is because in naturally ventilated vernacular house infiltration is very high. However, in pre-winter and winter months, it is found that that zone 2 and 3 of the house show some variation in the indoor temperature depending on the orientation of the house. The effect is more visible in the minimum temperature profile of zone 2 and 3. The reason of this behaviour can be attributed to the change in solar 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 283 altitude angle and exposure of different zone of the house at different orientation. The orientation of real house at Tezpur is 3150N (i.e. - 45 0N if the building is rotated anti clock wise). It is found that at this orientation, zone 3 (bed room) is showing better thermal performance than zone 2 (living room/ bedroom). Hence, it can be concluded that the orientation of the vernacular house is wisely selected. The maximum duration of wind direction in this climatic zone is from south, south-east and south-west. Hence, zone 2 and 3 of the selected house are in the line of the wind in summer months. It can also be concluded from Table 1 that in this climatic zone summer season will be uncomfortable due to persisting high temperature and high relative humidity. Natural wind direction is used wisely in this case to minimise the discomfort due to high relative humidity. Large openings in the form of window and ventilators on the external façade of the house is also promoting cross ventilation. Table 4 shows the comparison of indoor temperatures range between the field measurements and simulation of zone 2 of the building for base case. Figure 4 Daily maximum temperature profile of zone 2 of the selected house with common attic Figure 5 Daily minimum temperature profile of zone 2 of the selected house with common attic Figure 6 Daily total heat gains due to infiltration in zone 2 of the selected house with common attic Figure 7 Daily maximum temperature in zone 2 of the selected house with different cases Figure 6 represent the heat gain/loss due to infiltration in zone 2 for different orientations. It is observed from this figure that the present orientation of the selected house is the best option because in winter heat loss due to infiltration is less compared to other orientations. It is found that in summer there is a large heat gain due to infiltration and this may be one of the reasons for high indoor temperature and subsequently discomfort. In summer, it is expected to minimise the heat gain by operating windows and ventilators during day time and increase heat loss in night (night ventilation). Night ventilation can be enhanced naturally by opening windows and ventilators fully thus allowing maximum infiltration of outside air. This can also be achieved by using mechanical ventilation at night. This will reduce the discomfort duration in summer considerably. To avoid discomfort due to heat loss in winter, the main activity should be to increase the heat gain and minimise the heat loss. Here also if opening and closing of windows and ventilators are regulated intelligently then discomfort due to cold can be minimized to a large extent. So, it is found that large window to wall area ratio in existing vernacular houses can be used 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 284 intelligently to overcome the climatic constraint and consequently increase comfort duration in this vernacular house. Based on the above analysis, vernacular house of base case is selected (with common attic space and orientation 315 0N) for further simulations. The various scenarios considered for the simulation are listed in Table 3. Figure 7 and 8 represents the daily maximum and daily minimum temperature profile of zone 2 respectively. Similar profiles are also obtained for zone 3. It is observed from these figures that increase of insulation has minimum effect on the indoor air temperature profile of zone 2 in summer season. This happened due to high infiltration minimises the effect of increase in insulation. It is also observed from Figure 7 and 8 that increase in window area with double glazing leads to increase in daily maximum and minimum temperature in summer and winter season. However, the increase is more prominent in winter season. This happened due to low altitude of sun in winter helping sunlight enters directly inside the rooms through window glazing leading to increase in indoor temperature supported by better insulation properties of double glazing. Figure 9 represent the daily temperature swing in zone 2 for different cases. It is observed from Figure 9 that when insulation is applied to the inside wall of the building, the indoor temperature swing becomes high compared to base case. The reason for this can be attributed to the low inertia of the insulation and also insulation is not allowing energy stored in the external wall to radiate to indoors. Low thermal inertia and high infiltration is responsible for large temperature swing. This situation may lead to discomfort in indoors if insulation is applied to naturally ventilated buildings. Hence, it can be concluded that the base case is the best option with respect to the daily indoor temperature swing of zone 2. Similar results are also obtained for zone 3. Figure 8 Daily minimum temperature in zone 2 of the selected house with different cases Figure 9 Daily temperature swings in zone 2 of the selected house with different cases It is also observed from Figure 9, that decrease in window area case is showing lowest swing in daily indoor temperature profile. However, this cannot be suggested as best design option, as it will also drastically reduce the day lighting level and natural ventilation which will lead to discomfort. In all the cases, high indoor temperature swing is observed from January to June months as during this period wind velocity is high, which enhanced the heat gain and loss due to high infiltration. It is found from the Predicted Mean Vote (PMV) and thermal model analysis, that these houses show low thermal comfort in winter and summer months (Singh et al., 2010a; Singh et al., 2011a; Singh et al., 2015). However, they show acceptable thermal comfort in pre-summer and pre-winter months. Simulation results also show that zone operative temperature is always lower than zone air temperature by 1 – 1.8 0C and zone meant radiant temperature is always lower than zone operative temperature by 1.2 – 1.8 0C throughout the year. Similar trend is also observed for all other cases also. CONCLUSIONS In this study, vernacular building of warm and humid climate zone of North East India is considered for design based thermal optimization by using the simulation tool TRNSYS. It can be concluded based on the analysis that the house having ceiling with common attic is showing acceptable daily indoor temperature swing. It is also found that the vernacular houses of this zone must have ceiling to minimize 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 285 the daily indoor temperature swing. It is found from the analysis that due to high infiltration in naturally ventilated building, insulation has almost negligible effect on the daily indoor temperature swing. However, it is found that increase and decrease of window and ventilator area has significant effect on the daily indoor temperature swing (window and ventilator area is most sensitive building design parameter). It also can be concluded from this study that increase and decrease of glazing area has maximum effect in the winter season when the sun altitude is less. Hence, it can be recommended that if the window be replaced with double glazing with proper shading mechanism then the indoor thermal conditions will be significantly improved. Thermal comfort analysis shows that buildings are thermally more comfortable in pre-summer and pre-winter season. However, this study needs to be further carried out by integration of airflow model with thermal model to obtain better results. REFERENCES Assimakopoulos, M. N., Mihalakakou, G., & Flocas, H. A. 2007. Simulating the thermal behavior of a building during summer period in the urban environment. Renewable Energy, 32(11): 1805 - 1816. Auliciems, A. 1981. Towards a psycho-physiological model of thermal perception. International Journal of Biometeorology, 25(2): 109 - 122. Bambrook, S. M., Sproul, A. B., & Jacob, D. 2011. Design optimisation for a low energy home in Sydney. Energy and Buildings, 43(7): 1702 - 1711. Bansal, N. K., & Bhandari, M. S. 1996. Comparison of the periodic solution method with Transys and Suncode for thermal building simulation. Solar Energy, 57(1): 9 - 18. Beckman, W.A., Broman, L., Fiksel, A., Klein, S. A., Lindberg, E., Schuler, M., & Thornton, J. 1994. TRNSYS The most complete solar energy system modeling and simulation software. Renewable Energy, 5(1-4): 486 - 488. Brager, G, S., & Dear de, R. J. 1998. Thermal adaptation in the built environment: a literature review. Energy and Buildings, 27(1): 83 - 96. Datta, G. 2001. Effect of fixed horizontal louver shading devices on thermal performance of building by TRNSYS simulation. Renewable Energy, 23(3-4): 497-507. Fuller, R.J., Zahnd, A., & Thakuri, S. 2009. Improving comfort levels in a traditional high altitude Nepali house. Building and Environment, 44(3): 479 - 489. Jelle, B. P. 2011. Traditional, state-of-the-art and future thermal building insulation materials and solutions – Properties, requirements and possibilities. Energy and Buildings, 43(10): 2549 - 2563. Kulkarni, K., Sahoo, P. K., & Mishra, M. 2011. Optimization of cooling load for a lecture theatre in a composite climate in India. Energy and Buildings, 43(7): 1573-1579. Orehounig, K., & Mahdavi, A. 2011. Energy performance of traditional bath buildings. Energy and Buildings, 43(9): 2442 - 2448. Ruiz, M. C., & Romero, E. 2011. Energy saving in the conventional design of a Spanish house using thermal simulation. Energy and Buildings, 43(11): 3226-3235. Singh, M. K., Mahapatra, S., & Atreya, S. K. 2007. Development of Bio-climatic zones in North East India. Energy and Buildings, 39(12): 1250 - 1257. Singh, M.K., Mahapatra, S., & Atreya, S. K. 2009. Study to enhance comfort status in naturally ventilated vernacular buildings of northeast India. 29th ISES Solar world Congress, Johannesburg 11-14 October, South Africa, Vol 2, Page number 1442 - 1450. Singh, M. K., Mahapatra, S., & Atreya, S.K. 2010a. Thermal performance study and evaluation of comfort temperatures in vernacular buildings of North-East India. Building and Environment, 45(2): 320 - 329. Singh, M.K., Mahapatra, S., Atreya, S.K., & Givoni, B. 2010b. Thermal monitoring and indoor temperature modeling in vernacular buildings of North-East India. Energy and Buildings, 42(10): 1610 - 1618. Singh, M. K., Mahapatra, S., & Atreya, S. K. 2011a. Adaptive thermal comfort model for different climatic zones of North-East India. Applied Energy, 88(7): 2420 - 2428. Singh, M. K., Mahapatra, S., & Atreya, S. K. 2011b. Solar passive features in vernacular architecture of North-East India. Solar Energy, 85(9): 2011 - 2022. Singh, M.K., Mahapatra, S., & Teller, J. 2015. Development of thermal comfort models for various climatic zones of North-East India. Sustainable Cities and Society, 14, 133-145. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 286 The Climatic Design in Chinese Vernacular Courtyard House Settlement – A Wind Environmental Simulation Yuan SHI, PhD pre-candidate Edward NG, Professor [School of Architecture, The Chinese University of Hong Kong] [School of Architecture, The Chinese University of Hong Kong] shiyuan@cuhk.edu.hk edwardng@cuhk.edu.hk ABSTRACT The application of Chinese vernacular courtyard environment adaptability design strategies in Chinese contemporary architectural design are becoming more popular among Chinese architects. Starting from the climate consideration in architectural design, using CFD simulation to understand the wind environment in the Chinese traditional vernacular courtyard and settlement based on North China climate conditions (taking Beijing as an example). Firstly, the wind environment in the original courtyard building settlement is simulated. Secondly, parametric studies on the effect of width-to-length ratio (W/L) and north building height-to-south building height ratio (H1/H2) on the wind environment in courtyard house are conducted. Finally, several variants of courtyard house are also tested. This study is consummating the deficiencies of previous similar studies and digging the key points of how the same architectural form provide wind environment adaptability in different seasons with totally opposite weather conditions. 1. INTRODUCTION With the development of building technologies, from 1950s, unified building forms of modernism (internationalism) swept the worldwide building industry and have been squeezing the survival space of the regional and vernacular buildings with traditional forms not only on cultural and aesthetic aspects but also on environmental aspects (Frampton, 1993). More dependence on unified modern building technologies leads to less local climatic considerations in architectural design. As building indoor environment is now universally conditioned by the HVAC system, energy consumption becomes higher and higher. In China, more than 1/3 of the total energy consumption directly derives from buildings. Considering both the severe global environmental situation brought by high energy consumption and the environmental quality needs from building occupants, the concept that to achieve better environmental quality, higher performance and more sustainable building by exploring climatic considerations and environmental strategies in the conventional vernacular building form has been put forward and got wider recognition (Givoni, 1998; Olgyay & Olgyay, 1963). As one of the most typical, conventional architectural forms-courtyard house was adopted in many contemporary architectural design by architects in China. The reason is that courtyard house has perfect climatic adaptability, especially with its natural ventilation and thermal environmental performance (Blaser, 1995). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 287 Although ‘learning from tradition’ is generally recognized among architects, the climate adaptability of those new architecture design based on tradition can hardly be achieved without drawing the ancient wisdom into the new building design in the right way and suitable place. The courtyard house is one of the most widely distributed architectural forms in China. By adjusting the shape (such as changing the aspect ratio, building height etc.), this form can easily be adapted to totally different climate conditions. Therefore, it is essential to know how different types of vernacular courtyard can be adapted to different regional climate. Using a specific courtyard shape in a wrong place or only forming a ‘seemingly the same’ enclosure space will not work or will even make thing worse. In this paper, computational fluid dynamic (CFD) simulations will be conducted for achieving better understanding about the wind environmental design characteristics of the Chinese vernacular courtyard and settlement based on North China climate conditions (taking Beijing and its specific courtyard form‘Siheyuan’ as the example). Related parametric studies will be conducted for consummating the deficiencies of previous studies on similar topic. 1.1. Climatic Condition in Beijing Beijing is located in northern hemisphere warm and semi-humid continental monsoon, which can be regarded as the typical representative city of the North China. Climate characteristics of Beijing can be summarized as follows: In winter, it is dry and cold, the mean temperature is approximate -2℃; in summer it is hot, sometimes humid, the mean temperature can achieve 26℃ or higher; in spring, it is warm and dry; in the fall which is the best season of Beijing, the temperature falls and the sunlight is bright (CMA, 2006). It can be observed that under Beijing’s climate condition, high air temperature with relatively high humidity will reduce both the indoor and outdoor thermal comfort level in summer if there is no enough ventilation; and in winter, the cold temperature climate condition will trial the building wind-protection and insulation performance. Figure 1 Summer time (left) and winter time (right) wind rose chart of Beijing. As to the wind, what can be seen from Figure 1 is that during the summer, highest frequency wind direction is from southeast (135°, N = 0°), the average wind speed is approximate 5.5m/s. In winter, frequencies of winds from different directions are relatively uniform. However the worst condition highest frequency of excessive wind speed appears at direction of northwest (315°), in which the wind speed can reach 8.0m/s or even more. High wind speed which exceeds the wind comfort threshold of peoples will decrease both the indoor and outdoor thermal comfort, meanwhile put negative influence on peoples’ activities (Penwarden & Wise, 1975). 1.2. The Vernacular Courtyard in Beijing - Siheyuan The courtyard is a central opening enclosed by buildings which is the most basic elements for Chinese traditional built environments, including cities, houses, and gardens (Xu, 1998). The vernacular courtyard house (also called ‘Siheyuan’ or ‘Chinese quadrangles’) is a conventional type of residence, which was widely adopted throughout the North China (Figure 2). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 288 Figure 2 Typical and the most basic layout of Beijing courtyard house with one central inner courtyard space. 2. LITERATURE REVIEW ON PREVIOUS STUDY In worldwide, many studies on CFD simulation for vernacular buildings in different climate regions have been conducted. However, regarding to the traditional courtyard in North China, although there are a lot of research and practice on the architectural design issues (such as spatial design, cultural context inheritance and heritage conservation, etc.), there are only two studies on the wind environment based on CFD simulation. In the first study (LIN, WANG, ZHAO, & ZHU, 2002), the effects of different courtyard shapes on their wind environment were tested based on only four different simplified “box” models without indoor space, the conclusion are descriptive not quantitative. Another study which focus on the wind environment in a single courtyard house was conducted by the author (SHI, 2013), which is the initial part of the study shown in this paper. The previous study draws the conclusion that the optimized courtyard shape should has the width-to-length ratio (W/L) of 1.0 and the north building height-to-south building height ratio (H1/H2 in Figure 3) of 1.2-1.4; when the overall amplification of courtyard building scale is three times of the original scale or more, appropriate precautions on wind-proof design must be taken to keep the pedestrian level wind environment comfort around the building. However, these results are acquired based on the simulation for a single courtyard house model without surroundings (Figure 3). Instead of the single courtyard house model, this study will validate the previous conclusion based on a courtyard array model and also take more design parameters into consideration. Therefore, this study can be regarded as the extension and also the improvement of the previous one. Figure 3 The single courtyard house model and the simulation domain used in previous study (Source: (SHI, 2013)). 3. WIND ENVIRONMENT IN TRADITIONAL VERNACULAR COURTYARD HOUSE Reynolds-averaged Navier-Stokes (RANS) equations with standard K-ε turbulent flow model is used to simulate the wind environment of the traditional vernacular courtyard house. The most basic courtyard house form in Beijing, which has one single inner central courtyard, is taken as subject investigated. The wind velocity field inside and around the courtyard house is studied with CFD simulation to understand the climate adaptive design strategies. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 289 One important thing should be clarified about this simulation study is that the buoyancy effect is not taken into account at this stage. It is true that buildings are receiving certain solar radiation, which will lead to spatial temperature difference thus affect the whole pattern of air movement. However, for the courtyard house in cold regions of North China, the thermal mass of buildings is considerable. Therefore, the air movement due to the temperature difference is relatively insignificant compare with the air movement due to building geometry. Figure 4 The courtyard house array model and the simulation domain used in this improvement study. Where H is the building top height of a single courtyard house in this array model. 3.1. Physical Model and Simulation Setting As mentioned, the simulation and parametric study are based on a courtyard array model (Figure 4). Simulation was conducted under summer and winter condition separately. The distance between courtyard houses (street width) is set based on the traditional urban texture of Beijing old cities (WANG, 2007). Following the climate analysis in Section 1.1, for summer time simulation setting, the initial temperature is 26℃, the wind condition is 5.5m/s with southeast direction based on the prevailing wind condition; and for winter time setting, the initial temperature is -2℃, the wind condition is 8.0m/s with northwest direction based on the most negative condition for thermal comfort and building wind-protection. Initial wind environment is generated based on the logarithmic wind profile with a reference height of 10m. At first, the wind environment in the original courtyard building settlement is simulated. Then, parametric studies on the effect of width-to-length ratio (W/L) and north building height-to-south building height ratio (H1/H2) on the wind environment in courtyard house are conducted. Finally, several variants of courtyard house with different entrance direction are also tested. For all simulations, both under summer and winter conditions, the wind speeds mean value and maximum value of all inner courtyard and indoor spaces in the courtyard house array is calculated and taken as the evaluation index. Therefore, for one type of courtyard geometry, there are total four indicators of wind speed. 3.2. Wind Environment in the Original Courtyard Building Design Figure 5 (left) is the wind environment in a courtyard house array in the summer time. In summer, the average wind speed in all inner courtyard space of the house array at the 2m-height position is about 2.0m/s under the background wind with a speed of 5.5m/s at 10m-height, which can provide satisfying natural ventilation. In winter time, Figure 5 (right) shows that the average wind speed of all regions of the nine courtyard houses at the 2m height position still keep the level of 2.5m/s, even the maximum wind 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 290 speed is lower than 4.0m/s under the background wind with a speed of 8.0m/s at 10m-height. Figure 5 Wind environment simulation results: wind velocity field at the 2m height level (m/s) in courtyard house array in summer (left) and winter (right). 4. FURTHER PARAMETRIC STUDY BASED ON PREVIOUS RESEARCH 4.1. The Width-to-Length Ratio (W/L) Wind environment in five courtyard house array models with different W/L ratio (range from 0.52.0) under summer and winter condition were simulated separately. Figure 6 shows the parametric simulation results comparison. Compare with the simulation of a single courtyard model, the overall wind speed for whole parametric test decreases, but the trend keeps unchanged. Taking the W/L ratio of 1.0 (original ratio of basic conventional courtyard house) as the baseline, what can be observed is that increasing the W/L ratio will slightly increase wind speed in the courtyard both in summer and in winter, but the effect of wind speed change on people is non-significant; when decreasing the W/L ratio, the wind speed is increased distinctly. In a courtyard where the W/L is equal to 0.5, even the maximum wind speed reaches the comfort threshold. Therefore, when architects attempt to design an enclosure courtyard space, quadrate shape whose W/L ratio approximates 1.0 is preferred. Long and narrow space along with the prevailing wind direction should be avoided. Figure 6 Comparison of wind speed in inner courtyard space with different W/L ratio (m/s). 4.2. The Building Height Layout (H1/H2) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 291 The second parametric study is regarding to the building height layout. Considering the prevailing wind direction both in summer and in winter, the north building height-to-south building height ratio (H1/H2) was selected as the variable in this parametric study. Six different H1/H2 ratios are simulated for comparison range from 1.0 to 2.0 under both of summer and winter condition (Figure 7), taking the original H1/H2 ratio of conventional courtyard house (1.2) as the baseline. Under summer condition, courtyard arrays with the H1/H2 ratio of 1.2, 1.4 and 1.8 have basically same performance better than others (H1/H2 = 1.0, 1.6 and 2.0). They have appropriate average wind speed for ventilation, and meanwhile the peak value doesn’t affect peoples’ activities negatively. However in winter, the H1/H2 ratio of 1.8 have maximum value of wind speed which will decrease the thermal comfort level and increase the heating energy consumption. Thus, the H1/H2 ratio from 1.2 to 1.4 is the preferred choices for designing courtyard space. Figure 7 Comparison of wind speed in inner courtyard space with different H1/H2 ratio (m/s). Above simulation study results based on basically validate the correctness of previous simulation based on the model of a single courtyard house that the optimized courtyard shape should have the widthto-length ratio (W/L) of 1.0 and the north building height-to-south building height ratio (H1/H2) of 1.2-1.4. There are some feedbacks on the previous study query that why the roof shape is not taken into consideration. The answer is that the roof shape of courtyard house was designed based on the daylighting (especially the accessibility of sunlight in winter) and summer shading, which is not the main factor of wind environment design. 4.3. Different Variants of Courtyard House Although most Beijing courtyard houses are built to follow the basic layout that the overall orientation along north-south (positioned in the north and facing south) and four buildings of a courtyard house are arranged along the north-south or east-west direction and the main entrance are located at the southeast, there are still different variants of courtyard house (MA, 1999). The most commonly seen variation is the adjustment of the main entrance location (shown in Figure 8). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 292 Figure 8 Several most commonly seen variants form of Beijing courtyard house. The basic layout doesn’t change, but the location of main entrance are different (Source: Based on (MA, 1999)). Using the same setting with above simulation, a courtyard house array model consists of four different variants of the courtyard house with different entrance location are simulated. Figure 9 shows the simulation results. The results show that the difference among four kinds of variants of courtyard house is also non-significant, which means that the entrance location is not the main impact factor of wind environment in courtyard house buildings. To further explain, in winter time, even the entrance has the same direction with winter time prevailing wind, by adding an additional corridor, the wind speed in the inner courtyard space still can be kept in the comfortable range of building occupants. Figure 9 Simulation results of the most commonly seen variants form of Beijing courtyard house: wind velocity field at the 2m height level (m/s). Left: summer time, right: winter time. 5. CONCLUSION AND DISCUSSION This study improves the previous study on the bioclimatic architectural design of the Chinese vernacular courtyard building. The optimized courtyard shape which has the width-to-length ratio of 1.0 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 293 and the north building height-to-south building height ratio of 1.2-1.4 is validated by the parametric study based on a new created Beijing courtyard houses array model. The simulation for several variants of courtyard house also shows that the location of courtyards’ main entrance will not affect the wind environment in the central courtyard space, if appropriate design measures are conducted. In the section 3.1, the author implies that higher wind speed is preferable in summer. Nevertheless, higher wind speeds are not chosen from results shown in Fig. 6 and 7. The reason is that for a residential building designed under climatic conditions with significant seasonal variation, it is important to make a balance between different needs in different season. It this case, both higher wind speed in summer or lower wind speed in winter is essential for a comfortable wind environment. Therefore, the trade-off has been made when determining the optimal choice. By understanding the key impact factors of wind environment in the courtyard house building form, architects can apply this vernacular architectural form and its climatic adaptability strategies into the environmental building design properly and wisely. This study focus on the influence of the building geometry, therefore, more future works will be conducted on the thermal aspects for comprehensive understanding of vernacular building climatic design strategies. REFERENCES Blaser, W. (1995). Courtyard House in China / Hofhaus in China: Tradition and Present / Tadition und Gegenwart. Basel: Birkhäuser. CMA. (2006). Atlas of surface climate diagram of China. Beijing: China Meteorological Administration. Frampton, K. (1993). Toward a Critical Regionalism: Six points for an architecture of resistance. In H. Foster (Ed.), Anti-Aesthetic. Essays on Postmodern Culture (pp. 16-30). London: Pluto Press. Givoni, B. (1998). Climate considerations in building and urban design. New York: John Wiley & Sons. LIN, B., WANG, P., ZHAO, T., & ZHU, Y. (2002). The Digital Simulation Study on Wind Environment of Traditional Courtyard. Architectural Journal(5), 47-48. MA, B.-j. (1999). The Building of the Quadrangle in Beijing. Tianjin, China: Tianjin University Press. Olgyay, V., & Olgyay, A. (1963). Design with climate: bioclimatic approach to architectural regionalism. Princeton, NJ: Princeton University Press. Penwarden, A., & Wise, A. (1975). Wind environment around buildings. Building Research Establishment Report. London: Dept. of the Environment, Building Research Establishment: Her Majesty's Stationery Office. SHI, Y. (2013). Wind Environment Characteristics in Chinese Vernacular Courtyard and its Design Application. In M. A. Schnabel (Ed.), Cutting Edge: 47th International Conference of the Architectural Science Association (pp. 493–502). Australia: The Architectural Science Association (ANZAScA). WANG, B. (2007). Beijing Micro-geographical Notes. Beijing, China: Joint Publishing (Beijing). Xu, P. (1998). Feng-shui models structured traditional Beijing courtyard houses. Journal of Architectural and Planning Research, 15, 271-282. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 294 Demystifying vernacular shop houses and contemporary shop houses in Malaysia; A Green-Shop Framework Amira Elnokaly, PhD Jun Fui, Wong, MSc University of Lincoln University of Lincoln aelnokaly@lincoln.ac.uk junsamuel89@hotmail.com ABSTRACT Vernacular shop houses in Malaysia have been thoroughly studied to understand their significance in environmental, cultural, economical and heritage values. UNESCO recognition in 2008 has further secured shop houses conservation works in Malaysia (UNESCO, 2008). However, contemporary shops in Malaysia do not share similar concerns of preservation and cultural significance. Popular view has perceived contemporary shop as lacking of both cultural and building performances standards. Thus, this research testifies the cultural and building performances in both contemporary shops and vernacular shop houses through cross-content analysis onto Malaysia Uniform Building Bylaw (UBBL), Green Building Index (GBI) and Special Area Plan (SAP). This research aims to critically investigate the correlation between vernacular and contemporary shop houses to develop a guideline strategy for green performance in shop houses. Through re-learning of vernacular shop house design and critical examination of governing policies, this research had highlighted some design issues that affects today practices. Policies are exploited to users’ interpretation that contributes to poorly built shop houses that have neither green nor cultural significance. This framework developed three distinct yet complementary areas in a bid to explore various green strategies and important criteria, which are building envelope design, green design, and cultural design to identify the correlation between green performances and cultural sensitive buildings. Hence, this research provides fundamental guides to portray future potential of high performance shop architecture in Malaysia. 1. INTRODUCTION Vernacular shop house (Malay: rumah kedai) is one of the unique architecture found in South East Asia particularly in Malaysia and Singapore built from 17th to early 20th century (Chen, 2007; Wan Ismail, 2005). The unique Chinese form of shop houses resulted from local influences and colonial’s modification in an attempt to adapt to tropical climates. Vernacular shop houses follow Chinese rules of thumb in architecture which are symmetrical (Hong, 2009), narrow layout, and air-well in between spaces (Wan Ismail, 2005). Contemporary description has defined a vernacular shop house as ‘built single, double or triple storey building’ (Mohd. Baroldin & Mohd. Din, 2012) with measures of 6 to 7 meter width and depth of 30 meters and it could extend up to 60 meters (Haromshah, 2009). However, these attached buildings are not built simultaneously but over the time, adjoined together (UM-NUS Joint Studio Programme, 2009). Singapore Governor Sir Raffles altered the Chinese shop house’s structure in 1822 by imposing five-foot ways (a covered pedestrian arcade) to accommodate wet weather in the region (Wan Ismail, 2005, p. 28; Abdul Mohit & Sulaiman, 2006). Furthermore, in late 19th century, backlanes for shop houses were required to allow accesses for sanitary and fire preventive measures, yet, had been reduced to limited use of rubbish collection in contemporary practice. Nonetheless, these changes have contributed to today unique shop houses physical forms. However, there are no evidences of continuous improvement to shop physical structure to adapt to present needs since mid of the 20th century. Hence, shop houses should revamp their present conventional structure to enhance the building performance toward greener design in a parallel response to sustainable development. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 295 1.1. Shop houses development to present Conservative organisations such as Georgetown World Heritage Inc. (GTWHI) and Badan Warisan have recognised heritage importance of shop houses that are dated back to before early 20th century (Figure 1). Mass developed shop houses after 1960’s are perceived as non-cultural importance (Figure 2) and categorised as contemporary shop houses. Since the 17th century, shop houses went into a series of evolution that represented the Chinese and hybrid cultural influence. The evolution or transitional changes are part of the process of adaptation of climates, local cultural, economic demands and fashion influence. Shop houses were popular urban fabric during the 19th century to the early of the 20th century because of socio-economical advantages (Chen, 2007, pp. 90-91). It is only found in Malacca. It is constructed with Dutch brick and lime plastered with timber structured roof. Simpler façade and symmetrical windows. It has one or two storey for residential use. Unlike other shop houses, Dutch styled shop houses are not connected with front arcade but confined with private entrances. Ornaments flourished with strong European influences. Shop houses employing tripartite arrangement windows. Decorative ceramic tiles are used and European decorative plaster molded to form bouquet and festoon shape on the façade. Timber is decreased and replaced by reinforced concrete. Earliest Chinese influenced shop houses found in Malacca city. The shop house embedded the notion of Chinese symbolism to promote spiritual harmonies. The structure is built with lime plastered brick and timber roofing. The architecture has close assembly to Chinese shops in southern China. Simpler shop house follow the Southern China style. Normally built as double storey with connected pedestrian arcade. Earlier style has smaller form and façade, constructed with timber. Masonry party walls are adopted. Buildings are decorated with classical elements such as pediment, moulded plaster and colonnade. Shop houses have abandoned timber construction and opt for masonry built. However, timber pitch roof is still practiced in the construction. Dutch-inspired gable was adopted for the facade of this shophouse. Source: Chen V. F. (2007), pg 90, Dutch Patrician. Minimal decoration display with decorated air-vent located below upper floors window. Continuous timber shutter design topped with Chinese decorated gable roof. Brick and lime plaster and unglazed roof tiles is widely used. Adapting Malay, Indian and European influences. Yet, ornaments are still limited in use but certain designs are adopted such as pilasters, arch windows, and keystones. New materials have been adopted such as glass and concrete, although timber is still widely used in construction. Geometrical shape inspired and few decorative that is limited to extended parapet and flat pole. Embedded building date as part of the decoration. Adaptation of Shanghai plaster for façade treatment and concrete shading devices. Reinforced concrete has fully adapted into building construction and abandoned timber as structural material. The trend further influenced contemporary practices without ornamentation but large flanks of overhang and shading design. Figure 1: Timeline depicts shop houses facade transition is influenced by socio-cultural and political changes. The facades show different materials used ranging from timber to concrete through the global technology advancement 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 296 The stylistic of shop houses have hybrid characteristic incorporating architectural vocabulary from the West, Chinese and Malay. However, the adaptation to the Western design was only popular after local exposure to the culture in late 19th century. Revival styles such as Neo-classical and Palladian in the 1920’s and 1930’s (Pile, 2009, p. 432) prompted this fusion design. The revival styles are more acceptable than modernism, although both were popular during the early 20th century. Hence, the built form fashion could be integrated into local identity (Abel, 2000) are of trans-cultural significance (Presas, 2005). Regrettably, shop houses that adapted modernism later in the 1960’s have not been classified as of historical significance because it lacks these unique characteristics of earlier shop houses and were continuously diluted by mass development. Moreover, contemporary shop houses are scarcely retained in original forms because of heavy modifications that resulted in difficulty to identify transitional forms in present architecture. The continuous modification of buildings structure is evidence of contemporary poor understanding of users’ needs. Figure 2: Classification of contemporary shop houses based on authors’ research and observation 1.2. Shop houses as everyday architecture Shop houses are simple buildings that do not stand as landmark or are of structural significant in urban definition. Their contemporary development and contribution to urban coherence and socio-cultural is poorly understood. Mass development is controlling the number of shop houses development in today urban fabric. However, studies found that this typology of everyday architecture is highly significant towards cultural development in heritage towns (Davis, 2006). Hence, contemporary shop houses would leave their marks onto Malaysian architecture and urban context that critically shaped the future of regional development. In contrast, the formations of vernacular shop houses have encapsulated everyday life and place identity (AJM Planning and Urban Design Group , 2011). The unique shop houses structures have remarkably shaped earlier part of many cities with Chinese settlements (Chen, 1993) in Malaysia. Collectivity (Terraced shop houses) of “individual” shop houses with distinctive embellishment has enhanced the language shared within the urban taxonomy. Shop houses constructed forming several rows have increased their significance as a cluster of buildings that shapes local community life. These buildings could not function as singular entity; despite of their building performances (Davis, 2006, pp. 236-237). Therefore, contemporary shop houses should be critically re-examined to ensure the significance of shop houses continuously upheld as unique everyday architecture in Malaysia. 2. Evolution of Shop Houses in Malaysia Contemporary shops in Malaysia are evolved from shop houses that dominated the urban landscape in the 19th and early 20th century. These retail buildings are significant in shaping local socio-economic aspects and formed parts of contemporary urban fabrics. The continuous development of new neighbourhoods or towns would observe shop houses as part of the common fabric for general small commercial activities in townscape and neighbourhood development (Maleki, et al., 2012). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 297 Policies restriction and socio-cultural acceptance play roles in how shop houses perform to fulfil community needs. The notion of shop houses being green architecture is not bound within the set of physical structure but is correlated within policies and regulation and their impact on socio-cultural aspects. Encouragement and support from authorities are needed to promote greater energy efficiency and sustainable building among public users (Yang, et al., 2014). 2.1. Locality and Regionalism in Architecture Shop house evolution lies within the acceptability of local towards foreign culture. The emphasis on foreign culture acceptability and adaptability to local identity is important to understand socio-cultural value in building performances. The relationship between architecture and its surrounding is simply an understanding of a place by oneself to create local identity (Abel, 2000, p. 143). Thus, architecture is the tangible resource of place identity that is influenced by socio-cultural aspects. The context of place identity is not only encased within socio-cultural aspect but as a holistic understanding of the place including climatic and topographic issues (Perera, 2013). Lee et al (2013, p. 604) identified place identity as reflection of local activities and its physical environment. The cultural importance, however, is only shown onto relevance and appropriation of “correct” culture that will enhance functionality and provides sense of orientation (Pelletier, 2012). However, in present shop house environment, buildings and its space has foregone the regional identity with globalised and homogenised image (Abel, 2000, p. 190). Kaye (1991 , p. 31) described the dilapidating of local shop houses as “hollow out of tradition”, emphasising the idea of empty shop houses’ façade. Hence, regional orientated contemporary shop houses could transform the building into contextually appropriate, instructive and encouraging as locally unique and functioning architecture (Too, 1990). Lewis Mumford summarised that regionalism is not preserving the past or imitating but to recreate the same cultural value that are encapsulated in vernacular architecture onto new buildings that represent contemporary community (Lefaivre & Tzonis, 2001). Thus, the notion of regionalism in Malaysia is to celebrate the local identity (Day, 2004, p. 238) and localise the modernity development in the country. Regional designated shop houses should be climatically responsive to enhance building performances (Ozkan, 2006, p. 108). Hybrid designed vernacular shop houses are learning examples of adapting foreign architecture to form local community identity (Abel, 2000) for contemporary practices. These shop houses will continue to readapt the changing norms of the society and practices. Hence, successful contemporary shop houses could adapt foreign elements to enhance aesthetic, functionality and building performances without sacrificing local identity. Concomitantly, the notion of modernisation would contribute into greater understanding of social problems that could be addressed with modern knowledge. Technology advancement and greater standardisation system could be beneficial towards improved design with enhanced understanding of material properties, construction methods and users behaviour. 2.2. Shop Houses and Urban Space In view of a good urban environment is a precondition for a good quality of life, the quality of that area (urban) is a reflection from buildings and minor developments within the boundary. These physical developments besides being functional should incorporate cultural identity, green initiative and efficiency. Mass developed standardised contemporary shop houses have disrupted the urban patterns (Said, et al., 2013) with monotonous façades they have intimidated other surrounding buildings. The destruction of community identity is in-search for new “signature” and enforces this synthetic image to represent the local identity (Kaye, 1991 ). Shop houses have encompassed other urban functionality in Malaysia including socio-cultural and economic importance. Street activities around shops area such as daily greengrocer market, weekly night market and community events that priorities local need (Ja'afar, et al., 2012; Lee, et al., 2013) and plays a major role in the public realm. The daily street activities are essential practice to form community conscience, which stimulates cultural identity and economical advancements in the area (Day, 2004). However, the sense of community would not be sustained without strong physical evidence in urban fabrics (Ujang, 2012). Contemporary shop houses are lacking in the physical attribution towards local cultures and community uniqueness (Said, et al., 2013, p. 422). Researchers (Samadi & Mohd., Yunus, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 298 2012; Said, et al., 2013) have suggested that modernisation development should preserve the cultural images to maintain the consistency of urban character. 3. METHODOLOGY This research is employing qualitative content analysis (QCA) in demystifying shop houses changes in physical design with socio-environmental influences. Similar researches were conducted in researching particular theme from documents as shown in Beharrell (1993) and Airken (1998) study (Bryman, 2008, p. 557). The significant of QCA is to produce wider and in-depth meaning from textual data by interpretation and relating it to the conducted research. This research would adapt relational analysis to explore and critically examine the relationship (Williamson, et al., 2003) of green building performances, socio-cultural aspect and shop physical design. Similar research was conducted onto shop houses in Singapore (Tut, 2011). In this research, source of data would draw up from three significant building standards, which are Malaysia Uniform Building By-law (UBBL), Green Building Index (GBI) and Penang’s Special Area Plan (SAP). UBBL is Malaysia building regulation law that administer all building construction standards in the country. Their minimum requirements would be generalised in this research as fundamental criteria for building construction. On the other hand, GBI is a non-compulsory rating system in Malaysia that is tailored to suit the country’s climate (Tan, 2009). GBI would be the yardstick for both types of shop houses in green performances. Lastly, SAP is a draft regulation in Penang for protection of heritage buildings. SAP allows this research to identify details of construction methods, material and structure of vernacular shop houses. The identified criteria would be expanded and analysed. Therefore, through the three manual of regulations, authors would narrate keywords pending to physical regulatory and environmental input such as ventilation requirements, indoor comfort and greenery obligation. The cultural aspect would draw up from SAP by general coding keyword and requirement such as original materials, cultural words and shop houses. The coding of the three manuals would identify significant set of criteria of the Malaysian regulation standards that correlates to agencies commitment towards building performance and green design. 4. RESULTS UBBL is drafted as preliminary law to regulate the built environment industry that draws up by architect’s council in 1984 (Ministry of Housing and Local Government Malaysia, 2012). UBBL was drafted based on local buildings by law and British building regulation to unify the standards in building construction. Many of these standards in UBBL are following either local Standard Specification (LSS) or American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) regulation to provide optimal quality in building construction and internal comfort. On the other hand, SAP drafted by GTWHI complies with UNESCO requirements in conserving the world heritage town (AJM Planning and Urban Design Group , 2011). SAP claimed to be championing the improvement in quality of life, derived from economy progression and developing sustainable and conserving heritage city (2011, pp. 6-1). The dynamic vision of SAP is fundamental notion that portray positivity that could be adopted in this proposed green framework. Thus, analysis of SAP would identify on how local cultural sensitive design incorporates into contemporary practices. In addition, SAP is in line with other building regulations and laws, which includes UBBL to avoid legal contradiction. SAP depends on UBBL to provide building standards such as fire regulation, building height and ventilation requirement. Lastly, GBI Non-Residential New Construction (NRNC) guideline released in 2009 is used as reference towards green performances. NRNC is derived from 51 requirements under six criteria; energy efficiency, sustainable site planning and management (SM), water efficiency, material and resources, indoor environment quality and innovation. In addition, Township guideline Version 1.01 is used to identify the cultural and green significance within urban context. Township guideline has 45 requirements that encompassed climate, energy and water, environment and ecology, community planning and design, transportation and connectivity, building and resources, and business and innovation. GBI Township is strongly focused on social and economic value with promoting the drive force for local business, amenities and housing facilities. Hence, Township guidelines would correlate the socio-economic aspects with local environment to produce green urban development. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 299 Table 1: Cross Case Analysis of UBBL, SAP and GBI requirements Criteria Physical Design Building Physical constraints UBBL SAP GBI Remarks Building depth and width is not specifically mentioned but building width shall not fall less than 20 feet (ft.) according to Concannon 1951 ruling Section 44 (3): Shop house ht. shall not be less than 10 ft. for ground floor and 8.4 ft. for any upper floors. No storey restriction. Section 40: Minimum requirement for 2 storey requires 7 square meter (m2), subsequently each floor is entitles to 1 m2 N/A N/A Typical vernacular shop houses have 13-20 ft. (width) and depth expand from 30-60 ft. Sec. 4.4: Building ht. shall not be higher than 18 meter (m) or 5 storey ht. N/A Vernacular shop houses have 12-18 ft. ht. or 1-3 storey ht. Sec. 6 Item 9.0: Air-well shall be maintained as part of the design with flexible roof to allow day lighting and natural ventilation NRNC EQ8: Skylights are encouraged to promote day lighting in building design Five Foot Way (verandah-way) Section 38: Verandah-way shall not be less than 2.25 m with 600 mm depth columns. Ramp (gradient less than 1 in 10) or staircase (minimum 150 mm riser x 275 mm treads) to level the adjoining units Sec. 6 Item 2.2: Commercial activities shall not obstruct pedestrian use. Verandah-way dimension shall abide to local regulation Townscape CPD and TRC: 75% of linked pedestrian walkway shall be covered to promote pedestrian scheme Vernacular shop houses have 1 to 3 air-wells separating internal spaces depending on building depth with optional rear court feature Vernacular shop houses have 5 ft. depth or less verandah-way Accessibility Universal Design (UD) is required for disabled accessibility. The designates section is also covered pedestrian prioritised for verandahway Section 112: staircase for shop with direct access from street shall be enclosed with incombustible materials. Opening shall be provided at each landing for ventilation except building that is less than 3 storey could be unventilated Party wall shall not be less than 200 mm thickness (thk.) with masonry or in-situ concrete Public space sharing is emphasised to encourage pedestrian scheme in verandahway Township CPD: Emphasising UD to accommodate disabled users with pedestrian network (TRC4) and open spaces Car park facilities should expand to accommodate contemporary use Sec. 5 Item 12: Staircase shall be built close to air-well for ventilation and abide to UBBL regulation on material use N/A N/A Vernacular shop houses have thicker party wall (300 mm thk.) sharing between units as fire preventive measurement. N/A N/A Section 53: All material shall abide to fire preventive and material safety endorsed by MS Standard. However, green materials are not included, assuming other regulation to be used ASHRAE and MS standards are applied to regulate IAQ through mechanical or natural ventilation. Under ASHRAE Standard 63-73: building shall provide 0.14 m3 of air per minutes (cm) per occupant. Thus, any room shall have opening not less than 15% of total floor space with exception in Section 41 (mechanical ventilation) Building that exceed 4000 m2 shall require to have overall thermal transfer value (OTTV) less than 0.4 W/m2K. However, typical shop houses does not require to but must abide to ASHRAE Standard 55 Energy efficiency shall abide to MS1525:2007 standard Material is restricted for roof and finishes (lime plaster and tiles) in preserving urban coherence. NRNC MR: Recycled and green certified materials with regional sourcing to reduce unnecessary carbon footprint in transportation N/A Ventilation depends on passive design through air-vent, air-well, rear court, jack roof and facade opening design NRNC EQ: IAQ shall abide to ASHRAE Standard 62 in regulating ventilation system to prevent harmful pollutants and mould. Natural ventilation is optional criterion provided that effective air exchange is set. N/A Passive cooling combined with lightweight structures to reduce thermal mass and heat gain NRNC EQ6: Accorded to ASHRAE Standard 55 Kwong et al. (2014) claimed ASHRAE Standard 55 would create unnecessary internal cooling. Shall utilise natural lighting through air-well, opening and air-vent N/A Not specifically mentioned in regulation, but, UBBL has emphasis UD and public space requirement in shop houses Shall conduct Cultural Impact Assessment (CIA) to take urban context coherent with the building including physical landscape, economy and community aspect NRNC EE: 35% of GBI guideline encompassing energy efficiency through exploiting available green certified lighting and lower OTTV Township CPD: Encompassing diversity in community and mix land use by providing secure design, health and basic amenities Building Height (ht.) Air-well Staircase Party Wall Green Design and Socio-Cultural Building Material Indoor Air Quality (IAQ) Thermal Comfort Energy Efficiency Socio-Cultural 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad N/A 300 Urban development N/A Assessing the impact of cultural diversity, living heritage and townscape in CIA. Encouraging mix use in land development in promoting urban interaction NRNC SM: Integrate community with greater compact density with promoting green transportation. Under SM3 provision to provide basic amenities in the surrounding. Greater emphasis to car park spaces to facilitate contemporary needs either concentrated or incorporated into building space 5. DISCUSSION AND RECOMMENDATION The analysis of the studied policies has provided critical in-sight into contemporary practices. The cross examination between policies has provided fundamental knowledge into this proposed framework (Table 2). The framework comprises of 3 factors namely; Building Envelope Design (BED), Green Design (GD) and Cultural Design (CD) to determine fundamental values that should be embedded into future developments based on analytical research presented in Table 1. New technology and innovation in GD should be encouraged to enhance the building performances and fulfil the green agenda. Yet, passive design should always be prioritised to avoid ill-practises in this framework. Nonetheless, further research from public survey, climatic factor and physical simulation (Edwards & Naboni, 2013) is needed to testify the framework viability. Green-Shop Framework Item Building Envelope Design (BED) BED1: Building Physical Description Remarks 1) Shop house shall not build higher than 5 storey or 18 meters. 2) Shop house shall maintain higher ceiling height with minimal 12 ft. ht. to assist passive ventilation. 3) Shop house shall maintain 20 ft. or more width and more than 70ft. depth 1) Exceeding 5 storeys shop house is not supporting socio-economy with difficult accessibility. 2) Higher ceiling height allows stack ventilation and cross ventilation. 3) Deeper and wider shop house provides greater usable space with better air quality per occupant. 1) Studied recorded East-West orientated buildings have greater heat gain by 20-30%, hence, required thicker wall insulation or shading devices. 1) Comprising SAP and UBBL guideline to provide comfortable pedestrian friendly network. 2) UD is required for disable accessibility. 3) Pedestrian scheme could enhance socioeconomy with more space for community activities and engagement. While, concentrated car park space would provide safer zone for pedestrian use. 1) 2) Rear court space could be used for green space and allow greater ventilation. Back house activities could be contained within building, hence, promoting cleaner communal space. 1) As interconnect space, staircase is suitable as a daylight source for internal spaces. 2) - BED2: Façade Treatment 1) Shop house with East-West façade orientation shall have thicker wall insulation compared to North-South orientated façade. BED3: Five Foot Way (Verandah-way) and Accessibility 1) Verandah-way shall not be obstructed by any means to ease pedestrian use. 2) UD shall apply to shop house design for all user accessibility. 3) Concentrated parking space or using basement as car park space could maximise land use and allow pedestrian-prioritised scheme on street level BED4: Air-well and Rear Court 1) Restricted physical space, air-well would not be suitable for contemporary shop house design. Alternative solution such as skylight or light shaft could be employed. 2) Rear court shall be maintained for hygiene and optimising back lane functionality. 1) Enclosed staircase practice could be designed as light-well for shop's interior. 2) Staircase is encouraged to have more opening for ventilation and admitting daylights BED5: Staircase Access Green Design (GD) GD1: Building Material GD2: Energy Efficiency GD3: Passive Design 1) All building materials shall abide to fire safety use and endorsed by local standards. Green certified materials should be prioritised. 2) Materials shall not be restrict but would be encouraged to use local products with consideration of urban coherence. 1) Shop house shall be designed to maximised day lighting to reduce dependency of artificial lights. i.e. light shelf, light-well, light shaft 2) Users are encouraged to use green certified products as suggested in GBI guideline. 1) Shop house shall optimise passive design (i.e. orientation, insulation, ventilation) and reduce mechanical assistance whenever possible. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 1) 2) Using local products could reduce transportation's carbon footprints. Meanwhile, building materials could disturb urban coherence with unnecessary adornment. 1) 2) - 1) Research show local occupants have higher tolerance for regional climate. (Omar & S.F.Syed-Fadzil, 2011) 301 GD4: Indoor Air Quality (IAQ) 1) Passive ventilation should be prioritised through encouraging air-vent design, more opening space and jack roof design. 2) Mechanical ventilation shall not be part of alternative building design. Mechanical ventilation shall only be applicable to encourage air flow GD5: Thermal Comfort 1) Thicker insulation for wall and roofing to reduce heat gain. Shading devices (i.e overhang, louvres) should be employed. Shop house shall refrain from using tinted window 2) Shop house shall have greater thermal mass to reduce u-value on strategic part of the building (i.e East-West orientated wall/facade) GD6: Technology and Innovation 1) Green high technology and innovation are encouraged in accordance to advancing society. 2) New technology application shall adhere to local green standards. Cultural Design (CD) CD1:Cultural and Community 1) Developer shall conduct post-occupancy evaluation (POE) to understand regional (local) communities in Malaysia. 2) Shop house design shall be inspired by local design to reflect the rich cultural differences CD2: Socio-Economy 1) Shop house development shall asses economy value to provide basic amenities for residential area CD3: Urban Context and Identity 1) Mix land use development shall be maintained and integrated with local communities. 2) Shop house development shall consider all basic facilities to attract local community engagement 1) Studied recorded passive ventilation is sufficient to provide effective ventilation in shop house building. (Omar & S.F.Syed-Fadzil, 2011) 2) Mechanical ventilation could be adopted to remove pollutants. Mechanical fan and attic fan could be adopted to encourage stack ventilation or cross ventilation in the building 1) Thermal resistance shall increase to lower internal temperature, while, reducing solar heat gain with strategic employment of shading devices. 2) Shop house shall meet 0.4 W/m2K (u-value) requirements regardless of building size. 1) High technology shall not replace or supersede passive design whenever possible. 2) Technology adoption shall be thoroughly studied before application to avoid ill practices. 1) POE could empower architects engagement with end-users to understand local community's culture, place attachment and reduce unnecessary features or adornment 2) Place identity is important for new development in searching of sense of belonging and attachment. 1) Basic amenities and local working opportunity shall undergo studies to reduce unnecessary shop house development at mature neighbourhood 1) 2) - ACKNOWLEDGEMENT The authors would like to extend their gratitude to Arkib Negara Malaysia, Singapore’s National Archive, Badan Warisan Malaysia, Georgetown World Heritage Inc. and other individuals that provided valuable information and resources to support this research project. REFERENCE Abdul Mohit, M., & Sulaiman, M. B. (2006). Repeal of the Rent Control Act and it's Impacts on the Pre-War Shophouses in Georgetown, Malaysia. Journal of the Malaysian Branch of the Royal Asiatic Society, 79(Part 1), 112-120. Abel, C. (2000). Architecture and Identity: responses to cultural and technological change (2nd ed.). Oxford: Architectural Press . AJM Planning and Urban Design Group . (2011). Draft Special Area Plan: George Town, historic cities of the Strait of Malacca. Petaling Jaya: State Government Penang. Bryman, A. (2008). Social Research Methods (3rd ed.). Oxford: Oxford University Press. Chen, V. F. (1993 йил July). Malaysia's Architectural Heritage - a slide survey. Intan: Majlis Warisan Malaysia. Chen, V. F. (2007). The Encyclopedia of Malaysia: Architecture (1st ed., Vol. THE ENCYCLOPEDIA OF MALAYSIA). Kuala Lumpur: Didier Millet. Davis, H. (2006). The Culture of Buildings (1st ed.). New York: Oxford University Press. Day, C. (2004). Place of The Soul: architecture and environmental design as healing art (2nd ed.). Oxford: Architectural Press . Edwards, B. W., & Naboni, E. (2013). Green Buildings Pay: design, productivity and ecology (3rd ed.). Abingdon: Routledge. Haromshah, N. H. (2009 йил 25-March). Lecture Week 4: The Chinese Immigrants. Cyberjaya: Limkokwing University of Creative Technology. Hong, T. (2009 йил December). Classical Architecture: Through the Window of the Singapore Shophouse. Passage, pp. 89. Ja'afar, N. H., Sulaiman, A. B., & Shamsuddin, S. (2012). The Contribution of Landscape Features on Traditional Streets in Malaysia . Procedia - Social and Behavioral Sciences , 50, 643-656. Kaye, L. (1991 йил 11-April). Town Planners have to balance contradictory factors: The L-shape boom . Far Eastern Economic Review, pp. 31-36. Lee, Y. L., Said, I., & Kubota, A. (2013). The Roles of Cultural Space in Malaysia's Historic Towns: the case of Kuala Dungun and Taiping. Procedia - Social and Behaviour Sciences , 85, 602-625. Lefaivre, L., & Tzonis, A. (2001). The Suppression and Rethinking of Regionalism and Tropicalism After 1945. In A. Tzonis, L. Lefaivre, & B. Stagno (Eds.), Tropical Architecture: Critical Regionalism in The Age of Globalisation (pp. 14-58). Chichester: Wiley-Academy. Maleki, M., Zain, M., & Ismail, A. (2012). Variables communalities and dependence to factors of street system, density and mixed land use in sustainable site design . Sustainable Cities and Society, 3, 46-53. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 302 Ministry of Housing and Local Government Malaysia. (2012). Uniform Building By-Laws 1984. 12th. Putrajaya: MDC Publishers Sdn. Bhd. . Mohd. Baroldin, N., & Mohd. Din, S. A. (2012). Documentation and Conservation Guidelines of Melaka Heritage Shophouses. Bangkok: Elsevier Ltd. Omar, N., & S.F.Syed-Fadzil. (2011). Assessment of Passive Thermal Performance for a Penang Heritage Shop house. (pp. 203-212). Penang: Elsevier Ltd. doi:10.1016/j.proeng.2011.11.157 Ong, K. D. (1999 йил October). The Conservation of Pre-War Shophouses: a case study of Penang and Singapore. Kuala Lumpur: University of Malaya. Ozkan, S. (2006). Traditionalism and Vernacular Architecture in the twenty-first Century. In L. Asquith, & M. Vellinga (Eds.), Vernacular Architecture in The Twenty-First Century: theory, education and practice (pp. 97-109). Abingdon: Taylor & Francis. Pelletier, L. (2012). The Space of Fiction: on the cultural relevance of architecture. In P. Emmons, J. Hendrix, & J. Lomholt (Eds.), The Cultural Role of Architecture: contemporary and historical perspectives (pp. 58-67). Abingdon: Routledge. Perera, N. (2013). Critical Vernacularism: Multiple roots, cascades of thought and the local production of architecture. In N. Perera, & W.-S. Tang (Eds.), Transforming Asian Cities: intellectual impasse, Asianizing space and emerging translocalities (pp. 78-93). Abingdon: Routledge. Pile, J. (2009). A History of Interior Design (3rd ed.). Hoboken: John Wiley & Sons. Presas, L. M. (2005). Transnational Buildings in Local Environments (1st ed.). Aldershot: Ashgate Publishing Limited. Qi, j. K., Adam, N. M., & Sahari, B. (2014). Thermal Comfort Assessment and Potential for Energy Efficiency Enhancement in Modern Tropical Buildings: a riview . Energy and Buildings , 68(Part A), 547-557. Roaf, S., Crichton, D., & Nicol, F. (2005). Adapting Buildings and Cities for Climate Change: A 21st Century Survival Guide (1st ed.). Oxford: Architectural Press. Said, S. Y., Aksah, H., & Ismail, E. D. (2013). Heritage Conservation and Regeneration of Historic Areas in Malaysia. Procedia - Social and Behavioral Sciences, 105, 418 – 428. Samadi, Z., & Mohd. Yunus, R. (2012). Conflict of Image and Identity in Heritage Commercialization . Procedia: Social and behavioral Science, 50, 675-684. Tan, L. M. (2009). The Development of GBI Malaysia (GBI). Kuala Lumpur. Too, A. (1990 йил May-June). Didactic Streetfronts and Backlanes - the shophouse typology (Part One). Majalah Arkitek, pp. 48-56. Tut, C. G. (2011). Learning from the Singapore shophouse: towards a sustainable tropical architecture. Florida: University of Florida. Ujang, N. (2012). Place Attachment and Continuity of Urban Place Identity. Procedia - Social and Behaviour Sciences, 49, 156-167. UM-NUS Joint Studio Programme. (2009). Reclaiming Heritage: Shophouses and vernacular houses in Kuala Terengganu, Malaysia (1st ed.). Kuala Lumpur & Singapore: CORE, University of Malaya and CASA, National University of Singapore. UNESCO. (2008). Historic Cities of The Straits of Malacca: Melaka and George Town. Retrieved 2012 йил 29-November from http://whc.unesco.org/en/list/1223 Wan Ismail, W. H. (2005). Houses in Malaysia: Fusion of the East and the West (1st ed.). Johor Bahru: Universiti Teknologi Malaysia Press. Williamson, T., Radford, A., & Bennetts, H. (2003). Understanding Sustainable Architecture (1st ed.). London: Spon Press. Yang, L., Yan, H., & Lam, J. C. (2014). Thermal Comfort and Building Energy Consumption Implications - a review. Applied Energy , 115, 164-173. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 303 Session 3B : Innovative technologies PLEA2014: Day 1, Tuesday, December 16 16:05 - 17:45, Compassion - Knowledge Consortium of Gujarat Design Strategies on Heat Recovery of Cooking Stove in Rural Houses of China 1 Shimeng Hao 1 Zhenghao Lin 1,2 Yehao Song , Prof. PhD 1 Ning Zhu , PhD 1 Gui Zhang 3 Jialiang Wang , PhD [1. School of Architecture, Tsinghua University, Beijing, China] [2. Key Laboratory of Urban-Rural Eco Planning & Green Building, Ministry of Education, Beijing, China] [3. Department of Civil, Architectural, and Environmental Engineering, Missour University, MO, U.S.] ieohsong@mail.tsinghua.edu.cn ABSTRACT Wulong County is a high altitude mountainous region located in the southeast of Chongqing Province, included in the hot summer and cold winter (HSCW) climate zone of China. The indoor and outdoor temperatures are quite low during the winter and sometimes are intolerant for local occupants. The percentage of possible sunshine is only 13% in winter according to meteorological data, which makes it nearly impossible to use solar energy for space heating. Other ways of heat gaining without extra energy consumption should be explored for the rural houses of this area. This paper analyzed impacts of cooking activities on indoor environmental quality, and estimated the potential of heat recovery of cooking stove for space heating. We conducted in-depth observations of occupants’ behavior (including life patterns and cooking activities) and field investigations on thermal environment and indoor air quality. A series of design strategies were proposed based on these survey results. The strategies emphasized the utilization of heated walls and a proper room layout. INTRODUCTION The hot summer and cold winter (HSCW) climate zone, with 0.55 billion people living there, covers a grand area of the central China. The climate is far harsher than any other places of the same latitude. Wulong County located in Chongqing Province with an average altitude of over 1, 000 meters above sea level, is a representative area of the HSCW zone. The annual average temperature is 15-18°C. The extreme minimum temperature can reach as low as -3.5°C，while the highest temperature is 41.7°C, with high humidity all year round. The annual precipitation is 1000-1200mm. Most precipitation is April to June for four months, accounting for 39% of annual precipitation. People live in this remote mountainous area suffer from both extreme hot summer and cold, wet winter. The present existing rural houses fail to achieve thermal comfort especially in winter. According to the field measurements performed in February, 2012, the average indoor air temperature of a traditional timberwork house is 2.45°C. The indoor temperature even falls below zero sometimes. The situation with the modern concrete house is no better. The average indoor air temperature is 4.59°C with the minimum value of 1.4°C, which are far below the thermal comfort zone. The heating season is up to six months from October to the following March. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 304 
Figure 1 The climate condition of Wulong during a typical year. (Data sources, www.accuweather.com) An in-depth study has been conducted since 2011. Based on long-term field measurements and investigations, a number of low-tech and lowcost strategies have been proposed to improve the indoor environment quality. We found the potential of heat recovery from cooking activities and upgraded traditional heated-wall system as the heat source. In this paper, the present conditions of thermal environment, indoor air quality, energy consumption and occupancy schedule have been discussed in detail. And the innovative heated-wall system and a possible house layout are introduced. The design intends to rediscover localization by using locallyavailable materials and tradional building technologies in an innovative way. The indoor thermal comfort and indoor air quality can be improved and comply with the features of occupants' life pattern. The strategies and techniques we proposed have broad-range applicability in HSCW zone. Community discussions were held during the whole design process and a pilot building is to be built there for future assessments. METHODS Respectively, four field investigations on indoor environment qulity (IEQ) were conducted in August 2011, April 2012, January 2013 and February 2014, in a remote village of Wulong County, Chongqing Province, China. Both questionnaire interviews and quantified measurements were applied. A total of 105 households participated in the questionnaire surveys, which includes 47 valid questionnairs in summer and 58 valid questionnairs in winter. The investigation included energy consumption of the household, heating method, health condition and 24-hour occupancy schedule. The subjective thermal comfort questionnair survey was also carried out, including thermal comfort vote (TCV), thermal sensation vote (TSV), thermal satisfaction and expectation. Certain scales and remarks are attached to these votes. TCV has a five-point scale from 0 to 4 (0 represents comfortable and 4 is limited tolerance); TSV has a seven-point scale from -3 to +3 (-3 is very cold and +3 is very hot); Thermal satisfaction has four remarks from unsatisfied to satisfied valued -1, -0, +0 and +1, respectively (-0 means “just unsatisfied” while +0 means “just satisfied”). Mean while, the thermal environment measurements were carried out during these four surveys, indoor and outdoor environment parameters were recorded by auto-loggers for at least 72 hours each time. The thermal and luminous performances of rural houses in different seasons have been discussed and presented in our former papers. For more detailed information pleases refer to those previous studies. Field measurements of indoor air quality (IAQ) were only performed in February 2014. The indoor CO concentration, CO2 concentration and particle matter in four typical kitchens were monitored for 24 hours. Detailed Information of the Instruments are shown in Table 1.

30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 305 Physical Quantity PM2.5 CO CO2 Table 1. Detailed Information of the Instruments Instrument Accuracy Dust Trak8520 Q-Trak 7565-X EZY-1 ± 0.001mg/ m3 ±3% or ±50 ppm ±75ppm Data Intervals 1 minute 5 seconds 5 minutes OBSERVATIONS AND RESULTS Energy Sources and Consumption According to survey, wood and liquefied petroleum gas (LPG) are the main energy sources for cooking while wood, charcoal and electricity are used for heating. On average, 16.7% of the household income is spent on energy. During heating season, the everage energy consumption of wood and charcoal are 536kg and 126kg per household respectively. And due to the increasing heating demand, the monthly everage energy consumption of electricity is 305kWh per household while the amount is only 178kWh in summer. The energy price of wood is 0.25 RMB/kg and charcoal is 2.8 RMB/kg. Residential electricity price is 0.55 RMB/kWh. Figure2 is composed of photos and thermographs of traditional heating and cooking methods. The locals adhere the concept of ‘interval heating and spot heating’. Instead of heating the whole room, families or neighbours gather around a basin of charcoal or an open fire pit. High efficiency biomass stove are also applied in recent years. The traditional cooking stoves are widely used, burning woods and agricultural wastes. These inefficiency stoves cause the indoor air pollution. To make it worse, people often use part of the kitchen as living room in a traditional house, which leads to higher exposure to particle matters and harmful gases. As for concrete house occupants who usually have better incomes, using electric stoves and electric heating lamps is a common choice. Although it is less polluted, it significantly increases the energy consumption. Figure 2 Traditional heating and cooking methods used by local residents. Subjective Votes About 34.5% of the respondents were dissatisfied with the present heating methods for hygiene problems or high levels of consumption. 96.6% of the respondents expected their houses to be warmer during the winter. TSVs for indoor thermal enviroment were -1.76 and -1.78 (close to cold) for winter daytime and nighttime, respectively. TCV was 1.18 (between slightly uncomfortable and uncomfortable) and the result of thermal satisfaction vote was -0.38 (between unsatisfied and just unsatisfied). These results illustrate the necessity of improve the thermal conditions in winter. Low temperature and high humidity lead to higher risk of rheumatic disease and even higher mortality rate in elderly during the winter. About 47.4% of the respondents had arthritis or rheumatism disease. The prevalence rate increased with age. This rate increased to 66.7% in the age group of over 60 years old. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 306 Indoor Air Quality Figure 3 Daily concentration of PM2.5 and CO of a vernacular house with traditional cooking stoves. The indoor air pollution is quite serious due to the lack of ventilation design and incomplete combustion, particularly for tradtional houses. As shown in Figure 3, the peak value of PM2.5 concentration in a tradtional house was 48.055 mg/m³ while cooking which was ten times higher then that of the concrete house. The peak value of CO concentration was 28.3ppm, which was eight times higher then that of the concrete house. The peak value of CO2 concentration reached 662ppm in daytime while the valley value was 423 ppm in nighttime. These air quality indicators overrun the limitation of national standard thus increase the risk of hypertensive disease. Life Patterns A 24-hour occupancy schedule was formed based on questionnaire survey and observational survey, as shown in Figure 4. It inllustrates the potential of using the afterheat of cooking activities as heat source for living room and bedrooms. The activities in living room and bedroom peaked right after cookings. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 307 Figure 4 24-hour occupancy schedule in winter based on field survey. DESIGN STRATEGIES Design Intends The design aims to support a more comfortable and healthier living condition without extra energy consumption and investments, by the properly building layouts and the innovative heated-wall system. It also takes the unique life patterns of local occupants into account, making efforts to minimize the total heating demand by continuing the concept of ‘interval heating and spot heating’. A house menu is provided to maximize the flexibility and personalization. The economic cycles of households are also considered. Occupants can run small business, keeping rural livestocks, or processing farm products at home. Several design strategies and techniques have been applied as follows. Figure 5 Detailed drawings of the heated-wall system Figure 6 Running modes of the heated-wall system 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 308 Innovated Heated-wall System Heated-wall system is a traditional technique for heating in the northeast of China but rarely known by southerners. The cooking stove is connected to a hollow wall, letting the hot smoke exhausted from the wall cavity. In this way, the wall is heated during cooking and radiant heat to adjoining rooms. The wall is usually located between kitchen and bedroom. According to the study helded by Tsinghua University in 2008, the heated wall can maintain a relatively comfortable temperature for about two hours after the stove stops burning. The innovated heated-wall system we proposed here adds a fireplace on the other side, sharing chimney with the hollow wall, As shown in Figure 5 and Figure 6. The cooking stove uses biomass instead of coal, reducing the reliance on fossil fuels. The fireplace can be alternated by a high-efficiency biomass stove. Buffer zone With a proper layout, we can optimize the use of heated-wall system (see Figure 8).. The concept of “buffer zone" is applied to resolve the paradox of cooling in summer and heating in winter (see Figure 7). The core zone of the plan is composed of four spaces surrounding the heated-wall system. Insulation layers are only used around the core zone which is the main living area in winter. The core zone is surrounded by a buffer zone, which is semi-open space buffering the core from cold outdoors and playing the main living area in summer. In this way, the indoor environment quality is improved without extra energy consumption and little investments. Figure 7 Plans and working model of design proposal Figure 8 Routine of the operation of heated-wall system harmonizing with daily life 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 309 Localization The construction can be performed by local labours, and the new building technique can be passed on and as a means of livelihood for them. All strategies are affordable, using locally available materials, and cost efficient even for people in less affluent area. Community discussions were held during the whole design process and a pilot building is to be built there for future assessments. CONCLUSIONS In this paper, based on in-depth field investigations and measurements, the thermal environment and indoor air quality of rural houses in southwest of China were analyzed. The authors disscussed the impacts of cooking activities on indoor environmental quality and the potential of heat recovery through occupant’s life patterns. A series of low-tech and low-cost design strategies were proposed based on survey results. The strategies emphasized the utilization of an innovated heated-wall system for space heating, locallyavailable materials and a proper room layout to optimize the benefits. ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China (NSFC), Design Strategies of Chinese Vernacular House in Hot-summer and Cold-winter Climate Zone (Grant No. 51278262) and State Key Laboratory of Subtropical Building Science, South China University of Technology, Research on Ecological Strategy and Technology of Livable Environment in Subtropical Area. REFERENCES JGJ134-20012001. Design Standard for Energy Efficiency of Residential Buildings in Hot Summer and Cold Winter Zones. Beijing: China Academy of Building Research and Chongqing University. (in Chinese) Song, Y.H., Wang, J.L., Hao, S.M., and Y.L Song. 2013. The energy-related impacts of social factors of rural houses in southwest China. ISES Solar World Congress. Song, Y.H., Hao, S.M., Wang, J.L., and J.J Li. 2012. A comparative investigation on sustainable strategies of vernacular vuildings and modern buildings in southwest China. Peru PLEA Conference. Hao S.M., Song, Y.H., and W. Zhang. 2011. Field study of summer thermal environment of rural house in the southeast of Chongqing. Eco-city and Green Building, 8(4): 90-93. (in Chinese). Li, Y., and S.M. Hao. 2012. Investigation on light environment of eural house in the southeast of Chongqing. Eco-city and Green Building, 9(1): 113-116. (in Chinese) GB/T18883-2002, etc. 2002. Indoor Air Quality Standard. Beijing: China General Administration of Quality Supervision Inspection and Quarantine. (in Chinese). Shan, M., Li, D.K., and X.D. Yang. 2011. Study on influences of combustion of fire place on indoor environment and thermal comfort in rural residences. Building Science, 27(6): 10-14. (in Chinese). Feng, G.H., Zhang, X.Y., Feng, G.Y., and X. Sheng. 2010. Structural design and performance analysis of heating wall with inner phase change materials. Renewable Energy Resources, 28(5): 139-143. (in Chinese). Wang, P.S., yang, M., yang X.D., and M. Shan. 2014. Thermal performance of a traditional Chinese heated wall with the in-series flow pass: Experiment and modeling. Energy and Buildings, 84(12): 46-54. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 310 Efficient Building Design Model Generation and Evaluation: The SEMERGY Approach N. Ghiassi, MSc U. Pont, PhD A. Mahdavi, PhD [Department of Building Physics and Building Ecology, Vienna University of Technology] J. Heurix, MSc S. Fenz, PhD [Institute of Software Technology and Interactive Systems: Vienna University of Technology] ABSTRACT The discordance of the available Building Performance Evaluation Tools with the capabilities and expectations of the design community and complexities related to data availability and accessibility are among the most important technical barriers against sufficient adoption of such tools to support and guide design decisions. On the other hand, the complexity of the design problem, a consequence of the large number of variables and options involved (e.g., financial, environmental, technical, and legal factors) calls for more effective approaches towards sustainable design optimization. The SEMERGY project seeks to overcome such technical challenges through development of a user-friendly design optimization environment, tailored to suit the specific requirements and skills of the novice and professionaldesign community. The key feature of the SEMERGY environment is the incorporation of semantic web technology toward efficient search for and compilation of input information required for comprehensive analysis and evaluation of candidate design options supported by multi‐objective decision support methods. The present contribution briefly presents the underlying concept of the SEMERGY environment. It particularly focuses on SEMERGY's beta release, designed for optimization of retrofit projects in view of potential construction options. The current release accommodates requirements of novice user through a simplified web-based Graphical User Interface, the workflow of which is presented and discussed in detail. INTRODUCTION AND BACKGROUND Despite the advances in the development of Building Performance Evaluation (BPE) tools over the past decades, the adoption of such tools to support and guide design decisions has been relatively slow and their implementation mostly limited to certification purposes (Hensen et al. 2004, Pang et al. 2012). Most important stated technical barriers against adoption of such tools for sustainable planning and informed decision making are discordance of the available tools with the capabilities and expectations of the design community (Attia 2011), and complexities related to data availability and accessibility, which render the process of data provision for BPE applications cumbersome, error prone and time consuming (Mahdavi & El-Bellahy 2005). The SEMERGY project is an ongoing effort to facilitate the integration of performance assessment methods in the building design process to support sustainable design decisions. Neda Ghiassi and Ulrich Pont are research fellows in the Department of Building Physics and Building Ecology, Vienna University of Technology (VTU). Ardeshir Mahdavi is the director of the Department of Building Physics and Building Ecology, VTU. Johannes Heurix is a research fellow in the Institute of Software Technology and Interactive Systems, VTU. Stefan Fenz is senior researchers in the Institute of Software Technology and Interactive Systems, VTU. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 311 Identification of different design alternatives through a performance guided process is hampered by the complexity of the problem, a consequence of the large number of variables and options involved (e.g., financial, environmental, technical, and legal factors). "Manual" approaches toward identification of optimal solutions (mainly by trial and error) are time-consuming, expensive, and inconclusive. Hence, more effective approaches to optimization operations are being pursued, involving optimization platforms and associated automated procedures (Coffey 2008). SEMERGY explores developmental opportunities toward effective evaluation environments for comparative assessment of alternative design and retrofit options (Mahdavi et al. 2012a). The missing link between users' simplified component representations (e.g., "external wall", "window") and complex specifications of real world products makes the efficient generation of building performance assessment models very difficult. In other words, it remains the task of end-users to map such simple notions of building components to appropriate real-world products that meet calculation procedures' informational requirements. The key contribution of the SEMERGY project is the demonstration of the potential of semantic web technologies toward populating the input data for building performance simulation models via the navigation of the extensive but currently ill-structured web-based information space. Currently, SEMERGY is focused (as proof of concept) on the scattered pool of building product and material data. However, data pertaining to building systems (e.g. heating and cooling systems, active solar components), as well as resources and documents concerning procedural, climatic, and financial (e.g. public funding) information that could be of value to designers and decision makers, can be processed and utilized in the same fashion (Mahdavi et al. 2012a, 2012b). To accomplish this task, the SEMERGY system deploys two main strategies. First, information regarding building materials, elements, and components are obtained from various resources of the web environment. This information is preprocessed, restructured and augmented to meet the informational requirements of the integrated performance evaluation procedures. Using this reorganized and enriched repository, SEMERGY identifies design alternatives through a rule-based procedure. These potential alternatives are checked against building codes (pertaining, for example, to maximum allowed Uvalues). Thus, the corpus of possible permutations of the initial design could be efficiently reduced to a computationally reasonable size. Once the ordered set of feasible alternatives is constructed, it is made subject to a comprehensive evaluation process. Thereby, normative demand calculations, environmental impact assessment and cost estimation procedures are deployed. Upon completion of the assessment of the alternative designs, a collection of the best performing solutions is generated and presented to the user. In summary, the SEMERGY system has links for i) user interaction (user interface), ii) applications and computational engines (reasoning interface), and iii) sources of information (semantic interface). The user interaction link is intended to involve both simple web-based templates for novice users and advanced building information models for professionals. The beta version of the web-based interface has been released and is openly accessible (SEMERGY 2014). The application link supports data exchange between the system and multiple analysis tools pertaining to energy calculation, lifecycle analysis, financial payback assessment, and optimization. The information link, which is the critical ingredient of the proposed architecture, is supported by Semantic Web Technology (Mahdavi et al. 2012a). Previous publications presented and discussed fundamental features of the technologies embedded in the SEMERGY environment (Mahdavi et al. 2012a, 2012b, Ghiassi et al. 2012, 2013, Shayeganfar et al. 2013, Pont et al. 2013, Heurix et al. 2013, Hammerberg et al. 2013, Wolosiuk et al. 2014) The present contribution focuses on the SEMERGY's beta release. Specifically, User Interface and workflow patterns are discussed in detail. SEMERGY BETA FOR NOVICE USERS Purpose and Structure The current SEMERGY environment addresses the requirements of novice or professional users interested in a quick estimation of the thermal and environmental performance of their intended design and the alternative design possibilities within the limits of their financial means. The proposed alternatives include sets of construction options, for various building elements, entailing real world products available on the market. SEMERGY beta addresses at present optimization of retrofit projects. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 312 However, SEMERGY for new buildings is under development. Figure 1 depicts the structure of the SEMERGY environment. The reasoning interface of the present tool, incorporates a normative calculator for heating demand, a cost estimator, a simple life cycle analysis method based on the OI3 index (IBO 2014), as well as a multi-objective optimization procedure with an embedded automatic generator of alternative design options. The semantic interface integrates an ontology of building products and materials. This linked repository is derived from two web-based product databases (MASEA 2014, BauBook 2014), enriched with cost data from various product reseller websites (e.g., OBI 2014), and other properties required for the alternative identifier logic. The data transition between the various system components is facilitated by the SEMERGY internal building data model (SBM), developed to comply with the requirements of the integrated and intended computation engines (Ghiassi et al. 2013). Figure 1 Structure of the Semergy Environment. Workflow: Data Acquisition General Information. The process starts by the entry of the location of the building to be refurbished. According to the location, the appropriate weather data and building codes are selected on the background for calculation and validation purposes. Next, the construction period and general construction method is selected as shown in Figure 2. This information allows the system to determine the relevent construction configurations of various building elements according to the common practice of the time of erection of the building. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 313 Left: Definition of building location by selection on the map or entry of adress; Right: Figure 2 Selection of construction period, type, and number of floors. Roof Properties. Typical building floors can be easily represented in 2 dimensions and therefore even complex plans can be drawn with a simple 2D user interface. However, relatively simple building plans can result in complex roof geometries that cannot be represented in 2D. As a result, requiring users to input precise roof geometry necessitates a more sophisticated 3D user interface and places a larger data input burden on the user. Therefore, it is important, in view of simplifying the information input requirements on the user, to explore other alternatives. An algorithm has been developed to autogenerate a set of potential roof forms and select the statistically-proven best-fitting geometry based on a minimal set of input data (Hammerberg et al. 2013). Such an approach supports the generation of a roof approximation, suitable for more detailed performance analyses such as dynamic simulation, in the meantime, SEMERGY estimates the volume and area of the roof according to the roof type, height, angle and floor plan formation acquired from user as shown in Figure 3. This data suffices the purpose of the currently incorporated computational engines. Building Systems Information. The heating system of the building is selected from a provided list of options. The current version of SEMERGY does not include building systems in the optimization process, but the next release is intended to address this issue. Building Geometry and Space Properties. The geometry of the building is entered via drawing of plans on a grid. Help messages guide the user through the geometry entry process. Each building element of a different composition (e.g., external wall, fire wall, internal wall, etc.) is drawn with a different line color as shown in Figure 4. This layering is used on the back ground to associate the various enclosure elements with the appropriate constructions and boundary conditions. Once the walls, doors and windows of each level are in place, the drawing is analyzed and various spaces (rooms) are identified. The user is then asked to select the function of each room (e.g., bedroom, unheated basement, etc.) and the height and sill height of the windows. Room functions are associated with templates of internal conditions, occupancy, lighting and equipments. This information is useful in case of integration of more sophisticated thermal performance computation methods such as dynamic simulation. For the time being, functions only determine whether a space is conditioned or not, to help identify the thermal envelope of the building. After the entry of all floor plans and assignment of space properties, the orientation of the building is set by rotating the plan on a map of the location. In order to capture skylights and dormers, the type, number, dimensions and orientation of the roof windows are entered by the user. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 314 Figure 3 Left: Simplified roof description; Right: Selection of building systems. Figure 4 Left: Acquisition of floor plans; Right: Capturing of glazed roof elements. Semantic Attributes of Building Components. In accordance with the construction year and type, set in the beginning of the process, the user is presented with a list of construction options for each building element. These constructions follow the common practice of the period, in which the building was erected and comply with the national norms on calculating the heating demand of historical buildings. The user may moderately readjust the layers of these default constructions to better represent the building (as shown in Figure 5). They may also determine whether or not a certain construction is subjected to optimization. Detailed physical properties and cost information pertaining to the selected material configurations are retrieved from the building product ontology embedded in SEMERGY's semantic interface. As such, by simple selection of a construction from a given list of options, all information required for thermal, environmental and economic performance evaluations of the building are provided in the background. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 315 Figure 5 Selection and readjustment of semantic properties of various building elements. Workflow: Evaluation and Optimization Base Case Evaluation. Once the complete description of building and its components is acquired, SEMERGY generates a model of the building in the SBM format, a space-based three dimensional representation of the building, compliant with the requirements of the integrated performance computation engines. This model is subjected to performance evaluations. The system then provides the user with a base case assessment of the building's thermal performance. The results are presented to in the form of an energy certificate. The available budget for the refurbishment project is given by the user and the optimization process is initiated. Identification of Potential Design Alternatives. According to the user-selected initial constructions, the integrated alternative identifier generates potential refurbished versions of building components by addition or removal and addition of new layers. These layers are selected from the ontology of building products by a rule-based logic that evaluates various properties of products for their conformity with the requirements of the construction subject to optimization. The resulting components are then re-evaluated in view of their compliance with building codes (e.g., maximum U-values). Optimization Procedure. Once the scope of the potential (refurbished) combinations for each building element is determined, implementing the genetic algorithm method to reduce the number of computations to a manageable size, the pareto optimal set of solutions are identified according to multiple criteria of thermal, environmental and economic performance of the building. Each solution is composed of a full set of constructions for different building components. However, not all these constructions are modified as not all components are subjected to optimization. Workflow: Communication of Results The user can navigate between various solutions by prioritizing different optimization criteria. This is done by the help of sliders (as shown in Figure 6), which cover the range of values of performance indicators (cost, heating demand and OI3 index) associated with different solutions. Obviously, selection of a value on one slider, affects the values displayed by others. Thus, the user may easily grasp the consequences of each decision (reduction of the budget or selection of more sustainable products) on all variables. A graph illustrates the pay-back time of the investment for each solution. Refurbished building elements of the selected solution are also displayed. The results of the optimization process can be downloaded as a PDF report. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 316 Left: Navigation through optimization results; Middle and Right: Final report including Figure 6 retrofit suggestions and the expected performance level. CONCLUSION & FUTURE WORK The present contributed illustrated the recent state of SEMERGY’s beta release. The current environment features a multi-objective optimization procedure for thermal retrofit projects and is supported by Semantic Web Technologies. This implementation demonstrates the capability of the proposed semantic approach in facilitating the utilization of the extensive, yet ill-structured informational AEC-ressources of the World Wide Web. Undoubtedly, intensive usability tests have to be carried out, to ensure the adequateness of SEMERGY to accommodate the requirements of novice and laymen users. In addition to usability tests, ongoing research includes improvment of the cost-estimation methods, integration of building systems and services in the computation process, addressing interoperability to CAD-, 3D-Drafting- and BIM-Applications, and augmenting functionality of the environment to support optimization of new buildings. Readers of this contribution are invited to use and test the free-of-charge SEMERGY Demo available under https://www.semergy.net. ACKNOWLEDGMENTS The SEMERGY project is funded under the FFG Research Studio Austrian Program (grant No. 832012) by the Austrian Federal Ministry of Economy, Family and Youth (BMWFJ). In addition to the authors, the SEMERGY team includes: A. Anjomshoaa, K. Hammerberg, I. Merz, T. Neubauer, C. Sustr, F. Shayeganfar, M. Taheri, A.M. Tjoa, D. Wolosiuk and A.Wurm. The SEMERGY project was recently granted funding for a third year of research after intensive evaluations by the FFG. REFERENCES Attia, S. 2011. State of the Art of Existing Early Design Simulation Tools for Net Zero Energy Buildings: a Comparison of Ten Tools. Louvain La Neuve: Université catholique de Louvain BauBook. 2014. [https://www.baubook.at/zentrale/] Coffey, B. 2008. A Development and Testing Framework for Simulation-Based Supervisory Control with Application to Optimal Zone Temperature Ramping Demand Response Using a Modified Genetic Algorithm. Montreal, Canada: Concordia University Ghiassi, N., Shayeganfar, F., Pont, U., Mahdavi, A., Heurix, J., Fenz, S., Anjomshoaa, A., Tjoa, A.M. 2013. A Comprehensive Building Model for Performance-Guided Decision Support. In: Mahdavi, A., Martens, B. (Eds.), Proceedings of the 2nd Central European Symposium. on Building Physics: 35-42. Vienna: Vienna University of Technology 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 317 Ghiassi, N., Shayeganfar, F., Pont, U., Mahdavi, A., Fenz, S., Heurix, A., Anjomshoaa, A., Neubauer, T., Tjoa, A.M. 2012. Improving the Usability of Energy Simulation Applications in Processing Common Building Performance Inquiries. In: Sikula, O., Hirs, J. (Eds.), Simulace Budov a Techniky Prostredi. Brno: Ceska Technika - nakladatelstvi Hammerberg, K., Jain, V., Ghiassi, N., Mahdavi, A. 2013. Generalizing roof geometry from minimal user input for building performance simulation. In: Mahdavi, A., Martens, B. (Eds.), Proceedings of the 2nd Central European Symposium. on Building Physics. Vienna: Vienna University of Technology Hensen, J., Djunaedy, E., Radosevic, M., Yahiaoui, A. 2004. Building Performance Simulation for Better Design: Some Issues and Solutions. In: Wit, M.H. de (Ed.), Proceedings of the 21th conference on passive and low energy architecture, vol. 2: 1185-1190. Eindhoven: Eindhoven University of Technology Heurix, J., Taheri, M., Shayeganfar, F., Fenz, S., Pont, U., Ghiassi, N., Anjomshoaa, A., Sustr, C., Neubauer, T., Mahdavi, A., Tjoa, A.M. 2013. Multi-Objective Optimization in the SEMERGY Environment for Sustainable Building Design and Retrofit. In: Mahdavi, A., Martens, B. (Eds.), Proceedings of the 2nd Central European Symposium on Building Physics: 35-42. Vienna: Vienna University of Technology IBO. 2014. Ökokennzahlen - OI3 Leitfaden. [http://www.ibo.at/de/oekokennzahlen.htm] Mahdavi, A. and El‐Bellahy, S. 2005. Efforts and Effectiveness Considerations in Computational Design Evaluation: a Case Study. In: Building and Environment. Vol.40, Issue 12: 1651‐1664. Elsevier Mahdavi, A., Pont, U., Shayeganfar, F., Ghiassi, N., Anjomshoaa, A., Fenz, S., Heurix, J., Neubauer, T., Tjoa, A. 2012a. SEMERGY: Semantic Web Technology Support for Comprehensive Building Design Assessment. In: Gudnason, G., Scherer R. (Eds.), eWork and eBusiness in Architecture, Engineering and Construction: 363-370. Reykjavík: Taylor&Francis Mahdavi, A., Pont, U., Shayeganfar, F., Ghiassi, N., Anjomshoaa, A., Fenz, S., Heurix J., Neubauer T., Tjoa, A. M. 2012b. Exploring the Utility of Semantic Web Technology in Building Performance Simulation. In: Proceedings of BauSIM 2012: 58-64. Berlin: Universität der Künste Berlin MASEA. 2014. Materialdatensammlung für die energetische Altbausanierung. [http://www.maseaensan.de/] OBI. 2014. [http://www.obi.at/decom/home.html] Pang, X., Hong, T. Piette, M.A. 2013. Improving Building Performance at Urban Scale with a Framework for Real-Time Data Sharing. Report. San Diego: Ernest Orlando Lawrence Berkeley National Laboratory Pont, U., Shayeganfar, F., Ghiassi, N., Taheri, M., Sustr, C., Mahdavi, A., Heurix, J., Fenz, S., Anjomshoaa, A., Neubauer, T., Tjoa, A.M. 2013. Recent Advances in SEMERGY: a Semantically Enriched Optimization Environment for Performance-Guided Building Design and Refurbishment. In: Mahdavi, A., Martens, B. (Eds.), Proceedings of the 2nd Central European Symposium on Building Physics: 35-42. Vienna: Vienna University of Technology SEMERGY 2014. [https://www.semergy.net] Shayeganfar, F., Anjomshoaa, A., Heurix J., Sustr, C., Ghiassi, N., Pont, U., Fenz, S., Neubauer, T., Tjoa A.M., Mahdavi, A. 2013. An Ontology-Aided Optimization Approach to Eco-Efficient Building Design. In: Proceedings of the 13th Conference of International Building Performance Simulation Association: 2194-2200. Chambery Wolosiuk, D.,Ghiassi, N., Pont, U., Shayeganfar, F., Mahdavi, A., Fenz, S., Heurix, A., Anjomshoaa, A., Tjoa, A.M. 2014. SEMERGY: Performance-Guided Building Design and Refurbishment within a Semantically Augmented Optimization Environment. In: Advanced Materials Research Vol. 899: 589-595, Switzerland: Trans Tech Publications 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 318 Integration of Outdoor Thermal and Visual Comfort in Parametric Design Emanuele Naboni Associate Professor, Institute of Architectural Technology, The Royal Danish Academy, School of Architecture, Copenhagen, Denmark ABSTRACT Parametric modeling tools are increasingly adopted in design practice. Various plug-ins for Grasshopper – the most widely used parametric tool – allow the creation of mathematically originated geometries from environmental data such as solar geometry, wind direction and velocity, radiation intensity, illuminance levels, etc. However, a critical look at the application of parametric methods in the practice of design reveals that their use is still predominantly based on aesthetical, structural and fabrication criteria. The opportunities that these tools offer to design strategies and components that are responsive to outdoor and indoor comfort conditions are starting to be explored at research level, but are rarely comprehensively integrated in the education and practice of architecture. To investigate the links between parametric form-making and outdoor comfort, a workshop at the Royal Danish Academy – aimed at the design of shelters – combined Parametric and Environmental Simulation Tools (ESTs) with the use of the most recent Grasshopper’s plug-ins. In search of thermal and visual comfort optimization, the students employed these parametric design tools to achieve responsive geometrical design solutions. KEY WORDS: Parametric Design, Environmental Simulation Tools, Outdoor Comfort, Design Creativity. INTRODUCTION The past decade has seen the emergence of intricately-articulated surfaces whose design and production were enabled by the capacity of parametric tools. However, the way in which these design solutions have contributed to human comfort (e.g. thermal) has often not been analyzed in detail. This is particularly evident when looking at the parametric design (PD) of urban shelters. Complexly shaped forms have typically responded to fabrication and aesthetical principles, without holding careful consideration of users’ comfort conditions (Turrin et al., 2012). Yet, the potential is there: their geometry and materials could positively affect thermal and visual comfort. To explore the missing link between PD and comfort, a series of “parametric shelters” were designed by students of the CITA Master at the Royal Danish Academy (Cita.karch.dk, 2014). In a period of two weeks, design teams composed by 2 to 4 people approached climatically “challenging” urban sites in New York, Berlin, Honk Kong, Shenzen, Singapore, Reykjavík, and Madrid, a rural site in Barcelona, and a desert area in Iran. Each of these locations presented climatic conditions that limited their usability. The design of the shelters was optimized through 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 319 the parametric control of their overall shapes. The rationalization and modularization of their geometry was obtained with the integration of Environmental Simulation Tools (ESTs). Design Comfort by Architectural Means The design of shelters should mitigate external climate influences and facilitate, among other functions, the thermal comfort and daylight quality (Fig.1) of the spaces below and those adjacent. Several studies showed the influence of comfort on the ratio of utilization of outdoor spaces and on users’ behavior (Nikolopoulou and Lykoudis, 2006). Outdoor climate studies indicate that the conditions expressed by the Physiological Equivalent Temperature (PET) – one of the most accepted models for outdoor comfort – are dependent on the radiation exposure and on wind velocity (Hoppe, 1999). Thermal comfort studies conducted by Bouyer et al. (2007) on envelopes of stadia showed the importance of geometry with respect to such parameters. For instance, designing the porosities of stadia (i.e., the capacity of the envelope and of the structure to control wind flow penetration) and the sky-opening factor (i.e., the sky view from the spaces sheltered) determines relevant differences in PETs (Turrin et al., 2012). Figure 1. Valldaura (Barcelona). (a) The site, which includes a vineyard, hosts various activities and is exposed to continuously changing weather conditions. (b) First phase of the design exercise, when various shapes are studied according to optimal protection or exposure, depending on instantaneous radiation. (c) The structure is parametrically defined in order to enhance or impede wind flows and solar penetration as a function of comfort. Comfort number hours are predicted to be extended by 40% from April to October (Students: Ida Katrine Friis Tinning, Inès Klausberger, Tadeas Klaban). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 320 TEACHING CONTEXT Integration of Design and Analysis in a Parametric Environment The workshop was centered on the use of Grasshopper, one of the most popular parametric platforms (author’s article, 2014). There are several plugins that have been developed for Rhino/Grasshopper (Table 1). Some of the most recently available plugins allow an interface with validated Environmental Simulation Tools (ESTs), such as EnergyPlus, Radiance and Daysim (Roudsari, Pak, 2014). These plugins and tools are free and open source, and users can customize them based on their needs. In the CITA Master workshop, a single model was used for design and comfort analysis, facilitating a smoother, more integrative, and efficient workflow (Roudsari, Pak, 2014). The benefits of integrating ESTs into the design process have been often discussed in previous studies (Weytjens et al, 2012). At the time of writing, however, discussion by non-developers of the actual application of such tools in the design process has rarely been documented (plugins such as Ladybug and Honeybee, used in the workshop, are very recent). Table 1. Parametric Comfort Workshop: Design Process and Tools Design Phase Climate Analysis Performance Goals Geometry design and performance behavior assessment Inverse computing for design adaptability Verification Design Action Weather file information was overlaid to digital models Selection of one or more coupled performances goals (e.g. reduction of radiation when temperatures is over 24 degrees and wind is limited) Design of a variety of geometrical shelter massing based on the repetitions of modules and assessment of their performances. Specific design of one component and analysis of its performance Dynamic behavior of modules are determined on the basis of instantaneous weather conditions and user activity Design Solution are simulated for verification of their performances Parametric Modeling Tools (PMTs) Grasshopper - Ladybug Outdoor Comfort Diagrams, Physiological Equivalent Temperature (PET) Grasshopper - Honeybee, Ladybug Grasshopper Grasshopper - Honeybee, Ladybug, DESIGN METHODS AND PROCESSES Climate analysis The design was founded on a thorough understanding of the weather data from the part of the students. Students learnt how the climatic conditions of their site transform over time in order to define adaptive design responses (Fig. 2). Even if the weather files were provided to students, it was considered that sites may have different microclimatic conditions than the one represented by the available data due to local factors such as urban fabric, materials, colors, soft/hard-scapes, etc. The use of the Ladybug plugin allowed to relate weather data to contextual characteristics (Grasshopper3d.com, 2014). This plugin to the parametric platform Grasshopper facilitates the graphical exploration of connections between the data available and geometries and materials designed. Multiple climatic variables (e.g., radiation and temperature) can be represented, overcoming limitations of tools such as Ecotect and Vasari. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 321 Figure 2. The design aimed to bring daylight in, and guide it through, a narrow court in Honk Kong. Indoor and outdoor illuminance levels controlled the position of façade reflectors. The courtyard, with the addition of the reflective components, receives an increases amount of daylight. Indoor rooms receive a well-calibrated light (a) The different positioning of the reflective components is based on sky conditions and users activities’ requirements (c) Rendering of two configurations offered by the system (Student: Chan Chun Yin). Setting numerical and time based performance goals Multi-functionalism was considered as one of the design drivers: shelters had to be flexible and with open plans. The students envisaged final users’ behavior, as it determined the comfort goals with measurable and meaningful indicators. Accordingly, and in order to contain the shelter design within achievable boundaries, a selection of numerical performance goals was made. The selection comprised two groups of indicators, respectively for thermal and daylight comfort. Each team started the design by setting time dependent performance goals for: - Thermal comfort: PET or isolated factors influencing the heat-balance models of the human body were used (e.g., temperature, mean radiant temperature, shadows’ hours, radiation, air direction and velocity, etc.). - Daylight comfort: Illuminance and luminance values were adopted as design generators (Fig. 3) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 322 Figure 3 Design of an envelope system in Shenzen. The open-space hosts exhibitions (a). The facade opens depending on the type of exhibition light requirements. (b) Illuminance values were recorded in museums in order to study performance goals. (c) Possibility of illuminance variations were controlled by the façade system during a winter day (plan views) (Student: Wenyu Wu). Geometric design and assessment of performance behavior In this phase, Ladybug was extensively used to consider the implications of radiation and sunlight-hours. Integration of the plugin with Grasshopper allowed an almost instantaneous feedback on design modifications. The students created preliminary design concepts and simulated how they meet well-defined performance goals. Two key factors of their parametric system’s design – geometry and material properties – highly influenced thermal and visual comfort. Another Grasshoppers’ plugin, Honeybee, was integrated in the process to support detailed daylighting and radiation simulation using validated ESTs (Roudsari and Pak, 2014). Students could run several types of accurate image-based analyses to produce diagrams for luminance, illuminance or radiation. Inverse computing for design adaptivity A series of geometric reconfigurations were made so that shelters were able to react to climate conditions and users’ thermal and daylight comfort. The focus was on components that constantly varied their configurations to adjust to weather variations and fluctuations of spatial programs. The parameterization process defined the dynamic movements of the components. Inverse computing techniques were used to determine such specific movements (Fig. 4). The systems designed are reactive to changing weather conditions and factors such as radiation, illuminance values, and wind speed and direction. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 323 Figure 4 Design of a shelter in New York. (a) A Grasshopper “recipe“. (b) Determination of the degree of opening of each parametric module. (c) The opening is function of solar irradiance and activities. (Students: Hulda Jonsdottir, Lukasz Wlodarczyk and Olga Krukovskaya) DESIGN RESULTS The qualities of all the shelters designed stemmed from their inherent link to the site. Projects located in New York, Madrid (Fig. 5), Barcelona, and Singapore (Fig. 6), experienced overheating. It was assumed that the air temperature of the spaces underneath was largely affected by the solar exposure of the spaces. The focus was therefore on reducing radiation when excessive. In opposite climates, with significant wind loads, such as the desert in Iran (Fig. 7), and the Icelandic city of Reykjavik (Fig. 8), protection from wind was considered crucial. Airflows in the spaces adjacent to the shelters depended on the incoming wind velocity, which is affected by their shape and openings. Finally, projects located in dense sites in Berlin (Fig. 9), Shenzen, and Honk Kong, focused onto creating visual comfort, while saving artificial lighting. The use of daylight as a primary light source enhanced environmental quality. Similarly to the principles informing the achievement of passive thermal comfort, daylight-oriented passive strategies aimed at reducing the use of artificial lighting with variable forms and reflectivity of materials. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 324 Figure 5 Radiation simulated at ground level is the parameter controlling the movement of a canopy hosting several street activities in Madrid. (a) Configurations of the system at different times of a day. Radiation is controlled only when necessary (i.e. when temperatures are above 24 to 30 degrees, depending on wind velocity) (Students: Huen Cheying, Lo Chen-Chi, Annika Nora Richmond). Figure 6 Parametric shelters in Singapore. (a) A series of sections shows how natural ventilation, radiation, and reflected light from surrounding buildings are integrated in one model where massing and components act as a system that increase thermal and visual comfort (Students: Mattias Lindskog, Lyn Poon, Karoline Wæringsaasen, Thyge Wæhrens). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 325 Figure 7 Iran. (a) Simulated wind flows and solar radiation are the parameters controlling the mechanic movement of flexible strips inserted into rhomboidal components. (b) These are hosted within a series of vertical wind shelters (Students: Zeynab Zaghi, Jens Jacob Jul Christensen). Figure 8. Due to high wind frequency and velocity, outdoor spaces in Reykjavik are quite uncomfortable for a large part of the year. The project looked at controlling wind flow in a public market. Pressure coefficients from rudimentary CFD models were used to control the façade permeability to air (Student: Kristjana Sigurdardottir). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 326 Figure 9 Berlin. A series of vertical fins were parametrically controlled linking their elastic deformation to targets of illuminance levels. Since the site is a large open space, creating differentiated visual conditions was seen as a stimulus for users’ wellbeing and engagement (Student: Anders Per-Kristian Hansson). CONCLUSIONS Parametric modeling significantly increases the opportunities for climatic design. The workshop attempted to break the mono-disciplinarity of most teaching approaches to parametric design where geometric inverse computing are often solely based on fabrication or structural roles, for which a larger tradition of conceptual morphogenesis exists. The series of designs elaborated by students illustrated how relations between structure, shape, and materials can be efficiently integrated with environmental considerations related to thermal and luminous comfort. The findings indicated that it is possible and beneficial to integrate emerging parametric tools into the digital architectural and environmental design. These tools allow a ‘transversal’ approach to design information, and facilitate a decision making process based on the feedback loop between formal and functional relationships. Parametric design must be released from the constraints of ‘parametricism’ applied without any variations to all climates (however impressive its effects could be) and exploited to produce intelligent designs that embrace the full complexity of the environment. A guided use of parametric tools seems to grant the opportunity to foster climatic awareness, intuition, and design skills in support of the students’ decision making. At the workshop, students gained key insight into the microclimate conditions at play, and into capitalising locally available resources. They learned how the built environment is a highly complex system that involves many interdependent sub-systems. AKNOLEDGEMNTS The author gratefully acknowledges the collaboration of Tore Banke, Daniel Nielsen, Paul Nicholas, Martin Tamke, Sergio Altomonte and the dedicated participation of the students attending the workshop. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 327 REFERENCES Bouyer, J., Vinvent, J., Delpech, P., and Carre, S., (2007), Thermal comfort assessment in semioutdoor environments: Application to comfort study in stadia. In: Journal of Wind Engineering and Industrial Aerodynamics Vol. 95, pp. 963-976. Cita.karch.dk, (2014). CITA: Center for Information Technology and Architecture - Copenhagen. [online] Available at: http://cita.karch.dk/ [Accessed 1 Jun. 2014]. Grasshopper3d.com, (2014). Ladybug + Honeybee. [online] Available at: http://www.grasshopper3d.com/group/ladybug [Accessed 31 May. 2014]. Hoppe, P. (1999). The physiological equivalent temperature--a universal index for the biometeorological assessment of the thermal environment. International Journal of Biometeorology, 43(2), pp.71--75. Matcha, H. (2007). Parametric Possibilities: Designing with Parametric Modelling. pp.849--856. Naboni , E. (2014). Sustainable design teams, methods and tools in international practice. DETAIL GREEN, Issue 1, May 2014 Nikolopoulou, M. and Lykoudis, S., 2006. Thermal comfort in outdoor urban spaces: Analysis across different European countries. Building and Environment, 41 (11), p. 1455-1470. Roudsari, M. and Pak, M. (2014) Ladybug: A Parametric Environmental Plugin for Grasshopper to Help Designers Create an Environmentally-Conscious Design. Turrin, M., Von Buelow, P., Kilian, A. and Stouffs, R. (2012). Performative skins for passive climatic comfort: A parametric design process. Automation in Construction, 22, pp.36--50. Weytjens, L., Macris, V., Verbeeck, G., 2012, “User Preferences for a Simple Energy Design Tool:Capturing information through focus groups with architects”, PLEA2012 - 28th Conference, Opportunities, Limits & Needs Towards an environmentally responsible architecture Lima, Perú 7-9 November 2012 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 328 Session 3C : Control techniques for energy management PLEA2014: Day 1, Tuesday, December 16 16:05 - 17:45, Grace - Knowledge Consortium of Gujarat Energy demand, Thermal and Luminous Comfort in Office Buildings: a computer method to evaluate different Solar Control Strategies Antonio Carbonari, AP University IUAV of Venice carbonar@iuav.it ABSTRACT Many contemporary office buildings are characterized by large glazed surfaces, often located without any consideration about orientation. Without a suitable solar control strategy, this fact implies several problems related to visual comfort, thermal comfort and energy demand, which is mainly related to HVAC and, to a smaller extent, to artificial lighting. Moreover, if the office room is large, the values of physical parameters influencing comfort are relevantly variable from point to point, mainly as a function of the distance from glazed surfaces. Typically, daylighting requirements of occupants located far from the windows can conflict with the thermal comfort requirements of occupants located next to the windows. In this work a case-study is analysed. It consists in a medium size office room located in a typical office building, in an urban context of the Northern Italy. Different solar control devices and related control logics are compared; their effects on global comfort conditions and energy demand are assessed. The considered devices consist of different kinds of movable external slats, some of which incorporating PV cells. This analysis is performed by means of a specific software: "Ener_Lux", already presented in previous PLEA Congresses. Once defined the kind of devices and the related operating logic, the program simulates the dynamic thermal and luminous behaviour of the physical system, provides various comfort assessment index values and calculates the primary energy demand for HVAC and lighting. INTRODUCTION Many Contemporary office buildings are characterized by large glazed surfaces. Without a suitable solar control strategy this peculiarity implies several problems, both with regard to visual and thermal comfort and with regard to energy demand. The latter is mainly related, in Mediterranean climates, to HVAC and to a smaller extent to artificial lighting. Moreover, if the office room is large, the values of physical parameters influencing comfort vary relevantly from point to point; this variability is mainly due to the distance of the occupant from glazed surface. Consequently, for instance, lighting comfort requirements of occupants located far from the windows can conflict with the thermal comfort requirements of occupants located closer to the windows, requiring a reduction of entering solar radiation, because of overheating and glare. These problems are particularly relevant in office buildings realised in Italy in the last decades, characterised by extended glazed surfaces, located in any façade with no care about orientation. In this Author is an assistant professor in the Department of Design and Planning in Complex Environments, University IUAV of Venice, Venice, Italy. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 329 work a case-study is analysed. It consists in an office room of medium size located in a typical office building, in an urban context of the Northern Italy. Different solar control devices and related control logics are compared, focusing on their effects on global comfort conditions and energy demand. In particular, the considered devices are listed below. 1. Different kinds of external adjustable slat systems, with mirror-like or diffusing surfaces. 2. A double skin façade incorporating an adjustable slat system split into two parts: the upper part is composed by slats having mirror-like reflecting upper surfaces, whereas the lower part is composed by packable slats with diffusing surfaces. 3. Two kinds of adjustable external slat systems incorporating PV cells. All of these devices are combined with an internal diffusing screen, aimed to avoid glare then the slats tilt allows the entry of direct radiation. As a general rule, the control logics are aimed to minimize energy demand and to allow daylighting for the maximum span. All of the analyses presented in this paper are based on computer simulations performed by means of software Ener_Lux, previously presented in PLEA Congresses, in particular at PLEA 2012 (Carbonari, 2012). THE SOFTWARE Software Ener_Lux is mainly aimed at the study of solar control devices and related control strategies. Therefore it takes into consideration the physical system composed by a room, its glazed surfaces, internal and external solar control devices (slats, blinds, overhangs and any element shading the opening) as well as the surrounding urban environment, including the building containing the room under investigation. Scheme of the Ener_Lux calculation flow. The figure shows the behaviour of the Figure 1 program when referring to a double slat array, one of which is provided with mirror-like upper surfaces. Once defined the kind of devices and its control logic, the program simulates the dynamic thermal and luminous behaviour of the physical system at hourly time-steps, and provides: Predicted Mean Vote (PMV), Predicted Percentage of Dissatisfied (PPD) (Fanger, 1970) and Daylighting Glare Index (DGI) values, Hopkinson et al. (1963), together with other controls about the visual environment quality. Then it calculates the energy demand for HVAC and artificial lighting. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 330 When adjustable devices are simulated, all the solar control actions aimed to maintain thermal and luminous comfort, such as slats tilting or screen lowering, are automatically simulated: in such cases, the program modifies the system geometry configuration and repeats the simulation of the hourly time-step. The check against visual discomfort conditions is performed only when the lamps are turned off, as shown in Figure 1. The indexes used for the assessment of the visual comfort are calculated by means of an algorithm simulating occupants’ visual field, as shown in Figure 2. Different kinds of glare are considered: veiling glare due to direct radiation impinging on the visual task, that can imply thermal discomfort too, big differences of luminance values between different points in the visual field, and glare due to large luminous sources (typically the sky seen through the windows), evaluated by DGI calculation. When one of these kinds of discomfort issues is detected, the program simulates a solar control action. In case of thermal discomfort, averaged in the room as a whole, the indoor air set-point temperature may be modified to reach the true value of PMV (that has to be between -0.5 and 0.5 according to the standard ISO 7730) and the hourly time-step calculation is repeated. However, in this work this feedback is not performed. Simulation of an occupant’s visual field at the same time on different illustrative days, Figure 2 in presence of double slat arrays. The slats of the upper array are mirror-like reflecting. THE CASE STUDY The case-study consists in an office room of medium size: 5.88 m wide along the façade, 6.18 m deep orthogonally to the façade, and with internal height equal to 3.27 m. The room is located in an office building of the industrial district of Venice (Marghera). The building presents an entirely glazed façade almost south oriented (with 22° West azimuth). In the local climate this orientation is the less favourable during the cooling period. Actually, this façade is equipped with a system of tiltable slats incorporating PV cells, as shown in Figure 3 (b). At this side of the building a room at the second floor was chosen, to take into account the shading effects due to surrounding buildings. However, considering the distances from these buildings, there are not remarkable differences in solar irradiation between the second and the top floor. For comparison, a top floor room too has been simulated: when equipped with diffusing slats (see later) its total annual primary energy demand is 4.4% lower than in the case study. The building structure is composed by reinforced concrete. Internal walls are in hollow bricks 0.08 m thick, with 0.02 m thick plaster layer on both the sides. Floors are built in hollow bricks and reinforced concrete: 0.24 m is the construction thickness, plus 0.06 m of screed and flooring and 0.02 m of plaster in the lower part. The only external surface of the room is the glazed one, composed by a double glazing of 0.006 m glass layers, and a 0.012 m air gap (overall U value: 2 Wm-2K-1). All the other internal enclosing surfaces are considered as adiabatic. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 331 Four occupants are located in the room, at different distances from the glazed surface. To assess the visual comfort level for each position the worst line of sight has been considered: i.e. the one implying the higher contrast in luminances within the visual field. Thus, the glazed surface has to be present in it but it is empirically assumed that it cannot occupy more than half the visual field; otherwise the occupant’s eyes adapt themselves to the luminance of the external landscape. Internal gains consist of: sensible and latent thermal flow from occupants (4 people  65 W of sensible thermal power and 65 W latent), office devices (4 computers and 1 printer for a time average total power equal to 300 W) and fluorescent lamps (luminous efficacy: 91 lm/W, total power: 732 W). To calculate the primary energy demand related to heating, ventilation and air conditioning (HVAC), it is assumed that the room is equipped with a full air centralized loop, and the daily time of utilization is from 09:00 (but the plant is activated one hour before) to 19:00. Although it is not the best efficient solution, it is assumed that the warm fluid is provided by a gas-boiler and the cold fluid by an electrically driven chiller (vapour compression chiller). Internal set-point temperatures are assumed to be 20 °C in winter and 26°C in summer (as prescribed by the Italian law), whereas in the middle season it is assumed equal to the daily average external temperature, because the clothing of the occupants is adapted to it. The relative humidity setpoint is assumed equal to 50% all over the year. (a) The office building in its urban context, image from Google Earth (b) the actual Figure 3 configuration of the building (courtesy of Zintek Srl). (a) Geometrical model of the physical system, the figure shows the position of the Figure 4 examined room inside the building (b) examined room model with workplaces. THE EXAMINED SOLAR CONTROL DEVICES AND THEIR CONTROL LOGICS The following solar control devices and related control logics are analysed and compared with reference to this office room. External adjustable diffusing slat system. The vertical distance between slats (0.5 m) is equal to their depth (normal to the façade when the slat is horizontal). Slats surfaces are diffusing and their total 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 332 reflection coefficient is equal to 0.6 for both the sides, whereas mirror-like reflection coefficient is assumed equal to zero. All the reflection coefficients used in this paper pertain to the total solar spectrum as well as the visible range. Slats are controlled by a seasonal logic: in each moment, slats are inclined at an angle that allows the entering of the only solar energy fraction that can contribute to cover the sensible thermal load, and avoiding overheating. Anyway, the entering solar radiation cannot be lower than the one required for daylighting, ensuring a minimum illuminance value (500 lx according to Italian standard UNI 10380) in the most critical workplace. Double array of external slats. This system consists of two slat arrays. The upper slat array starts at 1.7 m from the floor and its slats have a mirror-like reflecting upper side (total reflection coefficient: 0.9, specular reflection coefficient: 0.9), whereas the lower side is absorbing (total reflection coefficient: 0.1, specular reflection coefficient: 0.0), to prevent downward reflected beams from entering the room. These slats are controlled in order to redirect the solar beams upwards inside the room, avoiding direct radiation on occupants and visual tasks. To avoid the entry of direct radiation in winter, the vertical distance between slats is reduced to 66% of their depth (i.e. 0.5 m  0,66 = 0.33 m). The lower array is equal to the diffusing array described above. Also in this case the control logic is seasonal, but a little more complex. As a first step, the tilt of the upper array is adjusted so that the larger part of the incoming radiation is redirected upwards, whereas the tilt of the lower array is adjusted to allow the solar radiation to enter the room in the amount needed for heating purpose. When the lower array is completely closed and the entering radiation exceeds the required value, the upper array is adjusted to reduce it as well. This adjustment stops when the incoming radiation reaches the minimum value necessary for daylighting. Double skin façade. In this case one laminated external glass (0.01 m thick) is present at 0.9 m from the previously described glazing. In the 0.9 m void a double slat array take place, close to the external glass. It is similar to the one described above, but in this case the slats, protected by the external glass, can be smaller and less robust. In particular, the lower array of diffusing slats can by packable to allow the entry of a higher solar power when required. The slat control logic is the same as for the double array system described above. During the cooling period the void is ventilated and it is assumed at the same temperature as the outdoor air, whereas in the heating period the minimum ventilation necessary to avoid condensation problems is provided, and the temperature of the air in the void assumes an intermediate value between the indoor and outdoor air temperatures. Figure 5 (a) Schematic behavior of the double array of slats and (b) slats incorporating PV cells. External slats incorporating PV cells. Two different types of slats incorporating PV cells were studied. In the first configuration (type A), the vertical distance between slats (0.5 m) is equal to their depth, but the PV cells occupy only 66% of the slat upper surface (and at the outdoor side), because in this configuration the remaining part is shaded for the most of the time. The reflection coefficient value is assumed equal to 0.2 for PV cells and 0.8 for the remaining slat upper and lower surfaces; therefore the average value is 0.39 for the upper surface. In the other configuration (type B) the vertical distance between slats is the same but the depth is reduced by 33%. This way a fraction of the direct solar radiation too enters the room during some winter periods and lowers the need for artificial lighting. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 333 Moreover, the PV cells are less shaded, in particular as regards sky diffuse radiation, and their electrical production is higher. In both cases (A and B), initially, the control logic was finalized to maximize only the electrical production of the PV cells; therefore, the angle of incidence of solar beams on slat surfaces (the angle between the solar beam and the normal to the slat surfaces) was always minimized. Then, for type B, this logic was modified in order to allow the radiation needed for daylighting to enter the room, when available. This modification lowers relevantly the total primary energy demand, since the reduction in electrical production (limited to some hours of winter days) is negligible when compared with the achieved reduction in energy demand for lighting and consequently for cooling. In all of the solar control systems described above, when the entering solar direct radiation can cause glare, an internal diffusing blind is lowered; it consist of a dark (grey 80%) metallic foil microperforated for 50% of its surface. Its light transmittance is equal to 50%, but it reduces the solar heat gain by 10% approximately. As a matter of fact, the energy absorbed by the foil is re-emitted as infra-red (IR) radiation, that cannot transfer through the glasses, and only the half part, approximately, of the short-wave radiation reflected from the foil can do it. This blind is used only in the heating period, since in the rest of the year the slats block the solar beam radiation. ANALISYS OF THE RESULTS The solar control devices are compared under two points of view: room total primary energy demand and global comfort conditions. Therefore, the thermal flows provided to the room by the plant, are converted into primary energy as a function of current values of boiler efficiency, chiller coefficient of performance (COP), and global system efficiency. The electric energy absorbed by the HVAC system as well as by lighting system is converted into primary energy by means of the related Italian electric system conversion efficiency (equal to 36%). The electric energy generated by PV cells is converted into equivalent primary energy using the same conversion coefficient and is subtracted from the total primary energy demand. For this reason, in some periods, the total annual primary energy demand appears as negative as shown in Figure 7 (b). (a) Diffusing slats controlled by seasonal logic, room sensible heat flows (W) on a Figure 6 winter day (January 21) and (b) on a summer day (July 21). Energy performance. The room under investigation is characterized by relevant internal gains and more than the half part of these comes from lighting system (732 W), when turned on. For this reason the cooling loads, present in winter too, are dominant in the composition of primary energy demand. In facts, with the exception of the first one/two hours at some winter mornings, when lamps are turned on cooling loads usually take place in winter too. For this reason, the level of daylighting achieved inside the room is decisive to define the suitability of a solar control system. Probably this situation can change adopting different lighting system, mainly consisting in task lighting by means of individual lamps located at each workplace (the related power can be reduced from 732 to 320 W), but the simulated lighting system is actually far more diffuse in Italian office buildings. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 334 Among the devices without PV cells, the diffusing slats are the most energy efficient solution. The external double array with specular slats is less convenient, because of the lower total incoming radiation during the winter, consequently the artificial light is used for longer time and cooling loads are higher. The double skin system is fairly less convenient than the last one: the lower thermal losses reduce the energy demand for heating, but this benefit is overcome by the increase of energy demand for lighting and cooling. (a) monthly primary energy demand (per square meter of floor area) related to HVAC Figure 7 and lighting systems for various devices (b) total monthly primary energy demand, including PV 2 generated electricity (with negative sign) [kWh/(m floor∙month)]. The annual electrical production of the devices incorporating PV cells is in the same order of magnitude as the room total energy demand, thus they are more convenient than other solar control devices. The first kind of device (type A) is controlled only to maximize the electricity generation, consequently the entering solar direct radiation is always blocked by slats and the entering solar diffuse radiation is reduced too. Therefore artificial light is turned on for longer time and, for short periods, useful solar gains are minimized, whereas the consequent increase in energy demand for lighting and HVAC corresponds to a small fraction (29%) of the primary energy equivalent to electricity generation. Anyway, the best energy effective system is the one with less deep slats (type B), which allows a larger amount of solar radiation to enter the room, in particular when it is controlled in order to minimize the use of artificial lighting. In this case, in some months the room energy balance is positive: the primary energy equivalent to the PV electricity generation is higher than the room energy needs. (a) spatially averaged monthly PMV values with different solar control devices (b) Figure 8 spatially averaged DGI values in in a summer day (July, 21) with different solar control devices. Luminous comfort. A first comparison can be done on the basis of the total number of hours in which the DGI value is out of limits for at least one occupant and the internal blind is lowered. The limit value for DGI is assumed equal to 21 according to Italian standard (UNI 10840). Following this criteria, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 335 the external double array with mirror-like slats appears as the most comfortable configuration (the blind is used only in the 4% of the time) followed by the other two devices without PV cells, whereas all the devices incorporating PV present a worse behavior (the blind is used in the 37-57% of the time). In effect The DGI value exceeds the limit when the more luminous part of the sky is extensively visible, or when the internal average luminance is very low and the contrast with the visible sky is high, as it occurs in case of slats incorporating PV cells, particularly in type A, that assume every time the tilt necessary to block solar direct radiation. Another kind of discomfort can be caused by the solar beam radiation impinging on visual tasks; this is more frequent (during the winter) in case of double slats arrays. At last, the two devices equipped with specular surfaces provide the higher uniformity of internal illuminance. Thermal comfort. Using the internal air temperature as the indoor environment control parameter, the differences in the thermal comfort are mainly influenced by internal surface temperatures. Comparing different devices: comfort conditions are generally better with devices not provided with PV cells and controlled by a seasonal logic, as shown in Figure 8 (a). In these cases the temperatures of internal glazing surface and internal surfaces exposed to the incoming radiation are higher during the cold season and lower during the warm season. The devices incorporating PV cells intercept direct radiation for almost all the time, but in the cooling season the entering radiation reflected by the slats is higher than in the other cases. Consequently, the internal temperatures are lower during the cold season and higher in the warm season. The spatial uniformity of PMV values is generally high (the differences are lower than 17% in winter and 2% in summer); particularly in the case of double skin (the differences are lower than 11% in the colder month too). This is due to the glazing internal surface temperature that is closer to the internal air temperature. CONCLUSION Overall, the less deep slat system incorporating PV cells (type B) is the most energy efficient device, particularly when controlled in order to allow the solar radiation necessary for daylighting to enter the room whenever available. In this last case, the PV electric generation is slightly reduced during some winter periods, but the problems connected to artificial lighting (such as lamps energy demand and cooling load) are appreciably reduced. Moreover in the room considered, characterized by relevant internal heat gains, the reduction of useful solar gains concerns only some short winter periods. However all the systems provided by PV cells present worse performance regarding visual and thermal comfort. To avoid this problem it is possible to forecast some possible device evolutions with the purpose of combining the best energy and comfort performances, such as the following ones. 1. An array of mirror-like reflecting slats able to change the vertical distance between the elements, depending on the Sun position, thus entering solar radiation will be reduced only as a function of slats reflecting coefficient, but this kind of device is currently not available. 2. A double slat array with mirror-like reflecting slats in the upper part and slats incorporating PV cells in the lower part, controlled by the logic of type B. NOMENCLATURE PMV PPD DGI = = = Predicted Mean Vote (PMV) (Fanger, 1970), Predicted Percentage of Dissatisfied (PPD) (Fanger, 1970), Daylighting Glare Index (DGI) values, Hopkinson et al. (1963). REFERENCES Carbonari, A. 2012. Thermal and Luminous Comfort in Classrooms: A computer method to evaluate different solar control devices and its operating logics. Proceedings of PLEA 2012 - The 28rd Conference, Opportunities, Limits & Needs Towards an environmentally responsible architecture, Lima, Perù, 7-9 November 2012. Fanger, P.O. 1970. Thermal Comfort. New York: Mc Graw-Hill. Hopkinson, R.G., P. Petherbridge, and J. Longmore. 1963. Daylighting. Heinemann, London. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 336 Tuning Houses through Building Management Systems Jose Ripper Kos, PhD Massimo Fiorentini, PhD Candidate Federal University of Santa Catarina University of Wollongong Paul Cooper, PhD Felipe Miranda, Architecture Student University of Wollongong Federal University of Santa Catarina ABSTRACT This paper departs from an analogy of sailing race instruments to demonstrate the potential of automation systems on the house performance and, more important, on impacting households for a more sustainable behavior. Sailing instruments have positively influenced the results on experienced sailors’ speed and ultimately have confirmed their observations on nature cycles. We have presented two research projects for the development of Building Management Systems for a house that relies mostly on natural ventilation and thermal mass and another one, based on a complex conditioning system with a solar assited HVAC system, connected to a Phase Change Material thermal storage. Our argument is that if research on the sailing tournament America’s Cup instruments soon became available to other sailing boats, systems developed to the Solar Decathlon houses’ academic competition could, and should, be accessible to a great number of home owners. These two research projects give evidence that further research should be guided to more sustainable BMSs, which could significantly contribute to households’ behavior changes and ultimately support dwellers reconnection to natural cycles. INTRODUCTION Burry, Aranda-Mena, Alhadidi, Leon and Williams (2013) presented a challenging statement that “compared to architecture, performance is more transparent in a high-performance sport such as sailing where it is clear that ‘speed is good’.” Although their approach to sail racing and architecture focus towards a different perspective, we would like to use it as a starting point to elaborate our argument. Sailors can teach us about having their boat perfectly regulated. A connection between sailing and buildings has been presented by different authors, with a broad range of approaches (Murcutt, 2008; Lynn and Gage, 2011; and Burry et al, 2013). We believe that a house could be tuned in a similar way as sailors trim their sails and plan their racing strategies. Experienced high-performance sailors’ success depends on their ability to understand the boat functioning and environmental changes and cycles. We would argue if it would be possible to stimulate in dwellers similar abilities found in these sailors. “... And you work this house and you work most of my buildings like you sail a yacht. You have to work them so that you understand how to get the best out of the climate without having to aircondition." (Murcutt, 2008) In this remark, offered in a TV interview, Murcutt does not present a revolutionary concept. Actually, he raises a primitive concept forgotten in most contemporary houses. Contemporary homes are no longer able to communicate their operation systems. The systems’ complexity is usually hidden from the dwellers and does not support the latter engagement. Primitive communities’ homes, on the other side, clearly displayed their technologies that united them to their Jose Kos is a professor and Felipe Miranda, a student in the Department of Architecture and Urbanism, Federal University of Santa Catarina, Florianopolis, Brazil. Massimo Fiorentini is a Mechanical Enginner and PhD Candidate and Paul Cooper, a professor in the Sustainable Buildings Research Centre, University of Wollongong, Australia. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 337 place. Primitive dwellers, as sailors, understood their technologies, building performance and the direct environmental influence. High performance sailing boats have embedded a great variety of technologies. These technologies support sailors’ connection with the natural environment and boat performance awareness. Several of these technologies are brought by research for high-performance and high-cost competition enterprises, such as the most famous America’s Cup. “All through the 1980’s the America’s Cup was contested in Twelve Meter Yachts, and significant advances were being made in hull construction, sailcloth and panel layout, and in sailing instruments systems. But, perhaps more than the others advances, sailing instruments were beginning to change how boats were sailed. The information regarding wind angles and speed were better than ever, but being able to make calculations which could indicate how efficiently the boat was being sailed was what was changing the game for the world’s best sailors.” (Ockam, 2013a) Sailors use polars as predictions of the ideal boat speed across a range of wind angles and speeds. Several sailing instrument systems can compute and display, in real time, polar plots of the boat target speed in the given current conditions. They offer to the crew information that allows them to judge how they are performing against the boat’s speed potential. Their precision is quite important to the point that sailors such as Dennis Conner account for the polars’ precision to their success in races. (Ockam, 2013b) A dwelling performance is quite different from race sailing boats. The latter has to reach the destination before the other competitors. Boat speed is not the only parameter to define the winner, but it is surely decisive and sailors have always to search for the best speed in the particular conditions or strategies. In a house, users should seek for a sustainable comfort. Home sustainable comfort presents different variables to distinct dwellers and to their various activities (de Dear and Brager, 2002). Thus, there is no ideal performance modelling that could apply to everyone. Following this line of thought, it is important that users identify their comfort zone that could be applied to control natural and mechanical conditioning methods and even lighting levels. Cruising boats do not have to beat competitors, but are highly influenced by innovations developed by high performance sailing. These technologies afford cruisers with a combination of boat performance, weather data and location conditions. These instruments do not move away sailors from environmental awareness, as building technologies have done in the past. On the contrary, they facilitate an understanding of environmental data as an integrated natural network. Primitive dwellings have been developed as devices adapted and adaptable to environmental conditions. Dwellers could feel wind shifts, see changing colours and smells. Weather could be predicted and the house prepared for extreme conditions. Modern building technologies have supported comfortable homes, better protected from weather extreme conditions, less dependent to environmental cycles, as well as more energy dependent. We are so disconnected from environmental cycles that it is a difficult task to engage users in more sustainable habits. Sailing technologies can teach us to better understand environmental data in order to recognise our place on earth and better adapt our habits to environmental cycles. Most of our home technologies have been developed in times when energy and environment were not a major concern. We do not need to get rid of technology, but develop more efficient systems able to better connect our buildings and users to natural environment. Sailing technologies have performed this movement mostly because they have always depended on nature’s cycles. Information has been a key issue. We have lost the use of our senses to acquire it, as primitive dwellers did once. Therefore, systems integrated to environmental data could support engaging residents to dwell with more sustainable habits. This paper explores the development of two residential Building Management Systems and their user interface. The houses and their systems are quite different. They were designed for different continents (Australia and Brazil) but share some climatic similarities, as well as analogous relationships identified in high performance sailing systems and their potential impact on cruising boat technologies. Although the two management systems have broader applications, particularly related to energy generation and management, the discussions in this paper will focus on their thermal comfort systems. One of the houses relies on a hybrid and innovative HVAC system, while the other applies, with the exception of few days during the year, on passive conditioning and ventilation methods. BUILDING MANAGEMENT SYSTEMS Building Management Systems (BMS) applied to houses have become relatively cheap and 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 338 widespread. However, they are mostly limited to integrate few home controls usually related to lighting, thermal conditioning, media and security. These systems are able to incorporate powerful applications similar to those found in sailing instruments. These applications should be more explored to afford updated data about house performance, as well as weather data. They may also support understanding of the house functioning, which has been lost throughout history, as our houses became more technologically complex. In order to provide a meaningful environmental impact, the houses should not only be technologically holistic. They should appear holistic to their users. House performance depends heavily on users and their awareness. The reduction of energy consumption, for example, depends on technologies but also, and heavily, on choices taken by users (Schipper, 1989; Socolow, 1978). Building design and its technologies have the potential to involve and educate. Orr (1997) placed the argument that our buildings miss their potential to reflect a hidden curriculum embedded in design choices. He asked if "[t]hrough better design is it possible to teach our students [users] that our problems are solvable and that we are connected to the larger community of life?" Following his advice, designers should care about how dwellers perceive and understand their homes. Houses, as any building, should be adaptable to different environmental conditions and residents should aspire to have them tuned. Sensors, actuators and BMSs’ interfaces could act similarly to sailing technologies, in order to facilitate adjustments for tuning the house. They are able to provide updated data of the house performance and its systems, implementing automated functions towards dwellers awareness. If they become widespread, they could drive the development of affordable systems even for low-income housing. Although the experiments described below apply to relatively expensive houses, they demonstrate systems with a great potential for a relevant impact on more sustainable home behaviour. THE ILLAWARRA FLAME HOUSE The Illawarra Flame House is a small and high performance house developed for the academic competition Solar Decathlon (US Department of Energy, 2014), organized for the first time in Asia. Team UOW, lead by the Sustainable Building Research Center, University of Wollongong, designed, constructed, brought the house to Datong, China and won the Chinese Solar Decathlon, in 2013, attended by a total of 22 teams representing recognized universities around the world. Team UOW has demonstrated a remodeling and retrofitting of a common and archetypal Australian house built approximately 50 years ago. The aim is to inspire national building industry and the general community that it is possible to transform the vast majority of Australian homes into stylish, affordable, and sustainable homes of the future. By upgrading an existing house, Team UOW took up the challenge set by the U.S. Department of Energy, China National Energy Administration, and Peking University’s goals to "accelerate the development and adoption of advanced building energy technology in new and existing homes". With less than two per cent of Australia’s housing replaced each year the Australian team believes that this retrofitting approach has the greatest practical potential to achieve significant economic and environmental gains across the country domestic built environment. The Illawarra Flame showcases a radical, affordable and achievable blueprint and benchmark for retrofitting a typically Australian ‘fibro’ house. Fibro refers to cladding sheets constructed of asbestos fibre-reinforced cement. They are ubiquitous to Australian suburban streets. In addition to environmental concerns, recent increases in energy prices, the health implications of asbestos, and the poor thermal performance of these houses, have led to an urgent need for widespread upgrading. (Team UoW, 2013) The Illawarra Flame features a Solar Assisted HVAC system that integrates an air based Photovoltaic-Thermal (PVT) system, a Phase Change Material (PCM) thermal energy storage and a standard reverse cycle air conditioning system. The house holds two types of Photovoltaic panels, a 1st Generation Polycristalline 5kW array on a 5kW MPPT Inverter and a 4.6kW Thin Film CIGS array on a second 5 kW MPPT Inverter. The CGIS array constitutes the Illawarra Flame’s Photovoltaic-Thermal (PVT) system: a number of thin-film PV panels mounted on a steel sheet flashing that is fixed to the top of an existing sheet metal roof profile (Figure 1). This system creates a cavity underneath the steel flashing through which the working fluid, air, can flow and exchange heat with the PV panel. The 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 339 advantages of a PVT system rather than a PV system include an increase in the efficiency of the PV panels by reducing and regulating their temperature and the possibility to extract or release heat for heating or cooling purposes. This process increases the total energy extracted from the solar system, therefore improving the overall efficiency of the system. Figure 1 Solar Decathlon House PVT system The performance of the Illawarra Flame PVT system is enhanced by coupling it with a PCM thermal energy storage system. By using this thermal storage system it is possible to phase-shift the thermal generation so that thermal energy may be released at times when generation is not possible. The system can be used both for heating and cooling. Further cooling of the ambient air can be achieved in the PVT system, since during clear nights, it radiates heat to the sky. The Illawarra Flame BMS is based on an off-the-shelf residential control system, Clipsal C-Bus. This control system manages different systems implemented on the building, including the solar assisted HVAC system, electrical generation and distribution and automated high level windows for natural ventilation. To achieve this goal, readings from a Davis Vantage Pro 2 weather station, integrated in the control system, support logic decisions and inform users through graphic interfaces. (Figure 2) Figure 2 Solar Decathlon House BMS and SD house interface Managing and displaying information are key features of BMS systems. One of the objectives of the Illawarra flame BMS is to let households be aware of the house performance, educating them on the different aspects of managing energy in the house. The user can easily access the consumption of each electrical subdistribution circuit and try to target a reduction of the most energy consuming appliances. One can see how much energy can be harvested by the solar assisted HVAC system and which working mode the HVAC is operating at each time to maximise its efficiency. Integrating the readings from the weather station, the BMS displays outdoor conditions and indoor performance. Therefore, users can take the best decision to guarantee comfort with the use of natural ventilation or reducing the energy associated to conditioning with the different mechanical systems. The interface botton part displays a gray strip that identifies the comfort zone or better conditions for each variable. The coloured bars indicate the variables interval in the previous 24 hours and a filled dot, inside these bars, expresses the current measurement, facilitating quick readings by lay people. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 340 The BMS system controls on/off dampers, variable position dampers and variable speed fans, to set the system in the correct working mode. The system considers the generation and storage status with the house heating/cooling demands, optimising it to achieve the required thermal comfort with the maximum efficiency. The logic has been custom developed in Pascal language on the house C-Bus Pascal logic controller. A touch screen and a web supervisory engage the users in the systems performance displaying generation data of both arrays as well as thermal storage. The Illawarra Flame house comfort conditions are maintained using the feedback reading from 5 independently calibrated temperature sensors. Due to the small size of the house, the average temperature of the conditioned space is used for decision-making. The individual temperatures are fed back to the user though the touchscreen, as well as the average one. The comfort can be achieved using mechanical heating or cooling, using the different working modes of the Solar Assisted HVAC system or, every time outdoor conditions are favourable, using natural ventilation. In this case the BMS will open automatically the high level windows and advise the user to open the non-automated windows. Automatic opening of windows can be always be overridden by the user using the touchscreen or wall pushbuttons. Households can also define different temperature setpoints that overwrite the system Auto Mode. Predicted energy consumption is provided by each temperature user setpoint, who can compare with the prediction for the Auto Mode. THE FLORIANOPOLIS HOUSE The Florianopolis House is located in the south of Brazil, within an island, more than 50km long and around 20km wide. It is near a beach, which has few elevations and temperatures slightly lower than the interior of the island, ranging from 7,5°C and 31°C, with around 1600mm of annual rain precipitation, 85% of annual relative humidity and about 140 days of rain per year (Lamberts et. al, 2010). Rainwater is therefore an important resource to be highlighted. The rainwater falling into the pool, with bamboo trees on the background (Figure 3), is enjoyed in the living room and collected for water reuse. The pool amplifies changes on the cycles of nature, such as sunny or windy days. Figure 3 Views of Florianopolis House from inside the lot and from the back. Comfort inside the house is ensured by two main strategies: significant amount of thermal mass protected by insulation and shading devices as well as cross-ventilation in the direction of the region prevailing winds (North, Northeast and South). These strategies should be managed by the residents, to avoid opening the house in hours of extreme heat. On warmer days, upper windows are opened at night and, on cold days, closed windows let the warm sun in. The main facade, facing the pool is oriented N/NE, with large windows and doors that ensure ample lighting and sunlight in the colder months. Wooden panels on the upper balcony guarantee shading the rooms, providing more privacy to the bedrooms, when superimposed. The upper balcony provides protection to the hot sun during the summer, allowing it in to warm the colder days. The BMS developed for the Florianopolis House is integrated with a Davis 6250 Vantage Vue weather station and an Internet weather forecast provider. Connected to different home sensors, it 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 341 supports dweller awareness and their decisions. Some tasks are automated, however, the system most significant contribution is to provide a holistic comprehension of the house systems and the environment, building up users’ concern to perfect house usage towards a more sustainable living. Temperature sensors located in the living room on the lower floor, the top of the stairs, and the master suite show the temperature difference in these rooms and allow for comparison with the external temperature. (Figure 4) Different colours display the relationship between these four temperatures from dark blue (cooler) to red (warmer). The green circle displays humidity. A graph of the variation of the four temperatures (living room, roof, suite, and outdoors) allows checking of the performance of the house in the last 24 hours, confirming the adequacy of adopted strategies and data from the local weather station, associated with progressive variation of colours facilitates reading and comparison. Figure 4 Screen for temperature and upper window control and bar graph. With the association of temperature data from these sensors with information from the local weather station, such as speed and wind direction, one can plan a strategy for the next hours or days, with the support of the weather forecast. Three upper windows open above the roof, helping to extract the interior hot air. Those windows are motorized and controlled by the BMS. Therefore, it is possible to remotely open them. They close automatically on days with rain and strong southerly winds, to avoid water entering into the house. Combined with the other windows, they ensure cross-ventilation, which is particularly effective in such humid climate. The three-floor height central volume, formally emphasized by the use of solid brick, facilitates the extraction of hot air from the ground floor. In warmer weather, ceiling fans in bedrooms alleviate the thermal sensation, without mechanical air conditioning devices. A house with a significant amount of thermal mass requires some strategy for managing its openings. They are not a direct response to the current conditions, similar to sailing boat that often take a lower speed direction that would prove to be more effective in a longer term. Therefore, the decision is not only based on the difference between interior and exterior temperature, but also a prediction of extreme higher or lower temperatures. Local information is associated with a three day weather forecast, allowing dwellers to plan strategies to ensure internal comfort within the house in the following hours or days. Data displayed through the graphic interface information would suggest the choice of windows to be opened or closed or the use of additional systems, such as the roof evaporative cooling or, in extreme conditions, the air-conditioning. The bar graphs are aimed to support residents to learn with and about their house. The house systems become more transparent once the users find their location and performance, leading to more sustainable procedures and habits. The BMS and particularly the supervisory graphic interface are critical for this aim. The interface integrates numeric information, such as air temperature, with scalable vector graphics (svg). Therefore, one easily and quickly distinguishes temperature variation, translated into colour, in different areas of the house as well as outside temperature. The interface is optimally visualized on an iPad or a computer screen, although it can also be controlled through a smart phone. Data is organized in four main groups: energy, water, weather and security and background image changes also according to the menu item. Managing water and energy resources are also important 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 342 aspects of the system, but will not be detailed in this paper. The open source supervisory control and data acquisition (SCADA) system is a critical component for the Florianopolis House BMS. The open supervisory (SCADABR) has been developed at the Federal University of Santa Catarina for a broad range of uses. Choosing an open source tool was one of the research goals. A relevant objective of the BMS research project is to test sustainable management solutions for low-income housing. Reading information is critical to maximize house performance, adapted to the residents’ needs and sensors’ data increasingly adds complexity to the interface. Thus, data has to be organized by hierarchy importance relating to different dwellers understandings. CONCLUSION This paper does not aim to provide a result of the two systems’ impact evaluation on households or on energy use. One of the reasons is that the systems presented have just been implemented. In addition to that, the two houses are not targeted for regular families. The Illawarra Flame House was erected at the Wollongong University campus. The house will most probably be used by different guests. The other one is the home of one of the researchers, who cannot represent a regular use of a house. Therefore, the authors depart from an analogy of sailing race instruments to demonstrate the potential of automation systems on the house performance and, more important, on households’ behavior. Similarly to the most important sailing races, such as the America’s Cup, the Solar Decathlon competition has provided relevant and innovative experiments on home automation oriented to sustainable house performance. Our argument is that if research on America’s Cup instruments soon became available to other sailing boats, the systems developed to the Solar Decathlon houses could, and should, be accessible to a great number of home owners. The two systems illustrated in this paper convey a further development of Solar Decathlon research. The system accomplished for the Chinese Solar Decathlon continued to be improved after the competition and is presented in its current stage. The system for the Florianopolis House departed from the system of the Brazilian Solar Decahtlon 2012 house, embodying its concept and some of its features. Instead of hiding the complexities of all the houses’ systems, both of them attempt to present and understanding of the house functioning, engaging and challenging the users for a better performance. We argue that home automation systems have a relevant potential that is not explored by the market. In addition to reduce the effort to perform some of the daily dwellers’ tasks, these systems are noteworthy tools for more sustainable home behavior. Although the BMS design of the two houses followed similar principles, they highlight their houses specificities and display both their houses and BMSs’sustainable features. The complexity of the Solar Decathlon house has offered increased possibilities for testing more innovative approaches. On the other side, the main contribution of the Florianopolis house BMS to engage users is through history graphs, data translated to colors and the 3D section renderings. Apprehending the impact of habit changes in the house performance is not an easy task and is critical to engage users. Sailors can check their instruments in order to evaluate a different sail regulation. Our houses have a much slower response and instant data from energy consumption, for example, does not directly translate the impact of changing an activity. Overall daily energy consumption, as well, displays several different circuits that may considerably vary during the day. Thus, history graphs with hourly consumption of two different days may facilitate the identification of the impact of an activity’s change, comparing data from the current and the previous days. Different patterns of opening windows during two similar and consecutive days can present contrasting temperature results in the rooms. Similar history graphs with hourly temperature bars of different rooms are very educative about the house performance. Users can benefit from learning to better use their houses with a much more pleasant comfort outcome, reducing environmental impact. A good design has also a unity. A sustainable house requires sustainable households. Connecting the latter with the house unity should be a critical aim of BMSs. The Solar Decathlon house’s BMS presents some additional contributions. The comfort zone bar facilitates the readings of the house performance and climatic conditions. Usually, households do not directly identify comfort zones in their daily use of the house. Air conditioning temperatures, for 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 343 example, are many times set up after climbing stairs when arriving home. The comfort bar is not effective for all variables, for some, it may be decisive. Temperature, humidity, wind direction and spped, and even luminosity should be influential. They are often ignored and can be quite misleading even for more sustainable contemporary households. The possibility to identify if these variables are within the comfort zone supports more sustainable dwellers’ behaviour, as well as a confirmation of their observation of the influence of natural cycles in the house performance. The Illawarra Flame House BMS introduces also an analogous tool of the target speed celebrated by high performance sailors. The system calculates the energy consumption in the next 24 hours for the Auto Mode, as well as set point changes for heating and cooling. Therefore, the Manual mode is not only based on the assumption that the user has of a comfortable condition. The system actually provides a benchmark of an optimal Automatic Mode condition using every house resource to reduce consumption. Therefore, changes provided by the users could be predicted and compared against the ideal Auto Mode. They have ultimately all information for reducing consumption changing temperature set points. Sailing instruments do not reduce the connections between good sailors and natural environment. On the contrary, they support experienced sailors with a confirmation of their empirical observations of nature. Contemporary urban dwellers have diminished their connections to natural environment. Technologies developed over the last centuries have reduced the need to adapt their lives to natural cycles. Building standardization has also contributed to this tendency. Studies from primitive communities highlight these remarks. They have developed their homes and the way they inhabit them based on empirical observations. Their habits are directly associated to the building performance, particularly in cases of extreme climate conditions. Experienced sailors have not lost, throughout history, their ability to understand natural cycles. We have aimed, through this paper, to advocate that these systems have added precision and confirmation to sailors’ observations and they could support similar connections that urban dwellers lack. ACKNOWLEDGMENTS We would like to acknowledge the Brazilian Ministries of Education (CAPES) and of Science Technology and Innovation (CNPq) and the Sustainable Buildings Research Centre - University of Wollongog for their support to this research. REFERENCES Burry, J., Aranda-Mena, G., Alhadidi, S., Leon, A. and M. Williams. 2013. Design Trade-off: Sailing as a vehicle for modelling dynamic trade-off design, in J. Burry, Designing the dynamic: Highperformance sailing and real-time feedback in design. Melbourne: Melbourne Books. 17-23. de Dear R. and Brager. 2002. Thermal comfort in naturally ventilated buildings: revisions to ASHRAE Standard 55. Energy and Buildings. 34 (2002): 549–561. Lamberts, R., Ghisi, E., Pereira, C. and J. Batista (eds.). 2010. Casa eficiente: Bioclimatologia e desempenho térmico. Florianopolis: UFSC/LabEEE. Lynn G. and M.F. Gage (eds). 2011. Composites, Surfaces, and Software: high Performance Architecture. New Haven: Yale School of Architecture. Murcutt, G., and P. Thompson. 2008. Transcripts from TV interview to ABC Talking Heads, Retrieved from: http://www.abc.net.au/tv/talkingheads/txt/s2256196.htm. Accessed 9, December 2013. Ockam Instruments, Inc. 2013a. Twelve Meter Yachts and their Sailing Instruments. Retrieved from: <http://ockam.com/2013/02/23/twelve-meter-yachts/> Accessed 9, December 2013) Ockam Instruments, Inc. 2013b. Everybody talks about them, but what are they?. Retrieved from: http://ockam.com/2013/06/03/what-are-polars. Accessed 9, December 2013. Orr, D.W. 1997. Architecture as pedagogy II, Conservation biology. 11(3): 597–600. Schipper, L., Bartlett, S., Hawk, D., and E. Vine. 1989. Linking life-styles and energy use: a matter of time?. Annual review of energy. 14(1): 273–320. Socolow, R.H. 1978. Saving energy in the home. Princeton's Experiments at Twin Rivers. Cambridge: Ballinger. Team UoW. 2013. Project Manual. Solar Decathlon internal document. Datong. US Department of Energy, 2014. Solar Decathlon. Retrieved from: http://www.solardecathon.org. Accessed 9, December 2013. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 344 Adaptive comfort and control protocols for natural ventilation Santosh Philip1, Abe Shameson1, Nathan Brown1, 2, George Loisos1, Susan Ubbelohde1, 3 1 Loisos Ubbelohde, Alameda, CA 2 California College of Arts, San Francisco, CA 3 University of California, Berkeley, Berkeley, CA ABSTRACT Recently there has been a shift in international standards towards adaptive thermal comfort. ASHRAE had formally incorporated adaptive comfort into ASHRAE Standard 55-2010. In adaptive comfort, the range of indoor temperatures perceived as comfortable drifts upwards in warm weather and downwards in cooler weather. This is especially true where the occupants have control over their thermal environment (for example, with clothing or operable windows) and when occupants are aware of the variable outdoor conditions. This paper presents the ventilation control strategies used to provide adaptive comfort without air conditioning (compressive cooling) in a major retrofit for an office and classroom building on the University of Hawaii campus. The building retrofit is designed for cross-ventilation while taking care to acoustically isolate the spaces from each other and to dampen sound from outside. A supplementary fan system is designed and configured to complement the cross-ventilation strategy. The fan system is controlled to actively draw just enough air to provide adaptive comfort conditions when wind is calm. If the wind is sufficient, the controls sense the speed and direction and shut down the fans allowing cross ventilation to do the cooling. The fan controls sense the conditions inside, conditions outside the building, and weather conditions in the recent past. INTRODUCTION Kuykendall Hall at the University of Hawaii at Manoa (UHM) was constructed in 1964. It is made up of two building blocks connected by a bridged walkway. One block is a 56,000 sqft four story classroom building. The other block is a 20,000 sqft seven story office tower. Kuykendall Hall was originally built as a naturally ventilated building. Corrosion of window mechanisms and acoustic issues let to renovations in 1987 which made the entire building air conditioned. Due to the extreme high cost of energy in Hawaiian Islands, the university would like to transition many of the buildings on campus to be not air-conditioned or only partially conditioned. Kuykendall Hall was intended to be a pilot project in the pursuit of this strategy. After an analysis of different strategies, UHM chose to pursue an overall design that emphasizes natural ventilation and ceiling fans for cooling and comfort. Extensive computer modeling and physical modeling were done as part of this process. The modeling strategies had two objectives here. One was to model in an explorative manner so as to arrive at the design solution. The second was to analyze the design to understand the energy implications. EnergyPlus version 6 was used to simulate all zones of the building to understand the thermal and comfort implications on a detailed hourly basis for the whole year. Eppy - a scripting language for EnergyPlus (Philip, Tran and Tanjuatco 2011) was used to automate this process Since the performance of ventilation was critical to the success of the project, a boundary layer 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 345 wind tunnel testing was used to explore the design alternatives. We tested a massing model that included the site context within about 600 ft. of the building, and detailed models to study air flow within specific areas of the building. Models were placed on the wind tunnel turntable and tested under 9 different wind directions. The results of this study were presented at the 2014 AHSHRAE/IBPSA-USA Building Simulation Conference (Philip, S., Shameson, A., Brown, N., Loisos, G., Ubbelohde, S., 2014). The modeling described in that paper predicted the energy saving illustrated in Figure 1a and 1b, but is silent on the control protocols needed to achieve those results. The present paper describes the design configuration and the control protocol for comfort Figure 1: Expected Energy End Uses for the ASHRAE building, the existing building and the proposed buildings ADAPTIVE COMFORT The comfort chart depicted in Figure 2 (a) taken from ASHRAE Std 55-2010 ASHRAE 2010. The percentages in the diagram can be explained in the following manner: an indoor operative temperature falling within the 80% range should be regarded as acceptable or satisfactory to at least 80% of building occupants who are exposed to it, and the tighter 90% acceptable temperature range is likely to satisfy 90% of occupants. (a) (b) Figure 2: (a) The ASHRAE Std 55-2010 adaptive comfort standard as a function of prevailing outdoor temperature. ta(out) is the arithmetic average of the mean monthly minimum and maximum daily air temperatures for the month in question. (b) The ASHRAE 2010 adaptive comfort standard in naturally ventilated spaces showing the upper (80% acceptability) limit based on mean monthly ET* ASHRAE's adaptive comfort was defined using the mean monthly dry bulb temperature (ta(out)). In this project we have used the average of the last 15 days instead of monthly mean since we had access to realtime weather data. Note Figure 2(a) has OT (Operative Temperature) on the vertical axis and does not take humidity into account. In the original ASHRAE research project (RP-884) that led to the development of the adaptive comfort standard, mean outdoor effective temperature, ET*, was used to arrive at the comfort zone (de Dear and Brager 1998). ET* is defined as the temperature at 50% relative humidity which would cause the same sensible plus latent heat exchange from a person as would the 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 346 actual environment. ET* values can be calculated using WinComf© (Fountain and Huizenga 1996). The modified figure is shown in Figure 2 (b) When using ET*, the optimum indoor temperature is given as: optimum indoor temperature = 18.9°C + 0.255 * outdoor_mean_ET* (1) Acceptable temperature ranges around the optimum in naturally ventilated buildings were specified as ±3.5°C for 80% acceptability and ±2.5°C for 90% acceptability. This corresponds to the two acceptability deadbands shown in Figure 2 (a). At the time this project was done in 2011, there was a proposed addendum to Std 55 to take in to account effect of airspeed on comfort. In this addendum, an increase in airspeed increases the acceptable operative temperature. This project uses the upper 80% acceptable limit, with an airspeed of around 1 m/s, always available through ceiling fans, allowing us to increase the acceptable temperature by 2 degrees C., COMFORT THERMOSTAT Traditionally thermostats have dry bulb temperature (DBT) setpoints. The traditional thermostat has a DBT sensor that checks if the room DBT is within the comfort range of the set points. In a cooling only model the thermostats tests if the room DBT is above the setpoint. If it is above the setpoint, action is taken to bring down the temperature in the room. Such a thermostat does not allow us to control for adaptive comfort, where the setpoint may change each day So a comfort thermostat was designed to respond to adaptive comfort. The comfort thermostat we are using has a number of elements that are different from the standard thermostat. In a comfort thermostat, the metric used is operative temperature (OT) and not dry bulb temperature (DBT). OT is the average of mean radiant temperature (MRT) and DBT. In the standard thermostat, the setpoint rarely changes. In contrast the comfort thermostat sets the setpoint based on the temperature of the last 15 days. This takes into account the acclimatization of the body to the season. The setpoint for any particular day is calculated in the following manner: o out_ET* = average of the ET* of the previous 15 days (2) o optimum indoor temperature = 18.9°C + 0.255 * outdoor_mean_ET* ( same as 2) o 80% accept = optimum indoor temperature + 3.5°C (3) o ceiling fan (1m/s air speed) = + 2°C (4) o Thermostat setpoint = optimum indoor temperature + 3.5°C + 2°C (6) If the OT is above the thermostat setpoint, steps have to be taken to bring the conditions back to comfort. The following sections describe the logic of how this is done. CONTROL PROTOCOL Objective: to keep the Operative Temperature (OT) in the comfort range. Since the site is in Hawaii, we are primarily looking at the cooling to achieve comfort. Whenever the OT is outside the comfort range, the control protocol should take action to bring it back within the comfort range. o Whenever the conditions are not comfortable within the space, the occupants can operate the windows to become comfortable. o If the Operative Temperature (OT) is too high, open windows will allow natural ventilation to cool the space o If the Operative Temperature (OT) is too low, the windows can be closed to keep the cold air out. This is less of an issue in Hawaii's climate o If the windows are open and there is no wind, there will not be enough air changes (Air Changes per Hour - ACH) to cool the occupants. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 347 o o o O Now the forced air fans will start up and provide the needed ACH The volume of forced air increases until comfort conditions are reached. The control protocol watches for wind conditions. If the wind picks up again, with sufficient velocity to bring about comfort, the forced air fans will shut down. If the control protocol is not watching for wind, the forced air fans will simply shut down after running for some time. When the conditions reach a point of discomfort, the fans will start up again will start up again. CLASSROOM VENTILATION The ventilation in the classrooms is either through cross-ventilation or through forced ventilation. When the air changes are happening through cross-ventilation, the forced ventilation part of the system is shut down. Similarly when the forced ventilation is running, the cross-ventilation part of the system is closed down. The geometry of the building forced the adoption of this strategy where the building is in either one mode or the other. Figure 3 Configuration for natural ventilation and forced ventilation in classrooms Configuration for natural ventilation: The classroom building has a double loaded corridor layout with classrooms on two sides of a central corridor. This makes it difficult to get cross-ventilation since only one side of each classroom is exposed to the outside. The proposed solution is to let outside air come in through a louvered sound attenuation box beneath the window and allow it to leave the classroom via a duct that runs over the opposing classroom and out the other side of the building. The air could flow in either direction. This is illustrated in Figure 3 above. There is a velocity sensor in the duct. Readings from this air velocity sensor can let us calculate the Air Changes (ACH) in the room. There is a damper in the duct that can completely close the duct, when the system switches to forced ventilation. In Figure 3, the middle floor illustrates the air flowing in one direction and the lower floor shows the air flowing in the other direction. Configuration for forced ventilation: When there is insufficient wind to bring comfort through air changes, the cross-ventilation is shut down and forced ventilation starts up. During forced ventilation, the inlet air comes in through the sound attenuation box, below the window. The outlet air leaves through a duct at the opposite end of the room, that leads to a rooftop VAV fan. This is illustrated in the top floor in Figure 3. The outlet duct is the forced air duct. There is a damper in this duct that will be open during 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 348 forced ventilation and will close when when the space is in cross-ventilation mode. This damper can also modulate it's position, so to vary the flow of air. Multiple ducts from multiple classrooms lead to a single variable air volume (VAV) fan on the roof. Each duct has a variable air volume (VAV) damper. In effect, this is a standard VAV system working in the reverse direction. The air in the room enters the VAV dampers and exits the system through the fan. The dampers in each room will close and open in response to the operative thermostat in the space. The variable speed fan will change it's speed in response to the static pressure. Protocol for natural ventilation o o o the damper in the force air duct is completely closed The damper in the cross-ventilation duct is fully open Cross ventilation will come in through sound attenuation box and will exit through the cross ventilation duct. Or the air will move through in the reverse direction. Protocol for forced ventilation o o o the damper in the cross-ventilation duct is closed damper in the forced ventilation duct is partially or fully open The fan driven air will enter through the sound attenuation box, pass through the room and exit through the forced air duct, that leads to the rooftop fan Control strategy for classrooms Computer simulations in this project have shown that air changes in the space is critical in bringing about comfort conditions. The control protocol is built on this understanding that ACH in the space has to be estimated and modified if needed. To do this the ACH in the space has to be estimated under three conditions. The ACH has to be known 1) during natural ventilation. 2) during forced ventilation and 3) when the wind speed and wind the direction are known, the potential ACH from natural ventilation has to be estimated. It is done in the following manner: 1. ACH due to cross-ventilation. There is an air velocity sensor in the cross-ventilation duct. This duct is straight and long, so the readings are reliable and will allow us to calculate the ACH in the room. Let us call this "ACH_cross_vent" 2. ACH due to forced ventilation. This can be measured by an air velocity sensor in the forced air duct. It is possible that this duct is too short to get an effective reading from the sensor. In this case ACH can be calculated from the position of the VAV damper and the fan speed. Let us call this "ACH_forced_vent". 3. ACH_lookup_table. Detailed wind tunnel studies were done and a lookup table was developed to where for any wind speed and direction, we can lookup the ACH in a classroom. So based on data from the weather station, we can lookup the potential ACH_forced_vent in any classroom. (Once the building is operational, we can fine tune the lookup table based on air velocity readings in the cross-ventilation duct and weather data.) The most efficient state is when natural ventilation and the operative temperature (OT) of the room is less than the thermostat setpoint (T-setpoint). Here the damper in the cross-ventilation duct is fully open and the damper in the forced air duct is completely closed. As long as the conditions in the space meet the setpoint, this state will continue with crossventilation. If the OT within the space rise above the set point, the forced ventilation will start up with the steps listed in the sub-section “protocol for forced ventilation”. Figure 4 shows the logic of this. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 349 Figure 4 Turn fan on in classroom (a) Switchback Strategy 1 (b) Switchback Strategy 2 Figure 5 One the forced ventilation starts up, there is need to switchback to natural ventilation whenever natural ventilation can meet the comfort needs. The control protocol needs to know when to do the switchback. There are two strategies for this. The first one called “switchback strategy 1” In the “switchback strategy 1” the control protocol will retrieve the air speed and wind direction from the weather station. Using this air speed and wind direction the ACH_lookup_table will return the resultant ACH due to cross ventilation. Now if ACH_cross_vent is greater than ACH_forced_vent, there is sufficient wind to cool the space by natural ventialtion. So the system will switch over to natural ventilation. Otherwise it will stay with forced ventilation. Figure 5 (a) shows this control logic. Switchback strategy 2 has a simple logic. Continue with powered ventilation for some time after the set-point is reached. This extra time acts as a dead-band. After the waiting period, switch back to natural ventilation. Until the building is operational it is not clear which strategy will be more effective Response to high outdoor temperature If the outside temperature is too high during forced ventilation period, bringing a lot of hot air from outside will not contribute to comfort. Under such circumstances, the dampers should move to their minimum ACH position that will meet the indoor air quality. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 350 OFFICE VENTILATION The remodeled office has two corridors that run the entire length of the office building. Exploring design alternatives in the wind tunnel, it was discovered that if the long corridors were closed at one end and left open at the other end, all the rooms got sufficient ventilation. The prevailing wind would come into the corridor and pressurize the corridor. This pressurization would push the air into all offices adjacent to the corridor. If the wind came from the other direction, the corridor would get de-pressurized and it would draw air from the offices, allowing a flow in the opposite direction. All the offices are connected to an exhaust fan system that can vented by forced ventilation if the natural ventilation is insufficient. The forced ventilation system is similar to that of the classroom. It works like a standard VAV system operating in reverse. The air in the room enters the VAV dampers and exists the system through the fan. As the dampers close or open, the variable speed fan will change it's speed in response to the static pressure. The office ventilation strategy is fundamentally different from that of the classroom. In the classrooms either the forced ventilation is on or the natural ventilation is on. In the office, the natural ventilation can be on with the forced ventilation assisting it. Office control strategy o o if the Operative temperature (OT) is greater than the thermostat setpoint (T_setpoint) the VAV exhaust fans come on in case of high outdoor temperature, office forced ventilation strategy follows the same rules as the classrooms, in that the forced ventilation volume drops down to the minimum ACH needed for indoor air quality Switch off strategy in office Since the forced ventilation fans in the office simple assist the natural ventilation, the idea of switching back and forth between forced and natural ventilation does not arise. Once the setpoint is reached, the forced air fans continue to operate for a certain time period. This extra time period acts like a dead band. Then the forced ventilation fans switch off CONCLUSION The standard way the controls are designed is to use a mechanical system to actively bring about comfort. This is done by the mechanical system keeping the temperature of the space below the setpoint. In a sense the standard control protocol is very simple. In designing the control protocol for a passively conditioned building through adaptive comfort, we are breaking new ground. We cannot be sure that system will work as designed and have to be ready to respond to unplanned behavior of the system. The building has been extensively modeled and the design calls for monitoring of all critical sensor points in the model. Thus we can observe the building in realtime and tune the control protocol in response to actual behavior. We also troubleshoot any aspects that are misconfigured. The project is awaiting funding approval from the state legislature. Once the retrofit is finished, we should have a more complete understanding of effectiveness of this control protocol. ACKNOWLEDGMENTS This work has been funded by the "Commercial Building Partnership" initiative of the Department of Energy and administered by Lawrence Berkeley National Laboratory. The wind tunnel study has been funded by University of Hawaii, Manoa. The authors would like to thank: o The other members of the design team - Benjamin Woo Architects, Honolulu, Hawaii and Notkin Hawaii Inc. Honolulu, Hawaii o Gail Braeger and Fred Bauman of Center for the Built Environment, UC Berkeley for guidance on application of Adaptive Comfort to this project. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 351 o O Ed Arens for guidance and the use of the wind tunnel at Building Science Laboratory, University of California, Berkeley Cindy Regnier of Lawrence Berkeley National Laboratory for administering this project and keeping it on track. NOMENCLATURES ACH - Air Changes per Hour ACH_Lookup_Table – See subsection: Control Strategy for classrooms”, 3rd item on numbered list ACH_cross_vent - ACH through cross ventilation ACH_forced_vent - ACH through forced ventilation DBT - Dry Bulb Temperature MRT - Mean Radiant Temperature OT - Operative temperature RH - Relative Humidity T_in – Temperature inside T_out – temperature outside T_setpoint – Thermostat setpoint. This will change every day on the comfort thermostat UHM - University of Hawaii, Manoa REFERENCES ASHRAE (2007). ANSI/ASHRAE Standard 90.1-2007 - Energy Standards for Buildings Except LowRise Residential Buildings. Atlanta, American Society of Heating, Refrigerating and AirConditioning Engineers, Inc. ASHRAE (2010). ANSI/ASHRAE Standard 55-2010 - Thermal Environmental Conditions for Human Occupancy. Atlanta, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. de Dear, R. J. and G. Brager (1998). "Developing an adaptive model of thermal comfort and preference." ASHRAE Transactions 104(1A): 145- 167. Fountain, M.E. and C. Huizenga (1996). “WinComf : A Windows 3.1 Thermal Sensation Model - User’s Manual.” (Berkeley: Environmental Analytics). Philip, S., Tran, T., Tanjuatco, L. (2011). eppy: scripting language for E+, EnergyPlus (version 0.46) [Software - GNU AFFERO GENERAL PUBLIC LICENSE] Avaliable from <https://pypi.python.org/pypi/eppy/0.4.6>, https://github.com/santoshphilip/eppy/tree/r0.46 Philip, S., Shameson, A., Brown, N., Loisos, G., Ubbelohde, S. (2014) “Design and Modeling Strategies for Retrofit to Natural Ventilation” 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 352 Effective na tural vent ilati on in mo dern apartm ent buildi ngs P. C. T homas Ayshwary a Venk at esan Leena E. Th omas Team Catalyst, Sydney Australia pcthomas@teamcatalyst.com.au Team Catalyst, Sydney Austral ia University of Technology, Sydney ABSTRAC T This paper addresses the challenge of evaluating for natur al ventilation in modern apart ment buildings. A number of natur al ventilation design rules of thumb from published literat ure are listed. Their incorporat ion into one code for Australia (the Residential Flat Design Code, or RFDC) and India (the National Building Code, or NBC), in relation to apart ment buildings is examined. Practical limitations to converting these rules of thumb into effective natur al ventilation systems for apart ment building designs are discussed. Apartment designs in the moderat e locations of Sydney, Australia and Bengaluru, India are also reviewed to assess their effectiveness for natural ventilation. Simulation analysis presented indicate lar ge energy savings are possible if apart ments are retrofitted/designed to the proposed code requirements and designs compliant with thumb rules are capable of delivering effective natur al ventilation if users choose to operat e the apart ment in “fr ee running mode” during times when the outside dry bulb temperatur es lie in an appropri ate band. The paper also discusses how sub-optimal design solutions, affluence and adaptat ion to more stringent thermal conditions can negate the potential for natur al ventilation and calls for proact ive efforts to maintain climate responsive design standards an d education/policy to encourage the benefits of natur al ventilation over airconditioning. INT ROD UCT ION Apartment buildings have become one of the most affordable residential building configurations in cities around the world. In many locations, they purport to incorporate natural ventilation design elements, usually based on the minimum mandatory requirements of applicable codes and standards. Anecdotally, an increasing number of such apartments are also being fitted with air-conditioning systems. These are either fitted by the builder (usually in developed countries like Australia, or upmarket offerings in developing countries like India), or retro-fitted by the occupant. Easy access to air-conditioning brings with it the possibility of its increased use, even when conditions outside are conducive for natural ventilation to provide adequate thermal comfort. Therefore, the challenge faced in the design of modern apartment buildings is to ensure that the natural ventilation “system” is effective at providing thermal comfort at appropriate times, and to make the natural ventilation “mode” easily accessible to occupants. Evaluating the effectiveness of natural ventilation in real building designs has always been difficult. The analyst is limited to using simplified equations developed for idealised room configurations, or complex energy simulation programs that can solve heat transfer and airflow network equations PC Thomas is Director, Team Catalyst, Sydney, Australia. Ayshwarya Venkatesan is Architecture/ESD Specialist at Team Catalyst, Sydney, Australia. Leena Thomas is Associate Professor & Strand Leader for Environmental Studies at the School of Architecture, University of Technology, Sydney. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 353 simultaneously in the same time step. Computational Fluid Dyamic (CFD) can solve for velocity and temperature gradients in discrete volumes, but these are limited to instances in time with specific boundary conditions. It requires significant processing power and time to run CFD analysis in time step increments so as to make a judgement of the effectiveness of system A against system B. This paper came about because of a review (Thomas and Venkatesan, 2012) of the thumb rules for natural ventilation included as part of the Residential Flat Development Code (RFDC) carried out for the state government of NSW, Australia. The objective of the review for the RFDC was to develop a series of simple rules to allow a planner to approve a design on its merit. The method followed to develop these rules was as follows:  review the body of currently available knowledge listing the attributes for effective natural ventilation  turn them into a design calculation procedure simple enough to be carried out using hand calculations or a spreadsheet program, and  document an application process that can be reviewed and accepted by a planning official This approach has been implemented on apartment buildings in Sydney and Bengaluru. A two bedroom and a three bedroom apartment each from Sydney and Bengaluru have been selected for analysis. These represent the most common typology for Sydney and upmarket developments for Bengaluru. PRO CEDUR E FO R EVALUATIN G POTE NTIAL FO R EFFEC TI VE NATURA L VENTILAT ION The procedure devised and proposed to evaluate an apartment design for its potential for effective natural ventilation for the RFDC was based on common rules of thumb found in existing literature (Lechner, 2001). They include the following:  The ratio of room depth (W) to floor-to-ceiling height is to be W< 5H for cross ventilated rooms; however for a standard ceiling height of 2.7m, the room depth may be extended to 15m when other conditions are met.  The ratio of room depth (W) to floor-to-ceiling height is to be W< 2H for single sided rooms; however for a standard ceiling height of 2.7m, the room depth may be extended to 6m for living spaces. The ratio of room depth (W) to floor-to-ceiling height can be increased to 2.5H for rooms designed to have single sided ventilation when stack ventilation can be induced by providing a 1.5m height separation between inlet and outlet sections of the window.  The Effective Openable Area for windows is at least 5% of the total floor area. Effective openable area of the window is defined as the area of the window that can take part in providing natural ventilation to achieve occupant comfort. It considers the portion of the window that is openable. Window openable area is reduced by 50% when insect screens are fitted. In the case of a full height sliding door, the openable area is reduced, in elevation, when it is within 2m of a solid balustrade.  Windows are located so they are equally distributed on windward and leeward sides of the building. Windows on the leeward side (or “outlet” windows) may be allowed to be slightly larger than on the windward (or “inlet” windows) side of the building thereby utilising air pressure to draw air through the apartment. Similar guidelines are found in the proposed draft for the new Part 11 of the National Building Code of India (NBC), (BIS, 2012) on Sustainability. It proposes the following criteria:  Naturally ventilated building should take advantage of predominant wind, and enhance stack ventilation by providing low level inlets and high outlets  Floor depths of more than 15m are difficult to naturally ventilate  The total area of openings (inlet and outlet) should be 20 to 30 percent of floor area The NBC does not distinguish between single sided and cross ventilation. To our knowledge, none of the building codes in either country provide explicit recommendations for mixed mode operation, which is crucial to realizing the potential of thermal comfort with reduced energy use. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 354 CLI MATE ANALYSIS FO R BENG ALUR U AND SYDNEY Bengaluru, India has always been regarded as the “garden city” with a temperate climate, being situated on the Deccan Plateau at a height of approximately 914m above sea level. Sydney, Australia is also well known for its mild climate, and is at sea level on the south eastern coast of Australia. Figure -1: Temperat ure r anges for Bengalur u (top) and Sydney (below) An analysis of the range of dry bulb temperatures for the IWEC reference weather files for Bengaluru and Sydney is shown in Figure-1 (and later in Table-1), and clearly shows the reversal of seasons due to the two cities being located on the north and south hemispheres on Earth. Figure-1 also reveals the variability of temperatures between the two cities. There are only small differences between the daily average mean high (or low) and the maximum recorded high (or low) temperatures for any one month (ie., the distance between the box and whisker) for Bengaluru, but the data for Sydney shows large variations between the box and whisker for each month. This effect is most clearly seen in the Sydney data for the months of October and December. While there are many ways to analyse climate data in a comparative manner, the significance for this study is to determine the approximate number of hours when the outside dry bulb temperature falls within a nominated “comfort range”. This would allow us to “bookend” the theoretical number of potential “discomfort” hours at each location. Regulation, and design skills could then be focused on reducing the number of discomfort hours by incorporating the principles of effective natural ventilation, thermal mass, shading, orientation etc. It is difficult to find definitive regulatory information stipulating a “comfort range”. In Australia, the NSW government does not provide any specific guidelines in the RFDC. The NatHERS (Nationwide House Energy Rating Scheme), referenced in the National Construction Code (ABCB 2013) uses a setpoint of 25.5°C for Sydney. Above this temperature an air-conditioner is assumed to cool residential spaces for the regulatory energy analysis. Heating setpoints in living spaces are to be maintained at 20°C during waking hours, ie, 7:00am to midnight, while bedroom spaces are allowed to drop to lower temperatures at night (http://nathers.gov.au/). For Bengaluru, the National Building Code (BIS 2003) of India does not seem to differentiate between residential and other buildings, and defines narrow temperature ranges for winter (21-23°C) and summer (23-26°C) for all building and climate types (Indraganti, 2010). While the appropriateness of these temperature ranges can be argued, they are sufficient for the purpose of bookending the potential “discomfort hours”, based on ambient dry bulb temperature, for the two locations. Inferences from Figure-1 and Table-1 provide important insights to the two climates. From Table1, Bengaluru has less total “discomfort hours” (about 55% of the year), but they are divided almost equally into the too hot and too cold ends of the spectrum. Figure-1 indicates that highest temperatures 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 355 will not be more than a few degrees higher than the daily average highs for that month, e.g, April. In contrast, only 4-6% of the year can be considered to be too hot in Sydney, but they can include some extreme days, note the 40°C day in December. However, about 75% of the year falls into the too cold category. location Bengaluru Sydney location Sydney hours above hours "discomfort" 26C below 21C hours hours above 26C hours "discomfort" below 21C hours 2,265 2,520 4,785 26% 29% 55% 367 6,178 6,545 4% 71% 75% hours above hours "discomfort" 25.5C below 20C hours hours above 25.5C 482 6,008 6,490 6% hours "discomfort" below 20C hours 69% 74% Table-1: Dry bulb temperatures above and below cooling and heating setpoint temperatur es Therefore, for Bengaluru, apartment design must be equally adept at dealing with overheating (comfort ventilation) and cold climate (passive solar gain, insulation, reducing infiltration). For Sydney, the focus should be on managing cold conditions. This means an emphasis on insulation requirements, which is addressed by the BASIX (Dept of Planning and Environment, 2013) for Sydney and mandatory building code (NCC) for the rest of Australia. However, the inferences from Figure-1 suggest that while there are few hours of over heating, they can be quite extreme when they do occur. APAR TM ENT S SELE CT ED FO R ANALYSIS Plans for a two bedroom apartment and a three bedroom apartment were selected at each location. The plans selected for analysis for this study reflect these differening views of apartment living in the two cities. Each of the apartments has been evaluated using the procedure discussed earlier. The selection of apartment plans for analysis is predisposed to layouts that were cross ventilated and of optimum depth. This is because we wanted to identify plan typologies that showed the potential for natural ventilation, where the limiting factor for mixed mode operation (and energy savings) were due to the potential, and attitude, for user control. The plans selected for Sydney were recommended as better design practice solutions in the RFDC technical document. These plans performed well when analysed for effective natural ventilation with respect to the proposed RFDC criteria (Thomas and Venkatesan, 2012). The plans selected for Bengaluru were procured from websites of prominent builders who marketed their layouts as being luxiourious upmarket/green. The Bengaluru apartment plans were tested against the proposed draft NBC Part 11 criteria. Proposed RFDC Criteria (Sydney) 2 Bed corner Apartment (Cross Ventilation) 3 Bed Apartment ( Cross Ventilation) Draft NBC Part 11 Criteria (Bengaluru) 2 Bed Apartment (Cross Ventilation) 3 Bed Apartment (Cross Ventilation) Ratio of room depth W to floor to ceiling height H is W<5H. Yes Yes Apartment depth not greater than 15m Yes Yes Effective openable area of window is atleast 5% of the total floor area Yes Yes Opening Area of window is 20%-30% of the total floor area Yes Yes Inlet area equal to outlet area No No Table-2: Application of procedure to determine potential for Effective Natural Ventilation Table-2 indicates that all four apartment plans selected satisfy the building envelope design criteria for effective natural ventilation. Based on prior studies (Thomas and Venkatesan, 2012) the Sydney apartments (Figure-2), will therefore meet thermal comfort and minimum energy performance requirement for BASIX; and hence the energy use for air-conditioning will be low. Analysis from Table-2 indicates that users have the potential to effectively operate the apartments in a natural ventilation mode to further reduce energy consumption. Therefore, no further analysis has been carried out for these. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 356 Figure -2: Floor pl ans of apart ments analysed for Sydney, Austral ia Figure -3: Fl oor plans of apa rtments analysed for Bengaluru, India The three bedroom apartment in Bengaluru is proposed as a high rise block, as a part of a very large residential layout. The layout of the apartment shows good potential for cross ventilation capability. The apartments are being marketed as being fitted with split A/C units to living and all the bedrooms. They are also to be fitted with “clear float glass” with UPVC frames for all the windows. This last statement implies the use of single clear glass with no particular performance requirements. An energy simulation model for the apartment was constructed using DesignBuilder/EnergyPlus that separated the living and sleeping zones. Floor and ceiling zones were modeled to be adiabatic, thus representing a typical middle floor apartment. Thermostat setpoints were set as previously discussed at 26°C for cooling and 21°C for heating. A design COP of 2.5 was used for cooling, and single zone split systems were modeled. Nominal internal loads were modeled, and these were identical for both runs. The annual energy consumption was predicted for two cases:  for an uninsulated building envelope and clear single glass, and  retrofitted with NBC levels of insulation and glazing performance requirements (external wall insulation of R2.0; SHGC=0.3-0.4 for WWR of 60-40%) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 357 Figure -4: Representation of 3 bedroom apart ment for energy simulation; floor plate has been divided into bedroom areas (night), lounge/kitchen area ( day) and toilets (unconditioned) The results predict minimal heating was required for both cases tested. Therefore only the cooling energy consumption was considered, and this dropped by 44% when the original design was remodeled with NBC recommendations (see Table-3). A significant portion of this reduction can be attributed to the improved glazing. 3 Bed apartment with uninsulated building envelope and single clear glazing 3 Bed apartment retrofitted Insulation to fabric and Glazing performance as per Draft NBC section 11 reduction in cooling energy Cooling Energy, KWh/m2-yr 40.8 22.7 44% Table-3: Results of energy simulation for 3 bedroom apartment in Bengaluru Figure 5 below indicates the predicted cooling load across the year in the bedroom (night time) and lounge/kitchen (day time) zones for the draft NBC compliant apartment if it resorted to airconditioning to maintain temperatures below the 26°C setpoint during occupied hours. While this suggests reliance on cooling beyond peak summer in daytime zones, field work in naturally ventilated buildings (Indraghati 2010, Manu etal 2014) suggests subcontinental occupants are tolerant to higher temperatures such as 28°C based on an adaptive model of comfort especially as increased air flow via ceiling fans can further ameliorate discomfort. Bedroom J F M A M J J A S O N D Living J F M A M J J A S O N D Figure -5: Predi cted Cooling Load (kW) acr oss the year for Bedroom and Living Zones (Bengalur u). A final simulation run was carried out with the building allowed to operate in “free running mode” with controls that allowed windows to be open so that 30% of window area could take part in natural ventilation air exchange. The results indicate that this configuration allows the internal zone air temperatures to closely track the ambient outside temperatures (Figure-6, for both the Lounge Kitchen (Graph 1) and Bed 3 located in the south-east – the hottest corner (Graph 2). This confirms that effective natural ventilation is possible in the selected apartment design, when outdoor dry bulb temperatures are in the “comfort range”. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 358 Figure -6: Outdoor dry bulb temperatur e and zone mean air temperat ure in free running mode for lounge/kitchen (Graph 1) and bedroom (Grap h 2, Bed 3 in Figure 4), for 3 bedroom apart ment in Bengaluru BARRIER S TO EFF EC TI VE NAT URAL VEN TILA TIO N In Australia, apartments are located in the Central Business District (CBD) areas of the metropolitan cities, or in areas where land is very expensive, along transport corridors (near train and bus routes) and also in the less affluent suburbs. In many instances, apartments may be regarded as a stepping stone towards the goal of an independent house and land package. One reason for this view is because real estate is a reasonably liquid asset in places like Sydney, where a sale transaction can be completed in a manner of weeks. Apartments also attract strong investor interest based on potential rental returns. Therefore, a significant proportion of apartment residents are rental tenants whose options to retrofit do not extend beyond installing a pedestal fan or an electric heater. Strata laws prohibit the owner from making alterations to the external common building elements like walls and windows and the installation of flooring and curtains/blinds are the prerogative of the owners. While energy costs are increasing steadily in Australia and can add considerable budget pain to the middle class, most new apartments come with reverse cycle DX type air-conditioning sstems pre-installed. This easy access to air-conditioning, coupled with design solutions that are less than perfect that may present barriers to the effective use of natural ventilation. Three examples of such design solutions identified1 in some Sydney apartment buildings are shown in Figure 7. They include deep, narrow floor plates which do not allow for effective cross ventilation (Figure 7 a), and deep set “snorkel” windows and deep “notch” corridors that effectively impede air exchange (Figure 7 b). Figure -7a: Example of design solutions that reduce effectiveness of natural ventilation 1 Zanardo, M. 2012, Personal discussion and review of apartment plans for Sydney 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 359 Figure -7b: Examples of design solutions that r educe effectiveness of natur al ventilation In places like Bengaluru, apartments are permanent homes, with many new developments offering high levels of luxury within the security of a gated community. Such luxury developments do have adequate spacing between apartment blocks to have an increased potential to operate in a naturally ventilated mode. However, such luxury apartments are also generally pre-fitted with A/C systems, and this easy access to air-conditioning diminishes the incentive for occupants to adapt to ambient conditions. This is exacerbated for the modern knowledge-worker, who aspires to live in such luxury apartments; regularly works in air-conditioned offices, and whose tolerance for temperatures beyond the closely controlled temperature band in the office drops with increasing adaptation to air-conditioning. CON CLU SIONS In a world where there is increasing evidence of rapid anthropogenic climate change, it is critically important that apartment designs provide easy access to the real potential for reduced CO2 emissions, so occupants can minimise their use of non-renewable energy use with little extra effort. It is clear that the apartment designs selected for the two cities indicate that they pass the critical requirements to be able to provide Effective Natural Ventilation. The simulation analysis undertaken here predicts that  large energy savings are possible if apartments are retrofitted/designed to the proposed NBC requirements of Part 11, and  effective natural ventilation is possible if users choose to operate the apartment in “free running mode” during times when the outside dry bulb temperatures lie in an appropriate band However, it is argued that this potential for effective natural ventilation, and energy efficient living can be easily subverted. As discussed in this paper, sub-optimal design solutions, affluence and adaptation to more stringent thermal conditions can negate the potential for natural ventilation even in the relatively mild climates such as Sydney and Bengalaru. This calls for proactive efforts to maintain climate responsive design standards and education/policy to encourage the benefits of natural ventilation over airconditioning. REFER ENC ES ABCB 2013, Australian Building Codes Board, National Construction Code 2013 BIS 2005, Bureau of Indian Standards, National Building Code of India 2005, (SP 7:2005) BIS 2012, Bureau of Indian Standards, Draft Amendment No. 1 to National Building Code of India 2005 (SP 7:2005), to incorporate a new Part 11 Approach to Sustainability Dept of Planning and Environment, NSW, Australia, 2013, BASIX Thermal Comfort Protocol Lechner, N., 2008, Heating, Cooling, Lighting – Design methods for Architects, Indraganti, M. 2010. Thermal comfort in naturally ventilated apartments in summer: Findings from a field study in Hyderabad, India. Applied Energy, 87(3), 866–883 Manu, S., Shukla, Y., Rawal, R., Thomas, L., de Dear, R., Dave, M., & Vakharia, M. 2014. Assessment of Air Velocity Preferences and Satisfaction for Naturally Ventilated Office Buildings in India. In PLEA 2014: 30th Conference on Pa ssive and Low Energy Architecture. Ahmedabad, India: CEPT University, Ahmedabad Thomas, P.C., and Venkatesan, A., 2012, Review of Natural Ventilation and Daylighting Guidelines for RFDC, Dept of Planning and Infrastructure, NSW 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 360 Session 3D : Tools and methods/ framework PLEA2014: Day 1, Tuesday, December 16 16:05 - 17:45, Trust - Knowledge Consortium of Gujarat Numerical Simulation of Passive Cooling Strategies for Urban Terraced Houses in Hot-Humid Climate of Malaysia Doris Hooi Chyee Toe, Dr.Eng. Susumu Sugiyama Satoshi Yasufuku Universiti Teknologi Malaysia Hiroshima University Hiroshima University Tetsu Kubota, Dr.Eng. Hiroshima University ABSTRACT The objective of this study was to determine energy-saving modifications through passive cooling to urban terraced houses in Malaysia. Effects of two strategies, i.e. complete natural ventilation (NV) strategy and partial air conditioning (AC) strategy, were simulated using TRNSYS and COMIS. The complete NV strategy relied fully on naturally ventilated condition in the whole house for achieving thermal comfort in the master bedroom while the partial AC strategy was aimed at reducing the cooling load in the air-conditioned master bedroom by applying passive cooling techniques to the whole house. The results revealed that indoor thermal comfort was achieved in complete NV strategy by applying multiple passive cooling techniques that prevent external heat on the outer building envelope and night ventilation, even under heated urban climatic conditions. In partial AC strategy reductions of about 39% to 56% in the sensible cooling load compared to the current scenario were obtained by using several techniques including night ventilating other spaces and insulating inner surfaces of the master bedroom. INTRODUCTION Energy savings are important in the global building sector due to concerns about energy security and effects of global warming. In hot developing regions such as Southeast Asia, cooling demand in residential buildings is a major concern since it is predicted to rise sharply in the coming decades in line with rapid urbanization and population and economic growth (Liu et al., 2010; Sivak, 2009). It can be seen widely in the region that brick-walled buildings are becoming a common construction for urban houses in recent years. In Malaysia, a nationwide census in 2010 showed that 85% of the existing urban houses used brick and another 5% used brick and plank for their outer walls (Department of Statistics Malaysia, 2012). Unlike traditional lightweight constructions, the high thermal mass building envelope of brick houses might be difficult to be cooled in the hot-humid climate. It has been reported in 2009 that space cooling in brick houses accounted for 29% of the annual household energy consumption on average in the city of Johor Bahru, Malaysia (Kubota et al., 2011). It is thus crucial to apply passive cooling strategies wherever possible to these urban houses for energy-saving. Passive cooling encompasses techniques for solar and heat control, heat modulation and heat D.H.C. Toe is senior lecturer in the Faculty of Built Environment, Universiti Teknologi Malaysia, Skudai, Malaysia. S. Sugiyama and S. Yasufuku are graduate students in the Graduate School for International Development and Cooperation, Hiroshima University, Hiroshima, Japan. T. Kubota is associate professor in the Graduate School for International Development and Cooperation, Hiroshima University, Hiroshima, Japan. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 361 dissipation using naturally driven phenomena such as natural ventilation, radiative cooling, evaporative cooling and ground cooling (Santamouris and Kolokotsa, 2013). Passive cooling techniques have been studied in various climatic regions. However, few comprehensive studies were made outside moderate and hot-dry climates, including field monitoring and numerical modeling exercises with regard to existing Malaysian houses (Kubota et al., 2009; Mohd Isa et al., 2010; Sadafi et al., 2011). Some of the main climatic factors negatively affecting the efficiency of the different cooling approaches are high night ambient temperature, cloud cover, high humidity and insufficient wind speeds (Dimoudi, 1996). These conditions are usually prevalent in hot-humid climate. Due to dependency on climatic conditions, further local studies are required to predict effects of a passive cooling system before implementation. The objective of this study is to determine energy-saving modifications through passive cooling to Malaysian urban houses. The target houses are terraced houses, which formed majority (42% as of 2010) of the existing urban housing stock (Department of Statistics Malaysia, 2012). This study analyses the effects of two passive cooling strategies, i.e. complete natural ventilation (NV) strategy and partial air conditioning (AC) strategy, on thermal comfort and cooling load, respectively, through numerical simulation using TRNSYS and COMIS programs. METHODS Description and Modeling of the Case Study Terraced House One of the case study terraced houses from a previous field experiment (Kubota et al., 2009) was modeled in this simulation study. The selected terraced house represents typical modern terraced houses in terms of spatial design and building structures (Toe, 2013). The house measures 6.7 m by 13.1 m with a total floor area of 155 m2, which is an average sized double-storey terraced house (Figure 1). Floor-toceiling height of rooms is 3.05 m. The total nett air volume of the whole house is 538 m3; that of the master bedroom is 65 m3. The building was oriented towards northwest, which means that the external façade of the master bedroom faces northwest. It was constructed of brick and concrete and had single glazing windows. The entire house was not insulated. Description of the constructional layers of the terraced house and their reference U-values in the computer model is given in Table 1. The whole house was modeled as TRNSYS Type 56 ‘Multi-zone Building’ in three dimensions using the TRNSYS 3D plug-in in Google SketchUp interface (Klein et al., 2012). The building model comprised 17 thermal zones with corresponding air flow zones in COMIS to represent each partitioned room or functional space including attic spaces. All protruding elements on the building facades and immediate surrounding objects, i.e. neighbouring houses, that might shade the studied house were also modeled in three dimensions. A time base of 1 h was set for the transfer function to represent the thermal (a) (b) First floor Ground floor Figure 1 (a) Exterior view and (b) floor plans of the case study terraced house. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 362 Table 1. Constructional Layers and Reference U-values of the Terraced House Model. Building Element Constructional Layers Reference U-valuea (W/m2K) External and internal walls Party wall Ground floor First floor (family and bedroom zones) First floor (bath zones) Ceiling (master bedroom) Ceiling (other zones) Pitched roof 20 mm thick cement plaster + 100 mm thick clay brick + 20 mm thick cement plaster 20 mm thick cement plaster + 200 mm thick clay brick + 20 mm thick cement plaster 8 mm thick ceramic tile + 22 mm thick cement screed + 100 mm thick concrete slab + soil layer 15 mm thick timber flooring + 15 mm thick cement screed + 100 mm thick concrete slab + 20 mm thick cement plaster 8 mm thick ceramic tile + 22 mm thick cement screed + 100 mm thick concrete slab + 20 mm thick cement plaster 6 mm thick ceiling board 2.75 2.07 3.75b 2.81 3.29 4.55 3.2 mm thick ceiling board 5.54 20 mm thick concrete roof tile + 25 mm thick timber batten + 2.67 aluminium foil Flat roof 22 mm thick cement screed + 100 mm thick concrete slab + 20 3.37 mm thick cement plaster Window 6 mm thick single layer float glass 5.61 a 2 Includes convective and radiative heat transfer coefficients of 7.7 W/m K for inside surface and 25 W/m2K for outside surface. b Excludes soil layer. mass behavior of the brick walls. The party walls on both sides of the house were modeled as boundary walls with identical zone temperatures assumed on both sides of the walls. Meanwhile, the boundary condition for the ground floor was the constant soil temperature assumed to be the average air temperature at the site over the whole simulation period. Thermal properties of building materials and parameter/input values for air flows were obtained from Malaysian manufacturers or reference data to correspond with the local construction (Toe, 2013). Wind pressure coefficients were estimated using a parametrical model developed by Grosso (1992) known as CPCALC+. Coupling between the TRNSYS and COMIS models were implemented via Type 157 in TRNSYS so that air flow rates per zone and zone air temperatures were iterated in each time step until the mass and energy balance per zone reached convergence. Model Validation 14/7 15/7 16/7 17/7 18/7 0:00 13/7 0:00 Measurement 0:00 18/7 Measurement Simulation 0:00 17/7 0:00 16/7 Date / Time 0:00 15/7 0:00 14/7 0:00 13/7 0:00 Measurement Simulation 0:00 Simulation Outdoor 0:00 Operative Temp. (°C) Measurement 38 36 34 32 30 28 26 24 22 36 34 32 30 28 26 24 0:00 (b) Simulation Air Temp. (°C) Outdoor 0:00 38 36 34 32 30 28 26 24 22 36 34 32 30 28 26 24 0:00 Operative Temp. (°C) (a) Air Temp. (°C) Empirical validation of the terraced house model was performed using the above-mentioned field experiment data from June-August 2007 (Kubota et al., 2009). This study focuses on the results in the master bedroom because existing households used air conditioners mainly in master bedrooms (Kubota et al., 2009); the study interest is to reduce this cooling energy. Figure 2 shows temporal variations of the simulation results compared to the measurement data at 1.5 m height above floor in the master bedroom. Two ventilation conditions, i.e. night ventilation and daytime ventilation, are shown. Overall, the Date / Time Figure 2 Temporal variations of the simulation and measurement data in the master bedroom in (a) night ventilation and (b) daytime ventilation conditions. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 363 Table 2. Simulation Test Cases. Technique Natural ventilation (open window period) Forced ventilation Attic ventilation Thermal insulation Test Conditions Night ventilation (20:00-8:00); daytime ventilation (8:00-20:00); no ventilation (0 h); full-day ventilation (24 h) 10 ACH or 30 ACH night (20:00-8:00) in master bedroom 10 ACH night (20:00-8:00) or 30 ACH full-day (24 h) in attic Roof; ceiling; external wall – outside surface; external wall – inside surface; internal wall; party wall; floor (R-value: 4 m2K/W) External shading; internal shading (Shading factor, SF: 0.75) Solar reflectance: 0.8, longwave emissivity: 0.9 Low-E glass (U-value: 2.54 W/m2K, G-value: 0.44) or heat barrier film (U-value: 5.73 W/m2K, G-value: 0.48) Window shading High reflectivity roof coating Window glazing validation results are satisfactory in terms of air and operative temperatures with root mean square errors (RMSE) of 0.31-0.55 °C and coefficients of determination (R2) of 0.89-0.96. Simulation Test Cases and Weather Conditions Table 2 summarizes the simulation test cases of this study. The techniques were selected by considering their practicality to be applied to existing terraced houses through relatively simple building modification and/or behavioural adjustment. In particular, night ventilation is considered a potential passive cooling technique for brick houses while daytime ventilation emulates the window opening behavior of the majority of existing households (Kubota et al., 2009). The complete NV strategy relies fully on naturally ventilated condition in the whole house for achieving thermal comfort in the master bedroom. Meanwhile, the partial AC strategy attempts to reduce the cooling load in the master bedroom by applying passive cooling techniques to the whole house. It was assumed that air conditioning was used only in the master bedroom for nine hours per day (21:00-6:00) with a set temperature of 23 °C. This study deals with the sensible cooling load only. Internal heat gains from occupants (4 persons; seated at rest), lighting (5 W/m2) and common household appliances were considered in all simulations. It is noted that infiltration rates in the master bedroom average 0.1 ACH when no ventilation was applied for both complete NV and partial AC strategies. Weather conditions for the simulation were taken from an actual weather data set measured at the centre of a heat island in Johor Bahru, Malaysia to represent urban climate of typical terraced housing neighbourhoods (Kubota and Ossen, 2011). The geographical location is 1°29’19” N and 103°45’41” E at an elevation of 26 m above sea level. A wind velocity profile exponent of 0.25 was used to represent the urban location (Counihan, 1975). The simulation time step was set to coincide with the weather data at 10-minute intervals. The simulation was run using the above weather file for two whole months, i.e. January-February 2010. Subsequently, simulation results for a 10-day period of continuous typical fair weather days are analysed in this study. As shown in Figure 3, outdoor air temperature ranges from 2536 °C while outdoor relative humidity ranges from 50-90% over the period. The analysis period begins several days after the simulation start time, thus allowing the model to acquire sufficient thermal history. Output files were generated and post-processed in Excel spreadsheets after the simulation. Solar Rad. (W/m²) RH (%) Air Temp. (°C) 36 34 32 30 28 26 100 24 90 80 70 60 1200 50 800 400 0 Rain period 528 552 23/1 576 24/1 600 25/1 624 26/1 648 27/1 672 28/1 696 29/1 720 30/1 744 31/1 768 1/2 Date / Time (Hour of the Year) Figure 3 Temporal variations of weather data for the simulation analysis period. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 364 RESULTS AND DISCUSSION Complete Natural Ventilation Strategy Figure 4 presents the simulated indoor air temperatures in the master bedroom for the four natural ventilation conditions that represent night ventilation, daytime ventilation, no ventilation and full-day ventilation. It is noted that temporal variations of indoor temperatures in each simulation have similar patterns over the 10-day analysis period. Thus, results are shown in statistical summaries for the whole period. As expected, night ventilation provides the lowest indoor air temperatures among the tested open window conditions (Figure 4). This is due to the nocturnal ventilative cooling through open windows and thermal mass effect of the cooled building structures that lowers the night-time and peak indoor temperatures of the following day. Daily maximum (95th percentile) and minimum (5th percentile) indoor air temperatures in night ventilated condition are 1.7 °C and 1.3 °C lower than those of daytime ventilation, respectively. Nevertheless, the daily minimum air temperature in the night ventiled room is still 2.7 °C higher than the outdoors. Further passive cooling techniques are applied consecutively as shown in Figure 5 in addition to night ventilation and daytime ventilation, respectively. The most effective technique in reducing the daily maximum air temperature in night ventilated condition is roof insulation; the said temperature is decreased by 0.9 °C compared to applying night ventilation only (Figure 5a). Most of the solar heat gain in the master bedroom, which is on the first floor, probably comes through the roof due to its relatively large surface area and high noon solar altitude at the location. With less heat gain during the day and a cooler adjacent attic space for the whole day, the building structures maintain cooler and serve to reduce the minimum air temperature as well. Techniques that reduce solar radiation through roof into the building would be important. In fact, high reflectivity roof coating reduces the mean indoor air temperature most among all of the techniques in Figure 5a, i.e. by 0.6 °C. The high reflectivity coating probably improves nocturnal cooling additionally by virtue of the less heated roof surface on exposure to the sun and absence of thermal insulation at night. Nevertheless, all of the solar control techniques are Outdoor Night ventilation Daytime ventilation No ventilation Full-day ventilation 25 26 27 28 29 30 31 32 33 34 35 Air Temperature ( C) Figure 4 Statistical summary (5th and 95th percentiles, mean and ± one standard deviation) of simulated indoor air temperatures in different natural ventilation conditions for complete NV strategy. (a) (b) Night ventilation Daytime ventilation Roof insulation Roof insulation High reflectivity roof coating High reflectivity roof coating Ceiling insulation Ceiling insulation External shading External shading Low-E glass Low-E glass Internal shading Internal shading External wall-outside insulation External wall-outside insulation External wall-inside insulation External wall-inside insulation Forced ventilation (10 ACH night) Forced ventilation (10 ACH night) Attic ventilation (10 ACH night) Attic ventilation (10 ACH night) 27 28 29 30 31 Air Temperature ( C) 32 27 28 29 30 31 32 33 34 Air Temperature ( C) Figure 5 Statistical summary (5th and 95th percentiles, mean and ± one standard deviation) of simulated indoor air temperatures in (a) night ventilated and (b) daytime ventilated conditions with respective passive cooling techniques for complete NV strategy. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 365 (a) (b) Night ventilation Roof insulation External walloutside insulation + Roof insulation + External shading + External wall-outside insulation External shading + Low-E glass Night ventilation + High reflectivity roof coating + Forced ventilation (10 ACH) Master bedroom ACE 80% comfortable upper limit + Attic ventilation (10 ACH) + Internal shading Night ventilation 27 28 29 30 31 32 Operative Temperature ( C) Figure 6 (a) Thermal comfort evaluation and (b) conceptual illustration of combined passive cooling techniques for complete NV strategy. less effective in daytime ventilated condition compared to night ventilated condition (Figure 5b). The inflow of hot outdoor air through open windows during daytime increases the indoor air temperature and diminishes their cooling effects. On the other hand, forced ventilation with an air change rate of 10 ACH in the room at night lowers the daily minimum air temperatures most, i.e. by 0.6 °C and 1.4 °C in night ventilated and daytime ventilated conditions, respectively. Figure 6a shows the simulated indoor operative temperatures for combinations of the most effective technique for each of the building elements. The techniques are applied accumulatively and step-by-step from more effective ones to less effective ones in night ventilated condition. The results are evaluated for thermal comfort using an adaptive comfort equation (ACE) for naturally ventilated buildings in hothumid climates (Toe and Kubota, 2013). The 80% comfortable upper limits predicted using daily mean outdoor air temperatures of the analysis period average 29.6 °C. Figure 6a indicates that the daily maximum indoor operative temperature is reduced by 2.2 °C and meets the 80% comfortable upper limit when roof and external wall-outside surface insulation (R-value 4 m2K/W), and external window shading (shading factor 0.75) are applied in addition to night ventilation under the heated urban climatic conditions. Alternatively, the comfort limit is also met by substituting the roof insulation with high reflectivity roof coating, though daily maximum temperature is higher in the latter. It is implied that introducing these four techniques to existing urban terraced houses may satisfy indoor thermal comfort in naturally ventilated condition on fair weather days (Figure 6b). Partial Air Conditioning Strategy Figure 7 shows the simulated sensible cooling loads in the air-conditioned master bedroom by considering different natural ventilation conditions for the master bedroom and other zones. The cooling load is 50.2 MJ/day when daytime ventilation is applied to the whole house (Case 1), which represents the current behaviour of most households. By applying night ventilation to the whole house except the master bedroom, the cooling load is reduced by about 5% even when the master bedroom is daytime ventilated (Case 5). Building structures that are cooled at night keep adjacent indoor temperature low and reduce the cooling load indirectly. The highest reduction in cooling load, i.e. 8%, is seen when the master bedroom receives no natural ventilation and other zones are night ventilated (Case 6). Further passive cooling techniques are applied consecutively as shown in Figure 8 in addition to the ventilation conditions of Cases 1 and 6, respectively. For Case 1 the most effective technique in lowering Daytime ventilation Case Daytime ventilation 1 Daytime ventilation No ventilation 2 No ventilation No ventilation 3 No ventilation Daytime ventilation 4 Daytime ventilation Night ventilation 5 No ventilation Night ventilation 6 Master bedroom Other zones 0 10 20 30 40 50 60 Cooling Load (MJ/day) Figure 7 Simulated sensible cooling loads in different natural ventilation conditions for partial AC strategy. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 366 (a) (b) Daytime ventilation (all zones) High reflectivity roof coating No ventilation (master bedroom) + Night ventilation (other zones) Ceiling insulation Roof insulation Roof insulation Ceiling insulation High reflectivity roof coating Attic ventilation (30 ACH full-day) Attic ventilation (30 ACH full-day) Floor insulation Internal wall insulation Internal wall insulation External wall-inside insulation External wall-inside insulation External wall-outside insulation Party wall insulation Floor insulation External wall-outside insulation Party wall insulation External shading External shading Heat barrier film Low-E glass Internal shading Internal shading 0 10 20 30 40 50 60 0 Cooling Load (MJ/day) 10 20 30 40 50 60 Cooling Load (MJ/day) Figure 8 Simulated sensible cooling loads in ventilation conditions of (a) Case 1 and (b) Case 6 with respective passive cooling techniques for partial AC strategy. (a) Daytime ventilation (all zones) Case 1 No ventilation (master bedroom) + Night ventilation (other zones) Case 6 (b) Case 6 No ventilation (master bedroom) + Night ventilation (other zones) Ceiling insulation + Ceiling insulation + Internal wall insulation External wallinside insulation + External wall-inside insulation + Floor insulation AC + Low-E glass + External shading Master bedroom Internal wall insulation Floor insulation + High reflectivity roof coating + Party wall insulation + Internal shading Night ventilation 0 10 20 30 40 50 60 Cooling Load (MJ/day) Figure 9 (a) Simulated sensible cooling loads and (b) conceptual illustration of combined passive cooling techniques for partial AC strategy. the cooling load is floor insulation; the reduction is about 8% compared to the current condition (Figure 8a). Applying high reflectivity roof coating and roof or ceiling insulation give reductions of 6% and 5% each, respectively. For Case 6 ceiling insulation decreases the cooling load most by about 7%, followed by roof insulation and high reflectivity roof coating (Figure 8b). Besides, wall insulation is more effective on internal wall, followed by external wall-inside surface. Overall, all of the passive cooling techniques except floor insulation, party wall insulation and attic ventilation give greater reductions in the cooling load in Case 6 compared to Case 1, likely due to exclusion of hot outdoor air in closed window conditions during daytime. Figure 9a presents the simulated sensible cooling loads for combinations of the most effective techniques in the ventilation condition of Case 6. As before, the techniques are applied accumulatively in step-by-step basis. Compared to the current condition (Case 1), the cooling load of the master bedroom is reduced by about 39% to 30.9 MJ/day when the ceiling, internal wall, external wall-inside surface and floor are insulated (R-value 4 m2K/W) for Case 6 (Figure 9). The cooling load is lowered by 56% to 21.9 MJ/day when all of the techniques are used simultaneously, although the further reductions by high reflectivity roof coating, party wall insulation and internal shading are only about 3% or less each. It is implied from the above simulation results that changing from daytime ventilation to night ventilation is fundamental to gain better effectiveness of other passive cooling techniques for both complete NV and partial AC strategies. Due to the intense solar heat gain through the roof, roof insulation for complete NV strategy and ceiling insulation for partial AC strategy provide the greatest cooling effects. In particular, for complete NV strategy techniques that prevent external heat on the outer building envelope are relatively effective to keep the indoors cool (Figure 6b). On the other hand, for partial AC strategy insulating the inner surfaces is relatively effective to reduce the cooling load (Figure 9b). Since the master bedroom is air-conditioned in this strategy, these techniques aid to prevent the mechanically cooled indoor air from being transferred outward. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 367 CONCLUSIONS The simulation results of a typical Malaysian terraced house reveal that indoor thermal comfort may be achieved in naturally ventilated condition by applying multiple passive cooling techniques that prevent external heat on the outer building envelope and night ventilation, even under heated urban climatic conditions. When air conditioning is used in the master bedroom, reductions of about 39% to 56% in the sensible cooling load compared to the current scenario can be reached by using several techniques including night ventilating other spaces and insulating inner surfaces of the master bedroom. Further consideration of different building orientations, annual performance, cost-and-benefit effectiveness, and effects on indoor humidity as well as latent cooling load would be useful to realize their practical implementation in the urban terraced houses. Such modifications are expected to contribute largely to energy savings and carbon emission mitigation. ACKNOWLEDGEMENTS We gratefully acknowledge financial support from the Nichias Corporation, Ministry of Education Malaysia and Universiti Teknologi Malaysia for Research University Grant Program 2014 (Vot Q.J130000.2721.01K22), and The Hitachi Scholarship Foundation. REFERENCES Counihan, J. 1975. Adiabatic atmospheric boundary layers: a review and analysis of data from the period 1880-1972. Atmospheric Environment, 9: 871-905. Department of Statistics Malaysia. 2012. Characteristics of Living Quarters. Population and Housing Census of Malaysia 2010. Putrajaya: Department of Statistics Malaysia. Dimoudi, A. 1996. Passive cooling of buildings. In: Santamouris, M., Asimakopoulos, D. (Eds.). Passive Cooling of Buildings. London: James & James (Science Publishers) Ltd., pp. 35-55. Grosso, M. 1992. Wind pressure distribution around buildings: a parametrical model. Energy and Buildings, 18: 101-131. Klein, S.A. et al. 2012. TRNSYS 17: A Transient System Simulation Program. Madison: Solar Energy Laboratory, University of Wisconsin. <http://sel.me.wisc.edu/trnsys>. Kubota, T., Toe, D.H.C., Ahmad, S. 2009. The effects of night ventilation technique on indoor thermal environment for residential buildings in hot-humid climate of Malaysia. Energy and Buildings, 41(8): 829-839. Kubota, T., Ossen, D.R. 2011. Analysis of climatic conditions affecting urban heat island intensity in Johor Bahru, Malaysia. In: Summaries of Technical Papers of Annual Meeting, Architectural Institute of Japan, 2011, August 23-25, Tokyo, Japan. Kubota, T., Jeong, S., Toe, D.H.C., Ossen, D.R. 2011. Energy consumption and air-conditioning usage in residential buildings of Malaysia. Journal of International Development and Cooperation, 17(3), Special Issue: 61-69. Liu, F., Meyer, A.S., Hogan, J.F. 2010. Mainstreaming Building Energy Efficiency Codes in Developing Countries: Global Experiences and Lessons from Early Adopters. World Bank Working Paper No. 204. Washington, D.C.: The World Bank. Mohd Isa, M.H., Zhao, X., Yoshino, H. 2010. Preliminary study of passive cooling strategy using a combination of PCM and copper foam to increase thermal heat storage in building façade. Sustainability, 2: 2365-2381. Sadafi, N., Salleh, E., Lim, C.H., Jaafar, Z. 2011. Evaluating thermal effects of internal courtyard in a tropical terrace house by computational simulation. Energy and Buildings, 43: 887-893. Santamouris, M., Kolokotsa, D. 2013. Passive cooling dissipation techniques for buildings and other structures: the state of the art. Energy and Buildings, 57: 74-94. Sivak, M. 2009. Potential energy demand for cooling in the 50 largest metropolitan areas of the world: implications for developing countries. Energy Policy, 37: 1382-1384. Toe, D.H.C. 2013. Application of Passive Cooling Techniques to Improve Indoor Thermal Comfort of Modern Urban Houses in Hot-Humid Climate of Malaysia. Doctoral Dissertation (unpublished). Hiroshima University, Japan. Toe, D.H.C., Kubota, T. 2013. Development of an adaptive thermal comfort equation for naturally ventilated buildings in hot-humid climates using ASHRAE RP-884 database. Frontiers of Architectural Research, 2(3): 278-291. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 368 Thermographic Study on Thermal Performance of Rural Houses in Southwest China 1,2 Yehao Song , Prof. PhD 1 Ning Zhu , PhD Shimeng Hao 1 Junjie Li 1 3 Jialiang Wang , PhD [1. School of Architecture, Tsinghua University, Beijing, China] [2. State Key Laboratory of Subtropical Building Science, South China University of Technology, Guangzhou, China] [3. Department of Civil, Architectural, and Environmental Engineering, Missour University, MO, U.S.] hsm04@mails.tsinghua.edu.cn ABSTRACT The thermal performance assessments of rural houses are often inaccurate by thermal calculation or simulation due to complicated micro climates of rural settlements and the informal processes of selfbuilt structures. Infrared (IR) thermography is an effective and efficient tool to evaluate building and material performance. This study aims to show the possibilities of using IR imaging to better understand the thermal process of rural houses. Several typical rural houses with different kinds of building envelopes in the Southwest of China were selected. A series of thermographs were taken under various circumstances, including different seasons, time periods and weather conditions. Continuous outdoor and indoor air temperature measurements were conducted simultaneously. The results show that the correlationship between envelope surface temperature distributions and air temperature variations of adjoining rooms, as well as the heat gaining and losing processes of different building envelopes. INTRODUCTION The study on the thermal environment of rural houses is of great significance. On the one hand the rural structures are often well-acclimated with low energy consumption. On the other hand they may still need improvements to meet higher thermal comfort requirements. However, the thermal performance assessments of rural houses by using regular thermal calculation or simulation tools are often inaccurate. Because the microclimates of rural settlements are often complicated and the informal processes of these self-built structures cannot ensure the fully use of material properties. Furthermore, most rural houses are free running which means natural ventilation is enhanced. Especially in the southwest of China, the locals like having doors and windows open all day long even during the cold winter due to their living habits. Therefore, the simulation results which based on an enclosed-space model and laboratory parameters have low reliability. Infrared (IR) thermography is an efficient tool to obtain the superficial temperature distribution of the inspected object. It has a broad range of applicability and has been applied to buildings for a couple of decades [1]. IR inspections of building envelopes can be used to detect heat losses, insulation defect, thermal bridges, air leakage and moisture sources, HVAC and electrical installations can also be 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 369 inspected [2]. It’s a very cost effective tool for building diagnostic and retrofit. In addition, this technique can visualize the dynamic heat transfer process through the envelope. This study aims to show the potential use of IR imaging to better understand the thermal process of rural houses. A four-year field study was undertaken in Wulong County (Chongqing Province) since 2011. A number of research results have been published [4-7]. It is part of Hot Summer and Cold Winter (HSCW) climate zone of China. The outdoor temperature can reach 40°C in summer while it often falls below 0°C in winter, with high humidity all year round. In this study, several typical rural houses with different kinds of building envelopes were selected. A series of thermographs were taken under various circumstances, including different seasons, time periods and weather conditions. Continuous outdoor and indoor air temperature measurements were conducted simultaneously. The results show that the correlationship between envelope surface temperature distributions and air temperature variations of adjoining rooms, as well as the heat gaining and losing processes of different building envelopes. THE ASSESSED BUILDINGS The five representative rural houses we chose to take thermographs include two modern ones (built after 1990s) and three traditional ones (built before 1980s), as shown in Figure 1. House (a) and house (b) are three-storey reinforced concrete frame structures, infilled with 390mm x 390mm x190mm cement bricks. The exterior walls are 400mm thick approximately, covered with glaze ceramic tiles. The exterior windows are single-glazed aluminum alloy windows. The ground floors used as garage or store enclosed with metal shutter doors. House (c) was two-storey stone structure built in the 1950s, as the dormitory for slaughterhouse workers. The exterior walls are 500mm-600mm thick with lime plaster layers. The exterior windows and doors are single-glazed framed with wood. House (d) and house (e) are timberwork houses represent the most common vernacular architectural styles in the Southwest China. The exterior windows and doors are same as house (c). The exterior walls are 20-30mm thick wooden boards. House (d) was built in the1930s and house (e) was built in 1961. Some alterations have been made for house (e) and the exterior walls are partly replaced by exposed cement bricks. None of these exterior walls or roofs has thermal insulation layers. (a) Figure 1 (b) (c) (d) (e) Five typical rural houses of different envelopes (a) and (b) Cement brick with ceramic tiles; (c)Stone; (d) Wood; (e) Partly wood and partly cement brick METHODS Four field surveys were conducted on August 26th~27th (2011), April 13th~15th (2012), January 26th~27th (2013) and February 16th~20th (2014), respectively. Both thermographic images and visual images were taken every two hours from 8:00 to 20:00. The indoor and outdoor air temperatures were recorded every 30 minutes. To increase comparability of results, the indoor temperatures have been measured in rooms on the second floor and adjacent to the objective façades. The infrared thermographic camera used in this research is VarioCAM HR Inspect. The information of the instruments is shown in Figure 2 and Table 1 in details. Thermography is a very cost effective tool, and several methods were applied to prevent inaccuracies. 1. To mitigate the effect of incident solar radiation, we chose the façades facing north or northwest. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 370 2. Parameters that could impact the accuracy of the measurement like material emissivity, ambient temperature and distance from the target are also considered and corrected using software. 3. For building diagnostic, the measurements should performed before sunrise or after sunset to minimize the effect of incident solar radiation. In this case we chose late evening Physical quantity Surface temperature Air temperature Table 1. Detailed Information of the Instruments Instrument Range VarioCAM HR Inspect WSZY-1 7.5µm~14µm -40°C ~100°C Accuracy 0.05°C 0.1°C (a) (b) Figure 2 The instruments for measurements. (a) is infrared thermographic camera and (b) is hygrothermograph meter RESULTS Diurnal variation of different envelopes The inspection of exterior surface temperature can illustrate heat gaining and loosing process from dawn to sunset. After sunrise, the external building surfaces start to absorb solar radiation and surface temperature will increase. When environment temperature fall below external surfaces temperature, especially after sunset, heat will dissipated by radiation and the surface temperature will decrease. As the consequence of this heat exchanging process, the indoor air temperature fluctuates along with it. This process can be impacted by material, colour, weather condition and etc. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 371 Figure 3 '# Thermographs taken on February 16th (cloudy day) and 20th (sunny day), 2014. (a) is wooden board wall; (b) is stone wall with lime plaster; (c) are cement brick walls with glaze ceramic tiles (in the middle) and cement plaster (on the left). 3+,+90#:.;<=#>;0?#*;4+@# "# *+,-+./01.+23
$# 8059+# A556# 3+,+90#:.;<=# %# &#  !%# !$# !"# '(&&# )&(&&# )%(&&# )$(&&# )"(&&# )'(&&# %&(&&# Figure 4 '(&&# )&(&&# )%(&&# )$(&&# )"(&&# )'(&&# %&(&&# 345167#######################################################################################81997
Average surface temperatures of different building envelopes Figure 3 shows a series of thermographs illustrating thermal processes of four different building envelopes in winter affected by the parameters mentioned above. During the period of the measurements, the sunrise time is 7:20 am and sunset time is 6:40 pm. The outdoor air temperatures range from -2.3°C to 2.0°C on cloudy day and -2.1°C to 5.2°C on sunny day. The average outdoor air temperatures are 0.02°C and 1.17°C respectively.The average surface temperatures of different building envelopes were measured through these thermographs, as shownin Figure 4. It's important to note that 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 372 only opaque surfaces were taken into account for calculating the average temperature. That’s because it is hard to obtain accurate temperature of glazing unit due to the influence of reflection. Besides, the glaze ceramic tiles can also reflect sky and buildings around it which may lead to errors on the thermographs. Figure 4 can give us a direct impression that there are great differences in the average exterior surface temperature between different envelopes on sunny day, but little differences on cloudy day. On cloudy day, wooden envelope has the highest average surface temperature of -0.9°C, with the maximum value of 1.24°C. While on sunny day, the temperature of cement brick wall with mortar plastering is the highest, the average value is 1.36°C and the maximum value is 5.21°C. In contrast, the surface temperature of cement brick wall with glaze ceramic tiles is the lowest on both cloudy and sunny days because its light-colored and polished surface has high reflectivity. The average values are -1.02°C and -1.4°C respectively. The surface temperature of wooden and mortar plastering envelopes fluctuate strongly, with the values of 2.93°C and 3.63°C on cloudy day and 10.14°C and 9.14°C on sunny day, respectively. The amplitude of temperature fluctuation of the stone wall is the smallest due to large thermal mass and thermal resistance. On cloudy day, the highest exterior surface temperature is reached around 2:00 pm, while it peaks at 6:00 pm on sunny day except for tiled brick walls. The accumulation of solar radiation is one reason for the difference. Another significant factor is the northwest orientation led the envelope to a western exposure before sunset. Figure 5 to Figure7 shows the comparison of exterior surface temperature and outdoor/indoor air temperatures of these buildings. It is obvious that the fluctuation of exterior surface temperature is consistent with the outdoor air temperature. No heating equipment has been adopted in the rooms we measured which are adjacent to the envelopes. The interior air temperatures of these rooms are more stable and relatively low. The building with wooden envelope has the lowest temperature but has the highest amplitude of temperature fluctuation valued at 2.3°C. The average temperature is 0.42°C on cloudy day. While on sunny day, the average temperature is even lower (0.28°C). Despite the average temperature of exterior surface is higher than others, the distribution of temperature is uneven. The upper part of the wooden façade adjoining the testing room is shaded by eaves thus has a relatively low temperature. The average interior temperature of building with tiled cement brick walls is the highest (valued at 2.01°C on cloudy day and 2.22°C on sunny day). Even so, the interior air temperatures are far below the thermal comfort range of local residents. According to previous study, the 90% acceptable range in winter is 6.85-13.60°C in operative temperature [5]. From that mentioned above, we can see that the exterior surface temperature is mainly determined by the colour and smoothness of the outermost layer of wall. The light-coloured and smooth surface has a relatively low temperature (as lime plaster and glaze ceramic tile). On the contrary, the temperature of dark and rough surface is higher (as wood and cement plaster). As a consequence, for a higher surface temperature in winter, one should use a dark and rough outer layer for the wall. While in summer, a light and smooth surface is better to avoid overheating. One other thing to note is that the differences between different envelopes is greater on sunny days, but on cloudy days the differences are not distinctive. In consideration of that in this area, most of the days in winter are cloudy, a light-coloured and smooth wall surface is a balanced choice. Because the influence factors of interior air temperature are more complicated, the relationship between exterior surface temperature and interior air temperature is uncertain. The building with wooden board wall has the lowest indoor temperature while its surface temperature is relatively high. Whereas the glaze ceramic tiles has the lowest surface temperature but the indoor temeprature is higher than the others. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 373 )%# *+,-+./01.+#23
'# $# 510655.# /C+./D+# ,;9# ;9655.!%B# ,/E# ! &# !$# !'# "(&&# Figure 5 &'! 345167################################################################################################81997
Surface temperature distribution and indoor/outdoor air temperature of building with wooden walls
510655.! /<+./=+! ,:9! #! *+,-+./01.+!23
'(&&# )&(&&# )%(&&# )$(&&# )"(&&# )'(&&# %&(&&# '(&&# )&(&&# )%(&&# )$(&&# )"(&&# )'(&&# %&(&&# :9655."';! ,/>! $! ! %! "$! "#! ()%%! #)%%! &%)%%! &')%%! &$)%%! &()%%! &#)%%! '%)%%! #)%%! &%)%%! &')%%! &$)%%! &()%%! &#)%%! '%)%%! 345167!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!81997
Figure 6 &'! 510655.! /<+./=+! ,:9! #! *+,-+./01.+!23
Surface temperature distribution and indoor/outdoor air temperature of building with stone walls
:9655."';! ,/>! $! ! %! "$! "#! ()%%! #)%%! &%)%%! &')%%! &$)%%! &()%%! &#)%%! '%)%%! #)%%! &%)%%! &')%%! &$)%%! &()%%! &#)%%! '%)%%! 345167!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!81997
Figure 7 Surface temperature distribution and indoor/outdoor air temperature of building with tiled cement brick walls 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 374 Seasonal variation of different envelopes The seasonal variation is distinctive, as shown in Table 2 and Figure8. The differences between different envelopes are more significant at higher temperatures. In February, the average exterior surface temperature of wooden façade is 0.66°C higher than that of stone wall. In April, the average exterior surface temperature of wooden façade is still the highest, valued at 13.96°C. While the stone façade and the brick façade have the similar temperature valued at 12.99°C and 12.97°C respectively. The temperature gradient is approximately 1°C. In August, the temperature gradient reaches 5.13°C. Wood façade still gets the highest temperature of 27.45°C. And the temperature spans are wider in summer than the other two seasons. Table 2. Thermographic Records of Different Seasons (10:00 am) °C Avg Min Max Span SDev WINTER (19th February, 2014)   Wood -0.23 -1.13 1.81 2.94 0.34 Stone -0.89 -1.87 0.94 2.81 0.35 Tiled Brick -0.87 -2.24 0.49 2.73 0.74 SPRING(15th April, 2013)   Wood 13.96 12.75 14.51 1.75 0.31 Stone 12.99 12.54 13.75 1.20 0.18 Tiled Brick 12.97 12.56 13.62 1.07 0.15 SUMMER(27th August, 2011)   Wood 27.45 24.17 30.94 6.77 1.86 Stone 25.34 23.45 27.33 3.88 0.83 Tiled Brick 22.32 18.81 25.77 6.95 2.44 Figure 8 Seasonal differences. (a) is the building with wooden walls; (b) is the building with stone walls; (c) is the building with tiled cement brick walls 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 375 Building diagnostics The most common use of IR thermography is for building diagnostics. It can be used to identify air leakage through openings, heat losing areas and moisture problems. To minimize the effect of incident solar radiation, the measurements for detecting building defects were performed at late evening (at 8:00 pm, on February 20th, 2014), as shown in Figure 9. Thermographs (a) and (b) show windows or a door frame viewed from exterior. The “red lines” along the top of the openings show the locations for hot air exfiltration. Thermograph (b) also indicates the possible water damages at the foot of the stone wall. The surface temperature is higher in this area, since the heat is conducted through wet mass more rapidly from interior. The red part of the façade in thermograph (c) is exposed cement brick wall with no insulation. (d) is the thermograph of a reinforced concrete building. The brighter parts under the eave and balconies are exposed concrete slab. There’s no obvious air leakage around the openings, which may explain the fact that the indoor air temperature is higher than timberwork and stone houses even if the heat gains through the building envelopes are less. These resluts suggest that to improve building performance in this area, some measures should be adopted. Fill up the wall cracks and gaps around openings. The building foundation should be dampproof and waterproof. Thermal insulation mortar or polystyrene board can be used for building exterior walls. (a) Figure 9 (b) (c) (d) Detailed thermographs of building defects (a) shows the second floor window of a timberwork house; (b) shows the window and door of a stone house; (c) shows the exposed cement brick wall of a timberwork house; (d) shows windows and balconies of a reinforced concrete house CONCLUSIONS In this study, the infrared thermographic measurements of rural houses were conducted in the Southwest of China from 2011 to 2014. The results show that there are significant daily variation and seasonal variation differences between different building envelopes. The common defects have been revealed with the help of infrared thermography. Following conclusions can be made 1. The exterior surface temperature is mainly determined by the colour and smoothness of the outermost layer of the wall. Light-coloured and smooth surface has a relatively low temperature, while the temperature of dark and rough surface is higher. The differences between different envelopes is greater on sunny days, but on cloudy days the differences are not distinctive. In consideration of that most of the days in winter are cloudy, a light-coloured and smooth wall surface is a balanced choice for this area. 2. The fluctuation of exterior surface temperature is consistent with the outdoor air temperature. On the contrary, the correlationship between exterior surface temperature and indoor air temperature is not obvious. The indoor temperature is also related to the envelope structure, thickness of wall, type of openings and other factors. 3. The seasonal variation is distinctive. And the differences of surface temperature distribution 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 376 between different envelopes are more significant at higher temperatures. The temperature spans more widely in summer than the other seasons. 4. To improve building performance in this area, some measures should be adopted. Fill up the wall cracks and gaps around openings. The building foundation should be dampproof and waterproof. Thermal insulation mortar or polystyrene board can be used for building exterior walls. ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China (NSFC), Design Strategies of Chinese Vernacular House in Hot-summer and Cold-winter Climate Zone (Grant No. 51278262) and State Key Laboratory of Subtropical Building Science, South China University of Technology. REFERENCES [1] Barreira, E., and V.P. de Freitas. 2007. Evaluation of building materials using infrared thermography. Construction and Building Materials, 21(1): 218–224. [2] Balaras, C.A., and A.A. Argiriou. 2002. Infrared thermography for building diagnostics. Energy and Buildings,. 34(2): 171–183. [3] Martı́n Ocaña, S., Cañas Guerrero, I., and I. González Requena. 2004. Thermographic survey of two rural buildings in Spain. Energy and Buildings, 36(6): 515–523. [4] Song, Y.H., Hao, S.M., Wang, J.L., and J.J Li. 2012. A comparative investigation on sustainable strategies of vernacular vuildings and modern buildings in southwest China. Peru PLEA Conference. [5] Song, Y.H., Wang, J.L., Hao, S.M., and Y.L Song. 2013. The energy-related impacts of social factors of rural houses in southwest China. ISES Solar World Congress. [6] Hao S.M., Song, Y.H., and W. Zhang. 2011. Field study of summer thermal environment of rural house in the southeast of Chongqing. Eco-city and Green Building, 8(4): 90-93. (in Chinese). [7] Li, Y., and S.M. Hao. 2012. Investigation on light environment of eural house in the southeast of Chongqing. Eco-city and Green Building, 9(1): 113-116. (in Chinese) [8] Zhu Y.X. 2010. Environmental science in building. Beijing: China Architecture & Building Press, pp.47. (in Chinese). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 377 Aggregating building energy demand simulation to support urban energy design Jessen Page Sébastien Dervey Gilbert Morand HES-SO Valais-Wallis jessen.page@hes-so.ch HES-SO Valais-Wallis HES-SO Valais-Wallis ABSTRACT Designing energy efficient cities and in particular designing buildings as well-thought components of the urban fabric and active components of the urban energy system requires reliable information on the current demand in energy within buildings, its distribution in time and space and the possibility to impact on this demand. In this article we present a methodology developed to produce this information by aggregating the simulation results of a very large number of buildings. The methodology relies on the use of an existing simulation tool (bSol) and of selected default parameter values corresponding to pre-defined building typologies as well as data from existing weather and GIS databases for calibrating the tool. The core challenge being the adequate choice of default parameters related to the building’s environment (weather conditions and surroundings), its fabric (mass and envelope), its equipment and its occupants’ behaviour, we pay special attention to producing a sensitivity analysis of the tool’s results in relation to these parameters. This leads us to define a database structure for required default values and to start populating the database with robust values. We apply the methodology to recognise typical urban typologies relevant to energy planning and the parameters defining them, such as building typologies, mixity of use, urban density and morphology, existing energy infrastructure, potential for renewable energy production. For each urban typology we attempt to propose typical solutions at different levels of spatial resolution, ranging from decentralised solutions to more centralised solutions at neighbourhood, district or city level. While the methodology presented was developed within a Swiss project it can just as well as be applied to the challenge of planning and implementing energy efficiency in rapidly growing cities of developing countries, in particular to the regeneration of existing and planning of new city districts. INTRODUCTION It is widely accepted that an integrated multi-energy approach covering energy demand, supply, distribution and storage applied to a cluster of buildings (neighbourhood, district or whole city) is required for the optimal planning and design of urban energy systems. One of the main challenges in this field is being able to simulate the energy demand (in particular heating and cooling requirements) of large numbers of buildings at an acceptable level of accuracy. While building simulation tools are currently capable of simulating single buildings within a reasonable level of accuracy provided a great amount of information is made available to calibrate the tools used, simulating a large number of buildings and their aggregated demand profiles with very little input data remains a core challenge. With increasing computational power the issue is no longer computational run time but rather a reliable estimation of the discrepancy between simulated and real demand profiles (and the reduction of this discrepancy). This information is vital if we wish to rely upon 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 378 building simulation for the detailed design and operation of urban energy supply and distribution infrastructure. We present in this article a methodology to address this issue consisting in proposing well defined building typologies with a minimum amount of significant input parameters that can be used to calibrate an existing building simulation tool. STATE-OF-THE-ART Previous efforts to model a large number of buildings have been made in particular to support cities in their energy planning. Projects worth noting, for example in Switzerland, are EnerGIS [Girardin et al., 2010], MEU [Rager et al., 2013] and Zernez [Orehounig et al., 2013]. In the case of EnerGIS the energy demand from the building stock of the city of Geneva was represented by a set of 80 building typologies each corresponding to a thermal signature. The two latter projects use the CitySim [Robinson et al., 2009] simulation tool to produce hourly energy consumption profiles of buildings in either the neighbourhoods of the cities of Lausanne, Martigny, La Chaux-de-Fonds and Neuchâtel in the case of the MEU project or the alpine village of Zernez. At the European level the introduction of compulsory European Performance Certificates has generated a large amount of data on the annual energy demand of buildings as well as a variety of national calculation methods to estimate these. The TABULA project (and its follow-up project, EPISCOPE [Episcope, 2014]) proposes national building typologies typically based on the year of construction and on the size (with categories “single-family”, “terraced”, “multi-family houses” and “apartment blocks”) of the buildings, mostly limited to residential buildings, for each of the 13 countries involved. While the main objective of the project is to inform users regarding the potential for energy demand reduction in “normal” and “ambitious” scenarios of refurbishment and the associated costs it represents an extremely valuable dataset of default values that can be used for parametrizing dynamical simulation tools. METHODOLOGY The methodology presented in this article was produced in the on-going Smart Heat project [Smart Heat, 2014] whose objective it is to propose to the energy utilities of Verbier and Sierre a preliminary solution (system design and operation strategy) for meeting the thermal loads of the buildings of each town in line with their urban energy plan. OSMOSE, a tool for the design and analysis of integrated energy systems developed by the Ecole Polytechnique Fédérale de Lausanne (EPFL) [Fazlollahi et al., 2014] is used within the project to define the best possible energy conversion system for the use case, combining centralized and decentralized solutions, by using its thermo-economic optimization functionalities. In order to produce this output OSMOSE requires the knowledge of the dynamical behavior of energy demand of all buildings. For this we use the bSol software [Bonvin et al., 2007] to simulate single buildings or building zones. bSol produces a profile of hourly heat requirements to compensate for heat losses over the previous time step, but does not model the buildings’ energy production and distribution (HVAC) system. It calculates losses based on the thermal balance of heat transferred through the building’s surfaces, heat transferred via air exchanges and internal heat gains due to solar radiation, occupant presence and use of appliances. This calculation requires the user to input parameters related to the building’s fabric (e.g. U-values of the building’s envelope elements, thermal capacity, glazing ratio and g-value of windows), its use (e.g. occupant presence profiles, installed capacity of electrical appliances, temperature set-points, air exchange rates, use of blinds) and meteorological data (extracted from the METEONORM database [Meteonorm, 2014] for the site in question) such as outdoor temperature and solar radiation (taking into account the topography of the site). bSol has the avantage of existing in a stripped-down server-based version capable of treating, within seconds, batches of simulation runs containing values for the input parameters and returning simulation outputs for each building. This version was used to simulate the energy demand of large numbers of buildings in order to determine the energy demand of neighbourhoods, districts or the whole 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 379 town. This approach transfers the problem of large-scale energy demand simulation to that of 1) defining the right inputs parameters for a large set of buildings and 2) aggregating the results of single building simulation to produce realistic energy demand profiles at various levels of spatial resolution. The problem of producing input parameters for large sets of buildings is best solved by defining building typologies that can be easily related to statistical information available on a building by building basis. The Swiss federal office of statistics [Federal land registry office, 2014] provides georeferenced information for each building in the country. The Centre de Recherches Energétiques Municipales (CREM) combines this information with other sources of information (e.g. from the trade register) to provide their own geo-referenced database named PlanETer [Cherix, 2011], conceived to support municplities in the development of their energy masterplan and that is gaining in widespread use amongst municipalities in western Switzerland. We use the following information for each building: location, year of construction, year of refurbishment, building use, building footprint (typically provided as perimeter and surface) and number of stories. On the basis of this information we define building typologies based on the year of construction and the building use. Categories for the year of construction are defined as: A - “Previous to 1980”, B - “From 1981 to 1990”, C - “From 1991 to 2000” D - “From 2001 to 2010” E - “After 2010” 5 – “Hotels” 6 – “Hotels - seasonal” 7 – “Commercial” 8 – “Schools” 9 – “Administration” 10 – “Industrial” 11 – “Sports halls” 1 Categories for building use correspond to: 1 – “Residential” 2 2 – “Residential - seasonal 3 – “Restaurants” 4 – “Restaurants - seasonal” This allows us to represent the great majority of buildings to be simulated with 55 building typologies. To each of these typologies we associate the full set of input data required to run a bSol simulation (see figure 1). While some inputs are currently fixed for all typologies the values of the following input parameters depend on the typology:   for the building fabric: U-values of the roof, facade, floor, windows and window frames, glazing ratio of each facade, g-value of windows, total thermal capacity; for the building use: temperature set-point, installed maximum heating capacity, occupancy profile and installed electrical appliance capacity (for the calculation of internal heat gains). The choice of numerical values for these parameters has been made to coincide with national building regulations (e.g. SIA 380/1, SIA 2024) when possible and based on the authors’ expert knowledge when not. Each building is simulated as a rectangular polyhedron facing southwards with no obstruction other than the topography of the site (i.e. not considering neighbouring buildings). The choice of intervals for years of construction was based on the authors’ expert knowledge; its pertinence is assessed in the sensitivity analysis. 1 Buildings not covered by these categories (e.g. churches, farms) are currently considered individually but can also be added to the list of typologies. 2 The distinction between “seasonal” and “non-seasonal” categories is intended to account for the significant seasonal dependence of building use in touristic resorts, such as Verbier. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 380 Figure 1 Sample of the inputs associated to 11 of the 55 building typologies and required to run a bSol simulation of a building. DISCUSSION OF RESULTS It was not possible to acquire energy demand profiles of the buildngs being modelled within the Smart Heat project. The application of statistical analysis on measured data to validate the methodology presented here will be done in further research. In the meantime we have focused our efforts on carrying out a sensitivity analysis of our model to 1) confirm our choice of typologies based on year of construction and building use and in particular our choice of categories for these parameters, 2) understand which input parameters associated to each typology have a significant impact on the simulation results. This latter analysis is important as it will also highlight which data needs to be collected in the future to validate the model and to improve the values of input parameters. In order to confirm the choice and number of typologies we produced an analysis of variance (ANOVA) to assess whether the year of construction and the building use are indeed decisive factors allowing one to distinguish between simulations of annual energy demand. Each category of year of 2 construction is associated to a unique total U-value of the building envelope (A – 1.36W/m /K, B – 2 2 2 2 0.89W/m /K, C – 0.65W/m /K, D – 0.46W/m /K, E – 0.32W/m /K). Each category of building use is associated to a unique occupancy profile, value of installed capacity (of electrical appliances) and value of air exchange rate. The ANOVA was applied to the results of 165 simulation runs, i.e. 3 runs per typology corresponding to three states of temperature set-point: 19.5, 20 and 20.5°C. This corresponds to the typical precision of a building’s thermostat, independent of its year of construction and building use, and allows us to introduce a common variability to the simulations of each typology. A two-way ANOVA was applied on the variability of energy demand for each couple (year of construction, building use). -16 This produced an F-test value of 11.29 that has a probability smaller than 2*10 of following a distribution of Fischer-Snedecor with ν1=40 and ν2=110 degrees of freedom and confirms our assumption that the couple of parameters (year of construction, building use) is relevant in classifying the simulated energy demand of buildings and a reasonable choice of building typologies. In order to visualize the impact of the year of construction and the building use on our simulation outputs of interest (annual energy demand, peak and average power demand) we also analysed the probability distribution of annual energy demand as a function of these two parameters separately. Figure 2 shows the boxplots of these distributions for categories A to E in the case of a residential building. The parameters shown are the median and average power demand (over one hour), the number th of hours of heating over a year and the 95 percentile of the distribution representing a reasonable estimate for the “maximum power demand” that should be used to size the heat production system. The 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 381 average power demand of each category lies beyond the inter-quartile range of the previous category confirming our assumption that these distributions are distinctly different from each other that this choice of categories was appropriate. Figure 2 Distribution of power demand for categories A to E including the total number of heating hours, average power (circle) and 95th percentile (upper tail of the boxplots). We then run, for each category A to E, simulations corresponding to the 11 different building uses. 3 The annual energy demand value of each building use is represented by a symbol in the graph of figure 3. The average (mean) and standard deviation (sd) of the distribution of the 11 values per category of year of construction are given in the table on the right. The first observation is that the order of results stays the same for each category A to E. This proves that there is a systematic dependency of annual energy demand related to building use, in other words that the heat gains and losses associated to a specific building use are not affected by the building fabric in such a way that results could be interverted. They are therefore distinguishable from each other; the question is whether the spread due to other input parameters (discussed below in the design of experiments displayed in table 1) overwhelms the spread related to building use. Aside from outdoor obstruction building use proves to have the highest impact on simulation results. Also this impact (the ratio between “sd” and “mean”) increases for newer categories of buildings for which distinguishing between building use becomes essential. We conclude that distinguishing between building use is necessary for most years of construction and should not be limited to newer buildings but be applied to all categories A to E. In addition to determining the way a building is used, building use categories also determine the building fabric elements typically associated to a building use, confirming even more so the need to consider all 11 categories for each year of construction. 3 Values for mean power show similar results. Values for maximum power were not representative due to the fact that the 95th quantile corresponded, for a large number of building use categories, to the nominal power entered as an input into bSol. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 382 Figure 3 For each category A to E (represented by its total U-value) the left figure represents the annual energy demand of each building use (11 symbols) with the average and standard deviation over all 11 values given in the table on the right. In order to confirm the choice and number of input parameter associated to each typology we produced a sensitivity analysis in the form of a design of experiments (DOE) for the following 6 input 4 parameters: temperature set-point (over the interval : 18, 19, 20, 21 and 22°C), total thermal capacity 2 (0.1, 0.25 and 0.5MJ/m ), glazing ratio (from 0 to 50% more than the default values given in figure 1), building orientation (from -90° to +90° relative to the default value of 0°, i.e. facing south), outdoor obstructions (obstructions on the horizon ranging from 0 to 90°) and weather conditions (“cold”, “average” and “warm” years corresponding respectively to Meteonorm data for Verbier for an average year over the period 1961-1990 – with an average temperature of 5.4°C, for an average year over the period 2000-2009 – average of 6.4°C, and for Meteonorm’s forecast for 2030 – average of 7°C). Table 1 displays the results of the design of experiments for categories in the unique case of a residential building use. It shows, for each category A to E, the percentage change (bottom row) relative to the reference value (top row) for annual energy demand, average power demand and maximum power 5 demand for all 6 input parameters . The sign of the percentage indicates whether increasing the value of the input has an increasing or decreasing impact on the simulation output with respect to the reference value. The results clearly highlight the significant impact, for all categories A to E, of the temperature setpoint, glazing ratio and outdoor obstructions on simulation outputs while weather conditions show little impact and building orientation none at all. The glazing ratio has the interesting property of significantly increasing heat demand for older buildings (due to high heat losses) while significantly decreasing heating demand in new buildings (thanks to increasing solar gains). Similarily thermal capacity is most 4 5 The underlined value corresponds to the reference, i.e. the default value used for the typology. Non-representative values marked N/A result from the fact that the maximum power demand was limited by the nominal power entered as an input. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 383 important for new buildings. Table 1. Design of experiments for estimating the impact of input parameters on annual energy demand (E), average power (P) and maximum power (Pmax) demands. Temperature set-point Total thermal capacity Weather conditions E P Pmax E P Pmax E P Pmax [kWh/m2/a] [W/m2] [W/m2] [kWh/m2/a] [W/m2] [W/m2] [kWh/m2/a] [W/m2] [W/m2] A 231 20% 122 22% 81 23% 49 24% 29 28% B C D E 50 6% 37 5% 29 7% 23 9% 16 6% 83 7% 58 7% 45 7% 35 9% 28 11% Glazing ratio E P Pmax [kWh/m2/a] [W/m2] [W/m2] A 231 N/A 122 30% 81 31% 49 4% 29 -31% B C D E 50 N/A 37 70% 29 83% 23 83% 16 63% 83 N/A 58 N/A 45 N/A 35 N/A 28 N/A 231 -9% 122 -16% 81 -17% 49 -20% 29 -21% 50 6% 37 8% 29 10% 23 9% 16 6% 83 2% 58 2% 45 2% 35 3% 28 4% Building orientation E P Pmax [kWh/m2/a] 231 -5% 122 -8% 81 -9% 49 -8% 29 -10% 83 -2% 58 -2% 45 -2% 35 -3% 28 -4% Outdoor obstruction E P Pmax [W/m2] [W/m2] [kWh/m2/a] 50 0% 37 0% 29 0% 23 0% 16 0% 83 0% 58 0% 45 0% 35 0% 28 0% 231 45% 122 76% 81 95% 49 131% 29 166% 231 0% 122 -1% 81 -1% 49 -2% 29 0% 50 2% 37 -3% 29 -3% 23 0% 16 0% [W/m2] 50 -6% 37 -11% 29 -14% 23 -13% 16 -13% [W/m2] 83 -2% 58 -3% 45 -2% 35 -3% 28 0% Based on these observations one could conclude that: i. ii. iii. iv. v. detailed information regarding the three dimensional layout of buildings within a city is of real importance in providing reliable simulation results (although the orientation of each building is of little significance and buildings can be simulated as facing southward with no significant loss in relevance of simulation results) as well as is information regarding the amount of glazing on exposed facades and whether buildings are attached or not, as the choice of the temperature set-point is significant this input parameter should take on a variety of values whose distribution needs to correspond to surveys of real values, the thermal capacity of a building should be well estimated (although three values representing “light”, “average” and “heavy” are enough) in particular for new buildings, using averaged weather files of a particular site seem to suffice in providing reliable simulation results. CONCLUSION We present in this article a simple approach to simulating a large number of buildings at the level of a city neighbourhood or whole city based on the use of building typologies (caracterised by periods of construction and by building use) used in combination with a dynamical building simulation tool. The resulting hourly energy demand profiles can be used in combination with an energy production and distribution modelling tool (e.g. OSMOSE) or alone (e.g. when modelling building refurbishment measures) to propose a solution (system design and operation strategy) for meeting the thermal loads of 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 384 an urban area in line with targets of the municipality’s energy plan. The success of this approach relies greatly on the appropriate choice of values for the simulation tool’s input parameters. The sensitivity of the approach’s results relative to the choice of periods of construction and building use in caracterising the building typologies as well as to the input parameters associated to the typologies highly depends on the case study to which it is applied, in particular on its climatic conditions, construction practices, typical building use (density of occupancy, profile of occupancy, installed electrical appliances) and the national building regulations that impact on these. We propose in this article a general methodology that the user can apply in order to fine tune the approach to their needs. For our specific case study, an alpine ski resort in Switzerland, we can conclude that our typologies – 5 categories of years of construction and 11 of buildings use – were well chosen. In addition we are able to recognize which input parameters require most attention. In our case study providing detailed information regarding outdoor obstructions, glazing ratios is paramount. Entering temperature set-points as well as information related to building use (internal heat gains and air exchange) that are consistent with real values is also important. Finally information related to the thermal capacity and (off average) outdoor temperatures are mostly relevant for newer buildings with lower heating requirements. OUTLOOK We have not discussed the clustering of single-building simulated energy demand profiles as this research is still in progress. A successful methodology will need to integrate the variability of real annual energy demands around the simulated average and the stochastic nature of energy demand. The research presented here will allow us to determine for which parameters an uncertainty analysis needs to be done. Stochasticity of demand will be integrated a posteriori by modifying simulated profiles of energy demand. ACKNOWLEDGEMENTS The research presented in this paper is supported by funds from the The Ark Foundation (www.theark.ch). The authors gratefully acknowledge the support provided by the following partners of the Smart Heat project: Gabriel Ruiz, Loïc Darmayan from the CREM, Jakob Rager, Ben Pfeiffer, François Maréchal from the EPFL, Michel Chérix from GECAL and our colleague Michel Bonvin. REFERENCES Girardin, L., Maréchal, F., Dubuis, M., Calame-Darbellay, Nicole, and Favrat, D. (2010). “EnerGis a geographical information based system for the evaluation of integrated energy conversion systems in urban areas”. Energy, 35(2): 830-840, February 2010. Rager, J., Rebeix, D., Cherix, G., Maréchal, F. and Capezzali, M. (2013). “MEU: An urban energy management tool for communities and multi-energy utilities”. Proceedings to CISBAT 2013, Lausanne, Switzerland, September 4-6, 2013. Robinson, D., Haldi, F., Kämpf, J, Leroux, P., Perez, D., Rasheed, A., Wilke, U. (2009). “CITYSIM: Comprehensive micro-simulation of resource flows for sustainable urban planning”. Proceedings to the Eleventh IBPSA Conference Glasgow, Scotland, July 27-30, 2009. Orehounig, K., Dorer, V., Carmeliet, J. (2013). “Sustainable energy plan for a neighborhood”. Proceedings to CISBAT 2013, Lausanne, Switzerland, September 4-6, 2013. EPISCOPE project: http://episcope.eu/ (last visited September 2014). Smart Heat project: http://www.crem.ch/SmartHeat/ (last visited September 2014). Fazlollahi, S., Girardin, L., Marechal, F. (2014). “Clustering urban areas for optimizing the design and the operation of district energy systems”. Computer Aided Chemical Engineering 33, 1291-1296. Bonvin, M., Morand, G., Seppey, P.-A. (2007). “bSol : A straightforward approach to optimize building comfort and energy consumption in early design process”. Proceedings to the 10th IBPSA Conference Beijing, China, September 3-6, 2007. Meteonorm: http://meteonorm.com/ (last visited September 2014). Federal land registry office: https://www.housing-stat.ch/ (last visited September 2014). Cherix, G. (2011), “Le projet PlanETer : Planification Energétique Territoriale à l'échelle de collectivités locales”, Conférence SIG 2011, ESRI France, Versailles, October 5-6, 2011. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 385 Built Environment Sustainability Assessment of Poor Rural Areas of Southwest China WAN Li, PhD [The Chinese University of Hong Kong] wanli@cuhk.edu.hk Edward NG, PhD [The Chinese University of Hong Kong] edwardng@cuhk.edu.hk ABSTRACT Rural areas of China have rapidly developed throughout years. However, the development model of conventional rural modernization is not suitable for poor rural areas of southwest China, which is a mountainous area with scattered residents and very low development level. Unsustainable development has resulted in a series of environmental, social, and economic problems. This study reviewed the sustainable rural development theory, analyzed the development and construction situation of rural southwest China, and established an indicator framework of built environment sustainability assessment system for the poor rural areas of southwest China. This system has been established according to a new sustainable development paradigm that emphasizes endogenous development and suitability for rural southwest China. The system covered environmental, social, and economic dimensions of sustainability and considered natural and social conditions of rural southwest China. Case studies were conducted to validate the suitability of the new assessment index framework for poor rural areas of southwest China, and the capability of this index framework to recognize the various features of different cases. The outcome shows that the applicability and sensitivity of the new assessment index framework is better than the existing assessment systems in rural southwest China. This assessment framework can be applied to other similar rural areas. 1. INTRODUCTION After the sustainable principle was applied in architecture in the 1990s, several countries and regions established building environment assessment systems. These assessment systems play a significant role in urban development because they provide standards or guidelines of building design, construction and management. However, most existing building environment assessment systems are established for urban areas, and are based on modernization development models. Some mountainous rural areas, such as rural southwest China are undergoing tremendous construction and development without appropriate guidelines and assessment systems. Copying the urban development model has led to serious problems in these areas. The lack of a suitable sustainable development model and the corresponding built environmental assessment system in these rural areas resulted in a problematic issue in China. This study aims to establish a framework of built environmental assessment system for poor rural areas of southwest China that is based on a suitable development model. 2. DEVELOPMENT AND ASSESSMENT SYSTEM OF RURAL SOUTHWEST CHINA 2.1. Status and problems of rural southwest China Southwest China (Figure 1) is a mountainous and rivery area with dispersed population. This 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 386 geographical pattern resulted in inconvenient transportation system that has seriously restricted the rural development of southwest China. Villagers face difficulty in to going outside for education, health care, purchasing, and trading. The conventional brick-concrete buildings and infrastructures are difficult to be built because of high transportation costs. At the same time, several unique minority cultures have been preserved in this area because of the closed living environment. However, these minority groups are relatively marginalized geographically and psychologically. Figure 1 Southwest rural China The social development level of rural southwest China has always lagged behind. It has been considered as one of the poorest regions in China. Almost half of the villagers have an educational level that is below junior secondary school. Unsafe water, poor sanitation, and inadequate health consciousness threaten human health. Moreover, southwest China is an area with frequent natural disasters. For example, earthquakes occur frequently in some areas of Yunnan and Sichuan. Other natural disasters, such as landslides and debris flow, happen often because of climate change and environmental degradation. Given the low disaster prevention and recovery capabilities, occurrence of natural disasters is another factor that causes poverty in rural southwest China. To solve these problems and to maintain economic growth, China implemented the New Countryside Construction policies in 2005. The increased funding for rural infrastructure construction is mainly used for the construction of irrigation facilities, roads, water supply system, power supply system, communication system, and biogas. This type of top-down implementation is large-scale, fast, and effective. In some of the rural areas, which are nearby cities and have flat terrain, this modernization development model greatly improved rural life and urban-rural integration. However, this modernization plan is unsuitable for the mountainous poor rural areas. Problems emerged in this rural development process. For example:  Modernization development model requires a considerable amount of inward investment that is difficult to sustain in rural southwest China.  Top-down planning and construction work often lack enough public engagement, unable to meet the actual needs of villagers, and lead to a lack of cultural identity and sense of belonging.  The timing and content of development have not been well organized to be consistent with the rural lifestyle. The original rural life has been disrupted. (Baoxing 2009)  Traditional values are influenced by external factors. Local tradition and minority culture have disappeared because of the effects of modernization. People focus too much on to 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 387 financial capital.  Neglecting sustainable development causes serious environmental damage.  Rural industrialization requires start-up capital and long-term technical support that are not easily attainable for majority of poor rural residents.  Although some of the rural houses have been rebuilt with industrial materials, people still cannot live a comfortable, safe, and affordable life because they cannot fully understand the design and construction technology of modern architecture. The building quality and performance are not satisfactory (Xiuyan 2008). Construction costs for these structures are higher than the costs to build local vernacular houses. Under this circumstance, a considerable number of rural residents chose to be migrant workers in urban areas, which have left behind the aged and the children. Hence, these mountainous rural areas lose vigor and cohesive force, and became the accessories of urban areas and symbol of backwardness. 2.2. Existing built environment assessment systems of rural southwest China Rural southwest China faces a series of problems in the process of development because the rural modernization development model is unsuitable for such mountainous areas. Therefore, rural southwest China needs an assessment system that could provide comprehensive understanding and guidelines for sustainable construction and development. Currently, only three standards relate to rural built environment in mainland China (Table 1). The Hygienic Standard for Rural Housing and the Hygienic Standard for Rural Household Latrine (MOHC 1998, MOHURD and AQSIQ 2003) mainly focus on sanitary conditions. In 2006, the Ministry of Environmental Protection of China established the National Eco-village Creating Standard (NCES) to encourage the creation of eco-villages. This standard considered economic, sanitation, pollution control, resource, sustainable development, and public participation issues. NCES covers a broader range of rural development. However, it is a very simple standard that contains only 16 indicators, and thus, the consideration of architecture-related issues are not enough (MEPC 2006). A comprehensive built environment sustainability assessment system for rural southwest China is needed. Table 1. Standards related to rural built environment in mainland China No Standard / Assessment System Published Institution 1 Hygienic Standard for Rural Housing 1988 Ministry of Health of China (MOHC) 2 Hygienic Standard for Rural Household Latrine 2003 3 National Eco-village Creating Standard 2006 Ministry of Housing and Urban-Rural Development (MOHURD) and Administration of Quality Supervision, Inspection and Quarantine (AQSIQ) Ministry of Environmental Protection of China (MEPC) 3. BUILT ENVIRONMENT SUSTAINABILITY ASSESSMENT OF RURAL SOUTHWEST CHINA 3.1. Sustainable development model for mountainous rural areas In Europe and other developed countries, the critique of rural modernization, which focuses on the problems of over-production, environmental degradation, and spatial inequality, has been proposed as early as the 1970s. Thereafter, a new rural development paradigm that is different from the modernization paradigm has been proposed. First, this paradigm shifted the development emphasis from “inward investment” to “endogenous development”. Second, the mode of delivery for rural development has shifted from a “top-down approach” to a “bottom-up model”. Third, the structure of rural development policy has moved from “sectorial modernization” to “territorially based integrated rural 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 388 development”. Table 2 summarizes the other key points of the new rural development paradigm. Table 2. Features of the modernization paradigm and the new rural development paradigm (Woods 2011) Modernization paradigm New rural development paradigm Inward investment Top-down planning Sectoral modernization Financial capital Exploitation and control of nature Transport infrastructure Production Industrialization Social modernization Convergence Endogenous development Bottom-up innovation Territorially based integrated development Social capital Sustainable development Information infrastructure Consumption Small-scale niche industries Valorization of tradition Local embeddedness Under this new rural development paradigm, rural development should be based on the local biocapacity and cultural context. The rural endogenous development strategies focus more on food localization, traditional craft industries resurrection, sustainable exploitation of resources, and social capital improvement. Different from the modernization paradigm, the new rural development paradigm compensates for the inconvenient transportation and insufficient financial capital, makes full use of local resources, limits environmental impact, respects local culture, and benefits human development. Obviously, this paradigm is more appropriate for poor rural areas of southwest China, which have inconvenient transportation and low development level. It can improve the quality of rural living environment, maintain the vigor and cohesive force of the mountainous poor rural areas, and increase life control of the rural residents. Rural areas can then be attractive places like urban areas. In the definition of sustainable development, “development that meets the needs of the present without compromising the ability of future generations to meet their own needs”, “human needs” comprise the core issues of sustainability. According to Maslow’s hierarchy of needs theory, the basic needs of humans have five levels: physiological needs, safety needs, love and belonging needs, esteem needs, and self-actualization needs. Among these levels, the first two needs are basic human needs, and the other three are psychological needs (Maslow 1943). In southwest China, sustainable rural development based on the new rural development paradigm means that the basic human needs should be met based on the local conditions and resources. Then, comfortable and sustainable environment should be provided to meet human psychological needs. Therefore, two levels of sustainable rural development should be implemented in southwest China:  First, to reduce reliance on outside resources and improve ability of life control and risks resistance, the basic human needs should be met without external support and environmental damage. Thus, self-reliance must be achieved in basic human needs under existing bio-capacity.  Second, to provide a sustainable environment to meet the human psychological needs, the natural environment should be recovered, bio-capacity should be increased, and development capability should be improved. 3.2. Built environment sustainability assessment of rural southwest china A Rural Built Environmental Sustainability Assessment System (RBESAS) should be established for poor rural areas in southwest China. Different from urban areas, the scope of rural built environmental sustainability assessment should include buildings and other infrastructures, public spaces, farmlands, and so on. The reason for considering the entire built environment is that the rural settlement is an organism comprising of humans and the environment. The building is not the most 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 389 important part of this organism. Rural residents do not spend as much time in buildings as urban residents do. The sustainability of a ural settlement is more related to the community and surrounding environment. Therefore, in this study, the built environment of poor rural areas of southwest China includes several components:  Buildings (residential buildings, public buildings, etc.)  Infrastructures (transportations, communications, power supplies, water supplies, markets, squares, landscape, etc.)  Production facilities (farmland, livestock pens, etc.) According to the two levels of sustainable rural development previously mentioned, the criteria of RBESAS should be divided into two parts:  Self-reliance capability: To meet basic human needs without over reliance in outside resources under existing bio-capacity, and at the same time, does not reduce bio-capacity.  Development capability: To increase the bio-capacity, and to meet human psychological needs for better development. Thus, the built environmental sustainability of poor rural areas of southwest China can be described as follows (Figure 2): Figure 2 Built environmental sustainability of poor rural areas Table 3 summarizes the relationship between rural development sustainability and rural built environment. Table 3. Rural sustainability and rural built environment element matrix Buildings Rural built environmental sustainability Self-reliance capability Biocapacity protection Self-reliance on basic human needs Development capability Ecology improvement Meet the psychological needs of human Rural built environment Infrastructures Production facilities ● ● ● ● ● ● ● ● ● ● According to the matrix of rural development sustainability and rural built environment element, a framework on issues and indicators can be established (Table 4). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 390 Table 4. Framework of RBESAS indicators 1. Land & resources conservation 2. Waste management 3. Pollution control 4. Food self-reliance 5. Water self-reliance RBESAS Self-reliance capability 6. Housing self-reliance 7. Safety and security 8. Health & well-being 9. Energy self-reliance 10. Economic self-reliance 11. Sustainable landscaping 12. Sustainable agriculture Development capability 13. Culture & context 14. Inclusiveness & participation 15. Education & information 1.1 Sensitive areas conservation 1.2 Agricultural land conservation 1.3 Soil and water conservation 2.1 Construction & demolition waste management 2.2 Operation waste management 3.1 Pollution-free construction & demolition 3.2 Pollution-free agriculture 4.1 Local food production 4.2 Diversified farming 5.1 Water quality 5.2 Water efficient irrigation 5.3 Water efficient buildings & appliance 5.4 Water reuse 6.1 Regional materials 6.2 Efficient use of materials 6.3 Indoor environmental quality 6.4 Housing affordability 7.1 Settlements location 7.2 Safety and security design 8.1 Living environmental sanitation 8.2 Community basic services 8.3 Community recreation facilities and open spaces 9.1 Embodied energy of materials 9.2 Energy efficient buildings & appliance 9.3 Local & renewable energy 10.1 Local economy improvement 10.2 Activation & empower 11.1 Biocapacity improvement 12.1 Circular agriculture 12.2 Biological controls 13.1 Protection of historical & cultural heritage 13.2 Keep local characteristics 13.3 Coordination with natural environment 14.1 Barrier-free facilities 14.2 Public engagement 15.1 Education space and facilities 15.2 Information facilities The issues of this system can be divided into two parts: self-reliance capability and development capability. The self-reliance capability category includes 10 issues. Issues 1, 2, and 3 consider environmental conservation and pollution control which aim to avoid environment damage and biocapacity reduction. Issues 4, 5, and 6 involve self-reliance of human physiological needs. Issues 7, 8, 9, and 10 include self-reliance of human safety needs. The development capability category includes five issues. Issues 11 and 12 consider environmental recovery and bio-capacity enhancement. Issues 13, 14, and 15 focus on the improvement of rural built environment concerning the human needs of love and belonging, esteem and self-actualization. This system considers both environmental protection and human development needs in the poor rural areas of southwest China. At the same time, the system includes all the three dimensions of sustainability, which makes it a relatively comprehensive built environmental sustainability assessment system. 3.3. Case study To validate whether the RBESAS index framework is appropriate to poor rural areas of southwest China, three cases will be analyzed by the existing NECS for rural China and RBESAS index framework. Comparative study will be done on three cases, which have different characteristics (Figure 3). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 391 Case I is a typical traditional village that possesses most of the typical limitations, such as low income, poor infrastructure facilities, and uncomfortable living environment. It is highly relies on local resources, but the development is very slow. Many villagers have decided to find employment in urban areas for better income. Therefore, the old village is losing vigor and cohesive force. Case II is a typical top-down rebuild village that is run by local government after a huge earthquake in 2008. Reconstruction followed the modernization paradigm of rural development. The infrastructure of this village has been improved, but the increased construction cost added immense burden to the villagers. The brick-concrete house is not energy efficient because of the use of materials with high embodied energy and the lack of bio-climatic design. The function and special design are not suitable for rural life because of inadequate public engagement. Case III is a post-earthquake demonstration reconstruction project that follows the concept of sustainable development. Local traditional building technology has been innovated to improve seismic performance and indoor environmental quality as well as to maintain energy and cost efficiency. Rural infrastructure and sanitation condition have been improved to provide better living environment. Villagers were fully engaged and empowered during the cooperation of reconstruction. This project also won several awards in China and abroad (Li, Ng et al. 2011). Figure 3 The three cases in southwest China The three cases are checked by indicators of the NECS and the RBESAS. The outcome shows the number of indicators applicable and not applicable for the cases. Among the applicable indicators, the number of requirements that can be achieved by the cases and those that cannot be achieved is determined (Table 5). Therefore, the study concludes with the number of indicators of each system that are suitable for these cases, and whether each system can identify the advantages and disadvantages of the different cases. Table 5. Analyze outcome 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 392 The analyzed outcome shows that the applicability of the NECS in poor rural areas of southwest China is relatively low. A quarter of the indicators are not applicable to these cases. RBESAS showed great superiority in the applicability of these cases. Almost all RBESAS indicators are applicable for the case villages because this indicator framework has been established according to the existing situation of rural southwest China. The system considers the entire built environment, including buildings, infrastructures, and production facilities. On the other hand, the sensitivity of the NECS is relatively low. According to the assessment outcome, the performance of Case I is the poorest. Case II is slightly better than Case I, and Case III is slightly better than Case II. However, the difference among these three cases is relatively minimal. All the three cases are unable to meet more than 50% of the indicators. The RBESAS can distinguish clearly the different advantages and disadvantages of the three cases and recognize well the significance of endogenous development model. The outcome of analysis shows that the sustainability of Case I is slightly better than Case II because the latter has less consideration on energy and economic self-reliance, local culture, and public engagement. The sustainability of Case III is better than Cases I and II. Case III improved rural living environmental quality without posing damage on the environment and entailing high costs, and preserved the local culture. Moreover, villagers felt that they are the real masters of their home land because they were fully engaged and empowered during the reconstruction. 4. CONCLUSIONS Having a comprehensive understanding of built environmental sustainability of poor rural areas is one of the significant steps to solve the problems between rural development and environmental conservation in southwest China. The results of this study show that RBESAS provides an appropriate indicator framework of built environmental sustainability assessment system for poor rural areas of Southwest China. First, this framework was established based on the concepts and theories of sustainable rural development and sustainable architecture. Second, scope and issues of this system were established according to the existing situation of poor rural areas of southwest China. These ideas ensured the scientificity and adaptability of the assessment system, and were more suited for rural areas that follow the endogenous development mode. RBESAS indicator framework also appropriate to other rural areas which follows sustainable development paradigm. More specific research needs to be done to identify evaluation method of each indicator according to local situation of different rural areas. REFERENCE Baoxing, Q. (2009). "Sheng tai wen ming shi dai de cun zhen gui hua yu jian she (生态文明时代的村镇 规 划 与 建 设 )." Retrieved 7 May, 2014, from http://www.mohurd.gov.cn/jsbfld/200903/t20090316_187287.html. Li, W., et al. (2011). "MOHURD No.1 Site: Post-Earthquake Village Reconstruction and Demonstration Project in Ma'anqiao Village." Eco-city and Green Building 6: 58-62. Maslow, A. (1943). "A Theory of Human Motivation." Psychological Review 50: 370-396. MEPC (2006). National Eco-village Creating Standard. Beijing, MEPC. MOHC (1998). Hygienic Standard for Rural Housing. Beijing, China Architecture & Building Press. MOHURD and AQSIQ (2003). Hygienic Standard for Rural Household Latrine. Beijing, China Architecture & Building Press. Woods, M. (2011). Rural. New York, Routledge. Xiuyan, L. (2008). Research and Analysis of Energy Consumption on New Kinds of Country EnergyEfficient Houses. Xi'an, Xi'an University of Science and Thechnology. Master. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 393 Session 4A : Passive Design PLEA2014: Day 2, Wednesday, December 17 8:30 - 10:10, Auditorium - Knowledge Consortium of Gujarat Climate-responsive Vernacular Swahili Housing Lorna Kiamba, MPhil (Cantab) Lucelia Rodrigues, PhD Benson Lau, MPhil (Cantab) University of Nottingham, UK University of Nottingham, UK University of Nottingham, UK ABSTRACT Vernacular architecture is manifested by a large variety of forms that have a diversity of explanations. This arises from the idea that people of different backgrounds and cultures respond differently to wideranging physical environments and the interplay of socio-cultural factors. In this paper, the authors introduce the development of Swahili architecture in Kenya as a response to warm-humid climate. Initially, to encapsulate generic design recommendations for this climate, bio-climatic design responses were derived using Mahoney Tables analysis and reinforced by guidelines obtained from previous research. This revealed conspicuously supportive arguments for lightweight air-permeable buildings in contrast to the existing Swahili form that is noticeably heavyweight. An exploration of the influence of socio-cultural factors and building materials enabled the authors to explain how these and other factors may have masked or overridden the sole effect of climatic parameters resulting in the heavyweight typology. Further, field study investigations of a series of Swahili buildings in Mombasa were conducted during the warmest periods of two years. In this paper the authors focused on the findings from the investigations of one of those buildings. Results showed indoor temperatures lower than corresponding outdoor temperatures by up to 7°C during peak times. Additionally, an occupancy survey conducted during the study periods found that up to 70% (during the warmest months) and 99% (during the coolest months) of the occupants found the studied building thermally comfortable. These analyses of the environmental response of this architectural typology revealed the suitability of plan, form and fabric characteristics. It was concluded that vernacular Swahili housing could offer insights into a different and valid approach for design of passive contemporary buildings within the local warm-humid climate. Keywords: Swahili architecture, heavyweight architecture, climate-responsive design INTRODUCTION Steeped in history, the East African coast covers the coastal region of Kenya, Tanzania and Mozambique. The region developed largely as a result of trans-oceanic trade with Arabs and Indians that was facilitated by alternating monsoon winds. This led to the establishment of a number of coastal towns whose inhabitants share history, language and cultural traditions of which some scholars claim to date to at least 100A.D (Ghaidan, 1975). The lifestyle of the immigrants combined with the impact of their religion, Islam, and much else of their tradition, had a strong influence on the local inhabitants. This interaction eventually resulted in a distinctive cultural mix referred to as ‘Swahili’ (Oliver, 2007). As a result, there evolved in these coastal cities, a civilization having a character of its own, not least in respect of its architecture. Today, the region continues to thrive as a trade and transport hub. Author A is a PhD researcher in the Infrastructure, Geomatics and Architecture Research Division, Faculty of Engineering, University of Nottingham. Author B is the Deputy Head of the Infrastructure, Geomatics and Architecture Research Division, Faculty of Engineering, University of Nottingham. Author C is the Course Director of MArch Environmental Design, University of Nottingham. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 394 WARM – HUMID CLIMATE This climatic region is located between 15° North and South of the equator. Major cities found within this region of Kenya include Mombasa and Lamu, both of which showcase a rich heritage of Swahili architecture. Little seasonal variation is experienced with no distinct seasons apart from instances of heavy or light rain - slight differences may arise from variations in altitude and exposure. Annual temperature averages range from 27°C to 32°C (daytime) and 21°C to 27°C (night-time) with on and off shore breezes occurring throughout the day. Humidity levels range from 55% to 100%, and with an annual average of 75% (Koenigsberger, Ingersoll, Mayhew, & Szokolay, 1974, p. 26). Due to the regional proximity to the equator, the sun is almost always directly overhead resulting in high radiation intensities, especially at the zenith and the western orientation. Recommended architectural responses in Warm-Humid climate A review of past building trends in warm-humid climates for a significant period of the 20th century reveals that the widely accepted house type was lightweight and elevated on stilts so as to enhance full cross ventilation – much like the traditional Malay house type (see Figure 1). The typology showed quick thermal response and would cool down rapidly after sunset while exploiting breezes to offer relief to occupants (Koenigsberger et al., 1974; Szokolay, 1996). Using the ‘Mahoney tables’ design aid, as introduced by Koenigsberger et al. (1974), recommended bio-climatic design strategies were derived using recorded climate data for Mombasa. This analysis revealed main design recommendations as illustrated in Figure 2. Results revealed a typology that is principally lightweight. It works on the premise that: since the temperature differences between the outside and inside show little variation, the only substantial relief that can be gained by users is from air movement for physiological cooling. Figure 1 Examples of building types found in warm humid climate regions Top: Traditional Malay house, Malaysia. Bottom: Mijikenda house, Kenya (images produced by the authors). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 395 Figure 2A, B and C: Recommended design strategies for warm humid climates (images produced by the authors). SWAHILI ARCHITECTURE The predominant building type is the Swahili house - usually a one, two, or three storey structure set in irregular rectangular plans as shown in Figure 3. Much unlike the aforementioned lightweight typology, it is characterized by heavyweight structures made of thick coral rag walls, timber framed doors and windows, timber balconies and flat coral rag or pitched palm leaf frond roofs (more recently, these were replaced with galvanized iron roofing sheets). Figure 3 A view of the selected two storey Swahili house (Old Post Office) in Mombasa, Kenya (images from/produced by the authors). It has been suggested that form is not merely as a result of physical constraints or individual factors, but rather the effect of an entire range of socio-cultural factors, modifications by climate and the availability of materials and technology (Rapoport, 1989). Indeed, initial analysis revealed that the Swahili house manifests a complex interaction of various factors that could be responsible for this deviation from the expected lightweight typology. It is suggested that without compromising on the aesthetics, physical and social functionalities, this architectural type was adapted to fit into the milder warm humid climate. An examination of these influences sought to explain how they may have worked towards creating Swahili architecture as we know it. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 396 Socio-cultural Influence It has been noted that Swahili architecture borrowed heavily from that of the hot-dry architecture of the Arab world as a direct result of cultural integration (Ghaidan, 1975; Mombasa Municipal Council & National Museums of Kenya, 1995). Indeed, the Islamic heritage of the Swahili people is strongly manifested in their architecture through the form and spatial arrangements of their settlements. Writing on the Swahili concept of space, Ghaidan (1975) suggested that spatial organisation is culturally determined. As is typical of Islamic settlements, aspects of privacy are significant and evident in use of screens and the ‘inward’ organisation of space. Owing to the strict requirements for visual and acoustic privacy it seems highly unlikely that the prescribed lightweight solution would have been deemed socially acceptable. Also, this might explain the prominence of architectural elements such as screened balconies/windows and courtyards which not only enhance privacy through provision of semi-private outdoor spaces but also promote shading and cross ventilation. In older settlements such as Lamu, buildings were primarily double storied and used mainly for residential purposes (Oliver, 1997). In more recent Swahili settlements such as those in Old Town, Mombasa there is a slight variation to this plan; commercial activities are on the ground floor and living spaces on the top - resulting in a mixed-use building (Mombasa Municipal Council & National Museums of Kenya, 1995). Also evident is the use of ornate architectural elements for aesthetic purposes. This is apparent in the use of ornately carved doors, highly decorated balconies and decorative frieze motifs (figurative representations are rare due to the discouragement of imagery in art by Islam). Climatic Influence The typical Swahili town consists of an irregular maze of buildings arranged in dense clusters with streets measuring an average of 1.5 to 6 metres wide and punctuated by a series of open spaces, as shown in Figure 4. Climate analysis revealed that the air temperature in warm humid climates is almost always very near to that of skin temperature. This leaves sensible air movement as the main means of relief through physiological cooling. In response, the streets - considered to be public living rooms and constantly abuzz with activity - are laid out to channel cooling sea breezes, as is the case in Mombasa and Lamu (Ghaidan, 1975). Subsequent work by Deogun, Rodrigues, and Guzman (2013) and field studies conducted by the authors reveals this to be a logical assumption. Aspects related to enhanced air movement and solar sun shading appeared to be the main strategies showcased in Swahili houses. As air movement is essential to cope with the humid heat, elements such as screened balconies, window shutters and courtyard spaces were used to facilitate effective cross ventilation and thermal regulation. Additionally, narrow streets and alleyways, balconies and small enclosed courtyards played a big role in mitigating high intensity solar radiation as illustrated in Figure 5. By closely aligning the buildings, mutual shading worked by the heat transfer potential through the external walls. Similarly, balconies and window shutters worked as sun shading elements by screening of direct sunlight and prevention of glare. For open spaces, vegetation was used to shield direct solar radiation. To lower the impact of direct solar radiation, it is advisable for buildings along or near the equator to be laid out with the shorter sides facing East and West. However, mainly due to the need to channel breeze or due pre-existing street layout constraints, this was not always possible. To counter this, walls would be effectively shaded to minimise direct insolation as explained earlier. In addition to this, the reflective qualities on the outer surface of the white lime washed walls also helped reduce the impact of incident solar radiation. The typical Swahili house is relatively deep, has thick walls and roofs and single or double banked rooms. In hot dry climates, use of a heavyweight walls and roofs is valid as the warmth accumulated in the thick fabric is dissipated during the significantly colder nights. In warm-humid regions, heavyweight walls act by minimising heat conductivity potential. Despite the slightly warmer night temperatures, heat that builds up in the heavy fabric during the day may still be dissipated by facilitating air movement during the night or during periods of cooler outdoor temperature - done mainly by opening of windows. Also, where houses had flat coral rag roofs, most were eventually covered with a pitched roof with open 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 397 gable ends (Oliver, 1997), thereby creating an attic-type space that effectively reduced the impact of direct solar radiation at the zenith. In Swahili houses, windows had integrated shutters that could be opened and adjusted as necessary to encourage air movement and heat dissipation (see Figure 6). To improve occupant comfort, windows were located at body level and could be up to 2.1 metres in height. Occasionally, the buildup of temperature and humidity indoors may result in an increase above outdoor conditions. However, when the outdoor temperatures are cooler, the shutters could be opened to allow for a comforting breeze. Similarly, the screened balconies would provide zones where one could enjoy the benefit of the sea breeze. Figure 4 Left: Sectional plan of Old Town, Mombasa. Centre: Children play within one of the open spaces. Right: Ndia Kuu Road, one of the narrow streets - note the mutual shading of buildings (Plan is author-modified from Google Earth, images from the authors). Figure 5 Shading configurations found in a vernacular Swahili setting (images from the authors). Figure 6 A typical vernacular Swahili house (Old Post Office) showing A: a screened balcony and B: Window shutter C: Sketch of a typical window in use (images from the authors). Materials and Construction Influence Records indicate that earlier settlements had houses made of palm fronds, mangrove poles and wattle, and mud brick. With the onset of immigration, one was held in higher esteem if they owned a ‘stone’ coral house (Ghaidan, 1975). Correspondingly, as benefitted by its readiness of availability and usability, coral became a major building material on the East African coast. It is available in two varieties: hard terrestrial coral and soft reef coral, used for structural and non-structural purposes, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 398 respectively. As is still done in some places today, the coral was burnt to provide the lime for mortar and plaster. The coral walls were made to be notably thick, ranging between 440 to 560mm thick (Ghaidan, 1975, p.24). Field studies by the authors have found evidence of walls of up to 700mm thick in the Old Post Office building that would be able to provide a time-lag of up to 8.5 hours (Kiamba, 2010a). This facilitates delay of peak indoor temperatures thus extending periods of thermal comfort indoors. As mentioned earlier, in more recent settlements, roofs are made of locally available mangrove poles with palm leaf fronds or galvanized iron sheets with an attic below. Externally, carved doors and intricately designed balconies enhanced the facades. Inside, the use of niches, carved into the walls for display of items and decorative friezes on the plasterwork were used to enrich the interior spaces. ENVIRONMENTAL RESPONSE Despite the fact that the Swahili typology (see Figure 7) is a deviation from the lightweight norm, further research has indicated that closely packed heavyweight buildings are potentially feasible for warm humid climates. The basis behind this premise is that the heavyweight buildings which are mostly closed up during the day would allow for significantly lower solar heat gains. At night, they would then be opened up to allow for sufficient cross ventilation that would in effect get rid of stored heat. Szokolay (1996) suggests that the inherent thermal capacity of the heavyweight structure facilitates heat storage, release and dissipation as opposed to a lightweight structure that closely follow outdoor temperature variations. Related studies by Tenorio (2002) reached a similar conclusion. In an initial field study, temperatures of up to 7°C below that of corresponding outdoor temperatures were recorded during peak outdoor temperature times within a typical vernacular Swahili building (Old Post Office). Further analysis through computer simulations confirmed this reduction in temperature swings inside the house (Kiamba, 2010a, 2010b). This suggested that the heavyweight construction had a significant impact on indoor conditions, especially during the warmest hours of the day. Figure 8 shows corresponding indoor and outdoor daily temperatures recorded on the ground floor of the Old Post Office building, during the initial study period in March, 2010 (Records over the last 50 years indicate that the warmest month of the year in Mombasa is usually March). In a subsequent and substantially longer period of study undertaken in March, 2014 (see Figure 9) corresponding indoor and outdoor daily temperatures recorded in the same location and building show a similar trend. The lower (daytime) and generally more stable indoor temperatures were suggested to be due to thermal inertia provided by the heavyweight fabric and solar sun shading that effectively reduced the rate of and amount of heat gain absorbed. Also, air movement through screened openings not only provided physiological cooling for occupants but also promoted indoor heat dissipation through cross-ventilation. Figure 7: How a typical vernacular Swahili house (Old Post Office) works (image produced by the authors). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 399 Figure 8 Recorded temperatures for corresponding indoor (Ti) and outdoor (To) temperatures for a typical Swahili house (Old Post Office) in Mombasa (graph produced by authors). Figure 9 Recorded temperatures for corresponding indoor (Ti) and outdoor (To) temperatures for a Swahili house (Old Post Office) in Mombasa (graph produced by authors). A post occupancy study was undertaken to determine occupant satisfaction levels, including that of thermal comfort, within the building. The respondents consisted of 10 occupants, both male and female, ranging from 55 to 18 years of age, of generally good health and who were accustomed to warm-humid tropical climatic conditions. Typically, occupants had a metabolic rate ranging from 0.7 to 1.6 met and a clothing insulating value of less than 0.6 clo. It was established that up to 70% of the occupants found the building thermally comfortable and satisfactory during both periods of field study (conducted during the warmest time of the year) and up to 99% during the cooler months. For 95% of occupants, access to ventilation controls (primarily windows) was satisfactory and enabled them to adjust indoor conditions as required. As the building is naturally ventilated all year round, air movement is largely dependent on channeling of breezes. Whereas mechanical fans were installed in the late 1970s, occupants noted that they preferred not to use them, citing that the incoming sea breeze was adequate. Indeed, the authors found that it was quite pleasant to sit by the windows or in the balconies as one would experience a refreshing sea breeze, especially in the afternoons. Occupants indicated that the only time fans tended to be used was if there was an increase in the number of users of a space, and even then, mainly in the afternoon period in the warmest periods of the year. During the study, recorded outdoor air velocity was 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 400 almost always greater than 0.25m/s with monthly averages of approximately 3.7m/s. Similarly, indoor air velocity measurements varied with recordings of 0.2 to 2m/s measured at window openings with higher readings in the afternoons, indicating that air movement levels could possibly have aided comfort. CONCLUSION Swahili houses serve as an example of a vernacular typology that developed as a result of the interplay of social-cultural and physical factors. An analysis of these factors has shown how this seemingly ‘out of place’ architecture came to be influenced by its interactions over a long period of time. In the initial part of the study, design recommendations were found to pointedly recommend a lightweight solution that capitalises on low thermal capacity and the physiological cooling effect of sensible air movement to make higher temperatures acceptable to users. On observation, Swahili architecture was found to be the anti-thesis of this strategy as it is densely packed, notably heavyweight and with comparatively fewer and smaller openings. Initial analysis suggests that the Swahili house exhibits a potentially suitable architectural and environmental response to the local context and climate. Identified design strategies included the use of: 1. 2. 3. 4. Heavyweight building fabric to reduce the impact of solar radiation. Mutual shading of the 2 to 3 storey high buildings to mitigate heat gain. Pitched roof with a ventilated attic space to reduce impact of solar radiation at the zenith. Screened balconies and shuttered windows to promote thermal regulation by channeling breezes and sun shading while meeting requirements for acoustic and visual privacy. 5. Light coloured envelopes enhanced reflective capabilities of exposed wall and roof surfaces. Having thrived over centuries, Swahili architecture is manifested by a distinct typology that enriches the fringe of the East African coast. Even so, with increasingly rapid urbanization and the influence of 20th century modern architecture, cities in the region continue to grapple with deterioration of their built environment. This ‘newer’ and mainly lightweight architecture has been marked by the introduction of active measures that are costly and unsustainable. This has created the need to find viable climate responsive design alternatives. It is possible that implementable solutions lie in the architecture of vernacular Swahili housing - this paper is the start of this investigation. Although initial investigations suggest that the typology in combination with the aforementioned strategies is potentially an appropriate strategy in moderating the impact of the external climate on indoor conditions, further analysis is currently underway to outline in greater detail the suitability of this approach to warm-humid climate. REFERENCES Deogun, I. Q., Rodrigues, L., & Guzman, G. (2013). Learning From the Swahili Architecture in Mombasa/Kenya. Paper presented at the Sustainable Architecture for a Renewable Future, Munich. Ghaidan, U. (1975). Lamu: A study of the Swahili town. Nairobi: Kenya Literature Bureau. Kiamba, L. N. (2010a). An Investigation of Building Thermal Performance: Old Post Office. MPhil Essay. University of Cambridge. Kiamba, L. N. (2010b). Thermal Regulation Strategies for Warm-humid Climates: An approach for Non-domestic buildings. MPhil Environmental Design in Architecture, University of Cambridge. Koenigsberger, O. H., Ingersoll, T. J., Mayhew, A., & Szokolay, S. V. (1974). Manual of Tropical Housing and Building: Climatic design: Longman. Mombasa Municipal Council, & National Museums of Kenya. (1995). A Conservation Plan for the Old Town of Mombasa, Kenya. Oliver, P. (1997). Encyclopedia of Vernacular Architecture of the World: Cultures and habitats: Cambridge University Press. Oliver, P. (2007). Dwellings: The Vernacular House World Wide: Phaidon Press. Rapoport, A. (1989). House form and culture: Prentice-Hall. Szokolay, S. V. (1996). Thermal Design of houses for warm-humid climates. London: James and James (Science Publishers). Tenorio, R. (2002). Dual mode cooling house in the warm humid tropics. Solar energy, 73(1), 43-57. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 401 ‘The Open Air Office’ Climatic adaptation of the office building typology in the Mediterranean JONATHAN NATANIAN (MSc) Technion – Israel Institute of Technology leyonatan@gmail.com ASTRACT The paper reports on a dissertation project undertaken at the Architectural Association School of architecture towards a Master’s degree in Sustainable Environmental Design by the Author. The starting point for the research was the troubling common practice of office buildings in Israel - racing towards the tallest, highly mechanical and fully glazed buildings, scarcely applying any environmental considerations and occasionally using the local green point system code to render them as ‘green’. This research aims at contextualizing the office building typology to the local Mediterranean climatic conditions of Israel; based on a theoretical framework and a detailed study of local common practices and climate, followed by an analytic optimization study, the ‘Open Air Office’ concept is introduced: one which uses an integrated environmental design approach to rethink some of the core values that drive office building designs in Israel today. INTRODUCATION Adoption without adaptation In contrast to a history of successful adaptations of international building styles in Israel which demonstrated high sensitivity towards the local climate, the adoption of the fully glazed office building typology in Israel- mostly throughout the late 80's and 90's- has been almost automatic (Figure 1), with very little awareness towards the environmental impacts of these buildings on all levels - from the cityscape to the occupant. (1) Figure 1 Gradual abandonment of climatic considerations throughout the history of Israeli Architecture. (Source, left to right: Le Corbusier, author’s sketch, Arieh Sharon, Emporis, and Telavivinf.com) Jonathan Natanian is a Teaching Fellow in the Faculty of Architecture and Town Planning, Technion – Israel Institute of Technology, Haifa, Israel. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 402 Starting from evaluating the performance challenges of this common practice in Israel against the potentials of the local climate, this research began by questioning how to improve performance by taking simple considerations into account in the early stages of office buildings design in Israel; using the common practice as a base case this study gradually explored the possibilities of optimizing both comfort and efficiency by applying preliminary, good and further on best practice environmental strategies that fit Israel’s climatic conditions. In conclusion, this research aimed at offering a platform for architectural design that will give a new angle to the correlation between the contemporary indoor office space and the local climate. Methodology Firstly, a study was conducted on new global workplace trends and explored the potential of new occupancy patterns to affect both comfort and performance levels of office buildings in the near future. Secondly, the local climate of Israel was studied, followed by relevant basic and advanced environmental strategies evaluation. In order to study the existing context, combined literature and analytic work were used to help gather insights regarding the current performance, layout and materiality of the local common practice office building. These insights helped highlight specific challenges and potentials, so as to define the base cases towards further studies. The analytic work, which was focused on thermal, daylight and Solar geometry aspects, moved through different levels - from the simplest preliminary environmental concepts to more advanced ones, evaluating their potential to improve both efficiency and comfort levels within the typical office space; Simulations were conducted by using Tas (thermal, by EDSL), Ambiens (CFD, by EDSL), Radiance (Daylight, by Berkley lab) and Ecotect (solar geometry, by Autodesk) software. The last part of this research explored and analyzed the unique opportunity of the Mediterranean climate to open up the building’s envelope towards the outdoors throughout different seasons of the year with correlation to the changing internal office layouts. The performance analysis for this ‘Open Air’ concept, was followed by an applicability study of one possible office building configuration, which proved its potential to work very well with the local climate and office culture while providing high performance and comfort levels within the office space. THE FUTURE WORKSPACE Nowadays, in contrast to the prescribed tasks of the traditional “office factory” through single fixed workstations, the economic shift towards the “knowledge” society has created a need for variety of alternative spaces, with higher levels of interaction and autonomy (Harrison et al. 2004); As new technologies enable people to work virtually anywhere and new interaction between users within the office space through virtual computing (Johnson et al. 2011), a new office layout terminology has been evolved which encourage higher levels of flexibility, collaboration and autonomy (Duffy, 1997); The Israeli work culture, characterized by a vivid, creative informal atmosphere, with strong communal routs is highly exposed to global trends, which are commonly taken into account during the design process. Therefore, the future trends which shape workspaces globally served as an important anchor for this research. ISRAEL’S OFFICE BUILDING DESIGN CULTURE Common practice design The research has focused on the climatic and urban context of the Tel Aviv metropolitan area, the economical center of Israel, due to its large conglomeration of office buildings. Building design. High rise office buildings are becoming common within the Central Business District of Tel Aviv. The Israeli market is highly speculative, predominated by “Shell & Core” buildings, being designed within a central core rectangular layout due to costs and internal space considerations. Materiality. Reinforced concrete frame & floor slabs is the prevailing framing method, cladding is usually done with different variations of curtain wall systems which combine glazing and opaque 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 403 cladding materials; Fully glazed buildings are very common and solar control is usually applied by reflective or tinted glazing systems combined with internal venetian or roller blinds. Office space layouts. The demand for enclosed private offices is still relatively high; the common practice consists of 4.5 deep office spaces with changing widths. Double-sided or single-sided open space offices are becoming more and more common, with depths ranging from 9m (single-sided) to as much as 30m (double sided); These observations helped generate the base case as shown in Figure 2. Performance criteria and benchmarks Although it is widely acknowledged (McKinsey and Co. 2009) that most buildings in Israel are big energy consumers, official energy consumption database or performance criteria for office buildings are currently unavailable. In order to establish a reliable reference point for the performance studies throughout the research, data was gathered from HVAC experts in Israel and was correlated with published cooling energy consumption data from similar climatic zones within the US. For this comparison, Tel Aviv climate had been considered as part of the A2 category (using ASHRAE international climatic zone definition) and energy consumption data of office buildings in Houston Texas was selected as a general reference. (2) 4.5 m (3) Figure 2 Common practice office space base case. 9m 15 m Figure 3 Common practice energy cooling demand for 3 typical office layouts1 (in kWh\m2, Source: TAS). Common practice performance Based on findings from the common practice design studies, three layout configurations have been 1 modeled, as shown in Figure 2. The boundary conditions for all the three base cases have been closely considered in regard to typical glazing, solar control, occupancy and materiality properties which have been identified as prevailing during the common practice studies. As a starting point, in order to establish a reference for further parametric optimization studies, a thermal simulation was conducted using TAS for the three different base cases facing West (or East West) orientation (a very common orientation for office buildings with proximity to the Israeli coastline). The combination of high exposure, together with their West-facing orientation, resulted in very high cooling energy demands across all three office layouts, mostly in cellular office layouts due to 1 Boundary condition in all cases included fully glazed 75% WWR designs, oriented West (or West-East orientation in double sided configuration), no external shading, w. internal semi opaque blinds. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 404 their high window to floor ratio (Figure 3). These results were found to correlate well with the data gathered from the actual practice in Israel, in which HVAC systems are being designed to supply cooling 2 energy of approximately 200 kWh/m annually. ANALYTIC WORK - OPTIMIZATION PROCESS Preliminary optimization The first level of optimization was set up to explore the hypothesis that a high level of efficiency and comfort could be gained by applying very basic environmental design strategies at the very early design stages of office buildings. The parametric approach included four different strategies which were assessed separately (Figure 4), and were conclusively merged into different typical combinations according to the common office building design scenarios in Israel (Figure 5). Set point temperature. When considering the adaptive comfort model, the common fixation of the o temperature set point on 23 C or even lower during cooling period seems inappropriate. In order to test the implication of set point changes on performance, the three base case models were thermally o simulated for three different set point temperatures (23, 25, 28 C); Simulations showed how this simple operational change could substantially improve the annual cooling loads in all office layouts by 20-30%. Orientation. Thermal simulations for the three different layouts revealed large amplitudes in annual cooling energy demand between different orientations. In contrast to the Cellular base case which is more thermally fragile, open space layouts reacted more mildly, with the double sided model showing very limited performance effect by orientation. Shading. Different external shading geometries were simulated and studied in Ecotect for all four major orientations. After daylight levels had been verified through simulations (Radiance), selected optimized shading configuration were modeled thermally TAS. Window to wall ratio (WWR). This study explored the balance between minimizing exposure for thermal considerations while insuring adequate daylight levels for the 3 base case layouts through different orientations. (4) (5) (A) (A) (B) (C) (B) (C) Figure 4 Parametric optimization study of annual cooling energy demand for 3 typical office layouts (in kWh\m2, Source: TAS). Figure 5 Annual cooling energy demand for different preliminary optimization scenarios (in kWh\m2, Source: TAS). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Scenario A - Optimized configuration (A+B+C+D). Scenario B - Fully glazed configuration, all parameters applied besides the reduction in WWR (A+B+C). Scenario C - Sea View/Western configuration, all parameters applied besides optimized orientation (A+C+D). 405 Preliminary optimization. Figure 4 shows the effect of each of the parameters on the base case cooling energy demand; the chart demonstrates how orientation is critical for cellular offices (A); when oriented North, cellular offices will perform better than the double sided open space (C) due to reduction of solar gains being replaced by the effect of other internal gains. The chart also shows the changing trends between single (B) and double sided (C) open spaces, the latter being less effected by orientation due to its double orientation exposure. The combination of different strategies (Figure 5) indicates similar trends in cooling energy demands between the three layouts, in which differences observed between the three base cases dissolved in different scenarios when solar gains were being effectively modulated by applying external shading. Good practice optimization Aiming at higher performance and comfort levels, the following studies evaluated the potential of more advanced strategies to further reduce the resultant temperature and cooling loads accordingly; Glazing properties. The thermal balance of glazing, being a dominant component, was studied; for each base case, five envelope properties were considered, each offering a different balance between heat gains vs. losses as well as different daylight intensity (Figure 6). Thermal mass and night ventilation. The possibility of applying high thermal capacity materials within the common office space layout was narrowed down to the 3 scenarios; (a) Exposed concrete floor, (b) a + concrete partitions, (c) b + exposed ceiling. Thermal simulations for the three scenarios revealed considerably lower days within comfort for scenario (c) during cold periods, indicating that heating might have to be introduced, and that the ‘cold’ outcome might work better in cases of high internal gains due to occupancy patterns or equipment usage. Natural ventilation. These studies evaluated the balance between space cooling and physiological cooling considerations; after determining the required air flow rates needed to optimize performance, CFD simulations had been used in order to evaluate resultant temperatures and air flow rates throughout the office space for a typical mid-season week. Natural ventilation studies proved that in periods when external temperature reaches above certain limits, the space cooling effect achieved by natural ventilation was often very limited; however, under the same conditions, physiological cooling effect by air movement throughout the space could effectively lower the resultant temperature towards comfort. (7) (6) Cooling ↗ Heating ↘ Figure 6 Annual cooling energy demand of five different glazing properties scenarios for Cellular base case (in kWh\m2, Source: TAS). Figure 7 Cooling energy demand for different good practice optimization parameters (in kWh\m2, Source: TAS). Good practice optimization. The study showed how further efficiency improvements could be achieved by applying advanced strategies (Figure 7); however, high internal gains will still generate an inevitable need for cooling (mostly during hot period), and by cutting off solar gains during cold period, 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 406 heating demand will increase. Nevertheless, during the process of their evaluation, strategies such as natural ventilation or thermal mass should also be measured against their added values for the quality of the space; i.e. by creating desirable air flow and stabilizing internal temperatures. Towards best practice Aiming towards the ‘zero carbon’ office space, this part of the research, within the recognition of air conditioning as the main energy consumer, focused on the passive cooling solutions that might reduce or eliminate its use. After four selected passive cooling strategies had been evaluated (based on data from contemporary research, literature and precedents), this section highlighted radiant and ground cooling as the most effective passive cooling strategies for Israel’s climate. However, considering the highly speculative Israeli market, the applicability of these strategies in Israel is expected to face strong market barriers, and most probably would be considered only for a ‘use value building’ (Harrison et al. 2004), custom designed for specific end user organization. The limiting potential of the Israeli air to effectively absorb excess heat indicates the need for hybrid or mixed mode systems in which low energy mechanical systems are coupled with natural forces; e.g. hybrid evaporation systems could be very efficient, as well as heat recovery mechanisms which could be coupled with the ground cooling system. These should be integrated aside renewable systems (which were not addressed in this research), mostly solar systems, in the light of Israel’s high solar radiation availability. THE OPEN AIR OFFICE CONCEPT (8) The inspiration for an ‘open’ approach for office buildings in Israel was drawn from the local building tradition; one in which the potential of outdoor and semi-outdoor activities to take place throughout a considerable time of the year had been widely recognized, mostly throughout the pre- AC era. The ‘Open Air’ concept aimed at reintroducing that potential to the contemporary office building typology (Figure 8). The need for diversity and informal reflection and meeting areas as part of the new office space, generated an excellent opportunity to reinvent these spaces in the Israeli office building typology through transitional or outdoor spaces. A dynamic semi-outdoor space, which could be opened or closed according to the user demand and the outdoor conditions, could serve as an extension to the internal office space and activity. Analysis methodology For the thermal and daylight analysis, a double sided section had been chosen with similar proportions as the previous 15m base case used throughout this research. The layout of 15m X 12m which included the fully optimized configuration (as previously studied) was coupled with a 3m transitional space projected towards to South, with openable full height glass partitions between them calculated to be opened when the adjacent semi outdoor space temperature was within the boundaries of comfort (Figure 9). 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Figure 8 the Open Air Concept evolution 407 Thermal simulations for the internal office space in a free running mode (Figure 10), revealed resultant temperature fluctuating very closely to the external ones and relatively high levels of comfort hours yearly; strong potential was evident for opening the office space to the outdoors in sunny days during cold-periods, as well as through most of the mid-period. Daylight simulations for the same boundary conditions showed the potential of the adjacent space to serve as an effective buffer from direct sunlight without compromising the required daylight levels in the office space. (10) (9) Internal space Semi outdoor space Figure 9 Case study model for the Open Air Concept analysis. Figure 10 Annual resultant temperature simulation (Source: TAS). External temperature (oC) Internal office space resultant temperature (oC) ASHREA 55 comfort band Thermal vs. visual comfort By applying adaptive opportunities and space diversity, the occupants interact both with building elements and with the building space. In addition to the basic environmental strategies applied (orientation, exposure, ventilation etc.), the application of the adjacent semi-outdoor space towards the South is taking performance further - by serving as a mediator between internal and external conditions, and by offering a good balance between thermal and visual issues. Figure 11 shows how the deep projection addresses glare and excess heat issues effectively while the adjustable louver-shelf serves as solar protection for the semi-outdoor space and redistributes light towards the depth of the office space. (12) (11) Figure 11 Adaptable envelope design. Figure 12 Exterior view of the open air case study showing the design expression of the environmental concepts. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 408 Activities The diversity of space which is required to address new patterns of office activities have been addressed and distributed according to visual and thermal considerations; e.g. presentation areas in less exposed spaces, informal areas in semi outdoor spaces with higher tolerance levels to the outdoor climatic conditions and fixed workstations, with higher control levels and stabilized thermal conditions. Design expression The external figure of the building (Figure 12), reflects the shift of mind from the sealed glass building, disconnected from its environment, to a new model in which different levels of exposure drive the building’s external image. The correlation between the external layout and the performance becomes complete and corresponds with the local architectural design language from earlier times in Israel’s history, in which climatic considerations helped contextualize the modern trends to the local climate. CONCLUSION This research showed that by adopting an environmentally responsive approach, the troubling performance of the Israeli common practice office could be dramatically enhanced - even up to a ‘zero energy’ balance. As overheating was identified as the main issue, optimization phases mostly focused on blocking solar heat gains while dissipating internal ones; while preliminary optimization studies showed how effective simple design decisions could be, In further optimization levels, when direct solar gains were effectively blocked, internal heat gains became predominant, and issues such as control, adaptability and thermal vs. visual comfort balance became critical to comfort and performance. Aside the more ‘technical’ aspects which evaluate comfort and performance through numerical prediction, the architectural performance of the space must also be accounted for and correspond to the office concepts of tomorrow. In the new work environment where borders dissolve (e.g. ‘home’ and ‘work’ or ‘virtual’ and ‘real’), the same levels of flexibility and adaptability will have to be applied to the spatial differentiation between ‘in’ and ‘out’; The ability of the coastal Mediterranean climate to dissolve these boundaries within comfortable outdoor climatic conditions, offers the unique opportunity to open up the sealed office ‘glass box’ to the outdoors. The open air office approach had been incorporated into this concept, in which new office organizational trends and spatial design values reintroduce the potential for working with the Israeli outdoor climate. ACKNOWLEDGEMENTS I would like to thank Professor Simos Yannas for his personal guidance and feedback throughout the process of my work on this research. I would also like to thank the Architectural Association for the bursary that supported this research. REFERENCES ANSI/ASHRAE/IESNA Standard 90.1. 2007. Normative Appendix B – Building Envelope Climate Criteria. International climatic zone definition Autodesk Ecotect Analysis. 2011. Autodesk. Dixon, M. and Ross, P. 2011. VWork: Measuring the benefits of agility at work. Unwired Ventures Ltd. UK. Duffy F. The New Workspace. 1997. Conran Octopus. EDSL Tas (v.9.2.1.1). 2011. EDSL Ambiens (v.9.2.1.1). 2011. Harrison, Wheeler and Whitehead. 2004. The distributed Workplaces. Spoon Press. Johnson, Counsell and Strachan. 2011. Trends in Office Internal gains and the Impact on space Heating and Cooling. CIBSE Technical Symposium, Leicester, UK. McKinsey and Company, Greenhouse gas abatement potential in Israel, Israel’s greenhouse gas abatement cost curve, November 2009. Radiance 2000, Berkeley Lab. US Department of Energy National Renewable Energy Laboratory (NREL). http://www.nrel.gov 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 409 Session 4B : Low carbon cities and neighborhood development PLEA2014: Day 2, Wednesday, December 17 8:30 - 10:10, Compassion - Knowledge Consortium of Gujarat Solar radiation availability in forested urban environments with dry climate. Case: Mendoza Metropolitan Area, Argentina. Mariela Arboit, PhD and Ernesto Betman, PhD Instituto de Ciencias Humanas, Sociales y Ambientales. (INCIHUSA – CONICET) Av. Adrián Ruiz Leal s/n Parque General San Martín. (5500) Mendoza, Argentina Tel: +54 0261 524 4310 Fax: +54 0261 4287370 Email: marboit@mendoza-conicet.gob.ar ABSTRACT The aim of this work is to advance the understanding of the solar potential of urban residential environments which, by their morphology, and the impact of urban trees, present values of irradiance that is very different from full solar collection. Morphological variables of urban settings and urban trees, a very distinctive feature of the Mendoza Metropolitan Area (MMA), have a fundamental impact on the feasibility of implementing strategies for solar energy utilization in urban environments. The results achieved will contribute, modify and gradually update urban and building legislation to implement higher levels of energy efficiency and minimum environmental impacts. This work proposes to study the potential of solar collection in urban environments, analyzing eleven urban configurations selected according to their building and urban morphological characteristics. Results obtained so far indicate that solar masking is critical for vertical surfaces, with a reduction of the available solar energy between 2% and 66% in the winter season. The first case corresponds to a high density homogeneous area, with wide street channels, at a third of building height, with little influence of surface shaded by the neighboring buildings and trees. In the second case, the impact of masking produced by the unleafed branches in winter is considerable and comes from species of first and second magnitude (10 m. high or more), constructions of low density and narrow street channels (13 to 16 m.). However, these drawbacks caused by urban trees are compensated by benefits in the warm season (Brager, et. al. 2001): controlling the intensity of the urban heat island, absorption of pollutants, cooling and humidifying the air through evapotranspiration, reducing thermal loads of buildings, occupancy of public open spaces, and an invaluable contribution to the urban aesthetic. INTRODUCTION The present work studies available solar radiation on the North facades of highly forested urban environments typically of the Mendoza Metropolitan Area (MMA), as part of the environmental and energetically sustainable development of dry land cities within Central Western Argentina. The Mendoza Metropolitan Area is the most important political, cultural and economic urban setting of the whole MMA. Geographically, MMA is sited on the mesothermal arid region of the Andean alluvial plane, with coordinates at: latitude: 32.85, longitude: 68.86 and an altitude of 827 m.a.s.l. The region’s main climatic variables are: DBT (min. yearly mean): .8 Cº; DBT (max. yearly mean): 32.30 ºC; RH (yearly mean): 56 %; Heating DD, (base 18 ºC): 1.384; Cooling DD, (base 23 ºC): 163; Annual hours: A 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 410 in comfort: 21.53 %, heating needed: 70.14 %, cooling needed: 8.33 %; solar resource (4.58Kwh/m2 day to 5.55 Kwh/m2). Regarding the relationship between habitat and energy, the unsustainability of the current development is aggravated by the fact that the situation inexorably deteriorates through time due to a lack of knowledge, management and planning actions. Some global figures are eloquent enough to describe it. At a global level, the habitat, which is the group of cities and buildings of the world, consumes 50% of the total energy used, leaving the other 50% divided into approximately equal parts between transportation and industry (de Rosa, 1988). Given this scenario, it is very important to take advantage of the natural resources and, in particular, to optimize access to the solar resource in order to diminish the consumption of non-renewable energy (Capeluto, et. al. 2003; Jenks, 1996; Owens S. 1986; Town and Country Planning Association. 1996). Given this concern, the following question arises: What is the impact of urban-building morphologies and of the different tree species that predominate in the urban environment with regards to the availability of the solar energy resource? In the MMA, several specific studies have been developed, considering the representative groups of the urban-building morphology. These groups were determined by the solar potential in low and high density environments, which are highly forested. The objectives of the studies were: i. preserving the physiognomy of the forested city, maintaining the scale and homogeneity of the constructions and the aesthetic and environmental contribution of urban forest. ii: enabling the maximum use of solar resource for space and sanitary water heating, through the control of urban morphology and of the forest. iii: improving habitability conditions of the building area. iv: contributing to the sustainability of local development by enabling the recycling of existent constructions that were well maintained and of good quality. In this way, extractive processes and solid waste emission (rubble) in the ecosystem are reduced. In order to complete these studies, a solar irradiance measurement was performed on the North facades of 11 selected cases according to previous results. The intention of the study was to offer a contribution, both conceptual and operative, that through the transfer channel in the future, the official sector will become aware of the seriousness of the situation and soon begin to implement new urban and building norms that could also reduce the consumption of natural gas and other non-renewable energy in urban buildings. METHODOLOGY TPCA SMA TPCA TPCA In previous works (Arboit, et. al. 2008) the Mean Insolation Factor (MIF) indicator was defined. Mean Insolation Factor (MIF) provides a measure of the potentially collecting North facing walls, not masked by neighboring buildings and trees, calculated as: the ratio of the sum of insolated collecting areas of North facing walls, times the sum of the energy received at each considered hour, during a heating season, to the sum of the total areas of the same surfaces, free of all masking, times the sum of the hourly impinging radiation during the considered heating season, as a percentage. Defined in Fig. 1. Solid Masking Area (SMA) (Constructions) Permeable Masking Area (PMA) (trees) PMA SMA PMA PMA PMA IS IS Total potential collecting area of the North facade (TPCA) Net Insolated Area (NIA): Addition of the values of insolated surfaces IS+ (PMA*P) TPCA: Total potential collecting area of the North facade (m2) SMA: Solid Masking Area (constructions): Potential collecting facade affected by shades projected by constructions of close buildings (m2). (Figure 1) PMA: Permeable Masking Area (trees): Potential collecting facade affected by shades projected by urban trees (W/m2). P: Monthly Permeability Factor. Solar permeability percentages of each vegetable species and for each month of the heating season (%). (Cantón, 2000) R(m d h): Energy by surface unit available in the North facades for each hour, day and month of the heating season (Wh/m2). Hourly impinging radiation on North facing walls for each month (Wh/m2). Sub indicators: m: Month of heating. It varies between April and September. Number of months to which the heating season is extended (n). d. Day of the month. It varies between 1 and 30. h: Hour. It varies between 9:00am and 6:00pm. Figure 1 Descriptive diagrams of TPCA, SMA, PMA and NIA. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 411 08 MIF = 30 14 :30 ∑ ∑ ∑ [TPCA m : 4 d :1 h :9 :30 ( − SMA m d h 08 NFA ×∑ ( + PMA 30 14 :30 m −d−h ∑ ∑ R m −d−h ( ×1 − P )))×R m − d − h ] ×100 (1) m : 4 d :1 h :9 :30 MIF = [exp(4.1969 − 0.1486 St.Wi 0.0127F.Tr ess + 0.8547 B.Morp − 0.4721 TOF) ] [1 + exp(4.1969 0.1486 St.Wi 0.0127F.Tr ess + 0.8547 B. Morp − 0.4721 TOF) ] (2) Considering the Mean Insolation Factor (MIF) on North facades and using a Multiple Linear Regression Model, the main variables that influence the access of the sun were determined (Arboit, et. al 2008). These variables are: Building Morphology (B. Morp.), Street width between construction lines (St. Wi), Fullness of trees (F. Trees) relationship of the existing (healthy) trees around a city block to the total number of trees that could fit around the city block, considering the dominant species and their corresponding distance between individuals, as percentages , Total Occupation Factor (TOF) Total built up area to total buildable area of corresponding parcels, as percentages . (eq.2). The easiest way to note the influence of the urban-building variables with access to the sun is by measuring global solar irradiance over the vertical plane of the completely sunny North facade and measuring the same variable of the facade that is partially sunny. The latter is affected by the urban building morphology (building morphology, streets width, fullness of trees, building height). The measurements were carried out with portable irradiance data-measuring equipment that had a constant spectral transmission factor for all wavelengths between 0.3 and 60 µm. The measurement period was established according to the latitude and longitude of the place of study, for a given day. Considering the local time, the solar noon was defined (≈ 1:30pm) and four and a half hours before and after the solar noon were taken into consideration. This approach left nine definite solar hours to measure from 9am – 6pm. Data was registered every minute during the month of July, August, and September of 2013. Based on what has been presented, the corresponding measurements were selected during a clear day for each urban environment. Clear sky conditions allow the evaluation, in its complete magnitude, of the influence of the trees and the urban morphology. Given that the measurements were realized during a winter period, the condition of unleafed branches has been evaluated. In this work we have utilized two irradiance sensors to evaluate each urban environment, one for the measurement of plain sunlight and another to register the shaded sections (Cárdenas, et. al. 2012). One must have in mind that in tree-covered urban environments irradiance can present important variations. In order to incorporate these characteristics it is suggested that a grid is drawn with two meters of separation and that a hemispheric photograph is taken towards the North at each corresponding point. In the hemispheric photograph one can overlap the sun trajectory to find the availability of the natural resource every hour of the day for that specific location in the environment. In order to calculate the received energy, one can apply a model of global irradiance on inclined surfaces (in this case 90º) that has been validated for different latitudes and climates. The proposed models for Liu & Jordan or Brichambaut can be appropriate (S. Benkaciali et.al 2012). In one of the case studies carried out in this work the application of hemispheric photographs is studied in order to implement this proposal. Selection of case studies: For the monitoring of global solar irradiance in North facades affected by solid and tree masking, 11 cases were selected, in which different measuring conditions. For the selection of case studies, the urban-building and forestal morphology of the MMA urban environments was considered, based on the Mean Insolation Factor (MIF). On Table 1, the case studies and the characteristics of urban and building variables can be observed. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad 412 Table 1 List of urban and building variable values of the sample. Cases N° Blocks Density 1 2 3 4 5 6 7 8 9 10 11 High Low High Low Mid High Mid Low Low Mid Low Street Tree Orien- width Magnitation (°) (m) tude 13 30.00 2ª 4 13.40 2ª & 1ª 13 30.00 2ª 5 16.40 2ª & 2ª 5 16.40 2ª 8 30.00 2ª 7 14.20 2ª & 1ª 29 20.88 1ª 29 20.18 1ª 29 20.18 1ª 29 20.88 1ª URBAN VARIABLES Urban forest Tree species Morus alba Morus alba y Fxaxinius excelsior Morus alba Melia azedarach y Morus alba Morus alba Cupressus sempervirens Morus alba y Platanus acerifolia Platanus acerifolia Platanus acerifolia Platanus acerifolia Platanus acerifolia Fullness of trees Several pairings 2 individuals Several pairings 2 pairings/ Young Individual Individual Several pairings Individual/pruning 2 pairings Individual/pruning 2 pairings BUILDING VARIABLES Front Building Setbacks Morphology Height (m) Regular Irregular Regular Regular Irregular Irregular Irregular Regular Irregular Irregular Regular 0 5 0 0 3 2 0 0 0 0 0 3rd 1st 2nd 1st 2nd 1st 2nd 1st 1st 2nd 1st ANALYSIS OF RESULTS Figure 4 shows the diversity of energetic situations according to the analyzed urban environment and it shows a preponderance of each of the variables in equation 1. Measurements in shaded surfaces registered a reduction from 2% to 66% of the total energy received in the plain sunlight sections. Total Energy in North Facades depends on the urban building environment; the 2% reduction corresponds to case 1 with an urban channel of 30 meters and homogeneous building morphologies that do not present shaded surfaces by neighboring buildings. In addition, its height is 10 m (with scarce influence of forest at the crown canopy level). The maximum reduction percentage, 66%, corresponds to Case 7, which is characterized by a wide street channel of 14.20 m, and the measurement surface is partially shaded by the forest at a second magnitude Morus alba and first magnitude, Platanus acerifolia, with the consequential reduction of the Figure 4 Total Energy %. available energy. For a more detailed study, Cases 1,2,3,6, and 7 have been considered, because they are urban morphologies in which each of the variables in equation 2 shows a well defined preponderance. In this way, the variables of street width, fullness of trees, building height and morphology will be analyzed and we will observe how they will influence the quantity of the received solar energy in North facades. Influence of the Street Width: The street width and their orientation are decisive of the solar potential of urban buildings, in this case, although the direction E W is the best orientation of the facade, and consequently, its best sunlight and maximum energy efficiency, the width of the 14.20 m street channel seriously affected the mentioned advantages. A narrow urban canyon when combined with the urban forest of first magnitude has a first order incidence in solar potential. The results of Case 7 indicate a 66% reduction of the available solar energy, the maximum reduction reached in this work. Figure 5 Measured irradiance in N facade, case N°7. 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Figure 6 Views of the facade. Figure 7 Trees seen from the sensor. 413 Masking produced by the unleafed branches can be considerable in winter, especially with species of considerable size (1st and 2nd magnitude and narrow street channels). In this case, it is observed (Figure 5) that the permeability that the trees in narrow urban channels offer towards radiation presents great variations according to the position of the sun. Influence of Trees: In order to study how trees condition the access of the sun we have selected a case that presents the influence of two different species of deciduous leaves. When analyzing Figure 9, we see that in this case the reduction of available energy is 55% and that the influence of each species is very noticeable. Figure 8 Environment. Figure 9 Irradiance measurement of the North facade, case N°2. In morning and mid day hours the reduction of the availability of the solar resource is 39% (Figure 10). This value corresponds to the Morus alba (species 1) bare foliage obstruction, while in the afternoon hours the resource reduction is 85% due to the high density of unleafed branches that Fxaxinius excelsior (species 2) presents (Figure 11). These results show that urban trees result in a first order variable when capturing a solar resource in highly forested urban environments, and the choice of the optimal tree species at the moment of urban planning should be considered when contemplating future design strategies. Figure 10 Irradiance measurement of the N facade, case N°2 (Influence of Morus alba, species 1). Figure 11 Irradiance measurement of the N facade, case N°2 (Influence of Fxaxinius excelsior, species 2). Influence of Building Height: In order to analyze the impact of urban building morphology in relation to building height, measurements have been made by vertically displacing the sensor that measures the partially shaded zone of the facade. Figures12 and 13. Figure 12 Irradiance measurement of the N facade, case N°1 (Third Floor, 7 meters high approximately) 30th INTERNATIONAL PLEA CONFERENCE 16-18 December 2014, CEPT University, Ahmedabad Figure 13 Irradiance measurement of the N facade, case N°3 (Second Floor, 4.5 meters high approximately) 414 Figure 14 Images of Case Study 3 98% of the total irradiance was available when the sensor that measures the partially shaded zone was placed on the third floor. Meanwhile on the second floor location, the measured values indicated a 78% availability of the resource. The