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
1m
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
4Sunrize1
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 log1 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 Wm-2K-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 drop