Solar World Congress : Proceedings of the Biennial Congress of the International Solar Energy Society Denver Colorado USA 19-23 August 1991. Volume 3 Part 1


EPA/600/R-
Apri1 1993
SOLAR
WORLD
CONGRESS
Proceedings of the Biennial Congress of
the International Solar Energy Society,
Denver, Colorado, USA, 19-23 August 1991
VOLUME 3. PART I
M.E. ARDEN
SUSAN M.A. BURLEY
MARTHA COLEMAN
PERGAMON PRESS

-------
BIBLIOGRAPHIC INFORMATION
PB93-174803
Report Nos: none
Title: Solar World Congress: Proceedings of the Biennial Congress of the
International Solar Energy Society. Held in Denver, Colorado on August 19-23, 1991.
Volume 3. Part 1.
Date: cl992
Authors: M. E. Arden, S. M. A. Burley, and M. Coleman.
Performing Organization: American Solar Energy Society, Boulder, CO.
Performing Organization Report Nos: EPA/600/R-93/058E
Sponsoring Organization: ''Environmental Protection Agency, Research Triangle Park,
NC. Air and Energy Engineering Research Lab.
Supplementary Notes: See also Volume 2, Part 2, PB93-174795 and Volume 3, Part 2,
PB93-174811. Sponsored by Environmental Protection Agency, Research Triangle Park,
NC. Air and Energy Engineering Research Lab.
NTIS Field/Group Codes: 97N*, 97J, 89B*
Price: PC A99/MF A06
Availability: Available from the National Technical Information Service,
Springfield, VA. 22161
Number of Pages: 672p,c
Keywords: *Solar energy conversion, *Solar houses, ^Buildings, ^Meetings, Solar
architecture, Solar cooling, Solar heating, Residential buildings, Commercial
buildings, Computerized simulation, Thermal insulation, Heat storage, Daylighting,
Convection, Comfort, Atriums.
Abstract: Topics covered in Volume 3, Part 1 include Solar Building Designs; Zero-
Energy Building Designs; Emerging Architecture; Vernacular Architecture; Passive
Commercial Buildings; Daylighting; Atriums; Passive Strategies and Materials;
Transparent Insulation; Convection and Mass; Comfort; Passive Cooling; Passive
Computer Analysis.
i

-------
PB93-174803
EPA/600/R-93/058e
April 1993
1991
SOLAR
WORLD
CONGRESS
(IN FOUR VOLUMES)

-------
Pergamon Titles of Related Interest
Bloss & Pfisterer ADVANCES IN SOLAR ENERGY TECHNOLOGY,
4-volume set
Bowen & Yannas PASSIVE AND LOW ENERGY ECOTECHNIQUES
Granqvist MATERIALS SCIENCE FOR SOLAR ENERGY
CONVERSION SYSTEMS
ISES INTERSOL 85
McVeigh SUN POWER, 2nd Edition
Sayigh ENERGY AND THE ENVIRONMENT: INTO THE 1990S
Stecco & Moran A FUTURE FOR ENERGY
Treble GENERATING ELECTRICITY FROM THE SUN
Yannas PASSIVE AND LOW ENERGY ARCHITECTURE
Pergamon Related Journals
(Free specimen copy glady sent on request)
ENERGY
ENERGY CONVERSION AND MANAGEMENT
GEOTHERMICS
HEAT RECOVERY SYSTEMS AND CHP
INTERNATIONAL JOURNAL OF HEAT AND MASS TRANSFER
INTERNATIONAL JOURNAL OF HYDROGEN ENERGY
PROGRESS IN ENERGY AND COMBUSTION SCIENCE
RENEWABLE ENERGY
SOLAR ENERGY

-------
1991
SOLAR
WORLD
CONGRESS
VOLUME 3. PART I
Proceedings of the Biennial Congress of
the International Solar Energy Society,
Denver, Colorado, USA, 19-23 August 1991
Edited by
M.E. ARDEN
SUSAN M.A. BURLEY
MARTHA COLEMAN
American Solar Energy Society, Inc.,
Boulder, Colorado, USA
PERGAMON PRESS
OXFORD • NEW YORK • BEIJING •
FRANKFURT • SEOUL • SYDNEY • TOKYO

-------
Pergamon Press pic, Headington Hill Hall,
Oxford 0X3 OBW, England
Pergamon Press, Inc., Maxwell House, Fairview Park, Elmsford,
New York 10523, U.S.A.
Maxwell Pergamon China, Beijing Exhibition Centre,
Xizhimenwai Dajie, Beijing 100044, People's Republic of China
Pergamon Press GmbH, Hammerweg 6.
D-6242 Kronberg, Germany
Pergamon Press Korea, KPO Box 315, Seoul 110-603, Korea
Maxwell Macmillan Pergamon Publishing Australia Pty Ltd., Lakes
Business Park, 2 Lord Street, Botany, NSW 2019, Australia
Pergamon Press, 8th Floor, Matsuoka Central Building,
1-7-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo 160, Japan
Copyright © 1991 International Solar Energy Society
All Rights Reserved. No part of this publication may be reproduced,
stored in a retrieval system or transmitted in any form or by any
means: electronic, electrostatic, magnetic tape, mechanical,
photocopying, recording or otherwise, without permission in writing
from the copyright holder.
First Edition 1991
Library of Congress Cataloging-in-Publication Data
ISBN: 0-08-041690-X
NOTICE
Neither the International Solar Energy Society® , the United States
Section of the International Solar Energy Society, nor any of the
cosponsors of this Congress make any warranty, expressed or implied,
to accept any legal liability or any information, apparatus, product, or
process disclosed, or represent that its use would not infringe privately
on rights of others. The contents of the proceedings express the opinion
of the authors and are not necessarily endorsed by the International
Solar Energy Society® , the United States Section of the International
Solar Energy Society, or any of the cosponsors of this Congress.
Printed in the United States of America
The paper used in this publication meets the minimum requirements of
@™ American National Standard for Information Sciences -- Permanence of
Paper for Printed Library Materials, ANSI Z39.48-1984
Copyright © 1992 International Solar Energy Society. Since
this work was, in part, funded by the U.S. Government, the
Government is vested with a royalty-free, non-exclusive, and
irrevocable license to publish, translate, reproduce, and
deliver that information and to authorize others to do so.
A
Pergamon Press Offices:
U.K.
U.S.A.
PEOPLE'S REPUBLIC
OF CHINA
GERMANY
KOREA
AUSTRALIA
JAPAN

-------
2479
INTERNATIONAL SOLAR ENERGY SOCIETY
The International Solar Energy Society is a worldwide nonprofit organization dedicated to the
advancement of the utilization of solar energy. Its interests embrace all aspects of solar energy, including
characteristics, effects and methods of use, and it provides a common meeting ground for all those
concerned with the nature and utilization of this renewable non-polluting resource.
Founded in 1954, the Society has expanded over the years into a truly international organization with
members in more than 90 of the world's countries. It has been accepted by the United Nations as a
nongovernmental organization in consultative status, and it is widely regarded as the premier body of its
type operating in the solar energy field.
The Society is interdisciplinary in nature and numbers among its members most of the world's leading
figures in solar energy research and development, as well as many with an interest in renewable energy and
its practical use. High academic attainments arc not a prerequisite for membership, only a special interest
in this particular field.
Organization
Day-to-day administration is provided by the Society's headquarters office, which since 1970 has been
located in Australia. The headquarters house the Secretary-Treasurer and the Administrative Secretary,
together with members of their supporting staff.
In countries and regions in which sufficient interest exists, Sections of the Society have been established.
These Sections, which are largely autonomous, organize meetings and other local activities and in some
cases produce their own publications. All Society members are eligible to belong to their respective
national or regional Sections, although in some cases this may involve the payment of an additional
Sectional fee. In recent years the number of Sections has increased slowly but steadily.
Activities of the Society are:
1. Publications of Solar Energy, a monthly scientific journal of an archival nature, containing scientific
and technical papers on solar energy and its utilization, reviews, technical notes and other items of
interest to those working in the field of solar energy.
2. Publication of a less technical magazine, SunWorld.
3. Publication of a newsletter for members, ISES News.
4. Organization of major International Congresses on solar energy at which numerous scientific and
technical papers are presented and discussed. These Congresses are held every two years in different
countries, normally in conjunction with equipment exhibitions, and are widely attended.
5. Publication of the Proceedings of each International Congress. Whereas copies of the Society's three
periodicals (items 1-3 above) are supplied to all members as part of their membership, copies of
Congress Proceedings are available (from the publisher) only on special order and at an additional cost
Special pre-publication prices are normally available to Society members.
6. More recently ISES has become increasingly involved with other major Non-Governmental
Organizations in matters relating to the application of renewable energy and other global environmental
problems, and is currently preparing its contribution for presentation at the United National Conference
on Environment and Development (UNCED - or popularly referred to as ECO 92).
Headquarters:
International Solar Energy Society Telephone: 61 3 571 7557
PO Box 124 Fax: 61 3 563 5173
Caulfied East, Vic. 3145 Telex: AA 154 087 CITVIC
AUSTRALIA

-------
2480
AMERICAN SOLAR ENERGY SOCIETY
The American Solar Energy Society (ASES) is the United States Section of ISES and presently
has over 4,000 members.
ASES seeks to promote the widespread near-term and long-term use of solar energy. To achieve
that goal, ASES:
•	Fosters the use of science and technology in the application of solar energy;
•	Encourages basic and applied research and development in solar energy;
•	Promotes education in fields related to solar energy; and
•	Provides information relating to all aspects of solar energy.
Activities:
•	ASES conducts the National Solar Energy Conference as a annual forum for exchange of
information about advances in solar energy technologies, programs, and concepts. The
conference features speakers who are national leaders in their technical and professional fields.
Workshops, exhibits and tours of solar applications highlight this annual event, which is attended
by more than 450 solar energy enthusiasts from throughout the country.
•	ASES publishes Solar Today, a bi-monthly magazine. Each issue highlights practical
applications of solar energy, presents the latest results of solar energy research, covers
developments in the nation's solar energy industry, and includes member discussion of solar-
related issues.
•	Each year, ASES sponsors a Roundtable in Washington, DC, bringing together energy decision-
makers in a highly visible public forum. Each Roundtable addresses an issue of critical
importance to ASES members and the nation.
•	To ensure worldwide dissemination of information about solar energy developments, ASES
annually publishes Advances in Solar Energy. This compendium of the latest R&D
developments is authored by ASES members who are nationally recognized experts on their
respective topics.
•	Technical, regulatory and educational issues are addressed in periodic White Papers, which
present critical analyses of important solar energy topics.
•	ASES educates the public and energy decision-makers on the benefits of solar energy through a
public relations campaign and information materials.
•	ASES has 16 state and regional chapters, which are independently incorporated organizations
providing services to their members appropriate to the local areas. Typical activities include
newsletters, technical meetings, public outreach activities, and government relations.
Headquarters
2400 Central Avenue, Suite B-l
Boulder, CO 80301
Telephone : 303-443-3130
Fax	: 303-443-3212

-------
2481
Table of Contents
Volume 1: Solar Electricity, Biofuels, Renewable Resources
1.1	Photovoltaic Thin-Film Materials and Devices	1
1.2	Photovoltaic High Efficiency and Applications	39
1.3	Photovoltaic Modeling and Batteries	73
1.4	Large-Scale Photovoltaic Applications 1	117
1.5	Large-Scale Photovoltaic Applications II	155
1.6	Small-Scale Photovoltaic Applications 1	193
1.7	Small-Scale Photovoltaic Applications II	231
1.8	Photovoltaic Applications	269
1.9	Photovoltaic Utility Issues	307
1.10	Photovoltaic Issues	343
1.11	Posters: Photovoltaics	393
1.12	Posters: Photovoltaic Systems	465
1.13	Solar Thermal Electric	531
1.14	Wind Energy Experiences	563
1.15	Wind Energy Systems Performance	599
1.16	Wind Systems Applications and Hydropower	639
1.17	Utility and Regulatory Issues	677
1.18	Solar Hydrogen Technologies	715
1.19	Biotechnology	757
1.20	Bio-Chemical Conversion	797
1.21	Biofuels	835
1.22	Radiation Instruments, Measurements	891
1.23	Radiation Models	931
1.24	Radiation Models, Simulation	969
1.25	Radiation Resources	999
1.26	Use of Radiation Data	1037
1.27	Renewable Resource Posters	1075
Volume 2: Active Solar and Solar Heat
2.1	Collectors I: Selective Surfaces	1139
2.2	Collectors II	1183
2.3	Collectors in	1219
2.4	Collectors IV	1257
2.5	Solar Domestic Hot Water 1	1297
2.6	Solar Domestic Hot Water II	1337
2.7	Solar Domestic Hot Water III	1373
2.8	Passive Domestic Hot Water	1409
2.9	Solar Water Heaters	1449
2.10	Active Heating I: Seasonal Storage	1485
2.11	Active Heating II: Heating System Performance	1523
2.12	Active Heating III	1563
2.13	Active Cooling I: Advances in Open Cycle Absorption	1599
2.14	Active Cooling II	1635
2.15	Active Solar	!	1673

-------
2482
2.16	Posters: Active Solar 1	1709
2.17	Posters: Active II	1781
2.18	Concentrating Collectors	1853
2.19	High Flux	1891
2.20	Solar Heat Storage	1929
2.21	Central Receivers	1967
2.22	Dish Collectors	2005
2.23	Line-Focus Collectors	2043
2.24	Detoxification and Materials	2083
2.25	Solar Heat	2121
2.26	Solar Heat Posters	2159
2.27	Desalination	2251
2.28	Solar Ponds	2289
2.29	Desalination and Solar Ponds Posters	2327
2.30	Solar Drying 1	2401
2.31	Solar Drying II	2439
Volume 3: Passive Solar, Socio-Economic, Education
3.1	Solar Building Designs	2497
3.2	Zero-Energy Building Designs	2529
3.3	Emerging Architecture	2565
3.4	Vernacular Architecture I	,	2599
3.5	Vernacular Architecture II	2635
3.6	Passive Commerical Buildings	2665
3.7	DaylightingI	2701
3.8	Daylighting II	2739
3.9	Atriums	2779
3.10	Passive Strategies and Materials 1	2817
3.11	Passive Strategies and Materials II	2851
3.12	Passive Strategies and Materials III	2887
3.13	Transparent Insulation I	2925
3.14	Transparent Insulation II	2957
3.15	Convection and Mass	2989
3.16	Comfort	3027
3.17	Passive Cooling 1	3065
3.18	Passive Cooling n	3101
3.19	Passive Computer Analysis 1	3137
3.20	Passive Computer Analysis II	3173
3.21	Passive Computer Analysis III	3209
3.22	Monitored Passive Modules	3247
3.23	Extended Passive Monitoring	3285
3.24	Passive Non-Computer Design Tools	3325
3.25	Sustainability 1	3363
3.26	Sustainability II	3399
3.27	Passive Posters 1	3437
3.28	Environmental Effects	....3521
3.29	National Solar Programs			3559
3.30	Developing Country Applications I	3597

-------
2483
3.31	Developing Country Applications II	3633
3.32	Technology Transfer	3667
3.33	Socio-Economic Posters	3705
3.34	Education	3757
3.35	Education Posters	3803
Volume 4: Plenaries, State-of-the-Art, Farrington Daniels Lecture

-------
Intentionally Blank Page

-------
2485
Contents of Volume 3
3.1 Solar Building Designs
O. Aschehoug, A. G. Hestnes, A.G. Lien, A. Nordgaard, H. Raaen and M. Thyholt 2499
An Advanced Solar Low Energy Dwelling for Norway
O. M0rck 2504
Hybrid Solar Low Energy Dwellings
P. V. Pederson 2510
Total Energy Design in Two Danish Building Projects with Combined Use of Solar Heating
and Energy Conservation
J. Palmer, R. Watkins, A. Seager, M. Trollope, P. O'Sullivan, N. Vaughan, D. Alexander,
H. Jenkins, P. Jones 2516
The U.K. Passive Solar Energy Performance Assessment Project
L. L. Boyer 2522
Solar Performance Analysis for an Underground Dwelling at High Altitude
3.2 Zero-Energy Building Designs
P. D. Lund 2531
Possibilities For Zero Energy In Solar Houses
A. Goetzberger and W. Stahl 2537
The Self-Sufficient Solar House Freiburg
A. Heinzel and K. Ledjeff 2543
The Self-Sufficient Solar House: Hybrid Energy Storage System
G. Bopp 2547
The Self-Sufficient Solar House Electrical Concept
T. R. McBride 2553
Energy Self-Sufficient Single Family Home
F. Sick, W. GrieBhaber 2559
The Self-Sufficient Solar House: Remarkable Simulation Results
3.3 Emerging Architecture
J. D. Balcomb 2567
Advanced House Concept
F. Moore 2573
SOLARGREEN... A Critique After Ten Years of Living
P. Cooper and K. Haggard 2579
Horizontal and Vertical, Thermal and Aesthetic Responses to Two Contrasting Microclimates
L. Yde 2585
Plus Energy Housing in Denmark
S. Carpenter and J.P. Kokko, E. White 2590
First Year Performance of the Advanced House
L. M. Holder III, USA 2596
Passive and Active Solar Wind Generator Repair Facility—With No Backup System in Canyon,
Texas

-------
2486
3.4	Vernacular Architecture I
D. Radovic	2601
Bioclimatic Aspects of the Reurbanization of the Sava Amphitheatre Area in Belgrade
S.U. Egarievwe, I.C. Ezema, J.E. Mbamalu, G.T. Dorgu	2605
Observed Thermal Behaviour of Modified Traditional Buildings In the Nigerian Hot-Humid
Climatic Zone
D.A.	Bainbridge and M. Myhrman	2611
Straw Bale Building Systems
R.M. Kumar, J. Cook, N.K. Bansal and G. Minke	2617
Bioclimatic Analysis of Indigenous Houses in Different Climates of India
R.M. Kumar, G. Minke, N.K. Bansal, J. Cook	2623
Construction With Low Energy Materials
A.N. Tombazis	2629
Solar Village
3.5	Vernacular Architecture II
E.	Durand	2637
Solar Architecture in France
R. Rom£n L., R. CorvaMn P., and R. Tala D	2644
Passive Heating In The High Andes
A. Stahr	2650
Renewable Energy Concepts in the Design of Antique Roman Urban Houses
W. Chongjie	2656
The Design and Technical and Economic Analysis of the Passive Solar Energy Classrooms in
the Rural Areas of Shandong China
D. Vuksanovic	2659
Traditional Bioclimatic Buildings in Montenegro
3.6	Passive Commerical Buildings
F.	Lichtblau, F. Lichtblau and W. Lichtblau	2665
Storey and Solar Glass Roof Addition to an Architectural Office
J. Bleem, A. Kirkpatrick, C.B. Winn, N. Khan and D. Mori	2670
Energy Efficient Building Design and Operation at the New Denver Airport
A. P. Waterfield and B. Norton	2676
Design and Performance Monitoring of Green Park Combined School
Ph. Andre, J. Nicolas, J. Fr. Rivez, V. Debbaut	2682
Analysis, Monitoring and Evaluation of a Passive Solar Commercial Building Including Mass
Walls and Direct Gain Features: The "Auditoires Ful" Building in ARLON (Belgium)
M. Donn and N. Isaacs	2688
Solar Non-Domestic Buildings in New Zealand
C. Zydeveld	2694
A Municipality's Experience with Good Insulation and the Application of Passive Solar
Principles for All Newly Constructed Housing Projects in the Last Ten Years
3.7	Daylighting I
V. Cartwright and J. S. Reynolds	2703
Daylight, Energy Conservation and Comfort in an Office Building
M. S. Baker and J. S. Reynolds	2709
Monitoring and Modeling a Climate-and Energy-Conscious Office Building

-------
2487
D. Brentrup 2715
School House Light: A Case Study Assessment of Lighting
M. Shukuya and K. Ohashi 2721
Indoor Illumination by a Lightshelf Combined with Horizontal Blinds
I. Cowling, S. Coyne and R. Dew 2727
A Design for a Permanently Mounted Daylighting Device
R. M. Kumar, J. Battachaijee and J. Cook 2733
Design Curves For Daylighting In Tropics
3.8 Daylighting n
M. Boubekri and L. L. Boyer 2741
Optimization of Window Design for Thermal, Lighting and Occupant Appraisal Considerations
P. Waide 2747
Estimating Hourly and Daily Glazing Effective Transmittance for European Locations
C. Cooksey, J. Loveland, M. Millet, A. Vanags 2753
"Limits of the Sky" Testing and Evaluation of the Current State-of-the-Art in Mirror-Box Sky
Simulation
R. Rhyner, C. Roecker and J. L. Scartezzini 2761
An Automated Heliodon for Daylighting Building Design
W. Glennie and V. Krishnamurthi 2767
A New Type of Control System for Daylighting
A. Fanchiotti, G. Gagliardi, A. Peigari, P. Polato and M. Vio 2773
Variable Incidence Spectral Transmittance Measurements for Determining Glazings' Solar
Parameters for Daylighting Purposes
3.9 Atriums
A. Liu, M. Nawab and J. Jones 2781
Geometric Shape Index for Daylight Distribution Variations in Atrium Spaces
E. O.Kainlauri, G. J. Lehman, M. P. Vilmain 2787
Comparative Studies of Five Atriums on the Effects of Orientation, Exposure and Design on
Daylighting, Temperature, and Stratification of Air
H.T. Gordon, R. Kammerud and A.G. Hestnes 2793
International Atrium Research
P. Kristensen and T. Esbensen 2799
Passive Solar Energy and Natural Daylight in Office Buildings
M.R. Atif and L. Boyer 2805
Effective Top-Glazing and Internal Wall Area for Efficient Daylighting in Atria
J. Jones, M. Luther and M. Nawab 2811
A Comparative Analysis for Two Geometrically Different Atria
3.10 Passive Strategies and Components I
R. L. Crowther 2819
Sunspaces for Function, Security, Outgassing, and Air Tempering
G. Robertson 2825
Thick Friendly Walls - Energy Efficient Commercial Building Design
L. Baojun, W. Jingyu and X. Xiaogeng, Y. Shilong and J. Xiangshan 2830
The Heat Capacity Analysis of Rapid Heat Collecting Wall

-------
2488
R. Mendenhall 2834.
Radiant Barriers in Residential Dwellings: Energy, Comfort, and Moisture Considerations in a
Northern Climate
J. Yoo, S. Cho and H. S. Chung 2839
Thermal Performances of Latent Heat Storage Material for Floor Heating System
C. H. Filleux 2845
Air Window Collectors
3.11 Passive Strategies and Materials II
A. Bashir and B. M. Gibbs 2853
Simultaneous Heat Storage and Utilization in a Raining Particulate Bed Heat Exchanger
K. Yoshimura, M. Tazawa and S. Tanemura 2859
Electronic Structure of Iridium Oxide Electrochromic Thin Films
E. Shaviv and I. G. Capeluto 2864
The Influence of the Geometrical Design Parameters on the Passive Cooling and Heating of
Buildings
L. Sandoval, J. Pineda, R. Costaiieda and L. Sanchez 2870
Cooling Ceiling-Pond In Hot-Humid Climate
S. Medved and P. Novak 2876
Reflective Venetian Blind-A Multipurpose Element For Passive Solar Heating
F. Scamoni, I. Meroni, C. Pollastro, P. Tirloni 2881
Development and Analysis of Both Passive and Hybrid Solar Components
3.12 Passive Strategies and Materials III
A. A. Valdez and C. G. Currin 2889
Solar Processing of Silicone Glazing
Y. Shan-qing, Z. Xiao-ping, Z. Guo-ping and M. Min-wei 2895
Transparent Heat Insulating Coatings ofAg-Sn02 on a Glass
L. Bao jun, X. Xiaogeng, W. Jingyu, Y. Shilong, and J. Xiangshan 2901
A Design of Solar Energy Heating with Optical Fibers
M. Lee, E. K. Rhee, B. G. Chun and J. Hwang 2905
Development of Night Insulation Devices In Passive Solar System
C. Choudhury, H.P. Garg, and J. Prakash 2911
Comparative Analysis of Packed Flow Passage Solar Air Heaters
A. C. de Cerutti, C. de Rosa, J. L. Cortegoso, A. Ravetto 2917
Solar Permeability of Urban Trees in the Dry Temperature Climates of Western Argentina
3.13 Transparent Insulation I
W. Stahl and W.-S. Wilke 2927
The Space Heating Concept of the Self -Sufficient Solar House Freiburg
H. A.L. van Dijk 2933
Translucent Insulation For Passive Solar Energy Application
J.W. Twidell 2939
The World's Largest Demonstration of Transparent Insulation for Buildings
H. X. Yang, B. J. Brinkworth, R. H. Marshall 2945
The Potential Utilization of TIM in Passive Solar Buildings in China
K. Jahn, D. Christoffers 2951
Valuation of Transparent Insulating Devices

-------
2489
3.14 Transparent Insulation II
P.O. Braun, J. Schmid, E. Bollin, W. Stahl, J. Vahldiek, K. Voss, A. Wagner 2959
Transparent Insulation Material Demonstration Projects and Future Prospects
E. Bollin, Q.S. Yuan, L.Z. Guang 2965
Transparently Insulated Walls for Buildings of Low Latitude - A Pilot Project in Shanghai,
China
M. Zupan and J. Bostjancic 2971
Solar Gains by Transparent Thermal Insulation - Comparative Measurements on Test Cells
over one Heating Season
G. Brouwer 2977
Performance Criteria For Transparent Insulation Materials in Buildings
N. D.Kaushika 2983
Honeycomb Passive Water Heating Systems
3.15 Convection and Mass
L. T. James and W. A. Gross 2991
Heat Transfer by Convection and Thermal Radiation in a Small Scale Trombe Wall Simulation
Enclosure
G.S. Barozzi, M.S. Imbabi, E. Nobile, and A.C.M. Sousa 2997
An Experimental And Numerical Study Of Passive Solar Ventilation In Buildings
T.L. Thompson, G.V. Mignon, N.V. Chalfoun 3003
The Black Globe Thermometer for Indoor/Outdoor Mean Radiant Temperature Measurement
F.D. Heidt and R. Rabenstein 3009
Measuring Airborne Heat Flows in Passive Solar Buildings with Tracer Gas Methods
M. Lee, E. K. Rhee and G. Song 3015
Development of Design Strategies For Thermal Mass In Passive Solar Direct Gain System
N. Kazic and P. Novak 3021
Air Movement Response of the Building on the Various Disturbances
3.16 Comfort
N.V. Chalfoun, T. L. Thompson and M. R. Yoklic 3029
"MRT™ " Update: A Study on the Thermal Effect of Dry and Wet Paving and Landscape
Materials on Restoring Human Thermal Comfort Conditions at Outdoor Spaces
A. Cordier, M Galeou, F. Monchoux and F. Thellier 3035
Predicting Local Thermal Sensation in a Building
J. H. Heerwagen, J. Loveland and R. Diamond 3041
Coping with Discomforts
V. Calderaro and A. Ciolfi 3048
Architectural Design Methodology for Environmental Comfort
B. D. Howard 3053
New Initiatives + Solar Design = Healthy Building Design
R. A. Hobday 3059
The Heliotherapists
3.17 Passive Cooling I
B. Givoni, UCLA, USA 3067
Modelling a Passive Evaporative Cooling Tower
J.Cook 3072
Errors And Assets In Outdoor Mist Cooling

-------
2490
H.	Wu	3077
The Impact ofMoisure Up-take From Evaporative Coolers on the Cooling Loads of
Residential Buildings
C.	An?ay, O. Guisan, B. Lachal, R. Meldem, W. Weber	3083
Passive Cooling of an Administrative Building in Sion (Switzerland)
D.	K. Irurah and J. S. Reynolds	3089
Capacitive and Resistive Insulation in Architecture for Nairobi - Kenya
M.S. Sodha, R.L. Sawhney, J. Kaur, S.P. Singh	3095
Thermal Performance of a Room Coupled to an Evaporative Cooling Tower
3.18	Passive Cooling II
I.	Melody, L. Maxwell, R. Vieira	3103
The Florida Cracker Style: A Passive Cooling Compromise
M. B. Gadi and I. C. Ward			<;	3109
An Architect's Approach to Predicting Thermal Performance of Housing in Warm Regions
with Reference to North Africa
D. Feuermann and W. Hawthorne	3115
On the Potential and Effectiveness of Passive Night Ventilation Cooling
L.M. Holder EI	3121
A Passive Solar House in a Hot Humid Climate Designed to be Cooler both in Winter and
Summer
G. S. Yakubu and S. Sharpies	3125
Modulated Solar Shielding of Buildings: A solar radiation control stategy for low energy
buildings in hot dry and semi-arid climates
J. L. Motloch and K. D. Song	3131
Validation of Video Image Capture/Interpretation Method to Approximate Solar Radiation
Reduction Due to Tree Shading in the Summer
3.19	Passive Computer Analysis I
J. E. Christensen and K. Johnsen	3139
Third Generation of a Thermal Simulation Program: tsbi3
D. Christoffers and K. Jahn			3145
Numerical Simulation and Analysis of the Thermal Performance of Buildings
T. Trijssenaar	3151
PASCAUD, Passive Solar in Computer Aided Urban Design
L. Jankovic	3157
Evaluation of Learning Models for Building Energy Simulation
L. Fulop, L. F. Jesch and S. Gilani	3163
Transparent Insulation Module For TRNSYS
F. Sick, J. P. Kummer....	3167
An Extension of the TRNSYS Multizone Component for Transparent Insulation Applications
3.20	Passive Computer Analysis II
A. Mavroulakis, J. M. Baleynaud, R. Javelas and A. Trombe	3173
Dynamic Processing of Direct Solar Gains in TRNSYS Building Simulation Environment
J-H Huh, H. Wu and J-J. Kim	3179
Control Strategies for Switchable Glazing in Office Buildings

-------
2491
E.	K. Rhee , K. Kim and K. H. Cho, Chung-Ang	3185
Development of Thermal Analysis Computer Program for Selecting Wall Components in
Energy Conscious Design
C. A. Caroutas and A. T. Kirkpatrick	3191
Natural Convection Heat Transfer in Rectangular Partitioned Enclosures
M. Schiler, E. James and M. Milne	3197
Interactive Graphic Input for Superlite
C. Carter	3203
Building Simulation With Simplified Solar Radiation Data
3.21	Passive Computer Analysis HI
S. Reilly, J. Gottsche, and V. Wittwer	3211
Advanced Window Systems and Building Energy Performance
B.	Haglund, B. Sumption and A. Zabrodsky	3217
Visualization of Daylighting in Buildings
F.	Parrini, S. Vitale, L. Castellano	3223
Modelling Turbulence For Thermal Analysis Of Stratified Storage Tank
N.V. Chalfoun, M.R. Yoklic, K.J. Kent	3228
Passive Solar and Energy Optimization for a Residential House Type in Tucson, Arizona: A
Case Study for the Solar Village Project
J. E. Christensen and J. Christoffersen	3235
Daylight Analysis Using Superlite, Superlink and tsbi3
Ph. Andre, J. Nicolas, P. Andre, J. F. Rivez, V. Debbaut	3241
Application of the Optimal Control Theory for the Definition of Energy/Comfort Efficient
Strategies in a High Thermal Inertia Passive Solar Commercial Building
3.22	Monitored Passive Modules
P. Wouters, L. Vandaele and T. C. Steemers	3249
The PASSYS Project: Passive Solar Building Research in Europe
S.J. Harrison and F. Dubrous	3255
Measurement of Solar Heat Gain Coeffecient Using a Solar Simulator Test Facility
J.W. Twidell and P.H. Baker	3261
The UK Contribution to the European Community PASSYS Programme
E. Hahne and P. Voit	3267
Passive Solar Testing With Outdoor Testcells
H. A. L.van Dijk	3273
The PASSYS Method For Testing Passive Solar Components
M. S. Jeon, H. K. Yoon, W. G. Chun, H. S. Jeon, P. C. Auh	3279
Thermal Performance Evaluation of the KIER Test Cell through Short Term Measurements
3.23	Extended Passive Monitoring
R. S. Dumont	3287
Measured Energy Consumption of a Group of 99 Low Energy Houses
H. S. Joen, W. G. Chun, M. S. Jeon, P.C. Auh, K. Subbarao, E. Hancock, J. D. Balcomb	3293
Application of the PSTAR Method to a Thermally Massive Passive Solar House
M. Glaumann	3299
The Influence of Weather on Heat Flow Through Glazing
C.	Spiropoulou and E. Andreadaki	3306
A Passive Solar House in the Center of Athens — Evaluation of the Thermal Performance

-------
2492
S. N. G. Lo and B. Norton	3312
The Effect of Occupancy Patterns on the Long-Term Performance of Roof-Space Solar Energy
Collectors
M. N. Ferguson and J. L. Scartezzini	3318
Full Scale Experiment of a Predictive Controller in a Passive Solar Building
3.24	Passive Non-Computer Design Tools
S.C. Carpenter and J.A. Baker	3327
Determination of Total Window Solar Heat Gain Coefficient
C.	Armenta-Deu, M.C. de Andres and J. Doria	3333
Analysis of a Simple Model to Predict Thermal Behavior in a Bed-rock for Daily Space
Heating
D.	Kilfbyle	3338
Design Methodology of a Gnomon
J. J. Duffy	3344
A Design Tool for Determining Collector and Storage Sizes for Passive and Active Space-
Heating Systems
N. Egrican and S. Uygur	3350
Flow Prediction and Entropy Generation in Passive Heating Systems
D. J. Harris and B. Norton	3356
Autoregressive Moving Average Analysis of the Long-Term Performance Data From Passive
Solar Schools
3.25	Sustainability I
R. S. Levine and E. J. Yanarella	3365
The Once and Future Sustainable City
R. B. Swenson	3370
The Marketing of Sustainability
P. Fisk III	3376
Integration vs. Conservation: A Renewable Energy Building Block for the 21st Century
S. Loken	3382
Materials for a Sustainable Building Industry
P. Schuller	3387
A Measure of Sustainability for Energy Technology and Public Policy
R. Schoen	3393
Standalone Resorts as Vehicle for Sustainable Tourism ... and Architecture...
3.26	Sustainability II
R.J. Koester	3401
Modeling And Thereby Defining Sustainability: From Thermal Networks To Geographic
Information Systems
M. M. Valdez	3407
Maintaining Traditional Sustainable Life-Styles in Rural United States
K. Haggard, P. Fisk, P. Cooper and K. Noland	3413
Passive Solar Architecture to Fractal Architecture: A Natural Progression Toward
Sustainability
D. L. Kezell and B. D. Wood	3419
Use of the Analytic Hierarchy Process in the Development of a Regional Energy...
D. Seiter, W. L. Doxsey, G. Vittori and P. Fisk	3425
Austin Sustainable Systems Rating Program

-------
2493
R. S. Levine, E. Yanarella, T. Radmard and D. Harper	3431
The Development of an Interactive Computer Aided Design Model for Generating the
Sustainable City
3.27	Passive Posters I
L. Guang-Liang and S. Zhan-Gong	3439
Solar Town Upon "The Roof of the World"
Y. Ishikawa	3444
Thermal Effect of Evaporative Cooling by Spraying Water on Membrane-Structured Roof
C. D. Pdrez Segarra, A. Oliva, and M. Costa	3450
Numerical Modelization of Turbulent Natural Convection in Enclosures. Applications in
Active and Passive Solar Systems
J.L. Bonin, C. Butto, J.Y. Grandpeix, J.L.Joly, V. Platel and M. Rigal	3456
Development of Capillary Pump Loop For Earth Based Solar Heating Applications
K. I. Jensen	3462
Transparent Cover Based on Evacuated Monolithic Silica Aerogel
A. Oliva, M. Costa and C.D. Perez Segarra	3467
Numerical Simulation of the Thermal Behavior ofTrombe Wall
M. A. Herrero and J. Pazos	3473
A Design Tool for Sunshine Availability and Shading Effects Evaluation
A. C. Pitts and R. S. Abro	3479
Enhanced Ventilation for Improved Comfort in the Climate of Pakistan
M. A. Sattler	3485
Shading by Trees: The Influence of the Clearness Index on the Tranmission Factor
0.C.	Morck, I. Bryn, N. Morel, G. Silvestrini, A. Santamaria	3491
IEA TASK XI: Integrated Knowledge Based Solar Design Tool (ISOLDE)
N. F. Youssef, G.B. Hanna and M.F. Abadir	3497
Humidity Properties of Rigid Foam Materials Used in Buildings
1.	B. Sommereux and B. Peuportier	3503
A Bioclimatic Design Aid Based Upon Multizone Simulation
I. R. Edmonds	3509
Optical Properties of Laser Cut Light Deflecting Panels and Applications to Daylighting
M. Dragovic and J. Grabovac	3515
Passive Solar Energy Multistorey Building for Three Climatic Conditions of Yugoslavia
3.28	Environmental Effects
P. Tarana, R. Stewart and R. Perez	3523
The Global Energy Legacy
J.N. Swisher	3529
Limiting Global Climate Change: Potential Policy Connections Between Developed and
Developing Nations
J. N. Swisher	3535
A Case Study of Utility Costs For Reducing C02 Emissions
M Saif-ul-Rehman	3541
Why Solar Energy Has Made Little Contribution to the Energy Systems of the Developing
Countries
F. Kreith, P. Norton and T. Lang	3547
The Potential of Solar, Renewable, and Energy Conservation Systems to Reduce Global
Warming

-------
2494
C. Ravin and N. Lenssen 3553
Toward a Sustainable Energy Future
3.29 National Solar Programs
H. L.Walton 3561
Energy Projections and their Renewable Energy Component
A. Kristiansen 3568
Environment and Energy in the Nordic Countries
H. P. Garg 3574
Solar Energy Technology, Development and Applications in India
D. R. Neill, R. Koehler, N. Huang, C. Yu, M. Stackmann 3580
HNEI China and Pacific Area Cooperative Programs
W. H.Park 3585
New and Renewable Sources of Energy (NRSE) Research and Development in Korea
E. F. Jaguaribe and Pio C. Lobo 3591
An Assessment of Brazilian Alternative Energy Policies
3.30 Developing Country Applications I
O.St.C. Headley, I. A. McDoom and A. T. Kai 3599
Solar Timber Driers in the Eastern Caribbean
B. Norton and K. J. Perera 3605
Appropriate Design of a Solar Energy Water Heater for Sri Lanka
D. Dalton, R. Robin and R. Williamson 3611
The Hopi Foundation's Solar Electric Enterprise: A Model for Renewable Industry
Development in Developing Nations
A. Bishay 3617
A Non-Conventional Prototype Farm Based on Renewable Energy Principles and Applications
A.S. Sheinstein and E. E. Shpilrain 3623
Prospects For Solar Energy Utilization In Antarctica
M. McClelland 3629
Portable Solar Cookers for Returning Afgan Refugees
3.31 Developing Country Applications II
N. Suharta and H. Notzold 3635
Renewable Energies Indonesia (REI)
J.T. Pytlinski 3639
Renewable Energy Sources for the Caribbean Islands
S. S. Nandwani 3645
My Eleven Year Experience With Simple Solar, Solar-Electric and Heat Storage Cookers
S. W. Ali and S.M. Hasnain 3651
A Novel Design for Solar Cooker Incorporating Phase Change Heat Storage Material
J. Essandoh-Yeddu and C. Y. Wereko-Brobby 3656
Status of Solar Photovoltaic Programmes in Ghana
D. Holm 3662
The Quantitative and Qualitative Environmental Impact of Solar Passive Housing in the Semi-
Arid Regions of Southern Africa

-------
2495
3.32	Technology Transfer
P. Chaturvedi	3669
Renewable Energy Technology Transfer to Asia & Pacific
R. Vories and P. Notari	3675
Identifying the Information Needs of Renewable Energy Professionals: An Evaluation of
SERl's Technology Transfer Publications
R. Alward, J. Ayoub, T.A. Lawand and E.Brunet	3681
Implementing Renewable Energy Programmes in Developing Countries
S. Arafa	3687
Integrated Rural Energy Systems and Community Development: Lessons From The Field
H. Suharta, I. N. Suharta, H. Tobing, L. M. Panggabean and G. Wessels	3693
Technology Transfer of Solar Thermal Pump to Indonesia Industries
J. S. Foster	3699
Acceptance of Solar Water Heaters By New Householders in Queensland, Australia
3.33	Socio-Economic Posters
G.K.	Magney	3707
Keys to Successful Solar Cooking
N. Ying-jing	3713
Promoting Solar Housing in China
M.A.M. Shaltout and R. Botros	3718
Photovoltaic Activities in Egypt: Applications and Research
H.	Futai and L. Mingru	3727
A Solar Energy Assisted Vibro-Fluidized Bed Dryer
G. Richard, T. King, P. A. Borgo	3733
Egyptian Renewable Energy Field Testing Project
W. J. Shadis, M. C. Davies, A. Hegazy	3739
Preliminary Assessment!Planning Guide for Energy Efficient and Environmentally Sound
Tourist Villages in Remote Areas of the Sea Coast in Egypt
S. Arafa	3745
Solar Powered Video Training System For Village Production In Africa
C.	Schwarzer, W. Bieger, N. Ricking and N.K. Bansal	3751
A Solar Cooker with a Rock Bed Cum Oil Thermal Storage
3.34	Education I
R. J. King	3759
GM Sunrayce USA: The Race for the Future
J. M.Gordon and S. Weintraub	3765
ZORAN: An Educational Software Package For Photovoltaic Devices
M. Millet and J. Loveland	3771
Public Daylighting Education—Seattle's Lighting Design Lab
N. M. Lechner	3776
New And Traditional Methods For Teaching Sun Mechanics (Sun Angles)
G.M.Singh, S.S. Bhatti	3782
Energy Sciences: An M.Sc. Programme for Needs of Developing Countries
E. Naumann	3787
Case Study - A Method to Teach Specialists in Renewable Energies
D.	E. LaHart	3793
Implementing a Statewide Energy Education Program

-------
2496
3.35 Education Posters
L. Di-Si	3805
Educational Significance of Launching Activities in Using Solar Energy in Middle School
J. Klima	3811
An Application of Computer Assisted Instruction to a Solar Energy Technology Course
L. Broman, J. A. Duffie and E. Lindberg	3815
A Concentrated Course in Solar Thermal Process Engineering
D.	L.Konkle	3821
Solar Energy Demonstration Exhibit, Leslie Science Center
S. A. M. Burek and J.C. McVeigh	3824
Energy Systems And Environmental Management A New Master's Course
E.	T. Borer, Jr. and H. Lorsch	3830
A Simple Spreadsheet Program For Classroom Use
J. A.Turdgano and M.C. Velasco	3836
Active Learning of the Criteria for Building Design According to Microclimate Conditions
J. Pinter	3842
Solar Energy in Hungary
S. M. Chagwedera	3846
Young Scientists' Exhibition: A Co-Curricular Approach to Renewable Energy Education in
Zimbabwe
L. Broman and K. Gustafsson	3849
An Educational Travelling Exhibition on Solar Energy

-------
3.1 Solar Building Designs

-------
Intentionally Blank Page

-------
2499
AN ADVANCED SOLAR LOW ENERGY DWELLING FOR NORWAY
O. Aschehoug*, A. G. Hestnes*, A. G. Lien*,
A. Nordgaard**, H. Raaen***, and M. Thyholt***
*Dept. of Architecture, The Norwegian Institute of Technology,
N-7034 Trondheim, Norway
**SINTEF Applied Thermodynamics/HVAC section,
N-7034 Trondheim, Norway
***SINTEF Architecture and Building Technology,
N-7034 Trondheim, Norway
ABSTRACT
This paper describes a building concept developed in Norway as part
of the work within Task XIII of the International Energy Agency's
Solar Heating and Cooling Programme. The Task, entitled "Advanced
Solar Low Energy Buildings", aims at developing residential building
concepts that use minimal amounts of purchased energy.
The Norwegian concept, a single family, two story rowhouse, reduces
the need for purchased energy for heating by 90% and the total need
by 85%, compared to conventional dwellings built to the building
code. This is accomplished through the use of a combination of
energy conservation measures such as super insulation, advanced
glazings, and movable transparent insulation as well as through the
use of active measures such as an integrated mechanical system using
a heat pump and heat recovery, and photovoltaics. The dwelling is
to be built in 1991/92. It will be extensively monitored.
KEYWORDS
Energy conservation; solar energy; residential buildings.
INTRODUCTION
IEA Task XIII, "Advanced Solar Low Energy Buildings", was started
in 1989 and will be completed in 1994. In this Task the 15 particip-
ating countries develop building concepts that minimize or
completely eliminate energy use for heating and also greatly
reduce energy use for cooling, ventilation, hot water production,
and lighting. The emphasis is on new technologies that still are in
the research phase. Cost effectiveness is therefore not expected
before the turn of the century.
The results of the work will be presented as guidelines for use by
the building community. In addition, a number of the building
concepts developed in the various countries will be constructed and
monitored within the time frame of the Task. These buildings will

-------
2500
be experiments in new solar technologies rather than demonstrations
of existing technologies.
BACKGROUND
A high latitude country like Norway faces severe limitations in the
use of solar energy. Most efforts have therefore so far been concen-
trated on the research and development of more conventional energy
conservation technologies. The main conditions influencing the
introduction of solar alternatives are:
-	the high latitude locations, with maximum insolation on vertical
or steeply sloping surfaces during the heating season,
-	the lack of sufficient insolation during parts of the heating
season, and
-	the high standard of thermal insulation, with strict requirements
for thermal comfort.
From most of this it would seem that solar heating would be a rather
unlikely option. However, as the heating season is long and extends
into periods when solar energy can give a substantial contribution,
it is possible to use solar heating for longer periods in Norway
than in countries further south.
CONCEPT DEVELOPMENT
A conventional two story rowhouse unit of approximately 120m2 floor
area was used as the basis for the development of a solar building
concept. A parameter study of various existing technologies applied
to this preceded the development of new concepts. Its purpose was
to see what space heating savings could be achieved by using
conventional techniques. The results were then used to define a low
energy dwelling to be used as the starting point in the development
of solar concepts.
kWh/month
2500
— Conventional
house
	Low enercy
house
2000
: Solcr Coin
1500
H.L. : Heat:nc Locd
\<-S.G,
1000
500
JUL AUG SEP OCT NOV DEC JAN FEB MA APR MAY JU.NI
Fig.l. Seasonal variations in heating load and solar gains for
the conventional dwelling and for the low energy dwelling.

-------
2501
This low energy dwelling has an energy consumption of 17 kWh/m2 for
space heating and a total consumption of 100 kWh/m2 in a climate of
about 4500 heating degree days, base 20°C, over an eight month
heating season. Reducing the consumption further required the
introduction of active measures on the water heating load and
household electricity consumption. Energy targets for such a
dwelling were set at 20 kWh/m2 annually for space and water heating
and the same amount for lighting, household equipment, etc.
The parameter studies of the low energy alternatives were carried
out using the program SUNCODE. The studies of the solar alternatives
were carried out with the programs PV-Fchart, Fchart, TRNSYS and
TSBI-3.
THE SOLAR DWELLING
The solar concept developed is based on the same dwelling unit with
a similar layout. It contains all the features concluded in the
parameter study of the low energy dwelling as being most effective.
Transmission losses are therefore minimized by using a super
insulated opaque envelope with U-values in the 0.10-0.15 W/m2K range
and windows with a U-value of 0.7 W/m2K.
As discussed earlier, there is no appreciable solar gain in the
months of November, December, and January. For this cold, dark
winter period the only viable option is increased insulation and
heat recovery systems. Because the solar apertures have less
insulation than the opaque parts of the envelope, it is necessary
to apply additional night or seasonal insulation in this period.
The windows are therefore equipped with automatic, movable shutters
that bring the overall U-value down to 0.25 W/m2K. The shutters are
fitted with a high performance transparent insulation material. They
are controlled by a system that responds to the overall energy
balance of the windows. When removed, the shutters provide addition-
al insulation and a solar wall effect for the opaque part of the
envelope.
The south wall is 60% transparent, providing daylight and passive
gains to an integrated sunspace. and to the main space of the
dwelling. The south part of the roof and the opaque part of the
southfacing wall are covered with a 30m2 photovoltaic array
providing electricity.
The sunspace is used to preheat the fresh air supply during solar
gain periods, when the shutters are open and the sunspace is not
heated. The sunspace is kept at minimum +15°C for comfort reasons,
the heating load for this only being 130 kWh annually.
The conservation measures and the solar and internal gains bring the
space heating load down to 15 kWh/m2. The dwelling could therefore
conceivably be fully selfsufficient with thermal energy for space
heating. This would, however, require a seasonal storage system
covering the three coldest months. Such a system is technically
possible with conventional technology, such as water or PCM storage.
But on the scale of a one family dwelling seasonal storage is very
space and cost demanding, and therefore decided not to be feasible
at this stage.

-------
2502
SECOND FLOOR
FIRST FLOOR
LOFT
TECHNICAL
Fig.3. Plans and section of the solar dwelling.
The dwelling is equipped with an integrated mechanical ventilation/-
heat recovery/heat pump/domestic hot water/thermal storage system,
that reduces all energy demands to electricity only. The photo-
voltaic system will cover about 40% of the need for electricity,
resulting in a total annual demand for purchased energy of approxim-
ately 3600 kWh, or 30 kWh/m2. This amount is small enough that,
given a predicted breakthrough in battery technology, it could be
stored, making the dwelling totally selfsufficient. However, this
is not included at the present time.

-------
2503
PV-ayatem
Exhaust air
Electricity load
Water heater
Heat exchanger

	/$A



h
Vy

buppiy |
at©r




Evapor
a
%

~*]—®—
Building

X


Condenser
Fig.4. System diagram.
TABLE 1 Energy Balance of Reference and Solar House. kWh
Transmission losses
Infiltration/ventilation
Reference house
15840
11530
Solar house
5084
4992
27370
Total heat loss
Auxiliary heating	12080
DHW load	4000
Heat pump/mechanical system
Household electricity	5870
PV contribution
10036
0
0
3630
2400
- 2450
Purchased energy
21950
3580
The dwelling, which is still on the drawing board, will be built at
the beginning of next year. Negotiations with potential builders are
now taking place. After completion it will be monitored for at least
a year, thereby hopefully verifying the calculated results.
REFERENCES
Aschehoug, 0. and colleagues (1990). Parametric Studies for a
Norwegian IEA Task XIII House. SINTEF-report no. STF62 A90018,
Trondheim, Norway.
Hestnes,A.G. (1990). Advanced Solar Low Energy Buildings.
Norwegian and International Work within the IEA Task XIII.
Proc. of CIB W67 Workshop Low Energy Buildings (2nd Generation).
Heidenheim, Germany.

-------
2504
HYBRID SOLAR LOW ENERGY DWELLINGS
Ove Marck, M.Sc., Ph.D.
CENERGIA Energy Consultants
Stationsvej 3
DK-2760 Malev
ABSTRACT
The use of solar energy at high latitudes to substantially reduce the heat
load of dwellings implies "integration of advanced design	as pointed
out by Hestnes, (1990). The paper describes a concept for hybrid solar, low
energy residential buildings currently being developed in Denmark. The con-
cept has been proposed to the CEC THERMIE programme as a demonstration pro-
ject of 50 dwellings to be built first in Denmark and a year later in
France.
The proposed concept integrates passive, active and hybrid solar utilisation
and low energy techniques. An innovative air heating system provides auxi-
liary heating from cascade controlled natural gas furnaces shared by groups
of houses. The concept has been designed with a large degree of climate
adaptability.
KEYWORDS
Hybrid solar, storage wall, active solar, air heating, heat recovery,
passive solar, district heating, central control system.
BACKGROUND
Highly insulated walls and roofs, super-glazings and heat-recovery on ven-
tilation air are the low-energy techniques that in the future will reduce
the heat-losses in common dwellings so much that passive solar energy utili-
sing the Direct Gain principle will be very limited. The solar heat gain
will quickly surpass the losses and result in a considerable ventilation
load to control the indoor temperatures.
Therefore effective storage and distribution are essential for solar heating
systems in low-energy buildings of the future.
Utilising an active energy transport between either solar collector and
storage or storage and distribution hybrid systems offer significant advan-
tages over simple passive solar systems:
-	Collected solar energy can be moved to where it is most needed, i.e.,the
northfacing rooms.
-	The release of heat is delayed to better match the demand.
-	Collected energy can be more efficiently stored, i.e.,temperature swings
are reduced.
-	Energy collection and distribution can be automatically controlled offe
ing improved comfort to the inhabitants.
-	Natural cooling through night ventilation is possible.

-------
2505
CENERGIA has been conducting R&D&A work in this area for the past 8 years. A
recent study made in connection with the evaluation of a demonstration pro-
ject concerning 55 direct gain low-energy solar dwellings very clearly de-
monstrates the need for effective storage and distribution of the gained
solar heat. Fig. 1. overleaf is presented in a report (in Danish) about this
project (Morck, 1990).
The figure clearly shows how the amount of excess heat in March is increased
for improved glazing and larger glazing areas. The net heating demand for a
low energy house is only slightly reduced with larger southfacing areas and
even increased when the acceptable comfort temperature is lowered to 23°C.
Only when part of the southfacing glazing area is replaced by an air collec-
tor for a hybrid solar system a considerable reduction in heating demand is
achieved.
GJ
3
2
1
O AS S2S S23 A25 A23 LM IIY
Copenhagen, test reference
year, March.
Excess heat
Heating load
Fig. 1. Heating load and excess heat for different design changes of solar
low-energy building in the Smakkebo building project.
Southfacing glazing Max acceptable

Area
U-value
indoor temperature

%
W/sqm.
degree C
0: Original design
100
2.0
25
AS: As build
100
2.0
25
S25: Changed, case 1
100
0.5
25
A25: Changed, case 2
150
0.5
25
A23: Changed, case 3
150
0.5
23
LM: As A23, but light
150
0.5
23
HY: As A23, but the 50 % of the southfacing glazing area replaced by an air
solar collector, the collected energy of which goes to a brickwall within
the house. (The heating load in March corresponds generally to 12% of the
annual heating load).
Experiences from the above mentioned project and several other projects are
that the occupants often do not accept high indoor temperatures.
The conclusion is that for low energy dwellings direct gain passive solar
can only provide a strictly limited contribution to the heating load even
when super-glazing is used.
The main conclusion of a another recent Danish investigation is that the
most economical heating system for low-energy buildings is a system designed
as a small district heating system for groups of 10-15 houses.

-------
2506
Based on the above considerations and conclusions a concept for a flexible,
climate adaptable solar, low-energy house design with a completely new hea-
ting system has been developed.
DESCRIPTION OF CONCEPT
The developed concept integrates the following principles:
-	Combined passive, hybrid and active solar heating
-	High insulation levels, low U-value glazing
-	Heat recovery on ventilation air
-	Air heating
-	Air channels for hybrid heating, air heating and ventilation integrated
in construction for economy and comfort
-	Low temperature district heating with cascade controlled gas furnaces
for groups of houses.
Domestic solar hot water system.
The 50 houses are placed in rows with 8-12 houses in each row. Each row of
houses is equipped with a boiler room which holds the common installations:
cascade controlled furnaces, storage tank, CTS system.
The concept can best be described by focusing first on the individual dwell-
ings and thereafter on the shared systems in the boiler rooms. Fig. 2. il-
lustrates the coupling of systems connecting the dwelling and boiler room
installations.
Dwelling
Boiler room
Air-to-air heat-«xchanger
Liquid
solar
collectors
Water to air heating element
Air heating
DHW use
Heat-meter
Distribution
NaturaJ
gas
furnace
Sclar system
pipings
Storage
Fig. 2. Connections between boilerrocm and dwelling installations.
DWELLINGS
Hybrid Solar:
The hybrid system consists of an air solar collector, air channels in the
floor slabs and a brick wall storage with passive discharge to the north-
facing rooms. A principle sketch, fig. 3., is attached. The slabs on this
diagram are also designed to hold all other air flow channels.

-------
2507
South
facade
Leca-concrete
air ducts
slab with
A i r
solar
col lector
North
facade
Brickwa 11
as
heat storage
Leca-concrete slab with ducts for	DUCt for air
hybrid solar system and air heating heating system
system
Fig. 3. Principle sketch of hybrid solar system.
Passive Solar:
The southfacing windows provide direct solar gains to the thermally heavy
south oriented rooms of the dwellings. Brick partition walls and leca con-
crete slabs constitute the majority of the thermal mass. The windows have
low-emissivity glazing with a U-value of 1.5 W/sqm/K. The glazed area corre-
sponds to 15 % of the floor area in each dwelling, 70 % of this is placed on
the southfacing surfaces.
Insulation:
The houses are insulated to a level well above the requirements in the
Danish building regulations. The U-value will be:
Walls: 0.17 W/sqm/K.
Roof.: 0.12 W/sqm/K.
Floor: 0.17 W/sqm/K.
Active Solar:
The active solar system consists of roof-integrated solar collectors, 4 sqm.
pr dwelling. The collectors are coupled to the storage tank in the boiler
room. As the size of the collectors indicate the active solar system is
mainly designed to provide domestic hot water heating during summertime.
However, as the collected energy is stored in the common water storage in
the boiler room, which also serves as a buffer for the gas furnaces, useful
solar collected in the heating season can be used for	domestic hot
water as Veil- as for heating.
Airtiahtness/Ventilation:
Special attention is given to weather-stripping, resulting in natural air
change rates less than 0.1 per hour. A mechanical ventilation system provide
the necessary ventilation. The system includes an air-to-air heat-exchanger
which recovers up to 80 % of the heat in the ventilation air.
Air Heating:
The houses will be heated by air, primarily distributed via ducts in the
slabs. An water-to-air heating element placed in the heat-exchanger heats
the air to the required temperature to heat the house. The system includes a
recirculation possibility in the situations where an increased air flow is
needed to carry the required heating energy. An advanced control system is
included to ensure good comfort conditions at lowest possible, heating energy
consumption.

-------
2508
Water Heating:
In each dwelling a modern, efficient water-to-water heat-exchanger is
placed.
District Heating Network:
Both the water heating and air heating system in each dwelling is supplied
from the district heating network. These pipings and the pipings from the
solar collectors to the boiler room are placed inside the buildings to uti-
lise losses in the heating season. A heat-meter measure the total consumed
energy for heating and domestic hot water in each dwelling. By experience
this stimulates user interest in saving energy.
SHARED INSTALLATIONS IN THE BOILER ROOM
Heating System:
The individual heat demands from each dwelling are very low. Individual na-
tural gas furnaces would be very inefficient in covering these loads. There-
fore a solution has been chosen where the heating and hot water loads for
groups of app. 10 houses are covered by two natural gas furnaces controlled
in cascade. This means that only when the load is higher than the nominal
effect of one of the furnaces, both furnaces will be active. An automatic
control system controls the furnaces and switch the load between them to
assure an even wear. To obtain maximum efficiency and easier and better con-
trol the furnaces deliver heat to a buffer storage tank. This storage tank
is also serving as a storage for the active solar heating system.
CTS System:
An advanced Central State and Control system monitor the active solar sy-
stems, the furnaces, the storage, the distribution network and the ambient
weather conditions.
The system is needed for optimum control of the furnaces, giving first prio-
rity for available solar gains, and to control the cascade operation of the
furnaces. At the same time it provides the possibility to deliver data to a
remote overall monitoring system.
The overall principle for solar heating and solar DHW is illustrated on fig.
4. As it appear from the figure, space heating is to be provided by a combin-
ation of direct solar gains, hybrid solar gains, and heating from the dis-
trict heating network. The network is distributing heat from the water stor-
age, which again is fed from a combination of active solar collection and
the natural gas furnaces. The domestic hot water is heated from the district
heating network.
Production Storage	Distribution Use
Air solar
collectors

Brick wall
storage
Radiant &
convective
heating

Direct solar
gains
1




Space heating
Liquid solar
collectors
\

Low temperature
district heating
of space and DHW
Natural gas cas-
cade furnaces
Water storage
tank

DWH use
Fig. 4. Overall system operation diagram.

-------
2509
BUILT IN CLIMATE ADAPTABILITY
The concept as such is considered suitable for all climates resulting in a
net winter heating load for dwellings. A number of parameters can be varied:
-	The sizes of the solar systems, both air and solar collector areas, and
storage volumes, as well as window areas and types.
-	The insulation levels
In warmer countries the hybrid system can be reversed
in summer time to cool the mass storage during nighttime.
The optimum mix of principles and parameters must be found by an economic
optimisation in each case.
To prove this adaptability.it is the intention to build a project in France,
which can be conducted in parallel to the present project. The design and
construction of two parallel projects in two European countries with differ-
ent climates could be of a considerable importance to gain experiences with
the optimisation of the concept.
TEST BUILDING
The concept includes specifically two new techniques which has never been
tried out in practiGe before; the hybrid solar system and the air heating
system, both making use of air ducts integrated in the slabs. Both these
systems have to be very carefully designed to avoid problems with leakages,
and to measure pressure drops over the ducts for fan optimisation. In both
cases also connections and manifolds can be formed in many different ways
and practical testing of some of these is a necessity. The project therefore
includes the design and construction of a test building to obtain practical
experiences with selected solutions before building a series of 50 houses.
The project plans	to build the test building and
obtain experience with the systems during the winter 1991/1992. These expe-
riences can then be used for the design of the actual project. It should be
mentioned that Danish Leca has already conducted the first experiments with
the insertion of air ducts in the light concrete slabs which they produce.
MONITORING
After the completion of construction,the project will be monitored over a
period of a little less than two years. The monitoring will comprise a se-
ries of short term detailed measurements of the individual elements which
will be combined with an overall monitoring of the project by the heat-
meters and the CTS-systems in each boiler room. The monitored results will
be compared to the design calculations and the performance will be evalua-
ted.
REFERENCES
1.	Hestnes A.G. (1990). Passive and Low Energy Architecture Design and
Research Issues at High Latitudes. 1st World Renewable Energy Congress,
Reading, U.K. September 23-28, 1990. proceedings: Pergamon Press.
2.	Morck O.C. (1990). Laveneraibebygaelse med solvarme. ISBN 87-503-8332-9,
Byggeriets Udviklingsrdd.

-------
2510
TOTAL ENERGY DESIGN IN TWO ORNISH BUILDING PROJECTS WITH COMBINED
USE OF SOLAR HEATING AND ENERGY CONSERVATION
Peder Vejsig Pedersen
Ove Morck
CENERGIA Energy Consultants
Stationsvej 3
DK 2760 MA10V, Denmark
ABSTRACT
An 80% saving .compared to normal gas consumption for heating and DHW is ex-
pected in the community, Tubberupvaenge with 92 houses near Copenhagen in
Denmark. Here a total energy design,with combined use of energy savings and
solar heating,has been introduced to obtain the most ambitious energy saving
building project in Denmark to date. Savings on heating	is
first achieved by common energy saving techniques like extra insulation,
improved windows, heat recovery of ventilation air, and attached sunspaces,
which,as something new,has integrated pebble bed storage in the floor. Be-
sidesisach of eight building blocks utilises 45 m2 of roof integrated solar
collectors as part of local solar heating system design, primarily for DHW.
And for the first time in Denmark a solar heated seasonal storage is used,
here with 1025 m2 of high temperature solar collectors, and a 3000 m3 water
filled seasonal storage insulated and buried in the ground. From 1991
another building project, Egebjerggard III,with 100 houses,will be built in
Ballerup near Copenhagen, again using the total energy design concept to
reach a 60% saving of the gas consumption and a 30% saving of the electrici-
ty demand and of the water consumption. A co-generation system with combined
production of electricity and heat, is combined with a normal gas furnace in
the heating plant at Egebjerggard III.
KEYWORDS
Total energy design, solar heating, energy conservation, solar collectors,
seasonal storage, co-generation, electricity savings.
80% SAVING OF HEATING DEMAND FOR 92 HOUSES IN HERLEV, DENMARK
TUBBERUPV£NGE II is a governmental supported housing project, located in the
municipality of Herlev, comprising 8 blocks with a total of 92 housing
units.. Financing of the energy system applied in the project, was supported
by the EEC and the Danish government, as a demonstration project. The Buil-
ders, KAB building society and Herlev Building Society, supported the
idea of combining savings and use of solar heating to reach a total saving
of 80% compared to normal energy consumption for heating. The TUBBERUPV£NGE
II project was finished during 1990, see fig. 1.
The TUBBERUPVAENGE II project is located 8 km north-west of the centre of
Copenhagen. The low density houses in Tubberupvaenge are made of prefabricated
cassettes from a factory in Jutland.
It is the most ambitious energy saving building project in Denmark :to

-------
2511
date.Savings of heating demand is first achieved by common energy saving
techniques and passive solar design. Besides, each of eight building blocks
utilises 45 m2 of roofintegrated solar collectors as part of a local solar
heating system design, primarily for DHW. And for the first time in Denmark
a solar heated seasonal storage is used, here with 1050 m2 of high tempera-
ture solar collectors, and a 3000 m3 water filled seasonal storage insulated
and buried in the ground.
DESCRIPTION OF CENTRAL SOLAR HEATING PLANT WITH SEASONAL STORAGE
During the heating season, a district heating system for the 92 houses in
Tubberupvaenge will distribute heat from a small district heating plant in
combination with a central solar heating plant with seasonal storage
(CSHPSS). A 3000 m3 insulated seasonal storage in the ground, is combined
with Danish produced high temperature 12 m2 solar collector modules
(Scancon) with a complete area of 1025 m2 placed on stands.
The seasonal storage is heated to 85 C by the end of August, and from Sep-
tember heat is delivered from the seasonal storage to the district heating
network. Calculations show that it should be possible to cover all the ne-
cessary heating demand until December directly from the seasonal storage.
When the top temperature in the storage is less than 50 C, the operation
strategy is changed so heat will be delivered through a small size nighttime
and weekend operated electrical heat pump, (30 kW). This will cool the sea-
sonal storage to a minimum of 10 C in the spring. Supplemental heat after
December is supplied from a small gas engine generator system and a gas fur-
nace system, see fig. 2.
An intelligent control system,from the firm Danfoss,ensures an efficient
operation and survey of the solar heating plant.
It was originally expected to use a pitstorage in the ground with sloping
walls of clay as a cheap seasonal storage design. But because an uncertainty
arose about what would happen when the clay dried out at high temperatures,
it was necessary to choose a more expensive seasonal storage design with
vertical walls.
With support from the EEC and the Energy Council in Denmark monitoring and
evaluation will take place until the end of 1991.
All important energy flows and temperatures in the CSHPSS and in one of the
building blocks is monitored with a datalogger system.
ENERGY CONSERVATION AND LOCAL SOLAR HEATING SYSTEM
The Total Energy Project in Tubberupvaenge includes the use of low rise,
high-density houses with a 30% lower consumption of natural gas for heating
than the normal size of 180 kWi per m2 housing area per year in Denmark.
Solar energy from the CSHPSS covers 70% of the reduced yearly heat demand in
Tubberupvaenge, which means that the yearly gas consumption will only be 35
kwh per. m2 housing area.
10% of the building area is taken up by large sunspaces with integrated peb-
ble bed heat storages in the floor.
The pebble bed heat storage in the floor of the sunspaces is charged by use
of a ventilator. It has been experienced that when the pebble bed heat sto-
rage in the floor is used, it is possible to achieve a 6 C higher sunspace
temperature in the evening than in the normal situation without a pebble bed
storage. At the same time it is seen that passive discharge have the same
effect as active discharge but over a longer period.

-------
2512
As heating system for the houses is used low temperature radiators (60/40
C), floor heating and a ventilation systems with heat recovery.
A common ventilation system with heat recovery for 6 houses is used. This
reduces the electrical effect for the ventilation to only 200 W for all 6
houses. Besides a timer also reduces the electricity use by reducing the
operation hours.
Local solar heating plants are placed in connection to each of 8 housing
blocks.4 roof-integrated solar collector of 40-45 m2 per building block
heats a 2500 litre heat storage jacket tank, placed in a small cellar. The
tank is, besides from solar, heated from the central heating system in the
winter, as well as being backed-up by an electric heating rod in the summer.
Heat from the storage tank in the cellar is transferred to local 150 litre
DHW tanks with heat-exchangers in each apartment, and to the low temperature
heating system.
A built-in electrical heating element in the storage tank makes it possible
to stop operation of the district heating network for the eight housing
blocks during 4-5 summer month, leading to large amounts of saved heat los-
ses.
A NEW TOTAL ENERGY PROJECT WITH 100 HOUSES IN BALLERUP
It has now been decided to realise another building project, Egebjerggard
III with 100 houses in Ballerup near Copenhagen again using the total energy
design concept. The same team as in Tubberupvaenge is involved in this pro-
ject. The builder is the KAB Building Society, contractor is Jespersen &
Son, engineer is Dominia A/S and responsible for the low energy and solar
heating design is Cenergia Energy Consultants.
It is here aimed to reduce the gas consumption for heating and DHW not to
exceed 65 kWh per m2 per year equal to a 60% saving compared to normal. The
electricity demand will be reduced by 30% to 30 kWh per m2 per year, and the
water demand will be 70% the normal size (140 litre per person per day).
These savings are combined with the use of a local cogeneration plant, with
a new storage design and a low temperature district heating design with pul-
se operation. It is aimed to use 7-8 m2 of roofintegrated solar collectors
for DHW and heating in this project. The extra costs for the mentioned total
energy design will be approx. 10% compared to a normal building project,
leading to a positive economy for the inhabitants. Also this projectis partly
financed by the EEC and the Energy Council in Denmark. In Egebjerggdrd
III an improved low energy design for the buildings will be used compared to
the standard in Tubberupvaenge II. This is especially true for the insulation
standard, the windows and the DHW system which uses DHW heat exchangers in
each apartment. In fig. 4. a diagram is showing the calculated savings in
Egebj ergg&rd III.
Heat from the small co-generation plant will be transferred to six locally
distributed heat storage buffer tanks, which also function as storage for
the local solar heating systems. The local storage tanks make it possible
to obtain a pulse operation, so heat is only distributed to the local stora-
ge tanks when all heatthere is used up. This means that it will be possible
to avoid nighttime operation of the low temperature district heating
network, and in sunny periods onlypulses,or none at all,are necessary.
Fig. 5. shows losses for the proposed district heating design compa-
red with heat losses for a normal district heating network. It is obvious
that large savings are possible by use of this technology.

-------
2513

¦ _ - "	- .	_ i V" ^ ¦ ,Ji,
Fig. 1. Photo of the building project Tubberupvaenge in Herlev where a low
energy design and 400 m2 of local solar collectors are used together
with a central solar heating plant with 1025 m2 of high temperature
solar collectors and a 3000 m3 seasonal storage placed under the
wide part of the solar collector field. This total energy system
is expected to save 80% of the "normal" gas consumption for heating.
Central solar collector field
^Local solar
heating
systems
Heating pla it
Boiler room
pump
Buiwr-
sumec
j77ie Total Energy Concept reduces pollution
hy 80%, compared with traditional heating
systems - with no extra costs for the residents'.
Fig. 2. Diagram of the central solar heating plant with seasonal storage in
Herlev, Denmark. Solar heat is supplied from the solar collectors to
the seasonal storage depending on the delivered temperature. The so-
lar collectors are manufactured by the Danish firm Scancon. Heat
from the seasonal storage can be used directly, but when the storage
temperature is less than 45 C it is used in combination with a small
heat pump system. A small co-generation system functions together
with a small gas furnace as back-up system.

-------
Fig. 3. Plan of building site for the EgebjerggSrd III in Ballerup. The
local roofintegrated solar collectors are indicated in the drawing.
Total energy design for 100 houses at
Egebjerg near Copenhagen
- 60% saving of gas for heating and domestic hot water - 30% saving of electricity
- LOSSES IN DISTRICT HEATING PLANT
/LOSSES FROM CIRC. OF DHW
/DHWDEMAND
.DISTRIBUTION NETWORK LOSSES
oUSER EFFECT
.HEATING DEMANO

120
9 100-
AWWWWWWWWWV
mrnmmm

Ordinary housing
area
Low energy design	Low energy design
without local solar	including solar
heating system	heating
Fig. 4. Normal energy consumption compared to the calculated energy consump-
tion for heating in EgebjerggSrd, Ballerup. It is the aim to reduce
the gas-consumption for heating to 40% of normal, and at the same
time reduce the electricity use to 30% the normal size.

-------
2515
Heatlosses for different district heating,
operation strategies.
Em
= 2
1: ordinary district heating ( 60/40°C ).
2: one pipe system.
3: one pipe system and pulse operation.
183.269 kWh/year
91.993 kWh/year
47.982 kWh/year
Fig. 5. In the Egebjerggard III project which will be built in Ballerup in
1991, a new district heating design will be used. Six local heat
storage buffer tanks, which also function as storage for local so-
lar heating systems, is supplied with heat from a one pipe district
heating network with pulse operation. In this way operation of the
district heating network is avoided during night and in sunny pe-
riods it will often not be used at all. In the diagram is showed
calculated heat losses for a normal district heating network with
two pipes (1), and this is compared to a one pipe system (2) and
a one pipe system with pulse operation (3).

-------
2516
THE U.K. PASSIVE SOLAR ENERGY PERFORMANCE ASSESSMENT PROJECT
J Palmer. R. Watkins, A. Seager, M. Trollope +
Prof. P. O'Sullivan**, N. Vaughan, D. Alexander, H. Jenkins, P. Jones *
+ DATABUILD, 4 Venture Way, Aston Science Park, Birmingham,
England
* The School of Architecture, University of Wales
College of Cardiff, P.O. Box 25, Cardiff, Wales
** Bartlett School of Architecture and Planning,
UCL, 22 Gordon St., London, England
ABSTRACT
The Energy Performance Assessment Project (EPA), funded by the UK Department of
Energy and managed by the Energy Technology Support Unit, is a project aimed at
accelerating the uptake of passive solar design in buildings. Its intention is
to provide scientific case studies of passive solar buildings for the
architectural professions, providing both feedback and feedforward. To this end
a methodology has been devised for assessing and reporting on the performance of
passive solar buildings. The methodology, which drew heavily on work in the U.S.,
defined the EPA method for the monitoring and computer simulation of the energy,
cost, and amenity issues of occupied passive solar buildings.
Initial findings from the first buildings studied show good solar performance for
both heating and daylighting. Results from early houses show that up to 35%
solar displaced space heating is possible with no loss of amenity or significant
increase in building cost. Strategies for the guaranteed use of solar gains, for
both heating and daylighting, have been shown to be difficult to implement
because of inadequate system controls.
By the end of 1991 the EPA project will have evaluated a total of 32 houses and
non-domestic buildings and the findings will be widely disseminated to building
designers. It is hoped that this will assist in the development and uptake of
good passive solar buildings.
KEYWORDS
Passive solar energy; daylighting;
ENERGY PERFORMANCE ASSESSMENTS
The UK Department • of Energy's Renewable Energy Research,Development and
Demonstration programme is managed by the Energy Technology Support Unit at the
Harwell Laboratory. The utilization of passive solar energy (heat and light) in
buildings is a component of this programme and an important part of this is field
trials of real buildings.
Early UK field trials were influenced by their research context (Birmingham
School of Architecture, 1985). They tended to be directed to the research issues
of building performance and the research community, rather than the building

-------
2517
Industry. It was from such considerations that the Energy Performance Assessment
Project and Energy Performance Assessments (EPA) were developed (Littler, 1983).
EPA's were intended to be lower cost evaluations of a wide range of buildings
with an emphasis on analysis, interpretation and presentation of data to the
industry. Such trials are required to provide both feedback and feedforward to
those involved in building design.
Some of the essential characteristics of EPAs soon emerged from an assessment of
traditional field trials methods and the substantial US passive solar programme
(Swisher,1982; Gordon, 1986; Subbarao, 1984; IEA Task VIII, 1985). For example,
EPAs:
*should be credible to the building industry,
*should deal with typical (but 'good') buildings,
*should be 'analysis-rich' and not simply 'data-rich',
*must not be solely concerned with energy issues but must also address economic
and human factors.
The requirements of the above characteristics had many ramifications, e.g.:-
*Reliance would need to be placed on proven, available techniques.
*Measurement goals and methods would need clear definition.
*Standardization would be necessary to allow development costs to be spread over
many buildings.
These rules guided the development of an EPA methodology which is intended to
provide an holistic assessment of the energy, amenity and cost of a building.
The EPA methodology is regarded as a 'Template of Components'. There are six
components dealing with the measurement, modelling and referencing of energy,
cost and amenity. Although the template is quite rigid in the way it defines the
EPA method, it also allows for flexibility in the response to individual
'components'. The techniques within each component draw upon accepted current
research methods and can be selected as appropriate for each case study. They
can also be updated as improvements are made in research techniques.
EARLY EPA CASE STUDIES
South Staffordshire Water Company
The building is a four storey, 3800m2 office building. Each floor has a
continuous band of fenestration using low emissivity glass to reduce solar gains
and heat losses, and with openable windows to give natural ventilation. The
fenestration incorporates light shelves which, along with the overhangs, provide
summertime shading to perimeter areas. An automatic lighting control system uses
time of day and exterior light levels to judge when to make available, or to turn
off, lights in the perimeter offices.
Together with the EPA, this building was also the subject of a US/UK
collaborative exercise using the Daylight Performance Evaluation Methodology
(DPEM) of Lawrence Berkeley Laboratory (Andersson, 1987) to predict savings in
electric lighting use.
Using DPEM the second floor south facing office was modelled for electric
lighting usage and a figure of 9.3 kWh/m2 was predicted, showing an annual saving
of 50% compared to a non-daylit office. However, further analysis showed that
electric lighting usage could be reduced by modifications to the lighting control
strategy, which maintained an unnecessarily high base of electric lighting

-------
2518
throughout the working day. The occupants are pleased with the daylight
features, but there is some dissatisfaction with the automatic lighting controls
(Heap, 1988; ETSU, 1991a)
JEL Headquarters. Stockport.
This is a two storey rectangular building containing both office and electronic
assembly areas. The south facade is 100% glazed with offices on either side of
a central double height atrium which is, in fact, the plant room. Glazing on
other facades is restricted. South facing rooflights are used to provide
daylight in the production space. Shading to prevent summertime overheating is
provided by external roller blinds and internal louvre blinds. A comprehensive
BEMS controls room temperatures, shading and heat redistribution.
Results from monitoring show that gas used for space heating was 66 kWh/m2 and
there was an equal contribution from solar gains. However, overheating was a
problem on days when the external temperature was high. Temperatures as high as
31°C have been recorded in the production area (ETSU, 1991b).
Looe School
Looe Junior and Infants School is a single storey building providing 1400 m2 of
gross floor area for 300 pupils. The cruciform plan allows all the teaching
rooms to be largely south facing, with entrances on the north facade. Each
classroom has 100% glazing on the south facing windows which incorporate a mini
'Trombe' wall. A thermally massive bench is provided internally to restrict the
temperature swings caused by the solar gain. Clerestories and light wells took
light to the rear of the rooms.
The results show a gas consumption of 184,000 kWh for heating which compares very
favourably with other low energy UK primary schools and is better than the
current standards. The temperatures recorded throughout the monitoring were in
range of 18 - 23°C for most of the occupied period, with little indication of
severe over or under heating.
Detailed monitoring of the mini Trombe wall showed that under conditions of
medium solar radiation it could be inhibited from working by cold down draughts
from the windows above. In low solar radiation conditions reverse flow of the
air could inject cold air at floor level (ETSU, 1991c).
Solar Cottage
This passive solar house was one of the first EPAs to be undertaken, and it
served as a test bed for the methodology. It is a detached three bedroom house
of 135 m2 floor area. The passive solar and energy efficiency measures included
a south facing conservatory and large south facing glazing.
Analysis of the data for the year of 1986 indicates that the house consumed
20,000 kWh (148 kWh/m2) of fuel for space heating to provide a whole house average
internal temperature of approximately 22.5 °C.
As in the U,S. Class B method (Swisher, 1982) the 'subtractive' method of
determining solar gains was attempted in this house. However, the method did not
seem to work on this very thermally heavy-weight house and therefore the
regression of daily heat gains and solar radiation was used to establish the

-------
2519
solar performance. This technique then became the standard EPA method for
houses.
The daily data allow for the estimation of the contribution of solar gain and
this shows a solar heat gain of 2600 kWh for the year, or 20% of the boiler fuel.
The data were normalized to average conditions for the location giving
approximately a 35% displacement of space heating fuel. The cost of the house
was only 2% greater than an equivalent non-solar house and provided the extra
space and amenity of the conservatory.
Oak Farm Road
This is a two-storey three-bedroom semi-detached house with a 77m2 gross floor
area. This building was costed as being 2% less expensive than an equivalent non-
solar dwelling. It is double-glazed throughout with large south glazing windows
and minimal glazing on the other sides. In order to avoid excessive heat gain
during the summer and to retain stored heat at night time, energy-saver blinds
were incorporated.
The results showed a fairly low annual fuel use (12350 kWh) which was as a result
of the low occupancy pattern.
An annual displacement of 816 kWh of space heating energy by solar gains was
determined (17%). When normalized for 20 year average weather conditions, this
value became 21%. An increase in this figure could be anticipated with a lower
incidence of drawn curtains and blinds in the lounge (S facing), higher level of
occupancy, and a longer period of heating. This assumption was verified by SERI
RES modelling of the house.
SOME LESSONS LEARNED FROM EARLY EPAs
Non-domestic Buildings
Generalization about non domestic buildings is difficult owing to the vast range
in size, function, and complexity, however, passive solar strategies are
generally limited to displacing space heating and electric lighting in the
perimeter zone of a building.
Solar displaced space heating.
Since these buildings are usually well insulated and have high internal gains
then the potential for solar gains is reduced, and the risk of summer overheating
is increased. Consequently, where experience with domestic design has been
transferred directly to non-domestic buildings, then difficulties have been
observed. With reasonable standards of insulation (opaque U values of 0.4 W/m2/K,
or less, window U values of 3.0 W/m2/K, or less) then balance temperatures of
close to winter ambient can easily be achieved. The lesson is simply that
designers must take full account of the differing thermal inputs.
Daylight displaced electric lighting.
There is great potential for using daylight in non-domestic buildings. Not only
does it save energy but it can also significantly improve the quality of the
environment. Care has to be taken to overcome potential problems of glare,

-------
2520
control and shading. In one case an office design, with internal and external
lightshelves to improve the uniformity of daylight, the environment was well
liked by users, but several instances of glare were noted, particularly where
VDUs were in use.
Similarly, although electric lighting use was low, it could have been further
reduced as the controls often provided electric lighting when it was not needed.
In another study, still in progress, top lighting is used in sports halls. In
one of these the adequate daylighting does not displace electric lighting because
an occupant sensor switch turns lights on at the beginning of use, irrespective
of available daylight!
Houses
For domestic buildings the dominant design strategy is to reduce the building's
heat loss and to make maximum use of the thermal component of solar radiation to
displace space heating. The simplest and most widely used strategy is to
redistribute glazing to give large south and small north windows. EPAs and other
trials have observed that this can lead to north rooms with insufficient daylight
where electric lighting is almost always required and where the occupants feel
the room to be gloomy and unpleasant.
Much concern is expressed about overheating, as a result of large south glazed
areas, but UK houses rarely suffer excessive temperatures. Generally speaking
it appears to be straightforward to protect the house itself from overheating by
the use of mass and shading (deciduous trees have been successfully used for
seasonal shading (ETSUd, 1991)).
Large windows facing public spaces may lead to the use of blinds and net curtains
for privacy, which can dramatically reduce the solar gains. One EPA house used
large south facing windows for direct gain and these worked reasonably well
facing on to the rear garden. Designs must always reflect the real needs of
occupants rather than solely the energy needs of the house.
Controls
The most common practical obstacles to successful Solar Displaced Space Heating
and Daylight Displaced Electric Lighting and to user satisfaction have been
related to control.
Often these controls are manual and reflect a design intention "to provide users
with control over their own environment". In practice these are often not used
as intended with a consequent reduction in energy savings and comfort. EPA
findings, supported by earlier US. research, indicate that occupants are most
likely to operate controls if the controls make sense to them and provide the
required outcome (Kantrowitz, 1984). EPAs have demonstrated several cases where
these requirements are not met. For example: a lighting control strategy in one
building was not understood by the majority of occupants with a consequent dis-
satisfaction and misuse; deep cill height benches prevented some users from
reaching opening windows; a further natural ventilation strategy required
openings to be operated by three separate sets of people.
Where controls are automated then EPAs have observed many instances where they
do not function as intended, occasionally with a substantial energy wastage. The
guidelines given above for manual controls also apply to automatic controls.

-------
2521
CONCLUSIONS
Energy Performance Assessments suggest that passive solar buildings typically use
about 25 to 50% less heating and lighting energy than similar conventional
buildings. The energy benefit is matched by positive user reactions confirming
that sunlight and daylight are appreciated for their qualitative impact on the
internal environment. The energy saving function is often frustrated by
unresponsive heating and lighting systems revealing that there is still need for
development of effective control systems that are specifically designed to
enhance solar gains without interfering with user requirements and behaviours.
ACKNOWLEDGEMENTS
The EPA project is funded by the Department of Energy through ETSU. The EPA
teams at Databuild and UWCC have all contributed to this paper in addition to the
named authors. The views expressed in this paper are those of the authors and
the EPA teams; they do not necessarily concur with those of ETSU or D.En.
Finally, we must acknowledge the contribution to EPAs of the building owners,
designers, and occupants.
REFERENCES
Andersson, B. B. Erwine, R. Hitchcock, R. Kammerud, A. Seager and A. Hildon
(1987). Davlighting Performance Evaluation Methodology - Summary Report.
Report No. 24002, Lawrence Berkley Laboratory, California.
Birmingham School of Architecture (1985). Energy Conservation Measures in the
Sutton Coldfield Deanery First and Middle Schools. Walmlev. Report to Energy
Technology Support Unit, Harwell, UK.
ETSU (1991a). Solar Building Summary Report South Staffordshire Water Company.
ETSU 1160/SBS/4, Energy Technology Support Unit, Harwell, UK.
ETSU (1991b). Solar Building Summary Report The JEL Building. ETSU 1160/SBS/3,
Energy Technology Support Unit, Harwell, UK.
ETSU (1991c). Solar Bulldinp Summary Report Looe Junior and Infants School. ETSU
1160/SBS/l, Energy Technology Support Unit, Harwell, UK.
ETSU (1991d). Solar Building Summary Report Copper Beech House. ETSU 1160/SBS/2,
Energy Technology Support Unit, Harwell, UK.
Gordon, H. (1986). An Approach to Performance Evaluation of Non-residential Solar
Buildings. US Department of Energy.
Heap, L.J. Palmer, J, Hildon, A (1988) Redistributed Davlighting - A Performance
Assessment. National Lighting Conference, Cambridge 1988, CIBSE.
IEA Task VIII Subtask A (1985). Passive and Hybrid Solar Low Energy Building:
Monitoring and Performance Evaluation Procedures and Guidelines for Subtask D
Construction Projects. International Energy Agency.
Littler, J. M.Watson, P.Ruyssevelt, B.Ford (1983). Minor Field Trials. Project
Definition Study. Energy Technology' Support Unit, Harwell, UK.
Subbarao, K. (1984) Building Element Vector Analysis: a new hour bv hour building
energy simulation with system parameters as inputs. SERI/TR254:2195, Solar
Energy Research Institute, Golden, Colorado, USA.
Swisher, J. K.S. Harr, D.J. Frey, M.J. Holtz (1982). Performance Monitoring of
Passive Solar Residences at the Class B Level. SERI/TP-254-1675, Golden,
Colorado, USA.

-------
2522
SOLAR PERFORMANCE ANALYSIS FOR AN UNDERGROUND
DWELLING AT HIGH ALTITUDE
Lester L. Boyer, Professor
Department of Architecture
Texas A&M University
College Station, Texas 77843
ABSTRACT
A quantitative solar/lighting design analysis and evaluation has been conducted for
an energy conserving earth covered dwelling in a cold climate at high altitude in the
Rocky Mountains. Both simplified and rigorous techniques, including computer
simulation, model studies, and on-site measurements have been employed to analyze
the design and evaluate performance. Heating analyses include heat loss, solar
heating performance, indoor temperature swings, balance point temperature, and
back-up heating. Daylighting analyses include daylight/sunlight availability,
building envelope penetration, interior distribution, and glare control.
The solar glazing provided in this dwelling yields a heating load reduction of about 45
percent, and with the night installation provided this savings is increased to about 75
percent, compared to the same dwelling without solar. However, due to other passive
techniques, including super-insulation and earth covered construction, substantial
savings are already inherent in the basic non-solar building. Interior light levels
expressed as a percent of exterior levels typically range from about 10 to 30 percent.
Somewhat higher levels occur near the glazing. Regardless of the type of sky
condition, about 10 percent is normally available deep into the interior of the spaces.
The designed light distribution and glare control features have produced good quality
daylight in ample quantities.
KEYWORDS
Passive solar; daylighting; performance analysis; earth shelter/underground;
dwelling.
INTRODUCTION
The Building
The two-story structure is cut into the side of a hill and has a rectangular plan of 980
sq ft (90 sq m) with one of the long walls and windows facing southeast (Boyer,
1986a). The roof and rear wall are totally earth covered, as are much of the two end
walls except for a corner door and a course of glass block under the roof. Three
concrete modules comprise the structure. The center module is 2/3 of a square dome,
and the end modules each form 1/2 of a center module. An outward sloping parapet
forms a curved rib structure to carry the dome segments along the entire length of
the window wall. Two structural interior partitions with arched openings separate
the three dome modules. Windows represent more than 25 percent of the long
window wall facing southeast.
Plan and elevation drawings arc shown on Fig. 1. On the inside, above the sliding
glass doors in the center module, is an open-air timber-framed catwalk for
temporary storage of personal outdoor gear which serves as a partial internal
lightshclf and partially filters direct sunlight arriving through the high windows
on that wall. Bedroom lofts occur at the lightshelf level in both end modules.

-------
2523
k*"
k
1
/
ib t
: F]>!; l t^ltir'^rlTrr;;
_i	!	i	i—L. :
! j 1 1 I | j
- ...i..l | | I | i i I
_LJ.
! I
i I
: • pin i m c. I
1 : "T^' i ' 1
I
I
1
i i..i I.;
i-L-

xn:
! I I
r" T
233


l&.LT.Ctt5 «rl>4
ijCxJl foo'i
o 2 it 6 e io
Fig. 1. Main floor plan and end elevation.
TABLE 1 Winter Design Conditions
Indoor Temperature 68F (20C) solar building
Outdoor Temperature -10F (-23C) 2.5% data
Wind Velocity	15 mph (7m/s)
Infiltration/Ventilation 0.5 air changes/hour
Sile and Climate Data
The site is located at 38 degrees north latitude and about 105 degrees west longitude.
Based on 2.5 percent ASHRAE data for nearby locations, a winter outdoor temperature
of -10 F (-23 C) is assumed. During summer a temperature of 91 F (33C) with a daily
range of 32 F (18 C) and a design wet bulb of 65 F (18 C) may be assumed. Winter
design conditions are shown on Table 1.
Very high solar gains are obtained at the south-central Colorado site which is at 9,200
ft (2,800 m) altitude. The probability of sunshine at mid-day throughout the winter
season is greater than 80 percent. In mid-morning of late December, the diffuse
exterior illumination levels on the horizontal plane are near 500 fc (5,400 lux) for
both overcast conditions and clear sky without sun. With direct sun added, the
resulting levels are about quadrupled. The climate produces about 8,000 heating-
degree days F (4,500 HDD C), and significant snowfall occurs each year.
The above heating-degree day data assume that the building will have a balance
point temperature (outdoor temperature at which heating is needed) of about 65 F (18
C). Such buildings were common in past decades, but buildings of today can be much
better (lower balance points), especially if highly insulated, earth sheltered, and
passive solar heated. Balance point temperatures in such structures of well below 50
F (10 C) may be expected (Germer, 1985). For such cases, the total annual heating-
degree days will be reduced by more than half of the normally stated values.
SOLAR HEATING ANALYSIS
Winter Heal Loss
The winter heat loss, assuming adiabatic solar glazing, is about 29,330 Btu/hr (8,600
watts). All the envelope surfaces contribute over 19,000 Btu/hr (5,600 watts) and the
infiltration/ventilation contribution is about 8,400 Btu/hr (2,400 w). See Boyer
(1990).

-------
2524
The non-solar glass, exposed concrete walls and earth covered roof each contribute
about equally to the total envelope loss. The slab edge loss is small due to insulation
under the floor slab and the buried horizontal insulation at the exterior perimeter.
The single largest contribution to the load comes from the 0.5 air change rate per
hour, which is about 100 cfm (170 cu m/hr). This ventilation rate would provide the
recommended fresh air quantity for six occupants on a continuous basis (Boyer and
Grondzik, 1987). However, this amount of outdoor air could be throttled down at the
fresh air intake when fewer occupants were present.
¦Solar Glass and Mass
Preliminary sizing of the solar collection glass area and the exposed internal mass
area for a direct gain solar system was accomplished with the help of guidelines as
shown in Table 2. Suggested target criteria from a United States map (Stein, 1986)
indicated a solar savings fraction of SSF=70 percent and an envelope insulation value
of R-20. The solar savings fraction is the difference between the auxiliary heating
TABLE 2 Sizing Requirements for Solar Glass and Mass
Target Criteria: Solar Savings Fraction = 70%
Exposed Wall Insulation = R-20
Location Dbl. Glass Area Solar Savings Fraction
	in Colo.	% floor sq ft	No insul. R-9 insul.	
Denver	12-23 118-225	27-43 47-74
Pueblo	11-23 108-225 29-45 48-75
Eagle	14-29 137-284 25-35 53-77
Exposed surface mass for direct gain system for SSF = 70%:
210 lbs/sq ft of solar glass = 44,100 lbs or 588 ft2 of 6 in. concrete. For 6 in.
concrete in direct sun, mass-to-glass ratio = 3.0; 210 ft2 glass x 3 = 630 ft2
TABLE 3 Performance Estimate for Direct Gain System
Double Glass Area
BLC LCR
Bide. Load Cocff. Load Coll. Ratio
SSF
Back-up
Heating
Design Collector 210 sq ft
patio doors +
hieh center elass
9,000 Btu/HDD
42.9 Btu/HDD-ft2
45
20,200
Btu/ft2»y r
Total South Glass 307 sq ft 7,680 Btu/HDD
Note: not all class is effective due to some
25.0 Btu/HDD-ft2
(HDD = F»day)
shading: some cast and
65
west
12,900
Btu/ft2«yr
facing.
TABLE 4 Determination of Balance Point Temperature
Design temperature difference (68 + 10F)
78 F
(25.5 C)
Heat loss due to envelope and infilt/vent
Internal gains: people, appliances, lights
Solar gains: ave. Jan. daily insolation on vertical
29,332 Btu/hr
2,265 Btu/hr
12,820 Btu/hr
(8,590 W)
(665 W)
(3,755 W)
Balance point temperature for solar building
27.9 F
(-2.2 C)

-------
2525
required by a conserving but non-solar building and thai required by a solar heated
building, expressed as a percent. It is not the percentage of the building heat
supplied by the sun. For the three nearest regional climate data sets available, an
SSF-70 percent might be achieved with something in excess of 200 sq ft (20 sq m) of
glass if night insulation is provided over this glass. Further, guidelines indicate that
for 6 in. (15cm) thick concrete, a surface area equal to about three times the solar
glass area would be needed in direct solar exposure for several hours each day
(Balcomb, 1982; Stein, 1986; Mazria, 1979; and TEA, 1980). About, 210 sq ft (20 sq m) of
deliberate solar glazing has been provided, and the requirement of 630 sq ft (60 sq m)
of exposed mass is easily provided by the adjacent floor area and that of the side and
rear walls of the center module. The night insulation provided for the patio doors
produces an R-6.7 value.
Heating Performance Estimate
Both the building load coefficient BLC and the load collector radio LCR must be
determined to estimate the actual solar savings fraction SSF (Balcomb, 1982; Stein,
1986). These values are presented in Table 3. The BLC is the total heat flow occurring
through the non-solar portion of the envelope, including infiltration, expressed per
day and per degree of temperature (per heating-degree day). This BLC value is
obtained as follows: 29,300 x 24 hrs/78F = 9,000 Btu/F.day (4,700 W hr/C.day or 200
W/K). The LCR is obtained through dividing BLC by the deliberate solar collection
glazing area, which produces about 43 Btu/F.day.sq ft (240 W hr/C.day.m2 or 10
W/K.m2). These values correspond to SSF equals about 45 percent.
Back-up Heating Requirement
If a balance point temperature of 50 F (10 C) is assumed, then the back-up heating
requirement can be determined as follows (50 F is the lowest base for which data are
commonly available); H = (1 - SSF) x BLC x HDD 50F = (1 - 0.45) 9,000 x 4,000
= 19.8 million Btu (5,800 kWh)
or 20,200 Btu/sq ft.yr (5.9 kWh/sq ft.yr)
This back-up heating requirement, which is also tabulated in Table 3, comprises less
than half of the classic good practice recommendation of 55,000 Btu/sq ft.yr
suggested by the Building Energy Performance Standards BEPS (Stein, 1986).
However, with an expected balance point temperature of less than the 50 F (10 C)
assumed, the resulting back-up heating requirement would be accordingly less.
Indoor Temperatures
Calculated winter indoor temperature fluctuations arc determined as illustrated in
Fig. 2. The average expected temperature without auxiliary heating is about 57F (14C)
with a range of plus or minus 8.IF (4.5C). The average indoor temperature is a few
degrees higher if all the glass exposure is considered.
The calculated balance point temperature for the building is determined from values
shown in Table 4. The calculated value is 27.9F (-2.2C). The measured indoor
temperature of the unoccupied building with solar radiation blocked at the patio door
collectors is well above freezing at 46F (8C). The building is typically unattended for
a 2-month period prior to such site measurements over the past six years.
Alternate back-up heating systems are provided. One of the systems is a small wood-
burning stove which must be continuously operated for a period of several days in
order to bring the inside temperature up from the unattended condition to the design
condition.

-------
2526
F Temperature
70--
60--
50--
30
20"
10-
65
57i
49
indoor design temp,
upper
72
upper
average
indoor
lower
solar
gains
(LCR=40)
internal
62
52
average
indoor
lower
solar
gains
(LCR=30)
internal
- 20
" 15
- 10
average January outdoor temp. (Denver)
For only
For all
designed
glazing:
collector:
includes
patio doors
kitchen
and high
and two
center glass
bedroom

lofts
. 0
. -5
-10
-15
Fig. 2. Calculated indoor temperatures on
clear January day without auxiliary
heating
DAYLIGHTING ANALYSIS
Illuminance
-•2000
¦•1750
¦•1500
¦ -1250
-1000
-• 750
-•500
.. 250
Ml M2M3M4
C1C2
Fig. 3. Comparison of computer simulation
and site measured interior daylight
levels in winter (medians)
The daylighting analysis software program DAYLITE produced by Solarsoft, Inc. was
used to simulate the center module (Ashton, 1984). Both clear sky and overcast sky
conditions were simulated using available exterior illuminance data for Denver
(Robbins, 1986). Calculations were made at 28 points on a grid pattern. Simulation of
the apertures wall included two sliding door apertures, an overhang, interior
lightshelf, and an exterior wall fin at each end. The curved ceiling was simulated
with a raised sawtooth configuration with a wide horizontal aperture band. Ceiling
and wall reflectances were 85 percent, and the glazing light transmission factor was
75 percent. Results for a winter day are shown in Fig. 3 (Boyer, 1986b).
Interior on-site measurements were also made with two portable meters in tandem
under four different sky conditions in mid-winter. These were representative of the
range of conditions used with computer simulation. As shown on Fig. 3, the range of
the medians of the four site measurement sets is larger than the spread between the
two medians of the computer calculations. The median value of all the data sets is a
daylight factor (DF) of between 20 and 25 percent.
Scale model studies of the center module have also been conducted in the sky
simulator at Texas A&M University. These results have been compared with those
from computer simulation and site measurement data (Boyer and Atif, 1988).
Differences between scale model data for both furnished and unfurnished situations
can be quite marked deep into the space. The scale model (furnished) and actual site
measurements are quite close. The computer simulation seems to overestimate
performance in the front of the structure and indicates a less balanced distribution
than actually occurs. Levels of daylight factor are mostly between 10 and 30 percent,
with the computer predicting some potential near 50 percent (Boyer, 1990).

-------
2527
SUMMARY
The simplified solar analysis approach indicates that for the glazing provided in this
dwelling a heating load reduction of about 45 percent could be expected, and with
night insulation provided this savings is increased to about 75 percent compared to
the same dwelling without solar. However, for passive techniques such as extra
super-insulated, or earth sheltered and covered, substantial savings are already
inherent in the basic building of perhaps 25 percent or more. The balance point
temperature for the structure was determined to be below the freezing point.
Temperature measurements in the unoccupied structure, even with solar radiation
blocked at the patio doors, have been well above the freezing point. To satisfy the
small back-up heating requirement, several redundant systems have been provided,
including wood-burning stove, propane units, and forced-air perimeter system. In
mid-winter, several days of continuous heating with one of these systems are needed
to bring the inside air temperature up from the unattended condition to the design
condition.
Interior light levels expressed as a percent of exterior levels typically range from
about 10 to 30 percent daylight factor. Somewhat higher levels occur near the
glazing. However, regardless of the type of sky condition, about 10 percent is
normally available deep into the interior of the spaces. Very good light distribution
and glare control features have produced good quality light in ample quantities.
REFERENCES
Amer. Soc. Heating, Refrigerating and Air-Conditioning Engineers (1989). A S H R A E
Handbook-1989 Fundamentals. Ch. 24, Atlanta.
Ashton, William (1984). DAYLITE Program. Solarsoft - Div. Kinetic Software, Inc.,
Burlingame, Calif. Also in Kolar, William (1984). DAYLITE: A measure of visual
comfort, Solar Age. May.
Balcomb, J.D., ed. (1982). Passive Solar Design Analysis, Vol. II and Vol. Ill of Passive
Solar Design Handbook. U.S. Dept. of Energy, Wash, D.C.
Boyer, L.L. (1986a). Design and construction of a remote, fully passive, two-story
earth covered residence at high altitude, Proc 2nd Intl. Earth Sheltered Buildings
Conf.. Minneapolis, L.L. Boyer and R.L. Sterling, eds., Texas A&M Univ., College
Station, June.
Boyer, L.L. (1986b). Daylighting prediction and measurement in an earth covered
residence, Proc. 11th Natl. Passive Solar Conf.. Boulder, Colo., Amer. Solar Energy
Soc., June.
Boyer, L.L. and W.T. Grondzik (1987). Earth Shelter Technology. Texas A&M Univ.
Press, Drawer C, College Station.
Boyer, L.L. and M.R. Atif (1988). Comparison of daylighting distribution in an
underground dwelling high in the Rockies using computer algorithms, scale
models, and on-site measurements, Proc. 3rd Intl. Conf. Underground Space and
Earth Sheltered Buildings. Tongji Univ., Shanghai, PRC, Sept.
Boyer, L.L. (1990). Effectiveness of solar heating and lighting in an underground
concrete and glass dwelling high in the Rocky Mountains. Proc. Intl. Symposium
on Unique Underground Structures. Colorado School of Mines, Earth Mechanics
Institute, and U.S. Bureau of Reclamation, Denver, June.
Germer, Jerry (1985). How low-energy homes stack up. Solar Age. Vol. 10, No. 8,
August.
Mazria, Edward. (1979). The Passive Solar Energy Book, expanded professional ed.,
Rodale Press, Emmaus, Penn.
Robbins, C.L. (1986). Daylighting Design and Analysis. Van Nostrand Reinhold, New
York.
Stein, B., J. Reynolds, W. McGuinness (1986). Mechanical and Electrical Equipment for
Buildings. 7th ed., Ch. 5, App. A, B, C, John Wiley, New York.
Total Environmental Action, Inc. (1980). Thermal Mass Pattern Booklet. Church Hill,
Harrisville, New Hamp.

-------
Intentionally Blank Page

-------
3.2 Zero-Energy Building Designs

-------
Intentionally Blank Page

-------
2531
POSSIBILITIES FOR ZERO ENERGY IN SOLAR HOUSES
Peter D. Lund
Helsinki University of Technology
Department of Technical Physics
SF-02150 Espoo, Finland
ABSTRACT
New building materials, solar technologies, passive solar design, and energy
conservation give novel opportunities in reducing the energy demands in buil-
dings. This paper discusses new approaches in reaching a zero-energy target in
single-family houses. Optimum sizing of solar conversion and energy storage
subsystems are discussed. The main stress is in the seasonal energy storage
question for which electrolytic hydrogen is considered as a possibility.
KEYWORDS
Solar energy; energy storage; hydrogen; simulation
INTRODUCTION
Recent developments in material, conservation, and solar technologies have opened
new possibilities for reducing the energy consumption in buildings. Optimum use
of new building materials, e.g. transparent insulation, for solar house design
may drop the yearly space heating demand by 50-75% even in a harsh climate.
Furthermore, significant savings may be obtained in the electricity consumption
through electric efficiency and solar photovoltaics. A very common feature for
these low energy buildings is the strong dependence of the performance on the
solar radiation input, as the main auxiliary energy demand will mainly concentrate
the winter months when the solar radiation availability is at lowest.
The next step forward from a low energy building design with mainly passive solar
design and energy conservation utilization would be to apply active solar heating
and photovoltaics to approach a total energy independence. Due to the mismatch
between the residential load and the solar insolation, a zero-energy house design
would necessitate effective storage of energy on a seasonal basis.
The major objective of this paper is to investigate new possibilities and
concepts for zero energy in houses based on solar energy and energy storage
technologies. The study is restricted to a single residential building conside-
ring both thermal and electrical energy needs. The analysis is based on a new
design tool that is able to find an optimum collector area and storage volume to
satisfy a given building energy load. For the building energy analysis, we use
numerical simulation models. Although the study is based on computer simulations,
experimental work to verify some of the findings is also under way in the frame
of the national NEMO-programme.
The zero-energy question is approached firstly by finding analytically an
approximate value for the optimum solar collector area and storage volume to
reach full independence of external heat and electricity. Secondly, possible
technology solutions are searched for to satisfy the storage functions. Both
thermal and photovoltaic technologies as well as thermal and electric storage are
considered. Thirdly, the optimized solar system is analyzed as an integral part
of a building through a comprehensive building energy simulation to assess the
interactions of the solar energy flows with other energy flows within the
building. Also, the influence on the overall building performance is then
considered.

-------
2532
STORAGE DEMAND
The demand of energy storage depends on the type of building and on the location
of the site. For example, in a Finnish climate (60 °N), the building energy demand
may be dropped by 50% through conservation, new building materials, and passive
solar design; 50-70% of the remaining load may then be satisfied by solar thermal
and photovoltaics directly, but 30-50% will need seasonal compensation. Figure 1
demonstrates the effect of passive solar and energy conservation on the space
heating load profiles for different designs of a 120 m2 house in a 60 °N climate.
The most advanced concept utilizes here a large 20 m2 sunspace with transparent
insulation. The building energy analyses were based on the FHOUSE simulation
programme (an improved version of NBSLD) and the load profiles are also employed
in the analyses to follow.
To study the effect of building type, electrical efficiency, and climate, a
variety of possibilities are considered shown in Table 1.
Reference house
100
Daily energy
flow [kWh/d]
space heating
space cooling
-20
0 50 100 150 200 250 300 350
Time in days
Passive solar house
-20
Advanced solar low energy house
-20
-40
-60
0 50 100 150 200 250 300 350
Fig. 1. Effect of passive solar design and energy conservation on
the space heating demand of a 120 m2 house (60 °N).

-------
2533
As to storage,	the thermal and electric storage technologies available today
are not able to satisfy the seasonal storage demand in a reasonably compact form
for a single house. Scaling up diurnal storage technlogies to a seasonal mode
would lead to a 40-70 m3 water tank (40 kWh/m) with 100 cm insulation all around,
and to a 30,000-40,000 kg lead-acid battery bank (35 Wh/kg), respectively. Compa-
red to present storage technologies, a 20-30 fold improvement in storage capacity
would be needed to fulfill a seasonal storage requirement.
The central role of the storage is also demonstrated through Fig. 4 that shows
the solar fraction vs. the storage capacity and array size. The performance curve
for a 2 kW array vs. storage capacity indicates that no substantial improvement
is obtained when moving from a diurnal storage [1-10 kWh] to a weekly or monthly
storage [50-200 kWh]. The real improvements in the solar fraction take place at
a much higher storage capacity [500-800 kWh]. The compensation of storage through
increasing the PV array size is very ineffective in this case as indicated by the
table in Fig. 4.
Solar
fraction,
100
80
60
40
20
0
2 kWp
10 kWh
storage:
[kWp] [%]
2
59.1
3
65.4
4
69.0
5
71.1
6
72.4
1	10 100 1000
Storage capacity, [kWh ]
Fig. 4. Solar fraction versus storage capacity and PV array size.
[1500 kwh/a electric load, 60 °N climate].
ADVANCED ENERGY STORAGE TECHNOLOGIES
The low energy density (kWh/m3, kwh/kg), or large heat losses in small scale, are
the major problems for storage in a zero-energy approach. More advanced technolo-
gies that have the potential for seasonal energy storage include (Lund, 1990):
electrolytic hydrogen (electricity)	i)=0.4, 33.3 kwh/kg
Ni-H2 advanced battery (electricity)	i)=0.8, 0.06-0.1 kWh/kg
thermochemical heat storage (heat)	t|=0.9, 0.3-1.1 kWh/kg
In case of hydrogen technologies, solar electricity is used to produce electroly-
tic H, in the summer and store it for the winter use. Hydrogen may be converted
back to electricity by an engine or a fuel cell.
A Bolar-Hj and a Ni-Hj-battery system was studied more closely for a residential
application through a numerical analysis method shown in Fig. 5. In this case,
precalculated or measured load profiles were used as input. The solar-storage
system was optimized for a residential electric load of 1500 kwh/a representing
a future technology case. The minimum electric demand is 2.1 kWh/d and the winter
maximum 11.0 kWh/d, respectively. The outcome of the analysis is shown in Fig. 6.

-------
2534
TABLE 1
Maior inDut
data

CLIMATE
Ambient temp. Solar
radiation
45 °N
6 °C
1120
kwh/m2,a
50 °N
10
1380

60 °N
12
1930

THERMAL LOAD (60 °N)
Space heat
Cooling
Hot water
Reference house
8070 kWh
0 kWh
3000 kWh
Solar house
4845
480
3000
Advanced solar
2490
3210
3000
low energy house



ELECTRIC LOAD
Appliances
& Lighting

Reference
4500 kWh


Electric efficiency
2800 kWh


Future technology
1500 kWh


OPTIMIZED COLLECTOR AND STORAGE SIZES
The solar collector and storage are the most important components of a solar
energy system both in respect to economic and thermal performance. A novel method
(Lund, 1991) for optimizing the solar collector and energy storage size for any
given load and climate was employed to find out the requirement of these com-
ponents for the cases given in Table 1.
Figures 2 and 3 show a summary of the findings. Here both the thermal and
electrical load of a house are to be covered 100% through solar and storage, i.e.
a zero level of purchased energy is reached. It is observed that the size of both
the solar thermal collectors [m'] and photovoltaic array [kW ] are reasonable for
a house application even with present technology and in a Harsh climate. In all
cases considered, less than 40 m of high performance flat plate collectors and
7 kW of c-Si cells would be needed. Consequently, these may also be considered
for possible direct integration into building structures.
Load type: deg N:
45 50 60 45 50 60 45 50 60
Reference house
Passive solar house
Low energy house
Coll
14 &31 38
11 17 26
8 12 18
Stor
[MWh)
2.9 4.1 6.4
2.0 2.4 3.2
1.3 1.6 2.6
Load
[MWhl
5.5	8.5 11.1
4.6	6.4 7.8
4.3 6.1 5.5
Fig. 2. Optimal solar thermal design of a zero-energy house,
[high performance collectors with UL=2.8 W/m K, i)g=0.83,
storage efficiency=0.75, temperature levels 80/40°C]
Load type: degN:
45 50 60 45 50 60 All
Reference
Electric efficiency
Future technology
PV
|kW0
3.4 4.9 6.5
2.1 3.0 4.0
1.1 1.6 2.2
Stor
[MWh]
1.4 1.8 2.2
0.9 1.1 1.4
0.5 0.6 0.8
Load
[MWh]
4.5
2.8
1.5
Fig. 3. Optimal solar electric design of a zero-energy house.
[c-Si solar cells (»)=0.12), storage round-trip efficiency=0.4]

-------
2535
The component sizes of a H2 system are about 30 % larger than that of a Ni-Hp
based system because of the lower round-trip efficiency. However, the volume of
a pressurized H2 storage is much smaller than that of a battery. The seasonal
storage supplies the load for about 90 days.
PHOTOVOLTAIC
ARRAY
AC/DC CENTRE
BACK-UP ELECTRO-
BATTERY LYSER
RJEL
CELL
Hourfy weather
data
Or"
Conversion into daily
solar radiation on a
tilted surface
Hourly reference
electrical load
Q
Daily
insolation
hQn
Building energy	I
analysis ~V J p ~
• internal
PC file
Daily net
heat load
Daily electrical
mm load
Energy savings
and conversion
into daily values
-Ct-
System input
parameters
QrJ
Daily building
energy balance
Solar-hydrogen
system performance
analysis
<1
Fig. 5. Methodology approach for solar-storage analyses.

-------
2536
H2-storage
ENERGY 40
FLOW
[kWh/d] 30
storage capacity [kwd]
photovoltaic
output
electric
load
mm
100 200 300
TIME in days
Ni-H2-battery
ENERGY
FLOW
[kWh/d]
storage capacity [kwd]
photovoltaic
output
electric
loa
100 200
TIME in days
300
Fig. 6. Daily energy flows in a fully self-sufficieny solar electric system,
[electricity demand= 1500 kWh/a, Ni-Hj: 2.0 kW /700 kWhf, H
2. 2.9 kWp/1085 kWhs
CONCLUSIONS
Possibilities for approaching a zero-energy level in a residential house have
been studied through optimization and performance studies. The analyses accom-
plished indicate that present solar energy conversion technologies are good
enough to enable a reasonably sized solar system for a house, but in case of
storage a 20-30 fold improvement in storage capacity would be required.
Solar-hydrogen technology seems to be a very promising approach for total
independency of grid electricity in a house. For a future energy-efficient house,
a 2-3 nr pressurized hydrogen vessel may be enough for seasonal storage in 45-60
°N climates. A PV array of 1-2 kW would be adequate, respectively. The higher
values are representative for a harsh northern climate.
REFERENCES
Lund, P. (1990), Advanced Energy Storage Systems for Solar Low Energy Buildings.
IEA Solar Heating and Cooling Task 13 expert meeting, March 28-30, 1990, Bregenz,
Austria.
Lund, P. (1991), Optimization of Stand-Alone Photovoltaic Systems with Hydrogen
Storage for Total Energy Self-Sufficiency, submitted to Int. J. of Hydrogen Ener-
gy-

-------
2537
THE SELF-SUFFICIENT SOLAR HOUSE FREIBURG
A. Goetzberger, W. Stahl
Fraunhofer-Institut fiir Solare Energiesysteme
Oltmannsstr. 22
7800 Freiburg, Germany
ABSTRACT
The Fraunhofer-Institute for Solar Energy Systems is planning and building a completely self-
sufficient solar one-family house in Freiburg, Germany. The entire energy demand for heating,
electricity and cooking is supplied by the sun. The combination of highly efficient solar systems
with conventional means to save energy is the key to the successful operation of the house. The
meteorological conditions in Freiburg (48 degrees latitude) are characterized by abundant solar
radiation in summer but long dark periods in winter. Therefore, seasonal energy storage has to be
provided. It is accomplished by electrolysis of water and pressurized storage of hydrogen and
oxygen. The energy for electricity and hydrogen generation is supplied by solar cells. Hydrogen can
be reconverted to electricity with a fuel cell or used for cooking with a catalytic heater. It also
serves as a back-up for low temperature heat. There are provisions for short term storage of
electricity and optimal routing of energy.
KEYWORDS
Transparent Insulation; solar house; self-sufficient; hydrogen/oxygen storage.
INTRODUCTION
The Fraunhofer Institute for Solar Energy Systems in Freiburg is attempting to realize,for the first
time in Germany, a completely energy independent solar house. (Goetzberger, 1987). The main
problem at such high northern latitude (48 °) is the strong variation of insolation between summer
and winter as shown in Fig. 1. Since seasonal storage of large amounts of energy today is both
technically and financially prohibitive for decentralized units, it was decided to reduce energy
demand by all available energy savings technologies without impairing the comfort of the
occupants. Only the small remaining energy deficit in winter will be provided by a modest storage
capacity for high quality energy. (Heinzel 1991).
In a conventional residential building 80 % of the energy demand is for space heating. On the
other hand, the average insolation on the envelope of such a structure is comparable to the heat
transmission losses. Therefore, optimized use of passive solar heating technologies should make it
possible to avoid seasonal heat storage. The institute has developed and demonstrated transparent
insulation systems as a means of efficient utilization of solar radiation. In addition, other
technologies that were developed at the institute are combined in an optimized total energy system
to realize this project.

-------
2538
The other components are a highly efficient collector system based on transparent insulation for
domestic hot water (Goetzberger, 1991) and a photovoltaic generator in conjuction with a
hydrogen/oxygen storage system.
Planning of the project started in 1988; the start of construction is scheduled for June 1991 and
completion is expected in 1992. A three-year period of measurements in the occupied building
with 150 m2 living area will follow. Since no conventional energy sources will be used, this house
will operate without emitting any pollution.
The objectives of the project can be summarized as follows:
- Solar energy is substituted for other, environmentally dangerous energy carriers
Integration of new concepts of solar architecture into an energetically optimized
structure is demonstrated
- Advanced technologies for energy conservation are applied
- New solar energy systems are demonstrated
It is not the intention to reach economic viability with this project, but it can be expected that
many of the new technologies and components to be tested will find their way into practical
application.
diffuse light
1-
J FMAMJ JASONDJ
Month »
Fig. 1. Yearly distribution of solar radiation on a horizontal plane for Freiburg (48 °) •
TRANSPARENT INSULATION, ARCHITECTURE
The transparently insulated walls (Tl-walls) (Goetzberger, 1984) influence the layout of the
building to a large extent. In a number of experiments (Stahl, 1989), it could be demonstrated that
Tl-walls not only minimize heat transmission losses, but also convert the facade into a source of
heat, thereby compensating losses from other parts of the building and ventilation losses.
Dependent on the circumstances the gain per heating period is 100 - 200 kWh/m2. In combination
with other already established means for the minimization of heating demand it is thus possible to
heat a building exclusively with Tl-walls.
In contrast to existing low energy houses transparent insulation systems lend themselves to an
open design and natural living. Diffusion of moisture through walls is possible by appropriate
ventilation slots in the Tl-elements.

-------
2539
The optimal shape of the building differs from usual building design. Whereas up to now a
compact shape was best, with Tl-walls the optimization of energy gain is pre-eminent. This
problem was investigated with a specially developed simulation program. (Wilke, 1991).
The building (Fig. 2) consists of two storeys, a well-insulated basement and a flat roof which
supports the structure for the thermal and photovoltaic collectors. The ground-plan has the shape
of a circular segment. Sharp corners and thermal bridges are avoided. The south facade with a
length of 22 m consists of transparently insulated walls interrupted by optimized windows. The
ground floor contains two small apartments. The first floor houses demonstration and lecture
rooms for visitors and the measuring equipment. All occupied rooms are oriented toward the
south, while floors and the staircase are facing north. The total utilizable area is 150 m2.
Ventilation is achieved by a highly efficient heat recovery system. It is in operation only for three
to four months in winter.
The heat energy demand of the building excluding solar gains is 3.500 kWh/a which is in the range
of the best low energy houses in the same geographic region. Solar gains by transparent insulation
and through windows reduce the heating demand for the average year to 300 kWh/a. This results
in a heating energy of 2 kWh per m2 and year.
Fig. 2 View of the south facade of the self-sufficient solar house
(designed by Planerwerkstatt Vorstetten).
DOMESTIC HOT WATER SYSTEM
In accordance with the principles outlined above, the domestic hot water system was designed for
highest solar fraction. For this purpose a collector having very high efficiency at low insolation was
required. The limited space available for the collector had to be taken into consideration as well,
since most of the roof area is taken up by the photovoltaic generator. In simulation runs several
types of collectors were tested, including a vacuum tube collector. The best performance was
obtained with the new bifacial-absorber collector which will be presented at this conference.
(Goetzberger, 1991). This type of collector will be used for the first time in a practical system.
With 14 n? collector area and a 1000 1 stratified storage tank a solar fraction of 90 % will be
reached. Demand is 1901 per day including hot water for the washing machine and dishwasher.
The remaining 10 % of the demand amounting to 230 kWh occurs in the months of December,
January and February, and is supplied by the the YL^/O^ storage system. Thermochemical and

-------
2540
phase change storage was excluded because of high complexity and high auxiliary energy demand
for pumps and controls.
ELECTRICAL SYSTEM AND HYDROGEN/OXYGEN STORAGE
All loads requiring high exergy are supplied by 50 m2 of solar cells oriented toward the south with
an inclination of 45 °. Short term storage is accomplished by a 20 kWh lead acid battery.
For seasonal storage a new hydrogen/oxygen storage system has been developed at the institute. A
pressure electrolyzer splits water into its constituents. Both gases can be stored for any length of
time almost without loss in commercial gas tanks. For economic reasons a pressure of 30 bar was
selected leading to a volume of 15 m3 for the H2 tank and 7,5 m3 for the 02 tank. The gases are
either recombined catalytically to give high or low temperature heat or are fed to a fuel cell to
generate electricity. (Ledjeff, 1986; Ledjeff, 1988). The additional storage of oxygen not only
improves the energy balance but makes a completely closed system possible with pure water as the
reaction product which is recycled into the electrolyzer.
TABLE 1 Efficiencies of the Main Components of the System
PV-generator 11 % Electrolyzer > 80 %
Inverter > 90 % Fuel Cell > 50 %
Catalytic Heater > 95 %
Gas Storage > 98 %
The entire system is under the control of an integrated data acquisition and processing system. A
safety battery provides back-up energy for this system. These components and others like pumps,
fans, valves etc. are supplied with d.c. current. Total demand for this purpose is 1.100 kWh/a. TTie
occupants can use 700 kWh/a of a.c. power generated with a high efficiency inverter. This is
adequate if currently available high efficiency household appliances are used.
Another 700 kWh/a are intended for process heat produced by catalytic combustion for cooking.
The electrical system and the H2/02 storage are described in (Bopp, 1991; Heinzel, 1991).
OVERALL SYSTEM SIMULATION
The self-sufficient house is a complex system of many mutually dependent energy paths. System
simulation is therefore crucial for optimization. The program TRNSYS was employed for this
purpose. (Sick, 1991). For many components new modules had to be written. The simulation
program proved very valuable for studying the influence of changes of components, of climatic
conditions and behaviour of the occupants. The resulting dimensions of the components can be
found in table 2. Energy flows from the generators to loads are given in table 3.
TABLE 2 Dimensions of Components
Hydrogen tank
15 m3/1.400 kWh
Oxygen tank
7.5 m3
Lead acid battery
20 kWh
Electrolyzer
900 W rated power
Fuel cell
600 W rated power
Inverter
3000W rated power

-------
2541
TABLE 3 Energy Flows from the Generators to Loads
Photovoltaic system
Transparent Insulation system
Solar collectors
4.500 kWh/a
3.000 kWh/a
4.000 kWh/a
a.c. consumers
d.c. consumers
Process heat for cooking
Auxiliary energy for domestic hot water
Auxiliary energy for space heating
Circulation pump in the collector loop
Fans in ventilating system
700 kWh/a
1.100 kWh/a
700 kWh/a
230 kWh/a
300 kWh/a
60 kWh/a
60 kWh/a
ENERGY PAY-BACK AND BENEFIT FOR THE ENVIRONMENT
Today conventional energy sources are required for the manufacture of regenerative energy
systems. This adversely influences the environment, but the operation of this system provides
energy without harm to the environment. It can be attempted to estimate the energy pay-back time
of the self-sufficient solar house. This is only possible with a large error margin because the energy
expenditure for production of many components is not well-known.
We estimate a primary energy equivalent for the self-sufficient house of 215 MWh compared to
70 MWh for a low energy house of the same size without solar systems. The low energy house
consumes primary energy equivalents per year of 5 MWh for heating, 5 MWh for domestic warm
water and 9 MWTi for electricity. Therefore, the self-sufficient house will recover the expended
energy within 10 years. Thereafter this house will provide its services with no adverse influence on
the environment.
The project is planned for a duration of 6 years. The scientific part including planning, simulation,
measurements and evaluation is supported by the Federal Ministry for Research and Technology
with a total of 5.8 Mio. DM. Most of it is for personnel, with 800 000 DM intended for newly
developed technical installations. The state of Baden-Wurttemberg has been supporting the
development of the Hj/O, storage system with 500 000 DM per year. The building lot was
furnished by the city of Freiburg at preferential conditions. TTie Fraunhofer Society, finally,
assumes the cost of the construction including insulation and low energy loss windows.
At this time it is not possible to estimate the cost of a second generation self-sufficient house
because the H2/02 system is still under development. If there is no I-L/02 system and the
remaining energy demand is supplied by four 501 bottles of liquid gas, then trie cost is estimated at
600 000 DM.
Bopp, G. (1991). The self sufficient solar house - electrical concept. To be published in the Proc. of
this conference.
Goetzberger, A. (1987). Projekt eines energieautonomen Einfamilienhauses mit
Wasserstoffspeicherung. 1. Int. Energie-Forum. Hamburg.
FINANCIAL SUPPORT AND COST
REFERENCES

-------
2542
Goetzberger, A., J. Schmid and V. Wittwer (1984). Transparent insulation systems for passive
solar energy utilization in buildings. Int. I. Solar Energy 2. 298.
Goetzberger, A., J. Dengler, M. Rommel and V. Wittwer (1991). The bifacial-absorber collector: a
new highly efficient flat plate collector. To be published in the Proc. of this conference.
Heinzel, A. and K. Ledjeff (1991). The self sufficient solar house - hybrid energy storage system.
To be published in the Proc. of this conference.
Ledjeff, K. (1986). Energy Storage in PV-systems. IEA Proc. Task VII. Wien.
Ledjeff, K. (1988). Wasserstoffsysteme als Energiespeicher fur PV-Anlagen. Tagungsb. 6. Int.
Sonnenforum. Berlin, 669-674.
Sick, F. and W. GrieBhaber (1991). The self sufficient solar house - remarkable simulations results.
To be published in the Proc. of this conference.
Stahl, W. (1989). Wall heating with transparent insulation - results from realized demonstration
projects. Proc. of 2nd European Conference on Architecture. Paris, 247.
Wilke, W.-S. (1991). Transparente Warmedammaterialien in der Architektur - Anwendungen,
thermisches Systemverhalten und optimale Raumklimakonditionierung. Dissertation FhG-ISE.

-------
2543
THE SELF-SUFFICIENT SOLAR HOUSE: HYBRID ENERGY STORAGE SYSTEM
A. Heinzel, K. Ledjeff
Fraunhofer-Institute for Solar Energy Systems, Oltmannsstr. 22, 7800 Freiburg, Germany
ABSTRACT
For an energetically self-sufficient solar house, a very efficient and maintenance-free energy storage
system is required. Therefore, a special hybrid storage system is under construction consisting of a
lead acid battery, a pressure electrolyser for hydrogen and oxygen generation, pressure vessels for
gas storage and a fuel cell for redelivery of electrical energy by conversion of the gases. This system
will meet the needs of the household for electrical energy and also for high temperature heat by
combustion of hydrogen.
KEYWORDS
Solar energy storage, electrochemical energy conversion, hydrogen technology, water electrolysis,
fuel cells, catalytic combustion
INTRODUCTION: THE CONCEPT
The storage of solar energy for residences makes special demands on safety and efficiency. Short
term storage is easy to realize with commercial batteries like the lead acid accumulator. But for long
term storage (for the winter), a small hydrogen system is the better choice. Hydrogen is a very
convenient energy carrier, comparable to natural gas. In contrast to batteries, any required storage
capacity might be realized without an increase in system cost in the same measure. The present
technology offers the possibility of:
- hydrogen generation by electrolysis of water (Fischer, 1989)
- hydrogen storage in pressure vessels or hydrides (Buchner, 1982)
- generation of heat by (catalytic) combustion of hydrogen (Ledjeff 1990)
- generation of electric energy by means of a fuel cell battery (Kordesch, 1984)
The task is the adaption of commercial devices to the special requirements, the setting-up of the
complete, automatically controlled storage system according to the safety regulations, the
minimization of the energy demand of the peripheral gas and water management systems and last,
but not least,the development of new, improved energy conversion devices.
The concept (Ledjeff, Heinzel, 1990) is a hybrid storage system (fig. 1) in the low power range. The
solar energy is distributed to the consumers, the lead acid battery and the pressure electrolyser. This
electrolyser generates hydrogen and oxygen under a pressure of 30 bar, both gases are stored in
pressure vessels. As required by the consumers, the hydrogen is fed to the cooking stove and the fuel
cell, respectively.
For the energy flowing along path 1, there are the losses of dc/ac conversion only. The 2. path is
calculated with an efficiency of 85 % for the lead acid battery. Cooking with hydrogen is possible
with 83%. The value corresponds to the conversion of solar generated current into hydrogen by

-------
2544
electrolysis. The consumption of electrical energy in winter via the fuel cell leads to the worst
95%
85%
55%
85%
electrolyser
catalytic
cooker
electrical
consumer
fuel
cell
dc/ac
converter
solar
cells
Lead
Acid Battery
Fig.l. The concept of the storage system
value for the efficiency of approximately 42 %. Depending on the consumption behaviour of the
occupants, an overall storage efficiency of 65 -70 % might be reached. In order to prove the
calculated data, a laboratory version of the storage system has been constructed and tested. The
following results base on experiences with this Iabscale system.
THE COMPONENTS OF THE STORAGE SYSTEM
The calculation of the power and storage capacity data for the laboratory system were done by
simulation program. In table 1, the data are compared to the new results for the energetically self-
sufficient solar house (Sick, GrieBhaber, 1991).
TABLE JJDMa.oith^StQrag.e..SysteiTLs
laboratory version (1987) 	solar house (1990)
consumption
electric energy
high temperature
heat
550 kWh/a
1,5 kWh/d
550 kWh/a
700 kWh/a (household)
1086 kWh/a (control systems)
4,9 kWh/d
1233 kWh/a
capacity of the lead
acid battery
10 kWh
20 kWh
nominal power of the
electrolyser
650 W
900 W
nominal power of the
fuel cell
200 W
600 W
tank volume
h2
2
4 m3
2 m
15 m3
7,5 m3

-------
2545
The electrolyser and the fuel cell were not commercially available in the calculated low power range.
Therefore, the devices already match quite well with the data for the energetically self-sufficent solar
house.
The Lead Acid Battery
A commercial lead acid battery with stirred electrolyte serves for short term storage of electrical
energy and buffer for high power. The nominal storage capacity of the test battery is 14,4 kWh. This
storage capacity meets the need of electric energy of the household for several days without solar
irradiation. The limit for the depth of discharge is 70%. Under these conditions, the cycle life of the
battery should be - according to the data of the manufacturer- 1500 cycles corresponding to a
theoretical lifetime of 25-30 years.
The Electrolvser
A specially designed alkaline electrolyser with reinforced cells normally is operated at 48 V, the
power is 850 W, the efficiency about 85%. 0,2 Nm of hydrogen and 0,1 Nm of oxygen are
generated per hour, the maximum hydrogen and oxygen pressure is 30 bar. The maximum current
which can be fed to the electrolyser is 40 A, the power then reaches 2 kW at 80°C and 2,5 kW at
20°C.
For the use in the solar house, some special adaptions of the commercial product are required. The
peripheral electrical parts (pumps, valves etc.) initially had an energy consumption of 60 W during
operation of the electrolyser. By replacing some of the electrovalves by pneumatic ones and
modifying the logic of the control the energy consumption could be decreased to 15 W.
But there is a second possibility. Instead of the commercial alkaline electrolyser, a construction with
a polymer membrane as solid electrolyte is preferred. No aggressive alkaline liquid is required in
this case, only pure, demineralized water is fed to that kind of electrolyser. Such an electrolyser for
operation under pressure was designed in the Institute. First tests were carried out in comparison to
the alkaline electrolyser and led to encouraging results. The power density is very high, the safety
standard is good. In future, the long term stability of the performance will have to be proven.
The Fuel Cell
In case of the fuell cell, only alkaline technology has been available up to now. Recently, first
membrane fuel cells were delivered for testing purposes (K. Prater, 1990). The performance of these
prototypes is convincingly good, but they are not generally available/and for solar application they
are still too expensive. Therefore, a commercial alkaline fuel cell battery was bought and tested.
The operation of this alkaline fuel cell revealed some problems: electrolyte creepage and difficulties
to realize an automatically controlled start-up procedure were the main items. The testing procedure
was stopped. The safety and reliability are not sufficent for a use in residences.
There are two alternatives for a possible solution. The first is to look for special-designed, non
commercial products. Two systems will be purchased in 1991. The energetically self-sufficient solar
house will be equipped with one of those fuel cell batteries. The second way is to start a research
program with the goal of developing membrane fuel cells for solar application. This long-ranged
research program requires cooperation with interested manufacturers. The current work is
concerned with membrane pretreatment, coating with catalyst, fabrication of membrane/electrode
composites and optimization of the operating parameters in a lab-scale fuel cell. The enlargement
and the construction of a stack is the future task.

-------
2546
The Catalytic Cooking Stove
There are two possibilities of generating heat for cooking purposes by burning hydrogen: The
combustion of hydrogen in a flame (app. 2000°C) or the catalytic combustion (400-800°C). The
advantages of a catalytic cooking stove are:
- the heat losses are reduced
- the formation of poisonous waste gases (NOx) is nearly completely avoided.
The range of temperature is determined by the catalyst system. A ceramic carrier material with
noble metals as catalyst operates at 400-600°C and a power of 4-6 W/cm . At noble metal catalysts,
no ignition is required for burning hydrogen/oxygen mixtures. With a porous sinter metal, 700-800°C
are reached and a power density of 15-20 W/cm . In this case, an ignition is necessary like for
conventional natural gas burners, too. Cookers with hotplates using both catalytic techniques are
presently under construction.
Results of Laboratory Test Operation
The laboratory test operation is important in order to gain informations on the reliability and
service life of the devices, the correct and safe management of the system by an automatic control
unit and the true efficiency of the storage system. For the calculation of the overall efficiency, the
energy consumption of the control unit is taken into account.
In summer, charging of the storage system is the main process. The excess solar energy is fed to the
battery and the electrolyser in parallel. The battery is charged with priority. Both very low and very
high solar currents will be fed to the battery because of the restrictions of efficient electrolysis to
medium power.
In winter, the discharge process means consumption of electrical energy from the solar generator,
the battery and, in part, from the fuel cell. This part will have the greatest influence on the system
efficiency and should be used to a minimal extent. This is possible if the consumers are willing to
control their consumption carefully in times of deficient insolation. Therefore, a realistic operation
in a solar house with occupants is necessary in order to gain realistic informations.
SUMMARY
The solar house is a demonstration of technical realizable solar systems (Stahl, Goetzberger, 1990).
From the present point of view, the high investment costs and the poor availability of the energy
conversion devices restricts the use of solar energy to special applications. It is a challenge for our
future work to construct energy conversion devices which are suitable for solar systems.
LITERATURE
Buchner, H. (1982). Energiespeicherung in Metallhvdriden. Springer Verlag, New York.
Fischer, M. (1989). Wasserstoff-Technik. Chem. Ing. Tech. 61,124.
Kordesch, K. (1984). Brennstoffzellen. Springer Verlag, Berlin.
Ledjeff, K. (1990). New hydrogen appliances. Proc. 8th World Hydrogen Energy
Conference. Hawaii, USA.
Ledjeff, K. and A. Heinzel (1990). Energy storage system for solar houses. Proc. 8th World
Hydrogen Energy Conference. Hawaii, USA.
Prater, K. (1990). The renaissance of the Solid Polymer Fuel Cell. J. Power Sources.
29, 239.
Sick, F. and W. GrieBhaber (1991). Das Planungsinstrument Computer-Simulation am Beispiel der
PV-Anlage im Energieautarken Solarhaus. 6. Nationales Symposium
Photovoltaische Solarenergie. Staffelstein.
Stahl, W. and A. Goetzberger, (1990). Das energieautarke Solarhaus. Sonnenenerpie. 4 10.

-------
2547
THE SELF-SUFFICIENT SOLAR HOUSE
ELECTRICAL CONCEPT
Georg Bopp
Fraunhofer-Institut fur Solare Energiesysteme
Oltmannsstr. 22, D-7800 Freiburg i.Br.
Tel.: 49-761-4014-0
ABSTRACT
At present, the Fraunhofer Institute for Solar Energy Systems is planning the construction of an ener-
getically self-sufficient solar house which will use solar radiation as its only energy source
(Goetzberger; 1991; Stahl, 1991). The electrical system of the self sufficient solar house consists of
two main parts, the electrical power system,and the control and supervisory system to control and
monitor the whole house. A PV-generator of approximately 4 kWp provides all the electrical power
for appliances and the control system and^indirectly>hydrogen gas for cooking. For short-term energy
storage there is a lead-acid battery of approximately 20 kWh. For long-term storage (summer to win-
ter), an electrolyser is used, which produces hydrogen and oxygen. In winter the gases are either
transformed into electricity for electrical appliances by a fuel cell or used directly for cooking.
The household is equipped with energy-efficient electrical appliances (lighting, washing machine,
dishwasher, TV set, etc.) for 220 V AC. All appliances must be commercially available and are cho-
sen according to the criterion of low power consumption.
The house has neither an electrical connection to the grid nor an external gas supply.
KEYWORDS
Self-sufficient, solar house, photovoltaic, hydrogen, oxygen, fuel cell, electrolyser, efficient electrical
appliances.
INTRODUCTION
The electrical system of the self-sufficient solar house consists of two main parts, the electrical power
system and a control and measurement system to regulate and monitor the whole energy supply sy-
stem.
APPLIANCES AND ENERGY CONSUMPTION
A block diagram of the electrical power system is given in Fig. 1. The photovoltaic generator supplies
the electrical energy to the consumer via an inverter. A lead-acid battery is used for short-term

-------
2548
storage. The long-term storage consists of hydrogen and oxygen tanks together with an electrolyser
and a fuel cell (Ledjeff, 1988). The household is equipped with energy-efficient electrical appliances
(lighting, washing machine, dishwasher, TV set,etc.) using 220 V AC. All appliances have to be
commercially available and are chosen according to the criterion of low power consumption.
Using these appliances and the warm water heated by the solar DHW system (for the washing machine
and dishwasher), it is possible to reduce the electrical power consumption to 700 kWh/a for a four
person household, as compared to 3270 kWh/a in a conventional system (see Table 1).
220
VAC
BZ
ZS
Fig. 1. Block diagram of the electrical power system
Ely: Electrolyser, BZ: Fuel cell, ZS: Control system
TABLE 1 Electrical Consumption of the Self-Sufficient Solar House (Pour Person Household)

Self-sufficient
German

solar house
average (1ZE, 198S

[kWh/a]
[kWh/a]
Lighting
88
380
Refrigerator
110
530
Freezer
110
780
Washing machine
146
380
Dishwasher
62
380
TV
28
220
Small appliances
155
600
Total
699
3270
In addition to the energy consumed by household appliances, further energy must be supplied, as ite-
mized in Table 2. Thus, in order to cater for the complete power supply for the house, including coo-
king, additional heating, supplementary heating of warm water, control and measurement and the po-
wer consumed by the hydrogen-oxygen components, the photovoltaic system must supply much more
energy than is required for running purely electrical household appliances.

-------
2549
TABLE 2 Itemization of the Annual Energy Demand which must be met hv the Photovoltaic System
AC electricity	700 kWh/a	Household
Hydrogen	300 kWh/a	Heating
233 kWh/a	Hot water
700 kWh/a	Cooking
1233 kWh/a
DC electricity
51 kWh/a
134 kWh/a
701 kWh/a
56 kWh/a
144 kWh/a
1086 kWh/a
Electrolyser consumption
Fuel cell consumption
Control and measurement
Collector pump
Ventilation fans
Total consumption
3019 kWh/a
A simulation model and program were set up for the whole house,including all system components, in
order to dimension these components and determine the energy values as given in Table 2 (Sick,
1991). The result of these simulation calculations is summarized in Table 3.
TABLE 3 Characteristic Data for the System Components
Photovoltaic generator:
Total area of the PV generator:
Lead-acid battery:
Hydrogen storage tank:
Oxygen storage tank:
Alkaline pressurized electrolyser:
Alkaline fuel cell:
108 PQ 36/45, 4 in series, 28 in parallel
48,6 m2
48 V, 450 Ah
15 m3, 30 bar
7,5 m3, 30 bar
2 kW peakpower, 27 cells in series, 48 V
800 W peak power, 16 cells in series, 12 V
ELECTRICAL POWER SYSTEM
As shown in Fig. 1, the electrolyser is connected directly in parallel to the solar generator. Switch Sj
is a rapidly switching transistor switch. The distribution of the solar electricity between the electroly-
ser and the battery, depending on its charging state, can be controlled by the ratio of the pulse to the
pause. Figure 2 shows the characteristic curves of the electrolyser for different operating temperatures
in the same diagram as the characteristic curves of the solar generator for different insolation values
and temperatures.

-------
2550
Hg-production 1/h
ELY
U,
100
200
300
'ELY"
59,4
. 60
MPP
parameter
TELY
MPP
56,7
,.C
• 55
54,0
' 50
48,6
45,9
¦45
43,2
40,5
-40
0
10
20
30
40 Iely/A
Fig. 2. Characteristic curve of the electrolyser (27 cells) and of the solar generator
(PQ 36/45, 4 in series, 12 parallel)
The intersections represent the operating points which become established when these two components
are running parallel. For moderate to high insolation values, it can be seen that the solar generator and
the electrolyser are matched to 90 % or more. The electrolyser is usually switched off during periods
of low insolation by active solenoid values because of its own power losses, so that the solar electri-
city then flows directly to the consumers and into the battery. Thus there is no need for a direct cur-
rent transformer to match the solar generator to the electrolyser, which would have an efficiency value
of at most 90 % if the open circuit losses were taken into account. The parallel circuit for the solar
generator and the electrolyser is also in good agreement with the experience and simulation calculati-
ons made in the course of the Hysolar Project at DLR in Stuttgart (Brenner, 1989)
Switch S0 is a shunt controller which is active when both the gas tanks and the battery are completely
full. The diode between the battery and the electrolyser prevents energy from flowing from the battery
to the electrolyser.
The fuel cell which compensates for the deficit in the solar generated electricity during winter using
the stored gases must be connected via a direct current transformer because fuel cells with the relati-
vely low power of 800 W are available only with an operating voltage of 12 V max. As the fuel cell
has a maximum efficiency of 55 % it is important that the direct current transformer used be as effi-
cient as possible so that the fuel cell performance is not reduced further. The best efficiency value for
commercially available 12 V/48 V direct current transformers is 80 %, so a specially developed model
was commissioned which achieved an efficiency value of more than 90 %
The pressure electrolyser and the fuel cell have already been set up in the laboratory with alkaline
technology. The use of the solid polymer electrolyte (SPE)1 is planned for the self-sufficient solar
house (Heinzel, 1991).
1 SPE is a registered trademark of the company, Hamilton Standard, USA

-------
2551
To convert the direct current from the photovoltaic system to alternating current; an inverter with very
low open circuit losses must be used, as the average power consumption in a household is only 1I25 of
the peak power (Schmid, 1987).
When the electrical circuit in the hydrogen-oxygen system is installed the guidelines to reduce explo-
sion hazard must be observed. The regulation IEC 31 (CO) 43 (Electrical installations in explosive gas
atmospheres) specifies the installation of an isolated circuit with an isolation monitor to identify short
circuits through body contact or to the ground.
CONTROL SYSTEM
A computer operated control and data acquisition system is needed to regulate and measure the whole
energy system, as illustrated in Fig. 3.
Self - Sufficient
Solar- House
measurement
\data
dato
aquistion
dato aquistion unit
process
electroly.
fuel cell
neating
water
control
system
laptop
control
data
Fig. 3. Control and data acquisition system for the Self-Sufficient Solar House
Each system has its own CPU so that the control can operate independently of the data acquisition, but
the two CPU can exchange data via a serial interface. An independent battery supplies the power, as
shown in Fig. 1. As almost all control systems and components,e.g. solenoid switches,are available in
24 DC versions, 24 V DC was chosen as the voltage level for the control system. The control and
data acquisition system, including two LCD control monitors, has a power consumption of only
80 W, but this results in an energy consumption of 700 kWh/a for continuous operation. This is the
same energy requirement as for the whole household, and must also be supplied by the photovoltaic
generator. By using self-regulating system components,the extent of the automization should be kept
as small as possible.

-------
2552
CONCLUSION
It is possible to build a completely solar-powered house which does not need additional energy such as
grid electricity or fuel. Using the hydrogen-oxygen system, surplus PV energy from the summer can
be stored for the winter. The fuel cell has an important role in converting the hydrogen and oxygen to
electricity.
REFERENCES
Brinner, A. and A. Siegel. (1989). Operation of a PV Electrolysis System with different coupling mo-
des. Ninth E.C. Photovoltaic Solar Energy Conference (Treiburg).
Goetzberger, A. and W. Stahl. (1991). The Self-Sufficient Solar House Freiburg. To be published in
the proc. of this conference.
Heinzel, A. and K. Ledjeff. (1991).The Self-Sufficient Solar House Freiburg: Hybrid Energy Storage
System. To be published in the proc. of this conference.
IZE (1988). Informationszentrale der Elektrizitatswirtschaft (Frankfurt).
Ledjeff, K. (1988). Wasserstoffsysteme als Energiespeicher fiir PV-Anlagen. 6. Int. Sonnenforum
(Berlin), 669-674.
Schmid, J. (1987). Photovoltaisches Wechselstromsystem fur die Energieversorgung. ETZ. 108.
1076-1079.
Sick, F. and W. GrieBhaber. (1991). The Self-Sufficient Solar House: Remarkable Simulation Results.
To be published in the nroc. of this conference.
Stahl, W. and W.S. Wilke. (1991). The Space Heating Concept of the Self-Sufficient Solar House
Freiburg. To be published in the nroc. of this conference.

-------
2553
ENERGY SELF-SUFFICIENT SINGLE FAMILY HOME
T. R. McBride
Southern Enviro-Shelter R. & D.
A non-profit Subsidiary of SOL-ERA Inc.
Route 2 Box 127, Jacksons' Gap, Alabama 36861
ABSTRACT
In this study, basic alternative energy producing equipment is tested in passive-active structures.
Based on the average consumption levels of an energy-conservative one family dwelling. (See fig.
1) Peak energy input averages are gathered during seasonal intervals from each device. During the
favored season of each device strategical experiments are done to increase maximum energy
production.
KEYWORDS
Structurally Integrated energy producing systems; Immediate environment control; Energy Con-
venient efficiency; Localized supply-demand equivalency matching.
INTRODUCTION
The purpose of this experiment is to investigate the abilities of specialized energy conscience
structures; to supply all energy needed co-existing with human behavior patterns. Using the natural
environment in and around the structure as the main source of energy input while keeping the
equilibrium of life sustaining substances intact.

-------
2554
Energy Self-Sufficient Single Family Home
IW 16
14
1.2
10
08
06
04
0.2
A

k
12
1.0
08
0.6
0.4
02
17.97 kWH/OAY
A
VI
InW
14 2 IWH/OAY
UM66ION 2 4 6 8 I0M
k W 1.6
14
1.2
1.0
0.8
0.6
0.4-
0.2
IW 16
M 2 4 6 6 ION 2 4 6 6 I0U
FALL
p n
19.00 kWM/DAV
• VACATION 80S bWH/OAY
« ;
i-ijftf
KU
IB.5B IWH/DAY
•— WEEKEND 16.75 IWH/OAY
M2 4 I tlON 2 4 i 9 IOM
HOUR or MY
U24CII0N24CSI0M
HOOK OF DAY
Fig. 1. Seasonal characteristics of Diversified Electrical Loads Daily
Profiles. These four charts show the pattern of energy use for a family
of four living in an all-electric three-bedroom home using 5540 Kilowatt
hours per year. (Courtesy DOE)
In genera], using the electrical loads daily profiles as a guide, the performance of the following
electrical generating devices are shown to co-relate with patterns of natural occurring environmental
phenomenon.
100 WATT Photovoltaic (P.V.) passive sun tracking D. C. generating system. The P. V. Solar cells
produce electricity year round. The best season is between March 20 - September 23 intensifying
during summer. Cells produce maximum energy during the hours of 9 a.m. - 4 p.m. on sunny days.
Energy production is highest during low demand use hours. This device is excellent for charging
deep cycle batteries for demands of DC or inverted AC circuits during peak energy use hours.
Notice Fig. 2: The P.V. average daily energy characteristics is superimposed on summertime use
pattern.
Summer
400
Wails
300
100
50
0
M2468 I0N2468 10 M
Fig. 2. Relationship of Photovoltaic output to pattern of energy use
on a sunny day. The output can be adjusted to actual demand
usage by adding more P.V.'s to increase desired current.

-------
Energy Self-Sufficient Single Family Home
2555
5 WATT thermo couples attached to a intrepid wood parlor stove. The thermo couples begin
producing a D.C. current when stove reaches a surface temperature of 250° F. The current is used
to maintain a battery charge to operate small D.C. equipment of a separate circuit. The stove is also
equipped with water heating thermosiphon circulator coil that subsidizes a flat plate solar water
heating collector. During the winter season when outside temperature is lowest and use pattern are
highest the extra space heating from carefully selected hardwood help the demand levels lower on
other system components. Figure 3 illustrates this relation superimposed over the wintertime use
pattern.
WINTER
3 WATTS
STOVE
SURFACE
temp. M 1 4 « 8 10 N
—THERMO COUPLE OUTPUT SURFACE TEMP. -
THERMO COUPLE OUTPUT HOURS
I I I f __
4 6 8 10 M TEMP.
• WATER TEMP. FROM THERMO COILS
' WATER TEMP. FROM SOLAR COLLECTOR
- WINTER ELECTRICAL USE PATTERN
Fig. 3. Related patterns based on a typical winter day using 25 pounds
wood fuel with outside temperature averaging 40° refueling at 5:30 a.m.
and 4:00 p.m. This method will maintain a well insulated 60 gallon tank
at a average temperature of 115° F.
200 WATT Hydro-electric alternator is dependent on the flow of water through the site or heavy
rainfall in the guttering system. The amount of water falling is measured in gallons per minute or
running pounds of pressure. This amount will vary greatly from different sites. There will be
variation from dry to wet weather seasons. Figure 4 will illustrate the varied electrical supply in
amps with the use patterns of Fall which is usually the driest season.
Ouipul 13
in amperes 12
11
10
9
a
7
Fall
KLjs
^3
p n
4—i*
35 Running pressure
50 using 1/2" notde
45
40
35
30
25
20
M 2468 10 N2468 10 M
Hour of day
Fig. 4. Water falling with a running pressure of 40# will yield a steady
24 hour electrical charge to the battery bank at 10 amps.

-------
2556
Energy Self-Sufficient Single Family Home
90 WATT wind charger will begin charging batteries with a 8 mph breeze. Wind power comes in
many variations such as windspeed, direction, temperature, height, gusty, or smooth. This of course
changes with the seasons making wind energy hard to predict. In Fig. 5 the task is made somewhat
easier and shows some energy characteristics of a typical windy spring day.
Watts
200
170
140
120
90
30
10
0
0
0
Spring
/I
£UW J
36
32
28
24
20
16
12
8
4
0
M 2 4 6 8 10 N 2 4 6 8
Hour of day
10 M
Wind Speed (mph)
Fig. 5. Air flow is less turbulent and more steady 5 feet above tree line.
Also a good tower can be used to mount P. V.s, water tank, solar
collectors, and communication devices, etc.
Secondary automotive alternator battery charging systems are used year round for storing excess
electricity while commuting. The energy is transferred to dwelling by means of a special connector
is garage area. The device will not drain cranking battery. Energy is on demand during high load
hours in sequence with at home use patterns. Figure 6 illustrates the storage amount in ampere
hours during a given milage trip.
Miles 50
45
40
35
30
25
20
15
10
20 30 40 50 60 70 80 90 100 110 120
Ampere hours
Fig. 6. Every 50 miles driven in a average car will supply approximately
120 ampere hour to the deep cell battery on board.

-------
Energy Self-Sufficient Single Family Home
2557
Methane powered motor driven alternator is a unique way to disperse organic waste. Twenty pounds
of dry matter which consists of dry leaves, vegetable scraps, straw, and activated sludge will yield
250 cubic feet of biogas. This gas can be used to run engines, cook, light, refrigerate, boil water,
etc. The gas produced when sludge is piped underground to a storage digester for on demand fuel
use when needed. Figure 7 illustrates the consumption rate per hour with various burn tasks.

Gas Consumption
Task
Spec.
Cu. Ft. per Hour
Cooking
2" dia. burner
11.5

4" dia. burner
16.5

6" dia. burner
22.5
Refrigerate
18x18x18
1.5 to 2
Boil Water

10 per gallon
Run Engines

(16-18 per horsepower)
per hour
Fig. 7. A 1 horsepower stuart engine will consume 18 cubic feet biogas
per hour coupled with a 250 WATT D.C. 110 volt generator. The generator
will produce 225 WATT while running on biogas. Fig. 7A shows generator
set on automatic demand during high use pattern periods.
OUTPUT
WATTS
1500 —
1250 —
1000 —
750 —
500 —
250 —
100 —
0 —

\
i
i i i i i i i i i i i i
M2468 10 N2468 10 M
Generator running time
• ••••••••• Consumption rate
. Winter Electrical Use Pattern
—180
—144
—108
— 72
— 36
— 0
Cu. Ft.
Per Hr.
Fig. 7A. Generator peak demand back up with use pattern.
In summary, weighing the balance of present times and events, with increasing evidence of the
degradation of the natural environment. In fact, the repercussions from unpredictable cause and
effect responses from the eco-system is an unstable environment. The massive dwindling of non-

-------
2558
Energy Self-Sufficient Single Family Home
renewable, terribly polluting energy resources has already reached to levels that ecological com-
patible energy must be established to maintain sustainability. Therefore, the ongoing success of
this experiment, and many other s similar around the globe, is a growing testimony that the age
of restoration is upon each of us. With this, the ultimate responsibility for the wholeness of the
immediate environment, is truly in our own hands.
ACKNOWLEDGEMENT
The author would like to express great gratitude and appreciation to all Pioneers in Environmen-
tally sound energy research and development programs. With special recognition to authors,
Joel Davidson, David Morris, Robert Rodale, Donald Marier and many others who have helped
to open the way to a much brighter, cleaner, healthier environmentally responsible society.
REFERENCES
Davidson, Joel, and Richard Komp, (1983). The Solar Electric Home: A Photovoltaics How-To
Handbook.
Davidson, Joel, (1987). The New Solar Electric Home: A Photovoltaics How-To Handbook.
Morris, David J., (1983). Be Your Own Power Company: Selling and Generating Electricity
from Home and Small-Scale Systems, Photovoltaics, Wind power, Hydropower, Cogeneration.
Morris, David J., (1979). Planning For Energy Self-Reliance: A Case Study of the District of
Columbia.
Morris, David J., and Karl Hess, (1975). Neighborhood Power: The New Localism.
Marier, Donald. (1981). Wind Power for the Home Owner: A Guide to Selecting, Siting, and In-
stalling an Electricity-Generated Wind Power System.
Maycock, Paul, and Edward N. Stirewalt. (1981). Photovoltaics, Sunlight to Electricity in One
Step.
Rodale, Robert. (1981). Our Next Frontier: A Personal Guide for Tommorrow's Lifestyle.

-------
2559
THE SELF-SUFFICIENT SOLAR HOUSE:
REMARKABLE SIMULATION RESULTS
F. Sick, W. Griefihaber
Fraunhofer-Institut fur Solare Energiesysteme,
Oltmannsstr. 22, D-7800 Freiburg i. Br.
ABSTRACT
Computer simulations were a major design tool during the planning phase of the self-sufficient
solar house, a project carried out by the Fraunhofer Institute for Solar Energy Systems in
Freiburg, Germany. The simulation program TRNSYS (Klein, 1988) was used to model and
simulate the complex energy supply system of the house that includes photovoltaics, lead/acid
batteries, a long-term hydrogen storage with electrolyzer and fuel cell, inverters, controls and
transparent insulation. The paper describes remarkable results of the simulation of the self-
sufficient solar house.
KEYWORDS
Computer simulation; TRNSYS; self-sufficient house; low-energy building.
INTRODUCTION AND APPROACH
The Fraunhofer Institute for Solar Energy Systems in Freiburg is building an exclusively solar
energy-supplied 1-family-house (Goetzberger, Stahl, 1991). The energy concept for the "Self-
Sufficient Solar House" is shown in Fig. 1. Wall heating with transparent insulation will provide
almost all of the heating demand. High efficiency solar collectors supply more than 90 % of the
domestic hot water (DHW). Electrical energy is entirely produced by PV modules and buffered in
batteries. A hydrogen/oxygen system serves as long-term storage.
During the planning phase computer simulations were a major design tool. They were necessary
for
- the comparison of system component variations,
- the evaluation of auxiliary energy requirements of the subsystems,
- the resulting short-term and long-term storage capacities needed,
- the correct sizing of the system components and
- the investigation of influences like weather conditions and user behaviour.

-------
2560
TIM Thermal Collectors PV Panels
|in
Ambient Air

Battery Controls Data
Acqu.
Battery
Fuel
Cell
Catalytic!
..-¦J
DC/DC Converter
Storage
Tank
optional:
r'T"""1
DHW
Cooking
— Electricity
— Gas
Exhaust Waste Cold
Air Water Water
Fig. 1. Schematic of the energy supply system in the self-sufficient solar house in Freiburg
The basis for the simulation of the total system is the modular simulation program TRNSYS
(Klein, 1988). It allows the connection of many single components to a complete energy supply
system. Another reason for choosing TRNSYS is the possibility to include user-written
components. For the simulation of the self-sufficient solar house, several new components had to
be developed: an electrolyzer, a fuel cell, a pressurized tank for gas storage, a DC/DC and
DC/AC converter, a catalytic heater, and a central control unit. In addition, the TRNSYS
lead/acid battery model was modified and a new PV array module was written. The single
components were connected to a comprehensive computer model of the energy supply system in
the self-sufficient solar house. Using this model as a "tool", the authors interpreted the simulation
results in an interactive way, decided on system modifications and repeated - if necessary - the
"experiment" on the computer. Working in this manner, it can be avoided, for instance, that
unnecessary surplus energy is produced. An optimum control strategy can be developed and the
influence of single parameters can be investigated. In the case of the self-sufficient solar house, a
lot of knowledge on the energy-related behaviour could be gained long before construction began.
Thus a number of otherwise possible mistakes can be avoided without causing additional costs.
RESULTS
One of the most interesting questions concerning this system is the capacity of the two storage
devices, i.e. the battery capacity and the volume of hydrogen and oxygen tanks. Converted to kWh,
the lead/acid battery should be sized to have a 20 kWh capacity and the gas storage should hold
1436 kWh of energy. Figure 2 shows the state of charge of both storage devices for the simulated
typical year. (Note that the simulation does not begin in January but in April in order to get
realistic starting values for the critical winter period. The same applies to all the following results
shown.)

-------
2561
Battery
Hydrogen Storage(U36 kWh)
( 20 kWh)
en 0.9
cn
5 0.8
o 0.7
° 0.6
-------
2562
X////i PV-Generator U705 kWh)
Totoi Load (326£ kWh/a )
AC Load ,( 701 kWh/a)
_ DC Load 11329 kWh/a)
~1 Hydrogen Load (123A kWh/a)
pip
-100-
NOV DEC
i i r
APR MAY JUN
FEB MAR
Fig. 3. Monthly sums of PV energy production (pos. values) and the several loads (neg. values)
An optimization would thus be most efficient if these loads could be reduced. One idea is to use
the waste heat of the fuel cell. The operating temperature of the fuel cell is 70°C and it must be
actively cooled in order to maintain this temperature. The operating times of the fuel cell should
correlate well with the times at which increased auxiliary heating energy is needed. One does not
need a big and expensive computer to realize this fact at least in a qualitative manner. However,
with the help of a computer, one can prove and quantify the correlation.
45
40
35
2 30
$
- 25
oi 20
-------
2563
Figure 4 shows the result: the left bars represent weekly sums of the energy that is required for
DHW auxiliary heating. These numbers were determined by a separate TRNSYS simulation. The
bars on the right hand side show the waste heat to be taken from the fuel cell. The very good time
correlation is significant. Summing up the numbers, the fuel cell could theoretically, i.e. assuming
100 % efficiency of the heat recovery system, provide the total auxiliary DHW heating energy.
Calculating with a 50 % efficiency, still roughly 60 % of the required auxiliary heating energy for
DHW could be provided.
PV Gen. Losses Losses Aux. Energy
[kWh] [kWh]
Irrodsotion
IkWh]
636
217
1831
Storage
1103
738
H? Cooker
"337"
233,
321.
DHW
Cooking
612,
1816
276
336,
272
118? j"
BoMery
192
906
986
Fans
| Pump
701
806
j Inverter
105
701.
1086
Sum
1365
Heating
Fig. 5. Energy flow chart of the self-sufficient solar house (annual sums)
From Fig. 5 one can see where the energy produced by the PV generator flows. The yearly sums
are shown. There is no relevant surplus energy which means that the storage devices are correctly
sized. The numbers in the two columns on the right hand side indicate weak parts of the system:
-------
2564
unit amounts to a high value relative to the entire load, although a control system with only 80 W power
consumption has been chosen. (However, it must be stated here that the control system includes the data
acquisition unit which is needed for research reasons only.) 54 % of the PV cell area are used only to
cover the losses of the system components - not counting the PV generator - and the internal consumption
of the controls. A lot of work on the system details has to be done in order to improve this percentage. It
is one of the goals of the project "Self-Sufficient Solar House" to provide improved components and
techniques to the public for more conventional low-energy applications.
REFERENCES
Klein, S. A. and others (1988). TRNSYS - A Transient System Simulation Program. Engineering Exp.
Station Report. 38-12. University of Wisconsin, Madison.
Goetzberger, A. and W. Stahl (1991). The Self-Sufficient Solar House Freiburg. To be published in the
Proc. of this Conference.
-------
3.3 Emerging Architecture
-------
Intentionally Blank Page
-------
2567
ADVANCED HOUSE CONCEPT
J. Douglas Balcomb
Solar Energy Research Institute
1617 Cole Boulevard, Golden Colorado, USA
ABSTRACT
A concept for an advanced passive solar house has been developed in conjunction with the
International Energy Agency Solar R&D Task 13. The house design integrates good
insulation using high-R stress-skin insulation wall and roof panels, good air-tightness
combined with heat recovery, a large semi-enclosed sunspace with advanced glazing for solar
gain, and a ducted floor slab for thermal storage to achieve excellent performance. A detailed
simulation model is used to predict annual auxiliary heat. The estimated solar heating
fractions are 98% in Denver and 73% in Hartford, corresponding to annual auxiliary heat
values of 1.2 kWh/m2 and 14.3 kWh/m2 respectively. The ducted slab floor heat storage is
effective for improving comfort in Denver Dut is not warranted in Hartford. This design
concept may prove to be suitable for many United States locations with cold winters and mild
summers, including much of the northern tier of states.
KEYWORDS
Advanced house, simulation analysis, IEA, passive solar, conservation, thermal storage.
BACKGROUND
The International Energy Agency (IEA) Solar R&D Task 13, in which 12 IEA countries
participate, is called Advanced Solar Low-Energy Buildings. The objective is to investigate
the implications of new technical developments on potential performance of housing in the
long term, that is, 10 to 20 years into the future.
Based on the experience of the last 15 years, we know that it is possible to construct houses
that require very little energy to maintain comfort. The advent of new materials, such as
super windows with loss coefficients less than 1.5 W/°C-m2, should only improve the
situation. The IEA study seeks to probe the limits. This paper presents the results of a
United States preliminary advanced house concept.
The house concept was developed for Denver, which is located in a cold, sunny climate where
summer cooling is not a primary concern. The results will thu9 be more comparable to those
achieved by other IEA countries than, for example, to results for a house designed for a
south-eastern United States location.
SPECIAL SITE FEATURES AND CLIMATE
The primary site for the house is slightly west of Denver, Colorado, near the eastern base of
the Rocky Mountains at an elevation of 1830 m (6000 ft) above sea level and a latitude of
40°N. The house was also simulated using weather data from Hartford, Connecticut, which is
at an elevation of 51 m and a latitude of 42"N. The site is unobstructed and slopes to the
south. The entrance is on the northeast, the living areas are on the upper level, and
bedrooms on the lower level.
-------
2568
The Denver climate is arid with, mild summers, cold winters, and appreciable snowfall, which
occurs typically in three-day storms. Between storms, the weather is generally very sunny
and the snow melts. Denver weather data were used for the analysis. Heating degree days
are 3390 "C-days (6102 °F-days), calculated on an 18.3 *C (65°F) base. The recorded extreme
low temperature is -32 'C (-25 *F) and the extreme high is 39*C (103*F). Average winter solar
radiation is about 63% of extra-terrestrial values. Temperature data for Hartford are very
close to those of Denver, but the annual solar radiation transmitted through south vertical
double glazing is only 52% of that in Denver.
HOUSE DESCRIPTION
The two-story house has a large sunspace that serves as the primary passive solar collection
element. House plans are shown in Figs. 1 and 2. South elevation is shown on Figs. 3. The N-
S section is on Fig 4.
Krtcnen
Deck
Balcony
Sunspace
Upper Level
Home.
Office/
Bedroom
Bedroom r u
Sunspace
Master
Bedroom
Lower Level
Fig. 1. (left): Floor plan of the upper level of the house. The living room, dining room, and
kitchen are on this level. The two-car garage is located on the northwest., The balcony
overlooks the sunspace, which is two stories tall. Floor-to-ceiling glass windows allow view
through the sunspace to the south. Fig. 2 (right): Floor plan of the lower level of the house. The
entire floor is concrete slab on grade. The portion under the house is an air-floor construction
covered by carpet pad and carpet. The return from the air-floor enters the sunspace up
through a triangular section that extends over the air-floor at the north, corner. The sunspace
floor is brick or tile mortared directly to the slab.
Fig. 3. (left). South elevation, showing
the sunspace glazing.
Net house floor area
165
m2
Gross house floor area
2046
ft2
Sunspace area
41
m2
Sunspace area
463
ft2
Gross volume
1003
m3
Gross volume
35416
ft3
Window area
53
m2
Window area
567
ft2
South window
31
m2
South window
334
ft2
-------
2569
Fig. 4 (right). North-south section,
c ut through the junction of the brick
walla that separate the sunspace fron
the house. This emphasizes the bridg.
that connects the two internal
sunspace balconies.
Sunspaca
Lower Level
Airfloor
North-South Section
ENERGY FEATURES
Stress-skin panels are used for the exterior walls and roof to give good insulation and a very
strong air-tight construction. Wall panels are 20 cm (8 in.) thick resulting in U = 0.19 W/C-
m2 (R-30). The roof panels are 25 cm (10 in.); insulation U-value is 0.15 W/C-m2 (R-38). All
footings and the back wall of the lower level (which is dug into the hillside) are insulated on
the exterior with 5 cm (2 in.) of high-density polystyrene foam. Good sealing of all joints
reduces the natural infiltration to 0.2 ACH. However, half of this air is drawn through the
sunspace and therefore does not tend to cool the house. A heat recovery unit is installed to
assure adequate air quality. House windows have wood frames with low-e coated films and
krypton gas fill to achieve U = 0.71 (R-8) in the center and U = 1.14 W/C-m2 (R-5) overall.
Sunspace windows are quad-pane with the two inner glazings of Teflon and an argon gas fill
to achieve high transmittance and good insulating value. The building loss coefficient, BLC,
is 89 W/*C (169 Btu/hr-"F) for the house and 103 W/°C (196 Btu/hr-*F) for the sunspace.
Primary solar heating is passive with collection in the sunspace and distribution to the house
by natural circulation. Air wanned by excess solar heat is re^aimed from the sunspace and
transferred by a fan through the hollow concrete-slab floor of the lower level, a sort of
hypocaust, returning to the sunspace. The sunspace construction includes some added mass
to store heat. The concrete slab floor has a tile covering, and a brick wall separates the house
from the sunspace.
Long-term heat storage is provided within the slab floor of the lower level and underlying
earth and in the air-floor. A circulation fan draws 20 m3/minute (700 cfm) of air from the top
of the sunspace, forcing it horizontally through the air-floor with return to the sunspace.
The lower 1 m of the sunspace south wall is a 20-cm (8-in.) unvented Trombe wall with a
single water-white glazing and a selective-surface metal foil adhered to the outside surface of
the concrete-block wall.
Natural ventilation through the sunspace is provided for by side-window inlets and awning
windows at the top. Windows provide for cross-flow natural ventilation through the house.
Provision to manually vent the sunspace is made in the design, using low side windows to
draw in cool outside air and high windows on the NW and NE to release hot air. The exhaust
windows will draw very well because of the venturi effect caused by the southwest prevailing
breeze blowing over the roof; they are equipped with motor-powered openers.
The system employed is air-floor, a patented design that uses metal forms to create series
of domed volumes between a lower concrete slab and an upper slab that is supported by
pillars 30 cm (12 in.) on center in each direction. The overall floor is 15 em (6 in.) thick. The
floor rests directly on the ground, which adds greatly to its heat capacity.
-------
2570
SIMULATION MODEL
A 30-node hour-by-hour simulation model has been developed to evaluate the unique features
of the house and accurately characterize such nonlinear behavior as zone-to-zone convection
and the air-floor. Each of the two levels and the sunspace is modeled as a separate zone
because they behave very differently and only the lower level is strongly coupled to the mass
of the air-floor and the north massive wall.
The air-floor is modeled as six horizontal linearly connected nodes with conduction both up to
the lower slab and down to the ground. The earth is modeled as several massive nodes in
1 h6 regions—back of the lower level, under the air-floor, and under the sunspace
SIMULATION RESULTS
Performance during an 8-day period of severe weather from 1 through 8 January is shown in
Fig. 5. The previous few days are sunny and 1 Jan. is the last day of a recurring cycle A
snowstorm occurs on 2 Jan. extending through 3 Jan. Clearing occurs on 4 Jan and the
ambient temperature drops to -24°C (-12 F). The average ambient temperature over the 8
days is -6 C (22 F).
Dog. C
SUNSPACE
UPPER LEVEL
32.2
29.4
LOWER LEVEL
23.a
RIR-FLOOR CHARGING
BACKUP HEAT
FAN OPERATION
Fig. 6. Simulated
performance during the first 8
days of January. This includes
the coldest weather on the
TMY record. Prior days are
fairly sunny. A typical winter
snowstorm moves through
quickly on 2 January. Average
ambient temperature during
these 8 days is -6 V and the
minimum is -24 "C on the 4th.
The fan operates 6 to 8 hours
on each of the last five days,
reducing sunspace
temperatures and increasing
the air-floor temperature.
Internal gains from people, appliances, and lights average 439 W and are distributed over
the day according to a typical residential time-of-day profile (36000 Btu/day). The thermostat
for auxiliary heat in the upper level is set at 20"C (68 F). Auxiliary heat for this 8-day period
is 60 kWh. Maximum auxiliary heat is 2.2 kW on 4 Jan and the maximum daily auxiliary is
30 kWh on 3 Jan. (both are yearly peaks). There is no heat in the lower level, but it remains
reasonably comfortable.
As Fig. 5 shows, the sunspace cools without solar input, reaching a minimum of 15"C (59 F)
after two cloudy days. The sun reappears on 4 Jan. and a warming trend begins. The house
requires three sunny days to re-establish a recurring cycle.
The floor of the lower level remains well above the lower-level air temperature during the
cooldown, supplying 17 kWh of heat to the house on 3 Jan. Fan operation is typically 8 hours
on a sunny day requiring 3.6 kWh of electricity to operate the 450-W motor. Heat transferred
to the air-floor is typically 19 kWh on a sunny day, of which 48% conducts upward to the
house and 52% conducts downward to the earth, although this varies considerably from day
to day. Conduction upward is impeded somewhat by a carpet on the floor slab. This allows
the floor to become somewhat warmer and thus store more heat without overheating the
house. The earth beneath the house serves as an important heat-storage reservoir.
-------
2571
Annual performance calculation results are shown in Figs. 6 and 7 for November through
April, the typical heating season in Denver. Auxiliary heat totaling 199 kWh for the winter is
used only sporadically as shown in Fig. 6. Peak use is during the first week in January as
shown on Fig. 5. The heat that is vented to maintain the upper floor temperature at less than
26'C (78 F) is also shown on Fig. 6. This totals 2005 kWh.
The temperature of the lower level remains within good comfort bounds without backup heat
or venting. The temperature of the sunspace ranges widely, from about 14'C (56 F) to 34"C
(94 F), and typically is quite warm.
Solar gains absorbed by the house total 29770 kWh over the analysis period. Internal heat
totals 2562 kWh (15.5 kWh/m2). The "useful load" of the house, defined as the heat required
to maintain a 20°C minimum temperature in the house in the absence of all other heat, is
15343 kWh. The load not satisfied by the internal gains is therefore 15343 - 2562 = 12781
kWh. Thus the solar heating fraction is (12781 - 199)/12781 = 0.984 or 98.4%.
The upper curve in the plot on Fig. 7 shows the charging of the air-floor by warm air taken
from tne upper sunspace. The lower curve shows the heat loss by conduction to the house and
the ground. Each total 2734 kWh for the Nov. - Apr. period. The fan operates for a total of
860 hours consuming 387 kWh of electricity. The ratio of heat transported to electricity
consumed is 7.1. This can be considered as a coefficient of performance of the air-floor
system.
backup heat
vented from upper level
Fig. 6 (left)
0 60 120 180
NOV DEC JflN FEB MflR APR
Fig. 7 (right). Simulated heat
flow to and from the air-floor
system. The upper plot (always
positive) is the heat transported by
the fan from the sunspace to the
concrete slab of the lower level.
This totals 1298 kW over the
winter. The lower plot, which is
usually negative, is the total
conduction up and down from the
concrete slab.
-1
-2
-3
NOV
60
DEC JFIN
120
FEB MflR
180
RPR
-------
2572
Several simulations were made for various modifications in design parameters. The results
indicate that the transmittance of the sunspace windows is as important to performance as is
conductance. There is no reduction in auxiliary heat comparing a triple pane super glazing
with two low-e coatings and krypton fill gas with normal clear double glazing (although
minimum sunspace temperatures are increased). Significant improvement was obtained orny
with a glazing that combines high transmittance ana low conductance.
Reductions in sunspace south window area increase annual auxiliary heat as follows:
31.0 m2, 199 kWh; 26.3 m2, 308 kWh; 21.6 m2, 429 kWh. One benefit of decreasing sunspace
window area would be to reduce sunspace temperature swings. We have not yet changed the
design based on these results, but believe that some reduction may be desirable.
The primary heat storage in the sunspace is in the concrete floor and underlying earth.
However, the massive wall between the sunspace and the house is also important. In the
present design this wall is 20 cm (8 in.) brick. A more massive alternative is two outer layers
of brick facing on each side of a concrete center making up a total thickness of 41 cm (16 in.).
With this wall, the predicted backup heat is reduced from 199 kWh to 118 kWh. Although
sunspace comfort is slightly improved (the minimum sunspace temperature is increased from
15°C [59 F] to 16.7°C [62F]) the benefit is so small that the conventional wall was selected.
If the fan that charges the air-floor is not operated, annual backup heat increases only
slightly from 199 kWh to 227 kWh. Based on tnis alone, the air-floor is not justified (the fan
energy is far more than the savings). However, comfort in the lower level is improved by
storing excess heat taken from the top of the sunspace underneath the house. Sunspace
comfort is also notably improved by reducing temperature swings and returning cooler air
near the sunspace floor. For now, the air-floor has been retained in the design; however,
further investigation would be needed to justify it.
The simulation will be extended in the future to study summer comfort characteristics. This
will require a more comprehensive venting model than is used in the current simulation. No
problems are anticipated, mostly because of the favorable summer climate of Denver. Night-
vent cooling should be quite effective and no air-conditioning load is anticipated.
Similar simulations were made for Hartford. The sunspace is typically much cooler than in
Denver. The solar heating fraction is 73% and the auxiliary heat required to maintain 20 "C
on the upper level is 2359 kWh. The sunspace temperature is seldom warm enough in winter
to cause the air-floor fan to operate; as a consequence the air-floor is not charged and the
temperature of the lower level averages between 16'C (61*F) and 18*C (64*F) during mid-
winter.
CONCLUSIONS
The simulations indicate tfie house should work well in Denver. Backup electric heat is
minimal, costing about $16 per year at current rates. Thermal comfort is good and
overheating can be controlled by natural ventilation. The air-floor system seems to work well
and helps maintain comfort in the lower level, but it cannot be justified on the basis of energy
savings. Utilization of the air-floor is enhanced by stratification in the sunspace. Further
work will be required to investigate the air-floor and to determine whether insulation above
or below the air floor would be advantageous.
The simulations indicate that the house would also work well in Hartford, in spite of the poor
solar climate. Backup electric heat is small, costing about $190 per year at current rates;
however the air-floor system is not effective.
ACKNOWLEDGEMENTS
The design follows the general topology used in First Village, Unit One in Santa Fe, New
Mexico designed by Susan Nichols and William Lumpkins. Many people gave valuable Advice
on various design issues, including Jon Becker, Rocky Mountain Solar Glass; Robert Clark,
Alpen; James Tracy, Advanced Foam Products (R-Control panels), Larry Jackson, Delta H
Systems, (Airfloor); Steve Andrews, Solar Spectra; Mark Donelson, Donelson Architects;
Karen Keating, The Keating Partnership; Pamm McFadden, Elements Design Group; and
especially Pat Foss, Foss Builders.
-------
2573
SOLARGREEN... A CRITIQUE AFTER TEN YEARS OF LIVING
Fuller Moore
Center for Building Science Research
Department of Architecture
Miami University
Oxford. OH 45056 USA
ABSTRACT
SOLARGREEN is a passive solar-heated house designed by the author for the USHUD Passive Solar
Residential Design Competition Cycle 5 program and has been described previously (Franklin Research
Center, 1979; Moore, 1980a and 1980b). Hie design won a design award and four construction awards. In
addition. It was one of the three winning designs selected for presentation to Congress to justify the final
release of award funding. It was one of eighteen winning designs selected for detailed presentation in the
USHUD book of the winning designs. The author constructed and has lived in one house of this design; at
least eighteen variations have been constructed in five states. The paper presents a critique of the design
by the architect/owner/occupant after ten years of living.
KEYWORDS
Solargreen, Passive Solar Heating. Greenhouse, Sunspace, Case Study, Architecture, Ohio
INTRODUCTION
The objective of the design of SOLARGREEN was to develop a low-cost, innovative passive solar heating
and cooling system as part of a marketable, aesthetically pleasing dwelling that could be easily
constructed using existing building practices. The basic design is a three-bedroom, two-story, 160 m2
(heated floor area; 1600 ft2) home with an attached greenhouse .
The design utilizes a concept, "thermal layering," to establish an interrelated series of living zones which
are ordered in response to time-of-day activities and the climatic loads of various seasons. In plan (see
figures 3 and 4) this distributes the most frequently occupied areas next to the solar collection and
storage wall, with circulation and utility zones along the north wall as a thermal buffer. In section (figure
5), the living area Is at grade level, while sleeping areas are below grade. This positions daytime activities
(living, dining, cooking, etc.) in the warmest area of the house, while sleeping areas (which require less
light, cooler temperatures, and greater privacy) at the lowest level, with sunlight penetrating into the
sleeping areas through the greenhouse.
Fig. 1. Exterior showing south-facing
greenhouse glazing.
-------
2574
Fig. 2. Interior showing thermal storage wall
common to interior and greenhouse.
FAMILY/STUDY/
BEDROOM
/ BREEZEWAY
J M.BR
AMD C
I
GARAGE
DINING
GREJENHCJUSt
M
GREENHOUSE
STAIR
Fig. 3. Upper floor plan.
Fig. 4. Lower floor plan.
The lower level walls are constructed of reinforced concrete with 5cm (2-in.) polystyrene foam insulation
on the outside. The upper level Is conventional 5x10 cm (2 x 4) wood stud construction with conventional
fiberglass batt wall Insulation plus 2.5cm (1 -In.) expanded polystyrene sheathing, and prefabricated
wood roof trusses with 25cm (10-in) fiberglass batt ceiling insulation.
OPERATION
Cool weather operation (see figure 5a) - in practice, this mode has worked very satisfactorily and is
utilized throughout the heating season.
Cold weather operation (see figure 5b) - this mode has proven to be Ineffective and seldom used because
the single glazing results In too much heat loss and thus reduces the available heat to ineffective levels
-------
2575
during the coldest mornings when It Is most needed. In addition, dust from the collector plate was
Introduced Into the rooms.
Warm weather operation (see figure 5c) - this mode (which uses convection induced by the solar
collector to draw outside air down through a perimeter underground rockbed cooling chamber before
entering the house) has proved to be Ineffective due to the large friction-Induced pressure losses which
result in negligible cooling performance (180 cfm at 64°F db 95% RH; 18m3/mln at 18°C db 95% RH).
These foundation vents have been sealed out of concern for radon gas (even though this has been
monitored on two occasions and found to be Insignificant). In addition, the ceiling vents are kept closed
and the top roof vent Is used to vent the greenhouse, drawing outside air in through the screened exterior
greenhouse door.
Hot weather section (see figure 5d) - as noted above, the foundation vent has been sealed permanently.
However, the author has found that substantial underground cooling effect is achieved by drawing outside
air in through the downstairs east and west exterior windows (not shown in this figure) into the bedrooms
where It is cooled due to contact with the underground walls, then up the stairs into the living area and out
Into the greenhouse before being exhausted by the "whole-house" fan located above the greenhouse.
INSULATING SHADE
The shade is a double layer of aluminlzed, scrim-reinforced, thermally-reflective, polyethelyne material
(an Inexpensive, I-R-reflective "space blanket" film manufactured by King-Seeley Thermos). The double-
layer configuration tends to self-Inflate (apparently due to differences in temperature-related air density
differences) producing an Insulating "dead air" space. Magnetic tape seals that were originally Installed
along the sides have never aligned properly (due to the width variations that occur across this 8.2m [27-ft]
width from temperature and wrinkling), and, because the tape rubber-like material becomes brittle when
cold and has peeled and cracked. A sloping "side ledge" was installed for the edge of the shade to Ite
against and to receive the stationary magnetic tape. In spite of the failure of the magnetic tape, this ledge
has served to support the edge of the shade allowing it to form an effective seal simply by the action of
gravity. To keep the natural sag of the curtain from pulling the sides off of this ledge, the edge of the
curtain was curved outward about 10cm (4 in.) and stiffened with 4.9m (8-ft) long horizontal wood battens.
These battens act to stiffen the shade laterally while leaving It flexible vertically for rolling up. Moore
(1979) has described the use of magnetic seals on insulating shades in general, and Langdon (1981) has
Illustrated and described this particular installation in detail.
All things considered, this large shade has survived ten years of operation well. It has become frayed due
to mechanical abrasion along the edges and has been repaired and reinforced with duct tape (which is
close to the same silver color). It has worn through and been repaired with tape at two places where the
control ropes rub near the top. Considering the low cost of the material ($1.00/m2 or 10t/ft2 in 1980), the
material has performed remarkably well and shown no evidence of sunlight degradation.
One often hears concerns expressed about the effort and inconvenience associated with the thermal
shade. We have never found this to be an inconvenience; the burst of sunlight is so exhilarating that this
dally operation is as natural as pulling back conventional drapes to welcome the new day. Even the slight
physical exertion of pulling the control ropes was a pleasant ritual. In addition to entering and
brightening the living, dining and kitchen space upstairs, the sunlight also flooded Into the downstairs
bedrooms, wakening the late-sleepers - what a nice alarm clock. The sun also washes the dark-colored
masonry thermal storage wall, beginning the daily warm-up cycle.
One unanticipated phenomenon related to the shade that hasn't been a problem for our house but might
be for others under different circumstances. After a particularly cold night, there will be a thick coating of
frost over the entire Interior glass surface. This is a result of the successful Insulating quality of the shade
which reduces the heat loss from the greenhouse to the glass, allowing the glass surface to drop below
both the dewpoint and freezing point temperatures. This thick frost melts quickly (when the shade Is
raised in the morning and the warm air from the greenhouse reaches It) and runs down the glass to puddle
on the bottom sill. This amount can be considerable (up to a cup of water per 0.9 x 1.8m [3-ft x 6-ft] glazing
unit) depending on how low the outdoor temperature drops and the amount of moisture In the greenhouse.
Our greenhouse is used for plants exclusively, and wall and floor surfaces are exposed concrete, so this is
not a source of concern. However, In other buildings where finishes (such as plaster) might be damaged
by such repeated exposure to water, provision should be made to accommodate this "run-off.
-------
Ln
«^j
groontiMM thru
LIVING
£» «oll radiation
op*n-0Of[CbM>nigfit
SLEEPING
COOL
PASSIVE HEAT
pfcftum dtrtrfegtoo
«lr flow* Ov«r frort
•
-------
2577
GLAZING SLOPE
Oh the present house, the sloped greenhouse glazing angle has not proves/ to be the summer overheating
problem reported on a similar house built to this design. (The latter house had several performance-
related modifications to the original design Implemented by its builder, including extensive obstructions
to the thermal storage wall and significantly reduced summer greenhouse ventilation; see Spears, 1983).
We found that the combination of adequate greenhouse ventilation, closing openings between the
greenhouse and the rest of the house, the total shading of the thermal storage wall by the greenhouse roof,
and the partial shading of the lower part of the greenhouse glass by adjacent deciduous trees was
effective in preventing unwanted gain. The air temperature in the lower part of the greenhouse rarely
exceeded 6°C (10°F) above the ambient air temperature; house plants literally thrived throughout the year.
GLAZING COMPONENTS
The greenhouse was glazed with patio-door-glass units two high. These were overlapped, shingle-style to
shed water without an exposed horizontal mullion to trap water. This detail worked well,with one critical
exception. The bottom edge of the glazing unit was left uncovered and exposed to sunlight. After eight
years, this exposure degraded the adhesive compound used by the manufacturer to bond the two glass
layers to each side of the desiccant-fllled aluminum spacer/frame. This resulted In the loss of perimeter
seal on most units and the introduction of moisture Into the airspace. This moisture apparently reacted
with the deslccant material in the frame and resulted in a white, chalk-like Deln^aeposited on both
Internal glass surfaces, subsequently degrading the transmittance and appearance.
In 1988, all of the glass was replaced with similar units, with the addition of some aluminum flashing
material added to shield the otherwise exposed edges from the sun. We found that even with this
provision, virtually all of the glass manufacturers were unwilling to provide any warranty on the glazing
units due to their unconventional sloped installation (local retailers were typically unaware of that
limitation). This fact alone makes vertical glazing the only logical choice for most residential situations
being designed by an architect who must assume responsibility for such specification decisions while
designing an energy-efficient, yet cost-effective, project. For commercial projects with budgets sufficiently
large to permit the use of commercial-grade glazing systems design for sloped Installations, this Is less of
a constraint on building geometry.
GREENHOUSE VENTILATION CONTROL
Ventilation out of the greenhouse is by allowing outside air to be drawn in through basement windows,
and exhausted through a long vent across the entire length of the greenhouse. This exhaust vent is
covered with insect screen and is closed during the winter time. In the original design, I planned to use a
series of long, hinged doors to seal the vent In the winter, expecting to have to use a ladder twice a year to
convert the system from its heating mode to cooling. However, preconstructlon conversations with
Douglas Kelbaugh about his experiences with a similar situation on his own house convinced me of the
necessity of making operation possible from the ground and thus more convenient for frequent operation
that Is desirable during the transition months.
t.
. ... \ '-v •
Fig. 6. The 11.9m-k>ng (39 ft-long) soffit vent at top of
greenhouse has fabric/zipper flap for winter closure
and pull-cords for operation (view from below).
-------
2578
The device that I developed for closing this vent Is a 11.8m (39-ft) long zipper flap vent that opens and
closes and the zipper Is operated using cords direct down to ground level with pulleys (bullet, standing,
and cheek blocks to all you sailors: see figure 6). The system was fabricated by a local sailmaker (they
typically work with all of these materials Including the long zippers). It has worked remarkably well over a
number of years, with one exception. The fabric used was tan canvas (at the time selected as much for
color as anything); this cotton material had several loose threads after Installation that have resulted In
occasional jamming. The remedy that I have successfuljrused on similar Installations on other projects Is
to have the device made out of dacron sailcloth which Is cut with a sallmaker's "hot knife" which effectively
seals the edges, eliminating the offending loose threads, and resulting In a reliable, manually remote-
operated vent closure.
CONCLUSIONS
In general, the primary passive solar heating strategy Is successful (that Is. more than half of the heat
used is attributable to the passive design, and overheating is not a problem). The vertical "thermal
layering" strategy of locating the living areas upstairs (at grade level) and bedrooms downstairs (below
grade) to utilize naturally-occurring convective stratification to effect the desired temperatures
appropriate to each zone Is excellent both In terms of both thermal performance and comfort. The more
conventional horizontal "thermal layering" inherent in the elevatlonal plan stretches the plan east-west
and locates all living areas adjacent to the solar wall Is a well-proved strategy that was reconfirmed on
this project
The exterior 5cm (2 In.) open-cell, expanded polystyrene rigid foam Insulation ("beadboard") Is not
effective In isolating the downstairs underground concrete wall (probably due to saturation with ground
moisture) resulting in condensation on downstairs walls; 5cm (2 In.) of extruded polystyrene foam
("blueboard") Is the recommended alternative. The undei-ground rockbed cooling chamber and the above-
greenhouse air solar preheater are not effective strategies and have been deactivated. The roll-down
reflective shade and the zipper vent are simple and effective component Innovations. The sloped glazing
performs well year around (does not overheat due to generous summer venting), but Is impractical to
construct; vertical glazing with overhang Is recommended as an alternative.
REFERENCES
Franklin Research Center (1979). The First Passive Solar Home Awards. USDOE, Washington. DC. 140-5.
Langdon, W. (1981). Movable Insulation. Rodale Press. Emmaus, PA. 228-30.
Moore, F. (1979). A Magnetic Perimeter Seal for Thermal Shades. Proceedings of the Fourth National
Passive Solar Conference. American Solar Energy Society. Boulder. CO. 395-397.
Moore, F. (1980a). "SOLARGREEN and Patoka Nature Center," Proceedings of the Internationale
Sonnenjorum 3, Hamburg. Germany, 546-553.
Moore. F. (1980b). "Thermal Layering: A Passive Solar Design Strategy," Proceedings of the Fifth National
Passive Solar Conference. American Solar Energy Society, Boulder, CO. 1234-1238.
Spears, J. (1983). "Comparison of Vertical High Performance Glazing to Sloped Double Glazing with
Movable Insulation," Eighth National Passive Solar Conference. American Solar Energy Society,
Boulder, CO, 253-264.
-------
2579
HORIZONTAL AND VERTICAL, THERMAL AND AESTHETIC
RESPONSES TO TWO CONTRASTING MICROCLIMATES
Polly Cooper and Ken Haggard
San Luis Solar Group, Santa Margarita, CA
ABSTRACT
In this paper we compare two passive solar residences which employ the same
thermal techniques at sites with quite different microclimates. South
glazing and skylights with skylids / coupled to concentrated thermal mass
(water tanks) and distributed mass (concrete slabs and thin masonry veneers)
achieve thermal comfort within two quite different architectural forms.
KEYWORDS
Microclimate, visual and thermal compositions with site
INTRODUCTION
At the most fundamental level, passive solar architecture is the
architectural approach that most thoroughly connects to location. Obviously
we must do this thermally if we are to use the thermal sources and sinks of a
local microclimate to condition the building, but it is equally important to
give the same care to social, functional, and compositional integration to
the site. Passive solar architecture should be an architectural evolution
that, besides saving energy and providing superior comfort, reminds us of our
obligations to the uniqueness of every site, the spirit of every place with
which we come in contact. For each situation we should evolve architecture
that works with and enhances each site so that we maintain the sequences of
wholes within wholes that is our natural heritage as living beings. Our
architecture should speak to the site as part of its whole composition and
the site should speak to the architecture as part of its thermal and visual
composition - the interplay of wholes within wholes.
Many of our firm's projects occur in relatively rural areas with a great
variety of topography, vegetation types and microclimate. Therefore, these
concerns are dominate in our design process. These two projects illustrate
this approach.
-------
2580
SITE -
Site I Gentle south facing slope of Lopez valley, major views to the
south-west, live oaks to the north, excellent winter sun, hot periods
in the summer tempered by occasional fog from the nearby Pacific
Ocean.
Site II Steep north facing slope,
hemmed in between two all
weather streams, tall mixed
forest up hill to the
south, cold wet winters
with early sunsets and
severe cold air drainage.
This site has almost twice
the rainall of Site I, hot
summers with temperature up
to 112a F.
-------
2581
SITE AND VISUAL AND THERMAL RESPONSE
This Building has a low horizontal profile so as not to break the flow of the
land nor visually obscure the oak forest at the top of the site.
The Building is raised in height to
catch sun over the top of the oak
and cypress forest up hill to the
south. The three story section
completes a south facing court which
provides comfortable outdoor space
in the winter by emphasizing morning
sun and providing protection from
cold air drainage in the afternoon.
-------
2582
SITE AND ARCHITECTURAL FORM
The overall configuration of the
building is to make its scale
appropriate to that of the courtyard
while maintaining its vertical
expression. Thermal requirements of
the lower two floors are met by
water tanks with selective surfaces
and low E glazing below south
windows. The upper two floors are
served by skylights with skylids
coupled to thin veneer mass walls of
concrete brick and ferrocement, as
well as the third floor which is
concrete on a structural metal deck.
The building is conditioned by low water tanks with selective surfaces under
south windows. A horizontal
provide additional aperture.
band of skylights with skylids on the roof
These south-facing openings are thermally
coupled to a tiled slab and to brick veneer on many of the interior walls.
-------
2583
ARCHITECTURAL FORM AND SITE
Major views are up to the south on
the first floor and diagonally out
each side on the third floor. The
second floor is a combination of
these conditions. This creates a
progression of orientations as one
ascends the building, creating an
interplay of visual site connections
within the verticality of the
building.
The major views are to the south and southwest. The crenulated south facade
allows continuity of view, and visually connects the valley to the interior.
-------
2584
-------
2585
PLUS ENERGY HOUSING IN DENMARK
Lars Yde
Folkecenter for Renewable Energy
P.O. Box 200, DK-7760 Hurup Thy, Denmark
ABSTRACT
The project consists of demonstration of a 590 m2 low energy house
comprising 260 m2 living and working area under 360 m2 highly
insulated solar greenhouse with plant production and recreational
areas.
Humidity in the greenhouse is regulated by use of heat pumps which
condense the water laden air to produce 50°C hot water for space
and domestic water heating. Instead of the normal practire of
ventilating all excess heat to the outside in summer, the heat
pumps maintain a temperature of 22°C with 6 0% relative humidity in
the greenhouse and excess heat to an amount of 365 kWh/m2 of
green house area is produced annually i.e. equivalent to the output
from a flat plate collector. Over and above the heating
requirements in the demonstration house there is an excess of
100,00 0 kWh which is available for space and domestic water
heating in adjoining houses.
KEYWORDS
Solar energy, greenhouse, mobile insulation, humidity control by
heatpump, export of heat.
INTRODUCTION
This demonstration project supported by the Commission of the
European Communities and the Danish Council of Technology have the
following 3 innovatory aspects:
1. Use of a building as heat production source with sale of
surplus energy.
2. Energy savings in comparison to a traditional greenhouse.
3. Cost savings on dwelling prices due to a climate shield.
EXPORT 0E HEAT.
lhree of the Centre for Renewable Energy's windmills (13, 22 and
75 kilowatt) will supply part of their electricity to the energy-
producing house, which in its turn will produce heat which can be
supplied to the adjoining houses. The energy-producing house
will be 300 m2 in ground area and produce an annual energy surplus
of over 100,000 kilowatt hours. Equivalent to the annual heating
-------
2586
demand of four normal family dwellings, or 12 low-energy (super-
insulated) houses.
BEADWALL TECHNOLOGY.
The energy-producing house will be built of normally available
materials. It will have the combined functions of a greenhouse and
office building. The roof of 360 m2 of south-sloping glazing
enables the sun's rays to heat up walls, floor and ceiling. The
house reaches its required temperature after only a few hours
sunshine, and with the help of the mobile insulation in the roof,
can maintain this temperature for at least two days without
further addition of heat.
The glazed roof is made up of two layers of tempered glass. The 20
centimeter space between the two layers can be partly, or
completely filled up with polystyrene beads. When completely
filled up with polystyrene insulation, the glazed roof has an
insulation value which is equivalent to that of the outer walls of
a well-insulated dwelling. The polystyrene spheres are blown into
and sucked out of the roof space and can in the daytime be used
for shading, while their function at night will be to maintain the
house's temperature.
FRUIT AND VEGETABLES.
Half of the house will be used as a greenhouse with production of
fruit, vegetables and flowers. These plants give off moist air,
and the energy that is in the moist air (heat of evaporation)
forms the basis of the house's heat surplus.
The plants evaporate water when the sun shines on them. Hanging
under the ceiling are rows of plastic pipes through which cold
water flows. The moist warm air from the plants forms drops of
condensed water on the plastic pipes and transfers the heat of
evaporation to the water in the pipes. The water drops are
collected in a gutter and returned to the plants.
The now slightly warmed water in the plastic pipes goes to a heat
pump where the heat is extracted. F rom here cold water is pumped
up to the ceiling again. The heat pump sends hot water to the
house's radiators and also to the mini district-heating network
which supplies the adjoining houses. The energy-producing house
has in addition a hot and cold store which evens out the
differences between supply and demand, and thus allows the size of
the heat pump to be reduced to a minimum.
In the night when the sun does not shine on the plants, their
evaporation is minimal, which means that condensation will not be
needed. During this period the heat pump will produce cold for the
cold storage, which thus will be ready for condensation of water
from the moist air during the sunny hours the next day.
The heat pump system supplies all heat needs and at the same time
solves two other problems. It controls the inside climate in the
house by reducing humidity and in addition can "export" surplus
heat to other parts of the house and the rest Of the village. It
is calculated that the house will be able to sell 180,000
-------
2587
kilowatt hours of heat to nearby houses. There is, however, a
requirement of 80,000 kilowatt hours of windmill electricity to
run the heat pump and other electrical needs of the house.
The cost of the energy-producing house runs to 3.3 million DKK
(approx. 550,000 USD). Measurement and analysis account for a
considerable portion of this sum. The actual cost of the house is
4,500 DKK (750 USD) per m2. This is cheaper than the normal cost
of a dwelling, which runs to 6,000 DKK (1,000 USD) per m2.
ACKNOWLEDGEMENTS
The development of the Plus Energy House is a team work between
Biologist Hamish Stewart, Biologist Jergen Hinge, Architect
Benjamin Bak, Architect Merete Scheller and Engineer Lars Yde from
the Folkecenter for Renewable Energy.
REFERENCES
The Use of Renewable Energy in Commercial Greenhouses. FC-print,
January 1989. ISBN 87-88-660-47-8.
Torkild Vest Hansen, Examinations of high insulating window
constructions with mobile night insulation. Laboratory for
Heat Insulation, The Technical University of Denmark, January
1977, Information No. 4.
A/ORP^BSTJYSK FOLK£C£f
-------
2588
PLASTIC PIPES
"" COMPENSATOR
SON SHIELD
HOIST AIR NX
FROM PLANTS
SON
TWO LAYERS OF TEMPERED
GLASS TO BE FILLED UP
WITH POLYSTYRENE BEADS
AT NIGHT
WATER RETURNED TO T HE PLANTS
HEAT PUMP
EXPORT OF HEAT
SPACE HEAT
<1—'
DOMESTIC
HOT WATER
HOT STORE
COLD STORE
Fig. 2. Energy System of Plus Energy House
When the sun is shining on the plants their water evaporation
increases. The moisture is condensed on a condensator (long
plastic pipes) by means of which evaporation heat is transferred
by the heat pump to the heat system of the house and for sale of
district heat.
The warm and the cold store equalize the differences between
demand and production, which is the reason why the size of the
heat pump can be reduced considerably.
-------
2589
ARCHITECTS BENJAMIN BAK %¦
NBMETE SCHELUX-
tNERGY ENG/NEBK, LAKS YPE
Fig. 3. Folkecenter for renewable Energy
-------
2590
FIRST YEAR PERFORMANCE OF THE CANADIAN ADVANCED HOUSE
Stephen C. Carpenter, P.Eng.
John P. Kokko
Enermodal Engineering Limited
368 Phillip Street, Unit 2
Waterloo, Ontario N2L 5J1
Elizabeth White, Consultant
Marsh Hill Farm
R. R. 4
Stirling, Ontario
KOK 3E0
ABSTRACT
The monitored performance of the Canadian Advanced House Project is presented. Comparisons are
made between design estimates and monitored results with reasons given for the discrepancies. The
house was found to consume 28% more energy than predicted. This was due to difficulties with
passive solar preheating of ventilation air and lower than expected integrated mechanical system
performance. Improvement in IMS performance is expected when the house is fully occupied.
KEYWORDS
Low energy housing; energy conservation; passive solar; heat recovery.
INTRODUCTION
The Advanced House is Canada's most recent demonstration of leading-edge residential energy-
efficiency, and her contribution to Task XIII of the International Energy Agency. The Advanced House
was constructed by the Fram Building Group, and is located in Brampton, Ontario. The principal
funding agents are the Ontario Ministry of Energy, Energy Mines and Resources Canada and Ontario
Hydro. The project objectives are demonstration of new energy-saving technologies and equipment and
assessment of their potential for significant reductions in total home energy gse and electrical peak load
reductions. Construction has been completed and monitoring is under way.
According to predictions, the Advanced House is expected to use a total of 12,500 kWh per year for
ajl purchased energy under normal occupancy, or 30 kWhr/m2 of floor area (including the basement).
Total energy consumption is projected to be approximately 30% of an equivalent conventional home
built to Ontario Building Code standards and 47% of an equivalent R-2000 home.
This paper describes the house design and construction, the monitoring system and results to date.
The results presented in this paper are for the period when the house was open for public viewing but
otherwise unoccupied. The house is expected to be sold to a private owner, and monitoring will
continue until the spring of 1993.
HOUSE DESCRIPTION
The Advanced House is located on a corner lot in a Brampton subdivision just northwest of Toronto.
Apart from an intentionally eye-catching two-storey sunspace on the south side, the house looks similar
to its neighbours. The energy conserving features are:
-------
2591
Hioh-Performance Windows are triple-glazed with two sputtered iow-E coatings on surfaces 2 and 5,
two 12 mm (0.5 inch) airspaces, argon gas fill in both airspaces, a butyl rubber spacer and solid wood
frame. The predicted U-value for this window design was 1.06 W/mtoC (0.19 BTU/hrft^F), according to
the FRAME and VISION computer programs.
Integrated mechanical system (IMS) (see Figure 1) supplies space heating and cooling, DHW, and heat
recovery ventilation. It uses a cold water/ice slush thermal storage system to store heat recovered from
grey water, ventilation exhaust, and excessive passive solar, and internal gains. The objective is
demonstration of a multi-functional mechanical system which can store recovered heat, thereby
increasing passive solar utilization and waste-heat recovery, and reducing overall building energy
requirements. The system seasonal C.O.P. was designed to be 2.8 - 3.0, with an output of 6 kW.
household return air mixed
with fresh ventilation air from
outside (via sunspace)
3. A toced ok system
moves household air
through a hot water
heathg coll (winter) or
o cold water cooing
coll (summer).
space cooling col
space heating cod
hot water for
household use
The hot water tank
provides water for:
A. domes tic use
B. space heating
bock-up heater'
household
waste water
to drart
heat recovery
from kitchen
heat recovery
from waste
water
ond bothroom
ertxxBtair
'6*8;
Ice Water
warm
exhoust air
Qrttaust air
Storoge
to outside
neat
Pump
Hot
Wotei
fonk
1. Hoot from waste woter and Indoor
etftoust air b recovered and used to
Increase the temporature of the Ice
woter.
2. When spoce heat Is required
the heat pump extrocts energy
from the Ice water tank (making
It even colder) and uses this to
warm the water In the hot water
tank.
rig. l INTEGRATED MECHANICAL SYSTEM SCHEMATIC
The IMS demonstrated in the Advanced House was developed by Allen Associates of Toronto. The IMS
consists of a cabinet which houses a 40-gallon domestic hot water (DHW) tank with a back-up
resistance coil, a heat pump, heating and cooling fan-coils, an exhaust-air heat recovery fan coil, and
a thermal storage tank. A brine solution conveys heat to and from the storage tank and heat recovery
coils. The IMS is located totally indoors. The liquid-to-liquid heat pump has the low-temperature side
connected to the thermal storage tank and the high-temperature side to the DHW tank. Waste heat
is recovered from grey water through an extended length of copper drain with a pipe carrying brine
soldered to it. Brine circulates continuously through the two heat exchangers during the heating
season. Ventilation exhaust-air heat recovery is accomplished through the separate cooling coil in the
exhaust duct. Excessive solar gains, internal gains and wood heat gains are all recovered via the
ventilation. Space heating is supplied by a fan coil plumbed directly to the DHW tank. Return air,
mixed with fresh ventilation air, is blown through the fan-coil and distributed to the house. When the
DHW tank temperature drops, the heat pump operates, cooling the thermal storage tank and reheating
the DHW tank. As energy drawn from the thermal storage tank is free heat, purchased energy is only
-------
2592
required to operate the heat pump. The purchased energy also contributes to space heating.
Domestic hot water is supplied directly from the DHW tank. When space cooling is required, cold brine
is circulated to the house circulating fan-coil unit.
Energy efficient lights and appliances. An average Canadian home is considered to use 8000 kWh per
year to run lights and appliances. Energy-efficient lights and appliances were chosen to achieve a 50%
reduction. Specific measures include: Elimination of all conventional incandescent light bulbs in favour
of compact fluorescent, high-efficiency fluorescent and halogen lamps. This reduces lighting energy
use by 70% - 80%. Use of high-efficiency appliances includes the California Sun Frost refrigerator,
designed for PV applications and rated at 20 kWh/month, or about 20% of the energy consumption of
a conventional refrigerator. Other major appliances were supplied by the European manufacturer AEG
and are estimated to provide 20% savings over North American brands.
Hioh levels of insulation. The nominal envelope insulation levels are as follows:
Ceilings
RSI
10.6
(R60)
Above grade walls
RSI
7.0
(R40)
Below grade walls
RSI
5.6
(R32)
Basement floor slab
RSI
1.2
(R7)
ABOVE GfiAOfc WAU. R-40
bufcVQPKX'
i iiroadid
rw bN«g Mudv*
brac*g
t 1TIMM
r«4- <-w Mvmg wuc
I l/r
• irt MlM pokt AV8
GM luwtoor w towt
HEADER BANO R-40
trt*
I l/7*Gtt*OM.M*fl»t*0rt
n ca> eort town MM
7* Mown tautniQn
Ft trtOOX po
EtoMtt. * pUM
FMUMtien
HftotliMCf
1M*i» 1/T pfpwoo rtrMUp
ABOVE GRACE WAU R-40
buWng p«p«f
urGMaN.um
n<*MiwgtuoCiM9Mibn
2 iO"spac«
ru*«ci«MioiM
• 1/rtOnnMCOA
• m m»m poty av8
iq* oy-af
| CM «ucfloo> m
HEAOCRBANOfMO
tlUNng
•"KMC*
buU*gp«p«<
I 1/r GM*da4.WWT«Up«d
r bloon
•MfTMInpUC*
4-p#1
PC*
ovit««4iaofM
mta
MOM
FlOOn SLA8 N 4 J
y cMr+e
r baudM msuuuon
(ma per tfwvrad
BClOW QRAOC WAU. R-52
Mcging io gred*
r tUMcud
flr«»
i9*eonc.
( ml pey dvivflne io Qrtf
3 '<7»m
r*rat«s*iroc
rr*v 7* btoam hutaion
• rnl*M)Mpo*AV*
\rrorymtt
The cross-section of construction details is
shown in Figure 2. The majority of the insula-
tion in the house is cellulose. Wet-blown
cellulose is installed above grade, dry-blown is
used below grade and in the attic. The advan-
tage of the blown-in-place systems is the ability
to totally fill all voids, reducing convection loses
within the wall, and other problems due to
improper installation of batts. Cellulose is less
expensive than the foam insulations and is
considered to be environmentally acceptable.
Semi-rigid glass fibre is used as exterior shea-
thing above grade, below the basement floor
slab and on the exterior of the basement.
Airtight construction. Overhangs, cantilevers,
living space over unheated areas, and other
design features prone to air leakage difficulties
were eliminated at the design stage. The
primary air barrier of 6-mil sealed polyethylene,
was installed directly behind the drywall. An
exterior Tyvek weather barrier and the interior
drywall provided secondary air barriers. Elec-
trical boxes were moulded thermoplastic with
gasketed cover plates. Vents and other pene-
trations were sealed according to normal R-
2000 building practise.
Two storey passive-solar sunsoace. The major-
ity of the house's south glass area is located in
the sunspace. Because temperatures are
allowed to float (i.e., the space receives only
passive and indirect heating and cooling), glass
areas can be increased without energy penalty.
Passive gains are stored in the concrete floor slab, in the backfill under the slab, and in the large mass
of fireplace masonry. The gains are transferred to the heated living space through the common walls,
and through the glass doors which connect the two spaces. The sunspace also provides preheating
of fresh ventilation air. Outdoor air is brought in at the ceiling of the sunspace. A high wall return is
used and the fresh, warmed air is pulled through the masonry and into the floor slab to provide heat
-------
2593
to or remove heat from storage. The preheated air is then delivered to the IMS for distribution to each
room. An automatically-operated skylight provides passive cooling when the sunspace exceeds a set-
point temperature. Reflective blinds and exterior shading elements provide summer shading.
Enerav-efficient fireplace. This fireplace is a prefabricated, contra-flow masonry unit with exterior
combustion air and glass doors, manufactured by Temp-Cast Masonry Heater Inc. The principle is
similar to the Scandinavian contra-flow masonry heaters. A fast, hot burn reduces emissions and
heats up the masonry, which in turn gradually radiates its heat to the room. The fireplace is located
on the interior wall between the family room and the sun space. The mass can absorb passive solar
heat on the sun space side and can also radiate wood heat to the sun space.
A PC-based data acquisition system is installed to monitor the performance of the house and its
systems. Monitoring points include:
12 temperatures and 4 humidities for tracking comfort levels and energy flows throughout the
house
12 electrical and water consumption meters for specific appliance energy and water usage
2 solar radiation, 1 outdoor temperature and 1 humidity for meteorological monitoring
3 air flows, 4 liquid flows, 4 electrical consumption temperatures, 2 humidities and 4 status
switches for detailed characterization of IMS operation
In addition, one-time tests were performed to determine the window U-value, insulation R-value, building
heat loss coefficient, building time constant, building leakage area and pollutant concentrations.
CONSTRUCTION
No special crews or training were used for the construction of the Advanced House, nor were the site
supervisors party to the design meetings. As a result, many special details were overlooked and some
work had to be redone. Despite these problems, the building envelope performance meets the original
targets (see below). Although this approach to construction is not preferred by the authors, it does
indicate that wide-scale implementation of these technologies would be possible without major trades
retraining.
Some of the variations from conventional construction are:
1. Steel beams in the basement bear on posts inset from the concrete foundation to avoid thermal
bridging.
2. Double framing of above-grade walls, and modification of partition and floor framing to
accommodate a continuous air barrier.
3. Application of plywood jamb extensions to window frames prior to installation to provide an air
seal and an insulation stop.
4. Floor-to-ceiling framing on the inside of the below-grade walls.
5. Stiffening of ceiling joists to allow for the weight of added insulation.
6. Use of underslab insulation.
7. Use of dry-blown cellulose in below-grade, interior walls.
8. Use of wet-blown cellulose in above-grade walls.
9. Duct layout, including the fresh air supply system, avoided ducts in exterior walls and used high
wall supply registers.
10. Air-sealing, according to standard R2000 practise, which was performed by Fram's R-2000
sealing crews.
In addition to the construction variations, several new components were used-most of which had to
be obtained from specialty suppliers and required special installation. Interfacing the new components
with standard trade practice was particularly difficult. For example, the fireplace required a flue with
a thermal break at ceiling level, and a decorative facing with an expansion gap. The IMS duct layout
was somewhat different from standard designs. The refrigerator required a condensate drain. The
automatic bathroom faucet needed power. The lack of on-site personnel to supervise trades on a daily
basis and to co-ordinate items such as these led to some construction delays.
-------
2594
PERFORMANCE RESULTS
The results of the one time tests on the building shell are summarized in Table 1. The building appears
to be performing very close to design predictions. Although the measured values of insulation
conductivity, window U-value and air tightness are slightly above the design predictions, the total
building heat loss coefficient is slightly below the design value. Part of this discrepancy is probably
due to the added thermal resistance of closets, kitchen cabinets, the garage and the like, which were
not accounted for in the design calculation.
TABLE 1. Results of Building Shell Testing
Design Measured
Cellulose Conductivity (Wet Blown) W/(m*°C) 0.038 0.046
Total Window U-Value W/(m2*C) 1.06 1.10
Building Equivalent Leakage Area at 10 Pa (cm2) 304. 385.
Normalized Leakage Area at 10 Pa (cmVm2) 0.47 0.60
Building Heat Loss Coefficient W/°C 209 199
Building Time Constant in hours 48 54
The air quality in the house was tested to ensure adequate ventilation was obtained. The exhaust
ventilation rate was set at 68 l/s in accordance with the Canadian ventilation standard F326.
Formaldehyde concentration was measured at 0.048 ppm in the master bedroom and 0.026 ppm in the
dining room, both values well below the Canadian guideline of 0.10 ppm. The radon concentration was
0.002 working level units, well below the conservative U.S. guideline of 0.02 wlu (4 picocuries/litre).
The energy consumption of the house for the period July 1990 to February 1991 is presented in Figure
3. The line shows the predicted values for the same water load and internal gains (but with TMY
weather data). In general, there is good agreement between predicted and monitored values. As is often
the case, the monitored values are slightly higher, there appear to be two major reasons for the
discrepancy.
First, the sunspace preheat does not appear to be performing as designed. Blower door tests on the
sunspace showed that there is considerable air leakage around the six sets of single-glazed double
French doors connecting the sunspace to the house. The net effect is that only 45 % of the air being
pulled into the sunspace is from the outside. The remainder is air leaking from the house into the
sunspace. Measurements showed that the air flow in the sunspace preheat air duct to the IMS is only
half the design value. The net result is that only 15 l/s of outdoor air is being preheated by the
sunspace, the remaining 53 l/s of supply air comes from infiltration through the building shell.
The second reason for the discrepancy is that the IMS performance is lower than expected. Average
heating output at the time of this writing was approximately 4 kW. Figure 4 shows the monthly
breakdown of IMS energy use. The pumps, fans, and controls in the IMS consume an average of 400
Watts continuously. Over the year, this represents an energy use of 3700 kWhr or 30 % of the total
house energy consumption. Parasitic energy consumption is usually less than 5% of total energy
consumption in a standard house; however, total energy consumption at the Advanced House is much
lower then a standard house, and parasitic consumption is much higher (probably because the
Advanced House has two pumps and two fans that run continuously). The cumulative effect is to make
parasitic energy consumption a very significant portion of the whole house load: possibilities for
reduction should be investigated.
The COP of the IMS is lower than expected, causing extra back-up element use. The system COP
(including compressor, pumps and controls but excluding the fans that would be required in a non-
heat pump system) ranged from 1.5 to 2 during the winter. The system COP for the summer months
was slightly higher (because of higher evaporator temperatures), but still fell short of the expected COP
of 2.5 to 3.0. Approximately 0.3 to 0.4 of this COP shortfall is due to the higher than expected parasitic
consumption. Some of the shortfall is also due to the lack of heat available for recoveiy due to the
house being unoccupied. This reduced the average temperature in the thermal storage tank, and
therefore increased heat pump energy consumption. Investigations are under way to resolve the
remaining differences. Fine tuning of the heat pump shows promise of increasing IMS output
-------
2595
Though there appears to be
good agreement on energy
consumption In July and Dec-
ember, the comfort levels were
not completely satisfactory. The
shortcomings were due to min-
or control problems, which were
remedied. In heating mode the
thermostat anticipator had to be
disconnected. In cooling mode,
the load management computer
was functioning erratically. The
computer was removed when it
was found that load shedding
was not necessary. The peak
electrical demand was 8.2 kW,
well below the load-shedding
setpoint of 10 kW.
CONCLUSIONS
The conclusions from this pro-
ject thus far are:
- the Advanced House is a
good example of how
various energy-efficient
technologies can be incor-
porated into residential
housing;
- the building appears to be
operating close to design
predictions. Although the
house energy consumption
is 28% higher than predicted
values, the house still
consumes only 38% of the
energy of a conventional
house of the same size, and
60% of a similar sized R2000
house; and
- the higher than expected
energy use is attributable to
high IMS parasitic power
requirements, low IMS out-
put, and difficulties with the
sunspace preheating of
ventilation air.
- improvement in IMS perfor-
mance is expected when the
house is occupied.
ACKNOWLEDGEMENTS
Monitored Energy Consumption
at the Advanced House
Jul 90 Aug 90 S»p9© Oct 90 Nov SO 0*c96 JtnlS F*»9!
I Total IMS (c* Fan») SIS Ught & foctptkt* m Applimcm (Relrig) Computer*
- PretScwd (un&r tame load oon&ikx*}
Fig. 3
Monitored Perfomance of the IMS
at the Advanced House
M9Q Aug 80 S*p9Q Oct 80 Hw9Q DtetO JmSt ftbB!
; MB IMS Comptmnor ggj 8iK*-up fyn.Puty .Cortrofc "j
Fig. 4
The authors would like to acknowledge the financial assistance of The Panel on Energy Research and
Development of Energy, Mines and Resources Canada
-------
2596
PASSIVE AND ACTIVE SOLAR WIND GENERATOR REPAIR FACILITY-WITH
NO BACKUP SYSTEMS IN CANYON, TEXAS
L.M. Holder III
4202 Spicewood Springs Road
Ste 214
Austin, Texas 78759
This is a demonstration project using renewable energy funds from the
Exxon overcharge money administered by the Energy Efficiency Division of
the Governor's Office of the State of Texas. The project near Amarillo in
the Texas panhandle is a 30' x 75' wind generator repair facility divided into
a 30' x 55' workroom and a 20' x 30' office and support area.
FLOOR PLAN
Fig. 1.
SOUTH ELEVATION
-------
2597
Fi9* 2* NORTH ELEVATION
This workroom is passively cooled as well as passively and actively heated.
Low inlet windows on the south combined with high outlet windows on the
north provide positive ventilation and night flushing during the cooling
season. The openings were designed to provide optimum size for effective
ventilation and minimize infiltration. Ceiling fans are used to increase air
movement within the space to provide human comfort at higher
temperatures. A large area of low "E" glass on the south face provides a
direct gain solar collector into the main workroom work area. East glass
allows early morning gain in the winter and is shaded during the cooling
season. The floor is a concrete slab over a 1250 ft.2 rockbed storage
connected to 234 ft.2 of air solar collectors mounted vertically to the
south face of the building. The collectors are connected to two
thermostatically controlled circulating fans each serving half of the collector
area. These fans circulate the air through the rockbed in 8" diameter pipes
with 1" diameter holes drilled at 12" on center. 6' away is another 6"
diameter pipe with 1" diameter holes at the same spacing, which returns the
air to the collectors. (Some pipes have holes on both sides 1' on center
which alternate). When the passive systems are not available, the thermal lag
of the rockbed storage provides a warm floor for human comfort.
WEST EI.EVATION
EAST ELEVATION
3
The building shell was designed to be well insulated with a low infiltration
rate and a ventilated roof with a radiant barrier to protect the building
against adverse climate conditions inherent to the Texas Panhandle. All
glass was carefully located with minimum glass on the north and west,
seasonally shaded glass on the east, and adequate glass on the south for
passive heating, natural ventilation, and daylighting. Low "E" glass in wood
frames was used to reduce infiltration and heat loss at night. Using a
computer model, it was ascertained that 6" batt insulation in the walls and
R30 insulation in the roof would be optimum. Spraying a radiant barrier
paint to the underside of the roof panels and using a ventilated ridge and
soffitt vent reduced the impact of the "Texas Sun" during the cooling
season, as well as, decreasing "black sky re-radiation" on clear winter nights.
Tyvek infiltration barrier was installed throughout the roof system and
joined to the wall system which consists of metal panels. Furthermore, the
overhead door was carefully selected for good quality weatherstripping and
insulated panels.
-------
2598
SECTION
Fig. 4.
Daylighting for the entire workroom area is provided by a continuous glass
strip above the solar panels. Therefore, the artificial lights were not
connected during the construction period, since the natural light provided
quality lighting. Floor to ceiling south faring glass areas adjacent to the
major work area provides more light where more detailed activities occur.
So, the lights need only be used on very dark days and at night. The office
area has the same shell characteristics as the workroom. Similiarly, the
south wall is glass for direct gain collection in the winter and daylighting
and natural ventilation when appropriate. The limited north glass is all
operable to provide outlets for night flushing. To moderate temperature
fluctuations in this space, the rockbed has been replaced by water tubes at
the south glass for heating and masonry on the inside of the north wall for
cooling.
All water is heated by an active solar water heater.
Using computer simulations, all features and systems of this project were
selected and optimized. The anticipated maximum temperature was lowered
5 degrees and the minimum temperature was raised 10 degrees using the
simulation.
"Stand alone" photovoltaic lights iluminate the parking lot and project sign.
A 2 kW photovoltaic array and a 10 kW wind generator are grid connected. It
is anticipated that even with welding and other equipment in the workroom
this project will approach being a net energy producer rather than a net
energy user.
The project is currently under construction and will be completed this
spring. Operation data will be available and can be presented in August 1991.
-------
3.4 Vernacular Architecture I
-------
Intentionally Blank Page
-------
2601
BIOCLIMATIC ASPECTS OF THE REURBANIZATION OF
THE SAVA AMPHITHEATRE AREA IN BELGRADE
Dr Darko Radovld, d.l.a.
Center for Planning of Urban Development
Belgrade, Yugoslavia
ABSTRACT
Paper presents some themes from the first part of the research which, as a
part of the project „The amphitheatre of the Sava river and surrounding
areas", has been done In organization and under patronage of the Serbian
Academy of Sciences and Arts . That phase was dealing mostly with
theoretical aspects of blocllmatism, and this paper presents the main
results of the accomplished, first stage of that research.
KEYWORDS AND PHRASES
architectural typology, blocllmatlc regulation,covered pedestrian spaces,
directives for blocllmatlc development, gudellnes for architectural design,
operative model, planning documents, spatial levels, urban structure, level
of determination
INTRODUCTION
The natural amphitheatre formed by the band of the Sava river is In the
very centre of the capital city of Yugoslavia. The largest part of this
complex is occupied by the central railway station, Its services and the
tracks of Yugoslav railways. Due to the growth of the town these
Installations are blocking the most Important directions and possibilities
of the development of Belgrade, as well as the very Important and valuable
access to the river Itself. Therefore, relocation of the station and its
facilities, and development of the urban tissue of central Belgrade and
reurbanlzation of the area are anticipated in all planning documents.
As a part of the interdisciplinary project of the SANU (Serbian Academy of
Sciences and Arts), the first phase of the research on blocllmatlc
potentials of the area and blocllmatlc potentials of the reurbanlzatIon
Itself has been done. Second phase, dealing with concrete space and
problems, with urban and architectural proposals, in progress.
THE OPERATIVE MODEL OF URBAN STRUCTURE AND ARCHITECTURE FOR THE
AMPITHEATRE
Thinking on the most general level, with just a couple of preliminary
analyses and empirical assumptions, we can discuss the general 'model of
future urban structure and architectural typology of this part of Belgrade.
-------
2602
Ecological architecture, with bioclimatics as an inseparable part of it, means
dealing with scientifically-based architecture. Improvisations and "creativity"
in that area very often are nothing but trivialities.
Wider climatic context of the ampitheatre shows delicacy that has to be dealt
with during future research and design development stages of the work. Over-
heated summers and severe, continental winters require operational "double-
mode". All features related to climate have to be understood as favorable
during one season and unfavorable during another. The emphasis has to be on
comfort in both the overheated and the cold season; on comfort of both closed
spaces and open ones, with the open spaces being the stage for a very lively
urban scene in the summer months.
In order to meet the demands of the winter season and produce energy-saving
results, the urban structure has to be based on the requirements of that season.
That is an approximation of what we can learn from the scarce urban history of
the area.
The ampitheatre itself is, especially from a bioclimatological point of view,
not favourable for urban development, at least not for all functions of the
city. In the past there was a swamp there, drained from the river only one
hundred years ago, for the development of the railway installations. This is
very low ground, close to the river and with a high level of underground
water. This tells us that at the present, we cannot expect favourable con-
ditions. But development of cities and other aspects of urban life are some-
times considered more important than ecological situations. This puts the
emphasis on how to solve the problem only; and that is the proper emphasis for
significant contributions of bioclimatism and its capability to improve given
situations.
Generally speaking, urban structures in temperate-to-continental climate areas
can be developed freely: at this point we would like to mention the important
theme of covered pedestrian spaces, i.e. arcades, colonades, passages and
other, historically proven, urban settings which can operate in the aforemen-
tioned "double mode" to improve outdoor conditions in all seasons of the year.
Dense urban tissue makes development of energy saving and wind-conscious
architectural types possible with solar access,as the primary emphasis for
residential buildings, provided
THE IMPLEMENTATION OF THE MODEL AND ITS ADAPTATION TO SPECIFIC CASES
Development of structures closely defined by the general model of the architec-
ture of the ampitheatre, without considerable variation, is to be expected. In
cases like this., bioclimatio architecture finds its particularities at "lower"
levels defined by the most concrete stirmuluses coming out from microclimatic
Variations3 to the simplest topoclimatic "formae-localis". Sometimes just the
smallest details, details of the kind that gives "soul", give the breath of
recognizability and uniqueness, are delightfully found in structures of
traditional, spontaneous settlements.
If we want to have all the data that can produce the needed level of refine-
ment, then very precise and strict analyses are necessary. Just on the basis,
it is possible to proceed to the next stages of thinking on the future urban
and architectural structures of the Sava Ampitheatre: to definitions of spatial
levels for bioclimatic intervention, bioclimatic synthesis of the operational
model for each of those levels and to numerous expected parallel adaptations
of those models to demands of each concrete situation.
-------
2603
The most general level of bioclimatic regulation of the area should be the
planning document covering the whole amphitheatre. Its purpose is to
provide a chance for these spaces to become bioclimatologicaly correct.
Fig. Amphitheatre Oven
During the process of bioclimatological regulation of the whole, sites with
distinctive natural conditions should be defined. After coordinating all
the requests of sito-specific planning, we could procede with more detailed
urban design stages. At that level, directives for bioclimatic development
of the space can be produced. Controled flexibility is a very important
aspect of this stage of the work. Designers here need to be conscious that
their work is creating long-lasting changes. This is the time for development
of parallel and repeated analysis adjusted to the demands of each new context.
Such distribution of the level of: ctoermination is not on the basis of the
bioclimatic architecture, only. The city is based on the same complex of
interrelated aspects of life and environment as civilization. Certain
details of the house or flat have to respond to the request of a single
stimulus. That justifies the fact that we can speak of climatic conditions
as just one among equally important aspects of the city as a whole. Going
down to the lower levels of architecture where determination with climatic
conditions arises, in certain cases,brings one to the final determination.
We are putting emphasis on this because it ensures full utilization of
bioclimatics in architecture. The necessary feed-back and self-correcting
mechanism is established and positive effepts at the macro-level multiply
as a result of the cumulative effect of quality details. At the same time,
the micro-level functions correctly, because that was made possible at
"higher" levels of planning; ie., the urban and architectural design phase.
-------
2604
SOME CONCLUSIONS
While thinking on fashion and other social trends that characterize a certain
period in history, theory mentions "the mysterious imperative" that is
forcing the whole epoch in a certain direction. Speaking about this time,
we can define, as one of the very obvious imperatives of social development,
the need to reintegrate human civilization into the whole of its environment.
The important aspects of that process are town planning, urbanism and
architecture.
The only answer to Erich Fromm's "dilemma", "To have, or to be", becomes
apparent. The task it to"redefine the very aims of Civilizationto
understand, as put by Fromm himself, that "the aim is not to control nature
but to control technology and irrational social forces and institutions
that are endangering survival of Civilization, if not human kind, itself".
The nature of our theme brought us back to the most significant ecological
problems. We think that is the proper end for this paper. The fact to be
outlined is that bioclimatic architecture, ambiental design, and ecologically
conscious design are imperatives pertinent to now and the future; a necessary
and important prerequisite for the whole range of steps to correct development
of our cities and architecture. Our contribution can be very important and
significant, but we should never forget that dealing with problems like
ecological crisis at the level of architecture will not solve the problem as
to what the origins of that crisis are.
-------
2605
OBSERVED THERMAL BEHAVIOUR OF MODIFIED TRADITIONAL
BUILDINGS IN THE NIGERIAN HOT-HUMID CLIMATIC ZONE
S.U. Egarievwe*, I.C. Ezema**, J.E. Mbamalu* and G.T. Dorgu*
~Department of Physics, Bendel State University,
Ekpoma, Nigeria
~~Department of Architecture, Bendel State University,
Ekpoma, Nigeria
ABSTRACT
The thermal behaviour of modified traditional buildings in the
Nigerian hot-humid climatic sone was investigated. The observed
effects of sand-cement, rendering, galvanised roofing sheets and
roof ceilings on the indoor temperature are presented. In
general, wrong orientation with respect to the sun contributed to
increased indoor temperature. Even though the thick earth walls
mitigated heat gain through the walls, the absence of or the use
of inappropriate ceiling materials contributed greatly to high
indoor temperature. Standard methods of orientation, shading,
insulation and ventilation are recommended for subsequent
buildings.
KEYWORDS
Thermal behaviour; modified traditional buildings; Nigerian hot-
humid climatic region.
INTRODUCTION
About 70 - 80% of Nigerians live in the rural areas (Barbour and
others, 1982; Segynola, 1987) which have a preponderance of
traditional buildings. Pure traditional buildings were mainly of
earth construction - wattle and daub, cob, pise, t.ubuli and adobe
with vegetable materials as roof covering. In recent times,
traditional buildings have undergone modifications through the
application of sand-cement, rendering on the walls and floors and
the use of galvanised iron sheets (popularly but erroneously
referred to as sine sheets) as roof covering with or without,
ceiling.
Host builders in the rural areas modify traditional buildings
with the major aim of increasing the life-span and improving the
outlook of the buildings. Highpriority is not given to the
-------
2606
attainment of thermal comfort through the appropriate use of
materials. Additionally, wrong orientation with respect to the
sun contributes to high indoor temperatures.
The problems associated with thermal comfort in tropical regions
have been highlighted by some authors (Olgyay, 1963; Ngoka,
1976). Consequently, some work have been carried out on
prediction techniques (Petherbridge, 1974) and on models of
buildings and passive solar strategies (Borda-Diaz, Mosconi and
Vazquez, 1989).
The present work investigated the effects of sand-cement,
rendering, galvanised roofing sheets, roofing ceilings (made of
raffia palm), orientation and shading on the indoor temperature
of modified traditional buildings in the Nigerian hot-humid
climatic zone. The experimental results were compared with those
obtained in nearby relatively modern buildings which have
dispersed with the traditional earth and vegetable constructions.
CLIMATIC ANALYSIS
The Nigerian hot-humid climatic zone lies within 6°N to about
10°N latitudes and 2.5°E to about 14°E longitudes. It consists
towns like Oyo, Ibadan, Ilorin, Irrua, Lokoja, Enugu, Makurdi,
Yandeu and Yola (Agarwal and Komolafe, 1933)
Two seasonal periods,with regard to rainfall, are identified in
the hot humid zone. These are the dry season which covers
approximately six months, from late October/early November to
late March/middle April, and the rainy period, which has two
peaks, one in June/July and the other in September/October.
During the dry season, the "Intertropical Eiiscontinuity Line"
(ITD) approaches the hot-humid zone. This causes the alternation
between southwesterly wind (Monsoon) and. northeasterly wind
(Harmattan). The northeasterly wind brings dust,increasing the
atmospheric turbidity (dust, water droplets, etc.), which
decreases solar radiation intensity.
Monsoon wind, coming from the ocean during the rainy season,
causes thunderstorms, cloudy sky, a drop in the diurnal
temperature range, and high water vapour content in the
atmosphere.
The temperature in this climatic zone is not excessively high but
warm due to high humidity. The mean daily maximum temperature is
30°C - 35°C and mean daily minimum temperature is 20°C - 25°C.
Humidity ranges between 70% - 80% during the dry months and 80%
- 90% in the wet season.
DATA COLLECTION
Mercury-in-glass thermometers were used to measure outside air
temperature, inside air temperature and the temperature of the
-------
2607
various parts of the buildings studied. The building parts whose
temperatures were monitored are: outside and inside wall
surfaces, roof-top and ceiling.
Three kinds of buildings situated in the same area were studied.
These are mud-house (Fig. 1) mud-house with sand-cement rendering
(Fig. 2), and modern concrete-building. The three kinds of
buildings have corrugated iron sheets as roofing sheets. Two
mud-houses, one without ceiling and the other with ceiling (made
of raffia palm) were investigated. The mud-house with sand-
cement rendering and the concrete-house have ceilings (made of
asbestors). Temperature readings were taken at one hour
intervals for some days and two hours interval for other days.
Fig. 1: Mud-house
Fig. 2: Mud-house with sand-cement rendering
-------
2608
EXPERIMENTAL RESULTS AND DISCUSSION
The hourly variations of outside air temperature Ta, and the
inside air temperatures for mud-house without ceiling T-^ and with
ceiling To are shown in Fig. 3. It is observed that the ceiling
reduces the inside air temperature during high sunshine (10.00 -
15.00 hrs. local time) and thus makes the house more comfortable
than that without ceiling. During low-temperature nights, the
ceiling also increase thermal comfort by preventing excessive
heat loses through the roof - see Fig. 3, 18.00 - 4.00 hrs local
time.
O
20
Local time (Hrs)
Fig. 3: Hourly variation of outside air temperature Ta inside
air temperature without ceiling T^ and with ceiling T2
The comparison of temperature variations for outside air Ta,
inside air Tr, outerwall Two, inner wall Tw^ and ceiling Tc or
roof sheets Tr^ for the three kinds of buildings are shown in
Figs. 4 and 5. The effect of ceiling can be seen from the lower
set of graphs, T„ and Tr. It reduces the variation of inside air
temperature. The absence of ceiling in the pure mud-house
accounts for high roof temperatures - see Fig. 4.
-------
2609
CU
Local time (Hrs)
Fig.4. Comparison of temperature variations for outerwall Tw_
inner wall Twi, ceiling T and roof sheets Trfj for (A) mud-
house, (B) mud-house with sand-cement rendering, and (C)
concrete-house.
6 10 14 18 22 2 6 6 10 14 18 22 2 6 6 10
18 22 2 6
Local time (Hrs)
Fig.5. Comparison of temperature variations for outside air T
and inside air T for (A) mud-house, (B) mud-house wit!
sand-cement rendering, and (C) concrete-house
-------
2610
To minimise interior overheating in the modified local buildings,
builders should pay attention to standard methods of orientation,
shading, insulation and ventilation.
REFERENCES
Agarwal, K.N. and L.T. Komolafe (1983) N.B.R.R.I. Report No. 4.
Nigeria Building and Road Research Institute, Lagos.
Barbour, K.M., J.S. Oguntoyinbo, J.O.C. Onyemelukwe and J.C.
Nwafor (1982), Nigerian in Maps. Hodder and Strougliton
Educational, Hong Kong.
Boarda-Diaz,N., P.I. Mosconi and J.A. Vazquez (1989) Solar & Wind
Technology. Vol. 6, No. 4, pp. 189 - 400.
Ngoka, N.I. (1976) Proceedings of the International CIB Symposium
on Energy Conservation in the Built. Environment.
Construction Press, England, pp. 253 - 259.
Olgyay, V. (1963) Design with Climate. Princeton University
Press.
Petherbridge, P. (1974) Data for the design of the Thermal and
Visual Environments' in Buildings in Warm Climate, CP8/74.
Building Research Establishment, England.
Segynola, A.A. (1987) Journal of Environmental Management Vol.
24, pp. 71 - 82.
-------
2611
STRAW BALE BUILDING SYSTEMS
David A. Bainbridge* and Matts Myhrman**
*ECOCULTURA, PO Box 51651, Riverside, CA 92517
**OUT ON BALE (un)LTD., 1037 E. Linden Street, Tucson, AZ 85719
ABSTRACT
Straw bales can be used to build low cost, durable, super-insulated, energy efficient buildings.
Many buildings built with plastered straw bales are still in service after more than 60 years.
Building systems for using straw bales have been developed for a variety of permanent and
temporary uses, these include: stacked and pinned, mortared stack, truss walls, and infill insulation
for a variety of new and retrofit applications.
KEYWORDS
Straw bale, super-insulation, energy efficiency, affordable housing, sustainable development
INTRODUCTION
Plastered straw bales are an excellent building material (Bainbridge, 1986; Myhrman, 1990).
Plastered straw bale construction apparently began in the late 1800's in the Nebraska Sandhills, a
nearly treeless area of grass stabilized sand dunes (Welsch, 1970). This method of building proved
to be economical, durable and more comfortable than either wood frame or sod houses (Welsch,
1973). Plastered straw bale construction remained in common use well into the Twentieth Century
in the Great Plains but has also been used in other areas of North America. Plastered straw bales
have been used to build houses, farm buildings, schools, hotels, government buildings, airplane
hangars, apartments, and churches. Many of these are still in use after more than 60 years.
Straw bales are still being used by builders who have discovered the benefits of this low cost and
energy-efficient building material, with wall RSI values >6 and R-values from 26-150 (Argue,
1980; Bainbridge, 1987; CMHC, 1984; Gagne, 1986; Johnson, 1990a; Myhrman, 1990; Strang,
1985). Straw bale construction can provide superior comfort and greater energy efficiency than
conventional building systems at lower cost per square foot.
The simplicity and economy of straw bale building is attractive for both developed and developing
countries. Little skill is required to build and plaster straw bale walls, and homes and buildings can
-------
2612
be completed quickly and economically. Bale building can enable the rural poor to quickly build a
house that is better insulated, more comfortable, more fire resistant, and requires less maintenance
than a much more expensive conventional house. Since the cost of labor may represent four-fifths
of the cost of the wall system and owner-builders can realize considerable savings. A bale house
was recently constructed in New Mexico for about $7 per square foot (Fish, 1988). Bale building
is also well suited for utility buildings and animal shelters (Anderson, 1989; Bainbridge, 1987;
Johnson, 1990a).
STRAW BALE BUILDING SYSTEMS
Recently completed structures demonstrate that todays builder can use several techniques to create
efficient, comfortable, long-lived structures with straw bales. Straw bales can also be used to
improve existing buildings. Further testing and research will help builders meet building code
requirements and design larger structures with more innovative designs.
The major options are: A) simply stacking and pinning the bales, B) adding mortar between the
bales (like big bricks), C) including a truss in the bale wall, and D) using the bales as non-structural
insulating infill in frame or pole buildings. The first option is the simplest and most economical.
Adding mortar between bales may be advisable when soft bales must be used. Timber or pole
frame structures with bale infill have been used to meet building code requirements. If you are
uncertain about the strength required for foundations, headers, beams, and/or detailing(consult a
structural engineer.
A. Stack and pin
Stacking and pinning is the simplest and most economical method. Most bales today are compact
and strong and can be used structurally. Bales should be set at least 8" above existing grade on a
slab or continuous concrete footing. The foundation should be reinforced to prevent cracks and
sealed on top with concrete sealer, asphalt, aluminum foil and/or plastic to prevent moisture from
wicking up into the bales. Rebar stubs projecting from the footing hold the first course of bales in
place. Metal flashing, sand barriers, a layer of diatomaceous earth (insect control not pool filter
grade) or borate may be used to provide protection from termites (Anon, 1991; Best, 1990).
Bales are normally used like bricks, laid flat and lengthwise in the wall. Accurate placement and
alignment makes it easier to plaster. Long rebar (#3 or #4) or wood stakes or dowels are driven
down through the bales at a 45 degree angle through the walls to reinforce them. Extra pins may be
used in corners and around openings.
A rot resistant or pressure-treated wood plate, commonly two boards with cross-ties, provides an
attachment point for roof rafters or trusses. Where uplift is a potential problem,allthread can be used to
tie the plate to rebar run under several layers of bales or with couplers to foundation bolts. Heavy
wire or metal ties angled from the plate to the foundation can also be used. A reinforced concrete
bond beam could be poured on top of the top of the wall for greater strength. Wiring and plumbing
-------
2613
can be run between the courses of bales or through drilled holes. Utilities can also be run in
moldings, interior walls, under the floor, or in the attic.
Cracks between bales should be filled with straw or an earth/straw mix before plastering. Most
stacked straw bale buildings have used wire mesh to reinforce a surface coating of stucco or plaster,
although the new polyfiber reinforcing may suffice. Bale walls have also been finished with earth
plasters, either plain or stabilized with asphalt. Gunite worked very well on straw bale walls
(Fiske, 1991). The pressure of the sprayer promoted a good bond with the bales. If the bales are
used structurally it may be desirable to let the bales settle before plastering.
Unplastered bales in temporary buildings may last for many years (Johnson, 1990b) although
livestock may eat unplastered hay buildings (Welsch, 1973). A four year old uncoated bale wall in
northern New Mexico looked as good as new and reports of lifetimes of up to 50 years have been
reported. Unplastered baled straw potato storage facilities commonly last 25 years (Sparks and co-
workers, 1963).
B. Bales with mortar between bales
Louis Gagne developed a mortared stack bale system where the bales are laid up with a 5 cm layer
of mortar in both horizontal and vertical joints. This requires more cement and labor. This system
was tested with support from the Canada Mortgage and Housing Corporation (CMHC, 1984).
Mortared stack plastered bales easily met building code requirements. A 4 m long, 2.5 m high wall
did not fail when loaded with 58 kN of compressive load and 3.2 kN of transverse force. The
structural consultants felt this method would be adequate for the following loads:
live loads due to use and occupancy 1.9 kN/m^
snow loads 2.5 kN/m^
wind load 0.67 kN/m^
dead load 2.0 kN/m^
The National Research Council of Canada tested the fire safety of a mortar stacked plastered straw
bale wall (CHMC, 1984). Bales easily passed the small scale fire test with a maximum temperature
rise of only 43.4°C over four hours, twice the requirement. The surface coating withstood
temperatures of up to 1010°C for two hours before a small crack developed. This study found that
this relatively complicated method of building bale houses cost less than half as much as
conventional brick veneer walls and one fifth less than a stuccoed wood-frame wall providing only
one-third as much insulation.
C. Truss frame
It is also possible to include bales in a truss frame building system. Pliny Fiske from the Center for
Maximum Potential Building Systems used a modified ladder truss to build farm buildings with
-------
2614
bales. The trusses can be engineered to meet building codes yet are still relatively easy to
incorporate with the bales.
D. Bales as insulating infill for timber frame or pole buildings.
Bales can be used to insulate a variety of structural frameworks, including post and beam, timber
frame, pole buildings, and metal buildings. Building codes have encouraged this use of an
engineered frame where the bales are non-structural. Cutting and fitting bales to fit around a frame
adds to the work and cost.
STRAW BALE CHARACTERISTICS AND HANDLING
Although baled hay has been used for buildings, baled straw is preferred. The stability and lack of
weathering that make some straws a problem for farmers make them good candidates for bale
building. Straws regarded as especially good for bale buildings include: rice, rye, and flax straw
(Bainbridge, 1986; Johnson, 1990a). Wheat, barley, and many other materials including
tumbleweed (Salsola kalil have also been used. Straw yield commonly ranges from one to more
than two tons per acre (Bainbridge, 1987).
Although straw bales are not free, they can be very inexpensive compared to conventional materials
for a comparable insulation value. Cost per bale was calculated at 170 for a large operation using a
mechanical pull—stacker in 1978 (Long, Taylor and Berry, 1978). Retail costs have run up to
$5.00 per bale in Southern Arizona. Transportation is a major part of the cost with longer haul
distances.
DETAILS
Wide overhangs are desirable. Windows and doors are often set to the outer edge of the wall.
Wood has commonly been used for rough frames for doors and windows, these are often stabilized
with wood dowels that penetrate the wall system. Concrete frames, either poured in place or cast,
could also be used. Metal headers with two pieces of angle iron and welded cross-bracing fit into
the wall system without disrupting the bale courses. If a south-facing window wall is needed for
passive solar gain, a well-insulated wood-frame can be used Trusses or conventional roof framing
can be set on a wood plate or bond beam tied into the wall. A dropped-heel truss will make it easier
to add sufficient insulation to the ceiling, RSI >6, R-30-60. The foundation should also be
insulated properly.
ENVIRONMENTAL BENEFITS
Straw bale construction can provide additional benefits in regions where straw is an unwanted
waste. In California, for example, almost a million tons of rice straw are burned each year and
these fires cast a pall of smoke over the Sacramento Valley for several weeks (Toenjas, Bell and
Jenkins, 1986). A million tons of grass straw are burned each year in the Willamette Valley of
-------
2615
Oregon (Stiak, 1989). Smoke from field burning is a health hazard and causes visual pollution.
This smoke also causes many highway accidents, and field burning in 1988 resulted in seven
deaths and thirty-seven injuries. Large quantities of straw are burned in other areas as well.
Butterfield (1985) estimates that 37% of the straw in England and Wales is burned. Lewin (1977)
estimated more than 200 million tons of straw are burned each year in the U.S. and 4.5 million tons
in Britain. Not all straw burning can be eliminated because field burning may be needed to control
pests and weeds. Straw bales are one of the few building materials that can be grown in one year in
a completely sustainable production system with favorable environmental impacts. Plastered straw
buildings can reduce the pressure on trees for fuelwood and construction material and facilitate
reforestation.
Global Warming and the Greenhouse Effect
Straw bale construction provides many benefits in the effort to control global warming and the
deterioration of the global ozone layer. First, the millions of tons of straw now burned could be
used to build affordable buildings. This would immediately reduce carbon dioxide and nitrous
oxide production by millions of pounds. In areas where rice straw is removed from the fields there
would also be a reduction in methane emissions. Straw bale building systems would enable
homeowners and building managers to dramatically reduce energy consumption for heating and
cooling. The energy efficiency of these structures can easily be double or triple conventional
construction. This would reduce fossil fuel combustion and protect the global atmosphere.
BARRIERS TO WIDESPREAD USE
The principle obstacle that proponents of straw bale building must overcome is simple ignorance.
Like many other valuable traditional practices it has been ignored by the research, development, and
education community. An education and publicity program is needed to eliminate this barrier.
Widespread use of bale building in residential and commercial construction in the developed
countries will require research to break the institutional barriers, such as building code standards
intended to regulate other materials. Although the tests by the Canada Mortgage and Housing
Corporation (1984) should provide sufficient information for many builders and building
departments additional research is needed including:
1. Inventory and evaluate existing buildings, including interviews with builders/occupants
2. Determine thermal properties and strengths of different straws and bales
3. Prepare detail and design drawings for bale buildings
4. Structural and fire tests, both short and long term
5. Test for possible problems with termites, rot, etc. and devise controls
6. Prepare specifications for handling, storage, and construction.
7. Develop building inspection criteria for code approval
8. Test new bale shapes and strategies for bale buildings.
-------
2616
ACKNOWLEDGEMENTS
A special tribute to those who have shared information on bale buildings and built them. With
special thanks to Bill and Athena Steen, Roger Welsch, Pliny Fiske, Steve MacDonald, Robert
Argue, Ken Haggard, and Polly Cooper.
REFERENCES
Anderson, B.H. 1989. The Danish solution to housing outdoor pigs. PIGS May/June 5:3
Anon. 1991. More than one way to kill termites. J. Light Construction 9(51:3-4.
Argue, R. 1980. The Well Tempered Home. Renewable Energy in Canada, Toronto, Ontario.
Bainbridge, D.A. 1986. High performance low cost buildings of straw. Aerie.. Ecosvs.. Envir.
16:281-284.
Bainbridge, D.A. 1987. Straw bale construction. In D. Andrejko and J. Hayes, eds. Proc. 12th
Passive Solar Conference. Am. Section Intl. Solar Energy Society, Boulder, CO. pp. 250-
253.
Best, D. 1990. Non-toxic alternatives to CCA. J. Light Construction 8(7~):7.
Butterfield, B. 1985. The Straw Manual. E. & F.N. Spon, London, England 212 p.
Canada Mortgage and Housing Corporation Housing Technology Incentives Program. 1984. An
Innovative Straw Bale/mortar Building System. Ottawa, Ontario 3 p.
Fish, C. 1988. Straw bale house. Sustainable Living in Dry Lands 2:2.
Fiske, P. III. 1991. Personal communication, Center for Maximum Potential Building Systems,
8604 Webberville Road, Austin, TX
Gagne, L. 1986. A Straw Bales/Mortar House Demonstration Project. Canada Mortgage and
Housing Corporation, Housing Technology Incentives Program. Ottawa, Ontario 42 p.
Johnson, D.W. 1990. Farm buildings using rectangular or round baled roughage. American
Society of Agricultural Engineers, Paper #904550. ASAE, St. Joseph, MI 10 p.
Johnson, D. 1990b. Old tricks work with new bale shapes. Dakota Farmer 108(41:34-35.
Long, J.T., W.D. Taylor & T.W. Berry. 1978. Hav harvesting costs in Texas. Texas Agric. Ext.
Service Publ. B-l 171. College Station, TX 15 p.
Myhrman, M. 1990. One man's straw bale odyssey. Permaculture Drylands J. 10:Spring
Sparks, W.C., G. McMaster, J.E. Dixon, D.W. Works & E.B. Wilson. 1963. Idaho potato
storages-construction and management. Idaho Agric. Exper. Station Bull. #410, Univ. Idaho,
Moscow, ID.
Stiak, J. 1989. Farmers playing with fire. Sierra 74(21:92-93.
Strang, G. 1985. Straw bale studio. Fine Homebuilding. 12/84-1/85:70-72.
Toenjas, D.A., M. Bell & B. Jenkins. 1986. Baler ammoniation of rice straw. Calif.Agric..
May-June:15-17. .
Welsch, R.L. 1973. Baled hay. In: L. Kahn, ed. Shelter. Shelter Pub., Bolinas, CA. p. 70
Welsch, R.L. 1970. Sandhill baled construction. Keystone Folklore Quarterly 15(11:16-34
-------
2617
BIOCLIMATIC ANALYSIS OF INDIGENOUS HOUSES IN DIFFERENT
CLIMATES OF INDIA
R. Muthu Kumar*, J. Cook*, N.K. Bansal* and G. Minke+
*College of Architecture and Environmental Design, Arizona State University, Tempe,
AZ 85287, U.S.A.
"•"Building Research Laboratory, Gesamthochschule, Menzelstr 13,
3500 Kassel, GERMANY
^Centre of Energy Studies, Indian Institute of Technology, Hauz Khas,
New Delhi 110 016, INDIA
ABSTRACT
Indigenous rural houses from India are examined, and their locations are analyzed using the modified
bioclimatic chart (Arens 1980). To illustrate how native people throughout the country are aware of
their climate, houses from three different climates are chosen for study: hot and dry (Jodhpur), cold
and cloudy (Simla) and composite (New Delhi). Climatic data are graphically presented along with
the details of the rural houses showing the climatic concepts adopted. Comfort needs are satisfied by
design, choice of materials and the living pattern of occupants.
KEYWORDS
Bioclimatic chart; climate, cold and cloudy; climate, composite; climate, hot and dry; climatic
zones, India; indigenous architecture, India; psychrometric chart; rural houses, India.
BACKGROUND
Over many centuries, building design and construction techniques have been developed in all climatic
zones, bringing forth structures that provide more or less comfortable living conditions without the
use of sophisticated technical devices. Most designers in the past were familiar with the climate in
which they were building. They were also aware cf ways by which they could benefit from certain
climatic features, and overcome those that are less favorable, merely by means of appropriate
building shapes, details, location and orientation. Additional elements, such as vegetation and water,
were also integrated into the building design to improve the microclimatic conditions. Indigenous
rural houses in India offer well developed examples of these traditional responses as they are least
affected by strict land planning and building codes.
Just as in all other climatic zones of the world, vernacular architecture in India generally exhibits
considerable ingenuity in the use of locally available materials and techniques, to produce buildings
that are well adopted to the local climate. Many such proven methods have been ignored in the
design of modern buildings, which consequently need special means for heating and cooling,
invariably incurring high costs for equipment and energy input.
METHODOLOGY
Recordings of air temperature, relative humidity averaged over a period of thirty years, for three
locations in three different climatic zones were taken from Indian Meteorological Department (1981).
The minimum and maximum values of temperature along with the relative humidity for twelve
-------
2618
months of the year are plotted on the modified bioclimatic bioclimatic chart (Arens 1980), which is in
the form of Psychrometric chart. From this chart, the bioclimatic need for human comfort is found
for each month. Representative indigenous houses from each of the three locations are chosen and
detailed survey is made to find out how people are tackling those comfort needs. Some interesting
concepts, building design techniques out of each of the three houses are presented.
STUDY AND DISCUSSION
To know how people are aware of their climate, three representative locations in India, in different
climatic zones, namely hot and dry (Jodhpur), cold and cloudy (Simla) and composite (New Delhi)
are chosen. Figure 1 gives a map of India with the three site locations and climatic zones. Table 1
shows the criteria for the classification of climates.
Table 1. Criteria for the classification of climates bv monthly means (Bansal and Minke. 1988)
No. of clear days
Mean monthly Relative humidity
temperature (°Q (%)
Climate
Precipitation
(mm)
Hot and dry
Warm and humid
Moderate
Cold and cloudy
Cold and sunny
Composite
>30
>30
25-30
<25
<25
<55
>55
<75
>55
<55
<5
>5
<5
>5
<5
>20
<20
<20
<20
>20
When six months or more do not fall within any of the above categories
Figure 2 gives an overview of climatic variables in the selected three locations. The Indian
Meteorological Department (1981) gives the values of mean maximum and mean minimum dry bulb
temperature for each month, which is averaged over thirty years, for all the three study locations.
The morning and the evening mean relative humidity values are also used
~ Simla
dBs
«®3S
ICCENO
QU HO' AH0 OUT
WARM AMD MUMO
£3 CCMPOSTt
JOOHPUR
26 IB N
9MLA
33 05' N
NEW DELHI
28 35' N
SLKSHNE H0UK (K) »
RELATTVT mumxty ($¦
mamhq jj
TOPOATUHt (c) m
—1 '—

-JY
_ rT-V

Fig. 1. Climatic Zones of India with three Fig. 2.
study locations (Bansal and Minke 1988)
Overview of climatic variables of
the three study locations
-------
2619
Table 2. Conditions and climatic needs of all the months in the three study locations
Month Hot & Dry (Jodhpur)
Well below the comfort
J temperature limits. Solar
radiation is required almost all
through the day,
W
\>
0
N
D
Below the comfort limits in the
mornings and evenings.
During the mid-day it is within
comfort limits. Thermal stor-
age is required to gain heat
during mid-day and to radiate
in the evening.
Within the comfort limits. No
additional elements are
required.
Cold & Cloudy (Simla)
Well below the comfort limits.
Solar radiation is required all
through the day. Additional
insulation is required to retain
the gained heat Protection
against cold winds.
Well below the comfort limits.
Solar radiation is required all
through the day. Additionally
insulation is required to retain
the gained heat.
Slightly above the comfort
limits. Morning and evening
ventilation needed. Shading is
required during the mid-day
Above the comfort limits.
Humidity is low. Evaporation
of water with cross ventilation
can bring significant effect,
shading required throughout
the day. Protection from hot
winds is required.
Well above the comfort limits.
Evaporation of water is the
only solution along with
thermal mass for thermal
comfort Shading is must all
through out the day.
Above the comfort limits.
Thermal mass and cross venti
lation in the evenings are re-
quired. Shading is required.
Hot winds should be avoided
Above the comfort limits.
Thermal mass and cross venti-
lation in the evenings are
required. Shading is required.
Dehumidification will be good.
Slightly above the comfort
limits. Thermal mass and
cross ventilation in the
evenings is required. Shading
is required. Dehumidification
will be added advantage.
Well within the comfort limits
Within the comfort limits.
Slightly below the comfort
limits.
Well below the comfort limits.
Solar radiation is required all
through the day. Additional
insulation is required to retain
the gained heat.
Well below the comfort limits.
Solar radiation is required all
through the day.
Slightly below the comfort
limits. Solar radiation is re-
quired in the mornings and the
evenings. Mid-day periods are
comfortable.
Within the comfort limits. No
additional elements are
required.
Slightly below the comfort
zone. Solar radiation is
required in the mornings and
the evenings. Mid-day periods
are comfortable.
Slightly below the comfort
zone. Solar radiation is
required during the mid-day
with cross ventilation to cope
with the humidity. Dehumidi-
fication will be an advantage.
Below the comfort limits.
Solar radiation is required
during the mid-day with cross
ventilation to cope with the
high humidity. Dehumidifica-
tion will be an advantage.
Below the comfort limits.
Solar radiation is required with
insulation.
Well below comfort limits.
Solar gain with good insulation
is required.
Well below comfort limits.
Solar gain with good insulation
is required. Protection against
cold winds necessary.
Composite (New Delhi)
Below the comfort limits.
Solar radiation is required all
through the day with thermal
insulation.
Below the comfort limits.
Solar radiation is required all
through the day.
Below the comfort limits in
the mornings and evenings.
In the mid-day it is within
comfort limits. Solar gain
necessary for mornings and
evenings.
Well within the comfort
limits. No additional elements
are required-
Above the comfort limits.
Low humidity. Evaporation of
water along with cross
ventilation will create comfort.
Shading required. Insulation
can also reduce heat gain.
Protection from hot winds is a
must.
Above the comfort limits.
Evaporation of water in the
afternoon along with cross
ventilation in the mornings
and evenings will create
comfort. All day shading
required.
Cross ventilation in the
evenings is required. Shading
is required during the mid-day.
Cross ventilation in the
evenings is required. Shading
is required during the mid-day.
Cross ventilation is a must,
along with low thermal mass.
Within the comfort limits.
Slightly below the comfort
limits. Solar gain in the
mornings and evenings with
thermal mass required.
Below the comfort limits.
Solar gain is required all
throughout the day..
-------
2620
RELATIVE HUMIDITY (*)
Table 3a (Bansal and Minke 19881
Jodhpur house
Reduction of solar heat gain
• by small surface-to-volume ratio, achieved by
circular plan of house
• by partial shading of wall, by overhanging
eaves and fence (on the east side)
• by avoiding openings other than the door
Reduction of heat transmission to the interior
• by thermally insulating thatch roof
• by wind breaker/ fence for protection against
hot winds
Increase of heat loss
• by having ventilated separate resting shed like
an open pavilion (used in the evenings)
DRY BULB TEMPERATURE (C)
RELATIVE HUMIDITY (X) RELATIVE HUMIDITY (X)
*WD(m/s)
SHADING
UNE
SHADNG
UNE
NEW DELHI
SIMLA
RAOIATION
(W/m2) ^
(W/m2)
10 15 20 25 30 35 40
DRY BULB TEMPERATURE (C)
15 20 25 30 35 40
DRY BULB TEMPERATURE (C)
Fig. 3a, b & c. Modified Psychrometric plot for Jodhpur, Simla and New Delhi
Table 3b & 3c. Elements and concepts adopted in the selected three houses (Bansal and Minke 1988)
Simla house
New Delhi house
Increase of solar heat gain
Control of solar heat gain
• by facing the longer side towards south, to catch
maximum sun in winter
• by providing a sheltered verandah for use on
sunny days
Increase of internal heat fain
• by utilizing heat from cattle accommodated
below the living space
• by utilizing kitchen stove as a fire place and
placing the living room between two warm
spaces (cattle room and kitchen)
Decrease of heat loss
• by low entrances to cattle room and living spaces
• by reducing exposure to cold winter breeze, with
canvas hung around the verandah, creating a
buffer zone
• by locating the house on the leeward side of hill
Dampen temperature variations
• by deciduous tree providing shade in summer,
allowing the sun to heat the house in winter
Reduction of solar heat gain in summer, decrease
of heat loss in winter
• by providing small openings
Control of internal heat gain
• by cooking outside in summer, inside in winter
Reduction of heat transmission to interior in
summer, decrease of heat loss in winter
• by thermal insulation of roof (in summer
additional hay bundles placed on roof)
Increase of heat loss in summer
• by courtyard cross ventilation
• by evaporative cooling through sprinkled water
DeCTease of heat loss in winter
• by shelter from wind due to compact design
Dampen temperature and humiditv variations
• by thick stone/mud walls
• by absorption/desorption of massive earth walls
WND(m/«)
SHADING
UNE \:
0.020
JODHPUR
0.016
0.012 1
RADIATION
(W/n>2) ^
o.ooa
0.004
roNE
450
6 )
0.000
-------
2621
J] R»jfng
PlaHofB let wotK pefi
I | Cow* I
Ground floor
West elevation
Top ti»d bf ropt
SfrOw r«d
ihatcN roof
Wootf«n follrr
Sor.ditent jJqSi at waili
th ntrfor jo nfj
Sechon - AA
Fig. 4. Architectural details of Jodhpur House
(Bansal and Minke 1988)
Second floor
First floor
Ground floor
lilllSE] ,
West elevation
South elevation
West elevation
Ground floor
South elevation
Seclion-AA
sn« f
Section-SB
Fig. 6. Architectural details of New Delhi House
(Bansal and Minke 1988)
Fig. 5. Architectural details of Simla House
(Bansal and Minke 1988)
-------
2622
Although hourly values of temperature and humidity are preferable for plotting the bioclimatic chart,
they are not available. Thus it is assumed here that evening relative humidity occurs with maximum
dry bulb temperature, and the morning relative humidity occurs at minimum dry bulb temperature.
Using these four values (mean maximum, mean minimum dry bulb temperature, morning and
evening mean relative humidity), two points are plotted and connected in the modified bioclimatic
chart for each month. Twelve lines are plotted for each location corresponding to twelve months of
the year. Based on the locations of these lines (Watson 1979) on the chart, the average outdoor
comfort condition for each month is derived Also the possible strategies to create comfort in each
month for all the three locations are given in Table 2. Figure 3 shows the modified bioclimatic plots
for all the three locations with the modified comfort zone marked on them. Even though the
bioclimatic chart is plotted for the outdoor conditions, it gives a possible indication of the strategies,
which should be adopted for creation of confortable interior spaces.
After finding the required needs for comfort, the concepts and elements adopted by the rural houses
in the selected three locations are discussed below in table 3. Figures 4,5 and 6 illustrate the
architectural details of the chosen houses.
CONCLUSION
It is very clear from Table 4 that people are totally aware of their climate and suit their living pattern
inside their house in response to the climate. The limited types of building material and technology
available to them have been used to the fullest extent. For example, material like glass with its
'greenhouse effect' is a wonderful material for heat gain, but it is expensive^ and the technology is not
available for the rural people. Hence they were small openings to avoid loss of heat, and large
verandahs open to the sun. On seeing how traditional people live in harmony with the nature,
designers in the present-day world, with the modern technology and material,can derive concepts for
building climate conscious future buildings.
ACKNOWLEDGEMENT
The authors take this opportunity to thank the funding agencies and the research team who worked
on part of this project from both the countries. The funding agencies are.
Ministry for Research and Technology, Bonn, Germany and
Department of Non-conventional Energy Sources, New Delhi, India.
REFERENCES
Arens, E., Gnalez, R., Berglund, L„ McNall, P.E., Zeren, L. 1980. A New Bioclimatic chart for
Passive Solar Design Design, Proceedings of the American Section of the International Solar
Energy Society, Oct. 19-26, vol. 5.2, pp 1202-1206.
Bansal, N.K and G. Minke 1988. Climatic and Rural Housing in India, Kernforscungsanalage,
Juelich.
Indian Meteorological Department 1981. Indian Meteorological Climatic Data Book, Indian
Meteorological Department, New Delhi
Koenigsberger, O.H., T.G. Ingersoll, A. Mayhew and S.V. Szokolay 1984. Manual of Tropical
Housing and Building, Part 1 Climatic Design, Orient Longman Ltd., Madras
Watson, D. 1979 (Ed). Energy Conservation through Building Design, McGraw Hill Book
Company, New York, Chapter 6, pp 96-113.
-------
2623
CONSTRUCTION WITH LOW ENERGY MATERIALS
R. Muthu Kumar*, G. Minke+, N.K. Bansal# and J. Cook*
*College of Architecture and Environmental Design, Arizona State University, Tempe,
AZ 85287, U.S.A.
+Building Research Laboratory, Gesamthochschule, Menzelstrl3,
3500 Kassel, GERMANY
^Centre of Energy Studies, Indian Institute of Technology, Hauz Khas,
New Delhi 110016, INDIA
ABSTRACT
This paper illustrates the experiences of hybridizing the strategies of energy efficient building
design and low cost construction techniques in the design of an office space. The site is in India, a
developing country where inexpensive labour and indigenous construction techniques are
available. The basic technology was taken from the native dwellings, modified with scientific
tools, and adopted for the design and construction. Two major principles illustrated are: a)
building materials with low energy content and b) systems designed to reduce the recurring energy
load for comfort conditioning inside the building.
KEYWORDS
Adobe; structure, catenary; earth air tunnel; low cost construction; material, low energy; passive
cooling; passive heating; soil blocks; vault, Nubian
CONCEPT
A prototype building has been constructed primarily to demonstrate the use of certain available
passive and low energy techniques. For a low energy building to be appropriate in third world
situations it should involve economy at every level starting from building materials, transportation,
construction techniques, and use of the building. Since cheap labour is available, a construction
technique that requires a high labour component is more appropriate than highly mechanized
systems. The techniques used here are very easily adoptable by small contractors and can be easily
applicable in remote villages. Then, for the life of the building, there should be a low energy
demand both in use and in physical maintenance. Lastly when the building is no longer useful, it
should be disposable with a minimum of energy cost.
BUILDING DESCRIPTION
Design Features
Location. The site of this building is on the Indian Institute of Technology campus, New Delhi,
India. It has a triangular geometry and lies behind the series of parallel academic blocks and near
the boundary of the campus adjoining the outer Ring road of Delhi. The topography of the site is
-------
2624
plain with a high silty soil. The site was originally covered by wild bushes but not trees. It is
approached by regular asphalted road for the academic blocks. A parking lot which can
accommodate 10 cars is by the side of the site. The building provides an integrated office cum
research space for a research group involved in building science research. Construction was
completed in August 1990.
Space Accommodation. The building is designed with the longer side facing North-South and the
entrance on the west side. It has a covered area of about 120 m^- The entrance of the building is
into the main hall which is a multipurpose space for discussions and exhibition (refer to Fig. 1 and
Fig. 2). The professor's room and the students' studio have access from the main hall. Other
utility spaces, laboratory and computer rooms are at the rear as a separate block connected to the
main hall. The area analysis is shown below in Table 1.
TABLE 1 Spaces accommodated in the building and their area
Function Area in sq. m.
Exhibition area and Conference room 25
Office space 22
Students work space 12
Computing equipment 18
Laboratories 18
Utility space 10
Structural Systems and Foundation. It is well known that earth as a building material can take only
compressive stress effectively. Hence the roofing system of the whole building is designed to be
spanned by domes and Nubian vaults. The main hall, professor's room and students' studio are
covered by domes whereas utility spaces along with computer room and lab are spanned by
Nubian vaults (refer to Fig. 3). The two smaller domes of diameter 3.8 m spring directly from a
sixteen sided polygon which in turn is developed from an octagonal base by corbelling at the
corners. The diameter of the big dome is 5.4 m which must be over octagonal vertical walls.
(This is the only place where reinforced cement concrete is used as a small tie beam). Even the
reinforced concrete lintels are avoided by using burnt brick corbel arches. The thiclaiesses of the
domes and vaults were designed conforming to the climatic needs and ease of construction rather
than for structural reasons which were less demanding. Surprisingly the factor of safety at the
base of the dome is calculated to be around 32. The foundations were carefully designed to take up
the horizontal thrust from the small domes and vaults and that is also one of the reasons for a high
plinth.
Studer
02.0 ,
Computer Q
18.0 O
Holl
Office
10.Q_
Kit.
' 25.0
Loborotory
/N
' ?
Office
02.0

Fig. 1. Plan of the building
Fig. 2. Section 1-1 and Front Elevation
-------
2625
Building Materials. The building material of loam or mud was chosen for the building to have a
low energy content (i.e. energy used in the production process of the material is low). Up to the
plinth level, burnt bricks are used to avoid dampness spreading to the structure. The
superstructure and the domes are built with stabilized soil blocks (stabilization with 6% cement,
made by a simple mechanical hand press, marketed by a local research organization). The mud
blocks (called adobes) for the Nubian Vaults are stabilized with only 4% cement (moulded by hand
using simple wooden or steel moulds). The exterior plastering is by mud mortar with conventional
cow-dung and sand. The water proofing is done by a spray of hydrophobizing agent which
reduces surface tension without covering the pores, and thereby not allowing the water to stay on
the plaster. Thus the 'breathing effect' of the mud wall is not lost, and the absorption and
desoiption property of the soil blocks is retained.
Construction Specification. The specifications used for various components of the building are the
following :
Foundation Plain cement concrete (1:5:10) with brick bats as coarse aggregate or filler
Plinth Burnt brick masonry with 1:4 cement mortar
Damp proof 2.5 cm thick plain cement concrete with 0.25 cm dia. coarse aggregate
Walls & domes Stabilized soil block (1 cement: 6 sand: 25 soil) masonry with soil cement
mortar (1 cement: 6 sand: 20 soil)
Vault Adobe (1 cement: 8 sand: 32 soil) masonry with mud mortar (1 cement: 8 sand:
30 soil)
Exterior finish mud-cow dung mortar (1 sand: 1 soil: 1 cow dung)
Interior finish white wash with slaked lime
Construction Techniques. The roofing systems are by domes and Nubian vaults (an old Egyptian
technique) which requires no form work for construction and thereby reduces construction cost
dramatically. The geometry of the domes is not semicircular like a conventional dome but an
inverted rotational catenary which is an extremely stable configuration. The effectiveness of the
rotational inverted catenary is mainly based on the profile and hence a special steel guide, which is
pivoted at the centre of the domes, has been fabricated for controlling the profile with the accuracy
in millimeters (refer Fig. 5). For the Nubian vault also a steel template, fabricated out of
reinforcing bars, are used to transfer the geometry uniformly over the wall.
-"Si
a
Fig. 3. Section 2-2 and Side Elevation Fig. 4. Section 3-3 and Rear Elevation
-------
2626
Profile of the domes and Nubian vaults. Since the profile of the structural domes and vaults are
crucial, a more accurate mathematical expression has been used in a computer program to arrive at
the co-ordinates of the curve. The mathematical expression used is
Y = a Cosh (X/a)
where, 'X' and 'Y' are co-ordinates in X and Y axis respectively
'a' is the Y co-ordinate of the lowest tip of the catenary symmetrical about Y-axis
and Cosh is the hyperbolic trigonometric function
The value of 'a' can be derived for a given height and width of the catenary.
Tests of Materials. Initially extensive soil tests for grain size distribution were carried out to find
the clay content. From those results stabilized soil block samples and adobes for different
percentage of stabilizers were made and tested, for both dry and wet strength at the end of the 7th
and 21st day. These tests are done based on the procedures given in building codes of India. After
arriving at the workable composition of mixture for making soil blocks and adobe, samples were
tested from every major batch to maintain the required dry strength of 32 kg/cm2 and wet strength
of 7 kg/cm2- For plaster, different samples were tested for different proportions of soil, cowdung
and sand to arrive at the workable consistency.
Design Difficulties. The semicircular arched windows in the rooms with domical roofs have to be
constructed first since the arches pierce through the domes. The conventional method of working
drawings was not sufficient to visualize the connection between two curved surfaces. Also in the
steel guide the curved steel profile to match the profile of the dome had to be designed as two
pieces with a special joint since the arched doorways were intruding into the domes.
A dome is stable if it is complete on all 360°. But in this building the doors and windows have to
puncture through the domes and vaults, which challenge its stability. To solve this a secondary
curve or arches are used to transfer the loads effectively.
The joints between two composite material are always vulnerable to cracking due to the thermal
shock and the water that tries to enter by capillary action. So the base of the domes, not only
contains a layer of burnt brick but also a sloping lining of special tiles over which the mud-
cowdung plaster of the dome is continued. The water collected between the Nubian vaults is
drained with sufficient slope. The joint of heterogeneous materials, in the valley formed between
the vaults, is avoided until it touches ground well away from the wall
190 mm
-------
2627
ENERGY EFFICIENCY
Various Phases
Production. Normally a burnt brick takes about 2 kWh of energy for production while a sun dried
brick takes nothing if it is made from the site soil. The industrially produced materials like cement
or steel take energy several folds. In this building these industrial materials are kept to a minimum.
Calculation establishes a saving of about 60,000 kWh in the production of building materials.
Transportation. Transportation costs are reduced tremendously because compacted stabilized soil
blocks and stabilized soil adobes are produced on site from the soil excavated from the site itself.
The mortar used for the compacted blocks as well as adobes also use the site soil. The quantity of
the burnt bricks used is very small (only up to plinth level).
Construction. Construction techniques used here are highly labour intensive and avoid the use of
expensive formwork. The techniques used here involve only human energy and almost zero
mechanical energy for construction.
Recurring Energy for heating and cooling. The climate of Delhi is Composite in nature i.e.. both
heating and cooling is essential in different seasons. The maximum temperature is around 45 C
(120 F) in summer while minimum temperature in winter drops to 4 C (40 F). Energy
efficiency for space conditioning is achieved by
- proper orientation for wind
- proper size and position of fenestration
- appropriate design of overhangs and shading screens
- control of moisture in the building by suitable walling material and its thickness in
specific periods of the year.
- specially designed earth-air tunnel system (refer to Fig. 5).
Earth-air-tunnel system. Since the climate of Delhi is composite in nature and the yearly average
ground temperature is in the comfort range, the thermal storage effect of the earth is used in this
system. The ground temperature is warmer in winter and cooler in summer when compared with
the air temperature. A calculated length of stone-ware sewage pipes are buried under the earth
surface, about 3.8 m deep. Acute bends are avoided and the tunnel feeds into the major spaces
coinciding with the direction of the wind so that distribution of the air from the tunnel inside the
room is uniform. Ventilation grills are placed at the supply outlet in the room at floor level. The
exhaust of air is through the conventional window openings on the leeward side of the rooms. The
joints of the stone-ware pipes have to be water tight to avoid any moisture penetration in the pipe
which reduces the possibility of fungal growth. The depth of the pipes, length of pipes and their
diameter depends on soil conditions, material of the pipe, design air change rate and the design
indoor condition. It is planned to even run the sensitive computers without any additional air
conditioning systems.
Davlighting. For all interior spaces enough daylight is provided by openings of suitable size and
orientation. Specially for working areas, diffusing skylights are provided to create emotional effect
as well. The distribution of the daylighting is enhanced-by white washing of the interiors with
slaked lime. It is found that the skylights provide about four times the lux levels for the same area
of opening when compared with side openings. Since the area of the opening is small the solar
heat gain is less.
Ventilatioa The fans (2 numbers at 400 W each), which are switched manually are provided for
the earth-air-tunnel system to draw air through the underground tubes to heat or cool the air
according to the season automatically. Additional openings are provided around die skylights for
hot air exhaust by stack effect.
-------
2628
Energy Saving
Energy Saved from materials production = 60,000 kWh
(This is not considered in the life cost of the building)
Energy Saved from heating/cooling = 45,000 kWh / year
(Cost of operating an air conditioning system)
Energy spent on fans of tunnel system
(0.800 kW* 10 hr/day*300 days) = -2,400 kWh
Net energy saved = 42,600 kWh
@ of Rs. 1 per kWh cost saved = Rs. 42,600
@ 10 % interest rate net cost saved per year = Rs. 21,600
Pay back period is = 9.7 years
COST SAVING
To give an idea about the construction cost saving from this building, a comparison is made with a
typical office space in the locality without air conditioning and conventional specification. This
building costs about Rs. 210,000 (US$ 10,500) while the conventional building costs Rs.
270,000 (US$ 13,500). Apart from direct cost saving, indirect saving in terms of comfort is also
achieved.
MONITORING
With the equipment available to the research group which is going to occupy the building, it is
proposed to monitor the building closely for a period of one year.
SCOPE OF IMPROVEMENT
To understand these type of structures, a three dimensional physical model using small scaled
bricks is recommended. This seems like the best means to clarify the dimensions of the joints
between the surfaces of the semicircular arched window and the dome. Cost can be reduced
further by reducing the number of arched openingintruding the dome. Windows are easier to
build in the vertical wall. In a new building technology a different management of labour is
required for construction. If the diameter of the domes is in a standard module then the
prefabrication of the steel guides can be omitted. If proper soil is found nearby the site the
percentage of stabilizers used can be reduced.
CONCLUSION
The design of this building successfully demonstrates the use of low-cost, low-energy locally
available materials and the construction techniques that can be easily adopted by the local masons to
create comfortable low-cost appropriate shelter. The techniques applied here have a lot of scope
for the actual field application in villages all over India. This is obvious from the fact that
HUDCO (Housing and Urban Development Corporation) is planning to take up this technology for
health care and primary school projects in remote areas.
ACKNOWLEDGEMENT
The authors take this opportunity to thank the funding agencies and the technicians and engineers
who worked in this project from both the countries. The funding agencies are
German Technology Centre, GTZ, Frankfurt, Germany and
Indian Institute of Technology, New Delhi, India.
-------
2629
SOLAR VILLAGE
by A.N.Tombazis
85, Marathonodromou street
GR-154 52 Psychico - Greece
project management: Solar Village Co S.A.
project manager: C.Kanaris
architectural design: A.N.Tombazis and Associates
energy consultants: V.Loftness, architect and D.Daskalakis E/M engineer
E/M design: LDK - L.Damianides - D.Kirimlides
supervision: N.Pandis - A.Tombazis
energy systems' design and evaluation: INTERATOM GmbH, Germany
passive systems' evaluation: Prof. M.Papadopoulos
active systems' evaluation: Prof. B.A.SotlropouIos
social study: Prof. B.Joerges - Z.Theos
ABSTRACT
Solar Village No 3;built near Athensjs the result of a joint Greek-German project in the field of low
temperature solar energy utilization. Design began in 1979 and construction was completed by
1988. A community for 435 families has been provided with solar energy and energy conser-
vation strategies. Various systems have been applied, with the main scope to be evaluated from
different points of view, as efficiency, applicability, architectural implication, public acceptability,
etc. Monitoring of the different systems is now under way and some first conclusions may be
drawn.
KEYWORDS
Energy conservation, solar energy, active systems, passive systems, monitoring, evaluation.
INTRODUCTION
Solar Village No. 3 built near Athens is the result of an intergovernmental agreement made in
1978 between BMFT of W.Germany and the Greek Government as a joint Greek-German project
In the field of low temperature solar energy utilization. A community for 435 families has been
provided with solar energy and energy conservation strategies. The execution of the project was
undertaken by Solar Village Company, a Greek public-owned company which was created spe-
cially for this purpose. In April 1989,the houses were awarded by lottery to those eligible for sub-
sidized housing from the Workers' Housing Organization (O.E.K ) Monitoring of the different
systems is now under way.
The site, located in the suburb of Pefki, 18 km north of central Athens, has an area of ap-
proximately 7,2 ha and an overall density of about 220p/ha (fig. 1). A mild climate prevails in the
area with a need for heating in winter (1180 degree days).
The primary aim of the project was to reduce oil consumption. Heating needs are reduced by
appropriate architectural design and passive solar measures on the one hand and optimized
non-conventional active systems on the other. Natural ventilation is an important feature in the
project. The heating systems applied have the objective to reduce energy consumption by using
solar energy In direct or indirect form. Various systems have been applied, with the main scope
-------
2630
to be evaluated from different points of view, as efficiency, applicability, architectural implication,
public acceptability, etc. The basic energy conservation design for SV 3 saves an estimated 65%
of conventional energy consumption. Solar heating, then, meets over 70% of the remaining an-
nual load, resulting in housing that consumes only about 10% of traditional Greek construction.
Fig. 1. Aerial view
Fig. 2.Central square
435 apartments are provided in three sizes of 60-80 and 100m2 with one, two and three
bedrooms respectively. A number of units are organized as two-storey row houses. The remain-
ing units are developed as apartment blocks ranging from three to six storeys. Nearly all build-
ings are oriented due south with few exceptions to break the monotony. Central functions are ar-
ranged around the main square (fig. 2) which is sunken in order to create intimacy and a sense
of belonging. Included are a small shopping centre, a cafeteria, an information centre, a multi-
purpose hall, a library and an energy centre. Due to increased insulations and double glazing,
heating loads were substantially reduced to approximately 200W/°C (2.2W/m2, °C) per unit.
401 apartments have been provided with active solar systems for DHW (domestic hot water) and
SH (space heating). 34 (22 houses and 12 apartments) are considered as highly passive, since
their heating load is covered by passive solar systems although it could be stated that in es-
sence, even if not called so, all buildings are passive (fig. 3 and 4).
Fig. 3- Passive solar houses
Fig. 4.Two-storey passive solar house
-------
2631
The entire village is organized in six housing regions, the energy centre and the community
centre considering the solar system applied and the degree of centralization (fig. 5).
Central heat pump
Highly passive
Air collectors
Decentralized heat pumps
Fiat plate collectors
Interseasonal storage
Energy Centre
Community Centre
Fig. 5, Solar Heating Systems
A Data Acquisition System (DAS),organized by Interatom.records data on a 24 hour basis on five
and 30 minute intervals. The system which consists of about 1500 points of measurement lo-
cated in the Energy Centre and various critical points in the Village is linked with 25 processing
units and a central computer. Measurements are taken both in vacant and occupied apartments.
The system records climatic data from the meteorological station, air temperatures of rooms, sur-
face temperatures, heat flux, hot-water consumption and electric power consumption for heating
and other uses in the Village.
Two teams of scientists from the University of Thessalonikl and one from Interatom in Germany
are analysing the results in order to draw conclusions for the use and evaluation of the systems.
In Thessaloniki, the first research team, lead by Prof. B.A.Sotiropoulos, Is dealing with the active
systems, whereas the second, lead by Prof. M.Papadopoulos, Is dealing with the passive sys-
tems. The fierman team is monitoring the active systems. The Measurement and Evaluation
Phase has been divided into three periods. The first one covers the period when no system was
operating yet and the necessary preparatory work was being done. In the second, the systems
were operating, but in empty buildings for nearly one year, so as to test them and deal with any
unexpected problems. Finally,the third was ¦ when systems were operating and buildings were
occupied. In this last and most important period, measurements will be taken for another year in
order to draw some scientifically valid results for the various heating systems.
Some unforeseen problems,technical or other,did occur, delaying the evaluation work. Therefore
only a few first results may be presented here.
Further, to the technical and cost evaluation of the project,special emphasis is being given to the
social aspects of the experiment, with a group of sociologists from O.E.K. assessing the quality
of social life and environment In the community, the attitude of the Inhabitants to the energy sys-
tems and the Village as a whole.
-------
2632
GENERAL OBSERVATIONS & CONCLUSIONS FOR HIGHLY PASSIVE REGION B
The highly passive region B includes six different solar systems: direct gain and water bench,
direct gain and trombe wall or water wall, sunspace and trombe wall, sunspace and water wall,
thermosiphoning air panel and air loop plenums, sunspace and air loop plenums in concrete
floor.
Independent passive water collectors provide DHW to all houses in the region.
During the last measurement period, when the Village was inhabited, apart from the DAS
measurements, thermographies were taken to determine the thermal behaviour of passive ele-
ments in houses and apartments. Some first results are listed below:
- room air temperatures remain in the comfortable or still comfortable zone even during typical
cold February days, without any backup heating (fig. 6)
UAKOB
UNPLEASANTLY
FEB.90
UNPLEASANTLY
COLD
[
UB0O2A2
FEB.90
I I WON C0MF-
I I STH.L COKF.
E3 COMFORTABLE

UNPL
EASAS
TLY
HOI

---

\
V

UNPLEASANTLY
COLD |

AIR TEMP. (°C)
AIR TEMP. I°C)
Fig. 6.Thermal comfort diagram (Feb. '90)
- external insulations of the examined buildings are very efficient as no thermal bridges occur
from walls
- reinforced concrete parts of the buildings (beams-concrete walls) present no thermal bridges
- even distribution of insulation on the roof provides uniform indoor temperatures in the upper
areas, with no penetration of humidity which would also cause thermal bridges
- some losses occur at construction joints and joints between apertures and building elements.
The problem is more acutfe in north-facing entrances and could be solved by external wind
protection
- some occupants do not use passive elements properly, with the result of poor heating per-
formance of the apartment
GENERAL OBSERVATIONS & CONCLUSIONS FOR THE ACTIVE SYSTEMS
Results for the performance of the active systems are more difficult to draw, mainly because of
the complexity of each system. Nevertheless, some first conclusions for the occupied period are
listed below for each region.
Housing Region A. Community Centre H. Energy Centre G
In winter 1989-1990,the system in Region G operated from 1.11.89 to 31.3.90, i.e. the heating
period was of 3624 hours. The recorded operation time of the system was 2194.5 hours, or
60.6% of the heating period time.
-------
2633
The partial operation of the DAS is mainly due to manual shutdown, electric current interruptions
and other unavoidable problems.
The recorded average ambient temperature ot the heating period was 13.35 °C. The average in-
door temperature in the "reference" residences was 20.47 °C.
The recorded useful thermal energy supplied by the heat sources was 595389 kWh. 0.44% of
this energy was supplied by the heat pump, 0.13% by the co-generation plant and 0.43% by the
boiler.
The average Primary Energy Ratio of the heat pump was 1.73. The electric, thermal and total
Primary Energy Ratios of the co-generation plant were 0.33, 0.42, 0.75, respectively. The
average Primary Energy Ratio of the system in the Energy Centre was 1.3.
The average Primary Energy Saving of the heat pump in comparison to the boiler was 47%, i.e.
132306 kWh of fuel energy saved. The Relative Operating Energy Cost Saving of the co-
generation plant was 32-45%.
Region C - Air Collectors
In that region only one of the three houses connected to DAS was occupied during the heating
period.
The energy for SH (Space Heating) was only 18% of that predicted In the design phase. But, as
far as Indoor temperatures are concerned, the contribution actually offered in the house, includ-
ing the passive solar gains and the internal sources, raises the offered energy up to 71-73%
(figure 4).
On the other hand, the low Indoor temperature raises the question about the efficiency of the
control system. This is due to the Inaccuracy of the room thermostats which, although set to
20 °C, do not activate the auxiliary heating, when the indoor temperature is lower. It is important
to know, however, that no severe complaint has been expressed by the occupants against the
SH System. This could be attributed to the "warm walls" effect resulting from the well insulated
exterior surfaces, which finally produce a feeling of comfort despite the relatively low air tempera-
tures.
Region D - Decentralized Heat Pumps
The SH system was in operation from 11 November to 27 March. The great loss of data that oc-
curred during this period, especially in December and January, has prevented an accurate
evaluation of the system's performance. However, from the analysis of the overall system's be-
haviour during that period, one can conclude the following:
The heat pumps' performance factor lies in a range of 3.40 in December to 3.92 in March and is
higher than the design prediction.
• The SH mixing valve adjustment was in com-
plete disagreement with the design setting
during most of the heating period. This led to
low SH feed temperatures, especially at ambient
temperatures below 8 °C, and consequently to
low indoor temperatures. Nevertheless, the
mean air temperature of the building for most
days with sufficient available data was above1
19.5 °C.
- The total heat losses of the system have
been estimated as 4% of the SH load.
- The auxiliary energy consumed was higher
than that of the design prediction. The contribu-
tion of the boiler to the SH load was 50.3% for
the period of November to March, compared to
the 31% of the design prediction for the same
period.
- The performance factor of the overall system
for the whole period can be seen In figure 7.
25000
20000
15000
10000
5000
OCT NOV DEC JAN FEB MAR APR
Energy input
Energy to building
Performance factor
~
Fig. 7,
Performance factor of SH system
(Building UDB)
-------
2634
Region E - Flat Plate Collectors
A first general remark is that the actual amount of solar energy available during each month has
been constantly higher than that predicted by the design.
The collector field efficiency was between 16% (August) and 37% (December) with a mean value
of 26% for the whole period. This value, which is far from the expected 42%, is mainly attributed
to the generally low efficiency of the SET-vacuum tubes with boosters type of collectors.
Recirculation losses of this system lie in a range of 220 to 1450 Kwh/month with the lower values
encountered as expected during the warmer summer months. The amount of auxiliary energy
for DHW was constantly higher than the design values even during the summer months. The
DHW storage efficiency was 79% for the whole period and that of the overall DHW subsystem
30%, while the design value was 81.5%.
It seems that the heat gains (solar, internal) had been overestimated in the simulation calculation
of the load during the design period. The actual SH load illustrated seems to be close to the
theoretical heating demancf in spite of heat gains by persons and sun radiation through the win-
dows.
The solar fractions finally achieved as far as the DHW subsystem Is considered, were in general
of the same order as those expected by Interatom's calculations.
Region F - Interseasonal Storage
The annual Interseasonal Storage
(I.S.) efficiency was high (0.766). The
I.S. total annual heat losses were 17.2
MWh which represents the 23% of the
total annual energy transferred from
the solar collectors subsystems to the
I.S. tank. During long periods the
heat losses of the I.S. tank could not
be calculated due to the great amount
of missing data and to erroneous
values of the temperature sensors.
Nevertheless it seems that the ap-
proximation of the Heat Transfer
Coefficient (KA Value) 90 W/K for the
non-occupied period was far from the
real value. In the term "heat losses of
the I.S. tank" the heat losses of the In-
terseasonal Storage to Space Heating
Heat Exchangers (IS/SH HE) are in-
cluded. The main heat output from
the IS subsystem was to the Space
Heating subsystem while only 10.1 MWh were transferred to the IS/DHW HE. At the end of
January the temperature of the IS Tank was near 30 °C and so the IS subsystem could not con-
tribute significantly to the Space Heating demands. The SH subsystem was working sufficiently
(efficiency: 0.903).
CONCLUSIONS
It would not be appropriate at this moment to draw general conclusions for, as already men-
tioned, the monitoring of the whole project is still continuing. However, one could say that the
overall results so far are very positive.
Q202b
IS to SH
Fuel
Q8 04 b -1
Heat Los.
of SH j
Tank V
Q2 02 db
Aux.
Q 700b
SH Load
Heat Quantities
in MWh
Fig. 8. Energy flows in the SH subsystem
-------
3.5 Vernacular Architecture II
-------
Intentionally Blank Page
-------
2637
SOLAR ARCHITECTURE IN FRANCE
Eric DURAND, Architect
FRENCH AGENCY FOR ENERGY MANAGEMENT
500, route des Lucioles
Sophia-Antipolis - 06565 VALBONNE CEDEX
FRANCE
INTRODUCTION
The aim of this paper is to illustrate, by review of several projects, the
status of passive solar architecture production in France during the 1980's.
A recent census made by this author on solar architecture, including active
systems, gives the approximate number of 2000 solar structures built each year.
BUILDING TYPES
Single-Family Houses: These projects can be achieved profitably by a meeting
of the minds between a client and an architect, i.e. when one of them con-
vinces the other to build a solar house.
Public Housing: Many public offices in France have made the committment to
build passive or active solar dwellings, with more or less success. Two of
them, Aude and Dr&me, have managed to build solar housing exclusively. The
main objectives of these public offices is to offer the best quality of life
and comfort for the tenants with the lowest cost and reduced heating requirenents.
The Non-Residential Sector: For this presentation, examples of passive
solar buildings for which that architecture specifically achieves reduced
energy costs were chosen.
CONCLUSIONS
The main conclusion and information which this author wishes to impart is this:
passive solar design is possible without any extra cost both for housing and
for non-residential buildings. A good design including passive solar features
shows that a review of all parameters of the design will allow for the "payment"
of a solar feature, e.g. the sunspace or solar collectors for hot water, etc.
Another conclusion is that these dwellings save 20 to 50% of the energy con-
sumption of a standard design, with the high satisfaction of the inhabitants.
And finally, this author's most important point is that passive solar
architecture concerns everyone. Yes, sunspace, the device of solar architecture,
is for the people.
PRESENTATIONS
The following pages present six solar projects.
-------
2638
CHATRENET HOUSE
Architect: Mr CHATRENET
This house had the first price of a solar architecture competition (it is locate! near Toulouse (South-West of
France)).
This house is a very good example of a synthesis from the passive solar techniques :
° A Nor,h- Wesl buffer space to protect the house from the strong winter winds,
o A very closed North facade.
o A large South East and West sunspace which has a very efficient solar protection for summer irradiation
(opaque roof and several apertures for natural ventilation).
This house is built with local materials such as row earth bricks for the South walls. The other orientations
walls are eomponed with standard burnt bricks.
pie inertia of the house is developped throuh an East-West axis. Cheap to construct, this very comfortable
house in any season has also a low energy heating load.
vtum
REFERENCES: SYSTEMES SOLAIRES N* 53-54. Dossier AFME/CAS
-------
2639
MAISON MONORGUES : Improvement Architecture
Architect: Mr ALASSEUR
Located near the sea in the Bretagne Region.
It was a funny old house; it is now a pleasant comfortable passive solar new house and also very cheep for
heating...
The architect imagined a complete improvement and enlargement of the existing house with passive solar
features:
o Awide sunspace facing South and East used for pre—heating of the new—air.
o The good use of insulation,
. internal insulation, cheep (North),
external insulation to bring inertia and arrange thermal bridges.
The enlargement has a timber frame structure.
Auxiliary heating is a condensation gaz boiler.
h
~a
REFERENCES: SYSTEMES SOLAIRES N* 53-54. Dossier AFME/CAS
-------
2640
PUBLIC HOUSING - LA BAUME D'HOSTUN (DROME)
Architect: Mr CHAMBAUD
This project is an arrangement of 5 attached houses.
Each house composed with two levels is designed with passive solar standards.
Completely closed North facade, very glazed South elevation with a large sunspace arranged in the two floors.
Downstairs, it is an enlargement of the living room, upstairs, a conservatory, buffer space between bedrooms
and outside: the preheating of new air through the sunspace and other solar or insulation features allows a
"43 %" energy winning compared with standard housing (solar contribution + 35 %)'!
This programm is one of those engaged by the public office of the Drome who build more than 250 passive
solar dwellings since 1988. The "Aude Office" leader of bioclimatic housing programm has built
approximatively 1 000 dwellings since 1981.
|oo|
u
Fa9od*
Nord
REFERENCES : SYSTEMES SOLAIRES N* 53-54. Dossier AFME/CAS
-------
2641
COLLECTIVE HOUSING - FORMJGUERES
Architects : Mr WURSTEISEN and Mr KNYSZEWSKI
o Location : Pyr6ndes-Atlantiques - Altitude: 1 500 m - 2 720 sunshine hours.
This block of building is composed of 6 dwellings and the offices of the state police local force. It has been built
without any extra cost. The solar architecture of this new building is in complei accordance with the traditional
village.
Most of the energy demand is assured by renewable energy,
o Each flat possess a chimney, an insert, or a wood boiler.
o A solar hot water system with 2sq meters per flat; auxiliary energy is electricity but with very low
consumption.
The design of this building is rearly a perfect pasive solar example.
o Buffer spaces are arranged in the North. The North facade is very closed and no main rooms are facing
North.
o A double floor sunspace is integrated in each flat with the pre-heating of new air (natural or with fan),
o All the main rooms, living-rooms or bed-rooms have windows on South-East or South-West
orientation.
o The summer comfort is good thanks to the opaque roof of the sun spaces, roof eaves, shutters and the easy
South-East ventilation. The good insulation and the heavy inertia also contribute to the comfort.
5| , /M
iO\
f ,"¦
fcj1 atsfis
mm
* N
Wf

REFERENCES: SYSTEMES SOLAIRES N' 53 -54. Dossier AFME/CAS
-------
2642
SOLAR GYMNASIUM OF SAINT-LUPICIN (JURA - FRANCE)
Architect: Mr JACQUIER
This gymnasium has been built in continuation of a near high school:
o The South facade recieves the solar air collectors,
o The North facade has two apertures for day lighting.
o The East facade possess a large glazing area as well to recieve the early solar gains in mid-season and also
for the marvellous external view.
o The West facade is nearly closed to prevent over heating as the gymnasium is used most often in the
evening.
Day ligting was carefully studied. It is homogenous and very well distributed.
o Large glazing rays in the South (between the collectors) for lighting and passive heating,
o North apertures for diffuse lighting.
o The East gain aperture allows diffuse lighting which penetrates deeply in the gymnasium.
The active solar system is composed of 168 sqm of solar air panels connected to a 60 m3 pebble storage.
Auxiliary heating is assured by electricity (hot water storage during hoff-peak hours).
With high insulation and the solar systems, the gymnasium is very economical in energy.
During the first year monitoring solar gains were UK) % of the sport hall demand !
WEST FACADE
SOUTH FACADE
-------
2643
SALLE POLYVALENTE DE SAINT-JULIEN-SOUS-MONTMELAS
Architect: Mr ROCHE
This multi-purpose building located in rural area has been designed with energetic, functional and poetic
ideas.
The passive solar design is based on direct gains and insulation :
o 60 sqm of double-glazed area on the South facade,
o Insulated motorised rolling shutters.
North buffer spaces are arranged to protect from the dominant winds.
The structure is armoured concrete of the walls and glued laminated frame work.
Auxiliary heating is assured by gaz boiler.
The energy saving by solar gains is approximativcly 40 % heating load.
ft
Fofod>

1
r
SALLE 2 a
SALLE 1
bu<*»ou
cuts.
Fo?od*
REFERENCES : SYSTEMES SOLAIRES N° 53-54. Dossier AFME/CAS
-------
2644
PASSIVE HEATING IN THE HIGH ANDES
R. Romin L., R. Corvalan P., R. Tala D.
Department of Mechanical Engineering
University of Chile
Casilla-2777 Santiago - Chile
ABSTRACT
In this paper we report on the genera1 design, performance and model results of a passive
solar house built in the high Andes at 3600 m above sea level. The house has a 25.3 m2
Trombe wall and a built surface of 63.5 m2. Spot measurements gave a solar contribution
of 3500 MJ for October and 2100 MJ for November, in close agreement with the numerical
model.
KEYWORDS
Solar energy; passive heating; Trombe wall; passive heating modelling.
INTRODUCTION
The Atacama desert in northern Chile has exceptional conditions for the utilization of
solar energy. In a small geographical region one has average KT values of over 0.7, more
than 300 clear days a year and also many places that are important energy consumers.
In 1979 CODELCO-Salvador, the Chilean Copper Corporation, decided to build a new
house at its industrial water source (La Ola). The house, built during 1981 is 100 km
east of El Salvador at 26° 27' S., 69° 04' W. and 3600 m above sea level. The house
has a heating system consisting of a modified Trombe wall with coal stove backup, and
electricity is provided by an array of eight photovoltaic panels. The design of the solar
heating system was made by our group, and general architecture as well as construction
was made by CODELCO. In this paper we will discuss the general aspects of the design,
and its performance as well as that predicted by a numerical model developed to test
several design ideas.
The emplacement at La Ola is on the fringe of a salt flat. Climate is very extreme, with over
5600 degree-days (Centigrade) of heating a year, daily minimums of — 20 C and average
windspeeds of over 4.5 m/s. Solar energy availability is high, with monthly average KT
values from 0.63 to 0.75. Important aspects to consider in the design are that heating load
is present throughout the year; there is high solar availability, and that the low latitude
(—26° S) posed an interesting problem in a system that had to be passive.
SOLAR HEATING SYSTEM
The house has a built surface of 63.5 m2, an interior volume of 141.5 m3. The reinforced
concrete Trombe wall has a collection surface of 25.3 m2 is 25 cm thick (average) and due
to the fact that there is an important heating load during "summer" (December through
March), the wall had to be built with a 70° tilt from horizontal. All window glazing
is double, ceiling insulation is the equivalent of 100 mm expanded polystyrene and the
-------
2645
original design considered 25 mm exterior wall insulation. A simplified cross section of the
house is shown in Fig. 1. Calculated overall heat loss factor is 1.1 W • m-3 • K_1 with
exterior wall insulation and 1.6 without.
In the system, air flows between the wall and exterior glazing, penetrating the house along
the top through non-return plastic film vents, flows through the habitable space, and
returns to the wall under the flooring and in contact with a considerable mass of large size
boulders. The wall also has three 90 x 90cm windows that permit some direct gain, and
heating backup is in the form of a wood and coal stovcthat is principally used for cooking.
The house that was replaced used the same stove, annual fuel consumption was 20.000 kg
and minimum interior temperatures were frequently below zero.
Qcoir
.Gem
Qcok
Q«
Qcsit
Qtoi
«3|ir
Qcoir
Ore
Fig. 1. Cross Section of La Ola Solar House
SYSTEM MODEL
The house was modelled as a three-dimensional system with conductive, convective and
radiative (both short and long wave) heat exchange with ambient and other elements in the
system. We applied a finite difference method (F. Kreith, 1973) and reduced the system
to a. 54 element nodal netwqfk. Each principal building element was reduced to a sufficient
number of nodes, conductive heat flow was considered unidimensional, and,on the surfaces,
both convective and radiative heat exchange was considered. In this paragraph we give a
brief outline of the principal characteristics of the model.
Radiative Heat Exchange
Short wave
The program automatically generates, for any latitude, time and date direct, diffuse and
reflected solar radiation. Global energy on any plane can be generated from these com-
ponents. Direct radiation is estimated using Hottel's method (1976), and diffuse from the
relationships proposed by Liu and Jordan (1960):
with:
h
T
= T • I„
= A0+A1 .eK/c°'z
(1)
(2)
-------
2646
L = U ¦ I0 (3)
rd = 0.271 - 0.2939 • t (4)
in the case of cloudy days, radiation was considered totally diffuse, isotropic, and we used
the following empirical equation (Munoz, 1980):
Hr = 30.33 • EXP(3.4 • sin h) W • m" 2 (5)
for
0.35 < Kt < 0.55 and h— solar altitude
For energy transmitted through glass, variable transmission was considered according to
incidence angles and glass properties; energy reflected from the ground varied according
to albedo.
Long wave
We have two types of long-wave radiative interchanges, those between a surface and the
sky and ground, and those between different elements in the house itself. Sky temperature
was estimated from Bliss (1961) as 15 K lower than ambient. For radiative interchange
between two surfaces 1 and 2:
Q1-2 = Ai • F^-2 • a • (T* ~T%) (6)
were Fj_2 is the equivalent conductance and takes into account both emissivities and form
factor of the surfaces under study.
Convection
The model takes into account convection between wall surfaces and ambient, as well as
within the Trombe wall. For the first case we used classical relations, and in the second case
we relied on experimental values our group obtained earlier (Czerny and Longas, 1982).
These values correlate the Nusselt number versus Rayleigh according to the following
general expression:
Nu = c¦ Ram (7)
The values of c and m depend on the Ra number and the symmetry of heating inside the
wall. For a typical aspect ratio of 0.036, and for asymmetrical heating, c = 0.024 and
m = 0.819.
Conduction
The general heat conductive transfer equation is:
d dT, d ,, dT, d <37\ . „ dT
diiK + Ty[K d^] + dz{K + 9 ~ °E 'e " ~dt (8)
Supposing unidimensional heat flow, and that for a given element, heat transfer properties
are homogeneous, then this equation can be reduced to:
-------
2647
General operation of the model
Using the basic methods described in the previous paragraphs, and taking into account
the physical properties of the materials, it's possible to simulate the thermal behavior
of the house for any period and any sequence of clear and cloudy days. The program
incorporates a "thermostat" that connects and disconnects the auxiliary heating. Both
the setpoint of this "thermostat" as well as its differential are adjustable, thus permitting
wider temperature swings and better use of the thermal mass in the house. Physical
parameters can easily be changed, as well as tilt of the wall, emplacement and building
characteristics, thus allowing us to rapidly evaluate the results of constructive changes.
EXPERIMENTAL AND MODEL RESULTS
The poor accessibility to the "La Ola" house and lack of adequate instrumentation at
that time prevented us from conducting a systematic measurement campaign. We had the
opportunity to make spot measurements of solar radiation (both horizontal and tilted), air
speed in the system and temperature several times from October trough November 1982.
Besides this, average daily inside and ambient temperatures were monitored during a time
that the house was uninhabited (thus with no internal heat gain). These measurements
gave us for November: KT = 0.77;airspeed between 0.2 and 0.25 m/s in the air channel of
the wall, average daily efficiency for the solar system 77 = 0.25, and monthly solar "am of
2100 MJ. For October the solar gain was 3500 MJ. These results agree fairly well with
those reported by Trombe, (1976), even though air circulation was rather lower.
TABLE 1. Comparison of Simulation and SLR Method
Month
Heating Load
SHF
Heating Load
SHF

M J / month
(Sim.)
MJ/month
(SLR)
JAN
3688.1
0.657
7940.2
0.433
FEB
4296.5
0.750
8121.2
0.503
MAR
6024.0
0.827
9765.4
0.577
APR
7809.8
0.706
12125.8
0.543
MAY
9324.0
0.646
14501.5
0.509
JUN
10018.1
0.561
14310.J
0.502
JUL
11258.7
0.490
17697.9
0.437
AUG
10673.2
0.505
17045.4
0.439
SEP
9190.8
0.494
15472.6
0.421
OCT
6541.1
0.575
11455.4
0.449
NOV
5015.4
0.489
9844.4
0.385
DEC
4722.7
0.433
9752.6
0.354
The house has been inhabited since then, and the system has worked with no problems.
Minimum inside temperatures have never dropped below 14 C, and fuel use is about one
fifth of what it was in the old house.
In Table 1 we compare the monthly heating loads and useful energy for both the simulation
and using the SLR method (Balcomb, 1978). These results are for the house "as built".
We can see that the simulation predicts lower heating loads than the SLR method, and our
predicted SHF's are higher. This is due to the fact that the temperature swing allowed
for the SLR method is much lower, and also other hypothesis implicit in the SLR method.
-------
2648
Even then we believe that this method is excellent for preliminary design of a passive
system.
In Fig. 2 we show the annual overall performance of the Trombe wall. Heating load is
present throughout the year, and the solar system provides about 45% of the annual heating
demand. In practice this has been shown to be conservative, since wider thermal swings
are acceptable, and ambient temperature varies from 14 to 26 C. Since the Trombe wall
retains heat very well, operative temperature (which takes into account radiant exchanges)
MJ/day
Thermal Load !
!
Solar Heati.ngj
Aux. Heating |
350
300-
250-
200
150-
100-
>
50-
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DF.P Months
Fig. 2. Annual Performance
is higher than for a "normal" house. Figure 3 compares the SHF for both 90° and 70° tilt.
The comparison shows that there is a net advantage in using the 70° tilt all year round,
but that the difference is dramatic from October through February. Finally, in Fig. 4 we
see the effect of varying the thermostat "setpoint" on system performance. SHF improves
very quickly as setpoint decreases, but both the T.W. useful energy output and efficiency
are much less sensitive to this parameter.
Trombe Wall Contribution / Thermal Load
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Months
Fig. 3. Effect of Wall Tilt on SHF
-------
2649
MJ/day
250
©Trombe Wall Contribution
"100
200
150
'¦ SHF
0,5-
100
® Trombe Wall Efficiency
l0 J-0
20,5 Thermostat Temperature
17,5
14,5
Fig. 4. Effect of Thermostat on SHF
In addition to the results already mentioned, we carried out many other tests using the
simulation program. From them we can draw the following general conclusions:
• Exterior wall insulation is indispensable. Using it, the SHF rises to over 70%.
• If night insulation of the Trombe wall is also used, there is almost no need for auxiliary
heating.
• With exterior wall insulation, the house has enough thermal mass to assimilate a
sequence of three overcast days in winter with very little auxiliary heating.
REFERENCES
1. J.D. Balcomb. 1978. SLR Method in Los Alamos Reports LA-UR-78-1159 and
LA-UR-78-2570.
2. R.W. Bliss. 1961. Atmospheric radiation near the surface of the earth. Solar Energy,
5, 3.
3. H. Czerny, M. Longas. 1982. Estudio experimental de la conveccion natural en un
ducto rectangular con calentamieento asimetrico. Mech. Eng. Thesis, University of
Chile.
4. H.C. Hottel. 1976. A simple model for estimating the transmittance of direct solar
radiation through clear atmospheres. Solar Energy, 18, 129-134.
5. F. Kreith. Principles of Heat Transfer. 1973
6. B.Y.H. Liu, R.C.Jordan: Daily insolation on surfaces tilted towards the equator.
ASHRAE Transactions, 526. 1962.
7. R. Munoz G. 1981. Busqueda de una correlacion para la radiacion solar en base a
variables atmosfericas. Mech. Eng. Thesis, University of Chile.
8. F. Trombe, J.F. Robert: Caracteristiques de performance des insolateurs equipant
la premiere maison II chauffage solaire du CNRS. Publication du Laboratoire CNRS
d'Odeillo. 1976.
-------
2650
RENEWABLE ENERGY CONCEPTS IN THE DESIGN OF ANTIQUE
ROMAN URBAN HOUSES
A. Stahr
Fachgebiet Industrielles Bauen, Technische Universitat Berlin,
Eschershauser Weg 15 i, D 1000 Berlin 37, Germany
ABSTRACT
The demand for active energy production (heating) and the contribution of passive (solar)
energy to achieve a reasonable room temperature at all seasons was calculated for house
VI.2.4 (Casa di Sallustius) in antique Pompei. For constructional details still unresolved by
archaeologic research, two different assumptions were made: A "minimal" insolation concept
(example A) and an energy saving concept in accordance with the technology known and
applied then (example B). Energy demands and energy savings were calculated for both
assumptions and were compared to those of a standard modern one family house in this region.
If the mean room temperature should be 18° C (20° C at daytime and 16° C at nighttime), the
annual heat loss is 190,268 kWh, and the annual additional heating requirement 97,705
kWh for example A. For example B, the annual heat loss is 14,342 kWh, and the annual
additional heating requirement 673 kWh. The specific heating energy demand is 23.04
W/sq.m.K for example A, 4.93W/sq.m.K for example B and 4.15 W/sq.m.K for a modern
house in the region.
The contribution of passive solar energy for a standard contemporary house in this region is
about 50%. For the antique building analyzed in this study, comparable values are already
arrived assuming a "minimal" insolation concept (example A). Solar profit can be as high as
95% for antique houses, when all means of energy saving known then would have been
applied.
KEYWORDS
Passive solar energy; energy saving concepts; antique roman architecture; atrium house;
Pompei.
OBJECTIVE
Did climatic considerations, esp. energy saving and the use of renewable energy, play a
significant role in the construction of antique Roman urban houses, and how does the energy
calculation for an antique house compare to that of a standard modern one family house in this
region ?
METHODS
This question was studied at Pompei, an ancient Woman city, where, because of the Vesuv
outbreak of 79 A.C, the original city layout and the construction of the houses has been
excavated to an extent sufficient for this study. The general city plan, the layout of the
different properties, and the layout and construction of the individual houses were analyzed.
-------
2651
The demand for active energy production (heating) and the contribution of passive (solar)
energy to achieve a reasonable room temperature at all seasons was calculated for house
Vl.2.4 (Fiorelli, 1875). The main layout and the construction of the outer walls and the floor
have been pretty well established by excavation (Eschenbach, 1984; La Rocca and others,
1981; Mascoli and others, 1981; McKay, 1975; Overbeck, 1875; Richardson, 1988). Other
constructional details are still a matter of discussion. For these yet unresolved details, two
different assumptions were made: A simple "minimal" construction (example A) and an
energy saving concept in accordance with the technology known and applied then (example B).
Energy demands and energy savings were calculated for both assumptions using standard
formulas for contemporary constructions (H6nmann and others, 1989; Rehberg, 1982). The
energy calculation was then compared to that of a standard modern one family house in this
region. All energy calculations were made for the local climate (Palz, 1984; Deutscher
Wetterdienst, 1967; Muller, 1980).
RESULTS
City Planning
The city was erected on a lava plateau about 30 m abo sea level. The location was probably
mainly chosen for economic reasons: The editerranean sea was only 500 m away; there was
a navigable river, and it was at the crossing-point of two major trade-routes. Furthermore,
defence was facilitated through a steep declination in two directions.
The major road net ran from northwest to southeast (cardo), crossed by the decumani which
ran from northeast to southwest. The street layout thus followed the prevailing wind
directions: At daytime from the editerranean (SW) and in the evening from the mountains
(NE). This ensured a rapid and constant exchange of the urban exhaust.
Property
The property layout followed a standard plan, taking mainly economic considerations into
account: The property was usually situated between two parallel streets; the street front was
ordinarily commercially used.
Standard House Layout
The standard house layout showed a seasonal pattern: There were different living- and
diningrooms for winter- and summertime, a cool cryptoporticus and sunny rooms and
terraces on the roof. Integrated into this energy concept was the waste heat of the craft shops
on the property and heat insulation by front buildings.
House Construction and Energy Calculation
The constructive principle tfere thick walls with small openings, providing heat insolation
during the daytime and heat storage for the night.
House Vl.2.4 is a rather typical example for an urban house at ancient Pompei. It is an atrium
house standing isolated in a southwest/northeast orientation, the street facade facing
southwest. It is a one-family house for 10 to 15 persons with a floor space of 358.95 sq.m.
and a cubic capacity of 2,460.94 cu.m.
Well established constructional details
The construction of the outer walls and the floor have been pretty well established by
excavation.
Construction of the outer walls:
The wall facing the street was a 42 cm. thick monolithic fair-faced masonry from tuffstone on
a basal layer of limestone squarestones. On the inner side, it had a 3-layer 4.7 cm. thick
plaster (limesand / lime and limespar / lime, soap and wax with pulverized chalk). The k-
value for this wall is 1.11 W/sq.m.K. The other outer walls were equally thick, mainly made
from lava rubbish masonry, and were plastered on both sides (k-value = 0.97 W/sq.m.K).
-------
2652
Construction of the floors:
The less important rooms (for slaves, procoetum, stores) had a floor construction of 12 cm.
impermeable gypsum mortar on tamped earth (k-value = 1.30 W/sq.m.K). The other rooms
had a 4-layer floor construction: A 7.5 cm. basal layer of pebble was followed by a 22.5 cm.
layer of hard-core rubbish and lime mortar on tamped earth and loam. This was followed by
12 cm. of screed from Roman concrete, which consisted of pulverized brick and lime. The top
flooring was a limestone stab covering or mosaique stones. Resp. k-values were 1.61 and
1.64W/sq.m.K (limestone stab covering or mosaique stones).
Assumptions for a "minimal" construction for house Vl.2.4 ^example A)
Windows and doors:
Openings for windows and doors were small, with no glazing for the windows. Wooden shutters
from fir, oak or cypress with curtains or hides provided a certain degree of energy
conservation at night, esp. during the winter months (k-value = 5.88 or 1.41 W/sq.m.K, for
open or closed shutters resp.).
Roof construction:
The wooden roof construction was a combination of shed roofs. The rafters were topped with
unjointed clay tiles, the joist floor had a timber lining (k-value = 1.59 W/sq.m.K).
Floor construction:
The general floor construction and the resp. k-values have been outlined above.
Enerov calculation for example A
Energy calculation for example A shows for the winter half-year an energy loss through
transmission (heat transmission loss) of 1,270.31 W/K and a heat ventilation loss of
6,998.91 W/K at an air exchange rate of 7.9/h.
Assuming that the window-openings were shut at nighttime with wooden shutters, the air
exchange rate for the night was calculated as the heat ventilation requirement by the joint
ventilation in relation to the minimal exchange rate (0.8/h).
The total heat loss for example A is calculated as 8,268.22 W/K or 198.46 kWh per day,
resp. The specific heating energy demand (total heat loss per square meter floor) is calculated
as 23.04 W/sq.m.K.
The net collector surface is 127.64 sq.m. with an absorption factor of 0.50 for the outer
walls.
In the Pompei region, the winter months (November -Aprih are the sole contributors to the
all-over annual heat loss. If the mean room temperature should be 18° C (20° C at daytime
and 16° C at nighttime), the annual heat loss is 190,268 kWh, and the annual additional
heating requirement 97,705 kWh.
Assumptions for an energy savinn concept for house Vl.2.4 f=example
Windows and doors:
In this assumption, the windows had a single glazing of Lapis Lazuli in a lead tray (k-value =
5.26 W/sq.m.K). At nighttime, the closed wooden shutters on the outside were combined with
curtains or carpets on the inside (k-value = 1.39 W/sq.m.K).
Roof construction:
For example B, jointing of the clay roofing is assumed. For the main rooms it can be assumed
that the timber lining of the joist floor was plastered. The ventilated space within the roof
structure could have been insolated with an aboat 5 cm. layer of fern and reed (k-value -
0.48 W/sq.m.K).
Floor construction:
The floor construction for example B only differs in respect to the winter diningj-oom:
Between a 2-layer gypsum mortar was a layer of coal (k-value = 0.14W/sq.m.K).
Energy calculation for example B
Especially because of the window glazing, the air exchange rate is reduced to 0.5/h, leading to
a heat ventilation loss of 442.97 W/K.
The total heat loss for example B is 1,771.33 W/K or 42.51 kWh per day, resp. The specific
heating energy demand is calculated as 4.93 W/sq.m.K.
The annual heat loss is 14,342 kWh, and the annual additional heating requirement 673 kWh.
-------
2653
CONCLUSION
During the summer months, sufficient room cooling by ventilation and sun shading were the
major climatic requirements for the houses in antique Pompei. The efficiency of the cooling
measures in the antique house was at least comparable to that in the contemporary house
analysed.
Against heat losses during the winter months, cold insulation of the floors and celings were
less efficient than today. The insulation of the walls was comparable to modern demands, with
k-values close to contemporary constructions.
The constant daytime heat exchange during the winter months in example A with unglazed
window openings naturally leads to an extremely high heating demand. Assuming glazed
windows, though (example B), energy demands are close to those for contemporary houses in
the region.
The contribution of passive solar energy for contemporary houses in this region is about
50%. For the antique building analyzed in this study, comparable values are already arrived,
assuming a "minimal" insolation concept (example A). For a mean room temperature of 16° C
resp. 14° C in this example, the contribution of passive solar energy to the overall heat
demand increases to nearly 69% resp. 75%. When all means of energy saving known then
would have been applied (example B), solar profit could theoretically be as high as 95% for
antique houses.
TABLE 1 K-Values for the Different Constructional Elements (W/sq.m.K)
house Pompei VI .2.4
example A example B
contemporary conventional
house in the same reoion
outer walls
- street facade 1.11 1.11
- other sides 0.97 0.97
windows
- daytime 5.88 5.26
- nighttime 1.41 1.39
floors
3-layer floor
-with a limestone
s/ab covering 1.61 1.61
- with
mosaique stones 1.64 1.64
1 -layer
gypsum mortar 1.30 1.30
2-layer
gypsum mortar
with a coal layer - 0.14
roof
timber joist roof 1.59* 0.49*
-plastered - 0.48
1.54
1.54
2.90
1.50
0.64
0.46
*= unjointed clay tiles
*=jointed clay tiles
-------
2654
TABLE 2 Heating Energy Demands (W/sa.m.K)
house Pompei Vl.2.4
example A example B
mean room
temp, in C
specific heating
energy demand
percentage of
solar profit
percentage of
additional heating
requiremen
18 16 14
-23.04-
49 59 73
51 41 27
1 8
4.93
95
contemporary conventional
houses in the same region
1 8
4.15
52
48
J.05

3
0-
urnm
*
-------
2655
Fig. 2. Casa di Sallustius. Sectional views.
(All measurements in meters)
REFERENCES
Deutscher Wetterdienst (1967). Klimadaten des Standortes Neapel. Deutscher Wetterdienst,
Offenbach.
Eschenbach, H. (1984). Pompeji. Erlehte antike Welt. VEB Buch- und Kunstverlag, Leipzig.
Fiorelli, G. (1875). Descrizione di Pompei. Napoli.
HOnmann, W. and E. Sprenger (1989). Taschenbuch fur Heizuna und Klimatechnik. R. Olden-
bourg Verlag, Munchen-Wien.
La Rocca, E., M. and A. de Vos (1981). Guida archeoloaica die Pompei. Arnoldo Mondadori
Editore, Roma.
Mascoli, L„ S. De Caro, A. Jacques, P. Pinon, G. Vallet, F. Zevi (1981). Pompei e oil
architetti. Gaetano Macchiaroli. Napoli-Roma.
McKay, A.G. (1975). Rfimische Hauser. Villen und Paiaste.Ranoi-Verlao. Feldmeilen
Muller, M.J. (1980). Handbuch ausaewShlter Klimastationen der Erde. Forschungsstelle
Bodenerosion der Universitat Trier, Trier.
Overbeck.J. (1875). Pompeii in seinen Gebauden. Alterthumern und Kunstwerken. Verlag
von Wilhelm Engelmann, Leipzig.
Palz, W. (1984). Atlas uber die Sonnenstrahlung Europas. Band I. TIJV Rheinland, Koln.
Rehberg, S. (1982). Gberschiagiges Berechnunosverfahren fur die passive
Sonneneneroienutzuna in Wohngebauden. Technische Universitat, Berlin.
Richardson, L. (1988). Pompeii: An architectural history. The John Hopkins University
Press. Baltimore-London.
-------
2656
THE DESIGN AND TECHNICAL AND ECONOMIC ANALYSIS OF THE PASSIVE
SOLAR ENERGY CLASSROOMS IN THE RURAL AREAS OF SHANDONG, CHINA
Wang Chongj ie
Shandong Institute of Architecture/Engineering
5 Heping Road Jinan 250014 China
INTRODUCTION
The temperature of the interior of classrooms in winter should be at 16°C to
18° C according to The Standards for Architectural Design of Middle and Primary
Schools GBJQ9-86, formulated by the authorities concerned. However, for a long
time, the primary schools in cold rural areas of China couldn't reach the Stand-
ards because of the coal shortage and economic difficulties. The heating prob-
lems of the rural primary schools in winter have not been solved. The low
temperature of the interior of classrooms effects the normal teaching order and
both teachers' and students' health. Keeping warm in primary schools in rural
areas during the winter is a noticeable problem. The parties concerned should
pay close attention to the problem.
However, a very good way to use the inexhaustible supply of solar energy
would be to solve the problem of keeping warm in the countryside's primary
schools during the winter. In Shandong Province, the northern part of China,
there are comparatively rich solar energy resources. The concrete action would
be to build passive solar energy classrooms. Such buildings adopt a rational
juxtaposition of classroom orientation and the environment. The clever arrange-
ment (from interior rooms to exterior style) enables the buildings themselves
to collect, keep, store and distribute solar heat. Such construction can help
the building heating problem. The technical applications have a very practical
significance for improving heating conditions of middle and primary schools in
the northern rural areas of China, and for saving conventional energy resources.
This essay will discuss, research and analyze the passive solar classrooms in
Shouguang County, Shandong Province, China.
THE GEOGRAPHICAL POSITION AND THE SOLAR ENERGY RESOURCES
Shouguang County is situated by the Gulf of Laizhou, 40° 20' north latitude
and 119° 30' east longitude in the northern part of China. It is clear, dry and
cold in winter. The average temperature is -4°C in the coldest months. It has
2800 hours of sunshine annually in the area, and its annual sunshine rate is 62%.
The monthly sunshine is about 200 hours in November, December and January. The
sunshine rate is 63%. Shondong Province is one of the areas which have rich
solar energy resources in China. The advanced geographical position and the rich
solar energy resources provide the precondition for building solar energy class-
rooms. The area lacks coal resources. The price of coal here is high. It is
quite difficult to heat with coal. So, building passive solar energy classrooms
has realistic significance and applied prospects.
THE HEATING FACILITIES OF THE BUILDINGS COMPRISED OF STANDARD CLASSROOMS
The building has two heat collection possibilities: direct gain and thermal
storage walls. The whole built-up area of the building is 118 m^. In general,
considering the location of the classrooms in the school and the arrangement of
the pupils, the entrance of the classroom should be in the south. The entrance
-------
2657
should attach with a porch. The transitional space between the interior and
the exterior of the classroom can be used to reduce the cold outside air.
The porch, in fact, is a little sun shine room which has the good function of
keeping the area warm. The construction plans consider the classroom's normal
usage, i.e. morning, noon and afternoon use. In order to absorb sunshine and
raise the classroom's temperature as goon as possible in the morning, the class-
room will be turned to the east 5°. Such design can also restrain the flat
radiation of sunshine after 3 p.m. In order to ventilate the classroom, the
north wall should have two windows (900 x 1200 mm). Climbing plants can be
planted in the garden in front of the heat-collection windows facing south to
prevent sunshine and to lower the temperatures in summer.
The direct gain method of heat collection should be given priority while
the thermal storage walls are secondary. Both heat collection methods are
suitable for the classroom's characteristics. They are of benefit to the
initial investment and the management.
Direct Gain
The sunshine goes through the front glass window on the south allowing
the temperature of the classroom to go up. The general area of the window
is 10.86M. The bigger the window, the more heat is absorbed; however, the
window is the weak, link to keep warm in the enclosing walls. Its heat
consumption takes two thirds of the total heat consumption of the building.
To deal with the problem, double wooden windows are used. The heat resistance
of the windows is equal to three times that of the single wooden windows.
In order to reduce the permeation of the cold air and to improve the air
tightness, the plastic seals should be put on the outside of the window.
This will help retain the heat on the inside and prevent wind and sand on the
outside. The environmental quality of the classroom is improved.
Thermal Storage Walls
Thermal storage walls are a common form for passive solar energy buildings.
The design is to build thermal storage walls 7.2 sqm at the building's two
sides with direct gain windows. There are upper and lower vents on the wall»
allowing air to move into the interior room quickly. In summer, the upper vent
should be closed and the upper window opened to release the hot air. The
entrance to the classroom is taken into consideration when designing the
sunlight room. Besides the necessary structural components, big classrooms
should be used in order to obsorb heat quickly. A vent should be put on the
interior door.
The thermal insulating layer of the sunlight room is the key to the
temperature of the interior classroom. When we choose the thermal structures
and materials, the structure should be combined with the traditional methods
and the thermal material should be obtained on the spot. In that way, the
effect is optimum and the cost is low. The concrete action is: Walls: east,
west and north walls are thermal compound.[brick wall (240MM) + thermal layer
(100MM) + brick wall (120 MM)]. The thermal materials are plastic foam which
is available locally. The rock's unit weight is r=40kg/m3. and its thermal
conductivity A. =0.02w/m.k. In order to preserve the thermal material from
dampness, the foot of the wall is made of cement and the thermal materials
in plastic bags are placed in the wall. The brick wall 240mm and brick wall
120 will be linked together with reinforcing bar ^ 6. The roof: the roof is
made of common roof truss, purlin, reed and day tiles 100(mm) thick plastic
foam is placed in the ceiling.
-------
2658
The two small windows on the north wall should be adjusted for thermal
insulating in the winter. The entrance to the classroom would have a thermal
curtain placed on it.
THE ANALYSIS OF THE ECONOMIC BENEFIT
What is the economic benefit for building the passive solar energy
classroom? How much more investment would be necessary to build these
classrooms than to build ordinary ones? How many years can the investment
be retrieved? The designers and customers are concerned with the questions
very much. The principle of the design is not to seek the solar energy
energy-saved rate unilaterally but to consider the possibilities of
building and making popular such buildings in rural areas of China. Local
raw materials should be used to reduce the cost of the structure.
The extra investment for solar energy classrooms would be for thermal
insulating materials, bricks, wood and reinforcing bars. Every square
meter of the building should increase RMBV31 more. The investment for the
buildings in 15% higher than for the ordinary ones. The fixed nimber of
years of the investment recovery is concerned with many factors. But the
main factor is the cost of coal in the local area. One ton of the coal
needs over RMBY200 in winter. However, each classroom needs over 1.5T coal.
The stoves also cost about BMB 60 every year. The extra investment for solar
energy buildings will be reclaimed in about four to five years.
The project has been completed. The test shows that the average tempera-
ture of the classroom to be 12°C in January. The lowest temperature at 7
o'clock a.m. of the interior room is 8°C if the sunshine continues for four
days. The temperature of the building is 8 to 10°C higher than the ordinary
building without stoves and is 2°C higher than the ordinary building with
stoves.
The passive solar energy buildings are praised by the customers. They
are received warmly in the local areas. The local people ask repeatedly
for solar energy classrooms and buildings to be constructed. We also
believe the buildings have good prospects in the rural areas of Shandong and
the northern parts of China.
-------
2659
TRADITIONAL BIOCLIMATIC BUILDINGS IN MONTENEGRO
Dusan Vuksanovic
University of Titograd, Dept. of Civil Engineering
Cetinjski pu bb. 81000 Titograd, Yugoslavia
ABSTRACT
Before the general acceptance of an approach which ignored the physical and
cultural context of each building,there was a traditional builder's method
which, despite poor technology, created an architecture which satisfied those
needs and was adapted to climate and soil in a fascinating way. The purpose
of research in traditional building methods is to rediscover those forgotten
attitudes and elements of the design concept. Some interesting cases of rural
buildings and settlements in the Montenegro area, where environment influenced
building more directly than in urban conglomerations, are presented in this
paper.
KEYWORDS
Environment; adaptation; traditional building methods; bioclimatic response;
local materials; terrain; orientation; ambient
INTRODUCTION
Architecture of any certain region is the result of natural influences and
cultural factors. It is an expression of the life style and its specific
meanings and values. Any of the single influences, i.e. climate, topography,
local building materials, building technology, socio-economic and political
milieu cannot be accepted as decisive alone. Their influence cannot be
analyzed separately. Rather, the conditions in a natural environment, taken
as a whole, are the determinants of house forms.
MONTENEGRO
The territory of Montenegro lies in the south part of Yugoslavia, between
41° 52' and 43° 32' north latitude. This is mountainous terrain which
includes a part of the Adriatic coast.
The climate of Montenegro is characterized by the extreme temperature varia-
tions over short distances. For example, the average temperature in January
in the south is 5°C, yet there are places in the north and northeast with
low temperatures of -28°C to -32°C. Annual duration of the sun ranges from
1700 hours in the north to 2700 hours in the south. Annual mean global solar
radiation in the coastal area is 4200 Wh/m^ per day, with average maximums in
July of 7450 Whm^ per day and minimums in December of 1600 Wh/m .
-------
2660
REVIEW OF THE TRADITIONAL BIOCLIMATIC FORMS
Pastrovici (Adriatic coast)
Basic form of the Pastrovici house originates from the primitive shelter, a
single pitch roof leaned on a rock. Western Europe influences, from Venice in
the first place, caused the development of the house in vertical direction up to
the final solution with the ground floor plus two floors (Fig. 1, Fig. 2).
1
Fig. 1. Development of the house in Pastrovici
fj!

mil iiiii
llll

~~ir^

Fig. 2. Typical Pastrovici house: 1 - second floor,
2 - first floor, 3 - basement, 4 - cross section
The concept of a house is predominatly summer oriented, as expected, having its
best interpretation in the large terrace ("taraca") with pergola ("odrina") in
front of the house, on the first floor or high ground floor level (Fig. 3).
/\
Fig. 3. Position of the terrace in case of the
village street (Tudorovici)
-------
2661
The terrace is slat paved with the stone bttnch ("pizun"), as a fence, oriented
to the interior of terrace (Fig. 3). The pergola is usually with wine. Local
climatic conditions allow that terrace can be used as a center of the social
life during more than half of the year - "a summer living room".
Very obvious traditional builder's bioclimatic solution is the single roof,
which organically fol lows the lines of relief by its form and slope. At the same
time it is adapted to resist better the north wind ("bura") influences.
One of the examples which illustrates the authentic building sensibility for the
landscape is the groupe of houses in Srzentici. It makes a kind of extension of
the surrounding terrain by its bioclimaticly incorporated forms and contour, so
it is almost difficult to distinguish the built structure from the natural rocks
(Fig. 4).
Fig. 4 Bioclimatic relation of grouping of houses in Srzentici
to the surrounding terrain
Krtoli (Adriatic coast)
The model ofaKrtoli house reflects relatively independent status of these
settlements and production process based on the local resources. The house with
accessory buildings and the yard form "dom" - basic housing and economic unit of
the village. The house with the mill or the wine cellar in the ground floor is
usually in one building and the kitchen with the cellar or the stable is in
other one (Fig. 5).
Fig. 5 Spatial forms of the Krtoli "dom"
-------
2662
The concentrated morphology of Krtoli village makes it similar with the little
coastal towns in appearence (Fig. 6). Besides the defensive needs in the past,
such spatial organisation was strengthened by accessibility from the sea, which
was important condition in traditional economy of this region.
Fig. 6 The urban disposition and the view of a village
in Krtoli (Bjelila)
Houses in Krtoli, placed either longitudinal or transversal to contour lines,
are oriented to the South, Southeast, or Southwest. Two-story stone buildings
have external stairway and double pitch roof, cladded with Spanish tiles (Fig.
7). The windows are placed immediatly under the cornice, as at Pastrovici house.
Wooden shutters do not escsist in each case.
stone - thermal capacity
o ~ a
n n

o
a
vrrV i
irnrrr
tit
fir
n
shutters - protection from the sun
slope of roof - wind, type of
cladding
Fig. 7 Elements of the Krtoli house architecture
Rijeka Crnojevica (Skadar Lake basin)
For living on Skadar Lake shore it is very ImportanttD ncte lake area is 1/3
bigger during the rainy winter.whJoh makes some surrounding settlements flooded.
The exceptionally claer example of adaptation to conditions on the lake shore
and to summer heat conditions is accomplished in Rijeka Crnojevica. This small
town was the main port and the trade center ("pazar") 0f the Montenegrian
principality in the XIX century.
The house in Rijeka Crnojevica represents the specific type of
Mediteranien-oriental house. It has some features of the pile dwelling: it is
-------
2663
placed near the water, so its first floor is flooded during the high lake
level. The usual , functional organization of the traditional houseis:first
floor has economic or both economic and trade character;and the flat is on the
second floor (Fig. 8).
Fig. 8 Characteristic architecture of Rijeka Crnojevica
Large terracffi under the roof, in position immediatly above the water,provide a
pleasant and attractive respteduring the hot summer days. As a open living room,
terrace represents a succesfuly modulated form on the interior - exterior
relation
Region of high mountains (North and Northeast continental part)
Typical isolated settlements are "katuni" - summer cattle-breeders1 houses,
usually located near the canions. Their architecture reflects the influences of
severe mountain environment: stone walls and very pitched shingle roofs or roofs
with plain limestones. Expressive scatterectess of the villages from this region
points out their origin' as cattle-breeding settlements.
Fig. 9 Houses in residential area ("mahala") of Rozaje
-------
2664
In the case of house from Rozaje a high level in adapting of urban environment
to the natural conditions is achived. All elements of the house and its
farmstead are made from conifer wood, which ejrsists there in abundance, and the
pitch roofs represent the response to the snow, also following the terrain
contours in a exceptionaly good way (Fig. 9).
COMMON FEATURES
Shortcoming? .of the old traditional houses are not small in numbers: small
openings, often porous stone, weak bondings. Quality of the stay they offer is
adapted to the different life style - without sunny living rooms. But, reasons
for going in for their design concept are not in simple taking on traditional
forms. These well oriented, simple houses, built from local materials, located
on the firm soil represent worthy models of energy and environment-coneious
architecture.
Amazing reach in adaptation to the natural environment these houses, their
groupings and whole settlements originate mostly from the organic principle of
growth during time. All of these facts enabled the rising of architecture
like "the crystallised piece of landscape" (Fig. 10).
Fig. 10 "Crystallisation of the landscape"
REFERENCES
Olgyay, V. (1962). Design with Climate. Princeton University Press, Princeton.
N.J.
Cook, J. (1988). The Post-Industrial Culture of Regionalism. Proc. Conference
Passive Solar Architecture 1988, Ljubljana, Yugoslavia. 1-8.
Krainer, A. (1986). Equality in Variety, A Rewiev of Bioclimatic Growth of
Buildings on Yugoslav Territory. Proceedings PLEA'86, Pecs, Hungary. 168-219.
Gakovic, S. (1979). The House of Pastrovicl. Center for Urban Development
Planing, Belgrade.
Djurovic, G. (1981). Revitalisation Project for Thirtheen Ambient Locations In
Tivat Community. Center for Urban Development Planing, Belgrade.
Vuksanovic, D. (1990). Autochthonous Bioclimatic Building Approach. Proc.
International Congress Energy and Environment. Opatija, Yugoslavia. 259-266.
-------
3.6 Passive Commerical Buildings
2665
STOREY AND SOLAR GLASS ROOF ADDITION TO AN ARCHITECTURAL OFFICE
Florian (Autor)+Franz+Wendelin Lichtblau
Dipl.-Ing'e, Freie Architekten
SoeltlstraBe 14, D - 8000 Mlinchen 90
Statics: Dr.-Ing. Pittioni / Mlinchen
Technics: Dipl.-Ing. Struve / Rosenheim
Glass Elements: Fa. Okalux / Marktheidenfeld
with Dipl.-Ing. Koster / Frankfurt
Sponsored by the Economic Ministry of Bavaria
Translation: M.G. Montry / Munich
ABSTRACT
In Munich, in 1987 / 88,we constructed a pure wood-skeleton storey addition
under a completely glazed saddle back roof. Special glass elements combine
translucent insulation by capillary plates (TIM) with a shading lamina struc-
ture. For the first time, all functions of sun-protection, light and energy
passage, as well as light diffusion and heat insulation are fulfilled by only
one component, dry glazed in a wooden construction.
Three years of scientific measurement and personal observation in the 'Col-
lector House' confirmed all our assumptions made about 'intelligent domesti-
cation' of natural processes. Without any technical furnishings, the self-
regulating conditions of our workspace under the 'pilot-roof' can be consi-
dered ideal, primary energy savings amount to about 60 %! The new technology
opens a lot of extremely effective applications in Passive Solar and Day-
light Architecture.
KEYWORDS
'Intelligent Architecture'; passive heating and cooling; daylighting con-
cepts, health; transparent / translucent components; regulation concepts and
components; construction methods and materials; measurements, observations...
From the South over SoeltlstraRe The 'East Office'
-------
2666
PROJECT DESCRIPTION
Building Type. Extent of Construction Activity
MUNICH. In the early sixties we built, with the most modest approach, a ground
level office building of concrete panel /wood frame construction with a flat
roof over a two layer brick and concrete basement. To this and to newly layed
foundations we added storey additions in 1987 / 88: an upper floor with gal-
lery, of pure wood skeleton construction, under a moderately inclined saddle
back roof. The rebui 11space added 500 m3 to the existing 600 for a total of
about 1100 m3.
Location and Considerations
The neighborhood, since the thirties, has been zoned for single family resi-
dences with generous greenspace. Building density and design were tightly
controlled. Planning considerations are oriented towards general municipal
and state building codes.
Process, Construction, Costs
Planning and acquisition of permits for the initial version (facade opening,
air col lector / gas heating system) began at the end of 1986. X, V and Z pro-
portions were derived from the existing structural conditions, the column and
construction grid, for weight reasons as well, was minimized to a 10 cm
cross section.
In the middle of 1987 changes were made to accomodate the newly developed
roof gluing- In addition came the related calculations and application for
a government grant. After partly adventurous bid solicitation, construction
began in September. Quick assembly was a necessity as the office itself was,
at this time, without a roof.
Primary construction consists of laminated larch, horizontal members being
pegged with zink-coated T profile steel to continuous columns. Further con-
struction consists of mineral insulation and pine plywood with larch trim,
screwed to seal strips. To this was added larch flooring, larch stairways
with steel details, and window frames of Z profile steel. The installation of
the 72 mm thick glass roof elements ensued as dry glasing directly attached
to rabbeted wooden rafters, and as visible fastening by means of screws over
a watertight profile, to create a second drainage level, with aluminum trim.
Former Condition from Southwest Working Model, M 1:20
-------
2667
After a little less than four months of winter construction, lasting until
February of 1988, our storey additions were, with good fortune, complete. We
moved in in March. Construction methods and materials remain, to the present,
troublefree. The final costs for construction alone had increased about a
third from 300.000, - DM for conventional construction to 400.000, - DM for the
'pilot version'. We succeeded in getting proportional public funding,but on
the condition that, after completion,we were to carry out a one year program
of measurement and analysis. Balance sheets and results confirm all idealis-
tic assumptions.
ENERGY, COMFORT, ENVIRONMENT
Energy Consumption, Passive Solar Use
Heating of conventional construction, corresponding to specific net energy
consumption requirements (without the amount of heat produced by the equip-
ment) is calculated at 150 kWh / m2 per year. The consumption for artificial
lighting of conventional structure is similarly uneconomical.
To take advantage of passive solar energy we glazed the entire 165 m2 of
roof. Special glass elements were developed and fitted: translucent insula-
tion with capillary plates (TIM) under a shading lamina structure. These ele-
ments combine for the first time all the functions of sun protection, light
and energy passage as well as light diffusion and heat insulation into one
construction component. And this was accomplished without any technical fur-
nishings. The building itself becomes a solar collector full of light! The
specific heat energy consumption decreases to under 40 %. In addition come
corresponding energy savings on artificial lighting. Primary energy savings:
more than 60 per cent!
Thus, one can do without technical equipment to air-condition the building.
In addition, for heat flow, thermal ventilating flaps were installed. If
necessary, waste heat under the ridge can be recycled. For the time being, we
use only a temperature regulated blower to support heating of the ground
floor, which seems to work extraordinarily effective. Air ducts should extend
from two points under the ridge into the basement and should be equipped
with reverse blowers.
The gleeing of the facade is set back to the necessary measure to allow for
view and ventilation. In addition to the solar heating we use gliding, ther-
Mostat-guided gas single ovons with chimney connection as already existed in
the lower floors.

from the South
Upper Floor, 0l_j—1 I m
-------
2668
The Glass roof Components
consist of three panels. Sun protection or sun passage (depending on the time
of day and year) is assured by self regulating, (specific to location) stiff-
installed aluminum segment reflectors in the outer airspace, a plexi-capil-
lary structure forms a diffused scattered light passage in connection with
the translucent heat insulation (TIM). Seasonal sunpaths form the 'software'
in a fixed system.
The technical values of the element: an overall energy passage degree 'g'
from 0,20 to 0,65, heat passage degree k = 0,8 W / m2K, assessed sound absorp-
tion of 52 dB. With such elements were achieved during the current heating
period an equivalent 'k' value of minus -0,65 to -1,0 W / m2K. Solar gain is
therefore exceptionally high.
Influence of Surroundings, Comfort Level
The surface of the glass roof is not shaded by surrounding trees or buil-
dings. The quality of the evironment makes microclimate improvements unneces-
sary. Nonetheless, we cultivated intensively, so that within short time the
facade will vanish completely behind climbing vines.
All aspects of comfort, with regard to indoor temperature, air composition,
light direction and sound protection, have, through ecologically adapted
methods and materials, reached a high level of refinement. The intelligent
use of heating and air-conditioning equipments sharpens our awareness of na-
tural processes and correlations and creates a moderate'bracing climate'which
we find, after an adaption period, physically and mentally stimulating as well
as good for our health.
The self regulating light passage of the 'Koster' segments, in connection
with the diffuse scattering effect of the capillary plates and white vai—
nished interior walls create a shade- and dazzle-free daylight of virtually
high density, regardless of ambient conditions. The space use is extremely
variable. Our Mediterranean plants grow in the remotest corners.
PERSPECTIVES OF THE PLANNING
The acknowledged design intentions for our project were primarily:
- to continue, with decisive characteristics, a consistent dialogue between
'old' and 'new' until they grow together to form an insoluble whole.
- to ensure quick construction and flexibility of space creation and facade
sre s-s-roicbt-n-C-
¦1CT
Solar Glass Roof, OjuiiT——icm
Latitudinal Section
-------
2669
opening through modular order, building principle, and immediateness of mate-
rials and construction.
- to do justice to future energy savings requirements by means of reasonable
costs and evironmental and user friendly construction, and
- to search for, consciously, new methods of realisation, after a careful re-
search and planning process, that can also be offered to future clients, and
all who are interested, as visible proof of our priorities.
Next and generally, this thesis: that the resultant problems of the sins of
our civilization can at least be dulled by the simple application of ecologi-
cal knowledge. Inhibition is a function of our state of consciousness and not
primarily of technical aspects. There are truly many things that could be
changed immediately for the better, if only we wanted to exercise the lost
art of anticipation.
And we have ascertained that modern technology and ecological construction
methods are not mutually exclusive. Rather, the appropriate use of that tech-
nology, especially in our muddled situation, even seems to be indispensable
for the balance and / or completion of natural processes. In other words, to
'domesticate' them.
We believe, with regard to realisable state-of-the-art technology, that our
'Collector House' is an optimal form for utilizing suitably, in accordance
with human needs, the gift of sunlight. Both subjectively felt and objective-
ly measured values would speak emphatically in support of our belief.
If the above mentioned glass elements and construction components were made
commercially available, then endless possibilities would open up for most ef-
fective deployment in new building, renovation and redevelopment. The amorti-
zation factors add up to a considerably decreased expenditure for: construc-
tion planning and construction, installation and maintenance of technical
equipment, heating, ventilating, air-conditioning and lighting energy.
Therefore, without the additional use of unecological and unnecessary techni-
cal apparatus, we arrive at the most efficient use of space design, daylight-
ing and solar energy attainable. And all to the benefit of the human being.
As simple and logical as our end result may seem-it was hard-won. The risks
involved at the time of the undertaking piled up uncertainly and problemati-
cally. Looking back today, we are compelled to add that in no way would we
have forgone the rocky path that led to our result. The absolute value of our
journey has been preserved because, for once, we let the others do the tal-
king - we took the riskj. F. L. X ' 90
Before the Party
Night Architecture
-------
2670
ENERGY EFFICIENT BUILDING DESIGN AND OPERATION
AT THE NEW DENVER AIRPORT
J. Bleem, A. Kirkpatrick, C. B. Winn, N. Khan, D. Mori
Department of Mechanical Engineering
Colorado State University
Ft. Collins, CO 80523
ABSTRACT
Potential savings, costs and paybacks for various energy conservation and load management
measures (ECMs) were considered for a typical concourse building at the proposed new
Denver airport. Several common measures such as variable air volume (VAV) air handling
equipment, efficient windows, daylighting and use of an energy management system have
been incorporated into the design. Many of these measures were evaluated using the DOE-
2.ID building simulation code. In this paper, the results of these evaluations are presented.
KEYWORDS
Energy efficiency; daylighting; efficient motors; efficient belt drives; building simulation
INTRODUCTION
In response to projected increases in air traffic in the western United States, a new
international airport is currently under construction in Denver, Colorado. The current
airport design consists of several concourses and a main terminal building. One of the
concourse buildings was selected for review of several conservation / load management
measures, including daylighting, more efficient motors, efficient drive belts, occupancy
sensors, glass overhangs, upgraded insulation and exhaust air heat recovery.
The typical concourse building design is comprises of a central "core" and two "subcores".
The core is comprised of five sections: an unconditioned atrium, a mezzanine level, a
concourse level, an apron level and a basement. The subcores project as "wings" from the
central core section, and lead to the areas where contact is made with the planes. The
overall design is aimed at complementing the Rocky Mountain skyline near Denver through
use of a stepped roof outline with a large amount of glass. A simple plan view of the core
and subcore sections is given as Fig. 1 and a simple elevation view is given as Fig. 2.
-------
2671
Subcore Model
N
Core Model
J ^

\_r

I—I = 100 feet
Fig. 1. Concourse plan view

Z=75

Atrium

Z=28
Z=14
X=0

Mezzanine Zone

Concourse Zone

Z=0
Z=-13
Apron Zone
X=362

H = 10 feel

Fig. 2. Concourse elevation view
-------
2672
MODELLING
A large portion of the analysis was performed using DOE-2.1D, a public domain hourly
building simulation code. The DOE2.1 program has been used previously to evaluate
several large scale buildings and atrium designs. Landsberg et al. (1988), Bazjanac and
Winkelmann (1988) and Goldstein (1980) have modelled various large buildings with
atriums. DOE-2.1 was also used to model the terminal building for the new airport
(Kirkpatrick et al., 1990).
Input decks for the core and subcore sections of the concourse were developed based on
information available from concourse drawings, terminal building reference materials, and
discussions with airport design team personnel. Loads were based on a lighting peak of 1.9
W/ft2 (combined average), equipment peaks of 1.0 W/ft2 (subcore) and 3.5 W/ft2 (core) and
a peak occupant load of 1 person per 22.5ft2 (average). As part of the analysis, the design
was reviewed and found to comply with ASHRAE-90.1 (1989). U-values for walls, windows
and roof are 0.08, 0.35 and 0.05 (Btu/ft2 h °F) respectively (more stringent than those
suggested by ASHRAE). All windows are double pane with shading coefficient of 0.40,
visible transmittance of 0.36 and solar transmittance of 0.42. Mixing of conditioned air in
the mezzanine with unconditioned air from the atrium is based on an equivalent U-value of
2.6 Btu / ft2 h °F. Atrium venting is based on winter and summer setpoints of 70°F and 78°F
respectively. The design calls for outside air flow of 15 cfm per person, in accordance with
ASHRAE-62 (1989), and dampers are set at a fixed minimum of 20%. Thermostat settings
are modelled at 72°F winter and 74°F summer. To estimate costs, standard commercial rates
for electricity from the local utility were used. Gas rates were estimated assuming direct
wellhead purchase. Electrical rates are $9.76/kW month (demand) and $0.02604/kWh
(energy) and the wellhead gas rate is $2.22/mcf (825 kBtu/mcf).
The HV/ C system is mostly VAV (incorporating variable frequency speed control), with
some constant volume units in areas with more constant loads, and includes an air-side
economizer with 100% outside air capability. Chilled and hot water are supplied from the
airport central plant, which is comprised of gas-fired hot water boilers and gas-driven
chillers. An energy management and control system (EMS) is specified in the design to
optimize control of several items, including chilled and hot water flow, chilled and hot water
temperatures, flow lockout (to minimize simultaneous heating and cooling), and space
temperature setup and setback.
Thermal storage is also incorporated in the proposed design. The present design includes
use of ice storage for providing cooling to the planes. Loads associated with pre-
conditioning air for the planes account for approximately 12% of the peak cooling and 10%
of the peak heating loads. To prevent use of unsightly cooling towers as condensers for the
thermal storage system, the main chilled water loop is to serve as a heat sink.
RESULTS
Davlighting
Installation of daylighting controls was considered for 13 distinct areas within the building
design: four perimeter areas on both the concourse and mezzanine levels, a central area
(under the atrium) on the mezzanine level and two areas along the perimeter of each
-------
2673
subcore. The modelling was performed assuming stepped control, considering that artificial
lighting is 100% off when natural lighting levels are adequate to provide at least 30
footcandles in daylit areas. When natural light is not sufficient to provide 30 footcandles,
50% of the lamps come on to assist natural lighting. If 50% of the lights and natural light
combined cannot maintain 30 footcandles, 100% of the lights will be turned on. It is
assumed that controlling photocells will be distributed evenly throughout the areas
considered for control, and that all lamps controlled are 250 Watt high pressure sodium
lamps. For perimeter areas, control is considered for areas within 20 feet of the windows.
Savings are due almost exclusively to decreased lighting energy usage and demand. Since
the chilled water serving this area is supplied by natural gas energy (via the gas driven
chiller), only a very small portion of the savings (less than 3%) is due to reduced fan power
at the lower cooling loads afforded by daylighting. The total area considered for daylighting
controls is approximately 248,000 ft2. Energy, demand and cost savings are estimated as
1,040,000 kWh/yr, 4,400 kW/yr (sum of 12 monthly peak reductions) and $70,000/yr,
respectively. Average energy savings varied from about 25% to 44%, depending on
orientation of the space considered for control. The largest percent savings occurred in the
area beneath the atrium. With an estimated installed cost at $0.90/ft2 of controlled area
(from local contractors), the total implementation cost is estimated as $223,000. In
developing this estimate, it is assumed that only stepped control is used (i.e.,no dimming
capability is installed), and that the wiring will be done during installation of the lighting
system (i.e.,not a retrofit). Simple payback on this measure is estimated as 3.2 years.
Motors
Savings associated with this measure are due to the difference in operating efficiency
between "standard" and "high efficiency" motors. Efficiencies of standard motors were
assumed to correspond to those given by ASHRAE-90.1 for installations completed after
1992. Load factors, which represent the fraction of rated (nameplate) power at which
equipment operates at full mechanical load, are estimated based on brake horsepower and
nameplate horsepower values from mechanical schedules or are calculated from flow and
pressure drop values with estimated pump efficiency. Usage factor, or fraction of total
possible operating hours during which the equipment is actually running, and demand factor,
an indication of coincidence between peak use by the motor driven equipment and the
building peak, were determined based on results from DOE-2 simulation runs.
Supply fans, return fans, chilled water pumps and hot water pumps with nameplate ratings
of 1,235 hp, 225 hp, 290 hp and 150 hp are considered for the current design. For all
motors considered (1,900 hp), energy, demand and cost savings were estimated as 197,000
kWh/yr, 460 kW/yr and $9,600/yr, respectively. The total cost premium for the 67 motors
considered is estimated as approximately $31,280, giving a simple payback of about 3.3 years.
Efficient Belt Drives
The key assumption for this measure is from data supplied by Gates Rubber Company
(Gates, 1990), which indicates that an average power reduction of approximately 5% can be
realized by replacing standard V-belt drives with synchronous drives (also referred to as
HTD or cog drives). Supply fans are V-belt driven centrifugal type fans, while return fans
are axial direct drive. For 39 fans, energy, demand and cost savings were estimated as
-------
2674
144,000 kWh/yr, 370 kW/yr and $7,300/yr, respectively. Based on estimates from local
contractors, the implementation cost for this measure is estimated as $47,400, giving a simple
payback of about 6.4 years.
Occupancy Sensors
Considering the usage schedules and other characteristics of spaces for which design is
complete, no significant savings are possible due to this measure. All areas considered will
require design lighting levels continuously during peak (occupied) periods and overall
building lighting controls are designed to shut off lighting during unoccupied periods.
Glass Overhangs
Potential savings due to installation of glass overhangs with width to window height ratios
from 0.2 to 0.8 were estimated using DOE-2. The results indicate that energy and demand
savings for this measure are quite small; on the order of about 1-2% of total electrical usage.
Simple payback on this measure was more than 50 years. This is primarily due to the
excellent shading characteristics of the glass in the current design. Reducing solar gain
through use of glazing with good shading characteristics appears much more cost effective
than installing structural shades.
Upgraded Insulation
Potential savings due to installation of additional insulation were also estimated using DOE-
2. In this analysis, it was assumed that the wall R-value would be raised from 12.5 (existing
design) to 19.0 and that the roof R-value be raised from 20.0 (existing design) to 30.0. Cost
reduction is estimated as only about $6,000/yr. This small effect is due to the fact that the
bulk of the annual load is caused by ventilation, infiltration, internal gain and solar loads;
conduction losses are small in comparison. Considering costs (material only) of SO.SO/ft2 for
R5.5 fiberboard wall insulation and $0.75/ft2 for R10.0 urethane roof insulation, the total
material cost would be about $350,000. Simple payback on this measure is over 60 years.
Heat Recovery From building Exhaust
In the new airport HVAC design, outside air flows are relatively small, limited to a minimum
of 20% when outdoor ambient conditions are not appropriate to allow use of the air-side
economizers. About 5% of this air is used to pressurize the building, 5% flows out through
restroom exhausts and 10% is relief air. The distance between intakes and exhausts is also
significant, increasing the cost of recovery equipment. Finally, since the cooling is to be
provided by central plant gas chillers, electrical demand and energy savings due to this
measure are negligible. Simple payback for this measure would be greater than 15 years.
SUMMARY
Energy conservation is clearly one of the major focuses in the design process for the new
airport. Designers and utility representatives are striving to develop a state-of-the-art
conservation / load management facility by incorporating as many cost effective measures
as possible. This study has shown that significant savings can be realized by installing
daylighting, efficient motors and efficient drives. Occupancy sensors, glass overhangs,
-------
2675
upgraded insulation and heat recovery do not appear cost effective at this time. DOE-2.1D
was the key analysis tool used to handle the many complexities of this large building. The
daylighting algorithms were particularly useful, allowing effects of lighting, heating and
cooling to be considered simultaneously.
ACKNOWLEDGEMENT
This work was supported by Public Service Company of Colorado. The authors are pleased
that the utility is very interested in energy conservation and its support for this effort is
gratefully acknowledged.
REFERENCES
Bazjanac, V. and Winkelmann, F. (1988). Daylighting Design for the Pacific Museum of
Flight: Energy Impacts. LBL-3617.
Energy Efficient Design of New Buildings, American Society of Heating, Refrigeration
and Air Conditioning Engineers (ASHRAE), Standard 90.1-1989.
Gates Rubber Company, Remote Access Databank, 1990.
Goldstein, B. (1980). New Crystal Palace. Progressive Architecture. 12 76-85.
Kirkpatrick, A., Bleem, J. and Winn, C.B. (1990), Design of an Energy Efficient Terminal
Building for the New Denver Airport, 1990 World Renewable Energy Congress,
Reading, England.
Landsberg, D., Misuriello, H. and Moreno, S. (1988). Design for Energy Efficient Atrium
Spaces. ASHRAE Transactions. 94 310-328.
Ventilation for Acceptable Indoor Air Quality, ASHRAE, Standard 62-1989.
Motor Efficiency Data from Marathon, Baldor, Lincoln, AO Smith and Magnetek
(1989 catalogs).
-------
2676
DESIGN AND PERFORMANCE MONITORING OF
GREEN PARK COMBINED SCHOOL
A.P. Waterfield and B. Norton
PROBE, centre for Performance Research On
the Built Environment, Department of
Building and Environmental Engineering,
University of Ulster, BT37 OQB, N. Ireland.
ABSTRACT
The innovative design of a school building is described and the
approach adopted to achieve design analysis and non-invasive
long-term performance data is outlined. Included are examples of
graphed data from the early stages of the monitoring process.
Roof-space collector; schools; gas-fired warm-air heating system;
Energy Management System; remote performance monitoring.
The recently completed Green Park Combined School in Newport
Pagnell, Buckinghamshire, England (see Fig. 1) is the result of a
design process which sought to reconcile passive solar design
with the diverse functional constraints on school architecture
(Norton and Waterfield, 1990).
KEYWORDS
INTRODUCTION
roof-space
south zo

hall
north zone
main entrance
Fig. 1. Axonometric of Green Park Combined School
-------
2677
The 1400 m2 single-storey school houses 360 pupils aged five to
twelve years. The building encloses a central courtyard which
provides natural daylighting to the interior. The design incorpo-
rates, as its principal passive solar feature, three "roof-space
collectors" (see Fig. 2) which preheat the fresh air feed to the
gas-fired warm-air heating system. Thus, ventilation heat losses,
which may account for over one third of the total annual primary
energy consumption of a typical U.K. primary school, are effec-
tively reduced. Though located in a suburban residential area,
the arrangement of the site involves no overshading of the
south-facing facades.
The pertinent meteorological data for the site are as follows:
* prevailing wind is from the south-west
* total annual insolation on a horizontal plane is 3402 MJ/m2
* average daily temperature is 1.6 °C (min) and 21.8 °C (max)
* there are 2308 "U.K. standard degree days"
A roof-space solar energy collector (RSC) (Norton et al, 1987)
essentially comprises a pitched roof to a conventional roof-
space, glazed on its southerly aspect. Fresh air entering at the
eaves is solar-heated before being conveyed to the heating sys-
tem. Two schools in the U.K. currently incorporate such systems;
Perronet Thompson School (Williams et al, 1989) and Green Park.
THE ROOF-SPACE COLLECTORS
diffuse radiatioq
direct N
radiation
exhaust air
solor
warmed
heat
recovery
unit
\35 'overheating
i Jan' /
mois! oir Irom
w.c etc.
iccirculated air
warm air to
heoted zone
gas fired
worm air
heater
Fig• 2* Schematic of Airflow in a. roof~spac© collsctor.
-------
2678
In the latter case, return air is either recirculated to the
heaters or vented at the roof ridge via a heat-recovery unit.
Waste moist air from cloakrooms and W.C. is also passed through
the heat-recovery unit before being vented (see Fig. 2).
As with conservatories, air withdrawn from the roof-space
collector need only be at a temperature above ambient, that is,
not necessarily above that of the building, to provide a benefi-
cial effect, since any warming of the air passing through the
roof-space will provide a reduction in the ventilative heat load.
In summer, the collector may be isolated from the rest of the
building by simply shutting down the fans. At a temperature
setpoint, the roof-spaces are vented automatically, via heat-
actuated controls to fans in the gable ends.
Roof-space collectors can have a low capital cost, since they
utilise an existing architectural feature, the construction of
which does not differ greatly from that of a conventional roof.
The only additional requirements are twin-walled polycarbonate
"glazing", fans, ductwork and controls. The running cost should
also be low, being only that incurred by running the extra fans
over that normally associated with a warm-air heating system.
THE MONITORING SCHEME
Intrinsic to the Allen-Martin Energy Management System, installed
to monitor and control the environmental conditioning of the
school, were most of the sensors required for long-term energy
performance monitoring. A full list of logs is given below.
PARAMETER
LOG RATE
RECORDING
Gas Meter Reading
(daily)
therm
Electricity Meter Reading
(daily)
kWh
Kiln
(daily)
on/off
Gas Hot Water Boilers
(1/2
hr)
on/off
Pumps
(1/2
hr)
on/off
Gas Warm Air Heating
(1/2
hr)
on/off
Fans
(1/2
hr)
on/off
Electric Water Heaters
(1/2
hr)
on/off
Room Temperatures
(1/2
hr)
°C
Roof-space Temperatures
(1/2
hr)
°C
Ambient Temperature
(1/2
hr)
°C
Return Air Temperatures
(1/2
hr)
"C
Heat Recovery Supply Fans
(1/2
hr)
on/off
Extract Fans
(1/2
hr)
on/off
Roof-vent Fans
(1/2
hr)
on/off
Fresh Air Dampers
(1/2
hr)
on/off
Pyranometers (x2)
(1/2
hr)
volts
In addition, air flowrate measurements were made in the heater
ducts, to establish mass and heat transfer rates. All energy
flows may now be determined, directly or indirectly. At 00:00
each day, the EMS is interrogated automatically from the Univer-
sity of Ulster, using a modem. Thus is accomplished the remote
monitoring from Northern Ireland of a school in England.
-------
2679
i Wlllltlllll
15 IS 21
0 IliilltlHIli>»
I I I I l I r
12
Local Uma (hoir)
Variation of RSC
temperatures
with ambient and
Insolation
Mon 25.02.91
E
5
S_^
C
a
a
TEMPERATURE
x
north RSC
+
south RSC
*
hall RSC
><
ambient
INSOLATION
3. Variation of roof-space collector temperatures with
ambient temperature and insolation (school occupied)

i j

£:
\
¦

BOD
700
600
9 12 15
Local time (hoir)
18 21
Variation of RSC
temperatLres
with ambient and
Insolation
Sin 24.02.91
¦100
500 E
\
5
400 ^
a
300 S
c
200
TEMPERATURE
z
north RSC
+
south RSC
M
hall RSC
x
ambient
INSOLATION
4. Variation of roof-space collector temperatures with
ambient temperature and insolation (school unoccupied)
-------
2680
Variation of zone
temparatires
with ambient and
Insolation
Mon 25.02.91
TEMPERATURE
north zona
resaLrca area
south zone
ambient
^lllllilltlll 0
18 21
11 I I I I 111 I 11 11 I
D 3 6 9 12 15
INSOLATION
Local time (hoir)
Fig. 5. Variation of internal zone temperatures with ambient
temperature and insolation (school occupied)
u
o
o *
k.
CD
a
£i

: i i :
A
¦

| 4tj
[ |

800
700
-600
500
400
[-300
200
100
Variation of zone
tamperafires
with ambient and
Insolation
Sin 24.02.91
E
5
9 12 15
Local time (hoir)
18 21
TEMPERATURE
x
north zone
+
resoirce area
*
south zone
D
hall
x
ambient
INSOLATION
Fig. 6. Variation of internal zone temperatures with ambient
temperature and insolation (school unoccupied)
-------
2681
EARLY RESULTS
Figures 3-6 show graphs of the variation, with ambient tempera-
ture and insolation (global, on a south-facing inclined slope) of
internal zonal temperatures and roof-space collector temperatures
for two consecutive days, during which the school was unoccupied
and occupied respectively.
Sunday 24.02.91 was a cool and cloudy day, during which the
ambient temperature varied little, and the insolation level was
low. Consequently, the roof-space collector temperatures varied
little (see Fig. 4), while following the pattern of insolation.
Hardly any variation is observed in the temperatures of the
internal zones (see Fig.6), reflecting the latter's greater
isolation from solar radiation effects, compared to the RSCs.
Further, internal temperatures are uniformly low, reflecting the
lack of occupants.
Monday 25.02.91 was a cool but sunny day, during which insolation
reached peak values for this location, but ambient temperatures
remained low, (dipping considerably during the early hours of the
morning, indicating clear sky conditions). Consequently the
roof-space collector temperatures (see Fig. 3) varied considera-
bly over the course of the day, following the insolation pattern
closely. Internal zone temperatures (see Fig.5) rose uniformly
from stable base levels to thermostatically determined values,
coincident with periods of occupancy, before falling away again
later in the day.
CONCLUSION
Overall, these short-term diurnal results seem to indicate that
the school is behaving according to expectations. Continuing
thermal performance monitoring will enable the long-term energy
contributions form the roof-space collectors to be determined.
ACKNOWLEDGEMENTS
The support of DGXII and DGXVII of the Commission of the European
Communities, Brussels, Belgium is acknowledged. The views ex-
pressed herein do not necessarily reflect those of Buckingham-
shire County Council, whose'involvement is also most gratefully
acknowledged. The authors also acknowledge the financial support
of the Department of Education in Northern Ireland, through their
Distinction Award Scheme.
REFERENCES
Norton, B., and A.P. Waterfield (1990). Green Park School - inte-
gral roof-based solar air heating. Sun at Work in Europe. 5. 2.
12-14.
Norton, B., A.L. Crompton, S.N.G. Lo, and S.D. Probert (1987).
Hybrid Roof Space Solar-Energy Collectors. Proc. European
Conference on Architecture (Munich). 423-428.
Williams, A., J. Bower, C. Ratcliffe-Springall, D. Thomas, and
R. Nex (1989). Perronet Thompson School. Building. 45-52.
-------
2682
ANALYSIS, MONITORING AND EVALUATION
OF A PASSIVE SOLAR COMMERCIAL BUILDING
INCLUDING MASS WALLS AND DIRECT GAIN FEATURES:
THE "AUDITOIRES FUL" BUILDING IN ARLON (BELGIUM)
Ph. Andrd, J. Nicolas, J.Fr. Rivez, V. Debbaut
Fondation Universitaire Luxembourgeoise
Avenue de Longwy, 185
B-6700 ARLON (Belgium)
ABSTRACT
The FUL Research Institute in Arlon (south of Belgium) has decided to use passive solar heating
techniques for its new building ("Auditoires"), occupied since October 1986. This two storey:,
building includes a direct gain heated central zone and two indirect gain side zones consisting of
mass walls with integrated interior windows. The insulation level of the building is higher than
Belgian standards. The only glazed facade is south facing and includes roller blinds for night
additional insulation. A gas fired heating system and several mechanical ventilation subsystems
assist the passive solar features.The building has been selected as a candidate for the analysis of the
mass wall passive solar heating system in the context of the "Passive Solar Heating Working
Group" of the IEA solar task XI project.
KEYWORDS
Passive solar; mass walls; monitoring; simulations; design optimization; IEA task XI
INTRODUCTION
In 1985, the FUL Research Institute situated in Arlon, in the southern part of Belgium, decided to
plan a new building in order to provide rooms for the educational service of the Institute :
amphitheaters, offices and meeting rooms. A passive solar design, combining mass walls and direct
gain zones with a high level of thermal insulation, was selected, and the energy savings potential of
such a strategy was investigated in the context of the IEA task XI project, which focussed on
"Passive and Hybrid Solar Systems for Commercial Buildings". Therefore, the building has been
instrumented and monitored during two years'of occupation, from June 88 to June 90.
The aims of the analysis were;
- to measure and analyze the performances of the built mass wall system.
- to propose an optimization strategy with respect to the requirements of both the user profile and
the Belgian climate.
BUILDING DESCRIPTION
The building is descibed in (IEA XI, 1989). The essential passive solar features of the building are:
a large south facing glazed facade, representing 66% of the total facade; a direct gain central zone;
-------
2683
two indirect gain side zones consisting of white painted mass walls including interior windows. The
building is highly insulated, heated by a gas-fired boiler associated with a radiators circuitry, and it is
vented, in case of overheating or pollution,by a partly automatized ventilation system. Finally, a
shading device is provided to the south facing glazing. Globally, the building is seen as thermally
very massive.
ANALYSIS METHODOLOGY
The objective of the study was first to determine whether the building was working or not and to
what extent the passive solar design of the building was responsible for the performance.
Furthermore, the different components of the building, specially those concerned with the main
passive solar features, ie. the mass wall, were to be investigated in detail through extensive
monitoring and computer simulations in order to detect whether their influence on the global building
performance was optimal or not. The analysis started with the subdivision of the building in
components. The buiuing components'definition was closely associated with their participation in
the passive solar behaviour of the building:
environmental components : climate and occupants
glazing and shading device
buffer sunspace
accumulation wall
building zones
building envelope
auxiliary heating system
mechanical ventilation system
control system.
MONITORING
Objectives and Methods
The objectives of the analysis and the definition of the building components determined the
instrumentation of the building. The glazing, buffer sunspace and accumulation wall received a
special attention as they are the main passive solar features of the building. The detail of the
monitoring plan is given in Andre (1989). The monitoring period was first fixed to one year,but
several problems with the auxiliary heating system made a second year to be necessary in order to
get reliable information from the measurements. Furthermore, some "one-time" measurements were
planned : infrared thermographic analysis, pressurization test, heat fluxes measurements. The
measurements were planned with the following objectives :
1.Evaluation of the thermal performance of the building. Investigation of the passive solar
contribution of the mass wall to the heating load.
2.Validation of the Belgian computer model "MBDSA" (Nusgens and Cotton, 1989) against the data
recorded in situ.
3. Application of identification techniques for the determination of the thermal behaviour of a mass
wall. The heat exchange coefficients of the transfer occurring in a mass wall were to be
investigated as the values of these coefficients can be substantially different from those of a more
classical design.(Andre, Nicolas and Rivez, 1989)
4.Evaluation, by means of a Fanger - like methodology, of the thermal comfort in the building.
In order to meet these different objectives and following the division of the building into several
subsystems, the a priori selection of measurement points considered :
- the division of the building in ten zones in order to allow the comparison of measurements with
simulated results (a similar zoning was selected for the modelling);
- the organization of the instrumentation according to the subsystems listed above.
-------
2684
Results
The table 1 lists the auxiliary heating load for the first four years of occupation of the building
(October 1986 to September 1990).
Heating season
Heating load
Heating load/m2

kWh
kWh/m2
1986/1987
108587
163
1987/1988
99181
149
1988/1989
96198
144
1989/1990
97036
146
Table 1 : Heating load for 4 years
These figures can be seen as quite high for a passive solar building. The first year figure is
probably the result of building drying and the consequence of a winter colder than the other years.
The generally high level of the heating load is partly due to a bad control of the auxiliary heating
system. Some measurements of the heating system itself have shown that the night and weekend
setback strategy of the heating system was not effective, at least for one of the two distribution
circuits. Another explanation of the poor overall performance comes from the high value of
infiltration losses. A pressurization test performed by the Belgian Building Research Institute has
shown that the rooms connected to the roof have a high (more than lvol/h) infiltration rate due to the
lack of an air tight layer in the roof. Figures 1 and 2 show the correlation between the auxiliary
heating load and,respectively, the ambient temperature and the global horizontal solar radiation,
calculated on the basis of monthly average values. Both correlations are not very strong (R2 : 0.70
for the ambient temperature; R2 : 0.68 for solar radiation) and this relative weakness suggests that
an external factor is influencing the energy consumption of the building.
o
0
eo
120
1*0
200
0
Fig.l: Correlation q aux / solar radiation Fig.2: Correlation q aux/t ambient
The high level of the heating load is associated with a poor contribution of the mass wall to the
heating of the rooms situated behind it. Several reasons may explain the lack of performance of the
wall:
- The wall is white. Consequently, most of the solar heat is reflected outside or on the floor.
- The wall is located too far form the glazing : the roof works then as an overhang device when the
sun is high
- The opaque part (34 %) of the facade works as a shading device with respect to the wall when the
sun is far from southern azimuth .
- Ventilation ducts are situated in the sunspace. They are painted red and , consequently, absorb
more solar energy than the wall itself. Furthermore, they act as a shading device for the wall.
-------
2685
A regression analysis shows that the wall temperature is more correlated to the external temperature
than to the solar radiation intensity. The mass wal is therefore suspected to work more as an
insulating device than as a solar collection feature in the Belgian climate.
The air temperature has been recorded in every zone of the building. Averaging the recorded values
and calculating the standard deviation for two years of monitoring yield.the following figures (table
2).
Zone
Average temperature (K)
Standard deviation (K)
Auditoriums
19.5
0.8
Central hall
21.5
2.3
Offices
21.7
1.5
Visitors offices
22.7
2.0
Meeting-rooms
20.7
1.4
Table 2: Average temperature and standard deviations for several zones.
This table shows that the average temperatures in the building are quite high, considering the
setpoint of the auxiliary heating system to be 21°C during the day. The auditoriums are the coldest
zone of the building (because of the reduction of the direct gains), with the lowest deviation around
the mean, indicating the damping effect of the mass wall. All solar insolated zones exhibit an
average temperature above 20°C with large deviations around the mean, due to both direct solar
gains and heating setback.
The overheating trend can be investigated by the computation of temperature frequency distribution
diagrams. The figure 3 shows such diagrams for several rooms. Indirect gain zones (auditoriums,
offices) exhibit distributions centered around the heating set point with some underheating in the
auditoriums (because of the amont of thermal mass) and some overheating in the offices (because of
the interior windows which increase the direct gain contribution). The monitoring shows that
overheating occurs during a large proportion of occupation hours in the directly insolated zones.
Even though the building occupants have the possibility to lower roller blinds in case of overheating
feeling, they don't use them very often, probably because they want to keep in touch with the
surroundings and to benefit form the daylighting associated with solar radiation. Openable windows
are located in the roof of the offices area and occupants could use them in order to increase the
infiltration rate of the rooms. But they don't use them very often either. To reduce these
overheating problems, an additional mechanical ventilation system has been installed in the offices
area. Again, the occupants don't use it because of the noise created by the ventilators.
Auditoriums Hall Offices
Fig.3: Frequency distribution for typical zones.
-------
2686
SIMULATIONS
Besides the experimental analysis, a simulation work has been performed on this building in order to
1. Validate and/or calibrate a computer model that could be used as a reliable tool for the simulation
exercise.
2. Provide a reliable tool in order to perform a systematic analysis of the building in order to
establish whether the design is accurate and to derive optimal design guidelines considering both
energy and comfort related evaluation criteria.
The program "MBDSA" (Multizone Building Dynamic Simulator) has been used for the realization
of this analysis. A preliminary work has concerned the validation of MBDSA against in situ
measurements and a comparison with the predictions of the SUNCODE program. Two typical
periods have been selected for the validation exercise: a cold and foggy winter period and a hot and
sunny summer period. The winter simulations showed the big influence of the modelling of
the heating system. The summer simulations showed the generally good accuracy of the
MBDSA results compared to the measurements. Consequently, the validation of MBDSA is agreed
upon,and the program has been used for a systematic parametric analysis of the building.
The program MBDSA has then been used for a systematic parametric analysis of the building. The
parameters that were investigated are the following : (Andre, 1989)
- the orientation of the building
- the glazing type
- the buffer space geometry
- the mass wall absorption properties
- the mass wall thickness
- the roller blinds activation strategy
- the south facade glazing area
- the ventilation rate across the wall
- the insulation of the wall
- the storage material
- the climate
The evaluation criteria that allow the comparison of different designs were chosen' as :
- the total heating load for the evaluation of the energy savings
- the number of overheating hours in the offices (the most frequently occupied zone of the
building)
for the evaluation of thermal comfort.
The comparison of the variations has been performed with respect to two special designs :
1. The "base case" : the real (as built) FUL building design. The comparison between a given
variation and the base case leads to the optimization of the passive solar system.
2. The "reference case" : the design of an equivalent building (same volume, floor area,
architectural shape) realized according to the Belgian construction standards (insulation level,
fenestration,...) but without any special passive solar features.
The conclusion of the parametric analysis reveals that the energy optimum design of a mass wall in
the Belgian climate should include :
- a south orientation
- a double low-e glazing
- a narrow buffer space (wall and glazing close to each other)
- a black wall
- a 25 % glazed ratio in the south facade
- a night activation of roller blinds
- an optimized control strategy of the heating system.
-------
2687
This design optimization leads to a 15% reduction of the heating load.
The simulation technique has also been used in order to give an estimation of the passive solar
contribution to the heating load of the building for both the base case and the optimized case
(resulting of the parametric analysis). The following solar fractions are obtained : (table 3)
Case
Solar fraction
Base
0.29
Optimized
0.33
Table 3: Solar fraction for two designs.
CONCLUSIONS
Measurements on the FUL building show that this building doesn't work as it should. The passive
solar contribution to the heating load is reduced by several design mistakes. The mass wall doesn't
collect enough energy, and the glazing quality is too poor to keep the solar gains inside the sunspace.
Consequently, much of the solar gains entering the sunspaces are lost either by the wall itself
(reflexion towards outside, or by the glazing or by the ventilation ducts (heated by the sun and cooled
by the air pushed outside). The behaviour of the wall is nevertheless satisfactory in summer when it
works as a damping device for the temperatures inside the auditoriums.
Overheating is a major problem in this building as well, specially at the uppeiffloor, where the direct
gain contribution is much more important than at the ground floor. Temperature rises up to 35°C
have been recorded,and most of the cooling devices (shutters, windows, ventilation) are shown not to
be appreciated by the occupants.
Simulations of the FUL building have shown the theoretical solar fraction to be 29 %. This is not so
bacLi result. A survey of the different parameters that influence the performance of the building and
a systematic parametric analysis yield an optimization of the design. This optimized design results in
a 15 % energy savings compared to the base case. Once optimized, the building performs well, with
a theoretical annual heating load close to 340 MJ/m^ gross.
Finally, the analysis has shown that a good combination of experimental results and computer
simulation could work as a method for defining the optimum design of a mass wall building in the
Belgian climate. Simulations allow extrapolations to be made of what has been measured while
measurements establish a link with the real world.
REFERENCES
Andr£, Ph. (1989a). Advanced Case Study. Auditoires FUL fArlonl : Monitoring Report. IEA
Report, FUL, Arlon, Belgium.
Andrd, Ph. (1989b). Parametric Analysis of the FUL Building. IEA Report, FUL, Arlon,
Belgium.
Andr6, Ph., J. Nicolas and J.Fr. Rivez (1989). Application of Identification Methods for the
Determination of Heat Exchange Coefficients in a Passive Solar Commercial Building.
Proceedings ISES 1989 Congress. Kobe, Japan.
IEA Task XI (1989). Basic Case Studies. Auditoires FUL - Arlon, ETSU, Harwell Laboratory,
Oxfordshire, Great Britain.
Nusgens, P. and L. Cotton (1989). MBDSA - Guide de l'utilisateur. Laboratory of
Thermodynamics, University of Lifcge, Belgium.
-------
2688
SOLAR NON-DOMESTIC BUILDINGS IN NEW ZEALAND
Michael Donn & Nigel Isaacs
Energy Research Group, School of Architecture,
Victoria University of Wellington
P.O. Box 600, Wellington, New Zealand
ABSTRACT
The city of Nelson opened its 1800 square metre library at the beginning of March 1990. The architects -
Upstream Design Group - worked closely with the Energy Research Group at the School of Architecture,
Victoria University of Wellington, to optimise the design with respect to the internal environment and energy
use. The resulting light, airy, and comfortable building belies its origin as a used-car sales lot. Extensive use
was made of computer thermal modelling (SUNCODE) and a 1:50 scale model for testing daylight. The paper
discusses the interactions in the design process and the final design. Energy savings from passive solar design
were not expected to be great, but major savings were made by removing the capital requirements of air
conditioning and an additional electricity substation. Preliminary monitoring results indicate the summer
temperatures are very satisfactory, but lighting did not meet the design expectations for energy consumption.
INTRODUCTION AND BACKGROUND
During the 1980's the New Zealand government supported a large number of investigations into energy use in
non-domestic and domestic buildings. Although data collection has resulted in a significant information
resource, there has been little attempt to build low energy solar non-domestic buildings.
This paper reports the design and subsequent performance monitoring of the first non-domestic building
which has attempted to build on the lessons of this extensive research base - a new public library for Nelson
city. The design process was constrained by the feasibility of solar heating, daylighting and natural ventilation
of a deep plan building on a less than ideal, urban site in a provincial town.
New Zealand spans nearly the same range of latitudes as the United States and enjoys a similar lifestyle.
However, there are some basic geographical differences between New Zealand and North America that have
significant effect on energy and building design. New Zealand consists of two main islands, and the climate is
heavily influenced by the sea, which moderates year-round temperatures and maintains a rather high humidity
level. Heating degree days range from 1200°C-days (2100 F-days) in Auckland (the main centre of population)
to 3000°C-days (5400 F-days) in the far south. The total land area is comparable with California, Britain or
Japan,but of the total population of approximately 3 million, most live, in the warmer parts of the North
Island. There is a strong tradition of intermittent heating of homes - as a result,heating energy is only slightly
greater (on average) in the cooler South Island than in the warmer north. (Swisher & Sterios 1987)
Energy costs are not high, as New Zealand is reasonably resource rich. The electricity system is largely based
on renewable resources - hydro and geothermal, although future growth will have to come from coal or
natural gas. On-demand electricity costs around SUS2.3/GJ and night tariff (off peak) SUS6.4/GJ
($NZ1 = $US0.60). Natural gas is only available in the North Island ($US6/GJ),but LPG is available
throughout the country (SUS10.6/GJ).
-------
2689
SOLAR BUILDINGS IN NEW ZEALAND
Solar domestic buildings have been widely studied in the New Zealand context (Breuer, 1988; Donn & van der
Werff, 1990; Isaacs & Donn 1990; Tucker, 1987) at a theoretical and practical level. They have been the
subject of a major promotional campaign in a design awareness and demonstration programme by the NZ
Government. Simple solar heating measures are cost effective, and considerable experience has been gained in
their implementation.
The average New Zealand dwelling is a 100 square meter single storey stand-alone wooden frame building
selected from a standard range, on 1000 square meter plot. Physical constraints on the orientation and
placement on site are of as much immediate consequence as solar design potential to most people "building"
their homes. The programmes and research have been effective because they have now been sold to the major
home builders - both private and government sector. The government promotes an annual award for energy
efficient design, and one public utility has even found it desirable, in a period of major oversupply of
generation capacity, to promote energy efficient housing.
However, apart from a number of "Atrium" interiors (Shum 1990), some of which have had overheating and
other environmental problems, solar energy is not consciously used in non-domestic building practice.
NON-DOMESTIC BUILDING ENERGY USE
Non-domestic building energy performance has also been intensively studied in a number of sector-wide
(Baird & co-workers, 1984; Baird, Donn & Pool, 1982; Isaacs & Donn 1987) and detailed studies (Donn &
Pool, 1984; Bruhns & Baird 1988; Baird & Pool; 1987). These investigations showed that the area energy use
index for non-domestic buildings ranges from 300 MJ/m2 to 1000 MJ/m2, and with heating and cooling
comprising 20% to 50% of the total energy use, but only 6% to 30% of the total cost. Lighting, provided by
expensive electricity, can be up to 80% of the energy running costs.
These results point to the importance of daylighting as a substitute for electricity. Despite these results,
confirmed by case studies from many reports published in other countries (e.g. Databuild 1989), few, if any,
New Zealand non-domestic buildings employ daylighting or other solar techniques to reduce energy use.
NON-DOMESTIC SOLAR BUILDING - NELSON LIBRARY
Nelson is a small provincial city (Latitude 41° 16' S, Longitude 173° 15' E) in the north of the South Island.
The climate is surprisingly benign (2400 (base 18°C) Degree Days) as the region is spared many of the storms
that characterise other areas near Cook Strait. It is the sunniest region in New Zealand - averaging 2418 hours
of bright sunshine a year and 4.2 kWh solar radiation per day. Summer temperatures over 30°C are not
uncommon, with 36°C the highest reported. Winter temperatures can be cool, with a reported low of - 7°C.
The architects, Upstream Design, noted that Nelsonians enjoy their high sun hours, warmth and sea breeze -
dressing accordingly and quick to complain when they seem to miss out on regular exposure to the sunlight.
The central city site became available late in 1988, when a car company merged its two Nelson sales yards.
The building was in two parts - an old repair area that was demolished, and more modern display showroom
and offices that were extended. A new concrete slab floor was laid through the entire 1800 square metres.
Design
The architect's design goals for the library were to reflect social aspects of the users with broad use of "pacific
colours", fresh air and a high level of natural light. The 40 page brief was overwhelmingly concerned with the
provision of space suitable for library activities involving books - mainly issues of storage. Requirements for
the quality of the human environment - temperature and lighting - were succinct, but explicit:
Liehtinp: uniform 400 Lux at working level of 750 mm above floor level. Daylighting shall be
uniform, avoiding the need for supplementary electric lighting under normal outside conditions.
Heating: a winter working temperature of 20°C with an outside minimum temperature of 5°C,
with three air changes per hour under automatic control.
-------
2690
In addition, the electricity supply was limited to a peak power demand of less than 160 kW, with the possibility
of increasing to 200 kW only by installing a new service main. Above this, a large capital expenditure was
required for the construction of a new substation.
The energy interaction of these requirements necessitated an iterative modelling process, parallel to the
overall design process. Following an initial discussion of limitations of the site, the building and the Nelson
climate, the architects prepared a preliminary sketch design. From this the Energy Research Group
constructed a computer thermal model (Wheeling and Palmiter, 1985) and a 1:50 scale model for lighting
tests. Accommodation and cost constraints dictated that there was to be little change to the area and basic
shape of the building from that time, but a range of design options were explored for the roof, lighting and
heating.
Lighting. The daylighting model was tested in an IES standard
artificial (overcast) sky. Various placements of clerestory windows
were investigated. No supplementary lighting was required in much of
the final design for 75% of the working year. To meet the requirement
for control of sunlight, models of a central courtyard and window
shades were tested. To ensure sunlight added to a high quality overall
light level, external light shelves (Figure 1) which doubled as solar
shading were tested, and incorporated.
The building is of a deep plan, with extensive glazing in the original
building to the south, and with further glazing to the north and west in
the extended building. As the building is only single storey, complex
systems of daylight or sunlight collection were not considered. The
brief required the books to be protected from the direct sun, but this
was reduced in significance by the librarian's recognition that the life
of certain types of book - notably adult and children fiction, and
magazines - was related to their handling while on loan rather than
exposure to sun while on library shelves.
Heating and overheating. The range of temperatures in Nelson throughout the year indicated that special care
would be required both during the hot months and the cool months. A research project recently completed by
the Energy Research Group prepared weather files for 22 New Zealand and South Pacific centres (Van der
Werff, Donn & Amor 1990). Thus it was possible to model the building using climate data from Nelson
Airport.
Zone
Spaces
Area
(m*)
P
Lights
ower (k\
Misc
V)
People
1
Adults, Children, Lounqe - inc children's WC onlv
1,227
14.5
0.9
8.2
2
Workroom and administration and branch
221
2.7
2.9
1.2
3
Stack areas
112
1.4

4
Entrance lobbv
77
0.9

0.5
5
Services - staff toilets, lockers, etc
69
0.8
0.1

6
Small internal courtyard
42

TOTAL
1,748
20.3
3.9
9.4
7
Void (ceiling space)
1,671

Table I SUNCODE thermal model zones
Table I summarises the assumptions used for the thermal computer model about zones, floor areas and the
power consequences of the lighting, miscellaneous equipment and the users. The zones were selected to
provide a thermal model that was as accurate a reflection of reality as possible. For example, a ceiling void
zone was required to model heat flows through the roof due to high levels of solar radiation. Lighting power
use was based on the results of the lighting model studies, and is a maximum limit. The miscellaneous power
requirements were provided by the Nelson City Council, and the occupant power is based on the librarian's
estimation of the throughput of users.

Figure 1 External shading
-------
2691
The library was modelled as open to the public from 10 am to 6 pm two days a week, 10 am to 8 pm three
days and Saturday morning from 9.30 am to 11.45 am, with some staff commencing work at 8 am most
mornings. The power figure for occupants was also scheduled to test cooling performance on a hot day with a
large number of borrowers, such as on hot Saturday mornings. Minimum internal winter temperatures were
"set" at 20°C, and maximum internal summer temperatures at 26°C. Heating was modelled as under time-clock
and thermostat control. Studies were undertaken to determine maximum heating power demand.
For the base case, the building was naturally ventilated with the central courtyard bringing fresh air deep into
the building. The critical question was the ability to provide fresh air for cooling in the heat of summer.
Preliminary SUNCODE runs suggested the hours between 12 noon and 4 pm in mid-summer would be the
most critical times. Wind speed and direction were analysed for the average year for these hours and revealed
the predominant wind was from the north, averaging 6.5 ms'1 September through February, and 3.8 ms'1
March through August. These are measured in open country, so in the open suburban area around the library,
winds of 60% of these speeds would be expected. For the base model it was assumed 25% of glazing was
openable and half that was available for air movement. From these assumptions the air flow was calculated for
each zone and time period, and scheduled in the SUNCODE input files.
A number of parametric variations of the base case building were run through SUNCODE to investigate the
effects of: double glazing, tinted glass to the north and west, net curtains, external shading, no insulation and
double insulation. Then all runs were repeated with clerestories modelled, including the corresponding
reduction in lighting energy use. The models were run initially for one year, with the results reported as
monthly values of heating energy use, peak power, maximum and minimum temperatures. Where problems
were identified, supplementary runs reported performance in detail for the relevant days or even hours.
After the initial set of parametric runs, two further iterations were conducted. The trade-offs between summer
benefits and winter costs were examined. Tinted glazing reduced indoor temperatures in the first round by 1°C
from 30°C to 29°C. The second iteration had better shading and ventilation, a new clerestory providing air and
light, plus building operation which opened windows for ventilation in summer when the temperature was
above 15°C. This time the reduction due to tinted glazing was only 0.5°C, increasing the heating energy use
from 161 GJ to 184 GJ with no effect on the heating plant (158 kW).
The final design (Figure 2) incorporates a number of energy-aware aspects: increased thermal insulation
(3.3 m2C/W); extensive use of clerestory windows (turned to face south and maximum length 16.4 m); a
central courtyard for natural light and air; electrically operated high level windows for increased levels of
daylighting and ventilation; manually operated wall-mounted floor level grilles to promote natural ventilation;
electric ceiling fans to provide air movement and fresh air distribution; ceiling-mounted radiant heating panels
under optimiser control; a ventilated ceiling void; and external shades (increased during analysis on the north
and west from 1.2 m to 1.5 m) for high summer sun that double as reflectors for additional daylighting.
Performance
The library has now been in operation for a year, and is largely working to the satisfaction of all users - staff
and public. It is being monitored for one year as part of the government's "Energy Management
Demonstration Programme" with the intention of showing others involved in the design and operation of
libraries how energy costs can be minimised by careful design and management. The results of this analysis
are expected by late-1991.
Lighting. Lighting has not met the design expectations for energy consumption. Most of the electric lights are
switched on regardless of the availability of daylight. Three switches control lighting for the public areas of the
library and make it difficult for the librarians to match lighting supply to daylight availability. The capability is
designed into the lighting circuits to match lighting to need, but this is not brought out to the switches.
Compounding this situation, Nelson is a temperate climate and while regularly experiencing overcast
conditions, but it still has sun more than 57% (de Lisle and Kerr, 1965) of the time throughout the year. The
IES standard overcast sky is a better source of daylight than the blue sky of a sunny day. In the Nelson library,
the north facing windows are designed to admit large amounts of sunlight. Therefore the contrast in internal
lighting conditions between southern zones lit by the (dark) blue sky and the northern zones of the same
library space lit by the sun is too great for most users. The librarians' normal reaction is to turn the lights on.
-------
2692

L LLQ'Zl SCjyJ J QMt
E 3
Figure 2 Floor Plan
Heating and overheating.
The monitoring programme began in
April 1990. Daily minimum,
maximum and 9 am temperature
readings are being taken in three
locations in the library, along with
electricity use from the heating, light
power and night storage meter--
Infiltration has been measured using
the NAHB/AIMS (Song and Fan,
1989) passive sampling system. The
building was also run through a
three day long unoccupied
monitoring routine. This consisted of
a steady state co-heating test, plus
an overnight cool-down.
14-Od-tO OS-Hov-90 23-*>«-»0 1S-Ok-«0 02-Jot-9I
Dot*
Thus far the results of the thermal
analysis are inconclusive. Figure 3 Front M< • • ¦ Workroom CMdr«n» Ub Exttmol o»«-og»
shows the remarkable stability of the
average internal temperatures during
what was identified as the time critical period: summer. The only area of any concern is a small area in the
work room which overeats in summer - due to lack of cross ventilation.
'V
22-4«t-91
-------
2693
CONCLUSION
The conclusions from the energy and lighting modelling were:
1) Savings of at most $2000 per year at current electricity prices could be made though passive solar heating;
2) Daylighting by reducing the need for electric light would give at most a saving of $2000 per year.
3) Through perimeter windows and opening clerestories plus appropriate shading the temperatures inside the
library could be kept to within 2-3°C of the maximum outside air temperature on sunny summer days.
4) A new electricity sub-station (cost $30,000) was not required.
5) Full air-conditioning (budget allowance $70,000) was not needed.
It was interesting that the greatest savings were capital, not running costs. Of particular interest in the study of
a building of this scale was the difficulty of using the clerestories for heating; they were very successful for
daylighting when faced North or South, but they provided only slight reductions in heating energy use in winter
and considerable cost and complication for shading in summer.
The integration of the design and analysis resulted in improvements in both the architectural and energy
consequences of the design. The architects were able to use their skills and imagination,while the analysis
promoted energy desirable consequences and provided reasons not to use less desirable design features. The
close working relationship during the design process gave benefits that were measurable in the form of the
energy savings, and other less quantifiable consequences for a community building. The final vote must come
from the users of the building - and nothing but positive comments have been received.
REFERENCES
Baird, G. and Pool, F. An assessment of Motivation and Methods Underlying Energy Management in 22
Commercial Buildings (1987) Proc ICBEM Conference Lousannc.
Baird, G. Donn, M.R. and Pool, F. "Energy Demand in the Wellington Central Business District" (1982)
NZERDC Report 77.
Baird, G. Donn, M.R., Pool, F. Brander, W.D.S. & Aun, C.S. "Energy Performance of Buildings" (1984)
CRC Press, Boca Raton.
Breuer, D.R. (Pacific Solar Design) "Energy and Comfort Performance Monitoring of Passive Solar, Energy
Efficient New Zealand Residences." NZERDC , Auckland, Report 171, July 1988.
Bruhns and Baird, G. "Production of a Database of Commercial Buildings Sector Energy Consumption (1988)
NZERDC Report P136, Wellington.
Databuild "Passive and Hybrid Solar commercial Buildings. International Energy Agency: Solar Heating and
Cooling Task XI." Energy Technology Support Unit, Harwell, 1989.
de Lisle, J.F. and Kerr, I.S. "The Climate and Weather of the Nelson Region, New Zealand" New Zealand
Meteorological Service, Wellington, 1965
Donn, M.R. and van der Werff, I "Design Guidelines, Passive Solar in New Zealand", (1990) Energy and
Resources Division, Ministry of Commerce, Wellington.
Donn, M.R.. and Pool, F. 1983 Annual Building Energy use Survey for the Wellington CBD (1984), Ministry
of Energy, Technical Publication 026, Wellington.
Isaacs, N and Donn, M.R. "Design and Construction of Low Energy housing for the Elderly - Monitoring and
Evaluation Report 1" (1990) Min. of Commerce, Wellington.
Isaacs, N. and Donn, M.R. "Closing the Loop - Improving Energy Management in Schools." 1987 School of
Architecture, Victoria University of Wellington CRP No 49, Wellington.
Shum, Katrina. "Atrium Buildings - A place for space", unpublished final year research report, Victoria
University School of Architecture, Wellington, 1990.
Song, Ban-Huat and Fan, Jordan. "Easy Way to Monitor Air-Infiltration" Pollution Engineering Vol XXI, No
12 pp 82-84. November 1989
Swisher, Joel & Sterios, Peter. "Development of Passive Solar Heating in New Zealand." Poster Session 8.5,
ISES Solar World Congress, Hamburg, 1987
Tucker, Alan. "Thermal Performance Tests on an Unoccupied Passive Solar House", University of Canterbury,
Dept of Mechanical Engineering Technical Report No 37, 1987.
van der Werff, I., Amor, R. and Donn, M.R. "Standard Data Files for Computer Thermal Simulation of Solar
Low Energy Non-Residential Buildings" 1990 Victoria University School of Architecture, CRP 53.
Wellington
Wheeling, T. and Palmiter, L. "SUNCODE-PC Building Load Simulation Program" Ecotope Inc., Seattle, 1985
-------
2694
A MUNICIPALITY'S EXPERIENCE WITH GOOD INSULATION AND THE APPLICATION OF
PASSIVE SOLAR PRINCIPLES FOR ALL NEWLY CONSTRUCTED HOUSING PROJECTS
IN THE LAST TEN YEARS
Chris Zydeveld
Municipality of Schiedam, Emmastraat 1, 3111 GA Schiedam,
the Netherlands.
INTRODUCTION
Already during more than ten years the municipality of Schiedam encourages
developers and builders to construct houses with a low energy demand for
heating. The municipality refuses cooperation with other than energy conscious
participants .
After visiting ISES, ASES and other meetings we could convince city council of
the potential of high insulation levels and (other) passive solar principles for
energy saving.
Even during energy crises lower heating bills never formed the only motivation
for being energy conscious. From the beginning, air pollution and greenhouse
effect caused by the burning of fuels gave the main motivation. So our energy
conservation program continued smoothly when energy prices lowered again. And
it still goes on.
After the first project, it never was very difficult to get the cooperation of
the building trade. Some developers eveapreferred co build in our town because
of the advertising value this gave them in other towns. (In Holland, our town is
well known now for its energy saving (and other environmentally related)
activities).
PROCEDURE
In the beginning the main passive solar principles, as we learned them in the
USA, were described in a little booklet that was approved by city council as a
rough guideline for builders, developers and architects.
After the first experience with energy conscious projects the aim was clearly
defined. We refrained from the formulation of insulation standards, but choae
® performance standard instead. Low fuel use was the aim, so the standard
was defined in cubic metres of natural gas, our national heating fuel. The
calculated energy use for an average house in an average heating season had to
be lower than a certain amount. During the years this standard descended to
under 50Z of the national average and projects were realised with energy savings
of 80 - 90Z compared to the national average for new constructions.
When average fuel use for heating purposes of newly constructed homes in our
country was about 3000 cubic metres of natural gas, we realised projects that
did less than 1500. Nowadays with the national average between 1500 and 2000,
our average for new construction is 700. In one project no more than between 200
-------
2695
and 300 cubic metres are used for heating (Figure 1/2). (For a rough comparison
one can compare one cubic metre of gas with one liter of kerosene).
Fig. 1. Low energy homes, sun side
Fig. 2. Low energy homes, cold side
NEW CONSTRUCTION IN NEW DEVELOPMENT AREAS
Making jokes about the national b.utch character, we often describe energy
conscious building as a mixed application of two main principles. The first is:
"try to get as much free (energy) as you can" - that translates to passive solar
elements. The second is "give away as little (energy) as you can" - this
translates to insulation and other conservation measures.
During the years architects found their own solutions for our climatological
conditions. Some projects are easily recognized as passive solar designs (Fig.
3/4). Other projects can,not be recognized at all, because preventing heat loss
-------
2696
instead of solar gain was the main design element. If one looks at the map for
a projected new development for 1600 new homes, solar orientation is easily
recognised (Fig 5).
w

Fig. 3. Solar orientation, south view
Fig. 4. Solar orientation, north view
NEW CONSTRUCTION IN EXISTING AREAS
We first started energy conscious building in new development areas where
conditions are easiest. Later on we applied our standard for all new
construction in the whole town: when old homes were replaced by new, or for new
constructions in older areas, between other buildings. Although conditions on
such sites can be more difficult, some successful projects could be realised
(Fig. 6/7).
-------
2697
Fig. 5. New development area

Fig. 6. New construction in old surroundings
-------
2698
;»H
Fig. 7. Other point of view
RETROFITTING
After tackling new construction we also aimed for retrofitting. A project of
more than 400 flats, where occupants only could heat one of their three or four
rooms was retrofitted. Heating bills^with all rooms heated now, are roughly the
same as before (Fig 8/9).
tV#-. "
Fig. 8. Before
-------
2699
Fig. 9. After retrofitting
OBSERVATIONS
In spite of the differences in emerging architecture, some general observations
can be made.
It took us some time to realise that we did not need water filled oil drums or
other artificial masses, because the main building materials in our country are
brick and concrete. Our buildings tend to be high mass buildings from the
beginning.
It was not always easy to get outside insulation accepted because of the unusual
view when the insulation gets covered with the necessery protection layer.
Esthetic, or marketing circumstances often tend to insulation filled cavity
wall solutions, in spite of our preference from the energy saving viewpoint, for
insulation completely on the outside of the mass.
City planners do noteasily accept the idea of total solar orientation, from
(unnecessary) fear for monotony.
Air tight houses save energy and can be very comfortable. Forced ventilation
with an-air-to-air heat exchanger is more and more chosen as the solution for
a good indoor-air quality and humidity control.
CONCLUSION
Our experience shows that a combination of passive solar principles with good
insulation can be applied in day to day building practice without significant
problems.
Thousands of energy conscious homes in Schiedam prove every heating season again
that enormous energy savings can be obtained; the knowledge is available.
It is a shame that this knowledge, which is so easy to obtain in general, is
so badly used. We could do much better, thereby saving our earth.
-------
Intentionally Blank Page
-------
3.7 Daylighting I
-------
Intentionally Blank Page
-------
2703
DAYLIGHT, ENERGY CONSERVATION AND COMFORT IN AN OFFICE BUILDING
Virginia Cartwright and John S. Reynolds
Department of Architecture
University of Oregon
Eugene, Oregon 97403 USA
ABSTRACT
A 2330 m2 [24,000 sf] office building in the USA Pacific Northwest was designed to incorporate daylighting,
passive solar heating, and night flush cooling. Daylighting measurements were made during the fall, winter
and spring of 1989 -1990. Electricity consumption was measured from March 1989 to January 1991. The
results of the daylight measurements, and a comparison of the predicted and actual energy performance are
summarized.
KEYWORDS
Daylighting, night flush cooling, daylight factor measurements, electricity metering, energy conservation
LIGHTING DESIGN CONSIDERATIONS
The headquarters for the Emerald People's Utility District, a publicly owned electric utility with a strong
commitment to both energy conservation and to alternative, renewable sources of energy,is located near
Eugene, Oregon. Eugene is characterized by predominantly cloudy skies and relatively mild temperatures
from October to June. The summers are usually warm and dry, with a significant diumal temperature swing.
Daylighting was identified as an important design objective early in the design process due to its popularity
with the workers and its contribution to energy conservation. The decision to incorporate daylighting
provided the following design guidelines: (A) windows to face either north or south, to avoid the
glare and summer heat gain trom east and west windows, (B) work stations to be located within
Fig. 1. Building Cross Section
-------
2704
2.5H of the windows, where H is the height of the top of the window above the floor [1], (C) the floor
plan to be relatively narrow in the north-south dimension, about equal to 5H plus the width of a central
corridor, and (D) ceilings to be high, with windows up to the ceiling, to maximize daylight
penetration. This was achieved by using a one way structural system. The beams run perpendicular to the
window walls, spanned by precast concrete hollow core slabs. The windows are placed up to these slabs.
See cross section in Fig. 1. The contrast between plentiful daylight at the window, and less daylight in the
interior zone was reduced by (E) the use of light shelves at both north and south windows, and (F)
more glass area provided above than below the lightshelf. The resulting windows are T
shaped, see Fig. 2. Further, (G) window size to be such that an average of 4% Daylight
Factor (DF) is attained. Also, (H) individual dayllghtlng controls such as Venetian blinds to be
installed on the windows below the lightshelves. The design should incorporate (I) central clerestories
to provide Increased daylight for the interior zone of the second floor. To maintain visual comfort, (J)
no direct sun through the central clerestories Is to reach any work station. Baffles parallel to
the clerestories interrupt the direct winter sun, creating considerable visual interest on the ceiling, but
providing only a diffuse light on work areas. The supplementary electric lighting is (K) Indirect
fluorescent lighting to provide greater visual comfort, especially with computer terminals. The
high ceilings facilitate using indirect lighting which blends unobtrusively with the daylighting. To enhance
energy conservation, (L) the Indirect lights are in rows parallel to the windows, with daylight
sensors to turn off lights in proportion to available daylight. Each work station has an individually controlled
task light of 13 watts to supplement the general lighting.
Fig. 2. T Shaped Windows
THERMAL DESIGN CONSIDERATIONS
A desire for energy conservation led to the following thermal design strategies: A) well Insulated wall
and roof construction, and the provision of supplementary passive solar heating. The south facing
clerestories bring that in a B) total south facing glass area to 16% of the floor area in the second
floor. This should lead to a solar savings fraction (SSF) of about 27%[2]. (Movable night insulation at the
windows would have increased the SSF to about 43%, but the added cost and operational requirements
led to its rejection.) The use of C) exposed thermally massive surfaces stores sunny winter
afternoon excess heat. This is provided by the concrete slabs and the concrete block "fin walls" which
support the deep beams at their intersection with the north and south exterior walls.
To conserve cooling energy in summer, D) the building is flushed with night air. In this dry, warm
summer climate, a large diurnal temperature change is reliable. Rather than operating compressive
refrigeration by day to cool a building, in this region, fans can be operated by night to flush the interior with
cool air, and remove the stored heat of the daytime. Providing E) an adequate surface area of
thermally massive material of 1.5:1, mass area: floor area, was deemed sufficient [3] to meet this
need. The structural system was designed so that night air is also flushed through the hollow cores of the
precast concrete slabs for more thorough night flushing.
Vertical baffles of acoustically absorbent material are hung from the ceilings to reduce excessive sound
reverberation in this predominantly open-plan office building. The acoustic baffles are oriented
perpendicular to the windows to minimize interference with daylight distribution.
In the Pacific Northwest, a high degree of daylighting in an office building shifts the primary thermal task from
summer cooling to winter heating. Providing adequate levels of winter daylighting can lead to excessive
levels of summer daylighting, contributing to overheating. This is more a problem for south facing than for
north facing windows at this 44° latitude. In this building, F) deciduous vines on trellises are used to
protect the south glass from the sun, from early May through mid October. The thermal advantage of
deciduous vines is that they admit sun at the spring equinox (cool weather) and block sun at the fall equinox
(warm weather). The changing color of the leaves (Virginia creeper, Parthenocissus inserta or quinquefolia)
-------
2705
affects the color of the interior daylight from pale green in the early spring, to deep green in the summer to
brilliant scarlet in the fall. This process keeps the workers in touch with seasonal change in a beautiful way.
An additional energy conserving measure was the provision of a variable air volume system with economizer
cycle, and a total of 22 electricity sub-meters which allow separate tracking (for most of the building) of
energy consumed by lights, heaters, fans, refrigeration units, water heater, and miscellaneous power.
ENERGY PREDICTION VS. PERFORMANCE
During the design phase, both a daylighting model of typical bays [4] and an energy simulation [5] using
DOE 2.1 C were incorporated to assist design decisions. Details of trellises, clerestory baffles and the T
shaped windows resulted from these design tools. These energy predictions are compared with metered
performance during a period from March 1989 through January 1991 in Fig. 3. Several factors influence this
performance. There are continuing problems with the computer software so that daylight sensors are not
turning off the electric lights to the extent possible. Second, exterior shading blinds protect the south
facing clerestories from summer sun (during the period before the vines' grow to their full size). These 80%
solar-rejecting blinds were left in place during the entire winter of 1988-89, which included one of the
coldest Februaries on record. The blinds also diminish daylight through the clerestories. They were raised
late in October 1989, and still admitted sun through spring 1990. The winter of 1989-90 was by contrast
unusually mild.
It appears that the building is using more electricity for lighting than was predicted, yet some conservation is
evident when installed lighting is compared to lighting energy consumed. As installed, the indirect electric
lighting is predicted to average 21.5 w/m2 [2 watts/sf]. For this 2330 m2 building, the "baseline" prediction
assumed all lights on during all working hours, for a total of 118,000 Kwh/year. Yet the metered electric light
12 month total is about 64,000 Kwh an average of 11.1 w/m2 (1.08 watts/sf). The electric lights are on more
than expected. This could be the result of the location of the daylight sensors which are quite close to the
dark trellises. The shading cloth blind at the central clerestory may also be playing a part in reducing the
daylight levels in the building.
A typical designed-to-current-code Pacific Northwest office building might be expected to use about 158
Kwh/m2 year (50,000 Btu/sf year}, this building was predicted to use about 90 Kwh/m2 year (28500 Btu/ sf
year), and actually used about 126 Kw/m2 year (40,000 Btu/ sf year) in the metered period.
Kwhx1000
120-
100l
80..
60.
40-.
20...
II
If
Kwh/m2
50
-40
-30
.20
ill J..I..1.I l.l

HLAB HLAB HLAB HLAB HLAB HLAB
Heating Lighting Misc. Power Cooling Water Heater Auxil.(Fans)
H= High prediction, "basecase"no daylight or mass credit
L=Low prediction, full daylight and mass credit
A= Actual metered performance, March 1989 - February 1990
B= Actual metered performance, February 1990 - January 1991
Fig. 3. Annual Energy Usage, March 1989-January 1990
MEASURING DAYLIGHT
From October 1989 through April 1990, monthly measurements were made. Interior daylight measure-
ments near solar noon were compared with simultaneous exterior readings to determine the actual daylight
factor. The locations measured include ground floor north-facing work areas, and upper floor work areas on
-------
2706
both the north and south sides. Gossen Pan Lux cosine-corrected light meters were calibrated to each
other before the measuring process began.
The results of the measurements are summarized in Fig. 4. The target average daylight factor was 4.0 %.
The measured average daylight factor varied as follows:
Jan. fog,visible sun Mar overcast April dear
1st Floor 2.55% 4.3% 2.9%
2nd Floor 4.0 % 2.7 % 3.77 %
It is evident that the trellises reduce the daylight factor on the south side during overcast days. On days with
visible sun, the south clerestories perform very well, especially for the south side of the second floor. On
the north side of the first floor, conditions were best on the overcast day, on which the lightshelf "sees" the
bright north half of the sky dome. The horizon to the north is obstructed by trees which reduces the
available light coming through the lower windows on clear days.
The October, November and December measurements were done with many electric lights on and the
building occupied. The interior fighting including electric lamps resulted in a rather even range of daylight
factors:
South 2nd North 2nd North 1st
At window 5.7% 4.1% 2.5%
At most interior 3.7 % 1.2 % 0.9 %
undini!
c ncrc.***)
Fig. 4. Daylight Factors, January, March and April
(distance from window in 3 foot increments)
REFINEMENTS FOR FUTURE DESIGNS
In ths process of measuring the daylight and observing occupant usage of the building, several things
became clear. First, indirect lighting can have a number of failed fluorescent lamps (burned out, or
disconnected) without such non-functioning lamps being evident. Thus, the subtlety of indirect lighting
seems to allow for "energy conservation through neglect". This seems a valid strategy as long as the
diminished fighting levels remain adequate.
Second, in the presence of large quantities of daylight, it is not always evident that indirect electric lights
are on. Workers might be willing to override the computer controls and turn off electric lights when daylight is
plentiful, but they wonl do so if they can't see that the lights are on. Any attempt to increase user
awareness of indirect electric lights will work against the strategy of conservation through neglect.
-------
2707
Third, the light sensors are located high and near the trellises on the side side ol the building. They get a
"view" of the dark trellis and thus give a talsely low light reading to the computer. The software should be
adjusted to compensate for the position of the light sensors.
Fourth, the baffles that are installed parallel to the clerestories to intercept direct sun seem to be unevenly
serving the center of the upper floor by reflecting too much light down to the south, and not enough light to
the north. Fig. 5. Perhaps a tilted or a translucent set of baffles would pass more light on through to the
north side.
Fig. 5. Baffles at central clerestory
Fifth, the trellises provide some shading even in December, and quite a bit of shading in March which is still
a cool month in this climate. Rather than consisting of vertical slats arrayed in a horizontal plane, a future
trellis design might utilize tilted slats to pass winter solstice sun, acting more as an exterior reflecting
lightshelf at all other times.They should be of a light color. Fig. 6.
m
Fig.6. Alternative trellis configurations
Sixth, the indirect lighting strip along the inner edge of the light shelf directs about half of its light to the
ceiling above the light shelf, where it cannot serve work stations. By day these lights are almost never on.
But at night, the area at the wifidow is thus distinctly disadvantaged as it sits in the shadow of the light shelf.
The indirect lights at the light shelf might be redesigned to provide some direct light or a cove light could be
added at the lightshelf intersection with the exterior wall making nighttime lighting much more even. Fig. 7.
Fig. 7. Redesigned lightshelf
-------
2708
Although the building is not yet reaching its full conservation potential, the users express considerable
satisfaction with their environment. The softness and evenness of the daylight is appreciated, and we rarely
saw task lighting in use during our November and December measurements of the occupied building. Also
popular is the persistent coolness in summer. Some local dissatisfaction with winter heating has led to task
heaters under some desks. After these were installed, the overall heating electricity use in that zone
decreased, probably due to lowered thermostat settings.
ACKNOWLEDGEMENTS
We appreciate the continuing cooperation of the staff at Emerald People's Utility District, for access to the
building beyond normal working hours, and for energy performance data and interpretation. This project
was designed by John S. Reynolds AIA (Equinox Design), with Richard Williams AIA (WE Group, PC).
Virginia Cartwright was dayfighting consultant to the project.
REFERENCES
[1] Stein, B„ J.S.Reynolds and W.J.McGuinness (1986), Mechanical and Electrical Equipment for
Buildings. 7th Edition; John Wiley & Sons, New York, pp. 160-161.
[2] Ibid., pp.181-187.
[3] Ibid., pp.189-192.
[4] Reynolds, J.S., and M. S. Baker, "An Office Building for an Electric Utility, U.S. Pacific Northwest,"
Advances in Solar Energy Technology. Volume 4, Pergamon Press, I988.
[5] Reynolds, J.S., "Client and Climate in the Design Process," Proceedinos of Solar '87. 12th National
Passive Solar Conference, Portland, Oregon, 1987.
-------
2709
MONITORING AND MODELING A CLIMATE- AND ENERGY-CONSCIOUS OFFICE BUILDING
M. Steven Baker, Oregon Department of Energy
John S. Reynolds, University of Oregon
ABSTRACT
This paper describes the design process, energy modeling (using DOE-2) , and subsequent monitoring of
an unusual climate- and energy-conscious office building in the USA Pacific Northwest. This two-story,
2,400 m' (24,000 ft2) headquarters for an electric utility is extensively sub-metered to allow for separate
monitoring of lighting, heating, cooling and miscellaneous loads. It has been occupied since March 1988.
Design, construction and monitoring support came from the Bonneville Power Administration (BPA)
"Energy Edge" research program.
KEYWORDS
Daylighting; task heating; night flush cooling; energy-conscious office building; computer modeling;
building monitoring; DOE-2 modeling.
BUILDING DESIGN
The headquarters building for the Emerald People's Utility District, near Eugene, OR, was designed to use
extensive daylighting (Cartwright and Reynolds, 1990), sensor-controlled stepped dimming of electric lights,
seasonal shading using vegetation, and thermally massive construction for winter passive solar heating and
summer night cooling (Reynolds and Baker, 1988). This was accomplished with design features such as an
elongated cast-west plan with high ceilings, locating all work areas within 25' of daylight openings, trellises
with deciduous vines over south windows, exposed precast-concrete ceilings and concrete block for most
wall surfaces. In addition, the hollow cores of the precast roof and second floor slabs arc used to preheat
and prccool air (Reynolds and Mintorogo, 1988). The design process included an unusual degree of user
involvement (Reynolds, 1987), the use of schematic design rules-of-thumb, scale model daylighting studies,
and performance predictions using DOE-2 (Baker, 1989).
From the onset of design, it was obvious that DOE-2 would not be able to closely predict performance for
this design. The interzone daylighting (using clerestories and light shelves), seasonal trellis shading, an
exposed thermally massive structure, and an air flush thru the hollow cores of the precast slabs all were
beyond the capacity of DOE 2.1C. As a result,electrical submeters and a full hourly monitoring system were
installed.
MONITORING VERSUS ORIGINAL DOE-2 MODELING
As an energy-conscious client, the utility agreed to install separate submeters for the building's major spaces
for electric lights, electric heaters, and miscellaneous power, submeters also measure electric use by fans
and the central HVAC system cooling compressor. BPA has also installed a full hourly monitoring system.
Fig. 1 shows a cross-section through the office building indicating major spaces.
-------
2710
The original DOE-2 predictions for a typical year are compared with two years of monitoring data from
March 1989 through January 1991 in Fig. 2. Energy use for space and service water heating are quite close
to the initial model; a mild winter contributed to lower heating demands the first year. Annual cooling
energy use is much lower than the model; the summer was about average. About half of the total cooling
load is accounted for by the HVAC unit for the small computer room, so most of the building is receiving
very little mechanical cooling. Far higher than the model are both miscellaneous power and fans (discussed
below). The second year was a much colder winter and hotter summer. Table 2 compares degree days for
a normal year and these two monitored seasons.
spaced 2x6 wood
trellis for ivy
' planter beyond
North
metal roof
6" rigid insulation
corefloor planks
light baffles
acoustic baffles
4'0" o.c.
duct enclosure
spaced
wood
4" topping slab on
8" corefloor
flight shelf and
light cove
mass—~
concrete
fin wall
Section (north-south) through the Emerald PUD Building
Heating Lighting Misc.Power Cooling Auxil.(Fans) Water Htr.
Fig. 2. Energy Use for March 1989 through January 1991
-------
2711
ANALYSIS OF A TYPICAL SPACE
The monthly readings from 16 submeters and 7 main meters have provided insights into energy use and the
relationship between the users and the systems. In this paper, a first floor space on the north side (Customer
Service) is examined in some detail. This is the simplest of the major building spaces, and most easily
replicated in future buildings. It has flat ceilings, lighting from one side only (north), and no toplighting.
It thus receives no passive solar heating directly. Also, the daylight sensors in these zones have functioned
most reliably in reducing electric lighting, so the potential savings from daylighting are more evident here.
This space is 138 m! (1380 ft1) in floor area and is organized as an open office plan containing 8
workstations: 5 adjacent to the outside wall (north perimeter zone) and 3 adjacent to interior walls (interior
zone). North window area is 29 m2 (291 ft2); the average daylight factor is
29 m2 window
DF„ = .02 = 4.2%
138 m2 floor
Table 1 shows the monitored results from the three separate meters for this space. Also shown are the
building's total energy usage for the central cooling compressor, the primary variable speed drive fan
(packaged VAV system), and the night flush fans.
Table 1 Electricity Use (Average kWh/day)
Average kWh/day from submeters
Customer Service Area Total bldg.

Degree Days

(138 m2)

(2400 m2)

Month
HDD65/CDD
Heat Recept
Lights*
Compressor VAV Fan
N.F Fans
MAR 89
544/ 0
81
3
12
0
72
0
APR
278/ 1
40
3
9
0
79
0
MAY
271 / 14
27
3
8
4
135
0
JUN
87/38
13
3
10
17
112
11
JUL
31/47
11
3
8
24
121
17
AUG
21/81
3
4
9
19
144
14
SEP
65/60
19
5
9
36
139
27
OCT
336/ 0
23
7
14
0
118
3
NOV
522/ 0
56
8
16
0
96
0
DEC
733/ 0
74
10
16
0
197
0
JAN 90
674 / 0
63
9
17
0
78
0
FEB
654/ 0
63
12
17
0
76
0
MAR
498 / 0
37
8
14
0
101
0
APR
340/ 0
22
7
11
1
101
0
MAY
296/ 0
15
5
9
3
185
0
JUN
106/26
5
6
8
28
137
10
JUL
16 /192
5
4
8
86
105
29
AUG
10 /157
5
5
11
126
129
35
SEP
39/42
1
5
13
96
109
14
OCT
394/ 0
25
8
12
9
75
0
NOV
532/0
51
11
15
0
65
0
DEC
952 / 0
73
10
16
0
85
0
JAN 91
803 / 0
79
11
18
0
85
0
* 2,780 connected watts of lighting x 10 hrs/day x 5/7 days/week = 20 kWh/day at full use
-------
2712
Table 2 Climate Degree Days (Eugene)
Year Heating (°F) Cooling (°F)
Normal 4799 261
1989 4216 241
1990 4640 417
ENERGY USE OBSERVATIONS
Monthly electric lighting use ranges down to 40% of what would occur if all installed lighting were operated
during working hours. Yet, electric lighting could be reduced even further. The daylight monitoring team
repeatedly observed electric lights on during relatively bright days. Reduced electric lighting is made
difficult by a complicated EMCS (energy management control system) and occupant concerns with frequent
light level changes. As a result, some lighting circuits are now off the EMCS.
Potential savings are also reduced because of the stepped dimming system used. In order to keep abrupt
electric light level changes to a minimum, the control scheme must maintain hysterisis-the daylight levels
must exceed the actual light level desired by a substantial amount before turning off a step of electric light.
Otherwise, the occupant will experience a large number of articial lighting changes during partly cloudy
weather. The control scheme embedded in the DOE-2 computer program does not account for hysterisis or
control setting delays, so it should be considered a "best case" savings for stepped dimming. Table 2
compares the results of two DOE-2 runs for the Emerald PUD building—an idealized (unrealistic) stepped
dimming system and a stepped dimming system with hysterisis (the target light level where step change
occurs increased by 15 footcandles).
Table 3 Comparison of Davlighting Control Strategy on Projected Lighting Reductions
Percent Lighting Energy Reduction
Time Period (month) Stepped dimming (ideal) Stepped Dimming (hysterisis)
January
38
21
July
92
90
Annual
73
62
The monitoring program includes detailed site visits every six months to audit the building for any
equipment and use changes, interview occupants, measuring lighting levels, and other data. It is evident that
the indirect lighting is subtle enough that some "conservation through neglect" is occurring. When
fluorescent bulbs bum out in the indirect luminaires that supplement daylight, it is not clearly evident. This
saves energy, but also indicates that lights may be on unnecessarily.
In this building, "task heating" appears effective at reducing energy use. The high ceilings yield cooler
temperatures near the floor-fine in summer, but not in winter. The installation of under-desk electric heaters
(evident in Fall 1989, Table 1) increased the miscellaneous power loads, but resulted in greater simultaneous
reductions in the energy used by perimeter baseboard and VAV electric heaters serving the Customer
Service area.
-------
2713
Monitored energy use was larger than expected for the central VAV and exhaust fans. Much of this is a
result of a faulty variable frequency drive controller that used excess energy at standby and low load
conditions. The VAV controller operated erratically from May 1989 through September 1990 and would
force the fans to run at full power even when the EMCS requested only part load operation. This load drew
its maximum from March through September 1990, averaging about 124 kWh/day. This dropped abruptly
to 75 kWh/day in October 1990 when the controller was finally repaired, and has stayed at this level to the
present. On an annual basis, the night flush fans account for a very small portion (under 5%) of this load.
Measured annual miscellaneous power loads totaled about .18 kWh/m! (1.8 kWh/ft3), more than twice the
originally predicted consumption. Installing task heaters contributed to this extra consumption. However,
other studies of new office buildings in the Pacific Northwest (Katz, 1989 and Taylor, 1989) have found
such loads to be about 4 kWh/ftJ year. The initial assumption used in the DOE-2 model became low over
time. Table 4 reports the results of equipment audits gathered during site visits for both the Customer
Service area and the building as a whole. Changes in connected miscellaneous power are clearly evident.
Table 4 Comparison of Miscellaneous Connected Power (kW)
AS-BUILT DOE-2 MODEL
A major effort has been expended to develop an as-built DOE-2 model of the building. The original DOE-2
model used during design development had difficulty modelling daylighting, trellis shading, and the night
flush system. Daylighting was modeled using FUNCTIONS in the DOE-2 LOADS module, based on results
from a daylighting scale model of the proposed building (Baker, 1988). Based on these scale model studies
combined with DOE-2 predictions, the building design was changed to reduce the glass area below the light
shelves. This change was not reflected in other daylighting models due to time and cost restraints. Further,
the trellis was designed and built using larger wooden members than those assumed in the daylighting scale
model studies, resulting in more shading than was assumed. The deciduous vines have been very slow to
mature, providing much less summer shading. These changes are being handled using the Superiite program
for calculating daylighting availability using ray tracing (LBL, 1987). The results of Superiite analysis for
the as-built condition are then usdd to build revised daylight "FUNCTIONS" for the DOE-2 input files that
more accurately reflect daylight availability in the actual building geometry. These Superiite results are being
compared with actual daylighting studies on site for some level of verification.
As would be expected, a number of building design and energy system changes were made between Design
Development and Final Construction. Also, based on detailed energy audits, connected equipment and
schedules of building and system use continue to change over time. These changes are constantly being
reflected in an as-built DOE-2 input model of the building.
The core flush system for preheating and night cooling continues to pose a significant modeling problem.
DOE-2 does not support a plenum with large heat capacity, which would most closely correspond with the
actual system. Problems also exists in the control algorithm builtin to DOE-2 for night flushing. We have
been trying to alleviate this problem using user-defined FUNCTIONS available in the SYSTEMS module
of DOE-2 version 2.ID.
Heavy construction can be modeled in DOE-2 using either the FLOOR-WEIGHT parameter or custom
weighting factors. Unfortunately, DOE-2 has limited capabilities to acccurately model very heavy buildings.
The large exposed precast hollow-core ceiling and block walls pose significant modeling problems.
Customer Service Area
Audit #1 - June 1989
Audit #2 - January 1990
Audit #3 - August 1990
.985
1.646
1.870
-------
2714
The as-built DOE-2 is being fine tuned using monitoring data to better understand energy use in this
complex, energy-conscious building. Table 5 compares predicted energy use for the original DOE-2 input
model with an as-built model that primarily reflects equipment, system, and schedule changes. This
comparison indicates the range of error that can be introduced from standard building and system changes
between "design development" and final construction.
Table 5 Comparison of Original and As-Built DOE-2 Model Results (Annual kWh)

Original
As-built
Space heating
99,500
48,100
Space cooling
15,600
21,100
HVAC fans/pumps
8,050
8,200
Domestic hot water
5,460
3,200
Lights
46,000
72,500
Miscellaneous equipment
13,900
64,700
Total
188,000
217,000
REFERENCES
Baker, M. S. (1989). Modeling Complex Daylighting with DOE2.1C, Proceedings of the American Solar
Energy Society Annual Conference. Denver, Colorado.
Cartwright, V., and Reynolds, J. S. (1990). Integrating Daylighting and Energy Conservation in an Office
Building, Proceedings of the Solar Energy Society of Canada Annual Meeting. Halifax, Nova Scotia.
Katz, G., et al (1989). Major Projects Rule. Phase 2 Evaluation. Final Report, for Seattle City Light
Lawrence Berkeley Labs, Windows and Daylighting Group (1987). SUPERL1TE 1.0 Program Decsription
and Summary. Berkeley, California.
Reynolds, J. S. (1987). Client and Climate in the Design Process, Proceedings of Solar '87. 12th National
Passive Solar Conference, Portland, Oregon.
Reynolds, J. S., and Baker, M. S. (1988). An Office Building for an Electric Utility, U.S. Pacific
Northwest, Advances in Solar Energy Technology. Volume 4, Pergamon Press.
Reynolds, J. S., and Mintorogo, D. (1988). Night Flush Cooling for an Office Building, U.S. Pacific
Northwest, Healthy Buildings '88. Volume 2. Swedish Council for Building Research. Stockholm. 1988.
Taylor, Z., and Pratt, R. (1989). Description of Electric Energy End Use in Commercial Buildings in the
Pacific Northwest. End-Use Load Consumer Assessment Program, Bonneville Power Administration.
-------
2715
SCHOOL HOUSE LIGHT: A CASE STUDY ASSESSMENT OF LIGHTING
1908 Willson School 1938 Hawthorne School
1951 Whittier School 1964 Bozeman Junior High School
Dale Brentrup, Associate Professor
College of Architecture
The University of North Carolina at Charlotte
Charlotte, North Carolina 28223
ABSTRACT
The innovations and technical advancements that are associated with the built luminous
environment have framed, over the past century, the contemporary professional attitudes toward
architectural lighting technology and the related issues of energy efficiency in public school
buildings. This case provides an assessment of the daylighting and electrical lighting contributions
found in four classroom designs. A topical review of these designs, professional literature,
lighting products and research concerns for educational facilities of the particular period of their
design can lend precedents to this historic continuum. It can, as well, provide guidance in the
programmatic planning and design of new classrooms that are appropriate to their intended use
while providing for the energy efficient futures of these predominantly lighting dominated
facilities.
KEY WORDS
Case Studies, Classrooms, Daylighting, Electrical Lighting, Educational Facilities Planning
INTRODUCTION
"Chief among the architect's concerns must be the powerful influence that is exerted on the final
form of the composition by the relationship developed between the school-desk and the lighting
emanating from the school's windows into the classroom.... it may be said to lie at the very
threshold of the subject and ought to govern his plans, a feature seldom thought of at the outset.
Not until the building is finished and ready to receive its furniture, is it found how much more
suitably the schoolhouse might have been planned had the desk, its relationship to the window and
the question of its potential light been first decided." (Robson, 1874)
On initial observation the school classroom of today remains similar to its distant cousins of over a
century ago. The one and two room schoolhouses of the Great and High Plains of the United
States, for example, have not only similar spatial and planning characteristics to their more
contemporary high design counterparts but share a similar expression in how the articulation of
their architectural elements are formed together to provide for the specialized tasks of teaching and
learning. There are, however, many subtle changes that have taken place over that period that have
obscured our vision as to the particular characteristics that these very specialized environments
must possess.
A case study methodology has been followed to provide a vehicle for the investigation into the
implications of daylight and electrical lighting integration of four public school classrooms. Its
purpose is to focus attention on how the evolution of these two forms of lighting technology have
influenced the design of educational facilities. This case provides an assessment of the daylighting
and electrical lighting contributions found in each classroom and develops a dialogue between
regulatory constraint and material nuance. The classrooms that have been evaluated bridge an
eighty year span of contemporary educational facilities design and are exemplary of the architecture
of the period in which each was designed. They were selected for this study because they were
respectively designed in response to the same climatic region and continue to serve the educational
needs of the same community of Bozeman, Montana.
-------
2716
APPROACH
"An architect acquainted with technical papers on natural illumination (daylight) cannot fail to be
impressed with the backwardness in actual practice, as expressed in fenestration and planning."
(Wynkoop, 1945)
Not only do each of these buildings serve their community's educational needs today, but when
seen as a set of public school classrooms designed within different periods from one another each
professes differing positions to current lighting theory. It is that evolution of lighting theory, that
has proven to be a valuable asset in informing this critical review of their physical luminous
performance. Because each design has become a benchmark from which a point in lighting history
can be elaborated and current practice evaluated.
First, the current luminous performance of each classroom was compared in terms of their original
design intent and to the luminous design limits that were pervasive during periodsin which they
were designed. This was done to elaborate the changing role of daylighting and electrical lighting
design theory. It has provided an opportunity to elaborate the significant changes that have
occurred in the areas of material and regulatory intervention during each period. It has also
allowed for the tracing of the methods that have been employed to market acceptance of new
product development, while reinforcing design decision making and eliminating passive systems
design integration.
Secondly, because each of these classrooms had been updated in order to accommodate current
(1989) trends in facilities management theory and planning, they form a platform from which a
critical assessment of contemporary theory can be leveled. Each school in the district was retrofit
through the Institutional Conservation Program (ICP), funded by US DOE and administered by the
Montana Department of Natural Resources and Conservation. Lighting as an integral element to
energy efficient design was not considered an issue for system efficiency. This program was
primarily concerned with the issue of thermal energy efficiency in which the predominate
conservation scenarios included heating fuel, mechanical equipment and insulation efficiency.
Such conservation measures have been commonly associated with extreme climate design
principles in funded retrofit programs. At issue is whether classroom buildings are predominately
skin load dominate or are they internally load dominate and to what degree should their retrofit
address the balancing of those issues by luminous design. These school rooms provided, through
this example, a foundation from which a comparative dialogue was extrapolated that compared the
original lighting design performance to that performance that can be associated with and question
the value of popular contemporary wisdom.
OBSERVATIONS
"The controlling element in the design of any school building is the classroom and its approach.
Orientation is so important that no building can be completed without close study of desirable
exposure and the plans made accordingly. Heating, ventilating, lighting and sanitation are the
essential details of the design - serious design defects in any one would condemn the whole."
(Weeks, 1915)
The observations made by this investigation focus first on a technical and journalistic assessment
that reviewed the advancement of both daylight and electric lighting compared to those
evidenced by each original design. The results of this assessment formed a base from which the
illumination standards for each period were compared to the actual data from which simulated
parametric variations can be discussed. Other technical innovations that were evidenced through a
review of the product literature were evaluated and their implications to this research noted. Each
class was elaborated further through a comparative evaluation of the original design performance
against that of the now retrofitted 1989 classroom. A computer simulation of the as-designed case
was compared to those of their current configuration to develop a concluding dialogue.
1908 Willson School: "Daylighting of a school building is generally of more importance than the
artificial lighting." (Weeks - 1915).
The school is typical of the majority elementary and high schools built and published from 1890
through 1920. They are typified by a two to three story central hall plan with rectangular
-------
2717
classrooms. Plan dimensions no greater than 24 feet deep by 30 feet long with bilateral
perpendicular window walls having a ceiling height of fifteen feet were common
recommendations. This configuration allowed for the daylighting of the classrooms from at least
one side and placed in importance the relationship of the primary window wall in the shaping of
classroom spatial organization and use. Although the IES recommended unilateral lighting from
left as the preferred condition, if another window wall was utilized lighting from the back was
preferred. (Am. Arch't., 1918) Electric lighting typically was provided by pendent type general
diffuse luminaries requiring 3-6 foot candles of illuminance in accordance with the establishment of
the EES code of lighting school buildings. However, the primary recommendation was to maintain
between seven and twelve foot candles of daylight illumination within the space.
The daylight conditions of the Willson School classroom provided a daylight factor (df.) of two
with a total window to floor area ratio of 52% and a window to wall area of 40% for the North and
West window walls observed. The 50% reduction of glazing in the retrofit case yielded a df. of
one tenth of one percent and reduced the window to floor area to 15%. The lowering of the
window to a view level along with an average room cavity reflectence of 50% and a low lumen
maintenance factor of the suspended fluorescent system resulted in a classroom environment below
visual standards of recommended practice.
1938 Hawthorne School: "...The wattage to be employed (for artificial illumination) should be
equivalent to 2 watts per square foot....". (Wilson, 1938)
Typical of the work of the late Twenties, this plan type is seen through the Thirties. It is typified
by a two-story pavilion, doubled loaded corridor plan. It has rectangular classroom plans 22-24
feet deep by 30-40 feet with a unilateral window wall having ceiling heights of 12 feet (Wilson,
1918) as recommended by the National Council of School House Construction.
The illumination standards (ca. 1938) developed interior cavity reflectances of 50% for wall
surfaces and 80% for ceiling cavities. While "sunlight" illuminance (after Wilson) is recommended
to be 15 foot candles. Other issues for electric illumination are broadened in which new sources
such as fluorescent, luminous systems and controls are introduced. G.E. for example,
recommends a minimum of 20 f.c. and the utilization of double switching and photometric control
of glare. (Darley, 1937) While the American Recommended Practice of School Lighting was
suggesting power densities of 2 watts per square foot and doubling the 1918 standards to 18 foot
candles.
With the window walls oriented East and West each classroom in the Hawthorne School contained
44% glazing. The as designed condition yielded a df. of 1.5 with a floor area ration of 20%. The
existing retrofit condition developed a dt. of .01 with a window to floor area ration of .07. The
large drop in illumination is due not only to the up-fits but as well to the condition of the classroom
reflective cavity which is dominated by current teaching methods, not good lighting practice.
1951 Whittier School: "Daylight is bright enough everywhere if we are." (Gibson, 1948)
A post-World War II single story, double loaded linear corridor plan with unilateral and clear story
lighting is typical of classrooms of this period. Classrooms are typically 24-30 feet deep and 30-
40 feet long. They were planned for an activity area as well as individual student desks. This
school building type is most often associated with the "California School" (after Wynkoopo, 1945)
and the daylight/ventilation research of William W. Caudill at College Station, Texas A&M
University and is often referred to as the "Texas School", (after Gibson, 1948)
This researched school type provides a wealth of information to the advancing of daylight theory
and electric lighting technology. Fa?• instance common professional literature was publishing two
sides to the classroom illumination issue, posturing and seeing (Will, 1946) which marked the
application of refined methods of brightness control (Welch, 1946) and drawing attention to ihs
issues that supported facilities planning through total electrical lighting. Recommendations (G.E.,
1948) included ceiling lay-outs for typical plan types. Uniform electrical illumination is supported
by large increases in lamp wattages, while daylight factors of .5-2 are recommended with
brightness goals establishing contrast ratios and increases in room cavity reflectances.
-------
2718
The East and West window walls of the Whittier school, combine clearstory and side lighting
strategies to provide unilateral daylighting that is 70% glazed. This provided 36% window to floor
area ratio and resulted in daylight factors of two. The inherent problems with the as designed
space are a direct resultant of issues of orientation, sun control, exterior and interior consideration
of glare. The existing, retrofit response was to eliminate the clearstory and reduce the side lighting
to 31% of the wall area. This action resulted in a daylight factor of .02 and a window to floor area
ratio of 13%.
1964 Bozeman Junior High School: ".... it may cost more to use Mother Nature than to rel
-------
2719
Therefore, the daylight providing window wall was eliminated as an integral element in the
vocabulary of classroom design. However the criticism that was leveled against the use of daylight
may be more of an indictment of the misinformed educator and design professional than a
functional reality when considering the powerful advocacy for electrical lighting implementation.
At issue for contemporary design consideration are the following issues:
1. The function of the school classroom has not changed dramatically from those that
have described their fundamental programmatic usages delineated 100 years ago.
2. The technological advancements over the same period have changed dramatically,
affecting the users and designers ability to stay current. This leaves the design
professional strongly dependent on:
a) Contemporary literature: to advance the state of his knowledge of design
integration.
b) Current product literature: to advance the state of technical application and
understand the necessity for change.
c) The user: to identify specific conditions that must be addressed in order to advance
the state of facility productivity. Yet it must be noted that most administrators
would suggest that it is the responsibility of the design professional to bring
forward that nuance. Although facilities programming is not considered as basic to
design services.
3. The need to "modernize" can blind the balance that must exist between high and low
technological integration. The resultant problem is that retrofitting or designing new
building stock that are electric energy intensive, non-productive and uncomfortable as
well as tending to perpetuate larger scale environmental problems will pass on to yet
another generation missed opportunities for quality design in which the learning
environment could act as a model for living.
CONCLUSION
"The approach involves questions of planning, as required by school boards for educational
purposes, of day lighting principles, and of structural device." (Wynkoop, 1945)
This paper has articulated a study that was undertaken to provide an understanding of how the
issues that surround classroom lighting have evolved in one particular city. Moreover, and
possibly more importantly, it was undertaken as a method of teaching integrative lighting design
methods. At issue were the questions that must obviously follow from the uncovering of our
Modem Technological Archaeology. The demystifying, materializing and re-articulating of
previous design values in order to give historic context, dispel the myth and seek new solutions for
those issues that surround daylighting, electric lighting and thermal design methods of inquiry in
order to promote their design integration. Of equal importance is the realization that precedent
dominated studies must inform contemporary design practice not only to acknowledge past values
but to extrapolate greater value in order to evaluate the repositioning of accepted design method in
all lighting dominated buildings.
This study had a two fold objective. First, it became a learning laboratory for the teaching of
lighting and energy efficient design methods. Its purpose was to fundamentally ground the
principles of energy efficient design methods for future reference by the students of architecture
and engineering in which case it was intended that this process should provide a historic,
contextual and analytical basis for understanding the issues that frame any building design
problem.
Secondly, it becomes a model that recognizes the importance of establishing precedents that can
provide a guide for evaluative, judgmental decision making process for those people who are
responsible for formulating new guidelines for the design of new buildings. A process that begins
to establish a linage of understanding, within the context of a specific building type, its historic use
of light, relative energy demand and the importance of developmental research, that recognizes the
fundamental principles from which new concepts and products emerge.
-------
2720
The time line implicated in this case study serves to illustrate the degradation or at least eroding
away of an important design latent set of values. These design values implicate three general
groups of people who have influenced the design of our public school environments.
The Design Professionals: (the architect, consulting illuminating and electrical engineers)
The Users: (the administrators, educators, public school boards and parents)
The Pervasions of current wisdom: (the publishers and critics of popular professional
journals and their advertisers of innovative materials, systems and controls)
The evolution of these four classrooms alone indicate a symtomatic condition that identifies the
inability of a professional group to understand how the programmatic influences of a set of
changing technological advancements can subtlety overpower design method.
Because the aspirations of this study are inherently tied to these four schools and to the community
that they serve the conclusions of this study are best put forth in the paraphrasing of the words of
the president of Montana State University (ca. 1945), Earnest O. Melby:
"Each schoolroom should teach the principles of design. Under no circumstance
should children be permitted to learn in inadequate light or under undesirable
conditions of temperature or atmosphere. The school should employ the best
scientific knowledge concerning heating, lighting and ventilation."
The integrative nature of a building's design must be at once, responsive to the thermal as well as
luminous needs of their occupants and that they must do so by the knowledgeable articulation of a
design pallet that is made up of both passive and active intervention.
ACKNOWLEDGEMENT
The process and product of this case study investigation would not have been possible without the
support of the School of Architecture at Montana State University and the assistance of Bozeman
Public Schools.
REFERENCE
American Architect: 1918; "Light in the School House"; August, Vol. 114, p. 236-
241.
Caudill, W.; 1953; Architectural Record. "Studies on Natural Light and Ventilation in
Schools"; Dec., Vol. HI, No. 5; p. 164-166.
Darley, W. G.; 1937; American Architect and Architecture. "Design of School
Lighting"; April, Vol. 150; p. 77-82.
Ketchum and Sharp; 1958; Architectural Record. "Classrooms: The Window Wall";
Aug.
G.E. staff; 1948; Progressive Architecture. "Modernization Policy Benefits of
Planned Lighting to See"; August.
Robson, E.R.; 1874; School Architecture: The Planning. Designing. Building
and Furnishing of School Houses. Leicester University Press; New York: Humanities
Press;, (reprinted 1972)
Reid, Kenneth; 1951; School Planning - The Architectural Record of a Decade.
F. W. Dodge Corporation; New York, N.Y. (The following are reprinted articles from
Architectural Record sighted from this text.)
Gibson, Charles D.; 1948; "Daylight is Bright Enough if we Are"; March; p. 269.
Welch, K.C.; 1946; "Applied Brightness Control in School"; March; p. 211.
Will, P. & Harmon, Dr. D.B.; 1946 "Eyes and Ears in Schools"; Feb.; p. 179.
Melby, E.O.; 1945; "Neighborhood Schools - What Educators Desire"; June; p. 133.
Weeks, William H.; 1915; Architect and Engineer. "Some Lines of Progress in
School Architecture"; August; p. 112.
Welch, K. and H.G. Daverman; 1953; Architectural Forum. "Low Cost of Electrical
Light"; Oct., Vol. 44, No. 4; p. 187-200.
Wilson, William K.; 1938; Architectural Record. "School Lighting Standards-Time
Saver Standards"; April, Vol. 83; p. 152-155.
Wynkoop, Frank; 1945; Architectural Record. "Advances in the Art of School-Room
Daylighting"; July, Vol. 98, p. 90.
-------
2721
INDOOR ILLUMINATION BY A LIGHTSHELF
COMBINED WITH HORIZONTAL BLINDS
Masanori Shukuya and Kiyofumi Ohashi
Department of Architecture, Musashi Institute of Technology,
28 -1, Tamazutsumi 1-chome, Setagaya-ku, Tokyo 158, Japan
ABSTRACT
This paper describes a method for calculation of internal daylight illuminance due to a lightshelf
combined with Venetian blinds. We made an experiment using a scaled-down model of a side-lit
room to validate the method. A comparison of measured and calculated daylight illuminances on the
work plane under a typical partly cloudy sky condition on a summer day showed that the method
we propose is reasonably accurate. A case study shows that the installation of a lightshelf on a
window can allow the lower part of the window to fully transmit diffuse daylight and the upper part
to use direct sunlight for indoor illumination.
KEYWORDS
Lightshelf; Venetian blind; daylighting; calculation of indoor daylight illuminance; experiment using
a scaled-down model.
INTRODUCTION
Lightshelves have been used for indoor daylight illumination. Figure 1 shows their typical
form(Lam, 1986): a lightshelf has an interior part for shading direct sunlight/direct solar radiation
from occupants in the perimeter zone, while at the same time daylight, including direct sunlight, is
reflected on the upper surface of the lightshelf and towards the ceiling. This interior part of the
lightshelf, however, may become an occupants' annoyance because of its hanging overhead. For
rai
.•'¦¦r.r.Z SHiB

Fig. 1. Typical form of a lightshelf. The
interior part of the lightshelf may
possibly annoy occupants because
of its hanging overhead.
Fig. 2. A lightshelf combined with Vene-
tian blinds. The interior part of a
typical lightshelf shown in Fig. 1
is removed; instead, the Venetian
blinds are installed above the
lightshelf.
-------
2722
this reason, we have proposed to combine a lightshelf with Venetian blinds in order to optimize the
performance of a lightshelf as both a daylighting device and a shading device as shown in Fig.
2(Shukuya, 1986).
Prediction of indoor daylight illuminance due to the lightshelf is necessary to have a better under-
standing of its daylighting performance and then to be able to design it properly. This paper
describes a method for daylighting calculation of the lightshelf, its comparison with a scaled-down
model experiment under real sky conditions and a comparative case study of the daylighting
performances of the lightshelf and conventional Venetian blinds.
CALCULATION OF INTERNAL DAYLIGHT ILLUMINANCE
We have developed a method for calculating the internal daylight illuminance due to a lightshelf
combined with Venetian blinds, applying direct sunlight factor, sky light factor and reflected day-
light factor, which were originally defined to calculate the daylight illuminance obtained from a win-
dow with an ordinary shading device under partly cloudy or clear sky condition(Shukuya, 1983).
We use "lightshelf daylight factor" in addition to the above three factors to calculate the contribution
of the lightshelf to indoor daylight illumination. The lightshelf daylight factor is defined as the ratio
of the internal daylight illuminance obtained from the lightshelf to the luminous emittance of the
lightshelf surface. The lightshelf daylight factor consists of three components as shown in Fig. 3;
the first is concerned with daylight incident on the slats of the Venetian blinds above the lightshelf;
the second is concerned with daylight incident on the ceiling surface, transmitting through the win-
dow with Venetian blinds; and the third is concerned with internally reflected daylight inside the
room. Daylight incident on the ceiling that comes from the lightshelf is not uniformly distributed;
the ceiling near the window is brighter than that far inside from the window. Therefore, the ceiling
surface as a light source was subdivided into portions when calculating the second component of the
lightshelf day light factor.
Using the lightshelf daylight factor Lp, the internal daylight illuminance Ep [Ix] is expressed by the
following equation.
Ep = DF Edh + Sf Esh (—) + RF Mrh (—j) + Lp (—) ^
where Df is direct sunlight factor; Edh is direct sunlight illuminance on horizontal surface outside
[be]; Sp is sky light factor; Esh is sky light illuminance on horizontal surface outside [lx]; fs is form
factor subtended by the sky at the window above the lightshelf; Rp is reflected daylight factor; Mrh
is luminous emittance of the ground [Im/m^]; fG is form factor subtended by the ground at the win-
dow above the lightshelf; Mjls is luminous emittance of the lightshelf surface; and fi_s is form factor
subtended by the lightshelf at the window above the lightshelf. Assuming the surface of the light-
shelf to be ideally diffuse,
Fig. 3. The lightshelf day light factor that consists of three components. The first
and the second are direct components; the first is concerned with the win-
dow above the lightshelf; and the second is concerned with the ceiling as a
secondary light source. The third component is indirect, namely, internal-
ly reflected component.
-------
2723
Mls = Pls(Edl+Esl) (2)
where pls is diffuse reflectance of the lightshelf surface; Edl is direct sunlight illuminance on the
lightshelf surface [lx]; and Esl is sky light illuminance on the lightshelf surface [lx]. The (fs/0,5),
(fG/0.5) and (fLs/0.5) that appear from the second to the last term in eq.(l) are weighting coeffi-
cients to take into account the effects of the solid angles of the sky, ground and lightshelf subtended
at the window above the lightshelf. When the lightshelf is horizontal, the form factor fs equals 0.5
and hence the weighting coefficient (fs/0,5) turns out to be 1.0. The sum of fo and fts always
equals 0.5; if there is no lightshelf, then fo equals 0.5 and £ls equals 0.
OUTLINE OF THE EXPERIMENT
To validate the method described above, we made an experiment using a one-tenth scaled-down
model of a side-lit room shown in Fig. 4. Both room width and depth are 8 m. The ceiling height
is 2.8 m. Since the objective of this experiment was to examine the daylighting effect of the light-
shelf, the model has a window with the lightshelf only. The lightshelf we tested is flat; its length is
750 mm (75 mm in the model); and its reflectance is approximately 0.9. Slat's width-spacing ratio,
reflectance and angle of the Venetian blinds above the lightshelf are 1.36, 0.7 and 30°, respectively.
The reflectance of the ceiling, wall and floor are 0.9, 0.5 and 0.2, respectively. The window
orientation is due-south.
The experiment was made under real sky conditions. Daylight illuminance on the work plane that
is 750 mm (75 mm in the model) above the floor was measured at four points along a line
perpendicular to the window: A, B, C and D. The total daylight illuminance on horizontal surface
outside was also simultaneously measured. All measured illuminances were recorded at one-minute
intervals. Since we could not measure the direct sunlight and sky light illuminances separately in
this experiment, we split the measured total daylight illuminance into direct sunlight and sky light
illuminances using an empirical relationship between total daylight and direct sunlight illuminances
that was obtained from our previous daylight measurements.
COMPARISON OF MEASURED AND CALCULATED
INTERNAL DAYLIGHT ILLUMINANCES
Figure 5 shows a comparison of measured and calculated daylight illuminances on the work plane
under a typical partly cloudy sky condition on a summer day in Tokyo. The upper graph shows the
outdoor daylight illuminance fluctuating between 60 and 90 klx, while the sky light illuminance
being stable at around 30 klx. The middle graph shows the internal daylight illuminances measured
2000 11000
Fig. 4. A side-lit room assumed in the experiment; its one-tenth scaled down
model was used. The internal daylight illuminance on the work plane
level was measured at four points: A, B, C and D.
-------
2724
and calculated at point A that is 1 m inside from the window and the lower graph those at point C
that is 5 m inside from the window. The solid line denotes the measured illuminance; the square
and triangular plots are the calculated illuminances. The calculation for square plots was made by
using eq.(l); the calculation for triangular plots was made without using the "lightshelf daylight
factor"; instead, the direct sunlight factor, sky light factor and reflected daylight factor were used as
originally defined, although the ground reflectance, which is necessary to calculate the luminous
emittance of the ground, was modified with the form factors subtended by the ground and the light-
shelf and their respective reflectances so that it included the effect of daylight reflection by the light-
shelf. Although the calculation was made at one-minute intervals, plots are shown at 10-minute
intervals.
At point A that is 1 m inside from the window, the square and the triangular plots are almost the
same and both of them agree well with the measured illuminances. On the other hand, at point C
that is 5 m inside from the window, the square plots agree well with the measured illuminances, but
the triangular plots do not. When we do not use the lightshelf daylight factor, then it turns out that
the distribution of the daylight on the ceiling is ignored and hence the internal daylight illuminance
on the work plane that is deep inside from the window is overestimated. This implies that the use
of the lightshelf daylight factor is eligible to calculate the internal daylight illuminance due to the
lightshelf with a reasonable accuracy, especially where deep inside a room.
120
90
60
30
0
CM

Outdoor daylight illuminance
August 21, 1987
i i i
v/i/^^AaAVV
i i
—v^__Jotal daylight
• Y w V/ » y 1
1 1 1
^^Sky light ^S*Nxvr>:
i i
1 m inside from the window
.So C by eq. (1) 0
a without Lp
i i I
I i
5 m inside from the window
i
T
a .
T
i
a *aaAAaa
c —ffl—m-n nnraciF
10
Fig. 5.
11
12
13
Time [h]
14
15
16
A comparison of measured and calculated internal daylight illuminances at
point A and C. The upper graph shows total daylight and sky light il-
luminances on horizontal surface outside; the middle graph shows the in-
ternal daylight illuminances at point A and the lower graph those at point
C. The square plots denote the calculated illuminances by eq.(l); the tri-
angular plots those without using the lightshelf daylight factor.
-------
2725
COMPARISON OF THE LIGHTSHELF AND VENETIAN BLINDS
We made a computer simulation to compare the internal daylight illuminances obtained from a win-
dow with Venetian blinds shown in Fig. 6-a) and from a window with the lightshelf shown in Fig.
6-b): in case b), the window below the lightshelf has no Venetian blinds since it is shaded from the
direct sunlight by the lightshelf. A room we assumed is the same as that shown in Fig. 4. The
reflectances of the ceiling, wall and floor were assumed to be 0.7, 0.4 and 0.2, respectively. Slat's
width-spacing ratio, reflectance and angle of the Venetian blinds were assumed to be 1.36, 0.6 and
30°. The reflectance of the lightshelf was assumed to be 0.7.
Figure 7 shows a result of the comparison at noon, June 22, Tokyo. The direct sunlight and sky
light illuminances on horizontal surface outside were assumed to be 45 and 35 klx, respectively.
a) Venetian blinds b) lightshelf with Venetian blinds
Fig. 6. Two windows for a comparative case study. The window below the
lightshelf has no Venetian blinds when it can be shaded from the direct
sunlight by the lightshelf: in this case, if the profile angle of the sun is
higher than 54.5 , no shading is needed.
E
E 3
-------
2726
The internal daylight illuminance obtained from the window with the lightshelf is approximately
twice that from the window with the conventional Venetian blinds only. It is important that the
installation of a lightshelf on a window can allow the lower part of the window to transmit the
diffuse daylight and the upper part to use direct sunlight for indoor illumination.
CONCLUSION
We have developed a method for calculation of internal daylight illuminance due to a lightshelf com-
bined with Venetian blinds. Comparison of the measured and calculated internal daylight illumi-
nances showed that the method is reasonably accurate. To demonstrate the effectiveness of the
lightshelf, we made a computer simulation of the internal daylight illuminances obtained from the
lightshelf and from the conventional Venetian blinds. It showed that the installation of a lightshelf
on a window can allow the lower part of the window to fully transmit diffuse daylight and the upper
part to use direct sunlight for indoor illumination.
ACKNOWLEDGMENTS
We would like to thank T. Kataoka and A. Kaneko, who were then graduate student of Waseda
University and fourth-year student of Musashi Institute of Technology, for their contributions to
making an experimental model and performing the measurements. This study was supported in part
by the Grant in Aid for Encouragement of Young Scientists: No. 62750590, the Japanese Ministry
of Education, Science and Culture, in 1987.
REFERENCES
Lam, W, M. C. (1986). Sunlighting as formgiver for architecture. Van Nostrand Reinhold, 95 -
108.
Shukuya, M., and K. Kimura (1983). Calculation of the work plane illuminance by daylight
including the effect of direct sunlight through windows with horizontal or vertical louvers, Proc.
CIE. paper D304,
Shukuya, M. (1986). Lightshclves. J. Arch. Build. ScL 101/1249. 56 (in Japanese).
-------
2727
A DESIGN FOR A PERMANENTLY MOUNTED PAYLIGHTING DEVICE
I. Cowling, S,Coyne and R.Dew
Queensland University of Technology
Box 2434, Brisbane, Australia, 4001
ABSTRACT
Daylighting has perhaps been one of the quieter areas of Solar Energy research over
the last decade or so but nevertheless a number of groups around the world are
actively involved in this field. Normally when daylight enters a room through a window
the majority is directed downward to hit the floor within a few metres of the window.
One area of daylighting involves redirecting the light in such a way that it travels
upwards towards the ceiling and the back of the room to then reflect down onto work
surfaces deep in the room. The results can be significant savings in artificial lighting
cost and a more pleasant environment for people to work in.
This paper discusses the design and performance of a daylighting device that could be
permanently mounted in a window to redirect sunlight into a room irrespective of the
elevation of the sun. Light enhancement ratios of 4 to 6 times are reported in the back
parts of rooms 10 to 15 metres deep. Other potential applications of the concept are
also discussed.
KEY WORDS : daylighting, device, redirection of light, energy savings, results.
INTRODUCTION
In its broadest sense daylighting is the effective use of daylight in the overall
illumination of a room or space. This can be achieved using a range of techniques,
from just the strategic placement of large windows to the sophisticated systems
involving mirrors tracking the sun. The aim in each case is to use. natural light as much
as possible through the room, reducing the need for artificial lighting.
One method of enhancing natural light levels deep in a large room is to redirect some
of the light entering through a window to travel horizontally or upwards into the room to
reflect off the ceiling and upper parts of the rear walls down to the work surface below.
Selkowitz and Griffith (1986) have estimated that daylighting techniques can achieve
reductions in the use of artificial lighting in a room by up to 60%. In office buildings
and other primarily day-time use buildings this can represent a significant fraction of
the total electricity usage.
-------
2728
There are other advantages, both psychological and physiological, of enhancing the
level of daylight in a room. Numerous studies (Lewy, 1982; Erikson and Kuller, 1983;
Aydinli and Krochman, 1985) show that living and working for long periods of each day
in areas receiving low levels of daylight can upset a number of biological cycles that
rely on various components of the sun's radiation, including the uv component, for
their regulation. The bulk of research (Wells, 1965; Longman and Ne'eman, 1974;
Wotton and Barkow, 1983) also supports the view that people prefer daylight to
artificial light, despite the disadvantages of glare, excessive heat, fading etc.
This latter point highlights an important factor associated with installing light redirecting
devices. It is the visual comfort of the occupants (in terms of view, control of sky-glare,
uniformity etc) that will eventually determine the commercial success of a daylighting
system rather than its energy saving capabilities.
A number of different designs have been proposed for redirecting sunlight upwards
towards the back of a room. They vary in complexity from the simple light shelf
(Selkowitz et al, 1983) to designs with specially structured surfaces (Ruck, 1985;
Howard, 1986). Other systems exist (Whitehead et al, 1984) that concentrate and
collect the light externally before dispersing it through a system of light pipes through
the building. Devices using specially constructed holographic film are also under
investigation (Ian and King, 1986).
This enhanced lighting effect deep in a room is most pronounced when direct sunlight
falls on the window. This dependence on direct sunlight however means that the
design has to work as required for all possible angles of incidence of sunlight that can
occur throughout each day of the year. If the daylighting device cannot cater for this
large range of angles, then some form of tracking must be built into the system.
This paper describes the design and performance of a device for redirecting sunlight
into a room. The device was designed to have two properties that were considered to
be important if it were to be acceptable as a commercial product.
The first was that it should have a large acceptance angle. This means that
irrespective of the angle at which sunlight is striking the window the device should be
able to accept and redirect the light as required into the room without needing to be
tilted or adjusted in any way. If the device therefore is to be mounted in the vertical
plane of the window, coincident with the glass, it must have an acceptance angle
equivalent to sun elevations ranging from around 20° up to 75°. Under these
circumstances the device should be able to be permanently installed in the window
and left to perform as required all year round.
The second design requirement was that the devices should have a simple external
profile similar to a pane of glass, namely flat parallel outer surfaces. This ensures that,
with no external dust collecting crevices, it could be easily cleaned. It also would
enable the device to be readily butted against, say, an existing pane of glass or
sandwiched between two glass surfaces for protection if required.
-------
2729
PRINCIPLE AND DESIGN OF THE DAYLIGHTING DEVICE
The basic component of the device is a
single solid section of dielectric material with
a sloping curved base and a v-shaped
trough as the top surface (Fig. 1(a)). The
individual sections stack together to form the
module as shown in Fig. 1(b), creating an
enclosed air gap between each section. The
angles are such that light entering a section
passes through it by total internal reflection,
remaining always within the dielectric until
emerging generally in an upward direction.
A critical part of the design is the curvature
of the base of each section. It is this
curvature that enables the device to work
over a large range of incident angles. In
addition, however, the curvature of the base
means that emerging rays fan out over a
range of angles into the room, rather than as
a beam.
Fig. 1. Profile of the daylighting device
showing (a) a single section and (b) the
resulting module.
Figure 2 shows three rays being traced
through different parts of a section for three
incident "sun" angles, 20°, 45° and 70°.
From the diagram it can be seen that most
of the emerging light would travel deep into
the room to reflect off the ceiling or high on
a rear wall. A small amount of the light for
this design (around 10%) does emerge in a
slight downward direction.
For most applications it is envisaged that the
modules would be fitted in the top 25% to
30% of the window wall. By the very nature
of its action the daylighting device is not
uniformly transparent and will severely
distort the view through it. The device is
preferably placed in the top part of the
window for two reasons. Firstly the majority
o
20
O
45
O
70
Fig. 2. The path of rays through the
device for sun elevation angles of 20°,
45° and 70°.
-------
2730
of redirected light would stay above head height for people in the room, thus reducing
the possibility of glare. Secondly all downward and horizontal vision out through the
window would be retained, (generally agreed as the most desirable), with some
upward vision still possible from points near the window.
In commercial use it is anticipated that the modules would be 10 - 15 mm thick.
EXPERIMENTAL MEASUREMENTS
The first prototypes manufactured for testing were extruded in a high quality acrylic,
noted for its clarity and high transmission. It was also reported to have a high
resistance to ultra-violet radiation.
The major tests undertaken are what are called enhancement measurements.
Detectors are placed in a grid pattern around the floor of a model room and for a
particular angle of incident light illuminance measurements are taken firstly with the
modules present in the top part of the window and then with a full window. The ratio of
the light readings taken with and without the modules is called the light enhancement
ratio.
0 20 40 60 80 100 q 20 40 60 80 100
Depth into room (cm) Oepth into room (cm)
Fig. 3. Light enhancement ratios down the centre of a model room.
-------
2731
A typical set of light enhancement results are shown in Figure 3. The room grid for
this set of results was 10 (window to back wall) x 3 (side to side). However, no results
were taken where light fell directly on to the detectors near the front of the room. This
is particularly apparent at the lower elevations where the large window used restricts
readings to the back of the room only. In the graphs shown here only the readings
obtained from the detectors down the centre of the room are presented.
The results all show typical light enhancement factors of 4 to 6 for all angles of
incidence through the back half of the room, with some values approaching 10 times.
The model room used here has an equivalent depth of around 10 metres if the floor is
taken as a normal work surface. Light enhancements of this level provide light levels
of greater than 400 lux all through the room, indicating that under these conditions little
or no artificial lighting should be required in the room for many general office tasks.
OTHER DESIGNS
Other designs are being developed for different applications. One such design is for a
module that could be mounted at a small angle to the building (around 30°) resembling
a shade over the window. This device would catch much more of the light from the
high elevations and would be more useful in locations where external structures block
light from the lower part of the sky.
Another development aims to nullify any imbalance of light across a room when the
sun is at a large oblique angle to the window. Additional surfaces within the device
enable light to be reflected obliquely across the room, providing a more uniform
illuminance in this dimension all through the day.
Another design under consideration is for a device that automatically controls the
amount of light able to pass through into the room. This provides a way of maintaining
a more constant level of illuminance inside, despite the varying external conditions. In
addition, we are looking at the feasibility of a module for a south-facing window,
(southern hemisphere), that only receives diffuse skylight.
COMPUTER MODELLING
Computer programs have been written to model the performance of the daylighting
device and to determine the subsequent lighting effect within a room. As reflections
from high in the room provide the major lighting component deep in a room.it was felt
that existing daylighting software (eg Superlite, DOE-2) may not be able to provide
sufficient accuracy for comparison with our experimental results.
Further developments of this program will allow an integration of the daylight
contribution with that of any artificial lights in the room. This will then form the basis of
a potential energy savings program.
-------
2732
DISCUSSION
Initial testing of our daylighting devices has proved very encouraging, and more
rigorous testing covering all aspects of commercial use and acceptability is continuing.
An important aspect of the commercial viability of the device is the potential energy
savings and envisaged pay-back period. Discussions with manufacturers have
indicated an estimated cost of the modules of A$200 per m2. Assuming energy savings
from lighting alone of only 30%, calculations indicate a payback period of the order of
2 years for commercial applications. A consideration of the benefits of daylight
providing a more healthy and pleasant working environment, reflected in an increased
productivity, however, provides an even greater financial return suggesting a payback
period closer to 1 year.
ACKNOWLEDGMENT
The authors would like to sincerely thank the Energy Research and Development
Corporation of Australia for their financial support of this project.
REFERENCES
Aydinli, S. and J. Krochmann (1985). CIE Journal. 4. (2), 39-48.
Erikson, C. and R. Kuller (1983). Proc CIE. 20th Session.
Howard, T. (1986). Proceedings of the International Davliahting Conference. Long
Beach, California, 306-309
Ian, R., and E. King (1986). Proceedings of the International Davliahting Conference.
Long Beach, California, 279-287.
Lewy, A. J. (1982). Am. J. Psychiatry. 139. 1496-1498.
Longman, J. and E. Ne'eman (1974). Journal of Architectural Research. May, 24-29
Ruck N. (1985). Building Research and Practice. 13. 144.
Selkowitz, S., M. Nawab, and S. Matthews (1983). Proceedings of the International
Davliahting Conference. Phoenix, Arizona, 267-269.
Selkowitz, S., and J. Griffith (1986). Lighting Design and Application, 16. (3), 34-47.
Wells, B. W. P. (1965). Building Science. 1. 57-68.
Whitehead, L., D. Brown, and R.Nodwell (1984). Energy and Buildings. 6. 119.
Wotton, E. and B. Barkow (1983). Proceedings of the International Daylighting
Conference. Phoenix, Arizona, 405-410.
-------
2733
DESIGN CURVES FOR DAYLIGHTING IN TROPICS
R. Muthu Kumar*, B. Bhattachaijee* and J. Cook*
*College of Architecture and Environmental Design, Arizona State University, Tempe,
AZ 85287, U.S.A.
#Civil Engineering Department, Indian Institute of Technology, HauzKhas,
New Delhi 110016, INDIA
ABSTRACT
A simplified tool for the analysis of fenestration for daylighting in the tropics is proposed. Indian
sky conditions are used. The basic equations for the daylighting are first studied. TTie influence of
building variables such as fenestration area, room height, sill height, working plane height,
reflection coefficient, aspect ratio of the openings, etc. have been analyzed by computer program.
The results have been discussed, design graphs and other relevant conclusions have been presented
which use four correction factors and an average daylight factor to provide an estimation of average
illumination.
KEYWORDS
Daylighting; Tropics, daylighting; design tools, daylighting; fenestration parameters; sky
condition, tropics; aspect ratio, daylighting; location, daylighting; reflection coefficient
INTRODUCTION
The importance of daylighting in buildings in tropical climates is well established. But most of the
literature pertaining to daylighting, deals with skies which do not match a typical tropical sky,
particularly to conditions in India. However Saxena and Narashimhan (1969) have presented a
sky condition which is suitable to the Indian situation, obtained as a result of experiments
conducted under Indian conditions. The Indian Standard Code of practice has also been developed
from the same source. The methods available for daylighting analysis (Hopkinson, 1963;
Koenigsberger, Ingersol, Mayhew and Szokolay, 1984; Lynes, 1968; Walsh, 1961; Indian
Standard Institution, 1976) at present, mainly concentrate on the calculation of illumination at a
specified point and do not discuss the influence of various window parameters. Design graphs
which can be used during forward analysis of the building design are difficult to find. The
objective of this study was to develop correlation amongst the building parameters and the indoor
illumination due to daylight, so that window descriptions for the desired spread and penetration can
be achieved.
METHODOLOGY
After developing a computer program based on the equations proposed by Saxena and
Narashimhan (1969), it is validated and sensitivity is tested through a full sized experimentation
carried out in an existing room. Several computer runs are made changing one building parameter
at a time. The results are carefully analyzed and plotted to give design curves which can be used to
get average daylight factor in a room with certain design specifications.
-------
2734
PROGRAM STRUCTURE AND LIMITATIONS
The program is developed in a structured fashion with easy user interface. The main program
simply calls various subroutines and functions, which is the style of top down programming. The
data file is edited through the menu and the output is written in a separate file. Since the objective
of the study is that only one parameter has to be changed at a time, an extensive menu is created.
Thus the user is not troubled in editing the data file since the program itself edits the data file
through interactive statements. The subroutines and functions are tested individually and compiled
separately. However all have been linked to give the final executable file. The program is efficient
in calculating the daylight factors for both horizontal and vertical plane points. The output also
contains daylight factor along with Sky Component (SC), Externally Reflected Component (ERC),
Internally Reflected Component (IRC) and Lux Level based on the specified design sky. The
effect due to shading and external obstruction has also been considered. However the program has
the following limitations, a) it cannot process for rooms of odd shapes other than rectilinear forms,
b) the fenestration must be rectangular, c) the ceiling must be flat, d) the external obstruction
should be regular in shape, and e) the disturbance of daylight due to some internal obstructions like
partitions cannot be simulated.
EXPERIMENTAL COMPARISON
To check the validity of the equations used and also the software, daylight factor contours in a
given room have been compared with experimental values. A class room in Indian Institute of
Technology, New Delhi, latitude 28 °N, is taken for the study. The class room is 5.9 m x 4.7 m,
has a full glazed window on one side, which is shaded by a wide overhang. The building block
adjacent to it formed an external obstruction. The experimentation has been carried out using two
lux meters of different range. The class room has been divided into grids of lm x lm. At each
grid node, over the working plane, lux levels have been measured along with the simultaneous
measurement of outdoor illumination from the sky. The necessary room dimensions, window
dimensions, sill height, working plane height have been measured. The reflection coefficient of
surfaces have been found by measuring incident and reflected light. The daylight factor contours
are plotted to find out the differences; a meagre 7% variation. "Ilie reasons for the variation were
thought due to a) the maintenance factor taken for the glass may not be correct b) the design sky
taken in the software may be low and c) the light diffused through the entrance corridor could not
be calculated with the program.
COMPUTER EXPERIMENTATION
Having developed the program the following parameters related to fenestration for daylighting has
been taken for analysis: a) area of opening, b) room height, c) sill as compared to the working
plane height, d) reflection coefficient of walls, e) area of room, f) aspect ratio of the opening, and
g) location of window. Computer outputs were obtained by varying a single parameter while
others remained constant and finally interpreted to give necessary results.
Fenestration Parameters
Area of Openings. Openings include both windows and ventilators, which differ themselves only
in sill height and opening height. The area of openings is normally expressed as a percentage of
floor area. The area of openings were varied from 5% to 30% in steps of 5%.
Room Height. The illumination level varies with the room height even for a fixed fenestration area
since the IRC changes as the surface area of the room changes with increased room height,
according to split flux principle (Hopkinson 1963). The room height is varied from 2.8 m to 4.0
m in steps of 0.2 m, which is the normal range of room heights.
-------
2735
Sill Height as compared to working plane height. The sill height of windows does not always
coincide with the working plane height, to get the full utility of opening. Hence to take account of
the above, analysis have been carried out for the following conditions. Working plane height
varying from -50% to +100% that of window height, in steps of 25% such that working plane
height is below and above the sill height respectively for negative and positive increase.
Reflection coefficient of walls. Reflectances of walls normally range from 0.4 (grey or dark
surface) to 0.85 (polished, white washed surface) which affects the daylighting inside the room
especially in the IRC. For simplicity and by normal practice, the ceiling and the working plane
reflectances arc kept constant at 0.35 and 0.7 respectively while wall reflectances are changed and
analyzed. But the program developed also has the flexibility of changing the ceiling and working
plane reflectances.
Area of the room. Typically designers have been dealing with various room sizes in their practice.
To cater to this and to get the effect of room size on daylighting for a given fenestration percentage,
the effect of room area has been analyzed. Rooms of 24 m2,48 m2,72 m2 and 144 m2 were taken
for analysis. External obstructions have not been considered The overhangs were assumed
according to the room area from 0.6 to 0.9 m out from the windows and immediately above the top
of window.
Aspect ratio of the openings. In general, it is known that high windows provide more penetration,
and correspondingly the spread is diffused. To quantify this three cases of width to height ratio
were used 1:1, 1:1.3 and 1:1.6.
Location of window. For a particular fenestration percentage, many combinations of window
locations and number of windows can be proposed. For example windows may be in a larger wall
or in shorter wall, centrally located or at the end, etc. To analyze these effects, about nine various
locations have been considered and studied.
Procedure of analysis
A rectangular room has been taken as the base case. From the daylight factor computed at every
grid node, the iso-daylight factor contours have been drawn. The area between each contour has
been measured using a digital planimeter (Planix 7) three times and the results are averaged. The
average daylight factor between the contour has also been obtained. Later weighted average of
daylight factor has been calculated as follows.
_ A
DF = 1/A J DF dA
where, A is the area and DF is the average daylight factor. When area and DF are discrete
— n
DF = 1 [L Ai (E DFj A,), summed up for "n", areas
RESULTS AND DISCUSSION
From the Fig. 1. it has been seen that the average daylight factor is directly proportional to the
fenestration percentage, and also the average daylight factor is inversely proportional to the room
height. As the room height increases, the surface area is increased and hence the IRC decreases
according to the split flux formulae i.e., the absorption of light is higher when surface area
increases.
In Fig. 2. a correction factor (CF1) is defined by taking the ratio between the average daylight
factor obtained from a position of the working plane height with respect to the sill height,
compared to the average daylight factor when the working plane height coincides with the sill
height. The correction factor decreases as the working plane height is lower than the sill height.
The reduction is because of a) the portion of sky visible is less and b) the portion of the sky visible
has less brightness than the horizon as the altitude increases according to die equation,
Be = Bz- Cosec 9 (Indian Standard Institution 1976)
-------
2736
5 10 15 20 25 X
Fenestration X (% floor orea)
Fig. 1. Correlation between fenestration area
height and average daylight factor
50 75
(bdow)
Working Plone height compared with
window sill height (X of window height)
Fig. 2. Correction factor for window sill
Fig 3.
0.25 0.50 0.75
Reflection Coefficient of Walls
Correction factor for reflection
coefficient of walls
In Fig. 3. another correction factor (CF2) is
defined as the ratio between an average daylight
factor obtained with the given surface
reflectances, and the average daylight factor of
a surface with 0.6 as reflectance. As the
reflectance of wall increases the average
daylight factor also increases.
Correction factor (CF3) in Fig. 4. is defined as
the ratio between the average daylight factor
obtained in any room area, and average
daylight factor obtained in 24 m2 floor area.
The average daylight factor increases with area
of the room and fenestration area.
From Fig. 5. it is clear that the penetration is high if width to height ratio is high. Also the change
in width to height ratio affects only 60% of floor area's daylight and the rest is unaffected. The
average daylight factor rer-ains the same for various width to height ratios of the opening.
Fig. 4.
48 72 96 120
Area (sq. m)
Correction factor for different area of
the room

QS
pect rot
io

r~ - j.u
/r" M

Fig. 5.
0 20 40 60 80 100
% floor oreo having a specified
illumination
Distribution of daylight with different
aspect ratio of openings having same
area
-------
2737
Finally from Table 1 the idea of distribution can be obtained. The correction factor (CF4) is
defined for various window locations, as the ratio between the average daylight factor obtained for
that arrangement of windows, and the average daylight factor obtained when the window is one in
number and is at the center of the longer wall. The average illumination decreases when the
window is in the shorter wall. It also decreases when the window is at the corner of the room
since the area of sky seen from a point is less in these cases. If the number of windows is
increased, the spread of daylight is better while the penetration observed decreases.
TABLE 1. Influence of window placement on davlighting
Correction
Arrangement Factor
% area*
Correction
Arrangement Factor
% area*
Correction
Arrangement Factor
% area *
CF4 = 0.91
30%
CF4 = 0.71
38%
1 ™ 1 CF4 = 0.80
26%
CF4 = 0.86
24 %
CF4 = 0.77
22%
1^™" 1 CF4 = 0.94
28%
CF4 = 0.73
38%
P" CF4 = 1.00
24%
I" """I CF4 = 0.87
44%
* floor area with day light factor less than the average daylight factor of the room. In other
words it is the measure of the light distribution
CONCLUSION
As indicated in the goals of this study some simplified graphs for fenestration analysis pertaining to
daylighting have been developed. The developed graphs can be used effectively by the designers
in tropics for forward analysis stage of design, as follows.
To find the average illumination in a given room
a) From Fig. 1. one has to get the average daylight factor for the given fenestration percentage
and room height
b) The correction factor (CF1) for sill correction is multiplied with the above value from Fig.
2.
c) The correction factor (CF2) is applied from Fig. 3. to incorporate the effect due to
reflectances of the walls.
d) Then the correction factor (CF3) for the area of room from Fig. 4. is also multiplied with
the above value of daylight factor.
e) Finally form Table 1 (whichever is nearer to the given arrangement) correction factor (CF4)
is selected and also applied.
f) Now by multiplying the final value with the design sky (say 8000 lux) the average
illumination in the room can be obtained.
I = DF x CF1 x CF2 x CF3 x CF4 x design sky
where, I = average illumination, DF = average daylight factor, CF1 to CF4 = various correction
factors
-------
2738
Even though a number of parameters are considered here and graphs have been developed, there is
a lot of scope for further work in the area. The equations for sky component for odd shaped
openings may be developed. The effect of glare in daylighting, which is important in the tropics
has to be incorporated. Effects due to external reflectances, including shading device reflectances
have to be analyzed. Further, the influence of the aspect ratio of the room would also deserve
further work.
REFERENCES
Central Building Research Institute 1982. "Fenestrations for daylighting of side-lit rooms - A
simplified approach", Building Research Notes, Number 26, Central Building Research
Institute, Roorkee.
Central Building Research Institute 1982. "Design for Daylighting", Building Research Notes,
Number 35, Central Building Research Institute, Roorkee.
Egan D.E. 1983. "Concepts in Architectural Lighting", McGraw Hill Book Company.
Hopkinson R.G. 1963. "Architectural Physics: Lighting", London: Her Majesty's Stationary
Office, London.
Hopkinson R.G. 1967. "Sunlight in Buildings", Proceedings of the C.I.E. International
Conferences, Bouwcentrum International, Rotterdam.
Hopkinson R.G. and J.D. Kay 1969. "Lighting of Buildings", Faber & Faber, London
Indian Standards Institution 1976. "Indian Standard guide for Daylighting in buildings (IS 2440)",
Indian Standards Institution, New Delhi.
Kittler R. 1965. "Standardization of outdoor conditions for the calculations of the daylight factor
with clear skies", Vontrag Vor der C.I.E. Sunlighting Conference in New Castle.
Koenigsberger O.H., T.G. Ingersol, Alan Mayhew and S.V. Szokolay 1984. "Manual of Tropical
Housing and Building, part one: climatic design", Orient Longman, New Delhi.
Lynes J. A. 1968. "Principles of Natural Lighting", Amsterdam Elservier, Amsterdam.
Plant C.G.H. and M.W. Archer 1973. "A Computer Model for Lighting Prediction", Building
Science, Vol. 8, pp 243-249.
Saxena B.K. and V. Narashimhan 1969. "Equations des composante d'eclairement dues a la voute
celeste pour une repartition de luminances correspondant a un ciel tropical", Lichtlechnik, Vol
19, pp 42-52.
Walsh J.N.T. 1961. "The Science of Daylight", Macdonald, London.
-------
3.8 Daylighting II
-------
Intentionally Blank Page
-------
2741
OPTIMIZATION OF WINDOW DESIGN FOR THERMAL, LIGHTING
AND OCCUPANT APPRAISAL CONSIDERATIONS
M. Boubekri*, L.L. Boyer**
* Center for Building Studies, 1455 de Maisonneuve, Concordia University
Montreal, Quebec, H3G 1M8
** Department of Architecture, Texas A&M University
College Station, TX 77843-3137
ABSTRACT
A sunlighting or a direct gain passive solar strategy is based primarily on the direct penetration of
sunlight into the living or working space. Although sunlight is often considered an amenity, it
can cause sometimes problems such as overheating, glare and adverse emotional response on the
part of building occupant. This paper examines these three issues by taking the case of the multi-
fold problem of sizing an office window fully exposed to sunlight penetration. Three outcomes
are considered, namely the solar heating, the visual problems resulting from the direct sunlight
penetration expressed in terms of glare, and the occupant appraisal of the space condition. These
performances are at times conflicting and compromises must be made.
INTRODUCTION
Sunlight admission to buildings is usually considered an amenity. When the purpose of the
design is to heat the building, and if sunlight is directly admitted to the living space, then the
design strategy is called a direct gain passive solar strategy. When sunlight is admitted to provide
additional natural light, then the strategy is called a sunlighting strategy. In any case, there are
positive as well as negative effects attached to this direct intrusion of sunlight. Some of the
problems are overheating, high glare, and visual noises caused by large floods of sunlight
beams penetrating the indoor space.
Failing to effectively take into consideration the multiple effects of sunlight early during the
design could lead to negative results manifested primarily by dissatisfaction of the building
occupants. This latter is a design parameter that can be very important for designers,because it
has a potential economical impact especially in work environments where occupant performance
is a salient and a critical concern.
Sunlighting studies have dealt with sunlight penetration in terms of duration of penetration. The
drawback of most of such studies is the fact that they tend to neglect the important fact that the
amount of sunlight penetration is an important visual stimulus. As such, different amounts create
different lighting conditions as well as engender different emotional or psychological responses
from the building occupants in addition to generating different solar heat loads.
There has been a lack of serious psychophysical studies of sunlight. It has been suggested in the
1965 sunlighting conference of the Commission Internationale de l'Eclairage (CIE) that this lack
of psychophysical and emotional studies of sunlight effects is attributed to the complexity of the
problem itself (Hopkinson, 1965). The emotional or psychological effects that sunlight createsare
in competition, or even in a direct conflict with the visual discomfort inherent to its presence,
hence making their evaluation rather complex and difficult. It had been proposed, but not yet
tested, that positive sensations created by the presence of sunligiit could,in fact,offset to a degree
the visual discomfort (Hopkinson, 1965). The discomfort glare level could be more tolerable
than the same glare level from artificial light or glaring sources in diffuse daylight conditions.
-------
2742
This study intendsto explore the trade-offs design solutions involved in the case of an office
module where the window is fully exposed to sunlight radiation. It focuses on the thermal
performance, discomfort glare, and the occupant subjective response characterized by the
feelings and emotions experienced.
RESEARCH APPROACH
It is posited in the conceptual framework of this investigation that the emotions of space users,
glare and the thermal performance are all affected by the same common variables, namely the
window size (Fig. 1). In the presence of sunlight, the amount of sunlight penetration, measured
in this case as the percentage of floor area that is sunlit, is a continuous design variable as well.
The model suggests that, in addition to the size of the window and the amount of sunlight
penetration, the observer's position in relation to the window affects not only glare but also the
affective response that people experience. Based on previous studies, it is further acknowledged,
though not examined in this study, that the emotional response may differ depending on the
individual's characteristics as well (Mehrabian & Russell, 1974).
Amount of
Sunlight
Penetratior
AT
Window
Size
SPACE OF OPTIMAL |V.
SOLUTIONS
-¦ '
Observer
individual T.
Characteristics
** IAFFECTIVH
ESPONSE
IllpltlP
Sideways
lilitif
$

DISCOMFORT
GLARE
Observer
Position
Facing
SOLAR
HEATING
FRACTION
Fig. 1.. Conceptual Framework Evaluating the Performance of a Window
Fully Exposed to Sunlight
From the perspective of the environmental designer, the thrust of the model is represented by the
direct and indirect links existing between the different design independent and dependent
variables involved. Figure (1) illustrates these relationships. The proposed model suggests that
the window size affects directly the thermal performance of a building. Also, the window size
combined with the amount of sunlight penetration both mediated by the observer's position in
relation to the window affect directly or indirectly the perceived discomfort glare and the
occupant affective response.
EXPERIMENTAL PROCEDURE
The effect of window size on glare discomfort, affective response and the solar performance
was examined in one room of the size of a typical private office (21 ft x 11 ft x 8.3 ft). The room
had one window facing southeast and situated in one of its longest sides. The size of the window
area was measured as the percentage of the window area to the wall area . Four window sizes
-------
2743
were tested (10, 20, 40, and 60 percent) (Fig. 2). The corresponding amounts of sunlight
penetration was 3.5,15, 30 and 45 percent of the floor area.

Fig. 2. Change of Window Size and Amount of Sunlight Penetration
Glare and occupant appraisals were performed in a real environment, simultaneously, therefore in
the same conditions. The sample of forty office workers who participated on a volunteer basis
was divided into two groups of 20 each and assigned to the sitting positions in the room as
indicated in Figure 2. Each group was divided into four groups of five subjects each, and each
subgroup was randomly assigned to a window size.
The perceived glare was assessed on a seven point scale ranging from 7 "intolerable" to 1
"satisfactory". The affective response was assessed using a seven point scale mood inventory
composed of thirteen words i.e. peaceful, calm, relaxed, excited, stimulated, exhilarated, in
control, significant, important, bored, dull, rushed, hectic). This mood inventory is based on
Russell's circumplex model of affect. The circumplex model of affect proposed by Russell
(1980) indicates the existence of the two basic dimensions of affect of arousal and pleasure.
Along with these two primary dimensions, other combined affective states exist; the main ones
are those described by the dimension of excitement and relaxation.
The solar performance was assessed through computer simulation using the ENERCALC hour
by hour simulation program (Degelman et al, 1989). The office module was assumed to have
only one exterior wall containing the window facing a south orientation. The other three walls,
the roof, and the floors wer* considered adiabatic. The solar building was assumed to be a heavy
structure (180 psf), super-insulated, double glazed with R-5 night insulation.
RESULTS
The computer simulations indicate that the heating performance improves with the increase of the
size of the south-facing window. The larger the window, the larger the solar heating fraction
and the lower the auxiliary heat required to heat the space. (Fig. 3).
From a frontal viewing position, the effect of the four window sizes tested on perceived glare
was moderately significant but not statistically significant from a lateral position. Therefore, from
a frontal position, the experimental results indicate that perceived glare is maximal for medium
window sizes and is less for either small or large window sizes . Regression analysis indicates a
quadratic relationship with a quasi bell-shape tetween perceived glare and window size (Fig. 3).
Detailed discussion of these findings are presented elsewhere (Boubekri, 1990; and Boubekri
and Boyer, 1991a). The presence of sunlight in the room seems to create rather positive effects
on the occupant, hence allowing the increase of glare tolerance. These glare assessments have
been based on an observer facing the window and sitting in a shaded part of the office. When
sunlight is striking the eye, glare level is extremely high and would be considered intolerable
regardless of the size of the window. Moreover, these glare levels are obtained under conditions
-------
2744
where glare is potentially the highest,because the direction of view is normal to the window.
When the observer is not facing the window, glare is expected to be considerably lower.
The fourteen variables describing the affective response caused by the change in sun penetration
and window size were factor analyzed. The data indicated the existence of two meaningful
underlying factors were derived from a the maximum likelihood factor analysis. The words
describing these two factors are in agreement with the two affective dimensions described by
Russell's circumplex model of affect (Fig. 3). The first common factor, namely relaxation, is
best described by five words, three among which are loaded positively ("peaceful", "calm",
"restful") and two negatively ("rushed" and "hectic"). The second common underlying factor,
excitement, is best described by the words "stimulated", "excited" and "exhilarated," all
positively loaded. Details of the factor analysis results are presented in various technical
publications (Boubekri, 1990; Boubekri et al, 1991b).
W
J
<
U
«2
z
o
<
X
<
J
U
as
a
z
<
w
tc
<
J
u
December 21, Noon Hour
Relaxauon
Auxiliary Heat
Reduction
Perceived Glare
Neutral State
50
40
30
20
10
<
U
X
X
a
<
z
o
H
U
3
Q
Ed
06
20 30 40 50 60 70 80 90 100
WINDOW SIZE, % WALL AREA
Fig. 3. Trade-offs between Solar Heating Fraction,
Glare Control and Feeling of Relaxation in Albuquerque
To examine the existence and the nature of the effect of window size and amount of sunlight
penetration on the two underlying common factors, analyses of variance and regression analyses
were conducted. When the observer is sitting sideways, window size does not impact
significantly glare or the occupants' feeling about the condition of the room. However, from this
position the different amounts of sunlight penetration cause different levels of emotional
response represented by the feeling of relaxation. Regression analyses indicate that 52 percent of
this relationship is explained by a quadratic model showing somewhat a bell-shape relationship
(Fig. 3). Relaxation is predicted to be maximal for sunlight penetration between 15 and 25
percent of the office floor area.
This study has shown that relaxation is not affected by the change of window size, but rather, by
the change in the amount of sunlight penetration. The highest feeling of relaxation in the office
-------
2745
module occurs when sunlight penetration (sunlit area) ranges between 15 to 25 percent of the
floor area. The feeling of relaxation remains above average (above 4 on the seven-point
relaxation scale) whenever the amount of sunlight penetration does not exceed 40 percent. This
relationship between relaxation and sunlight penetration has been derived based on an observer
sitting in a shaded area of the office and sideways to the tinted grey glass window. It is also
important to note that sunlight was only seen and not felt on the skin,and that thermal comfort
was achieved in the room. This relationship may be different if the conditions were otherwise.
Figure 3 illustrates the trade-offs, or the pluses and minuses between the solar heating
performance, glare control and occupant feeling of relaxation as they are impacted by the
window size on December 21 during the noon hour. As mentioned previously, the time of day
and day of the year are important parameters because they allow the designer to relate the amount
of sunlight penetration to the window size. Figure 3 indicates that window sizes equal to or
larger than 60 percent provide better solar heating performance and lower glare. These large
window sizes appropriate for solar heating and glare control also allow very large amounts of
sunlight penetration during most hours of the day which is detrimental to occupant emotional
comfort. Conversely, optimizing the window size for emotional comfort results in low glare
levels for the observer facing the window, but also only poor solar heating performance.
Optimal sunlight penetration for relaxation (below 4 on the relaxation scale) is only between 15
and 25 percent, and the maximum amount to create a positive feeling in the room is about 40
percent. These values of sunlight penetration are indicative a priori of small window sizes,
imposing, therefore, a severe limitation on direct gain solar strategies of which the purpose is to
maximize the solar heating fraction. This third dimension, namely the occupant feeling of
relaxation, is particularly important in spaces such as the office environment where the well-
being of the occupant is a critical design issue.
Though window size does not directly affect the degree of relaxation, the amount of sunlight
penetration is directly related to the size of the window. In addition to this latter, the
geographical latitude of the location of the building and the time of the day also dictate the
amount of sunlight penetration into the room. A sample calculation procedure of the optimal
window size for 30 percent sunlight penetration in the tested office module for 21 of December
in Albuquerque at 10:00 a.m. has been performed assuming a rectangular window and that
sunlight is striking the floor only.
The relationship between sunlight penetration and window size for different times of the day is
depicted in Figure 4. Sizing the window for a direct gain passive solar strategy will result in
very large amounts of sunlight penetration exceeding 40 percent, which is the maximum
recommended in order to have a positive feeling of relaxation. The reader should be reminded
that the ratio of the gross exterior wall area to the gross floor area in the tested office module was
about 0.75. The figure indicates that small window sizes, less than 30 percent of the wall area,
allow adequate amounts of sunlight penetration for positive feeling of relaxation during most
hours of the day, (from 10:00 AM to 2:00 PM during December 21 in Albuquerque). This range
of window sizes could be appropriate for climates where solar heating may not be needed.
-------
2746
<
W
06
<
06
O
o
~J
b
*
Z
o
I-*
<
06
H
W
z
W
eu
u
J
z
«3
Southern climates where
low SHF is all that is needed
1*— s
IS Northern climates where!!
Window size less than 30 %
for glare control and relaxation
/

II
Albuquerque
11 ft Room Width
3 ft Sill Height
10:00AM i
li:00AM
12:00 AM
:.v ¦
1 high SHF is needed
Window size greater than
60 % for glare control and SHF
Maximurt) recommended sun penetration:;
^ ~- ~VMffW-vp f* N
gj§|S—~ 8:00AM*
Ml 0 9:00AM W
100
WINDOW SIZE, % WALL AREA
Fig.4. Sunlight Penetration in Relation to Window Size for Albuquerque
CONCLUSION
In terms of trade-off design decisions, the optimal size of a south-facing in a direct gain strategy
window is recommended to be equal to or higher than 60 percent of the wall for a typical cold
climate. However, this window size is also limited by the fact that the maximum sunlit area
should not exceed 40 percent of the floor area in order to have a positive feeling of relaxation.
This is particularly important in a space such an office. It is assumed that the observer is located
in a shaded area of the room where thermal comfort is achieved as well. This study showed that
a window size equal or higher than 60 percent is recommended in a direct gain passive solar
heating strategy if the amount of sunlight penetration is less than than 40 percent of the floor
area.
REFERENCES
Boubekri, M. (1990). Ph.D. Dissertation, Texas A&M University, College Station, Texas.
Boubekri, M., and L.L. Boyer (1991a). J. Lighting Research & Technology (in review).
Boubekri, M., B.R. Hull, and L.L. Boyer (1991b). J. Environment & Behavior Cin review).
Degelman, L.O., Y. Kim, and B. Kim (1989). ENERCALC: version 89.07- Energy
calculations for buildings: a user's guide. Department of Architecture, Texas A&M
University, College Station, Texas.
Hopkinson, R.G. (1965). The psychophysics of sunlighting. Proc. CIE Bowcentrum
sunlighting Conference. University of New-Castle Upon-Tyne, pp 13-19, April 13-19.
Mehrabian, A. Russell, J.A. (1974). An Approach to Environmental Psychology. Cambridge,
Mass.: MIT Press.
Russell, J.A. (1980). J. of Personality and Social Psychology. 39(6): 1161-1178.
-------
2747
ESTIMATING HOURLY AND DAILY GLAZING
EFFECTIVE TRANSMITTANCE FOR EUROPEAN
LOCATIONS
Paul Waide
School of Mech. Eng.
Cranfield Institute of Technology
Cranfield, Bedford
MK41 8LT, U.K.
ABSTRACT
A semi-empirical methodology is presented for determining the
monthly mean hourly (or daily) glazing transmittance given a-priori
information of monthly mean solar radiation quantities. The method
is subsequently applied to solar radiation data from 2 9 European
locations to produce an extensive data base of effective
transmittance values expressed in the frequency domain.
KEYWORDS
Transmittance, Minutely Data, Instantaneous Diffuse Fraction,
Parametric Clear Sky Insolation Model, Typical Reference Year,
Fourier Transform.
INTRODUCTION
Solar energy system design calculations require the collector
glazing transmittance-absorptance product as an input parameter,
either as a monthly mean hourly or monthly mean daily value.
Existing techniques for determining this variable are applicable
only for daily averages (Klein, 1979), however, daily averaging over
hourly variations obscures the important diurnal pattern which
results from cyclical solar geometry.
If the angle of incidence of the solar flux is known, it is'
straightforward to solve the glazing transmittance using standard
expressions (Duffie and Beckman, 1980). However, in-situ glazing
systems experience solar flux from all forward angles, and as glazing
transmittance is highly sensitive to incidence angle, the angular
distribution of incident solar flux is critical to the overall
"effective" transmittance. The existence of expressions for
equivalent angles of incidence (Duffie and Beckman, 1980;
Brandemuehl and Beckman, 1980) makes it pragmatic to split the
incident solar flux into the three subdivisions of beam, diffuse and
ground reflected flux and to calculate the effective transmittance
of each component separately.
Instantaneous effective transmittance is linked to the instantaneous
distribution of the diffuse, beam and ground reflected components,
and their respective equivalent incidence angles, as equation [1].
This incorporates Hay's (1979) anisotropic diffuse sky model to
permit the lumping of the circumsolar component of the diffuse
radiation with the beam radiation and the treatment of the remainder
of the diffuse component as isotropic. Horizon brightening has less
bearingupon the effective incidence angle of diffuse radiation,
because typically it is greatest at times of highest air mass, when
-------
2748
multiple scattering is greatest and lateral isotropy is strong.
T
T.
GGT
-5 (1 -d) (l+kd)rb
+ — (l+cosB) (l-£(l-d)) (1-cosB)
T, 2 T, 2
[1]
where: - GGT - d
k(,l-d) rh
(1+cosB) (l-ic(l-d))
+ (i-d)r^ P(1-cosB)
£ 2
The normalised form of equation [1] is most appropriate, because the
angular dependence of the normalised glazing transmittance is
approximately the same for systems with the same number of glazings
and refractive index, i.e. it is insensitive to the extinction
coefficient of each layer. However, before it is possible to
integrate equation [1] over time, it is necessary to determine the
temporal variation of the horizontal diffuse fraction from which the
beam/diffuse/reflected fractions on a tilted collector are
calculable.
ESTIMATION OF THE DIFFUSE FRACTION
The primary parameters which effect the horizontal diffuse
irradiance are the climatological factors of cloudiness, turbidity,
multiple reflections between surface and sky, and the geometric
factor of path length or air mass. Of these, the cloudiness is most
important and can be considered to be largely random and sometimes
rapidly changing, while the turbidity undergoes relatively slow and
calculable changes, and the air mass is determinable from the solar
geometry with a smaller dependence upon atmospheric pressure. The
long-term monthly mean diffuse fractions represent the deterministic
behaviour of the diffuse fraction and are the best estimate of
future values, because the short term stochastic fluctuations are
approximately equally distributed around the mean. Ideally long-
term solar records would be used to evaluate the monthly mean
diffuse fraction for each hour, but these are seldom available or
accessible. An investigation of European Typical Reference Years,
TRY's, (Waide, 1991) revealed that within a given month of hourly
values there is significant inconsistent variation around each
hourly mean, implying insufficient values to reflect the underlying
long-term population. Additionally, where more than one reference
year existed for a specific location (Waide, 1990), there was marked
disagreement between monthly mean hourly values. This is to be
expected because the generation methods employed place greater
emphasis upon net monthly quantities and typical short term
sequential behaviour than the faithful representation of long-term
-------
2749
hourly means. Where comparisons between TRY's were possible, the
monthly average diffuse fractions and the monthly average clearness
indices, D, K, displayed general agreement, suggesting that these
quantities were representative of their long-term means and thus
could be used as a-priori information for the subsequent model.
A wealth of relationships exist for determination of the horizontal
diffuse fraction, given a-priori knowledge of the clearness index,
however, these mostly treat the air mass effect indirectly and are
based upon studies of integrated hourly values. Because it has been
found that diffuse fraction relationships based upon small
integrated time steps result in significant differences from larger
integrated values, the approach favoured here uses empirical
expressions to compute the monthly mean instantaneous clearness
index and diffuse fraction as developed from minutely solar
radiation data (Suehrcke and McCormick, 1988-89). These expressions
were formulated from studies of minutely data that explicitly
considered the underlying bimodal distribution of solar radiation,
and the influence of air mass. To assess their validity for a
European climate, they were compared to solar distributions
generated from a year's worth of minutely data (Waide, 1991) for
Garston, U.K., latitude 51.7°, (Littlefair, 1988). The Garston data
agreed well with Suehrcke and McCormick's expressions and any
differences were not outside the bounds of experimental error. This
result implies that the model is correctly representing the
underlying process and thus has universal application.
To be able to solve the monthly mean instantaneous clearness index
from these expressions, the mean annual beam radiation extinction
coefficient must be known. This is determinable from the parametric
insolation model of Bird and Hulstrom described by Louche et al
(1987), and empirically modified for a European climate as Louche
et al (1988),' see appendix.
RESULTS
The model was applied for 29 European locations using the TRY's for
each. In each lcation normalised transmittances for:- single and
double glazed systems with vertical, 45°, horizontal tilts; and
eight azimuths were modelled. Producing an extensive data base of
values. Typical graphical results, which illustrate the dependence
of the clearness index on air mass and the complex nature of the
long term monthly mean hourly transmittance, are shown in figure 1.
The non-linear effect of the sun rising and setting behind the
collector
during the summer months can be clearly seen. The data base was
compressed by using an FFT to convert the time series into the
frequency domain and discarding irrelevant coefficients. Circular
convolution theory permits multiplication in the frequency domain
by solar radiation coefficients to compute the energy transferred
per unit glazing area. A simple Fortran programme has been written
to pick out and inversely transform requested values and
distributions. Correlations for monthly mean daily values as a
function of the clearness index and month, are presented in the
thesis (Waide, 1991).
-------
RBFRACTIVB DCDBX - IJM
EXTINCTION COBPF. ° 0«]?T/MM
NUMBER OP LAYERS - S
LATEX THICKNESS - 4.09101
TILT OP GLAZING - 90.M«.
OIUBNTATION - SOUTH
BBAM EXTINCTION COBPP. - 0X9
NORMAL BBAM TRANS MITT ANCE - Q-Hi
NORMALISBD DIFFUSB TRANS. - 0AT4
MEAN NORMALISBD TRANS. = 04)09
STANDARD BRROR IN MEAN - 0.0096
GROUND REFLECTIVITY - 0J0
-------
2751
REFERENCES
Brandemuehl, M.J., and Beckman, W.A., "Transmission of Diffuse
Radiation Through CFC and Flat-Plate Collector Glazings", Solar
Energy, Vol. 24, pp511-513, 1980.
Duffie, J.A., and Beckman, W.A., "Solar Engineering of Solar
Processes", Pub. Wiley, New York, 1980.
Hay, J.E., "Calculation of Monthly Mean Solar Radiation for
Horizontal and Inclined Surfaces", Solar Energy, Voi.23, No.4,
pp301-307, 1979.
Klein, S.A., "Calculation of the Monthly-Average Transmittance-
Absorptance Product", Solar Energy, Vol. 23, pp. 547-551, 1979.
Littlefair, P.J., "Measurements of the Luminous Efficacy of
Daylight", Lighting Research and Technology, Vol.20, No.4, ppl77-
188, 1988.
Louche, A., Maurel, M., Simonnot, G., Peri, G., S Iqbal, M,f
"Determination of Angstrdm's Turbidity Coefficient from Direct Total
Solar Irradiance Measurements", Solar Energy, "Vol.38, No.2, pp89-96,
1987.
Louche, A., Simonnot, G., & Iqbal, M., "Experimental Verification
of some Clear Sky Insolation Models", Solar Energy, Vol. 41, No.3,
pp273-279, 1988.
Suehrcke, H., and McCormick, P.G., "The Diffuse Fraction of
Instantaneous Solar Radiation", Solar Energy, Vol. 40, No.5, pp423-
430, 1988.
Suehrcke, H., and McCormick, P.G., "The Distribution of Average
Instantaneous Solar Radiation Over the Day", Solar Energy, Vol. 42,
pp303-309, 1989.
Waide, P.A., "Applicability of Short Reference Years For Passive
Solar Building Simulation", Phase I+II, Energy Technology Support
Unit, Harwell Lab., Dept. of Energy, U.K., 1990.
Waide, P. A., Ph.D Thesis, Cranfield Institute of Technology,
Simplified Methods For Passive Solar Design, (in preparation), 1991.
-------
2752
APPENDIX: EXPERIMENTAL PROCEDURE
For each glazing system:- calculate Tj_
For each location:-
i) Determine The Annual Beam Radiation Extinction Coefficient,
Read TRY: either for each hour with sunshine duration > 56
minutes, or for each hour with 1± > 0.8 Ilo
where Ij.0 = 0 . 9751E0I.0TryT0TgT„T.
calculate hourly P. from:-
P. = (-l/m)ln[0.8832TryTgTB+81.779/(1367EovrJ]
weight each value by the concomitant hourly
direct normal insolation, and sum as running
total
At end of year, divide running total by the annual clear sky
direct normal radiation to produce the radiation weighted
annual value of p..
ii) Calculate Monthly Mean Hourly Diffuse Fraction
Read TRY: for each month determine K and D from the data.
For each minute of each hour of a representative
day of the month calculate from Suehrcke &
McCormick (1988):-
dc = 0.0336+0.0477m
kc = 0.877exp(-0.0933m)
% = H0//G0exp(-p„m)dt
k = K [% (1-D) exp (~P.m) +D]
d = l-(l-dc) (k/kc)4'4 for k < kc
= (kcdc+k-ka) /k for k > kc
Integrate d over each hour to get d, the radiation weighted
mean hourly diffuse fraction.
xii) Calculate Monthly Mean Hourly Transmittance
Determine effective angles of incidence for each radiation
component, and thence xb,td, Try.
Calculate rb and solve [1]
-------
2753
"LIMITS OF THE SKY"
Testing and Evaluation of the Current State-of-the Art in Minor-Box Sky Simulation
C.Cooksy, with Loveland J., Millet M , and Vanags A.
Department of Architecture, JO-20
University of Washington
Seattle, Washington 98195
ABSTRACT
It is necessary to press on towards the establishment of standards in order to face the problem of
perfection. -Le Corbusier One of the most significant outgrowths of the 1989 Passive Solar
Conference technical session on artificial skies for the simulation of daylight was the awareness of a
lack of design standards for the construction of mirror-box overcast sky simulators. Many of these
skies are being constructed across the United States. Most of these are based on the design
specifications of the University of Washington's mirror-box, constructed in 1976. The presentation
of the various attempts at calibrating these new skies to the CIE Standard was very disappointing.
None of the skies were attaining the CIE Standard Luminous Distribution. Based on these
troublesome results and the upcoming construction of a new room-sized mirror box at the University
of Washington, the authors set out to design a new mirror-box specification. This required the
testing of geometric and surface reflectance variables so as to isolate and eliminate the present
troublesome conditions. This paper documents the results of the testing, design and construction of
a new specification for a mirror-box artificial sky at the Department of Architecture, at the University
of Washington.
KEYWORDS
Daylighting, Energy Efficiency, Simulation, Modeling, Overcast Sky
INTRODUCTION
All machinery is derived from nature - Vitruvius Mirror box artificial skies have been described
as physical models of a mathematical model of the real sky1 This study focused on elements which
affect the physical model in its attempted incarnation of the mathematical one. Artificial skies are
used in conjunction with building models to simulate and predict the quantity and quality of daylight
in buildings under the real sky. They provide a means of examining and comparing the effects that
various design elements (alternate fenestration patterns, e.g.) can have on the amount, distribution
and character of daylight in a proposed building or space. "Die work presented here was conducted
to examine the effects that certain variables in artificial sky design can have on the amount,
distribution and end-use effect of daylight-simulating electric light in a mirror box type artificial sky.
Using information and experience gained in that research, a mirror box sky was then designed and
built for use by the Seattle professional and academic communities.
In 1989 the University of Washington's Architecture Department began work on the design and
construction of a mirror box artificial sky for Seattle City Light's Lighting Design Lab, a
lighting/energy education facility intended for use by the local academic and professional
communities. As the artificial sky project neared the construction phase, word came that several of
the new mirror box skies at Florida A&M2, the University of Oregon 3 and the University of
Minnesota4 - (see Fig. 1) had problems developing the desired luminous distribution pattern (LDP).
1 Kim, J.J. (1989). informal conversations
^ Crenshaw, R. (1989) informal conversations
-------
2754
SLCT PL AM
Florida A & M University
SE£3.
PLAkj
University of Washington
u
5E£
PL AM
University of Minnesota
FLAM
University of Oregon
Figure 1. Existing Sky Plans and Sections.
Low altitude luminance was much too low and low and/or anomalous blips in the LDP appeared in
mirror box vertical and horizontal comers. These grumblings moved us to postpone construction of
the Seattle City Light (SCL) sky until such problems could be further studied. Two characteristics
of existing artificial sky development and design also inspired the work that followed. First, all
mirror box skies in the literature had been designed and built without preliminary testing for LDP.
The established process was to calibrate the sky after construction, to compare the results of
calibration with the desired LDP and then to adjust it in various ways to attempt to bring the pattern
into line. Second, there was a great variety in box proportions and design details, many of which,
we suspected, could influence LDP. (see Fig. 1.) Given the absolute character of the LDP in the
mathematical model and the seemingly fixed geometry of elements required to reproduce it in a
physical model, it was surprising that so much variation could be tolerated.
RESEARCH EQUIPMENT AND METHOD
There is also another kind of machine, ingenious enough and easy to use with speed, but only
experts can work with it. - Vitruvius
The heart of this project was the testing of artificial sky physical elements to determine their impact
on luminance distribution patterns. Design of the project research equipment and method was
approached with several principles in mind. We should be able to isolate the variables investigated
so that cause and effect could be accurately related, and all aspects of the test process should be
reproducible. Certain physical limitations also influenced equipment and process of design. Much
of the adaptation of the test sky and all of the metering would be conducted by one person, so a
certain scale and simplicity of design was important. We also wanted to be able to test materials and
construction methods anticipating the SCL sky project. A test sky, other research tools and data-
gathering methods were designed to accommodate those ends (see Fig. 2).
3 Cartwright, V (1989) informal conversations
4 Ubbelohde, S. (1989) informal conversations
-------
2755
UUUUUMMWW
Y~T
\
,6LL841MM-Z-
T
— M-unf. fi ryy
,
RAN
Fifur* 6. Test Sky Plan «nd section

. *?rvcta FRA!£
ffmoN
Figure 2. Test Sky Plan and Section
SENSITIVITY TEST METHOD AND EQUIPMENT
The sensitivity tests involved a building model and illuminance sensors. The model was designed to
show a range of DF from both sky and sidelight sources. To maintain simplicity, a building with a
square plan and a single sidelight and skylight was designed. Assuming a 1/2" = 1' scale, the model
represented a building with a 10' ceiling height and a 20' depth and width, (see Fig. 3.)
The sidelight was large - 7' height and 10' wide with a sill height of 1 1/2'. A much smaller - 2 1/2'
square - skylight was cente in the ceiling. The metering was to be done with a Campbell
Datalogger and Licor photocell sensors. When calibration problems became apparent at the end of
the second round of tests, a Tektronics illuminance probe paired with the J-16 photometer was
substituted. The data-gathering station points established were intended to give a range of readings
with contrasting reliance on side- and skylight sources. Since we had five working Licor sensors,
four were to be used at the interior stations and one placed on the model roof to obtain an available
light reference value. The first three interior station were centered side-to-side in the model and
deployed along a line running firom the window wall to the back wall. Position one was 1/3 of the
distance from front to back, position two 2/3 that distance and position three against the back wall.
The fourth station was centered front to back against the side wall, (see Fig. 3.) We assumed that
the first and perhaps the fourth stations would be mostly sidelight-illuminated, the second mostly
skylit and the third illuminated about equally from the two sources. The model used in the A- series
of tests was built of white Crescent board with these interior reflectances: ceiling = 75%, walls =
60% and floor = 25%. These reflectance value choices were based on the assumption that a typical
building model, particularly in daylighting design studies, would have interior reflectances in this
range. Unfortunately, there was no very accurate way to develop an ideal standard (such as the CIE
Standard Overcast would be for the LDP tests) for daylight factors under these interior conditions.
Because the test data were seen as having an internally generated frame of reference- one sky
variation was to be compared to another - this was not an immediate concern. When the last round
of tests began (at the same time that the Licor/Datalogger equipment was abandoned), it was decided
that such a standard would indeed be useful. Since we could derive a CIE Standard Overcast based
10*
Figure 3. Building Model Plan and Section.
-------
2756
set of figures for comparison by using a BRE Daylight Pprotractor5 if we avoided the issue of
internal reflections, a black model was built. The black model was identical to the white, except that
it was constructed using black Crescent board for the interior surfaces. This essentially eliminated
interior reflections and allowed the use of the BRS protractor#10 (for unglazed apertures) to
establish sky-component-only daylight factors. The white model continued to be tested in some of
the final test sky variations for comparison with black model data.
THE RESEARCH PROJECT: TEST SKIES
The clouds are electric in this university -Adrienne Rich The artificial sky variations chosen for
testing had several sources. All, of course, were assumed to have some impact on luminance
distribution patterns. Some, such as the diffuser variations, emerged in the course of testing. Many
were derived from articles about mirror box skies at other institutions or from conversations with
individuals who had used or considered building such skies. The tilted wall, for example,
mentioned also in discussions with Fuller Moore of the University of Miami, Ohio6, was first
suggested by Barbara Erwin of Seattle City Light7. We wanted to investigate as many elements of
existing skies as we could to establish a context of existing work. A few of the variations we tested
were based on our own preliminary design for the SCL sky. The standard wainscot/floor cavity
height (1/3 of the total wall height) was one such. Other test variation ideas, such as the octagonal
and white-wall boxes, arose in team discussions about the basic limitations of mirror box design.
Conceptually, the set of variations was broken down into groups defined by common elements of
variation. One group involved variations in plan/geometry: square, rectangle, octagon. A second
consisted of sectional variations defined by differences in floor height and/or horizon height and wall
angles. Surface transmittance and reflectance variables were common to a third group: floor and
wainscot color, wall surface material reflectance and diffuser transmittance. A preliminary set of
tests was conducted to examine metering and lamp cavity characteristics.
Life with us is very dull indeed. How can it be otherwise, when all one's prospects are nothing but
a single line, with no varieties except degree of brightness or obscurity? - Edwin Abbot
SFCTION/EI FVATION
FLOOR HEIGHT WAINSCOT WALL ANGLE LAIvP OVERRUN
PI AN
SQUARE OCTAGON RECTANGLE
~ oo
1IGHT TRANSMISSION/RFFl ECTION
FLOOR WAINSCOT WALLS DIFFUSER
Figure 4. Test Sky Variations
^ Longmore, J. "BRE Daylight Protractor".Department of the Environment, Building Research Establishment, Her
Majesty's Stationery Office, London, England, 1967
6 Moore, F. (1989) informal conversations
7 Erwine, B. (1989) informal conversations
-------
2757
TEST DATA AND IMPLICATIONS: LDP
Information in the form of sky luminance values and building model illuminance levels was gathered
for each test sky version. Data processing and analysis occurred in several stages. The first step
was to translate the skies' footlambert values into relative-to-zenith values for LDP evaluation and
the building model footcandle readings into daylight factors. In both cases this was a simple matter.
Since the LDP values represented the percentage of zenith :(unity) luminance seen at a given angle,
they were the quotients resulting from brightness (measured in footlamberts) at measurement angle
divided by zenith brightness (Btheta/Bz). Similarly the daylight factor values were obtained by
dividing illuminance values at interior reference stations by the available footcandles at the exterior
station. Once the LDP relative values were calculated, comparisons between sky variations could be
made. The values for each sky were studied relative to each other to evaluate the effects of given
variables on LDP. They were also studied relative to the CEE Standard Overcast values to determine
which elements or combination of elements seemed most likely to produce a sky design with the
desirable LDP.
TEST VARIABLE EFFECTS
From parametric studies of the test skies, the following conclusions were drawn about the effects of
focus variables on sky LDP:
Wainscotting/floor cavity: As noted above, the floor cavity seemed to be a light sink. The wainscot
nullified the effects of floor color variation - it mattered little whether the floor was white or grey
once it sank below the horizon level. A great shift in wainscot reflectance (from grey to white)
resulted in a doubling of the horizon luminance. A white wainscot had about the same effect as a
low-reflectance grey floor on footlambert levels at that altitude. A mirror wainscotting yields
essentially the same results as our grey wainscot. (This was not tested with alternate floor colors,
however. It may be supposed that a white floor/mirror wainscot combination would show higher
luminance than a white floor/grey wainscot sky at the horizon level.) A white wainscot above the
horizon line has a significant impact on horizon luminance. There is a point, as mirror
interreflections accumulate at low altitudes, at which the apparent brightness of an opaque white
surface is greater than that of a mirror. That point seemed, indicated by measurements in A-series
skies, to be at about 15 degrees altitude in both basic box and standard wainscot skies. A 15 degree
cutoff line for the wainscot was implied. This above-horizon wainscot does abandon all hope of an
infinite horizon but that, as noted above, is a fairly academic pursuit in any case.
Wall surface: Our tests showed opaque white walls to be a poor alternative to mirrors. There is very
little dropoff in luminance values at lower altitudes, and very little increase at higher altitudes until
the ceiling comes into probe view. At that point luminance levels nearly double. A more highly
reflective, smoother surface may make this option seem brighter. Its one great advantage is that
there is litde variation in the luminance at any given altitude between one azimuth angle and another.
It seems to suffer little from the double reflection corner or the crossover effect. In this respect it is
much superior to the mirrored wall skies.
Floor color: In grey floor/white floor comparisons it is seen that floor color has significant impact
only when the floor plane and horizon line are coincidental. In those cases, the floor color shift
effect is substantial. Whether the resulting desirable (by CIE standards) rise in horizon luminance is
actually desirable when achieved by this method is questionable. Reed and Nowak noted in their
Texas A&M study8 that the ground reflectance factor can introduce a misleading element into
building model studies, their implication is that acceptable horizon luminance values should not be
the result of untypically highly reflective ground surfaces.
Lamp overrun: The test sky conditions for study of this variable were unfortunately limited. Very
few relevant angles of measurement were available. The small body of applicable data gathered does
suggest that lamp overrun designs begin to solve the dark ceiling/wall corner problem Any move in
8 Reed, R.H. and M.A. Nowak, "Accuracy of Daylight Predictions by Means of Models Under an Artifical Sky,"
Illumination Engineering, vol. 50 (July, 1955), pp.336-346
-------
2758
that direction is a good one - this step, slim as the supporting evidence is (at least no evidence against
it emerged), is recommended.
Section variations: The A-Series examples begin to show what happens when the horizon is raised
and the luminous character of the sky becomes increasingly ceiling-dominated. In our tests the 2:3
wall height/width proportion works well in approximating the CIE standard LDP. At a 1:2 wall
proportion the ceiling's role begins to become too great The 60 - 90 degree altitude readings all
show about 100% relative to unity. The 15-, 30- and 45- degree altitude luminance levels are quite
good on average. The horizon luminance is still low. The critical element here is the altitude angle at
which the probe sees the ceiling, for at that angle relative luminance values will tend to be in the high
90's. The shorter the wall relative to width, the lower that angle will be. The tilted wall variation
shows some promise. Despite the very small angle of tilt it did indeed raise horizon luminance.
Altitude 15 was also brighter. Otherwise there is little difference between this and the basic box
LDP. It has been hoped that the tilt wall would help to diminish the crossover effect and other
wall/ceiling difficulties. That did not occur in the version we tested. A greater degree of tilt may
lead to a more successful sky in those regards.
Diffuser transmittance: The more opaque diffuser resulted in very slightly lower overall LDP values
relative to the zenith. The important distinction between diffusers is, as suspected, not in their effect
on LDP but in the respective level of available light they provide for illuminance photometry.
Plan variations: We had high hopes for the octagonal sky. Unfortunately the mirror used for the
insert panels was slightly bronzed and did not match the reflectance of the original four side panels.
The test was a failure. Shortage of time and funds precluded correction. This sky probably could
work better than a square plan sky in terms of CIE LDP but for our purposes it, like the tilt wall sky,
would entail construction and operational disadvantages outweighing its advantageous potential for
meeting the CIE standards. The rectangular plan skies perform much like the square. Their LDP is
less symmetrical because the boxes are bilaterally rather than quadrilaterally symmetrical. The 0 and
90 degree azimuth patterns are no longer identical. At 45 degrees altitude in test sky R-9 the probe
sees the ceiling at 0 degree azimuth (the short direction). This pattern becomes more emphatic
(compared the UW plan sky, R-6, with the UO plan sky, R-8) as the difference in wall length
increases. That tendency will give the same effect as having two floor heights (see sectional
variations above), one for 0 degree azimuth readings and one for 90. The general LDP is much the
same as in the square skies; however, with a crossover effect and inconsistent luminance at various
azimuths with the same altitude. Same road, more bumps.
Meters: The agreement between the 1- and 8- degree luminance probes was generally good. The 1-
degree probe has a disconcerting habit of registering clearly any irregularity in mirror surface. By
the time the A-series of tests was begun it was apparent that the more forgiving 8-degree probe was
providing data of sufficient purity. Neither the fineness of test sky craftsmanship nor the sky's
potential for CIE perfection warranted the sensitivity of the one degree probe.
TEST DATA AND IMPLICATIONS: DAYLIGHT FACTOR
But soft, what light through yonder window breaks? - Wm. Shakespeare
The building model sensitivity tests were intended to investigate the impact of sky designs variables
and the resulting changes in sky LDP on DF levels in a generic building model. Information on
magnitude and location of DF value shift and the design-based causes for these shifts was sought.
TEST VARIABLE EFFECTS
Wainscotting/floor cavity: the wainscot reflectance seemed to make the difference in these tests. The
model in no-wainscot (horizon/groundplane coincident) sky shows DF levels between higher (white
wainscot) and lower (grey wianscot) levels found in the floor cavity versions. The shift in levels is
fairly constant across all four reference stations, suggesting that the low altitude LDP shifts are
accountable. A mirror wainscot produces conflicting results. The distinctions are quite narrow, but
the white model tests show higher DF values for mirror than for white wainscot skies,while the
black model shows lower DF values in the mirror wainscot sky. One would expect that wainscot
-------
2759
color wuld have relatively small effect on DF, since it had so little on LDP. Perhaps this is what we
are being told by the conflicting data. A white wainscot above the horizon (the UO waisncot),
unsurprisingly, raises white model sidelight DF values and leaves skylight-based values unchanged.
In the black model tests no significant change is seen. There is a very small overall drop in DF
values with the UO wainscot. This is unexpected: we had thought that the 2,3 and 4 stations in
particular would show at least a siigjitrise in sidelight-based DF.
Floor color: This variable is discussed in some detail above under the A-2/A-7 comparison. The
conclusions about the importance of internal and groundplane reflections are supported by a test of
the black model in the R-8 and R-12 skies (whose difference was floor color). "Hie black model
shows little DF value change in that comparison while the white model, subject to internal reflection,
shows the substantial change (described above in the A-2/A-7 comparison) when floor color shifts
from grey to whtie.
White walls: A higher DF is seen in sky only readings. The sidelight-based DF is higher at all but
the #1 positions. The same phenomenon shown in the discussion above of proximity of reference
station to window wall is at work here.
Sectional variations: Generally, a higher floor yields higher DF values. This holds true until the
wall height/width proportion passes the 1:2 point. The closer the light sources, the greater is the
illuminance. The relationship between available light on the unobstructed horizontal plane and the
light on the interior work surface :(i.e., DF) appears to shift at some point in favor (if higher is
better) of the more distant source. The tilt wall variation, typical of skies that show high horizon
LDP values, again increases DF values in the sidelight readings.
Plan variations: The octagonal sky was not tested. Rectangular sky tests show no significant
variance in DF from corresponding square plan skies. There is a slight shift in DF in rectangular
skies depending on the orientation of the model (i.e., whether the sidelight faces a long or a short
wall). In the rectangular skies tested in this project, with floors off square by no more than 3:4, this
shift is both quite small and inconsistent in its appearance, and probably inconsequential.
THE SCL SKY: DESIGN AND CONSTRUCTION
You boil it in sawdust: you salt it in glue:
You condense it with locusts and tape:
Still keeping one principal object in mind:
To preserve its symmetrical shape.
Lewis Carroll
Having explored a range of design options and impacts, the research team, awash in data and no
longer relying on precedent alone, turned again to the design of the SCL sky. There were three main
influences on the SCL sky design. In no particular order of importance these were the test sky data
analysis, the SCL Lighting Design Lab program restrictions and requirements and the materials'
limitations and specifications. We had hoped that the test skies would direct us to some ideal mirror
box design specifications. These hopes faded quickly as test results were seen. Test sky data
analysis led to no definitive criteria for design. It did help to narrow the field of choices, however.
Some potential design elements were eliminated from consideration due to poor luminance
distribution patterns or values shown in test runs. The white wall, the very high floor and the base
model variations were discarded in this purge. Although the rectangular plan skies had in some
cases performed well in terms of LDIJ there seemed to be no advantage in straying from the more
nearly radial symmetry of the square plan. Several schemes, such as the tilt wall and octagonal
skies, were promising in terms of LDP but presented too many practical problems. Deemed worth
incorporating into the SCL sky were the lamp overrun, the high-transmittance diffuser and the
square plan. The daylight factor studies, even the black model tests, were not so directly useful in
indicating good design variables. Sometimes LDP and DF results conflicted - what seemed better for
LDP seemed worse for black model DF. The wainscot sky compared with the basic box
demonstrates this syndrome.
-------
2760
THESCLSKY: CALIBRATION
Calibration of the Seattle City Light sky was performed with basically the same equipment and
methodology as had been used in the test sky calibrations. The rotating circle/periscope mount panel
was installed in a freestanding table, a variation on the elevated periscope mount used for calibrating
the wainscot version test skies. The top surface of this table was at the same height as the SCL
chamber wainscot upper edge. As in the test sky calibration, the vertical periscope tube was centered
on the sky floor and azimuth and altitude intervals of 15 degrees were used. For this series,
however, azimuth readings were taken every 15 degrees around a full 360 degree circuit A series of
three calibration runs were made once the photometer probe mounthead been aligned and centered.
Several additions may work to improve the performance of the SCL sky. The addition of a U.O. -
type of wainscotting between the horizon and the 15 degree arc would probably produce both better
LDP values at those altitudes and higher- more accurate- DF values in building model tests.
Installation of fixtures and lamps just below and perpendicular to the existing fixture ends should
help to eliminate vestiges of the dark band seen so clearly in the pre-diffuser tests. A white ceiling
panel directly above the fixtures is also to be considered to produce a more consistent luminous
quality at all azimuth angles at high altitudes. Whether advantages made thereby would be worth the
sacrifice of easy heat evacuation through the open ceiling is a moot point Luminance distribution
inconsistencies addressed by both of these last two alterations seem to be largely overcome by the
diffuser alone. The bumpiness of the LDP in the 15 - 45 degree range are certainly due in part to
slight warps and out-of-square construction. The inherent flaws in using a cubic form to achieve a
radial LDP probably far outweigh the impact of these factors on failure to produce the CIE ideal.
CONCLUDING REMARKS
Much of the work of this project was done with future sky-builders in mind. The following
conclusions and advice are addressed mostly to them: It is easy to miss the CIE Standard Overcast
LDP in mirror box skies. Given mirror box geometry.it is, in fact, nearly impossible to generate
such an LDP. The building model studies seem to indicate that a condition approximating the ideal
can be good enough for DF prediction,and the test sky studies indicate that a sky with such an
approximately ideal LDP is possible. In general, the design that emerges has these qualities: a wall
h:w ratio of somewhere between 2:3 and 1:2, a square plan, a long lamp overrun (longer than the
SCL sky managed), a high-transmittance diffuser and a high-reflectance wainscot (if any). A few
specific recommendations and notes: Light leaks of the sort that occur in comers and around door
openings are inconsequential. Maintaining squareness in construction is important, as the tilt wall
sky showed - a little error in trueness of angle (and by extension, trueness of plane) goes a long way
in its effect on LDP. Although not well served by this project, the tilt wall section, the octagonal
plan and the opaque wall schemes have the potential to be successful mirror box skies.
It is typical in a report such as this to conclude with some proposals for further study. More
carefully controlled and exhaustive research is easily imagined and could certainly be done. A
rectangular mirror box artificial sky with a perfect CIE LDP is plainly impossible, however, or so
nearly impossible as to be unattainable by reasonable people. Left with the option of a good-enough
sky, many things are suddenly good enough. For designers aiming for CIE perfection there is, with
this and a wealth of recent and ongoing studies elsewhere, a sufficiently extensive body of
information that none need work in the dark nor suffer not-good-enough results. Suggestions now
available from this variety of sources all describe, unsurprisingly, similar design elements for
inclusion and similar pitfalls to avoid in the creation of mirror box skies. As a teaching tool whose
purpose is to raise student awareness of the relationships between building design elements and
natural lighting, an artificial sky needs not meet any particular standards at all to claim success. In
short, there probably has been enough done and said and measured already in and about the
particulars of these particular skies.
ACKNOWLEDGEMENTS
This project was made possible through the financial support of the Washington State Energy Office:
Kim Drury, Tony Ussibelli and Sandra Tarzan, and the cooperative efforts of Seattle City Light: Ed
Holt, Steve Poole and Diana Campbell.
-------
2761
AN AUTOMATED HELIODON FOR DAYLIGHTING BUILDING DESIGN
R. Rhyner, C. Roecker and J.-L. Scartezzini
Laboratoire d'Energie Solaire et de Physique du Batiment
Ecole Polytechnique Fed£rale de Lausanne
CH-1015 Lausanne, Switzerland
ABSTRACT
Daylighting building design offers real opportunities to achieve a part of the
electricity savings potential of Switzerland. Design tools, however, are necessary to
reach this goal in the practice.
A sun simulator was built to allow the study of the direct sunlight contribution to
the lighting of buildings. It is made principally of an automated heliodon and a
calibrated light source, installed in a carefully designed black chamber. In this way,
visualizations of daylighted architectural spaces, as well as monitoring of direct
daylight factors can be achieved in optimal experimental conditions.
The recognition of this equipment by lighting designers, ergonomists and architects
confirms the important-" of the latter one. It is a very good complement of the new
infographic lighting design programmes developed recently for the same purpose.
KEYWORDS
Daylighting, design tool, automated heliodon, artificial sun, direct sunlight, direct
daylight factor.
-------
2762
INTRODUCTION
Artificial lighting of buildings is responsible for a significant part of the national
electricity consumption in Switzerland. Daylighting building design, associated to
efficient light sources and control offers, on the other hand, real opportunities to
achieve significant electricity savings at the national level (Brunner and
collaborators, 1986).
Daylighting design, however, has a strong influence on the architectural aspect of
buildings, as well as on the users'visual comfort (Lam, 1986; Moore, 1989). Architects
and lighting designers are the first concerned by this problem. They feel the need for
user-friendly design tools, which could contribute to spread this new building and
lighting technology in the practice.
A set of experimental tools, principally made of an automated sun simulator
(heliodon) and an artificial sky, is currently under construction at EPFL. Both
equipments are expected to complete computer infographic daylighting design
programmes developed for the same purpose. This paper gives an overview of the
automated sun simulator, which is the first part of this equipment and already in
operation.
DESCRIPTION OF THE EQUIPMENT
The aim of the sun simulator is the study of the direct sunlight contribution to
lighting of buildings. This part is usually far bigger than the diffuse sky vault
contribution; it is, on the other hand, responsible for the most frequent discomfort
situations.
The sun simulator is expected to allow the following operations :
- measurements of direct daylight factors within scaled models of buildings,
- evaluation of visual comfort indicators for particular sun positions and
daylighting systems,
- visualization of the qualitative aspects of the interior of buildings illuminated
by daylight.
Figure 1 gives an overview of the different components of the automated sun
simulator. Each of them will be described briefly.
THE LIGHT SOURCE
The light source, as well as the optical system, were chosen in order to reproduce the
main physical features of the sunlight (spectral distribution, homogeneous
illuminance, parallel light beams).
A short arc discharge lamp (OSRAM 2.5 kW HMI), combining high luminous
efficacy (0 = 240'000 Lm, T| = 96 Lm/W), daylight close spectrum (Tc = 5600 K) and
high color rendering index (Ra > 90), was used for that purpose.
-------
2763
light source
optical element
black curtains
£
-------
2764
obtained with the projector alone reaches up to 10'200 Lx with a uniformity of only
33%), showing a significative improvement compared to previous similar tentative
of optimization of such an equipment (Fontoynont, 1987).
2750 Lx 2585 Lx
Fig. 2 Optimal illuminance distribution measured on the work surface of the
equipment (disk of diameter 1.5 m).
THE AUTOMATED HELIODON
The heliodon is a mechanical device, which is able to reproduce the relative motion
of the sun (simulated by the light source) for an observer placed on the moving
work table (defined by a 1.3 m x 1.3 m black aluminium cross). Six different designs
were considered before to start its construction. The final geometry, shown on figure
3, was chosen accounting for the following main criterion :
- absence of potential self-projecting shadows on the model,
- simplicity of the mechanical construction,
- movement continuity,
- minimal movements of the model center,
- maximal distance between the light source and the work plane,
- weight limitation.
The heliodon has two movable and motorized axis (DC powered motors). An ad[hoc
gear technology (harmonic drive) was used to obtain a soft and gentle movement, as
well as a maximal torque. It has the following main mechanical characteristics :
- accuracy of positioning : <1°
- movement speed range : 0 to 12 rpm
- maximal model size : 1.0 x 1.0 x 0.6 m
- maximal model weight : 25 kg.
-------
2765
Fig. 3 Schematic view of the automated heliodon (the moving work table is made
of a simple cross reducing the parasitic reflections).
THE CONTROL AND MONITORING UNIT
The two motorized axes of the heliodon are driven by a PC/AT 386 compatible.
Different types of movement, accounting for the site latitude and longitude, the time
of the year, the building orientation, are possible. These movements can be
continuous, going from point to point, or reaching a defined position corresponding
to selected sites, months, days and hours. The speed of the different movements can
be chosen.
A serious effort has been dedicated to the way the users can interface with the
machine. Accounting for the fact that architects, as well as lighting designers, are
expected to be the principal ones, graphic inputs involving self explanatory symbols
(icons), pop-up menus and mouse interactions have been employed at purpose.
To complete this equipment, different visualization and data acquisition facilities
have been added to the control software. They include :
- an endoscope with corrected lens for architectural applications (real
perspective),
- a high resolution video camera,
- a series of miniaturised photometers (sensor diameter of 4 mm).
The hardware is connected to the PC/AT 386 compatible allowing automatic data
sampling of daylight factors, as well as output data plotting and treatment.
THE BLACK CHAMBER
All the experimental equipment, which has been described so far, was installed in a
black chamber, built up for that purpose (see fig. 1). The black chamber was designed
to offer the best experimental conditions during the operation of the sun simulator.
It is made of black curtains, defining a space of 8 x 5 x 5 meters, which minimizes the
-------
2766
incoming of parasitic external light (vertical illuminance smaller than 0.25 Lux even
during a clear day). Internal reflections, due to the lighting of the curtains by the
light source, were lowered by the choice of a black and matte tissue (reflection
coefficient : 2.7%). The floor has been covered by a carpet of similar photometric
features (reflection coefficient : 1.4%), reducing the risk of perturbating the
measurements by unwanted light reflections on the floor. For the same reason, the
heliodon has been painted with a dark black paint.
The careful design of the overall equipment has led to a very low expected
experimental relative error (lower than 15%). The latter one involves all the
different sources of error, which justifies all the procedures used to minimize their
overall contributions.
CONCLUSION
A sun simulator, made of an automated heliodon and a calibrated light source, was
built as an experimental daylighting design tool, principally aimed for architects. It is
expected to allow an optimal use of direct sunlight within buildings in order to
achieve lighting electricity savings and improve visual comfort.
The careful design of this equipment has allowed to minimize the different sources
of experimental errors. This had an influence on the design of the automated
heliodon itself, but also on the choice of the light source and its optical system, as
well as on the construction of the black chamber hosting all the equipment.
The recognition of the usefulness of this tool by the architects, the lighting designers
and the specialists of visual comfort, confirms the importance of the latter one in
regards to daylighting design within buildings. It is expected in this way to be an
optimal complement of the new infographic lighting design tools (Ward and
collaborators, 1988) developed recently for the same purpose.
REFERENCES
Brunner C.U., Baumgartner A., Miiller E.A., Stulz R. and B. Wick, (1986). "Electricity
savings". National Research Programme "Energy", Swiss National Research
Foundation (FNRS), Zurich, Switzerland.
Moore F., "Concepts and practice of architectural daylighting". (1989), Van Nostrand
Reimhold, New York, USA.
Lam W., "Sunlighting as formgiver for architecture". (1986). Van Nostrand
Reinhold, New York, USA.
Fontoynont M., (1987). "Prise en compte du ravonnement solaire dans l'eclairage
naturel de locaux : m£thodes et perspectives", Thfese de doctorat, Ecole des
Mines, Paris, France.
Ward G.J., Rubinstein F.M. and Clear R.D. (1988). "A ray tracing solution for diffuse
intereflection", Computer Graphics. 22 (4).
-------
2767
A NEW TYPE OF CONTROL SYSTEM FOR DAYLIGHTING
W. Glennie and V. Krishnamurthi
Lighting Research Center
Rensselaer Polytechnic Institute
Troy, New York 12180
ABSTRACT
A new lighting control system has been expanded to adjust the angle of Venetian
blinds as well as the output of electric lights in response to outdoor lighting
conditions. The prototype control system (ImCon) consists of a small charge-
coupled device (CCD) camera, computer, stepper motors and associated lighting
controls. By using an image of the room, lighting conditions can be matched
automatically to a variety of activities, with a degree of precision that cannot
be achieved using existing manual or automatic controls.
KEYWORDS
Lighting control; computer control; camera; image-based control; daylight;
shade control; window blinds.
INTRODUCTION
In addition to reducing the use of electricity in commercial buildings,
effective daylighting can bring delight and pleasure to a room's occupants.
However, the glare and overheating of uncontrolled daylight can make it
difficult or impossible to work efficiently or even to enjoy being in that same
room. While it is conceivable for an individual to achieve ideal illumination
at all times by constantly manipulating the levels of various lights and
positions of window shades or blinds, this is at best a considerable imposition
on his or her time. Without some encouragement, most people will not turn out
lights when they leave a room (Rea, Dillon and Levy, 1987), so it is unrealistic
to expect them to fine-tune the configuration of Venetian blinds every few
minutes as exterior conditions change. They will commonly set the angle to
prevent any entry of direct sunlight and will leave on all of the lights, which
is usually a major waste of energy. Once blinds are shut, they are seldom re-
opened (Rea, 1984).
In addition to wasting electricity, the lack of an effective suitable control
system can interfere with an individual's performance. Either the light levels
are excessive or inadequate or light is poorly balanced throughout the space.
Even with the best commercial lighting control systems, light levels are driven
to a single target value whenever an area is occupied. Ideally, the light
levels ought to change s activities within a room change. An ability for a
lighting control system to respond to varying activities within a space can best
conserve the most important resource, human effort.
EXISTING CONTROLS
The simplest strategy employed by current automatic window blind controls
without a photosensor is to set the angle of the blades so that there is no
possibility of direct sunlight entering the room. The setting may change during
the day and the daily pattern may change through the year, but the intent is to
err on the side of caution by eliminating the possibility of glare. In many
parts of the world, fewer than half of the daylight hours have clear skies, so
this type of control is unnecessarily conservative. For example, in the test
room, the daylight factor (fraction of total unobstructed exterior illumination
that reaches the interior) is 5% at a distance of 5 m (16.5 ft) from the
windows, if no blinds are in place. With the minimum possible shade obstruction
(40% reduction in effective glass area), this value falls to 3%. However, if
the blinds totally block the direct view of the sky, the daylight factor is
reduced to 1% or less, depending on the color of the blinds. These values are
based on the well-known formula from Hopkinson and Kay (1966):
DF = 10WH2/D(D'+H2) + 4GR/F(1-R) (eq. 1)
where DF is the Daylight Factor (%), W is the width of the window (3m), H is
the height of the head of the window above the reference plane (2.5 m), D is the
perpendicular distance from the reference point to the window wall (5m), G is
the actual area of glass (including any below the reference plane) (7.5 m2), F
is the floor area (31.4 m ) and R is the reflectance of the walls (0.7). When
-------
2768
the shades are in place, W and G are reduced to account for the reduced
effective glass area. Thus it can be seen that closing the blinds more than
necessary can eliminate two-thirds of the daylighting potential (DF of 1% rather
than 3%), when there is no direct sunlight to be blocked.
There are some more sophisticated systems which use photocells, such as the
Galaxy Sun Controller'*" from Levolor Lorentzen, Inc. When the light level in a
space falls below a specified minimum value, the blinds will open one step,
unless they are already fully open. Similarly, if a given maximum illumination
level is exceeded, the blind will close a bit. While this is superior to the
siipple "time of day" control, it is susceptible to the same problems that plague
all photocell-based lighting controls: difficult adjustment and response only to
the average light conditions within a room. In contrast, ImCon can recognize
and respond to the presence of even a small area of direct sunlight, which might
not be sufficient to trigger the photocell's response. As an additional benefit
of using an image, the system has the potential to change its operation on the
basis of where individuals are located within the room. Glare is "directional"
and the pleasure of sun streaming through the windows, when it does not detract
from the task underway, is one of the factors that makes a room comfortable and
pleasurable.
The failings of traditional photocell controls for electric lights have been
well documented. Verderber, Morse and Jewell (1989) provide the measured power
consumption and levels of daylight illumination measured at ten banks of
fluorescent fixtures in a large office building at one unspecified instant.
They also provide a curve of light output vs. power consumption for the electric
lights. Their measurements show a total reduction of only 28% from full load,
while the expected reduction, considering the daylight levels at the time, is
64%. Indeed, the situation is even worse than these figures indicate, because
one of the ten rows of lights is dimmed too much, providing illumination of only
290 lx (27 ft-c), rather than the design level of 377 lx (35 ft-c). Admittedly
the measurements were made at an "early stage of the building's operation," but
these problems are typical of photocell-based control.
In a detailed study of photocell configurations and control algorithms,
Rubinstein, Ward and Verderber (1989) document several cases where particular
illuminance levels are as much as 60% below the design level. Some of the
specific cases that they study are much more successful at maintaining the
target luminance at two specific locations. However, no matter what control
algorithms are used or what modifications are made to the photocell itself,
these systems can never be satisfactory, because, unlike a human eye or camera,
"sensors are 'blind' to glare and excessive luminance ratios that can cause
discomfort or low visual performance for occupants" (Jaekel and Rea, 1983). It
would take a large array of photocells to egual the degree of control that is
possible with ImCon's approach. Even so, this would not provide all of the
advantages of working with an image of the room.
IMAGING CONTROL SYSTEM (ImCon)
During the past two years, the authors, along with Mark Rea and Inderpreet
Thukral (also at Rensselaer Polytechnic Institute's Lighting Research Center),
have developed a new lighting control system that incorporates all of the
beneficial properties of existing lighting controls and has many substantial
advantages over any combination of current techniques. ImCon uses a small
charge-coupled device (CCD) camera connected to a personal computer to set
independent levels of several fixtures in a room. A complete description of its
components and fundamental* aspects of operation are given elsewhere (Glennie,
Thukral and Rea, 1991).
Like a system using a photocell, ImCon can respond to light levels in a room;
unlike a photocell, which integrates flux density over a wide area, the imaging
system responds to the actual brightness distribution, complementary to the
response of a human observer. Like a standard occupancy sensor, ImCon detects
the presence of people in a room; unlike infrared or ultrasonic devices, the
image shows the number and location of the occupants. This information allows
the system to respond intelligently to current activities and adjust the lights
and blinds accordingly. Like a multi-channel, pre-set light controller, ImCon
can store and recall several different light "scenes" (such as conference, desk
work, slide show, computer work); unlike the manual system, no operator
intervention is required to switch from one configuration to another. Changes
are made automatically on the logical basis of some perceived change in
activity. In addition, the imaging control is based on actual brightness
levels, not simply "percent of maximum" settings for individual sources.
Figure 1 shows the room in which the ImCon system is being developed. It was
chosen, in part, by the presence of two large, south-facing windows, which
provide an excellent opportunity to test the response of the system to direct
sunlight. A typical image from the camera, which is located high in one corner
-------
2769
View of Test, Room.
-------
2770
of the room, is shown in Fig. 2. Although the present system is a prototype, it
does demonstrate that the computer-based ImCon system can;
(1) provide a consistent and appropriate amount of light under changing
daylight conditions,
(2) change the desired brightness pattern in response to changing events in
the room, and
(3) turn the lights off when the room is unoccupied and turn them back on
when an occupant returns.
Thus ImCon has the potential to address all aspects of energy, building and
human efficiency related to lighting by acting as a "vigilant lighting
designer."
HARDWARE
Only two components needed to be added to the basic ImCon system to control the
Venetian blinds that were already located in the test room: one stepper motor
for each blind and their controller. A complete diagram of the equipment, which
is fully described in Glennie, Thukral and Rea (1991), is shown in Fig. 3.
Because of the modular nature of the digital to analog control hardware used in
this development system, it was relatively easy to incorporate two stepper
motors as additional outputs. Stepper motors were chosen due to their digital
operation; unlike a conventional motor which receives a continuous current, a
stepper motor receives pulses of a pre-determined duration which result in an
incremental movement called a "step." The advantages of these motors are
twofold:
(1) Since a computer is used as the basis for light control, the digital
operation is ideal.
(2) With a conventional motor, monitoring the position of the shade would be
a very complex operation involving feedback loops. With the stepper
motor, the system simply records the number of steps that have been moved
in each direction from a known initial position.
The stepper motor controller (Model SC-149 from Alpha Products, Darien, CT) can
operate up to four motors. It contains its own Z-80 microprocessor with read
only memory (ROM) to store the motor control commands and random access memory
(RAM) to store temporary variables and other parameters. This controller hooks
directly into the existing expansion unit that holds the digital to analog (DAC)
modules and relays that are used to control the electric lights. It is
especially appropriate because of the built-in intelligence of the controller,
which stores the position of each motor and can vary the stepping rate and other
parameters of operation. It also has a backup battery for the RAM so that
stored values are not lost when the system is switched off.
The stepper motors are of a unipolar type (Model MO-104, Alpha Products),
operating at 5 volts. They have a step size of 1.8 degrees and a holding torque
of 0.42 J (0.31 ft-lb), which is adequate to drive the blinds on a window that
is 1.5 m by 2.5 m (4.8 ft by 8.1 ft). While they are able to move the blinds
from one extreme to the other in 10 seconds (which requires almost 2000 steps),
their operation is much quieter if the rate is slowed to 50 steps per second,
and the resulting full stroke time of 40 seconds is not excessive.
As with many commercial Venetian blinds, the angle can be adjusted by turning a
long extension rod that happens to be removable. These rods were removed, and
motors attached through a coupling to the short shaft that remains. Because it
is critical to match the ancjle correctly (to avoid stress on the connection), an
adjustable bracket was fabricated and attached to the window frame.
SOFTWARE
Because of the "intelligent" nature of the stepper motor controller, it is quite
easy to implement any desired motion. It is simply necessary to transmit a
string of characters, such as "MV2 200," which means to open motor 2 by 200
steps. The only precaution which must be taken is to check whether or not the
blinds have reached a limit before executing the command, to prevent overload on
the motor or damage to the blind tilt control).
As far as actual control of the blinds in response to the condition of light
within the room, two algorithms have been developed. Both start with the blinds
fully closed (tilted up on the inside). After an initial image is recorded, the
first algorithm uses brightness data from a specified area on the floor to
-------
2771
detect a substantial change in luminance when the blinds are opened one "notch"
(100 steps or about 9 degrees). The second utilizes data from a large area of
the image and counts the number of pixels that exceed a specified maximum
brightness. In either case, when the limit is reached, the blinds are closed by
one notch. This position is maintained for two minutes, to prevent the
distraction of constant motion, although any time that the limiting condition is
triggered, the blinds can close. After the delay, the blinds will again be
opened by one notch. This cycle repeats constantly through the day, at least
while there is any appreciable exterior light, which can be determined from the
direct view of the window.
A third algorithm has been designed, but is not yet implemented. It uses the
desired brightness levels to check for the presence of excessive light. If the
brightness of any segment exceeds its desired value by some factor, and the
lights are already at the minimum level, the blinds will be closed by one notch.
Of course, all three of these algorithms require further study. Human subject
evaluations will be the next phase of this work, which will help to quantify
some of the values that we are currently pulling out of the air.
PERFORMANCE
At this time, preliminary estimates of energy savings have been made from a few
days of testing. In general, when the diffuse sky is bright (20,000 lx or 2,000
ft-c exterior illumination levels, not including any direct beam component),
measurements show that the electricity used by the light fixtures is reduced by
more than 75% of its full value. Even with a dull overcast sky (10,000 lx or
1,000 ft-c), the energy savings are approximately 40%. On a very dark day
(5,000 lx or 500 ft-c), the overall power reduction is less than 10%.
Extrapolating these results to a full year is not reliable at this point, but a
reasonable estimate of annual energy savings is in the neighborhood of 50%.
Note that this does NOT include savings due to turning the lights off when the
room is unoccupied, which is highly dependent on the usage pattern of the
particular space in which the system is installed.
As far as human visual performance is concerned, ImCon seems to be capable of
maintaining blinds in appropriate positions throughout the day. While the
system is successful at preventing glare, there are still particular instances
that need to be addressed to enhance occupant productivity and satisfaction.
For example, when someone is working at the computer, the blinds directly behind
that person should be tilted very far up or down, so that there is no veiling
reflection in the screen. This effect is a problem whenever there is even a low
level of exterior light. One near-term development will be to maintain limits
on blind position as well as luminance values for different light scenarios.
The present implementation includes an algorithm that detects the extra
brightness that appears in a particular area of the image when a slide is
projected, at which point the system turns off the lights and rotates the blinds
to the full up and closed position.
LIGHTS
RELAY
CCD
CAMERA
VISUAL
DISPLAY
STEPPER
MOTORS
DIMMING
CONTROLS
IMAGE
ACQUISITION
D/A
CONVERSION
MICROCOMPUTER
I/O BUS
Fig. 3, Hardware Schematic
-------
2772
MODIFICATIONS AND IMPROVEMENTS
Early results indicate that it would be beneficial to incorporate additional
motors that can actually raise and lower the blinds completely. On dull,
overcast days, of which Troy, NY, has no lack, the energy savings can be
increased from 10% (blinds tilted open as much as possible) to 40% (blinds
completely raised). The only concern is that the major motion of blinds sliding
slowly (or quickly) up and down would be a significant distraction, which may be
detrimental to human response (Boyce, 1980). It should be possible to establish
some reasonable guidelines for when this step is taken. For example, the ImCon
could start the day with the blinds at the top, and leave them there as long as
the window brightness does not exceed a specified maximum.
A second concern is the provision of some manual override for the blind
position. People do vary in their response and preference for outside light and
view, and some means of permitting that variation must be incorporated. Because
of the need to maintain blind position through the computer (to avoid forcing
the blinds past a limit), the control should be made through the system. It is
possible for this to take many forms, such as a special keypad, a switch located
at the window or some form of telephone key combination. Initially, control can
most conveniently be made through the computer that is the brain of the entire
system. However, it is anticipated that the production version of ImCon would
be much reduced in size and cost, so some other means will be required.
A final contemplated improvement is with the blinds themselves. Some existing
daylighting systems have used inverted blinds with a reflective upper surface.
This serves to redirect the sunlight to the ceiling where it can provide general
illumination without glare. Although previous examples have been based on
manual or "time of day" control, the same type of blinds may be added easily to
the image based control.
CONCLUSIONS
The original ImCon system design has proven to be quite easy to expand. By
using an image of the room and a computer to provide lighting control, a degree
of precision and response is possible that cannot be approached by existing
techniques. The automatic adjustment of Venetian blinds is relatively simple to
accomplish and successful at increasing the energy savings beyond that which is
possible for less sophisticated time-clock or photocell-based controls.
ACKNOWLEDGMENT
Major support for this work was provided by Public Works Canada, through a
contract to MSR Scientific Enterprises, Inc. Additional funding was from the
Niagara Mohawk Power Corporation Seed Research Program and Rensselaer's Lighting
Research Center. Our thanks are also due to Mark Rea, Director of the Lighting
Research Center, for his suggestions and inspiration. Joseph Kleinmann, Program
Head for Research and Development in Efficient Lighting; -Russ Leslie, Associate
Director; and Peter Boyce, Program Head for Research and Development in Human
Factors, also provided valuable input to this project. Finally, Judith Block,
Editor, made it all readable.
REFERENCES
Boyce, P. R. (1980). Observations on the manual switching of lighting. Lighting
Research and Technology, 12, 195.
Glennie, W. , I. Thukral and H7 S. Rea (1991). Feasibility demonstration of a new
type of lighting control system. To be presented at the Illuminating
Engineering Society Annual Conference, Montreal, Quebec, August, 1991.
Hopkinson, R. G. and J. D. Kay (1966), The Lighting of Buildings. Praeger,
New York.
Jaekel, R. R. and M. S. Rea (1983). A case study of a daylight-linked dimming
system for fluorescent lamps. Building Research Note No. 194, Division ot
Building Research, National Research Council of Canada, Ottawa.
Rea, M. S. (1984). Window blind occlusion: a pilot study. Building and
Environment, 19, no. 2, 133-137.
Real TTI ST; K7 F. Dillon and A. W. Levy (1987). The effectiveness of light
switch reminders in reducinq liqht usaqe. Liqhtinq Research Technology, 19,
81-85.
Rubinstein, F., G. Ward and R. Verderber (1989). Improving the performance of
photo-electrically controlled lighting systems. Journal of the Illuminating
Engineering Society, 18, 70-94.
Verderber, R. R., O. C. Horse and J. E. Jewell (1989). Building design: Impact
on the lighting control system for a daylighting strategy. IEEE Trans. Ind.
Applic., 25, no. 2, 198-201.
-------
2773
VARIABLE INCIDENCE SPECTRAL TRANSMITTANCE MEASUREMENTS FOR DETERMINING
GLAZINGS' SOLAR PARAMETERS FOR DAYUGHTING PURPOSES
A. Fanchiotti*, G. Gagliardi", A. Piegari* P. Potato**, M. Vio*
* University di Roma, Rome, Italy
SIV, Society Italiana Vetro, San Salvo, Chieti, Italy
'Thin Film Lab., ENEA, Rome, Italy
" Stazione Sperimentaie Vetro, Murano, Venice, Italy
$ istituto Universitario di Architettura di Venezia, Venice, Italy
ABSTRACT
A visible and near-infrared spectrophotometer, equipped with a variable incidence angle accessory, has been
arranged and used for measurements on coated and uncoated architectural glasses. The paper presents the
results obtained by using the experimental data so collected as an input to an integrated energy analysis com-
puter code, HEATLUX. The paper presents an example to show the impact of different glazing systems on the
global energy requirement.
KEYWORDS
Daylighting, glazings, non normal incidence, light transmittance
INTRODUCTION
Electric lighting requirements in buildings can be effectively reduced by daylighting. Transparent materials play
an essential role in admitting natural light into buildings. It is, thus, very important to assess their optical proper-
ties, as new materials are proposed and made available on the market, in order to estimate the consequences
of different design choices on .he energy performance of a buiding. Such behaviour can be evaluated by means
of computer programs performing energy analyses, which need accurate data on glazing systems. Several
methods for computing natural lighting in interiors have been developed by different laboratories. A research
project supported by the EEC has reviewed various existing daylighting computer programs. A new lighting
and thermal energy analysis simplified program, called HEATLUX, has been developed within the project
(Fanchiotti and Matteoli, 1989).
HEATLUX considers a room with rectangular apertures on vertical walls, closed with transparent glazings. An
unobstructed surrounding is assumed, and both externally and internally reflected light is taken into account.
The code provides a procedure for computing the natural illuminance at the points of a grid, as defined by the
user, on an hourly basis, as well as the associated solar gains. HEATLUX then estimates the necessary electric
lighting integration, includes it among the thermal gains, and finally computes monthly energy consumption for
heating, cooling and lighting.
LIGHT TRANSMITTANCE OF GLAZINGS
The presence of glazing on a window aperture reduces the light transmitted into the room. This reduction, the
more significant the greater the incidence angle, must be taken into account when computing thermal and light-
ing energy consumption. Different equations have been proposed for introducing such correction. HEATLUX
-------
2774
currently employs the widely diffused Rivera's formula (Rivera. 1958) for computing the light transmittance, ™
as a function of the incidence angle, 0:
tv(0) = 1.018tv(0) cosfl (1 + sin3e) (1)
However, this approximation, is not satisfactory for angles greater than 50 degrees and for some kind of glas-
ses. As an example, the comparison between measured light transmittances, tvM, and light transmittances given
by Rivera's formula, tvr, is reported in TABLE 1.
TABLE 1. Comparison between measured and calculated fRiverol
light transmittances for single glass plates
6
tvM(A)
tvR(A)
TvM(B)
tvR(B)
0
88.4
90.0
72.8
74.1
40
87.2
87.2
70.5
71.8
60
80.5
74.2
63.5
61.1
When the impact of different glazing systems on energy requirements have to be evaluated, the correct values
of their transmittances are needed. Therefore, tables of data, instead of equations, were given as input to the
HEATLUX program. These data were measured and partially calculated following the procedure described in
the next section.
LUMINOUS, SOLAR AND THERMAL CHARACTERIZATION OF THE DOUBLE GLAZINGS
The examined samples were four 6 mm thick commercial float glasses:
A) uncoated dear glass; B) uncoated green glass; C) dear glass with a reflecting coating; D) clear glass with
a high transmittance, low emissivity coating.
These samples are representative of classes of architectural commercial products with different luminous, solar
and thermal performances. In fact A is a glass transparent in the whole solar range, B Is an absorbing glass
partially reducing the light and solar radiation, C is used for daylight and solar control strongly reducing the
light intensity and the near infrared radiation by reflection and absorption, D reduces heat losses and therefore
increases thermal insulation without considerably modifying the luminous properties as compared to sample
A.
Double glazings were considered with the uncoated clear glass as outer sheet and one of the samples (A, B, C
or D) as second sheet. Coatings of samples C and D are turned towards the interspace. For brevity sake the
double glazings will be indicated as AA, AB, AC, AD.
A commercial spectrophotometer equipped with a suitable accessory was used for recording the transmittance
curves of the single glass sheets at different angles of incidence. Two specimens of the same sample are re-
quired for each measurement (Fig. 1). The second specimen neutralizes the lateral shear of the Incident beam
MM?U_C9MU!!MKI * 1
Fig. 1. Accessory for the transmittance Fig. 2. Transmission in a double glazing
measurement at the different angles of incidence (1; external sheet, i: interspace (not to scale),
(D: depolarizer, S: sample, P: polarizer). 2: internal sheet).
-------
2775
due to internal refractions in the first one. The spectrophotometer beam hits, therefore, the same area of the
sensitive window of the detector during the recording of the base line (100% line) and the measurement. Two
solidaly cog-wheels carry the sample holders where the two specimens are located and allow equal and op-
posite rotations. In this way the beam hits the two specimens with angles of 0 and - 0, respectively. The angle
of incidence can be modified thanks to a graduate cylinder solidal to the two wheels. Preliminary checks were
performed on the angle accuracy, using 6 mm ultra-pure silica plates, whose theoretical transmittance can be
worked out from refractive index values.
Since at oblique incidence the sample behaviour depends on the beam polarization, the polarization state during
the measurements must be known. The spectrophotometer beam in the sample compartment has generally an
unknown polarization, therefore to record transmission curves under specified conditions of polarization the
accessory is equipped with a depolarizer and a polarizer (Fig. 1). The polarizer is rotated until a type p or type
s polarized beam is obtained. By considering the square root of the value recorded by the instrument at each
wavelength, I, the type p or type s transmittance curve, Tp(x,0) or T
s( X, 0) is obtained. The angular transmittance for unpolarized light, corresponding to tfie real situation of solar
radiation, is obtained with an arithmetic mean: T(x,0) = (Tp(x,0) + Ts( X, 0 ))/2
The spectrophotometry measurements performed on the single glass sheets allow the overall spectral trans-
mittance to be obtained for each double glazing by computation (Fig. 2):
T(X,0) = Ti (X,0) T2( X, 0 ) / [ 1 - Rl (X,0) R2( X, 0 )] (2)
where Ti (X,0) ,T2(x, 0)are the transmittances of the component sheets 1 and 2; Ri(x,0) ,R2( X. e) arethe
reflectances of the component sheets on the interspace side.
The spectral reflectances of the single glass sheets were measured at different angles of incidence using a
home-made variable angle reflectometer (Castellini and colleagues, 1990).
The light transmittances of each double glazing corresponding to different values of 0 were obtained by the
following integration:
780 nm
780 nm
Tv(0) = JT(X, 0) D(X) V(X) dX// D(X) V (X) dX (3)
380 380
where T( x, 0) is the spectral transmittance curve of the double glazing; D(x) is the reference illuminant simulat-
ing the natural light; V(x) is the sensitivity function of the human eye. The two functions, D(x) and V(X), are stand-
ardized by ISO (ISO 9050, 1990).
The light transmittance obtained for each glazing is shown In Fig. 3.
Starting from refractive index, absorption coefficient and thickness of substrates and coatings and by consider-
ing a proper mathematical model of the glazing, spectral transmittances and reflectances can be calculated at
different angles of incidence (Born and Wolf, 1965). The comparison between measured and calculated values
allows the model to be verified and possibly used for angles of incidence where measurements are not reliable.
For our samples TABLE 2 shows the comparison between measured and calculated light transmittance of
double glazings at 0°, 40° and 55.6° (silica Brewster angle). The good agreement allows the values for angles
of incidence greater than 60° (not allowed by the spectrophotometer and reflectometer) to be calculated (Fig.
3).
TABLE 2 Measured and (calculated)llaht transmittances (%)
0 0 40 55.6
AA 78.7(78.7) 76.7(76.7) 69.5(70.2)
AB 65.5(64.9) 62.6(62.0) 55.8(55.6)
AC 35.0(35.1) 33.8(33.7) 31.6(31.1)
-------
2776
90 ¦
•0
so
40
30 ¦
Fig. 3. Light transmittances of double glazings.
The light transmittance in the range 0 - 90 degrees is one of the glazing parameters required for daylighting
purposes, but the main parameter characterizing the energy behaviour of an architectural glazing is the solar
factor g defined as the sum of the solar direct transmittance and the secondary heat transfer factor of the glaz-
ing towards the inside (ISO 9050,1990). For the determination of the solar factor, emissivities of the sheets sur-
faces and thermal transmittances (U-values) for each double glazing are required and determined according
to Proposed ISO Standard DP 292.
The HEATLUX code requires as input data the thermal transmittance and the solar factor for an angle of in-
cidence of 60°. The values of U and g(60) obtained for the examined double glazings, with an interspace 12 mm
wide full of dry air, are reported in TABLE 3.
TABLE 3 Parameters characterizing the thermal and solar behaviour
of 6-12-6 double glazings
Double
glazing
U
W/rr^K
9(60)
(%)
AA
2.9
60.2
AB
2.9
53.2
AC
2.6
44.8
AD
1.9
52.2
The behaviour of g as function of the incidence angle is going to be included in the next revision of HEATLUX.
EXAMPLE
As an example of using the experimental data on glazings discussed in previous sections, the code HEATLUX
has been used to compute the energy requirements of a room located in Rome. The room is square, with a 9
m side, and 3 m height. It has only two external walls, one facing south, the other facing north. There are two
types of windows in the gouth wall: either a small window (1.5x5 m), or a large window (1.5x8 m), each one
tested with the four glazing systems already presented. The program computes the energy requirements for
space heating and cooling, and for electric lighting, month by month, with set point temperatures of 20°C for
winter and 24°C for summer. The minimum required illuminance on the working plane is 500 Ix. The heating ef-
-------
2777
ficiency Is .85, the cooling system COP is 1.6, and the luminous efficiency of electric lighting Is 60 Im/W. The
results for the whole year are shown in the istograms of Fig. 4.
ROMA
South window
HE ATI WO COOU NO UGHT1HQ TOTAL
Energy consumption
-------
Intentionally Blank Page
-------
3.9 Atriums
-------
Intentionally Blank Page
-------
2781
Geometric Shape Index for Daylight Distribution Variations
in Atrium Spaces
A. Liu M. Navvab J. Jones
College of Architecture And Urban Planning
The University of Michigan
Ann Arbor. MI
ABSTRACT
This paper presents a new approach to model the effect of geometric shape variations and daylight distributions
within the cuboid atrium space. Computer simulation results combined with actual building monitoring show the
impact of the atrium geometric proportions on the daylighting distributions within atrium spaces. The results
demonstrate the use of simple geometric concepts in the construction of a daylighting prediction model under
overcast sky conditions. Two basic geometric proportions, the plane aspect ratio (PAR) and the section aspect ratio
(SAR), are employed to form the basis of the daylighting prediction model. The well index (WI), a combination of
PAR and SAR, is considered as an indicator and used to derive a mathematical model for the generation of
daylighting distributions in atrium spaces. Actual building monitoring results also confirm the simulation findings.
INTRODUCTION
Over the years, a few studies involving the idea of light-wells were presented to describe the effect of atrium ge-
ometry on the illuminance levels within a typical atrium and adjacent spaces [1-5], However, in practice, there
exists no simple and reliable atrium daylighting prediction model related to simple geometric concepts for designers
to quickly evaluate the atrium performance at the preliminary design stage. With the current design tools, designers
are unable to systematically evaluate the influence of allium shape variations on the use of daylight in atrium build-
ings. Furthermore, atrium buildings with different shapes and various glazing systems create a large area source of
light with very complex luminous environments. The application of daylight in atrium has resulted in,sometfaieSj
an uncomfortable visual environment and relatively high building energy consumption. Therefore, the idea of
having a prediction model relating to simple geometric proportions is becoming more important for designers to deal
with issues of atrium daylighting performance and then to make an optimum selection for sizing the glazing system
and its form with other design considerations, e.g., heat, noise, air quality.
Appropriate geometric proportions can be used to describe and analyze the related physical phenomena caused by
the shape itself. Geometric relations can also provide an effective mathematical model for the database generation.
Analysis of those phenomena alonj, with their related atrium shapes will be an effective method for studying the
advantages and disadvantages of a specific atrium form. The results will provide a better basis for making a
reasonable selection among different atrium shapes. For this reason, this study focused on the daylighting
perfoimance of the atrium space by the exploration of the influence of geometric proportions.
Despite the fact that atria have diverse forms and are very different from each other with regard to their interior
landscaping. Atrium spaces can be grouped into a number of often used geometric forms. Based on the classifi-
cation of the most preferred existing atrium space forms, these geometric fr ms consist of the cuboid, the cylinder,
the pyramid, or combinations of these forms. According to the author's in\ .^ligation of existing atrium spaces, the
cuboid is the most frequently adopted form in atrium space designs. Therefore, this study focuses on the cuboid
atrium, especially in the most often used four-sided atria involving the structural skylight system.
BACKGROUND
Due to the variations of glazing locations, two daylighting concepts are often used in atrium daylighting designs:
side-lighting, light from vertical walls, and top-lighting, light from roof cover. In this study, based on the selected
atrium type, the side-lighting concept is only used for analyzing the daylight contributions in adjacent spaces. The
top-lighting concept is concerned in the analysis of daylight distributions in atria. It includes four different generic
systems. The first one is the structural skylight system which has an entirely glazed rooftop and is adopted in most
atrium spaces developed in recent years. The second one is the skylight with a light-well, which introduces daylight
through a domed or flat skylight in the roof. The third is clearstory apertures. The last one is the roof monitors [4].
-------
2782
Based on the combinations shown in Fig. 1, type "A+0", a four-sided atrium with the structural skylight system,
is employed to examine the related daylighting performance with respect to its geometric features. This task was ac-
complished by using two basic geometric proportions and their combination: the plane aspect ratio (PAR), the
section aspect ratio (SAR) and the well index (WI). The PAR is the ratio of the atrium width to the atrium length
(w/l). The SAR is the ratio of the atrium height to the atrium width (ty\v)- For instance, PAR and SAR are equal
to 1.0 for a cube. PAR, SAR and WI are defined as follows [6]:
PAR:
SAR =
Atrium Width
Atrium Length
Atrium Height
Atrium Width
W
K w'
WI =
Well Height x (Well Length + Well Width )
2 x Well Length x WeU Width
_ Hx(L-fW) W
2xLxW L.
1 , Hx(L + W)
" 2 LxW I H
iQQQdl
Top-lighting 1
(Section)
Side-lighting CO E0EQ
(Plane)
H
Note:
Glazing
Solid wall i
Fig. 1 Classifications of daylighling systems in the cubic atrium shapes.
For a skylight with light-well system, the WI is an essential index and is employed to find the efficiency factor of
a light-well, which is based on the interreflections within light-well surfaces. It relates to the dimensions of the well
and is used for the calculation of the net transmittance of the skylight-well system [7]. This new approach assumes
that the skylight, in the selected atrium type "A+0", is similar in function to a glazed rooftop and the enclosing wall
surfaces are the same as surfaces of the lightwell.
The use of a grid system (reference points of calculation) is a more effective way for evaluating various daylight
distribution patterns. .When the same number of reference points are equally placed in different atrium spaces having
the same geometric proportions, the calculated Daylight Factor (DF) distribution patterns keep constant regardless of
variations in the distance between any two reference points (see Fig. 2). This finding was also validated within the
sky simulator using scale model photometry techniques. In this study, a grid system with 11x11 reference points
were used in order to compare various daylight distributions.
Overcast

PAR=1.0
SAR-1.0
90 WI=1.0
10 30 50 70
(a) with various atrium dimensions.
1 Offices * Institutional
~ Housing
£3 Store &
&
&
Shopping
Civic
Hotels
arcades
Dimensions:
%
—¦ —
35
30x30x30

25
60x60x60
90x90x90
15
Wall rcf.

=45%

1 3 5 7 9 11
(b) with a grid system (Grid: 11x11).
Fig. 2. Comparison of DF distributions in the central line
of the floor plane, from computer simulations)
jiMI
lifil, an.—j-
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
SAR
Fig. 4, SAR distributions in various building types.
DIMENSIONAL SCOPE
From the view of building designs, there are no theoretical limitations of the dimensions of atrium spaces.
However, based on the authors' investigation of existing atrium spaces, the often adopted range of atrium dimensions
is existential and can be employed to study the daylighting strategies often preferred by most designers. Atria are
frequently adopted in the following building types: offices, institutional and civic, housing and hotels, as well as
shopping arcades and stores [6]. Within this classification, the distributions of the preferred PAR and SAR are
shown in Figs. 3 and 4, respectively.
Figs. 3 and 4 indicate that different building types have different PAR and SAR ranges. The selected dimensional
scope is from 0.1 through 1.0 for the PAR and is from 0.5 through 4.0 for the SAR, respectively. Fig. 5 reveals
that the preferred WI are located between 0.1 and 3.0. This means that WI greater than 3.0 are less commonly
adopted in real designs. The WI range considered in this study is from 0.5 to 4.0. The prediction model derived in
this study is based on these delimitations.
-------
2783
¦
Offices
m
Institutional

& Civic
~
Housing

& hotels

Stores &

Shopping

Arcades
PAR
0 0.2 0.4 0.6 0.8 1.0
Fig. 3. PAR distribution in various building types.
45*
35
25-
1
lu
¦ Office buildings
| Institutional
& Civic bldgs
~ Hotels & Housings
E5 Stores & Schopping
arcades
-^1W1
1 2 3 4 5
Fig. 5. WI distributions in various building groups.
STUDIES OF PAR & SAR
For most atrium cases, the sky component (SC) and the internal reflected component (IRC) are two important
components in the analysis of daylight distributions. Their influences in various atrium shapes can be explained
through the studies of PAR and SAR relations. For a given geometric proportion, the curvature of DF distribution
patterns in the central line of the floor plane is exactly alike regardless of variations of wall reflectances. Various
wall reflectances only change the amount of daylight factor. Variations of the SC are more dramatic and are greater
than those of the IRC, especially when atrium depth is increased for the same PAR. Fig. 8 shows that the SC has
less of an influence on tall atrium spaces. As increasing depth, the curvature of the DF curves becomes flatter.
The influence of the PAR on DF distributions is shown in Fig. 6. For a given SAR, when the PAR gets small
(the plane shape becomes narrow) the slope of the DF distribution curve along the length shows a tendency to flatten
around the middle part of the calculated points. However, DF distribution curves along the width are nearly alike.
Fig. 7 shows the effect of SAR variations on the DF distributions. For a given PAR, when the atrium space is
shallow, the sky component becomes dominant. On the contrary, the sky component is not at all dominant as the
atrium space becomes deeper. The DF curve flattens as this occurs. The daylight is nearly evenly distributed in the
horizontal plane. As observed, when the SAR rises to a certain range, the DF distribution curves will keep constant.
However, the magnitude of the DF within this range is insignificant for the use of daylight because it is too small.
SAR=1.0
Overcast
PAR=1.0
30
PAR=0.75 (L)
30x40x30
— PAR=0.25 (W)
30x120x30
PAR =0.5 (W)
30x60x30
PAR =0.75 (W)
30x40x30
PAR=1.0(W)
30x30x30
- PAR =0.25 (L)
30x120x30
- PAR =0.5 (L) _ PAR=1,0 (L)
30x60x30 30x30x30
(b) along the length.
(a) along the width.
Fig. 6.DF distributions with various PAR's in the central line
of the floor plane. (data from computer simulations)
30x30x15
(SAR=0.5)
— 30x30x30
(SAR=1.0)
30x30x45
(SAR=1.5)
30x30x60
(SAR=2.0)
30x30x75
(SAR=2.5)
30x30x90
(SAR=3.0)
30x30x105
(SAR=3.5)
— 30x30x120
(SAR=4.0)
Fig. 7. DF distributions with vaiying SAR's
in the central line of the floor plane.
STUDIES OF WELL INDEX
The well index (WI) can be expressed as: WI = 0.5 x SAR x (1 + PAR). It indicates WI = / (SAR, PAR). Since
WI is a two-variable function, the WI can be employed to represent different combinations of PAR and SAR with
the exception of PAR=1.0. For example, a cuboid (30x30x60) with PAR=1.0 and SAR=2.0 has the same WI value,
WI=2.0, as the one (30x120x96) with PAR=0.25 and SAR=3.2. They have different geometric volumes and seem to
have different daylight distribution patterns in the horizontal plane.
A comparative study by computer simulations was conducted to see the variations of daylight levels at the central
point of the floor plane. A group of simulated results for WI=0.75 are listed in Table 1. It reveals that with varying
combinations of WI the calculated DF differs only slightly. As observed, an increase in the length of the horizontal
plane compensates for the decrease of DF caused by an increase of depth in a section. A comparison of related DF
curves is shown in Fig. 8. Another set of simulated results for WI=2.0 is listed in Table 2. The calculated DF
values differ little at the central point. The related DF distributions are shown in Fig. 9.
-------
2784
Table 1 Comparison of DF values for WI=0.75.
WI
SAF
PAR
DF(%)
LxWxH
Angle of visual sky
central pt
L-side
W-side
0.75
0.75
1.0
42.21
30x30x22.5
67 2
67.2
0.75
1.0
0.5
42.19
60*30x30
90
53
0.75
1.2
0.25
41.29
120x30x36
118
45
Table 2 Comparison of DF values for WI=2.0.
\W
SAR
PAR
DF (%)
LxWxH
Angle of visual sky
central pt.
L-side
W-side
20
2.0
1.0
11.1
30x30x60
28
28
2.0
2.7
0.5
11.74
60x30x80
41
21
2.0
3.2
0.25
13.37
120x30x96
64
18
(data from computer simulations)
(data from computer simulations)
Overcast wi = 0.75
~ PAR = 1.0
SAR = 0.75
PAR = 0.5
SAR =1.0
» PAR = 0.25
SAR = 1.2
Grid: 11x11
Fig. 8. Comparison of DF distributions for WI = 0.75.
(data from computer simulations)
WI = 2.0
~
PAR =1.0

SAR = 2.0
0-
PAR = 0.5

SAR = 2.7
m.
PAR = 0.25

SAR = 3.2
Grid: 11*11
Fig. 9, Comparisons of DF distributions for WI = 2.0
(data from computer simulations)
As observed, for small WI (=0.75), when the SAR becomes large (tall) and the PAR becomes small (narrow), the
calculated DF value decreases. There is a difference at both sides of the atrium floor. On the contrary, for large WI
(=2.0), when the SAR becomes large and the PAR becomes small, the calculated DF value increases. There is little
difference in DF levels at the central portion of an atrium floor. This phenomena is due to the effect of interior
surface reflectance. For lower WI, a shallow atrium, Sky Component is dominant; for higher WI, a tall atrium,
Internal Reflected Component is dominant.
A PREDICTION MODEL
Fig. 10 shows the relationship between the DF and the variations of PAR and SAR. As observed, for each speci-
fied SAR, say SAR=1, the DF increases when the PAR decreases. This is due to an increase in the skylight area.
As discussed previously, WI is a function of PAR and SAR. It also is a good index for daylighting prediction
because it produces very little DF difference in related geometric forms. Therefore, it is appropriate to consider WI as
an index in a daylighting prediction model. Fig. 11 illustrates DF distributions with relation to the WI.
80
60
EF
40
20
Overcast
Wall reflect an ce=45%
Floor reflcctance=20%
4-sided open atrium
EF
«— PAR=0.2
~ PAR=0.4
PAR =0.6
«— PAR =0.8
•— PAR=1.0

Overcast
Wall reflects nce=45%
Floor reflectance=209E>
4-sided open atrium

-t WI
SAR
1
Fig. 10. DF variations with various PAR and SAR
(at the central point of the floor plane
Fig. 11, DF distribution of an open atrium
with surface refleciance=45%.
(at the central point of a floor plane)
By using the curve fitting method, a polynomial regression function can be derived to fit these data points. In
this case, for example, the fifth-order polynomial regression function is:
y=103.56 - 121.09x + 64.203x2- 17.61x3 + 2.3934x4 - 0.12676x5
where y = the predicted DF, x = the WI.
Based on the derived polynomial regression function, the DF for overcast sky conditions at various WI can be
quickly calculated by using a programmable calculator. Similar functions would provide a quick approach to
evaluate two geometric design options using this database.
Based on the proposed prediction model, a DF table is calculated and listed in Table 3. From Table 3, designers
can quickly recognize the change of DFs as an indicator of daylight contribution according to simple geometric
-------
2785
proportions: PAR and SAR. For example, say an atrium space has an SAR=1.5, when the plane shape is changed
from PAR=1.0 to PAR=0.5; the DF at the central point will be increased from 18.1% to 27.1%.
Table 3 Daylight Factor Prediction Table.
(Overcast, Rw-45%)
Fig. 12.WI distributions with varying surface reflectances.
Daylight
PAR
Factor (%]
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

1.1
50.2
46.9
44.0
41.2
38.5
36.1
33.8
31.7
29.7
27.9

1.2
46.7
43.7
40.7
37.8
35.2
32.8
30.6
28.5
26.7
24.9

1.3
44.0
40.7
37.6
34.8
32.3
29.9
27.7
25.8
23.9
22.3

1.4
41.1
37.8
34.8
311
29.6
27.3
25.2
23.3
21.6
20.1
SAR
1.5
38.5
35.2
32.3
29.6
27.1
24.9
22.9
21.2
19.5
18.1

1.6
36.1
32.8
29.9
27.2
24.9
22.8
20.9
19.2
17.7
16.4

1.7
33.8
30.6
27.7
25.2
22.9
20.9
19.1
17.6
16.2
14.9

1.8
31.7
28.5
25.8
23.3
21.2
19.2
17.6
16.1
14.8
13.6

1.9
29.7
26.7
23.9
21.6
19.5
17.7
16.2
14.8
13.6
12.5

2.0
27.9
24.9
22.3
20.1
18.1
16.4
14.9
13.6
12.5
11.5

0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
%
80
60
EF
40
20
(data from computer simulations)
Overcast
4-sided open atrium
^ R=60%
^ v
R=45%
partially listed
-f WI
6
The graph and table described above are good for only one surface reflectance. Sometimes designers want to see
the DF change when the surface reflectance is modified. For this reason, it is necessary to have a graph which shows
the DF curves relating to corresponding surface reflectances. Fig. 12 is an example showing DF distributions with
various surface reflectances. Similarly, we can employ the curve fitting method to derive a polynomial regression
function for each curve. Once these functions are derived, a set of tables (similar to Table 3) can be generated to help
designers quickly check DF changes between atrium geometric forms and surface reflectances.
ACTUAL BUILDING MONITORING
Currently, the University of Michigan daylighting research laboratory is conducting atrium building performance
studies [vj. Horizontal illuminance sensors with fixed and open ficlds-of-view provided the capability to examine the
contribution of the surface reflectances within the atrium. The vertical illuminance on the south wall, plus hori-
zontal illuminance at various height were also measured (see Fig. 14).
Data is being collected every ten seconds with averaging over 30-second intervals and is recorded and stored in an
IBM PC/AT compatible computer every day. Photometric illuminance measurements were made by the open and
fixed field-of-view illuminance meters at the center of the atrium during different days and sky conditions. The
results show the ratio of the two sensors for the transmitted daylight through the glazing system with its obstruction
loss due to the structural framing system [8]. For overcast skies, the variations of light from the top to the bottom
of the atrium vertical wall surfaces are also shown in terms of the ratio of low to top sensors in Fig. 15. The
variation and the sudden increase in ratio on a clear sky condition are due to reflected sun light from the glass
windows inside the atrium. The results show the dynamic effect of the internal reflected components in Fig. 15.
The study shows the reflected components have a major impact on the daylight distribution within the atrium.
The measured vertical illuminance at various heights shows the contribution of reflected light from the glass
windows located on the east wall onto the surface of the south wall during the afternoon time when the sun altitude
is at its low angle. See Fig 16. The west wall does not have any windows so the morning illuminance levels on
the south wall are less than the afternoon measured levels. See Fig. 17.
CONCLUSIONS
Under overcast skies, observations about the relationship between atrium geometric proportions and related day-
lighting distributions indicate that the use of geometric proportions can help designers Find certain relations for the
prediction of atrium daylight distributions under other sky conditions, e.g., clear or partly cloudy sky. The well
index (WI) can be a good index for using in a prediction model because it produces very little DF difference at the
central point of related geometric forms. Future studies will extend from this base to clear sky conditions and to the
overall daylighting performance around an atrium space and its adjacent spaces. Data from actual atrium building
monitoring studies confirm the finding from the computer simulation studies using Superlite daylighting program.
This will be further discussed to aid the exploration of atrium daylighting performance in relation to shape variations
under clear skies. A new indicator other than Daylight Factor may be applied for clear sky conditions.
ACKNOWLEDGMENTS
This work is a part of author's current studies for Doctoral Dissertation. Special thanks arc due to Professors
Kurt Brandle, Norman Barnett, College of Architectural and Urban Planning, The University of Michigan, for their
encouragement and guidance during the past year. The actual building evaluation work was supported by the
-------
2786
Department of Plant Extension at the University of Michigan and the office of the Campus Architect. Special
thanks to Mark Luther and Robert Young for their assistance in building monitoring.
REFERENCE
1. Caitwright, V.," Sizing Atria for Daylighting", The 2nd International Daylighting Conference, Long Beach, CA,
1986.
2. Cole, RJ„ " The Effect of the Surfaces Enclosing Atria on the Daylight in Adjacent Spaces", Building and
Environment, Vol. 25, No. 1, Jan 1990, pp.37-42.
3. Kim, Kang-Soo, Development of Daylighting Prediction Algorithms for Atrium Design, Ph.D. Dissertation in
Architecture, Texas A & M University, May 1987.
4. Nawab, M. and Selkowitz, S., "Daylighting Data for Atrium Design", Proceedings of the Nineth National
Passive Solar Conference, Columbus, Ohio, 1984.
5. Aschehoug, O., "Daylight Design for Glazed Spaces", Proceedings of the 1986 International Daylighting
Conference, Long Beach, CA, Nov. 1986.
6. Bednar, Michael J., The New Atrium, New York: McGraw-Hill Inc., 1986, pp. 66 & pp. 140.
7. Illuminating Engineering Society, IES Lighting Handbook, Reference Volume, IES, New York, 1984, pp. 7-22.
8. Nawab, M., "The Application of Daylighting in University Office Building", 13th National Passive Solar
Conference of ASES, Boulder, CO, 1986.
9. Jones, J., Luther, M., Nawab, M. "A Comparative Analysis for Two Geometrically Different Atria", 16th
National Passive Solar Conference of ASES, Denver, CO, 1991.
Ch 13
¦>°Ch 16
sensors)
»Q"4 Chi
Ch 15 Opezt
Ch
(S - N section)
Fig 14. Arrangement of photometric sensors
at the Chemistry Building atrium of
the University of Michigan.
1000
800
600
400
200
(data from actual building measurements
at the Chemistry Building atrium on August 1989)
0
Ratio of fixed/open field - cloudy • Aug. 16
Ratio of fixed/open field • clear - Aug. 26
Ratio of vertical illuminance for lowAop floor
—
£
16
18
8 10 12 14
Time of day (hr)
Fig. 15. Comparisons of daylight measurements al various
reference points.
Aug. 26,1988 - Clear sky
(4 minutes average)
& Top(Oi. 11)
x Mid. (Ch. 13)
.~ Low (Ch. 15)
Daylight level drops at noon
(affected by wall reflectances)
3
Daylight level increases
- (sunlight reflected from -
the glazing on east wall)
Effect of structural
mem ben
10 12 14
Time of day (hour)
Fig. 16. Measured daylight levels on the south wall.
Fig. 17, The interior of the Chemistry Building
at the University of Michigan.
-------
2787
COMPARATIVE STUDIES OF FIVE ATRIUMS ON THE EFFECTS OF
ORIENTATION, EXPOSURE AND DESIGN ON DAYLIGHTING,
TEMPERATURE, AND STRATIFICATION OF AIR
Eino 0. Kainlauri, Ph.D., AIA
Professor of Architecture
G. J. Lehman and M. P. Vilmain
Research Assistants
Iowa State University
Ames, Iowa 50011
ABSTRACT
For the past three years, extensive studies have been conducted in the College
of Design atrium of Iowa State University (previously reported in the Proceed-
ings of the National Passive Solar Conferences in Cambridge, Massachusetts, 1988,
and Denver, CO,1989.) Since then, four additional atriums have been included,
while additional studies have continued in the College of Design atrium (on
acoustics.) As the four additional atriums all have different orientations,
exposures and design, the new studies provide more understanding on the effects
of these characteristics on daylighting, temperature and stratification of air
in these buildings. The additional buildings are: Parks Library Addition,
Durham Computer Center, Beardshear Administration Building, and the Soil Tilth
Laboratory.
The relationships between effects of sunlight, outdoor temperatures and humidity
are interpreted from recorded data which is shown by graphics. Among the re-
sults from the research are diagrams showing seasonal indoor temperature curves
in response to the outdoor temperatures, modified by heating and cooling systems.
Daily temperature variations are also recorded. Daylighting effects on top and
bottom floors of three of the atriums are shown together, while a comparative
diagram of the College of Design atrium is shown regarding a different season.
The Soil Tilth Laboratory research results are described only briefly, because
of the smallness oi the atrium and minimal interfacing with the rest of the
building. The three new major atriums, Parks, Durham and Beardshear, are illus-
trated by photographs. Occupant comfort and behaviour are observed especially
regarding areas in direct sunlight. Suggestions are given on proper design of
atriums in thier role as parts of passive solar buildings.
KEYWORDS
Atrium; daylighting; air stratification; temperature and humidity; orientation
and exposure; passive solar design; diurnal and seasonal variations; skylights.
INTRODUCTION
The College of Design atrium is five stories high, open ended, with a vaulted
skylight oriented from southwest to northeast the length of the building. The
Parks Library Addition is four stories high with an atrium at its southeast
quadrant, all glass exterior walls, and with a small skylight at the (inner)
northeast corner of the atrium. Durham Computer Center has three stories, with
and open ended atrium running west to east, and no skylights. The Beardshear
-------
2788
Administration Building is an older building with some multicolored skylights
over a four story atrium, and a central Rotunda with a dome. The Soil Tilth
Laboratory is three stories high, with a south to north skylight above a stair-
well/corridor area.
North
~"N
C. of Design Parks Library Beardshear Durham Soil Tilth
Fig. 1. Orientation diagram of the five atria. (No scale.)
RESEARCH PROCEDURE
The study procedure involves data measuring three times per day (morning, noon,
afternoon) once a week at designated locations in each building, for daylight,
artificial lignting, temperature and humidity, for a period of two semesters,
plus follow-up measurements for specific data. Photographic pictures are taken
monthly, at least one day, at above listed times. An agreement has been made
with Lawrence Berkeley Laboratory to provide comprehensive information and data
on two case studies (College of Design and Parks Library atria) to be included
in an expanded video-disk of 24,000 images on hypercards, operated by Mac com-
puter with a Panasonic video-disk player. The images will include plans and
diagrams, data sheets, building details, pictures and photographs -including
some time-lapse photographs taken periodically during the research. The picture
series will show both diurnal and seasonal changes in daylighting throughout
the year. The measuring instruments are primarily hand-carried, including a
light meter and probes for recording temperature and humidity. Outdoor condi-
tions are recorded during each measuring session.
RESEARCH RESULTS
Daylighting Information and Data
The College of Design atrium has a wide range of light levels depending on time
and location. Although the vaulted skylight has glare-reducing tinted glass,
fifth floor light levels can be-very high, especially near the ends where the
exterior is of clear glass. The southwest to northeast orientation appears to
be fortunate, although the afternoon direct sun produces glare and discomfort
in affected areas. At Parks Library atrium, the morning sun reaches deep into
the space, noon sun affects the near window areas, but afternoon daylight levels
are reduced to still comfortable levels. Light levels at the Beardshear atrium
are reasonably good at top level, but very low at the ground floor, mainly because
of the old, hard to clean color glass skylights and limited dome window area.
The large windows at Durham building provide much noon light on the third floor,
but less on the ground level. The deep window ledges resulting from thickness
of outside walls reduce direct sun and glare at high sun angle periods. At the
Soil Tilth Laboratory, the south to north atrium receives much light from the
top skylight, but the narrowness of space reduces the amount of daylight that
reaches to the bottom of the stair. The following pages show pictures of the
three major atria (Parks Library, Beardshear, Durham) and daylighting data.
-------
2789
Fig. 2. View into Parks Library atrium, early afternoon
3. Durham Computer Center atrium, Fig. 4. Beardshear Building atrium,
lookingeast. Rotunda, looking north.
-------
2790
5-Dec 12-Jan 19-Jan 26-Jan 2-Feb 23-Feb 2-Mai 23-Mar 30-Mar 6*Apr 13-Apr 20-Apr 27-Apr 11-May
I Qj 1st Floor Center I 5th Floor Center * OUTSIDE I
Fig. 5. College of Design indoor vs. outdoor light levels, 1990,
160
r 2500
140 |
• 2000
D 120 -
100 4-
1500
- iooo R
•• 500 F
20 -
4-Sep 11-Sep 18-Sep 27-Sep 2-Ocl 9-Oct 16-Oct 1-Nov 8-Nov 15-Nov 29-Nov 5-Dec 18-Dec
I C DURHAM
¦ BEARDSHEAR ¦ PARKS
» OUTSIDE
Fig. 6. Main floor light levels in three atria, noon, fall 1990.
4-Sep 11-Sep 18-Sep 27-Sep 2-Oct 9-Ocl 16-Oct 1-Nov 8-N'ov 15-Nov 29-Nov 5-Dec 18-Dec
| ~ DURHAM ¦ BEARDSHEAR ¦ PARKS » OUTSIDE |
Fig. 7. Top floor light levels in three atria, noon, fall 1990.
-------
2791
Temperature and Humidity Information and Data
All of the project atria are centrally heated and cooled. The Iowa State Power
Plant provides steam directly to Beardshear, and through heat exchangers to hot
water systems in the other four buildings. Chilled water is similarly provided,
but the Beardshear atrium uses primarily the natural system. At the College of
Design, the atrium functions as a large return air plenum from offices and stu-
dios, with return air grills at the ends of the main floor to the recirculation
equipment. During the heating season, temperatures fluctuate relatively little
daily and between bottom and top floors, but during the cooling season there
is a wider range of differences, particularly at the College of Design (75-90°F)
and at Beardshear (80-90°F). During the hottest days, the indoor temperature
in these two atria nears the outdoor temperature in top floor, amounting to 10°F
temperature difference between top and bottom floors at College of Design, and
similarly in Beardshear where the exterior doors on the ground floor and first
floor may remain open all day. However, indoor temperatures are somewhat cooled
in the atria by the cooled air entering from the other rooms. Seasonal changes
are greater, as temperatures from 65°F to 98°F have been recorded in the atria.
100
U -i i : '
4-Sep 11-Sep 18-Sep 27-Sep 2-Ocl 9-Oct 16-Oct 1-Nov 8-Nov 15-Nov 29-Nov 5-Dec lS-Dec
O PARKS * DURHAM ~ BEARDSHEAR — OUTSIDE
Fig. 8. Noon temperatures recorded in three atria, fall 1990.
100
4-Sep 11-Sep lS-Sep 27-Sep 2-Oct 9-Oct 16-Oct 1-Nov 8-Nov 15-Nov 29-Nov 5-Dec 18-Dec
— OUTSIDE * BEARDSHEAR ~ DURHAM O PARKS
Fig. 9. Noon relative humidities recorded in atria, fall 1990.
-------
2792
The indoor humidity conditions in the atria are generally within the comfort
range, except on some rainy days when it may be excessive.
Air Stratification Information and Data
Air stratification occurs particularly in the College of Design, above the 4th
floor, as the vaulted skylight does not provide enough "loft" during the cooling
season. Similarly, the dome above Beardshear Rotunda does not have enough vol-
ume to prevent stratification of air on the 4th floor. To a lesser degree, the
same situation applies to Parks Library, as the 4th floor is open to the atrium.
In Durham, the height of the top floor seems to be adequate to avoid stratifica-
tion, and at Soil Tilth Laboratory the cooling system seems to take care of it.
CONCLUSIONS
Atria are complex spaces where direct sun light may cause glare, temperature
rise and discomfort to the occupants, while daylighting in general is beneficial.
Relative humidity indoors usually follows the outdoors fluctuation, except that
airconditioning may cause a reversal. Air stratification is a problem in high
atria, but could be mitigated by adequate "loft" space* or appropriate exhaust
and recirculation of air. The orientation is important in order to provide most
daylighting for the periods when the atria are occupied, and to avoid glare and
discomfort caused by direct sun radiation. The shape and size are also impor-
tant, especially to avoid air stratification. Temperature and humidity controls
can be improved by appropriate HVAC system design.
In the examples of atrium design provided, research results show that a number
of problems exist. The orientation of the College of Design atrium is good for
daylighting, but excessive direct sun radiation needs to be reduced by louvers
at the southwest end windows. Air stratification could be reduced by adequate
air circulation by providing a "duct space" with an inside window wall at the
edge of atrium platforms at each end. The Parks Library atrium allows much sun
penetration in mornings, and some direct sun areas at noon, but lesser daylight
on afternoons, avoiding direct sun radiation. The Beardshear atrium has only
top skylights and small dome windows, and the central stairs in the atrium ends
prevent much of the light reaching lower floors. The skylights also need to be
cleaned and improved in order to increase daylighting. The Durham atrium seems
to serve well, unless the 3rd level space is put into a use that is affected by
direct sun radiation. At the Soil Tilth Laboratory, the south-north skylight
provides ample light and its function as a stairwell reduces occupant problems.
REFERENCES
Kainlauri, E.O., T.H. Hielkema, and R.E. Spears (1988). Research on the perform-
ance of a passive solar atrium. In Proceedings of the 13th National Passive
Solar Conference, by the American Solar Energy Society. Cambridge, pp. 127-
132.
Kainlauri, E.O., T.H. Hielkema, M.L. Jackson, and R.E Spears (1988). Atrium re-
search for the whole building performance. In Proceedings of the Symposium on
Performance of Buildings and Serviceability of Facilities, ASTM STP 1029.
Kainlauri, E.O., and R.E. Spears (1989). Interfacing of daylighting and artifi-
cial lighting in an atrium. In Proceedings of the 14th National Passive Solar
Conference, by the American Solar Energy Society. Denver, pp. 169-174.
-------
2793
INTERNATIONAL ATRIUM RESEARCH
Harry T. Gordon. AIA
Burt Hill Kosar Rittclmann Associates
Washington, I)C, USA
Ron Kammerud
Lawrence Berkeley Laboratory
Berkeley, CA, USA
Anne Grete Hestnes
Department of Architecture
Trondheim, Norway
(Working Group Leader)
ABSTRACT
International Energy Agency Task XI of the Solar Heating and Cooling Implementing
Agreement, is a cooperative international effort that is studying Passive and Hybrid Solar
Applications in Commercial Buildings. One focus of this research is the design of atrium
spaces that use passive solar and energy conservation features. Building measurements and
computer simulations demonstrate that atrium spaces can be designed to reduce heating,
cooling and lighting energy requirements in a range of climates from northern Norway to
the Southern U.S.
KEY WORDS
Atrium; passive solar; energy conservation; international; calibrated computer simulations;
building measurements; daylighting
INTRODUCTION
The authors are participants in a working group within IEA Task XI that is developing
design recommendations for atrium spaces that use passive solar and energy conservation
techniques. Other countries participating in this atrium working group include Sweden,
Finland, the United Kingdom and Germany.
The Atrium Working Group of IEA XI is developing conclusions from a four year effort
examining atrium design approaches and energy use in climates ranging from northern
Norway to the southern United States. The basis of our research is monitored energy
performance in eight atrium buildings, supplemented by detailed simulation of design
alternatives using a calibrated computer modeling technique developed by LBL. The
monitoring allows us to examine specific aspects of building performance, and the
simulations are used to extend these results to other climates and design configurations.
-------
2794
The monitored atrium buildings studied in Task XI house offices, retail, multi-family
residential, educational, and daycare functions. These buildings range in size from 490 M2
to 40,000 M2. Atrium configurations include Core, Linear, Integrated and Envelope (Ref
1). Although the monitoring protocols vary by country, the data includes heating
performance of the atrium and the adjacent building spaces, lighting and daylighting, and
ventilation cooling effects during overheating periods.
MEASURED PERFORMANCE RESULTS IN ATRIUM BUILDINGS
The monitoring approach in each atrium building was chosen to investigate the most
important aspects of that particular atrium design. Although the findings are not directly
comparable, the measurements do permit important conclusions to be drawn:
• In most climates, heating the atrium to a temperature in the range of 10-15"C
resulted in little or no increase in overall building heating energy requirements.
The heal put into the atrium partially offsets the heating requirements of the
adjacent spaces.
• In most northern climates, no mechanical cooling system is required to maintain
atrium coipfort. Ventilation of the atrium with outside air is sufficient to maintain
reasonable comfort; the rate of air change in those spaces that were measured is in
the range of 5-10 air changes per hour during summer days.
• Although daylighting of the atrium resulted in lighting energy savings in many
buildings, few took advantage of the potential to save energy in adjacent spaces,
even if there was sufficient daylight available.
CALIBRATED COMPUTER MODELING
A calibrated computer modeling approach was used to examine the energy and comfort
performance of a large linear atrium space at the ELA building of the Norwegian Institute
of Technology in Trondheim, Norway. In this approach, monitored energy performance
of the building was compared with simulation results for the building. The intent of this
"calibration" is to ensure that the simulation provides a faithful thermodynamic
representation of the actual building.
The simulations were performed with BLAST using TRY climate data for Oslo, modified
to reflect the long-term average difference between Oslo and Trondheim weather.
Agreement between the simulations and monitored data was achieved to within limits
determined by the differences between long-term average weather data and actual climate
conditions during the monitoring period.
Calibration of the simulation model increases confidence that the model is an accurate
representation of the building, i.e., that its design and operation are correctly described in
the simulation. The calibrated model for the ELA building in Trondheim serves as a
benchmark from which parametric studies can be used to reliably estimate the impacts of
changes in design and operating parameters and in climate location. In the present case,
the performance of the linear atrium building was examined for the climates of
Washington, DC; Fort Worth, Texas; Zurich, Switzerland; Oslo, Norway; and Rome, Italy.
The effect on energy performance was examined for (1) design parameters, such as glazed
area; (2) operating parameters, such as atrium heating set points; and (3) materials
properties, such as thermal resistance of the atrium glazing. In many of these cases.
-------
2795
comfort conditions in the atrium were also estimated using the Fanger thermal comfort
model in BLAS T. In these comfort studies, the activity level of the occupants (metabolic
rate) and their clothing level (CLO) were varied to estimate comfort conditions under a
range of possible atrium functions.
KLY FINDINGS
Although the majority of existing atrium buildings are not designed to do so. careful
atrium design ,and operation can result in signit icant energy savings in all climates. Most
energy-saving atria are allowed to float freely at some times of the year, and are
conditioned (heated or cooled mechanically) al oiher times. As illustrated in the figure
below, the relationship between the atrium and the adjacent spaces varies at different
times of the year:
• During mid-winter, atrium spaces provide a buffer from the climate tor the
adjacent heated spaces.
• In swing seasons, the atrium has the potential to contribute useful heat to adjacent
spaces, if the adjacent spaces require it
• In summer, the atrium (and sometimes the adjacent spaces) need to be cooled;
frequently this can be accomplished by natural or fan-assisted ventilation
*
o
-BLD6. HEAT LOSS
-MET ATRIUM D4ER6-Y.
(SOLAR &MM-HEAT 10$
-ATRIUM HAS POTENTIAL TO-
CONTMBUTE USEFUL HE ACT
TO ADJACENT SPACED
III)-):
J FMAMJ J ASOND
PEFJOD IW WHICH ATRIUM SolAP-
&UFFEK.
6A1M EXCEEDS AmiUM HEAT LD%
COMTftfr.
E*£E46 [COMTRIB.
BUFFER
fbTENTlAL.'i tOLAfc fpbTtvmAU'
Fig. 1.
RELATIONSHIP BETWEEN ATRIUM AND ADJACENT SPACES
-------
2796
Glazing properties (type, orientation, shading, etc.) are one of the most important factors
in determining building heating/cooling energy use.
• During the heating season, the 11-value of the atrium glazing is a more significant
factor than the solar transmission. Double low-e glazing is especially effective in
northern latitudes
Linear atria perform better if thev are not symmetrically glazed. In Trondheim,
Norway, making the north facing glazing opaque reduces atrium heating energy
use by about 25%. In Washington, DC making the south glazing opaque reduces
atrium heating energy use by about 25%, and cooling requirements by about 36%
/\
Hfc I
A B
TRONDHEIM
ATRium
MT&
tHEWt
%
in
A B A E>
Fig. 2. WASH. D.C.
In the linear atrium at Trondheim, Norway, atrium temperature set points between 5"C and
15"C appear to have a minor impact on overall building heating energy use.
Within this range, the increase in atrium temperatures improves the buffering
effects on adjacent offices, reducing their heating requirements
Above 15°C, higher temperatures resulted in substantial increases in atrium heating
energy requirements. An increase from an atrium set point of 15"C to 18°C resulted
in a 25% increase in total building heating energy requirements. In Washington,
DC, the result is similar
Tempering the atrium also reduces the size of required heating units in adjacent
off ices, reducing construction costs.

350

I
300
2
250
O

P
i
200
3
150
s

100
t
50

$
0
University Building, Trondheim
TOTAL
...AT.RI.UM...
OFFI9E
10 15
Atrium Heating Setpoint Temperature [*C]
Fig. 3.

350
.e

300
dl

250
t
?m
*

I
150
a
100

50
ft

0
University Building, Washington, DC
TOTAL
ATRtmi"
OFFICE
10 15
Atrium Heating Setpoint Temperature fC]
-------
2797
Even in mid-winter, unconditioned atria can be acceptably comfortable depending on
clothing value and metabolic rate of the occupants.
• If the atrium is used for an active function such as circulation, an occupant will be
comfortable with a light jacket even in Oslo or Zurich.
• If the atrium is used for more sedentary activity, such as seating, comfort
conditions mav not be acceptable, and spot heating (such as localized radiant
heating in sealing areas) should be considered
• In practice, comfort will be greater if the people are able to sit in a sunlit area
University Building, Zurich, Winter University Building, Oslo, Winter
Afrium Mnt U»tul / . .. ®'_ '.
Atrium Not Heated
¦Wilting with i Ixkct
•-wintiffj-wMout i jia-a"
•Sc3:«d with t Jacket
¦Strcd without t Jicta
-Wjlkisj w.:h i ixket
-_Se»ted »iih t Jjcket
- S
-------
2798
The atrium can also be used to provide cooling f or the adjacent building' spaces.
• A large office building near L.ondon, UK was designed to use the atrium to act as
a thermal chimney drawing ventilation air through the adjacent offices; this
resulted in the elimination of the mechanical cooling system and thus a building
that was less expensive to build than one with mechanical cooling, but no atrium.
Q

tz <13

<=S

C=2

<=3
ATRIUM USED TO ASSIST NATURAL VENTILATION
Fig. 6.
Daylighting of offices adjacent to the atrium can result in significant lighting energy
savings, especially on the upper floors.
• Measured and simulated data from a linear atrium in Trondheim, Norway shows a
potential annual lighting energy savings on the lop floor of 50% (based on 500 lux
during office occupancy hours)
If the same building was built in Dallas. Texas, the potential annual lighting
energy savings increases to 80%
REFERENCES
U.S. Participation in International Atrium Research, Gordon, et.al. (l'lth Passive
Solar Conference, June 1989)
ACKNOWLEDGEMENTS
The research work described in this paper was partially supported by the IJ.S Department
of Knergy.
-------
2799
PASSIVE SOLAR ENERGY AND NATURAL DAYLIGHT IN
OFFICE BUILDINGS
Poul E. Kristensen and Torben Esbensen
ESBENSEN, Consulting Engineers FIDIC
41 Havnegade
DK-1058 Copenhagen K
Denmark
ABSTRACT
A 4,500 m2 three-storey office block has been designed with special emphasis on utilization of
passive solar energy and natural daylight. Furthermore maximum visual and thermal comfort
were prime design objectives. The project was awarded the first prize in the architectural ideas
competition "Working in the City", which was organized by the Commission of the European
Communities CEC Departement Generate XII in 1988/1989.
KEYWORDS
Passive solar energy; daylight utilization; lighting control; natural ventilation; commercial
buildings.
Fig. 1. Model picture of the 4,500 m2 office building.
-------
2800
INTRODUCTION
In 1988 the CEC launched an architectural ideas competition for the design of non - domestic
buildings. The main idea behind the competition was to encourage architects to consider
utilization of passive solar energy and daylighting strategies as prime design objectives. Detailed
information about the competition is available in references (1), (2), and (3).
This project was designed by a team from Copenhagen, Denmark consisting of KHR A | S
Architects & Planners, ISLEF Building and Construction Compagny, Cenergia ApS and
ESBENSEN, Consulting Engineers FIDIC.
In offices, the energy use profile differs considerably from that of dwellings. Electricity use
for lighting and ventilation/cooling account for a large part of the total energy use. This is
illustrated in figure 2, which is taken from the competition material, see Reference (1).
^ Primary energy use pr. m2 facade
\ 'V®3'
total
200
heating &
lighting
- heating
cooling
lighting
so
0
Glazing ratio of facade
Fig. 2. LT chart for offices, climate Northern Europe,
East/West orientation, double glazing.
In fig. 2 the primary energy consumption (kWh/m2-year) for heating, cooling and lighting is
shown as a function of the transparent window fraction of the total facade area. The chart is
valid for the passive zones stretching 6 meters into the building from the facade.
The energy consumption is expressed as primary energy consumption kWh/irf-year, i.e. the
equivalent use of fossil fuel for producing heat and electricity. The total efficiency factor
anticipated for electricity production is 27%, and for thermal energy 67%.
The charts were developed by Dr. Nick Baker from the Martin Centre, Cambrigde University
GB, and Dr. Marc Fontoynont from Lyon, using detailed computer tools. For explanation of the
assumptions behind the LT calculation method, see References (1) and (5).
It is seen, that the potential for saving on electricity for lighting is very large, provided that a
lighting control system is installed, this beeing a prerequisite for the chart. The optimal
transparent window area for the Northern European climate is seen to be around 40%.
-------
2801
BUILDING DESIGN
The three-storey building has a total floor area of 4,500 m2. The building is organized with
three office blocks, which are connected with twp glaze atria. Each office has- a glazed
staircase room. To the South-West there is a glazed archade, which is situated in front of a
double high canteen in the ground floor and a meeting room in the upper floor.

AThIUM
ELELJ LMI^
~
MEETING ROOM/
CANTEEN
GLAZEO ARCHAOE
GSR* : GLAZED STAIRCASE ROOM
t- riTii'j
Fig. 3. Floor plan.
Ventilation is provided via a displacement ventilation system, which injects preheated fresh air
centrally in the building. Exhaust air is drawn out of the building via roof louvres in the three-
storey atria, utilizing the chimney effect as driving force. During wintertime, the roof louvres
are closed, and the atria acts as buffer spaces. Ventilation air is then drawn from the atria via
a fan-driven heat recuperation unit.
-J"
OFFICE .
OFFICE
HEAT
RECUPERATION
f
tec
V-
fc

Fig. 4. Section through atrium and adjacent offices.
-------
2802
In this building, the atria serve a triple purpose:they provide driving force for natural
ventilation, serve as buffer spaces to exploit passive solar energy, and finally the atria provide
natural daylight deep into the building.
The window area in the exterior facades is 30%. The window area in the upper part of the
atria walls is 40%, increasing to 60% at the ground floor. In this way an even distribution of
natural daylight is allowed in the offices, irrespective of whether the office is orientated towards
the exterior or towards the atria.
DAYLIGHT ANALYSIS
In order to analyse the daylighting qualities of the design, a 1:20 model of the part of the
building was built. In fig. 5 the monitoring data are illustrated for the first floor. The daylight
effenciency is characterized by the daylight factor = measured light level indoors divided by
measured light level outside the building on an overcast day.
The general light level required for offices in Denmark is 200 LUX. An ambient light level of
9,000 LUX is typical for an overcast day. It is seen, that on such a day, the 200 LUX threshold
is achieved in areas where the daylight factor exceeds 200 LUX/9,000 LUX = 2.2%. This is
seen to be the case in most parts of the offices.
ambient
ambient
atrium
r « » « « « *
x>:x:x>I'XvIvI'.V.t Measured daylight factor
6 ; 3 I 1 o
I ! Corresponding indoor LUX
- ^ ^ o level by 9,000 LUX
(overcast) outside
Fig. 5 Measured daylight factors in a 1:20 scale model.
The 500 LUX level for the desks requires a daylight factor larger than 5.6 to satisfy the
requirement on an overcast day. This threshold is exceeded for almost all desk positions.
-------
2803
It can be concluded that natural daylight will cover the total light requirement in most days,
and artificial lighting will only be needed on very dark days and during the winter days when
the sun is not up.
Based on the model measurements, and the CIE curves, reference (4), the annual savings on
electricity for general office lighting is calculated to 76% of the lighting required during working
hours 09.00 - 19.00.
ENERGY OPTIMIZATION
One of the prime objectives was to design a building with a substantially lower overall energy
consumption, when compared to traditional office buildings. In order to meet this objective,
various designs were analyzed using the LT Calculation method, which was provided as part of
the competition material.
The main steps of the optimization process are illustrated in figure 6. Here the primary energy
consumption (i.e. fuel used for electricity production, etc) is shown, seperated in auxilliary
energy for heating, ventilation/cooling and lighting. For a more detailed discussion of the
assumptions and calculations, see reference (4).
Primary energy consumption
MWh/year
800 - -
600 - -
400 - -
200--
lighting
cooling/ventilation
space heating
A. Traditional building with 20% of the
facade glazed.
B: Actual building with 30% of the facade
glazed, glazed stairwells etc., but without
glazed atria.
C: As "B", but with glazed atria.
D: As "C", but with lighting control
system and with actual fan assisted
natural ventilation system.
Fig. 6. Reduction of energy consumption through design for a 4,500 m2 three-storey office
building. Danish weather data.
A traditional, yet modern 4,500 m2 office building would have a total primary energy
consumption as indicated, almost evenly distributed between energy for lighting and energy for
heating. Cooling load is very small, since the summer months are cool in Denmark.
In case B the actual building shape is used, see fig. 2, with larger window area.Open spaces
between the office blocks allow daylighting into the building, but theese atria rooms are not
glazed. Lighting load is decreased, but heating load is increased due to the larger exterior
facade.
-------
2804
In case C, the three atria are now glazed, and the impact of this on the heating load is seen to
be very significant. In case D, an automatic control system on artificial lighting is added, and
the fan-assisted natural ventilation system is taken into consideration.
Now the potential for utilizing natural daylight is really exploited, with a reduction of almost
75 % on the electricity load for lighting. It is noted, that the heating load is increased compared
to the "C" - case, since there is now not so much "free heat gains" from artificial lighting to
offset heating load. However the total energy consumption is decreased, and of evenly great
importance, the electricity part of the total auxilliary energy needed is now very small.
CONCLUSIONS
All working areas and circulation areas receive ample natural daylight, for the benefit of good
visual comfort and a stimulating working environment.
Electricity for lighting is reduced to only 20% compared to a traditional office building. The
total primary energy consumption for the building is reduced to 50%. Environmental pollution
is consequently also reduced by 50%. Energy efficient building design can make an important
contribution to improvement of our common environment.
REFERENCES
Competition material for the CEC competition "Working in the City"
Available from Owen Lewis, Scool of Architecture, University College Dublin.
1988.
Research Digest number 3
Available from : see (1)
March 1989
Research Digest number 4
Available from : see (1)
August 1989
Commission Internationale de 1'Eclairage.
Publication No 40, CIE Central Bureau, Kegelgasse 27, A. 1030 Vienna.
Austria
Working in the City.
Selected entries from the 1989 CEC Architectural ideas competition.
Edited by Shane O'Toole and Owen Lewis
School of Architecture, Dublin Ireland.
-------
2805
EFFECTIVE TOP-GLAZING AND INTERNAL WALL AREA FOR EFFICIENT
DAYLIGHTING IN ATRIA
Morad R. Atif, Ph.D Candidate and Lester L. Boyer, Professor
College of Architecture, Texas A&M University,
College Station, TX 77843, U.S. A.
ABSTRACT
The objectives of this paper are: a) to determine the effective top-glazing area for daylighting in
atria, and b) to study the impact of the effective internal reflectance on the daylighting potential of
atria. Physical scale models of atria were built and tested in a sky simulator under clear and
overcast conditions. About 22 illumination levels were measured for 100 tests, in four and two-
story atria (linear and square). The two key variables were; a) horizontal and vertical south-facing
top-glazing area, and b) the proportion and the reflectance of the solid area of the atrium walls. The
measured Daylight Factors (DF) were compared to the daylight availability of selected cities, for
plant lighting requirements. The DF decreased with the increase of the Well Index and/or with the
decrease of the top-fenestration transmittance. The highly reflective vertical south-facing sawtooth
configuration helped in reflecting light into the atrium floor, but provided low daylight contribution
in the adjacent spaces under overcast sky, and required additional light from the sun in clear sky.
The optimum horizontal top-fenestration area was reduced about 50% with the highest effective
internal reflectance of the walls. This would enhance the daylighting value of an atrium in hot
climates by reducing the solar load.
KEYWORDS
Atrium buildings; daylighting; lighting; fenestration control; glazing; sunlight; color and reflectance;
environmental control systems; optimum design
BACKGROUND AND PROBLEM STATEMENT
The strongest motivation for atrium design relies on its role as a metaphor for the exterior (Navvab
and Selkowitz, 1984). The as-built atria are found either underlit for solar control, requiring
artificial lighting, or overlit with excessive amounts of cooling loads, especially in hot climates.
These problems have their origin in the lack of tools for effective top-glazing area which, if
oversized, causes a tremendous amount of cooling loads (Gillette and Treado, 1988). Moreover,
studies show that mass in the atrium walls can absorb solar radiation, caused by the esthetically-
required sunlight strategy (Utzinger and Bochek, 1985). In return, this solid area in the internal
walls has its effects on the daylighting distribution potential inside an atrium.
Shallow four-sided atria are the best candidates for an effective daylighting performance without
excessive cooling loads. In promenade atria where sunlight is often desired on the atrium walls
rather than on the atrium floor, designers often opt for vertical south-facing top-glazing
configurations for reduction of solar heat gain. However, the daylighting performance of such
fenestration is not well documented. Horizontal and vertical south-facing top-glazing
configurations are somewhat opposite design choices in a daylighting (and thermal) strategy, and an
estimation of their optimum areas could assist designers in sizing atrium top-glazing areas at the
preliminary design stage. The optimum top-glazing area is related not only to the top-glazing light
transmittance but also to other physical atrium parameters such as the percentage and reflectance of
solid area in the internal walls, that affect the daylighting behavior of atria.
-------
2806
South
Anglc=45
\\WW\N,
100%-opcn
Transmittancc=
28.0%
50%-open
Transmittance=
22.0%
Fig. 1. Schematic sections of the atrium
models showing examples of
top-glazing areas tested for the
vertical south-facing configuration
10 ft
H—~—
~ ~ ~ ~ o ~
~~~~~~
~ a ~ a a ~
Solid area= 66.7
Solid area=27.8%
Fig. 2. Schematic sections of the atrium
models showing examples of
different wall configurations tested
for the four-story linear atria
OBJECTIVES AND SIGNIFICANCE
The paper focuses upon the daylighting performance of atria with respect to two atrium physical
parameters: top-glazing area and internal reflectance of atrium walls (i.e., reflectance and percentage
of solid area). The premise of this work has been previously documented (Atif and Boyer, 1990).
The objectives of this paper are: a) to identify the effect of top-glazing configuration area on meeting
the illumination criteria at the atrium floor, and on the daylighting contribution in the occupied
spaces, and b) to determine the effect of the proportion and reflectance of solid area in the atrium
walls on the daylighting performance of atria. TTie daylighting value of an atrium will be enhanced
with an optimum fenestration, especially in hot climates.
METHODOLOGY
Description and Settings
Illumination measurements were collected in physical scale models in the sky simulator at Texas
A&M University, using an Illumination Data Acquisition System (IDAS). The IDAS consists of a
computer, serial analog module, and eight sensors. Each sensor has a unique calibration system
stored in the program disk. The scale of the model was l/24th of full size. The floor of the models
had a reflectance of 15 per cent (%). The illumination levels were collected for 100 tests, which
included variations in sky conditions, and in atrium physical parameters. The sky simulator was set
for both clear and overcast sky. The completely overcast sky, i.e., the C.I.E. (Commision
Internationale dEclairage) sky, had a zenith illuminance that is three times that of the horizontal
illuminance. The other widely accepted formulation for clear sky is that horizon illuminance is
about three times that of the zenith (Stein, Reynolds and McGuinness, 1986).
Variations in the atrium physical parameters included atrium proportions, top-fenestration, and
internal atrium walls. The study included four types of atria; two-story and four-story atria were
tested with both linear and square configurations (40x120x30 ft, 40x40x30 ft, 40x120x54 ft, and
40x40x54 ft) (12x36x9 m, 12x12x9 m, 12x36x16 m, 12x12x16 m). This translates into four Well
Indexes (WI): .50, .75, .90, and 1.35 in that order. The WI describes the well efficacy of a
skylighteid room and can be calculated as shown in Equation 1 (Boyer and Kim 1988; Saxon,
1983).
-------
2807
Well Index (Wl)=Well Height *(Well Width+Well Length)/2*Well Width*Well Length (1)
The second level of variation included the top-fenestration where two configurations were tested:
horizontal and vertical south-facing, each with three alternative areas: 37,50, and 100% of the total
horizontal area. All results assume clear glass with a transmittance of 100%. Figure 1 shows
examples of vertical south-facing top-glazing areas. The third level of variation included the
internal walls, represented by the reflectance of the solid area, and by the solid-to-total wall area
ratio. These ratios were 30 and 60% for the two-story atria, and 27 and 67% for the four-story
atria. Figure 2 shows the different wall configurations tested in the four-story square atria. The
internal reflectance values of the solid walls were 25,40, and 90%.
Data Collection and Evaluation Tools
Horizontal and vertical measurements were collected in the model for each test. Horizontal
measurements included: a) 9 test points on a work plane height at the atrium floor, distributed
through a grid of 20 by 20 ft (6 by 6 m) for the square atria and a grid of 20 by 60 ft (6 by 18 m)
for the linear atria, and b) 5 test points just below the top-fenestration. Vertical measurements (i.e.,
perpendicular to the center of each atrium wall) were collected for the second and the top
floor. The data were collected in footcandles (fc) (lux), and then transformed into Daylight Factors
(DF). The DF was used to assess the daylighting performance. Average DF's for horizontal and
vertical measurements (at each floor) were calculated for each test.
The average DF's were compared to the outdoor availability for three locations: Lake Charles, LA,
Phoenix, AZ, and Fresno, CA (Robbins, 1986). The transmittance of each top-glazing
configuration was calculated by averaging all the DF's measured just below the top-fenestration.
Legends in Fig. 3 illustrate the transmittance of top-glazing configurations tested. The transmittance
ranged from 19% to 85%. The internal wall effective reflectance was calculated by dividing the
sum of the products of each wall component area and its reflectance, by the total wall area.
Examples of the calculated effective reflectance for each wall configuration are illustrated in the
legends of Fig. 7. These values varied from 17.8% to 65.0%. The openings in the atrium walls
were simulated by a surface with a reflectance of 15%. The target illumination criteria at the atrium
floor was dictated by the lighting requirements for plants, which need a minimum of 92 fc (1,000
lux) for at least 12 hours a day (Mpelkas, 1987; Navvab and Selkowitz 1984).
RESULTS AND ANALYSIS
Effect of the Top-glaring Configuration
The average horizontal DF at the atrium floor decreased with the decrease of the top-glazing
transmittance, and/or with the increase of the WI (narrower and/or deeper atria). Figures 3 and 4
show the variation of the average horizontal DF as a function of the WI and top-glazing
configuration, in clear and overcast sky respectively. The horizontal DF profiles with horizontal
and vertical south-facing top-glazing configuration were similar, except for the magnitude. The
horizontal DF decreased about 85% of its original value in clear sky with no sun when the
transmittance dropped from 85.0 to 19.0%. The horizontal DF profiles with the 100%-open
vertical south-facing top-glazing indicate the role of a sawtooth with a highly reflective internal
surface in reflecting the light toward the atrium floor in clear sky.
The sky condition had an effect on the DF profile. The average horizontal DFs with horizontal top-
glazing fenestration were 10% higher in overcast sky than in clear sky. This was due to the
relatively larger amount of illuminance at the zenith. However, the horizontal DF's with vertical
south-facing top-glazing configuration were just slightly higher in clear sky. This is because both
surfaces of the sawtooth have combined in compensating for the relatively small amount of direct
light penetration with a sawtooth in overcast sky.
The vertical DF in overcast sky was usually about 50 to 60% higher than that in clear sky with no
sun. This is because of two reasons; a) clear sky (with no sun or with low sun angles) provide less
direct light than overhead light, and b) the light reaching the test points in clear sky is likely to be
subject to more reflections than that in overcast sky. Figures 5 and 6 show the average
-------
2808
DF (%)
501
40.
Vacation of iop-giazing
Tvne Areaf%") Transm.
¦ Horiz 100.0
B Horiz. 37.5
B S.Vert. 100.0
^ S.Vert.. 50.0
S S.Vert. 37.0
.75 .90
Well Index
Fig. 3. Measured average horizontal
Daylight Factor (DF) at the atrium
floor as a function of the Well Index
and the effective light transmission of
the top-glazing configuration, clear
diffuse sky
.75 .90
Well Index
Fig. 4. Measured average horizontal
Daylight Factor (DF) at the atrium
floor as a function of the Well Index
and the effective light transmission of
the top-glazing configuration,
overcast sky
DF (%)
50n
4C
30-
Variation of ton-glazing
Type Area(%') Transm. (%)
I Horiz 100.0
i Horiz. 37.5
E S.Vert. 100.0
. S S.Vert.. 50.0
^ S.Vert. 37.0
.75 .90
Well Index
Fig. 5. Measured average vertical Daylight
Factor (DF) on the atrium wall as a
function of the Well Index, and the
effective light transmission of the top-
glazing configuration, in clear diffuse
sky (2nd floor of two and four-story
atria)
DF (%)
50
775 .90
WeU Index
1.35
Fig. 6. Measured average vertical Daylight
Factor (DF) on the atrium wall as a
function of the Well Index, and the
effective light transmission of the top-
glazing configuration, in overcast
sky (2nd floor of two and four-story
atria)
-------
2809
measured vertical DF as a function of the Well Index and top-glazing transmittance, in clear and
overcast sky respectively.
Davlighting Performance as Related to Lighting Criteria
Daylight availability data were taken for March 21st and December 21st at noon. In clear sky with
no sun on March 21st,the minimum horizontal DFs at the atrium floor were about 10.5% for all
selected cities. This target of 10. 5% can be achieved in two-story atria with a top-fenestration
transmittance higher than 28%, and in four-story atria with a transmittance higher than 65% and
80% for linear and square atria respectively. For example, the average horizontal illumination on
the atrium floor of four-story atria with 100%-open sawtooth at the top in clear sky with no sun
would be around 52 fc (565 lux) for the linear atria, and 27 fc (294 lux) for the square atria.
In the absence of sun, or in the case of a north-oriented sawtooth, supplementary artificial light is
likely to be needed to maintain tall plants. The horizontal light from the direct sun of a clear sky on
March 21st at noon is 7.5 to 8.5 times higher than that from the diffuse sky for the selected cities.
Sunlight availability is evidently the highest in Phoenix, which should require a smaller top-
fenestration transmittance than Fresno and Lake Charles.
Under overcast sky on December at noon, the minimum horizontal DF's required are 8.1,8.3, and
10.0% for Lake Charles, Phoenix, and Fresno respectively. The lighting criteria can be met with a
top-fenestration higher than 22% for two-story atria, and with a top-fenestration transmittance
higher than 28% for four-story atria. However, the minimum criteria should be higher, due to the
limited time of daylight hours on December 21 sl The data for overcast conditions are more
important to the city of Lake Charles, due to the dominance of cloudy sky conditions.
DF
30-
20
10
0 20 40 60 80 100
Overall transmittance of the top-glazing (%)
Fig. 7. Average measured Daylight Factor (DF) versus measured overall top-glazing
transmittance of four-story square atria in clear sky, as a function of the
effective internal wall reflectance (reflectance and percent of solid area)
Variation in the atrium walls
Solid area (%) Reflect, of solid area (%) Effect, reflect
Not shown
Not shown
-------
2810
The daylighting contribution in the occupied spaces was small under overcast conditions, especially
in higher latitudes. Figure 6 shows that the average vertical illumination in the second floor of
four-story linear atria with a 100%-open horizontal top-fenestration would be equal to 112 fc (1,220
lux) in Phoenix, and 93 fc in Fresno. With a 100%-open sawtooth configuration, these values
decreased to 36 fc (390 lux) in Phoenix, and to 14 fc (152 lux) in Fresno.
Effective Internal Reflectance and Optimum Top-glazing Area
The horizontal and vertical DF increased with the increase of the effective reflectance of the internal
walls. Figure 7 shows the effect of the effective wall reflectance and the top-fenestration
transmittance on the horizontal DF profile in a four-story square atrium under a clear sky. The
horizontal DF increased about 40% of its value when the internal effective reflectance increased
from 17.0 to 65.0%. This is very significant for the sizing and the configuration of the top-
fenestration for efficient daylighting. The increase of the internal effective reflectance of the walls
can contribute to the reduction of the top-fenestration area needed for daylighting, thus reducing
solar heat gain in hot climates. A target DF of 10% can be reached by a top-fenestration
transmittance of 73% with the effective internal reflectance of the walls equal to 73%, and only by a
transmittance of 58% with the effective reflectance raised to 65%. In terms of design and energy
calculation, this translates into a reduction of more than 50% of the horizontal top-fenestration area
of a four-story square atrium.
CONCLUSION
Data on atria show that the top-fenestration area, and the proportion and color of the solid area of
the internal walls affect a great deal the daylighting potential of low-story four-sided atria.Whenever
possible, rectangular-shaped atria are preferred to square atria for daylighting. In general, sunlight
is needed when the top-fenestration transmittance is lower than 55% in clear sky. Therefore,
vertical south-facing configurations are always in need for the sun component to meet the
illumination criteria. A highly reflective sawtooth, and highly reflective walls will reduce the
optimum top-fenestration area for effective daylighting. This in turn will reduce the solar heat
gain, thus enhancing the daylighting potential of atria in hot climates. However, vertical
illuminations with low top-fenestration areas are very low at lower floors in overcast conditions
This will limit the daylighting contribution in the adjacent occupied spaces in regions with
predominant cloudy conditions. The paper did not investigate the effect of the different atrium
physical parameters of the study on other architectural issues, such as glare and perception.
REFERENCES
Atif, M. R., and Boyer, L. L. (1990). Impact of top-glazing and atrium surface area on
daylighting and cooling in hot climates. Proc. 15th Natl. Passive Solar Conf.. Austin, TX, pp.
169-174.
Boyer, L. L„ and Kim, K. (1988). Empirically based algorithms for preliminary prediction of
daylight performance in top-lighted atriums. ASHRAE Transactions. Part 1, pp. 765-781.
Gillette, G., and Treado, S. (1988). The daylighting and thermal performance of roof glazing
ASHRAE Transactions. Part 1, pp. 826-835.
Mpelkas, C. C. (1987). Indoor landscaping for healthy, beautiful workplaces. Architectural
Lighting. February , pp. 42-45.
Navvab, M„ and Selkowitz, S. (1984). Daylighting data for atrium design. Proc. 9th Natl.
Passive Solar Conf.. Amer. Solar Energy Soc., Columbus, OH, pp. 495-500.
Robbins, C.L. (1986). Daylighting: Design and Analysis. Van Nostrand Reinhold, New York,
pp. 332-563.
Saxon, R. (1987). Atrium Buildings. 2nd Ed.. Van Nostrand Reinhold. New York, pp.73-92
Stein, B., Reynolds, J., and Mc Guinness, W. (1986). Mechanical and Electrical Equipment for
Buildings. 7th edition, John Wiley & Sons, New York. pp. 884-886.
Utzinger, M., and Bochek, P. (1985). Influence of architectural design on the thermal behavior
of atriums. Proc. Research and Design. AIA Foundation. Los Angeles. CA. pp. 117-122.
-------
2811
A Comparative Analysis for Two Geometrically Different Atria
J. Jones M. Luther M. Navvab
Architecture and Planning Research Laboratory
College of Architecture and Urban Planning
The University of Michigan
Ann Arbor, Michigan 48109
abstract
The atrium is today's most popular element in commercial and institutional designs. Unfortunately, the
dynamic thermal response of these spaces is not well understood. A deficiency in most building load simulation
programs is their lack of incorporating such thermal processes as natural convection, thermal buoyancy and air
temperature stratification Conventional programs do not adequately model the performance of the air mass. While
this problem is less serious for typical space geometries, for voluminous spaces such as atria it leads to significant
inaccuracies in the modeling procedure. For this reason a detailed analysis of temperature stratification for two full-
scale, geometrically different atria was conducted. This study involves monitoring the thermal response of two
atrium spaces under various operating and climatic conditions. The daia is analyzed by comparing time series plots
and by applied statistics.
INTRODUCTION
During the past decade several new buildings with large atriums have been built on the University of Michigan's
Ann Arbor Campus. Unfortunately the thermodynamic behavior of the air in these atriums is generally not well
understood by designers. Part of the reason for this lack of understanding might be that most commonly used energy
simulation software do not accurately model these types of spaces. This is, in part due to the vertical temperature
stratification of the air which can have a large impact on the accuracy of the calculations for the building's estimated
energy consumption. In an effort to better understand the behavior of the air temperature stratification for these
spaces a detailed investigation was conducted. This paper presents the results of this investigation.
The objective of this study was to experimentally monitor these spaces under different seasonal and operational
conditions, determine which factors influence their thermodynamic response and apply the knowledge gained to
efficiently control the heating, ventilating and air-conditioning (HVAC) systems. In addition to the control of the
existing spaces, it was hoped that the results would indicate strategies for improving the design of new buildings.
EXPERIMENTAL SET-UP
The study was conducted for two geometrically different atriums. The first was a long and narrow atrium in the
Electrical Engineering and Computing Science building (EECS). The atrium has an east/west axis with dimensions
30 ft (W) x 300 ft.(L) x 75 ft.(H), (see Figure 1). This atrium has windows at the east and west ends and a skylight
for a roof. A cross-section near the center of the space was selected for monitoring. Thermocouples were suspended
at various vertical dimensions along both the north and south walls as well as in the center of the space, (see Figure
2). Outside air temperature, wind speed and global solar radiation were also locally monitored. Recordings were
made at five minute intervals during cold, moderate and hot outdoor conditions with the HVAC system both on and
off for several consecutive days. 10UTaiM.
LOCATION
OF MEASUREMENT
OCCUPIED
I ZONES I
iHHSffgOCCUPIEDl
sniilzoNEs
PLAN EECS BUILDING
Plan of EECS Building Section of EECS Thermocouple Location
-------
2812
CflP
7|K
I
© FLOOR plan CHKMUTRVBUILDING
J L

Fig:. 3.
Plan of Chemistry Building & Atrium
RADIATION O
KOKTOCLAM TXMT.
Thermocouple Location
Figiu 4.
Section of Chemistry Bldg.
In contrast to the long and narrow EECS atrium the Chemistry Building atrium is square in plan. The
dimensions are 94 ft.(w) x 94 ft. (1) x 77 ft. (h), see Figures 3 and 4. Volumetrically the two atriums are very
similar.
Temperature recordings in the Chemistry atrium were made for various vertical locations. Unlike the EECS
building, however, thermocouples were located along all four walls as well as in the center, (see Figure 4). Outside
air temperature, wind speed and global solar radiation were also recorded. Recordings were made at five minute
intervals during moderate and hot outdoor conditions with the air-conditioning system both on and off.
DATA STRUCTURE
Before beginning the analysis, the data was entered into the computer. Data for each building was organized
sequentially with recordings for the lower level first and upper level recordings last. Each case (line) had values for
the inside air temperature, outside air temperature, solar radiation, wind speed, and time. Dummy variables for
HVAC operation (l=on, 0=off) were created and added to the files. To prevent heteroskedasticity of the residuals
during the statistical analysis, dummy variables were created for each height. For example, with-six vertical
locations, six bivariate dummy variables were created. For recordings taken at the lowest level, the first dummy is
set equal to one and the others zero (1,0,0,0,0,0). For recordings taken at the highest point (level 6) the sequence
would be (0,0,0,0,0,1). These variables allowed the differences in temperature with height to be quantified.
ANALYSIS
As previously suggested the objectives of this investigation were to develop an equation that predicts inside air
temperature for a given height for each space, apply these equations toward optimizing the control strategies for the
HVAC systems serving the atriums and determine the relative influence of each variable.
In an effort to quantitatively determine the relationship between variables such as inside air temperature and solar
radiation, multiple linear regression was applied. Given a set of data, this approach tries to optimize the fit of a
multi-dimensional plane through the data by using a least-squares procedure. The result is an equation such as
Equation 1 that established the relationship between the dependent and independent variables. This relationship is
expressed in the partial slope coefficients (Bi, B2,...).
Eq.l Y = B0 + BiXi+B2X2 + B3X3 + ...+ BicXk
where:
Y = the predicted dependent variable
Bo = the intercept
B 1...BJJ = the partial slope coefficients
X^.-.X^ = the independent variables
-------
2813
A potential problem with using multiple linear regression on this time series data is that the residuals from the
analysis are likely to be autocorrelated. This violates an important assumption for linear regression, which, in the
case of positive autocorrelation, upwardly biases the significant tests for the partial slope coefficients. An alternative
analysis approach is to use Generalized Least Squares (GLS). This approach eliminates the previously mentioned
problem by incorporating the autocorrelation function into the model. The GLS approach provides reliable
significance tests from which interferences for the independent variables can be drawn.
RESULTS
A major objective for the research was to gain an understanding for the air temperature stratification in each of
the two atriums. Chastain and Colliver (REF. 1) define the degree of stratification as the difference between the
maximum T (at the highest elevation) and the minimum T (at the lowest elevation). From our experimental data a
temperature difference between the upper and lower zones was calculated for each case. Using the generalized least
squares analysis approach two equations were developed for predicting the stratification in each atrium. The equation
for the EECS and Chemistry buildings are respectively:
Eq.2
TDIFF
23.48 + .143TOA + .0005 RAD + .03 LogWind - .29 SOLZEN
Eq.3
TDIFF
2.88 + .194TOA - .0005 RAD + .15 LogWind - .14 SOLZEN
where:
TDIFF
predicted temperature difference form upper to lower zone (°F).

TOA
outdoor air temperature (°F)

RAD
global solar radiation (w/sq.m)

LogWind =
log transformation of wind speed (m/sec)

SOLZEN
Solar Zenith Angle
From these equations the predicted degree of stratification was calculated for each space for 0, 200 and 900 watts
per square meter global solar radiation and for outdoor air temperatures ranging from 30: F to 95' F. The calculated
values are shown in Figure 5.
PREDICTED
STRATIFICATION 10
(DEGREES F)
¦
EECS-900
-O
CHEM-900

EECS-200
-o-
CHEM-200
-A
EECS-0
-£r
CHEM-0
-10
30 40 50 55 60 65 70 75 80 85 90 95
F'S« 5- OUTDOOR AIR TEMPERATURE
Prediction of Stratification at Various Levels of Radiation
This figure is helpful toward understanding and quantifying the stratification for both spaces. For example,
while both appear to have approximately the same degree of inverse stratification (-5.5' F - colder at top) at about
30 'F outdoor air and no solar exposure the slopes of the lines are different. This is because the partial slope
coefficient for the Chemistry building (.194) is larger than that for EECS (.143). This suggests that the degree of
stratification in the Chemistry building is more responsive to change in outdoor air temperature. Conceptually this
might be explained by the air mixing that is more likely to occur in the Chemistry building during cold outdoor
conditions when compared to EECS. Because of the free flow of the down drafts from the cold air near the glass to
the lower level, the air temperature is more uniform than in the EECS building where the upper and lower zones
remain quite separate and distinct (see Figure 6). Also the cooling of the lower zone mass by reradiation to the night
sky is likely to further reduce the degree of stratification in the Chemistry building when compared to the EECS
building where the lower zone surfaces are less likely to reradiate to the glass surface thus maintaining warmer lower
zone temperatures and greater inverse stratification.
-------
2814
COLD NIGHT SKY
GLASS
CONDUCTION
NARROW
VIEW ANGLE
COLD NIGHT 5KY
XA35
CONDUCTION
COttf Alk
rilXJNG
Large
v «w Ang •
(grtattf
WAW AIR
rcrrt at on)
CHEMISTRY BLDG.:
Less Stratification, Greater Reradiatlon
and Greater Mixing Effects
EECS BUILDING:
Less Mixing, Greater Stratification
Less Reradiation
Fig,. 6.
A Comparison of Mixing, Stratification and Reradiation Effects
Note: The geometry of the EECS building has a narrower view
angle with the skylight than the Chemistry building.
PROFILE
ANGLE
PROFILE
ANGLE
EECS BUILDING: Solar Impact
Localized Effect
Greater Stratification
CHEMISTRY BLDG.: Solar Impact
Abundant Radiation (less localized)
Less Stratification
Fig. 7.
A Comparison of Solar Impact with Respect to Stratification Effects
-------
2815
Together, these factors have the effect of shifting the outdoor air temperature at which the inversion occurs
upward by about 10 F for the EECS building when compared to the Chemistry building. This shift has significant
consequences when cooiing loads are considered as the cavernous EECS building maintains a large amount of stored
energy in the lower zone.
During periods when solar radiation is available Figure 5 indicates that the EECS building stratification is larger
than in the Chemistry building for similar outdoor air temperature and radiation conditions. At 200 watts per square
meter , for example, EECS has on average about 5°F larger temperature differences than the Chemistry Building.
Larger variation between the two atriums is shown at lower temperatures while smaller differences in the plots are
shown for higher temperatures. As radiation increases to 900 watts per square meter the degree of stratification
increases for both spaces. However, the upward shift in the curve from 200 to 900 watts per square meter solar
radiation greater in the EECS building than in Chemistry. The larger predicted stratification in the EECS building is
again probably due to the two zone behavior (upper and lower) while the air, in the Chemistry building can more
easily mix. Also the portion of the space exposed to solar radiation is larger in the Chemistry building, this allows
deeper penetration of the solar radiation which contributes to the lower degree of stratification (sec Figure 7). The
results of this analysis suggest that stratification in the EECS building is greater than in the Chemistry building
during daylight periods.
The previous results describe the stratification magnitude as a function of outdoor conditions. A second analysis
was conducted to try and determine the average change in inside air temperature with height. Two equations were
derived in an attempt to quantify this effect. (Note: average here refers to both night and daylight periods combined).
Equation 4 predicts the inside air temperature in the EECS building while Equation 5 is for the Chemistry building.
Eq.4 TIN = 78.3 + .017TOA + .0014RAD + ,04Wind -.07 SOLZEN +
1.02HT2 + 1.7HT3 + 2.6HT4 + 4.3HT5 + 6.16HT6
Eq.5 TIN = 76.8 + .11TOA + .0005 RAD + ,13Wind -.06 SOLZEN +
.56HT2 + 1.0HT3 + 1.9HT4 + 4.5HT5 + 6.2HT6
where: TIN = Predicted inside air temperature (oF)
TOA = Outdoor air temperature (°F)
RAD = Solar Radiation (w/sq.m)
Wind = Wind Speed (m/sec)
SOLZEN = Solar Zenith Angle
HT2-HT6 = The average temperature increase for each vertical location as
compared to the lowest zone (HT1). The values of HT2 - HT6
arc bivariate (i.e., either 0 or 1).
The partial slope coefficients for HT2 thru HT6 are plotted for each building in Figure 8. This figure seems to
confirm early findings that the average degree of stratification is greater in the EECS building than in the Chemistry
building.
An interesting aspect of the plot is the distinct change in slope that occurs at about the 48 foot level in the
EECS building. This clearly shows the two zone behavior of this space (upper and lower). It also suggests that the
impact from the skylight is quite localized. The plot for the Chemistry building, on the other hand, shows no clear
point of transition in slope, but a more gradual shift from low to high slope as we move vertically through the
space. This also suggests a more thoroughly mixed air volume and less localized solar impact
Comparison of Predicted Average Temperature Increase with Height
7
6
5
A
Temperature
Difference * F
Relative to Lowest 3
Point
2
KEY:
0 sEECSBLDG.
0
60
70
8
16
28
30
56
46
Vertical Height in Feet
Fig. 8.
-------
2816
CONCLUSIONS
Admittedly this investigation is only a small step on the long road to better understanding the thermal behavior
of these spaces. Preliminary, however, we can say that for these two spaces stratification is much more prevalent in
the long and narrow EECS building. This is likely due to the distinct separation between upper and lower zone
behavior that occurs in these spaces. For the more open Chemistry building the air appears to be better mixed and
more uniform in temperature as we had hypothesized.
The two zone behavior in the EECS building suggests the need for rethinking the current practice of designing
and controlling these spaces as a single zone (see Reference 2).
This is an ongoing investigation. As additional data is gathered we hope to refine our results and apply them
towards better understanding these spaces.
REFERENCES
. Chastain J.P., Colliver D.G., The Influence of Temperature Stratification on Differences Resulting from the
Infiltration Stack Effect, ASHRAE Transactions 1989, Winter Meeting.
Jones J., Luther, M.B., Boonyatikam, S., Thermal Performance Evaluation in a Large Multistory Atrium, 15th
National Passive Conference , Austin, TX, 1990.
, Oberdick, W.A., Boonyatikam, S. Her-Fu-Wu, Energy Performance of Fabric Roof Structures, Architecture
Research Laboratory, The University of Michigan 1980.
Saxon, R„ Atrium Buildings Development and Design, Van Nostrand Reinhold Publishers, Chapter 8,1983.
-------
3.10 Passive Strategies and Materials I
-------
Intentionally Blank Page
-------
2819
SUNSPACES FOR FUNCTION, SECURITY, OUTGASSING, AND AIR TEMPERING
Richard L Crowther, FAIA
401 Madison St., Unit A
Denver, CO 80206
ABSTRACT
In the design of sunspaces, security, outgassing of various items, and air tempering offer
advantages not often considered. This paper addresses the basics of an effective design
approach within the purview of the foregoing considerations. The author has used these
benefits in architectural design.
KEYWORDS
Sunspaces: functions, security, outgassing, air tempering.
FUNCTIONS
Depending upon design, size, and proportions, sunspaces can accommodate various
intentional and unintended functions: interseasonal comfort, solar control as to tasks and
use and the effect of thermal mass on functional use and functional elements on thermal
mass. Factors of ventilation and sunspace plantscaping, glazing characteristics,
proportions, and configurations are relative to sunspace functional use. The secondary
effect upon use of interior spaces that adjoin the sunspace and the relationship of the
sunspace to external space use are other functional considerations.
SECURITY
Security has become an evermore important issue in residential architecture. Depending
upon type of neighborhood, proximity to neighbors, presence of a privacy wall, fences, or
landscaping that might shield an intruder, characteristics of the glazing and sunspace
construction, and an alarm system or watch dog, a certain degree of vulnerability will exist.
The sunspace can be a protector of interior occupied space. A number of effectual
deterrents and detection systems can be incorporated to optimize protection. Multiple
interventions against entry can be planned and designed.
OUTGASSING
Many items can be outgassed of chemical toxins that can be injurious to the respiratory
system and health. Items of particular concern are new clothing, footwear, clothes just
-------
2820
back from the cleaners, bedding, and other items that have been chemically treated in
manufacture. Also include magazines, books, plastic and electrical items, and other entities
that have a telltale chemical odor.
Care has to be exercised so as not to fade or otherwise harm the foregoing items by direct
solar radiation. Optimal natural ventilation along with either direct solar radiation (if not
harmful to the item) or elevated solar heat in a shaded area may be necessary to
significantly remove or reduce toxic emissions for a period of several days to several
months. Paint, lacquer, and other finishes on various items can be effectively outgassed
usually in a relatively short time.
AIR TEMPERING
Sunspaces can be effectively designed to retain solar thermal energy and use solar
radiation heat transfer to temper cold weather air for indoor ventilation. The proportions,
glazed areas, amount, type, and location of thermal mass as well as the interception of
direct solar radiation for heating controlled amounts of incoming outdoor air can in
principally unoccupied sunspaces (or occupied sunspaces relative to comfort) compatibly
raise the temperature of the ventilation air for occupied space comfort. In hot weather
locations with cool nights, sunspaces, self-shaded or otherwise shaded during daytime
hours from the more vertical position of the sun,can be used for nocturnal cooling.
When using air tempering to raise or lower the temperature of incoming air, discretion is
needed in the size, location of, and control over intake air grilles or openings. The location
and design characteristics and intended use for functions, security, outgassing, and air
tempering are individual to solar latitude and solar access, site conditions, and interseasonal
outdoor temperatures.
EVALUATION
Functions
Sunspaces in design can function as passive, hybrid, or active solar systems. Sunspaces
in design should correspond to the particulars of sunspace use. The greater the usefulness
as to space, energy, daylighting, correspondence with architecture and the external
environs, and to occupancy use including security, outgassing, and air tempering, the more
economic they will be.
Vertical glazed sunspaces with no overhead glazing are most suited to comfortable
occupancy. Direct solar radiation, daylighting, and thermal comfort are most easily
controlled by internal or external vertical or horizontal blinds, shades, or shutters. Draperies
are cumbersome, require considerable stacking room, and are not as effective for varying
levels of desired privacy, sun, and daylight control. Vertical glazed clerestories can bring
in direct sun or skyvault daylighting appropriate to the sunspace depth without clear sky
temperature or other thermal losses or the direct solar intensities experienced with overhead
glazing.
Depending upon the types of sun-loving or shade-oriented plants, greenhouses might be
designed with clerestories or overhead glass. Caution should be exercised in cold climates
-------
2821
as to the effect upon plants if the space is used as a transitory outdoor entry to interior
space.
Transitory solar atria of commercial buildings not consistently occupied can be naturally
ventilated during hot weather and,for the most part, not use mechanical heat in winter.
Factors of sunspace function are:
+ orientation, size, design, control, and other elements appropriate to occupancy functions
+ location and amount of thermal mass relative to the use of space
+ the sun's direct radiation and overheating controlled by design
+ daily and interseasonal solar and climatic aspects to not harm elements of the interior
architecture or its contents
+ design for secondary functional, thermal, and daylighting advantages for interior space
+ select reflective or thermal mass furniture and furnishings noninterfering with solar design
+ selected materials not subject to fading or other harm
+ select glazing, ventilation, and air tempering most appropriate to functional realities
+ sunspaces to be an economic and agreeable part of the architecture
+ the sunspace can be a place of relaxation as well as for tasks and activities
+ maintenance to be considered as a functional energy component of the sunspace
+ biophysical and psychoneural aspects and effects of the sunspace
Every project is unique. The total factors of design should be circumscribed and be
integrated and inter-responsive with the architecture. Optimizing the time-frame occupied
or nonoccupied uses of the sunspace will increase its cost benefit.
Security
The majority of homes and buildings have a potential of forcible entry. The value of
contents and possessions and potential risk to the life and well-being of the occupants are
individual to each existing (or to be designed and constructed) project.
The evaluation of risk depends upon many factors. Sunspaces can include the smallest of
residential interventions to the atria of commercial architecture and can be a psychologic
and physical obstacle to unauthorized entry.
While sunspaces can act as an effectual discouragement and barrier to forcible entry, all
other aspects that augment security should not be neglected in existing or new architectural
planning and design.
Factors of sunspace security are:
+ lessening the vulnerability of the location
+ avoiding motivating inducements to forcible entry
+ defensive design and construction of the architecture
+ concern with position, type, and vulnerability of openings
+ providing in-depth strategies to safeguard persons and possessions
+ making direct visual observance of possible intruders
+ installing deterrent noise alarms
+ installing electronic interception systems
+ installing closed-circuit visual monitoring
+ using time-delay factors against intruder objectives
-------
2822
+ devise capturing of intruder strategies
+ daylighting and nighttime lighting deterrents
Sunspaces can be appropriately designed using all or part of the foregoing factors as a
concern unique to the protection of each project. How to optimize the advantages of a
sunspace to accord with the total plan and design elements within such project is a primary
question. Sunspaces can be adroitly scaled and formed as an integral part of the
architecture or as a superimposed element leading into the architectural interior. Each
project offers creative possibilities in placement, concept, and design.
The general construction, type of glazing, means of ventilation, and security measures of
a sunspace can be designed for moderate to maximum security. A primary question is
how to buy time between an outer breach of a sunspace and an inner breach between the
sunspace and interior occupied space.
Assuming that the type of construction of the architecture is clearly adequate for security
and that a sunspace or sunspaces can be located and serve in place of more vulnerable
window and door openings, the challenge is how for a new or renovated project to design
an effectual, practical, and economic security sunspace.
Outclassing
How, when, and for how long to outgas materials and other entities can be perplexing. For
some items with strong telltale odors or known toxic chemicals, it is not difficult to make a
decision. The sensitivity of persons to the many and sometimes singular respiratory insults
of indoor space and its contents can result in mild to serious biophysical reactions. The
nose is a very sensitive instrument of detection. The ability of odor detection varies with
people. The ability to "detect" can be cultivated. Learn more about what noxious and
toxic chemicals might be expected from the products you may buy.
Factors of sunspace outaassina are:
+ outgassing of the architectural materials and interior elements
+ items to be outgassed should not present a risk to occupants
+ the sun's radiation may be directly or indirectly used
+ high ambient levels of heat in a shaded area should be a caution for some items
+ optimal outdoor air ventilation with a most effective exhaust should prevail
+ cold or hot outdoor temperature ventilation should not injure items to be outgassed
+ consider disassembling elements that need outgassing from those that do not
Air Tempering
Inasmuch as commercial buildings have significant internal thermal loads of lighting, people,
and equipment and generally have a large ventilation requirement, air tempering is of great
importance. Solar atria or other types of sunspaces can be effective for air tempering.
With proper ventilation of commercial buildings to avoid "sick building syndrome," air
tempering will be most needed during very cold and very hot weather. Using solar instead
of utility energy for air tempering can be both economically and ecologically advantageous.
Inasmuch as the "sun doesn't always shine," supplemental air tempering will be required
during periods of inadequate sunspace solar gain. This largely applies to commercial and
-------
2823
institutional buildings. Residences with few people, lower internal thermal loads, and lower
need for ventilation to provide oxygen and C02 removal are unlikely to have such
supplemental air tempering requirement.
Air tempering for auditoriums and large indoor congregations of people can be too
demanding for an atrium or sunspace. But this should not be ruled out without an
investigative possibility. Every commercial and institutional project has its unique
characteristics. What best can serve for the essential needs of air tempering requires
explicit air quality and thermal demand solutions. As with the majority of solar applications,
residential and smaller commercial projects offer the greatest range of opportunity.
Factors of sunspace air tempering are:
+ solar access and predictable insolation
+ calculating the interior volume, mass, and solar radiation interceptions and reradiant
loss of the sunspace as an air tempering system
+ calculation of outdoor temperature variation relative to air flow and thermal transfer of the
sunspace
+ manual or automatic methods for the introduction of tempered air into the interior
+ location of outdoor air intake versus intake of air into occupied interior space
+ concern with factors of air quality and sunspace and indoor levels of comfort
+ interface with passive and active natural as well as mechanical systems
+ consideration of unplanned and planned shading factors
+ time periods when the sunspace may be occupied or be used for outgassing
The ultimate sunspace for air tempering would be a proportional biosphere environs of
oxygenation and C02 adsorption by an adequate number of appropriate plants. The
tempered air would also benefit by an accompanying amount of humidity. However,
molecular adsorbing materials to control mold and filtration of sunspace-tempered air
mechanically distributed to occupied spaces would be desirable.
CONCLUSION
Each project has its individuality within planning and design. Sunspaces become part of
this individuality. Residential and commercial sunspaces (occupied or unoccupied) differ
in factors of possible function, security, outgassing, and effectual air tempering.
For space to be functional, reasonable comfort is needed, an absence of materials and
other sources of odoriferous and toxic outgassing is most advisable, and air tempering and
security depend upon design strategies that can be used to these ends.
Due to multiple advantages that can be incorporated into their planning, design, and
specification, sunspaces can be a valuable asset beyond uses most commonly ascribed
to them. The sunspace need not be located directly adjacent to other interior spaces that
it might serve for thermal needs, security, outgassing, and air tempering.
Interior space daylighting and secondary solar thermal gain through a sunspace are
benefits not to be neglected. An enclosed sunspace with glazing to interior space can itself
(due to its factor of depth) act as a shading device during the heat of summer (well
ventilated to exhaust heat from the sunspace) while allowing direct solar radiation in winter
to penetrate through to interior space. With carefully controlled air tempering it can perform
-------
2824
as a thermal buffer to the interior space. Reflective surfaces outdoors or within the
sunspace can have a thermal and visual effect from daylighting and solar reflection.
Sunspaces can be designed specifically for extensive and sustained outgassing, can
provide secondary solar gains and daylighting, and in being an isolated air space will not
affect the respiratory well-being of interior space occupants.
Orientation, location, size, proportion, design, openings, and all architectural resolutions
should not omit the effect upon any furniture, furnishings, products, or other items the
sunspace has as elements of thermal design and occupant use. Damaging and
uncomfortable as well as beneficial effects of direct solar radiation should be considered
in totality of the architecture and sunspace design.
The author/architect has used the conceptual design approach as presented in this paper
for a number of projects. These projects enjoy security, outgassing and air tempering
benefits, peace of mind, and protection of persons, contents, and belongings. The indoor
health benefits of outgassing toxic or odoriferous items in a sunspace separated from
occupied areas and tempering outdoor air for indoor comfort and ventilation are additive
values to the more customary use of sunspaces. The correlate opportunities of using both
residential and commercial sunspaces for various tasks, functions, and for security,
outgassing, and air tempering should not be overlooked as extendable values.
-------
2825
THICK FRIENDLY WALLS - ENERGY EFFICIENT COMMERCIAL BUILDING DESIGN
Graeme Robertson
Department of Architecture
University of Auckland
New Zealand
ABSTRACT
The environmental message of the last few decades can no longer be ignored - global pollution is an issue
that must be addressed by everyone, including architects. World wide, building industries have been
identified as major energy users, and, more particularly, significant producers of greenhouse gases.
This paper suggests that a growing awareness of the Greenhouse Effect, combined with energy price rises,
will create a definite need for architects to produce energy efficient designs generally but particularly in the
commercial building area. To further develop this attitudinal change developers and building owners are
showing an increasing perception of the advantages, in terms of worker efficiency, to be had from
producing environmentally connected commercial buildings.
Specifically reported is the on-going work in Auckland investigating the performance of 150 local high rise
office towers in terms of energy useage, greenhouse gas production and aspects of user comfort leading to
appropriate design guidelines and implementation methods.
In terms of architectural design theory a return to a functionalist style is argued as a reaction to the current
'eclectic' theories abounding. Environmentally sympathetic commercial building designs will become
acceptable, and all parties will benefit. The thick friendly wall will proliferate.
KEYWORDS
Greenhouse effect, commercial buildings, worker efficiency, environmentally connected, energy efficient,
thick wall architecture.
INTRODUCTION
Energy efficient commercial buildings designed to respond to the natural environment by using daylight,
natural ventilation systems and solar heating/cooling are relatively rare world wide. The architecture of
environmentally connected buildings in temperate climates is well recognised: long east/west shallow plan
forms to allow optimum solar control; well designed and positioned windows and, ideally, an increased
floor to floor height to allow optimum daylight use; efficient draught free ventilation systems; an external
skin with an appropriate thermal resistance to suit the climate zone and building use. All are design
considerations that can be summarised as "thick wall" architecture in contrast to the typical climate rejecting
glass walled towers of recent years.
Less than realistic energy supply and pricing policies of the last decade have reduced the movement toward
environmentally connected commercial buildings which was partially instigated by the first energy crisis of
the early 1970s. There are definite indications that both real and perceived public concerns relating to
climate change and energy efficiency generally will lead to increased emphasis on environmentally
connected commercial building designs. The growing recognition of the associated increased worker
efficiency with these buildings will also contribute to a more positive approach from architects and
developers. The Green Architecture movement provides the overall vehicle for better, more environmentally
responsible commercial buildings.
-------
2826
CLIMATE CHANGE
It is now well recognised (I.P.C.C. 1990) that emissions resulting from human activities are substantially
increasing the atmospheric concentrations of greenhouse gases: carbon dioxide, methane, chloro fluoro
carbons and nitrous oxide. These increases will enhance the greenhouse effect, resulting on average in an
additional wanning of the earth's surface of between 0.2° C and 0.5°C per decade, which is greater than
that seen over the last 10,000 yean.
In a temperate climate country such as New Zealand where electricity is the dominant energy form in
buildings, and 80% of this electricity is produced from hydro schemes, it has been estimated (Ministry for
the Environment, 1990) that if passive solar features were included in new commercial buildings then
energy savings of 50% could be achieved with minimal extra capital expenditure. This saving would then
translate through to a decrease in New Zealand's total carbon dioxide emission of approximately 3%, and
total greenhouse gas emissions of approximately 1.4%. Although these sector savings are relatively minor
when compared to those of many other developed countries, they cannot be ignored if New Zealand is to
meet its internationally stated aims. Growing tenant and public awareness, along with possible government
incentives, will encourage designers to respond to these possibilities.
WORKER EFFICIENCY
An energy efficient building design incorporating an environmentally connected external skin to maximise
daylighting possibilities, solar control and natural ventilation needs - the 'thick' wall - also largely satisfies
many of the requirements for worker efficient office spaces.
In terms of the dollar outlay over the 40 year life cycle of an office building, it is generally recognised
(Wineman 1986) that 2-3% is spent on the initial costs of the building and equipment, 6-8% on maintenance
and replacement and 90-92% is generally spent on personnel salaries and direct benefits. These figures are
quoted universally and have been tested in a recent Auckland study (Robertson 1990).
TABLE 1 Auckland C.B.D. Office Annual Costs (Tenants)
Component Percentage
Notes
Personnel Costs 80-87
Salaries & direct
benefits vary with the nature
and size of the business.
Leasing Costs
1.5-8
The variability of accounting
procedures produces such a
wide variation.
Lifts
Air conditioning/
Heating maintenance
0-5.5 Nature and age of the plant
contribute to variation
3-4
Cleaning
Building Maintenance
(Fabric)
Energy
Other
2-5
0.5-2.5 The variability of leasing
agreements is a factor here.
0.3-2.2
1.5-8.5 This figure includes
fire,office,communication
equipment,
along with rates
and insurances.
Clearly potential savings in the 'personnel costs' segment are of critical importance, and certainly outweigh
those possible in the 'energy' segment. If better performance can be achieved from the individual workers
then the 'personnel costs' segment can be reduced proportionately. Productivity is a broad concept which
-------
2827
include not only employee work performance but also associated organisational costs such as employee
turnover, absenteeism, tardiness, required overtime, vandalism, grievances and mental and physical health.
Historically much of the research (Lawler and Porter 1967, Locke 1970, Herzberg 1976) on productivity
has relied on measures of employee satisfaction as indicators of productivity.
Research suggests (Archea 1977) that environmental characteristics that influence this level of work process
include both architectural properties (size, heating, ventilation, airconditioning, lighting and ergonomics)
and architectural attributes (assessment of openness, noise, enclosure, lighting and temperature). Hedge
(1986) points out that from a survey of 896 office staff that even though the ambient environment was
supposedly being maintained around an 'optimum' level in physical terms, adverse reactions to this were
still quite pronounced (Table 2).
Table 2 Employee Reaction to Ambient Environment in Offices - Hedge 1986
Ambient
Comment
Percentage
Condition

Agreement
Temperature
Too hot
48
Too cold
21
Ventilation
Desire to open window

Too stuffy
61
Lighting
Prefer more daylight

Too bright
33
A recent Auckland study (Robertson 1990) involving in excess of 1500 workers in airconditioned offices
suggests a similarly high level of dissatisfaction with the quality of the office environment with clear
contradiction arising off aspects of temperature, ventilation and lighting, as shown in Table 3:
Table 3 Employee Reaction to Ambient Environment in Offices - Robertson 1990
Ambient
Comment
Percentage
Condition

Agreement
Temperature
Too cold
31
Too hot
48
Ventilation
Too stuffy
57

Too draughty
22

Don't know
21
Lighting
Too bright
34

Wrong angle
25

Too glarey
37

Too dark
28
The clear conclusion, because of the high level of dissatisfaction and contradiction, is that office staff desire
to be able to control their own environments more than the present air-conditioned offices allow. The
typical environmentally connected commercial building with the 'thick' wall characteristics satisfies
admirably this desire for more natural environments with better individual control.
The worker efficiency argument is attractive, as even a slight improvement in productivity will have a
significant impact on the overall economic structure of the organisation.
ENERGY EFFICIENT DESIGN OPTIONS
Recent local studies (Breuer 1988) have suggested viable energy efficient design options for office
buildings in our temperate climate. After running computer comparisons of a base building and one
incorporating various energy efficient design options, the clear conclusion is, as expected, that office
buildings in Auckland's temperate climate are internal load dominated, generating a considerable amount of
heat due to people, lighting and equipment. Excess heat has to be removed from the building in summer
and in winter.
-------
2828
Table 4 shows the summary of the thermal and economic performance in Auckland for each of the simple
energy saving measures
investigated.
Table 4, Energy Efficient Options,
Auckland Office Buildings
Option
E
kWh/n^/YR
PV-E
$PV/m2
C
$/m2
NB
$PV/m2
Efficient Lighting
Perimeter Daylighting
Skylights
Clerestories
Shading
Double Glazing
Extra Size Windows
Insulation
Where E =
PV-E
C
12.0
26.4
6.0
20.4
22.0
48.4
7.7
40.7
5.0
12.6
9.4
3.2
6.2
15.5
25.0
-10.0
3.0
8.2
10.1
-1.9
-1.5
-6.4
29.1
-35.6
1.5
8.0
2.8
5.2
5.0
6.3
4.6
1.7
Differential Thermal Performance
= Present value of energy saved
= Differential Capital Cost
(including labour)
NB = Net Benefits
The clear conclusion is that to improve the energy efficiency of a commercial building in Auckland it is
important to produce a form of building and an external skin design to maximise the possibilities of
daylighting because of the essentially cooling requirements of energy efficiency. Aspects relating to
reducing heat loss - insulation, double glazing - are predictably unimportant. To achieve good daylighting
without glare or damaging solar gains a 'thick' wall form of external skin is necessary.
DESIGN IMPLEMENTATION
If one accepts the justifications of climate change, worker efficient and energy efficient commercial building
design claims, then the production of "thick friendly wall" architecture becomes necessary. Walls become
filters to the natural environment rather than excluding as do so many of our recent glass towers. To
achieve this adequately, depth to the external skin is needed as shown in Fig. 1.
FIG 1. Appropriate "Thick Friendly Wall" Architecture
-------
2829
Developers and building owners will be attracted to the economic advantages of worker efficiency and
energy efficiency, and may well be significantly influenced by growing public awareness of the
Greenhouse Effect and the Green Movement generally. Government incentives or mandatory standards to
achieve the consequent reduction in greenhouse gas emissions may not be necessary if adequate education
is carried out. In a market forces economy incentives or even performance regulations by central
government are less than desirable or popular. There are real signs that the market will adopt the design
changes necessary without intervention by government and a new functional phase in the design of
commercial buildings will evolve. A functional style which is climate responsible and incorporates aspects
of the "thick friendly wall".
REFERENCES
Archea S. (1977). The Place of Architectural Factors in Behavourial Theories of Privacy, Journal of Social
Isai£S_2i
Breuer D. (1988). Energy Efficient Passive Solar Commercial Building Design, N7.F.R DP Report 173.
Wellington.
Hedge A. (1986) The Impact of Design on Employee Reactions to their Offices, Behavioural Issues in
Office Design. Van Nostrand Reinhold.
Herzberg F. (1976) One More Time: How Do You Motivate Employees? Concepts and Controversy in
Organisational Behaviour. Goodyear Publishing.
Intergovernmental Panel on Climate Change, Working Group 1 (1990) Scientific Assessment of Climate
Change. Report.
Lawler E.E. and Porter L.W. (1967) The Effect of Performance on Job Satisfaction, Industrial Relations 7.
Locke (1970) Job Satisfaction and Performance, Organisational Behaviour and Human Performance 5.
Ministry of the Environment (1990) Responding to Climate Change - A Discussion of Options for New
Zealand. Wellington.
Robertson G. (1990) The Marketing of Energy Efficient Multi-Storey Commercial Building Design,
Proceedings International Council for Building Research Symposium on Property Maintenance.
Managment and Modernisation. Singapore.
Robertson G. (1990) Survey of ISO Auckland Central Business District Offices. University of Auckland.
Wineman J.D. (1986) The Importance of Office Design to Organisational Effectiveness and Productivity.
Behavioural Issues in Office Design. Van Nostrand Reinhold.
-------
2830
THE HEAT CAPACITY ANALYSIS OF RAPID HEAT COLLECTING WALL
Li Baojun, Wang Jingyu, Xiong Xiaogeng
Yang Shilong and Jiang Xiangshan
Dept. of Physics, Shenyang Arch. Eng. College
Shenyang 110015, P.R. China
ABSTRACT
This paper is about a setting where a thermal insulation layer is set in the
south wall at the cold range, so that the south wall only collects heat, but
does not transfer it. During full shinshine hours solar energy can be rapidly
transferred into the room, and the heat loss of the south wall is greatly lowered.
The practical measurements have shown that the temperature in this solar house
is 2 to 5°C higher than that in a solar house of Trombe's.
KEYWORDS
Rapid Heat Collecting Wall, Trombe Wall; vent, greenhouse, Newton's law of
cooling; convective heat exchange
INTRODUCTION
The south wall of a passive solar house is generally a Trombe wall. It is
economic and there is little fluctuation of the room temperature. The dis-
advantage of the Trombe wall is that it absorbs and conducts heat slowly
and the solar energy is not absorbed fully during full sunshine hours. Espec-
ially in the cold range there is an amount of heat loss because the wall has no
thermal insulation layer. In order to overcome the above-mentioned dis-
advantages, this paper presents a new design concept of a rapid heat collecting
wall.
PHYSICS MODEL
The rapid heat collecting wall is made of bricks. A thermal insulation perlite
board 5 cm thick is put on the outside of the south wall. Black paint is
applied to the board to increase heat absorptivity. A glass plate 3mm thick
is covered outside the thermal insulation board, 15 cm spaced, to produce a
greenhouse effect. Two vents, 20x30 cm are
10 cm below the ceiling and 10 cm above the
The physics model is shown as vollows: ^^
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
separately installed in the wall
floor to make heat convection.
Glass plate 3 nan in thick
Greenhouse 15 cm in width
Black coating
Pearlite board 5 cm in thick
Red brick wall 37 cm in thick
Vent 20*30 cm at bottom
Vent door
Air circulation
Vent 20*30 cm at the top
Fig. 1. Physics model of rapid heat collecting wall.
-------
2831
ESSENTIAL PRESUMPTIONS
1) Sunlight directly transferred into a room does not affect air circulation.
2) Temperature in the room only changes with time. 3) Heat exchange by con-
vection in greenhouse is computed by two solid surfaces. 4) the ratio of hei^t
to width of the wall is more than 10, and it can be considered to be one dimen-
sion heat transfer, that is to say there is only thermal resistance through the
wall.
SOLAR ENERGY GOT BY THE SOUTH WALL
Solar radiant intensity;
HI 11 _
Isv= I„p cosh cos£ + ? I. 1-U4in£ sinh (1 )
Solar transmitted intensity through glass plate:
HG-S = (SSGDv.Xs + SSGi).SC.X^.F (2)
Solar energy got by the south wall:
HGW =dHGS (3)
SOLAR ENERGY FLOWING INTO ROOM
Heat Transfer "by Convection
While the air circulates through vents between the room and the
greenhouse* natural convection heat exchange can take place be-
tween the surface of south wall and the circular air, the quantity
of heat exchange is as follows:
q =0L(6 - t ) (4)
c °
Heat Transfer by Radiation
Because the dimension of glass window is much bigger than the dis-
tance between glass window and external surface of south wall,
transferred energy'is as follows:
Total Energy Transferred by Air Circulation
Whole energy transferred by air circulation equals to the sum of
the convection and radiation. In accordance with Newton's law of
cooling, total energy is as follows:
q = qc + q^ =dij ( 0 ~ t ) (6)
-------
2832
Heat Transferred into the Room Through the Vents
The air in the greenhouse passes through the vents at top and moves
into the room is thought as one dimension tapered flow. Heat flow-
ing into the room is as follows:
HGt = 2-^-t Pop.( (7)
MEASUREMENT RESULT
For reducing the space we only give out one drawing:
*
-------
2833
SSGd Heat of diffusion radiation transmitted glass, w/ml
qc Intensity of convective heat exchange, kcal/m'.h
c£c Factor of convective heat exchange, kcal/m2.h. *C
0 Temperature of external surface on south wall, °C
t Temperature of glass window, °C
qr Intensity of radiative heat exchange, kcal/m*.h
Gil Equivalent radiation factor, kcal/m^.h.k4,
q Total quantity of heat transfer, kcal/nfth
Surface heat transfer factor, C^.= dc+
HG^ Heat transferred into room through vents in unit time,
kcal/h
A Vent area
k Adiabatic exponent, k= Cp/Cy
P Airflow pressure in greenhouse
P„ Airflow density in greenhouse
$ Pressure ratio, ^ = P/P„
REFERENCES
Beikei, R. (1963)• Handtiuch der Physik, Vol. VIII, Pt. 2.
Chapman, S., and T. G. Cowling (1952). The Mathematical Theory of
Non-uniform Gases, 2nd ed. Cambridge Univ. Press.
Philips, D. (1964). Lighting in Architecture.
-------
2834
RADIANT BARRIERS IN RESIDENTIAL DWELLINGS: ENERGY, COMFORT,
AND MOISTURE CONSIDERATIONS IN A NORTHERN CLIMATE
Russel Mendenhall, Ph.D.
Snow College
Ephraim, UT 84627
ABSTRACT
The purpose of this study was to determine the conditions under which radiant barrier
utilization in attics is appropriate technology in building construction for a northern climate
in Utah and for use in a college building construction program. The majority of the
research about radiant barriers had been conducted in southern climates of 3,500 heating
degree days or less. Actual field studies in a northern climate such as Utah had not been
conducted to validate the effectiveness of radiant barriers placed across bulk insulation in
an attic. The climate region where this study took place was in Central Utah with
approximately 6,500 heating degree days.
KEYWORDS
Utility usage compared; attic moisture readings; field study; questionnaire analysis; attic
insulation.
INTRODUCTION
This study examined conditions under which radiant barrier utilization is appropriate
technology in building construction in a northern climate. This study determined the
following: 1) occupants perceived comfort level related to radiant barriers, 2) occupants
perceived condition of moisture artifacts associated with the use of radiant barriers, 3)
occupants perceived utility usage associated with radiant barriers, 4) utility usage between
houses with radiant barriers and houses without radiant barriers, and 5) the difference
between moisture content in ceiling joists in houses with radiant barriers and houses without
radiant barriers in a northern climate.
A review of the literature revealed that the majority of the research about radiant barriers
had been conducted in southern climates of 3500 heating degree days or less. The studies
had also been conducted primarily in highly instrumented laboratory conditions. Only a
few field tests in occupied houses had been conducted on the actual possible effects of
radiant barriers. In general, radiant barriers were most effective for energy conservation in
climates requiring cooling. However, actual field studies in northern climates have not been
conducted to validate the effectiveness of radiant barriers placed across bulk insulation in
an attic.
-------
2835
The procedures used to generate the research questions were as follows: 1) a questionnaire
was formulated based upon factors revealed in the review of literature, 2) the questionnaire
was validated by a panel of judges, and 3) the instrument was field tested.
Twelve houses with radiant barriers installed in their attics for at least one year were then
selected in Utah. Each of the selected houses had to meet specified selection criteria. Then
a control group of 12 houses without radiant barriers were paired with the 12 sample houses
with radiant barriers on the basis of size, similarity of construction, similar age, number of
occupants, and geographical proximity.
The questionnaire was hand-carried to the occupants of the 12 sample houses with radiant
barriers and completed by the occupants without assistance from the researcher. Further,
a demographics data collection instrument was hand-carried to the 12 sample houses with
radiant barriers and the 12 control houses without radiant barriers. The data collection
instrument was filled out by the researcher from the responses of the occupants. At the
same time, a letter was handed out which outlined details of the research.
The occupants of the sample and control houses were requested to sign a release for their
utility bill information. The utility bills were obtained directly from utility company records.
A t-test was calculated from the annual BTU's between the sample houses with radiant
barriers and control houses without radiant barriers.
From December 1, 1989, through February 1,1990, a moisture meter was used to measure
wood moisture content of the bottom truss cord or ceiling joists of the 12 houses with
radiant barriers and the 12 houses without radiant barriers. Three separate readings were
recorded in the morning near the first of the months of December, January, and February.
Moisture readings were taken centrally over three areas in the attic of each house, including
the kitchen attic area, bathroom attic area, and living room attic area. At each location,
three moisture measurement readings were taken in the wood truss at 12-24 inches above
the radiant barrier, one inch below the radiant barrier, and two inches above the ceiling
sheetrock. The data were collected and tabulated for findings reported in this study.
To summarize the major findings, the research questions will be listed and the findings and
conclusions will be stated.
Research question one. To what extent are radiant barriers associated with the occupant's
perceived comfort?
Sixty-seven percent of the occupants agreed that their houses were more comfortable and
the heat was more even throughout the house since the installation of radiant barriers.
Also, 60 percent of the occupants stated that their houses were cooler in the summer due
to the installation of radiant barriers. This was the most favorable comment related to the
use of radiant barriers and not surprising due to the fact that the review of literature
substantiated that radiant barriers are effective for energy conservation in cooling climates.
Research question two. What moisture artifacts (i.e.,nail rusting, sheetrock discoloration,
mold and mildew, etc.) appear to be associated with the use of radiant barriers?
In no case did an occupant confirm additional moisture artifacts due to radiant barrier
installation. In the open-ended statement at the end of the questionnaire, no occupants
indicated increased moisture effects inside their house due to installation of radiant barriers.
-------
2836
Research question three. To what extent will the occupants perceive a utility difference
associated with the use of radiant barriers?
The occupants agreed that radiant barriers did help in lowering the heating costs. One
household felt that they had a savings of $50 per month. A majority of the occupants did
not have air conditioning and did not have an opinion concerning savings on cooling.
However, the most mentioned and favorable statement at the end of the survey indicated
that seven out of twelve respondents said that radiant barriers did keep their houses cooler
in the summer. It may be noted that the home owners might be biased in favor of the
results because of the money spent on the product.
Research question four. To what extent will heating utility costs be affected by the
installation of radiant barriers?
A t-test was calculated from the annual BTU's consumption between the sample houses with
radiant barriers and control houses without radiant barriers. The t-test value between the
two groups was .998 and was found not significant at the .05 level. Further calculation
showed that the paired differential t-test was .846 which was also found not be significant
at the .05 level. As noted, the occupants did perceive heating savings, but no significance
was calculated. In reviewing the literature, there was an indication that the radiant barrier
might have a detrimental effect, reflecting away radiant heat which would normally help
heat the house in cooler months.
Research question five. To what extent will the moisture content of ceiling joists differ
during winter months in houses with radiant barriers versus houses without radiant barriers?
One house with a radiant barrier measured a moisture content over 8 percent. The
bathroom attic area measured 8.7percent, 11.6percent, and 11.8percent, respectively, for
the months of December, January, and February. On-site visits of the same house revealed
dampness on top of the insulation. Also, there was condensation over the kitchen attic area,
but no moisture reading was measured. The demographics of the house show that three
people occupied a 1,000 square foot, ranch-style house, masonite siding, electric radiant
ceiling heat with electric baseboard heat in the basement, and with no wood stove. The
occupants said there has always been excessive moisture in the house and condensation
around the windows. Moisture conditions are often characteristic of electric radiant heat
and warrant further investigation and monitoring of this type of house.
Another house showed signs of condensation but did not register any moisture content
measurement. Even though perforations were present in the barrier, there was a potential
for moisture in houses in northern climates.
Also, t-test results for comparing relative humidity inside the occupants houses and in the
attics were significant for the three months at the .01 level. However, the paired difference
in the t-test results were not significant at the .01 level. One must be cautious in
interpreting these results. Further research with a larger number of houses with controlled
variable would be necessary to substantiate these findings.
-------
2837
CONCLUSIONS
The conclusions of this study are based upon the findings obtained from the data collected
within the study. The study was based upon five research questions. The findings indicate
the following conclusions: 1) additional comfort is possible from the use of radiant barriers;
2) radiant barriers may be over-rated in the amount of comfort expected; 3) no additional
moisture artifacts exist inside houses in a northern climate. 4) radiant barriers keep houses
cooler in the summer; 5) research indicates no significant difference exists between utility
usage of sample and control houses; radiant barriers may not provide heating benefit in a
northern climate; 6) moisture condensation conditions can exist under the radiant barrier
in attics in northern climates; 7) moisture content in the ceiling joists of houses can be
found and measured in a northern climate; 8) houses with radiant barriers haw a tendency
to have higher relative humidity levels than houses without radiant barrier instaJhtion, which
warrants further investigation; and 9) it is questionable that radiant barriers are appropriate
technology to use in a northern climate.
-------
2838
REFERENCES
ASHRAE (1960). Heat transfer. Heating. Ventilating. Air Conditioning Guide. Vol 38.
ASHRAE, Inc., New York. p. 49.
ASHRAE (1985). Moisture in building construction. ASHRAE Handbook. 1985
Fundamentals. ASHRAE, Inc., Georgia, p. 21.2.
ASTM (1987). Increasing durability of building constructions against water-induced
damage. Annual Book of ASTM Standards. Vol. 4.07. Pennsylvania, p. 455.
Balcomb, D. J. (1983). Passive Solar Design Handbook. Vol. 3. American Solar Energy
Society, Inc., New York. p. 631.
Bourne, R. C.,& Hoeschele, M. A. (1988). Simulated Attic Radiant Barrier Performance
in Sacramento. Sacramento Municipal Utility District, CA.
Cleary, P., & Sherman, M. (1984). Seasonal Storage of Moisture in Roof Sheathing.
University of California, Lawrence Berkeley Laboratory, Berkeley, p. 4.
Degelman, L.,& Snider, R. (1989). The lowdown on radiant barriers. Custom Builder.
EEBA (1988). Energy Efficient Building Association. EEBA Alert. University of Southern
Maine, Technology Center.
Fairey, P.,Swami, M.,&Beal. D. (1988). RBS Technology: Task 3 Report. DOE No. DE-
FC03-86SF26305. Cape Canaveral, FL. p. 4.
Forestry Department (1986). Forestry Paper (Wood Preservation Manual). Mechanical
Wood Products Branch Forest Industries Division, p. 21.
FSEC (1988). Florida Solar Energy Center. Consumer Facts About Radiant Barriers. Rep.
No. FSEC-FS-37-88. FSEC, Cape Canaveral, Florida.
Kollmann, F. P., & Wilfred, A. C. (1968). Principles of Wood Science & Technology.
Springer-Verlag, New York. p. 108.
Hall, J. A. (1986). Performance testing for radiant barriers. Symposium on Building
Energy Efficiency. Arlington, TX. p. 5.
Hall, J. A. (1988). Performance testing of radiant barriers with Rll, R19, and R30
cellulose and rock wool insulation. Fifth Annual Symposium on Improving Building
Energy Efficiency in Hot and Humid Climates. Houston, TX.
Levins, W. P., & Karnitz, M. A. Heating Energy Measurements of Single-Family Houses
with Attics Containing Radiant Barriers in Combination with R-ll and R-30 Ceiling
Insulation. Contract No. DE-AC05-840R21400. Oak Ridge National Laboratory,
Oak Ridge, TN.
Levins, W. P., & Karnitz, M. A. Heating Energy Measurements of Unoccupied Single-
Family Houses with Atics Containing Radiant Barriers. Contract No. DE-AC05-
840R21400. Oak Ridge National Laboratory, Oak Ridge, TN.
Lstiburek, J. W. (1987). Applied Building Science. Building Engineering Corp.,
Downsview, Ontario, Canada, p. 15.
Roux, J. A. (1985). Modeling the Effects of Radiant Barriers on Heat Transfer Through
Fibrous Insulation.
-------
2839
THERMAL PERFORMANCE OF LATENT HEAT STORAGE MATERIAL
FOR FLOOR HEATING SYSTEM
Jein Yoo, Soo Cho, Hun S. Chung
New and Renewable Energy Research Center
Korea Institute of Energy and Resources
P.O. Box 5, Daedeok Science Town, Daejeon, Korea
ABSTRACT
The goal of this research is to develop a phase changing material for the low
temperature thermal storage of solar energy and to investigate its applicability.
Industrial grade calcium chloride hexahydrate(CCH) was chosen because of its great
potential for the Korean traditional heating system, ONDOL, which uses hydronic
radiant floor heating.
Polyurethane ONDOL-panel system was selected to test thermal performance. This
system has such problems as over-heating and fast-cooling which results in energy
loss and discomfort To evaluate the CCH's effect, two identical unit test cells
were built and installed with ONDOL-panels with and without CCH, respectively.
It was found that ONDOL-panels with CCH could reduce the room temperature
fluctuations and maintain the phase changing temperature for considerably long
duration, 2 ~ 4 times of heating hour, over no-CCH one. Therefore, it was concluded
that CCH having phase changing temperatute of 29°C is suitable for the floor
heating system.
KEY WORDS
Latent heat storage, Floor heating, Phase changing material
INTRODUCTION
Radiant floor heating system, ONDOL, has been used for more than a thousand years
in Korea. Floor was heated from the bottom by wood burning fire place. Nowadays
traditional ONDOL system was modified into hydronic radiant floor (HRF) heating
system. In HRF heating system, hot water heated by boiler circulates through pipes
buried in concrete mortar of floor. Therefore concrete floor itself is a sensible heat
storage material and also heat emissioning surface.
Recently new HRF heating system, so called ONDOL-panel was introduced in
Korean market. ONDOL-panel consists of thin galvanized steel plate, polyethylene
pipes and polyurethane as a insulation. They are mass-produced in factories and
can be assembled simply at each construction site. ONDOL-panel has negligible
thermal capacity due to the structure which results in drawbacks of temperature
fluctuations such as overheating and fast cooling. To solve these problems, latent
heat storage material may be one of the promising solutions.
CCH was chosen as PCM in consideration of phase changing temperature ( about
-------
2840
29°C), latent heat capacity, etc. It was found by the numerical analysis of heating load
of residential house in Korean weather that the surface temperature of floor should be
about 30°C. Therefore CCH was selected as PCM for ONDOL-panel.
Experiments were performed in two test cells sized 1.8x1.8x1.8 m. One test cell
equiped with commercially available ONDOL-panels and the other test cell equiped
with ONDOL-panels containing CCH. Comparisons were made in terms of room
temperature and surface temperature in test cells during winter heating season.
EXPERIMENTS
CCH and Containers
Calcium chloride hexahydrate (CCH) made from industrial grade was analyzed with
differential scanning calorimeter and its characteristics are listed in Table 1. The
detail of chemical process and characteristics of CCH developed in this research is
beyond the purpose of this paper, and so it is omitted here.
Table 1. Constituents and properties of CCH
Chemical composition
Physical properties
CaCI2 71.5%
H20 26.0 %
NaCI 1.5%
others 1.0 %
Melting 29.2 °C
Latent heat 170J/gr
Density: Solid 1.62gr/cm3 , liquid 1.50gr/cm3
Specific heat: Solid 1.44J/gr°C, liquid 2.32 J/gr°C
CCH containers were fabricated with extruded aluminium as shown in Fig. 1. The
capacity of a container was determined by the heating load of test cell with
assumptions of room temperature 20°C, floor surface temperature 30°C, ambient
temperature -5 °C and undergroud temperature 2°C. Total 20 Kg of CCH was filled
into 16 containers which holds about 1.25 Kg. The design of containers included the
concepts of enhancement of natural convection inside the containers which can
reduce the degradation of CCH and cold fingers which can reduce the super-cooling.
Aluminium pipes (6 mm diameter) filled with SrC^ 6H20 (SCH) are installed onto the
cold fingers in containers. SCH contacts CCH only at the small holes (2 mm diameter)
in the aluminium pipes and it could prolong the effects of SCH as a nucleator.
9cm
1. Aluminium cover
2. Rubber seal
1. Aluminium cover
2. Rubber seal
3. CCH container
A
Fig. 1.
Aluminium container for CCH
-------
2841
Test Cells and ONDOL-oanels
Two identical test cells sized 1.8x1.8x1.8 m were built to investigate the performance
of ONDOL-panel with CCH and without CCH. The overall heat transfer coefficient of
wall and roof in the test cells was 0.29 Kcal / m2 hr °C, since the scale modeling from
the regular size of the common room was considered. Schematic of ONDOL-panel
with-CCH and without-CCH was shown in Fig.2. The ONDOL-panel consist of
galvanized steel plate (1.2 mm thick), polyethylene tube for circulating hot water and
polyurethane for insulation.
Insulation (polyurethane)
Cover plate
(galvanized steel)
PE pipe
— Cover plate
(galvanized steel)
Aluminium CCH contanier
PE pipe
Backboard (zinc plated steel)
— Insulation (polyurethane)
Backboard (zinc plated steel)
ONDOL-panel without CCH ONDOL-panel with CCH
Fig. 2. ONDOL-panels without and with CCH
Experimental Procedure
The performance of ONDOL-panel with CCH was compared to that of ONDOL-panel
without CCH during winter heating season. 60 °C water was supplied to each
test cell with the flow rate of 0.7 lit./min, respectively. T-type thermocouples were
installed at 9 locations for room air, 10 locations for walls and 9 locations for floor
surface temperature. After the termination of charging heat for 2-4 hours depending
upon the test conditions, the characteristics of heat emission observed by
measuring temperatures. 11 tests were performed and analyses were
concentrated in view of temperature variations in the room.
RESULTS AND DISCUSSIONS
Figures 3 and 4 represents the temperature of room and floor surface for test cells of
ONDOL-panel without and with CCH, respectively. Hereafter ONDOL-panel without
CCH is reffered to Rm.1 and ONDOL-panel with CCH is reffered to Rm.2.
After the 2.5 hours of charging heat with 60 °C water, the room temperature became
about 22 °C. The floor surface temperature of Rm.1 reached sharply 45 °C and
that of Rm.2 reached slowly to 42 °C. After the termination of charging heat, the floor
surface temperature of Rm.1 dropped to 20 °C within 1.5 hours. But, that of Rm.2
took 7.5 hours. As shown in Fig.4, the floor surface temperature of Rm.2 maintained
above 29 °C for about 5 hours due to the latent heat of CCH.
-------
2842
O
w
«
E-<
O
40 H
30
~i—i i r t r l i i i i i r—1 i i i i i i i i i r
~ ~~~~ Floor surface
+++++ Room air
xxxxx Ambient
p2 20 ^ +^+++^++4
w
CL
w
E-i
oH
-10
+
+
+ +
xXXxxxxx^ I II llllll
x*xxxxxxxxx,
•xxxxxxxx
xxxxxxxx-i:
I I I | I I I | I I I | I I I | I I I | I 1 I
0 2 4 6 8 10 12
TIME, hr
Fig. 3. Average temperatures of the room and floor surface in Rm. 1
41444 CCH
~ ~~an Floor surface
+++++ Room air
xxxxx Ambient
30-
K 20-
-10 i i i | i i i | i i i | i i i | i i i | i i i
0 2 4 6 8 10 12
TIME, hr
Fig. 4. Average temperatures of the room and floor surface in Rm. 2
The temperature of CCH in Rm.2 shown in Fig.4 was measured inside of the
container. Temperatures were fluctuating within 2 °C. It is believed that electrical
noise caused this fluctuation because CCH is a strong electrolyte. However the
trends of absorbing and emitting latent heat at 29 °C were sufficiently observed.
Figure 5 shows the results of 11 tests in terms Of the necessary time to reach
20 °C of floor surface temperature for Rm.1 and Rm.2. It took only 2 hours at most for
Rm.1 to reach at 20 °C while Rm.2 took 4 to 10 hours.
-------
2843
12
11
10
9
8
7-
6-
5
4-
3-
2-
1
0

EZl Rm.1
ESI Rm.2
J
Fig. 5.
1 2 3 4 5 6 7 8 9 10 11
TEST NUMBERS
Elapsed time to decrease down to 20°C from the maximum temperature
Figure 6 explains the effects of ambient temperature to floor surface temperature . For
Rm.1 without CCH, the floor surface temperature was irrelevant to ambient
temperature since Rm.1 had almost negligible thermal mass in the room. The floor
surface temperature of Rm.2 exhibited a prolonged time to reach 20 °C as
ambient temperature increases.
Q
W
CO
w
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
-
a Rm.1

+ Rm.2
-
•«
~
~
+

~
~

c
a a p
~ n
a ~
- i i i i i i
1 1 1 1 1
Fig. 6'
-7 -5-3-11 3 5
AMBIENT-TEMPERATURE. C
Ambient temperature vs. heat emissioning time until 20°C
Figure 7 shows the floor surface temperature of Rm.2 with CCH. This depicts the
change of floor surface temperature in cases where carpet or rug covered the
floor. Aft&rc7.5 hourst a piece of insulation pad (50x50x2 cm) was
placed on the floor surface. The floor surface temperature rose up from 27 °C to 29
°C which is freezing temperature of CCH. This phenomena simulates the real living
situation in which a rug or Korean style bedding so called "IBUL" is used. As a result,
it turns out to be that CCH is a promising phase changing material for hydronic
radiant heating of which operating temperature is low , e.g., less than 60 °C. This
system may be a good combination with solar heating system where low
temperature water can be obtained.
-------
2844
t"1 20
1—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r
4 6 8
TIME, hr
Fig. 7. Surface temperature variation of the covered floor
CONCLUSIONS
The application of CaCI26H20 to hydronic radiant floor heating system was
investigated. The following results were found ;
0 In case the ambient temperature is -7°C ~ 7°C, heat emission of CCH in the
course of phase changing took 2 ~ 4 times longer than that of sensible heat
emmision.
0 CCH could be a promissing latent heat storage material for the low temperature
applications such as solar heating.
-------
2845
AIR WINDOW COLLECTORS
Charles H. Filleux
c/o Basler & Hofmann, Consulting Engineers
Forchstrasse 395, 8029 Zurich, Switzerland
ABSTRACT
Within the framework of IEA Task XI "Passive Solar Commercial Buildings" the application of air win-
dow collectors for commercial and industrial buildings has been investigated. The paper discusses
the general concept of air window collectors, refers to built examples and gives insight into relevant
design parameters. A TRNSVS module for the air window collector was developed to investigate the
sensitivity of various design parameters.
KEYWORDS
Air window collector; solar air collector; energy performance of air window collector; simulation tech-
nique; design insights into air window collector system.
INTRODUCTION
The double skin of glass as a place to accumulate and circulate solar heated air presents an intere-
sting concept when overheating from a conventional window would require the sunlight to be rejec-
ted. This concept of air flow windows has successfully been applied to a large number of residential
as well as commercial buildings all over Europe. The purpose of this work is to investigate in more de-
tail the energy performance of the air window collector in buildings with varying occupational patterns
(offices, meeting rooms, schools). Our study strictly applies for air window collectors only, although
many conclusions are also valid for supply air windows or exhaust air windows.
CONCEPT
The concept of air window collectors consists of a pair of spaced glazings, i.e. an exterior glazing
(usually double glazing for better insulation) and an interior glazing separated by an air gap from the
exterior glazing. For better thermal comfort double glazing is often used on the interior side also. A so-
lar absorber (e.g. a Venetian blind or solar absorbing roller blind) is located in the air gap. Air supplied
from the cold end of a storage unit enters the air space between the glazings. As it flows over the
sunwarmed absorber, it is heated. The warm air is then returned through the ducts to the thermal
storage. The heating efficiency of the air window collector benefits from the low temperature of air
supplied to the collector. If air is recirculated to the collector from the cold end of a stratified rock bed
storage unit, its temperature is approximately 20° C.
Air window collectors are also capable of providing privacy and glare control. Overheat protection is
obtained when the absorber is lowered and cooled by outside air which is then exhausted back to
the outside.
-------
2846
The main features of the air window collector are:
Solar heat collection
(active/passive)
Natural light
transmittance
Night-time insulation
Overheat protection
The air window collector is operated in an upward ventilation mode and is characterized by four
modes of operation according to the blind and damper positions:
a) Active collection (total radiation vertical south larger than 350 W/m2): With the blind lowered and
the fan operating, most of the accumulated heat will be transported by the moving air to the
storage. Only a small amount of direct radiation, together with some heat from the warmed inner
glazed surface enters the room directly.
b) Direct gain: When the radiaton does not reach 350 W/m2, i.e. when direct sunlight to the heated
space is acceptable, the blind is raised (automatically) and the fan stops.
c) Night: The blind is lowered to reduce the window heat loss and electrically operated dampers are
closed to prevent backward air circulation.
d) Summer time: The blind is lowered with its reflecting side facing the outside. Outdoor air cools the
absorber (preferably by thermosyphon).
Fig. 1. Two collection modes of operation of air window collector.
Left: active collection (a), right: direct gain (b)
-------
2847
ENERGY PERFORMANCE FOR OFFICES AND CLASSROOMS
Experimental Testing
Most air window collector applications in Switzerland have been for private homes. One exception is
the Haas office building in Jona/Rapperswil. It is a small office building (heated floor area = 213 m2),
earth sheltered and well insulated. All windows are south orientated to utilize passive solar energy.
During night-time, heat loss through the windows is reduced via manually operated insulating shutters.
The building is equipped with a relatively large, south facing collector area of 41 m2. The collector is
connected by ducts to an underfloor rock bed with a volume of 60 m3. The site built air window
collector is composed of two double glazings with a Venetian blind in-between. A fan circulates air
through the collector at either of two speeds (18 or 43 m3/h.m2con) depending on the absorber tempe-
rature. At the higher speed the fan power is 15.4 W/m2CO||.
Fig. 2. Air window collector in Haas office building seen from outside.
The whole building and system has been monitored and the results can be summarized as follows:
- The auxiliary heat requirement of the building is extremely low: less than 50 MJ/m2 of floor area.
This is about one tenth of the value in a conventional multi story office building in Switzerland.
- Good comfort conditions have been achieved, despite the rather large collector area.
- Electricity consumption for artificial lighting is larger by some 15 % than it would be the case with
direct gain windows.
- The storage volume proved to be too large by a factor of two to three and therefore storage
temperatures remained modest.
Detailed analysis of the air window collector performance over the heating season 88/89 gave the
following results: The total amount of energy collected by the air window collector during the active
mode (approx. 400 hours) was 3'380 kWh or 85 kWh/m2C0||. The average collection efficiency was
0.39 over the entire heating period (hours of active air flow only).
Simulation
A physical model (Fig. 3) of the air window collector has been developed and translated into a
TRNSVS module. The model uses an average temperature Tav for the air cavity, includes different
heat transfer coefficients for outside and inside losses and calculates the energy output from the
collector. It takes into account the heat capacity of the glazings and the frame and runs on the four
modes of operation described above.
-------
2848
SOLAR
RADIATION
iTAcoll
r
AIR FLOW TO STORAGE
mcp' 2 C^av " Tj)
•To
dir (Ag+Aw^T

EXTERNAL ENVELOPE
(DOUBLE GLAZING)
FR-taAcoii-Fr
^boi(Ag+Aw)
RADIATION
' TO ROOM
^top^coll
|WS^-AA/1
AIR GAP
w
z
s

INTERNAL ENVELOPE
(DOUBI .F. GLAZLNG)
AIR FLOW FROM STORAGE
Fig. 3. Physical description of air window collector.
Tav: temperature of node in the air cavity
Using the TRNSYS model described in Fig. 3 extensive parametric studies have been performed for a
70 m2 floor area classroom and a typical Swiss school building. They include the following: Size and
characteristics of the collector, responsiveness of control of the Venetian blind and airflow rate, orien-
tation and shading, pattern of occupancy, responsiveness of heating systems and controls, thermal
insulation standard of the building, amount of heat storage capacity and thermal mass of the building,
restitution of heat from the storage to the heated space.
Results of Parametric Studies
The total amount of auxiliary energy (heating and cooling; electricity consumption for lighting and fan
operation) is used to estimate the influence of each parameter variation. As an example result we
show the influence of collector size and construction (Fig. 4) on the auxiliary heat consumption. The
available algorithms allowed to study the number of glazings and the characteristics of the air flow
passing by the absorber. The shape of the absorber and the air gap geometry could not be
examined.
Glazing and
Standard Type A
Standard Type B
Construction Type
2XV
2IV
2IV
2IV
r 1

~
t
M

<- 21V / 21V Type A
¦O 21V / 21V Type B
A, 21V/11V Type A
~ 11V/21V Type A
-------
2849
Auxiliary Energy for Heating
[kWh / Heating Period]
600
500
400
300
Design Proposal
200
100
0
0 15 20 25
Glazed Collector Area per 100 W/K Heat Loss [m2Coll]
30
1
Fig. 4. Influence of collector size and construction on heating consumption.
Clearly, solutions with a single pane either on the interior or outer side are less energy efficient than
standard configurations.
Solar heat collection
Measured data and TRNSYS simulations have been used to determine the energy balance for an air
window collector for the heating period. For residential applications the value is approximately 140
kWh/m2C0||. For office buildings and schools it is lower (105 kWh/m2con) because of the higher
contributions of free heat from lighting and the occupants. The time of occupancy also affects the
efficiency of solar heat collection.
Energy balance of air window collector for schools
kWh/m^n-a
+120-
lo storage
balance
lo room
passive
'« +60
10 room
active
to ambient
Fig. 5. Energy balance of air window collector/pebble bed storage system for a class-
room in typical Swiss climate. Total active collection time was 485 hours.
-------
2850
DESIGN RECOMMENDATIONS
By designing an air window collector system several decisions must be made concerning the
collector size and glazings, air circulation and storage. The decisions affect daylight transmittance,
solar heat collection, overheat protection and dynamic insulation. These are briefly discussed below:
1 Allow for sufficient natural li
/
ght transmission
2 Ratio of a collector area to heated surface
area. Include conventional windows for
natural ventilation
3 Orientation to the west should be avoided
(overheating)
4 Use exterior solar protections in addition to
the Venetian blind to avoid overheating in
summer
5 For bette
will be u
ir
sc
the
id
/
/
/
/
/
/
rr
)n
rial insulatior
inner and a
d
ute
oub
ir s
/
/
5
/
/
le glazing
Lirface
(
6 The storage volume to collector area ratio
should not exceed 0,8 m3/m2col|
i fD
W y
1 M2 <0.8 M3
CONCLUSIONS
The analysis performed within Task XI have demonstrated that active solar heat collection by air win-
dow collectors and short time storage is a good alternative to the direct gain. Promising applications
include residential buildings, hospitals, office buildings and schools.
ACKNOWLEDGEMENT
The Swiss contributions to IEA Task XI has been supported by the Federal Office of Energy. Many
thanks to Markus H3nni for developing the TRNSYS module and performing the parametric studies.
REFERENCES
Filleux, Ch., and M. H3nni {1991). Task 11 Reference book - heating section, R. Hastings, editor, to
be published.
-------
3.11 Passive Strategies and Materials II
-------
Intentionally Blank Page
-------
2853
SIMULTANEOUS HEAT STORAGE AND UTILIZATION IN A
RAINING PARTICULATE BED HEAT EXCHANGER
A. Bashir and B.M. Gibbs
Department of Fuel and Energy
The University of Leeds
Leeds LS2 9JT England.
ABSTRACT
Due to the oil crisis in the 70's and the present Gulf crisis, there is a growing need to
develop alternative sources of energy other than traditional fuels. Advances in solar energy
technology can provide an abundant, affordable energy supply, globally, by ensuring a clean
environment. The present work is concerned with the development of a system in which heat
storage and its utilization can be achieved simultaneously. Experiments were carried out in a 1
m high vertical Raining bed heat exchanger; solids and gas were introduced from the top and
bottom, re spec tively; water was circulated through a centrally located copper tube. Sand sized
262 \im was chosen as a heat storage medium because of its chemical inertness and ready
availability. The efficiency of the system has been studied in detail, by employing internals,
consisting of aluminium hexagonal honeycomb discs and stainless steel pall rings. The rate of
heat transfer was found to be increased with internals, which is mainly due to the large surface
area provided by the internals and the turbulence created in the air path.
KEYWORDS
Heat storage;honeycomb disc;pall rings; particulate solids;sensible heat
INTRODUCTION
In the past few years, there has been serious concern over energy conservation schemes in the
industrial sector. The general objective of an energy conservation research and development
programme is to develop techniques and technologies which could enhance energy savings
and decrease pollution. Advances in solar energy technology can provide an abundant
affordable energy supply globally and ensure a clean environment.
The viability of solar energy can only be established by introducing a thermal store ,
otherwise, solar energy has to be used as soon as it is received. The annual solar insolation to
the earth's surface amounts to 5.4x 1018 MJ (Harker and Backhurst, 1988). But the technical use
of solar energy currently poses problems because of insufficient collection and storage. Solar
energy storage, in solids, can be divided into two categories, i.e. sensible heat storage and latent
heat storage; the present work is concerned with the sensible heat storage.
In sensible heat storage, the heat capacity of the storage medium experiences an increase in
temperature without undergoing a change in its phase. The main advantage of sensible heat
storage is its simplicity and the possibility of using the storage medium directly. Water is one
of the example of sensible heat storage; another attractive possibility for sensible heat storage is
to use a particulate solid. Harker and Kumar (1985), have presented a review of heat transfer
between gas and particulate solids, Mather (1962) has mentioned the use of pebbles for solar
energy storage.
-------
2854
Fig. 1. Schematic diagram of experimental setup.
1. compressed air; 2. water manomtter; 3. a.c. supply;
4. temperature indicator; 5. mild steel column; 6. water tank;
7. pump; 8. in-line heater; 9. copper tube; 10. honeycomb disc.
The present investigation was carried out to study simultaneous heat storage and utilization in a
gas-solid-liquid system. The purpose of this study is to determine the effectiveness of different
packing materials on heat transfer.
EXPERIMENTAL SETUP AND PROCEDURE
The experimental set-up consists of a 1 m high Raining particulate heat exchanger fabricated
from 4 mm thick mild-steel sheet (Fig. 1). The top part of the heat exchanger was connected
with the solids inlet assembly, whereas solids from the bottom section of the Raining bed were
collected in a fluidized bed. The Raining bed was equipped with a centrally located 0.028 m
diameter copper tube. Water and air flow were a in co-current direction, whereas particulate
solids flow is in a counter-current manner. The inlet of the water was connected with a
centrifugal pump through a rotameter. A by-pass line was taken from the outlet of the pump by
a tee junction for flow rate adjustements.
The study of the system was carried out at various liquid flow rates, with and without flowing
solids, by varying air inlet temperature and air flow rate. Four types of packings were tried, i.e.
three sizes of stainless steel pall rings (15 mm, 25.4 mm and 38.1 mm in diameter) and
aluminium hexagonal honeycomb discs. Honeycomb discs of 155 mm diameter were crimped
10 mm apart onto the copper tube. The inlet and outlet water temperatures, together with the
solids and air temperatures at various positions were measured, with a digital temperature
indicator equipped with a multichannel switch system.
-------
2855
xir4
9.00-
4.BH:
4.00:
3.81
3.01:
2.00:
2.00:
1.90k:
.80:
1.17
1.78 2.31
Solid Mas flux 0cb/i2b)
2.92
Fig. 2. Effect of solids nss flux on static hold-up
for honeycort disc at 6-0.
RESULTS AND DISCUSSION
In the absence of air flow, solids flow through the honeycomb appeared to be uniform and
smooth. The presence of the honeycomb discs did not alter the downward path of the
particulate solids. A negligible amount of solids was entrapped either in the central periphery of
the honeycomb discs or in the gaps between the outer edges of honeycomb discs and the
exchanger wall. Figure 2 presents a small value for static solids hold-up which is due to the
higher voidage of the honeycomb discs. The amount of static solids hold-up depends on the
angle of repose, packing voidage and the extent of solids internal friction. Hold-up values for
stainless steel pall rings have been discussed in a previous study (Bashir and Gibbs, 1991).
The pressure drop of a heat storage system is an important design parameter, and the pressure
drop caused by the solids flow in the Raining bed depends on the solids concentration and the
frictional forces exerted by the particles. In general, pressure drop may vary from zero upto a
maximum value corresponding to the weight of the solids hold-up. Figure 3 presents the
pressure drop results with honeycomb disc for various air flow rates. At low air mass flux, a
steady rise in pressure drop is observed, whereas at higher air mass flux, (> 1.61 kgm7s~l) the
increase is less than proportional. This is due to the fact that at higher air mass flux particles
become more scattered radially and some entrainment of particles is unavoidable with the exit
air stream.
To evaluate the performance of the system, its effectiveness was calculated over the entire
range of operating conditions. The effectiveness of packing is defined as the ratio of the heat
transfer to solids with packing to the heat transfer without packing. The lower curve in Fig. 4
shows the effectiveness value with a fixed distributor, whereas the upper curve represents the
value for a rotating distributor. In a previous study (Bashir and Gibbs, 1989), it is mentioned
that solids distribution is an important design parameter for raining particulate system. From
the results in Fig. 4, the rotating distributor is clearly superior over the fix ed distributor
because of a better dispersion for solids and gas. Furthermore, it can be seen that the
effectiveness decreases as the packing size is increased; this is due to less area being available
for gassolid contacting with larger pall rings. There is only a slight increase in heat transfer
-------
2856
LOS 10
Vi
a.ufti/to
1.06*|/£l
OO.OOkl/tti
1.61*|/fe
0.7E8t|/tf*
§
8
I
&
&
I I I I I I I I I | I I I I I t I I I | I I I I I I I I I | I I I i I I I » I | I I I I I I I I I I 1 1 I > I I I I I
.10 .» .40 .SB .70 .06 1
Air mm velocity (kg/«2s)
Fig. 3. Effect of air bus velocity on pressure drop
of the systea ultti honeycort hexagonal disc.
Xtt"1
18.00.
17.21:
tf.4t!
a.eoJ
§14.»
Il4.^
113.21:
—*— flrtatlm
—•— Flni

UJ
12.41:
u.eii
11.*
1111111111 ¦ 111111 < 111 ¦¦ 11111111111II111111111111111111 ¦ 1111
1.N 1.17 2.9 3.00 3.87 4.31 8.00
Packing size W w*
Fig. 4. Packing size versus heat transfer effectiveness
for rotating and Fixed solids distributors.
-------
2857
S
99
Fig. 5. Effect of air nss flux on heat transfer to water
Mith honeycoab at varioua air te^eratures, S-0, Nl-0.25 1/iin.
from air to water with pall rings packing, which is due to the poor contact of pall rings with the
copper tube. Contrary to this, with hexagonal honeycomb discs, in the absence of flowing
solids, there was a considerable increase in heat transfer to the water (Fig. 5). The resistance
offered by the fluid film surrounding the copper tube is reduced due to the turbulence created
with the presence of honeycomb discs. Average effectiveness (with honeycomb/without
honeycomb) for heat transfer to water was calculated to be 1.31.
CONCLUDING REMARKS
Sensible heat storage utilizing Raining bed is cheap and convenient for solar heating
applications. For a gas-solid-liquid system, for simultaneous heat storage and utilization, heat
transfer can be improved with hexagonal honeycomb discs. The pressure drop and static hold-
up are relatively low for honeycomb. Stainless steel pall rings showed an increase in heat
transfer to solids, but for water the increase is very slight. This suggests that a combination of
pall rings and honeycomb disc could be beneficial.
M.I aC
69.9 «
79.9 IC
HHIIHII»|M1*9t»IHIM»iailHMI»lllll
4.99 4.89 8.99 8.89 I.
Air aass flux (kg/i28)
-------
2858
ACKNOWLEDGEMENT
One of the authors (A. Bashir) would like to thank, Ministry of Science and Technology,
Pakistan, and State Cement Corporation of Pakistan for providing an opportunity for this
research.
REFERENCES.
Bashir, A. and B.M. Gibbs (1989). Heat Storage for Flowing solids in a Raining bed Heat
Exchanger. ISES, Solar World Congress, Japan.
Bashir, A. and B.M.Gibbs (1991). A Gas-Solid counter-current thermal storage system-
Hydrodynamic characteristics. ISES, Solar World Congress, USA.
Harker, J.H. and J.R. Backhurst (1988). Fuel and Energy, Academic Press.
Harker, J.H. and V.G.Kumar (1985). Energy transfer between gases and particulate solids.
J. Inst, of Energy.
Mathur, K.N. (1962). Heat Storage for Space Heating. Solar Energy, vol.6, no.l 10.
-------
2859
Electronic Structure of Iridium Oxide Electrochromic Thin Films
Kazuki Yoshimura, Masato Tazawa and Sakae Tanemura
Government Industrial Research Institute, Nagoya
I Hirate-cho Kita-ku Nagoya, Japan 462
ABSTRACT
Iridium oxide electrochromic thin films were prepared by reactive RF magnetron sputtering.
The dependence of the electrochromic performance of those films on the preparation conditions
has been studied. To make clear the cause of this, dependence, crystalline structure and
electronic structure of prepared films were investigated by using XRD and XPS measurements,
respectively.
KEYWORDS
Iridium Oxide; Electrochromism; sputtering; XRD; XPS.
INTRODUCTION
Recently Electrochromic devices haeattracted much attention as a solar energy control device in
glazings for passive solar system (Lampert, 1984). Among various electrochromic materials,
iridium oxide is one of the typical anode type materials. Iridium oxide shows rather good
electrochromic behavior concerned with durability and response time than those of WO3 which
are the most popular electrochromic materials for a prototype commercial device (Shay and
Schiavone, 1978). Moreover, iridium oxide is a 'non-color' material which means the color
of the film can be changes from black to transparent, while the color of WO3 being transparent
to blue. So iridium oxide is considered as a good candidate for next generation electrochromic
devices.
Iridium oxide thin films used to be prepared by various methods. Most typical methods are
anodically grown method (Gottesand, Beni and Shay, 1979; Sato and others, 1987) and
sputtering method (Schiavone, 1979; Kang and Shay, 1983), respectively. Some authors have
reported that the sputtering method has some advantage concerned with reproducibility and
durability of the performance of obtained films compared with anodically grown method,
though the detail of electronic structure and the mechanism of electrochromic behavior of
iridium oxide arc still in question.
In this study, we investigate the electrochromic structure of sputtered iridium oxide films and
its dependence of film preparation conditions on electrochromic behavior to clear up the factors
which determine their electrochromic properties. Electrochromic behavior of prepared iridium
oxide samples has been studied by conventional optical measurements and voltammograms.
After those characterization,we investigated crystal structures and electronic structures by
using XRD (X-ray Diffraction) and XPS (X-ray Photoelectron Spectroscopy), for the samples
which show characteristic electrochromic behavior.
-------
2860
EXPERIMENTS
Iridium oxide thin films weic prepared by reactive RF magnetron sputtering. Target was metal
iridium,and thin films were deposited on ITO (Indium-tin oxide) coated glass. The partial
pressure of oxygen and growth rate were controlled by adjusting total pressure, mixing of ratio
of oxygen to argon gas and induced power, respectively. Substrate temperature during
sputtering was controlled from 60*C (no-heating) to 200'C. The thickness of the deposited
films was monitored by crystal oscillator (INFICON), the thickness was calibrated by the
optical interferometer (WYKO T0P03D). "Die electrochromic bleaching and coloration of the
films were carried out with the aid of O.5M-H2SO4 as an electrolyte.
Optical measurements were performed using HITACHI U3400 optical spectrometer and XPS
spectra were measured by KRATOS XSAM800 system.
RESULTS AND DISCUSSION
As-sputtered films were light gray and their color could be changed by applying voltage using
H2SO4 as an electrolyte. Fig 1 shows the typical change of transmittance of the sample in
condition that growth rate was 30 A/min and substrate temperature was 60 *C. This sample
exhibits the best performance among prepared samples. For more quantitative analysis the
change of optical density of the same sample is shown in Fig. 2.
bleached
colored
500 1000 1500
Wavelength (nm)
c
o
Q.
o
1.2-
0.8-
0.4-
- bleached
colored
0.0-
—r-
500
T
T
1000 1500
Wavelength (nm)
Fig. 1. Transmittance of iridium oxide
electrochromic thin film in
bleached state (solid line) and
colored state (broken line).
Fig. 2. Optical density of iridium oxide
electrochromic thin film in
bleached state (solid line) and
colored state (broken line).
00.2
~
Growth
Rate
10 A/min
0
Growth
Rate
20A/min
•
Growth
Rate
30A/min
0 100 200
Substrate Temperature (°C)
Fig. 3. Dependence of AOD on the substrate temperature and the
growth rate.
Performance of obtained iridium Oxide films strongly depend on sputtering conditions,
especially growth rate and substrate temperature. We define AOD as the difference of optical
density at 1000 nm as an index of performance of electrochromic films. The dependence of
AOD on substrate temperature and growth rate is shown in Fig. 3. In this figure AOD
increases with the decrease of substrate temperature and with the increase of growth rate.
-------
2861
To obtain the information about the crystalline structure of prepared films, XRD measurements
were performed. Fig. 1 shows the XRD patterns of selected as-grown samples among iridium
oxide films prepared with various conditions. The preparing conditions of each sample, are
summarized in Table I.
Table 1. Preparation Conditions of Selected Samples.

Growth Rate
(A/min)
Substrate
Temperature (°C)
Sample A
10
200
Sample B
20
200
Sample C
10
60
Sample D
30
60
Figure 4 show the XRD pattern of each samples. Several peaks are observed in these patterns
but almost all peaks arise from indium oxide in substrate. Only peak at 34.2° is due to Ir02.
In Fig.4(a) this peak is quite sharp and it indicates Sample A consists of well crystalized IrC>2.
The intensity of peak due to I1O2 in Fig.4(b) and Fig.4(c) are smaller than that of Fig.4(a). In
Fig.4(d) there is no I1O2 peak. It shows that Sample D has amorphous structure.
Figure 5 shows the relationship between the peax height of I1O2 and substrate temperature and
growth rate. From this figure we can observe the trend that crystallinity of the film increase
with the decrease of the substrate temperature and with the increase of the growth rate.
<
£¦
<
30 40 50 60 20 30 40 SO
26 26
(c) (d)
Fig. 4.XRD pattern of the selected samples (a) Sample A, (b) Sample B,
(c) Sample C, (d) Sample D.
-------
2862

E
30
X
X
0
•
a
CC
20
X
0
0
JC
?
O
10
0
0
0
O

I
I
I
0 100 200
Substrate Temperature fC)
Fig. 5. IrC>2 Peak intensity dependence on substrate temperature and
growth rate, the diameter of the circle is roughly proportional to the
intensity of the peak from IrC>2, X indicates there is almost no peak
structure of I1O2.
To compare the electronic structures of these samples, XPS spectra were measured. Figure 6
shows the XPS spectra of the selected as-grown samples in valence-band region . Exciting
source was Mg Ka. The spectrum of Sample A is quite similar to that of single crystal I1O2
measured by Wertheim and Guggenheim (1980). It indicates that the electronic structure of
Sample A has much similarity to that of single crystal I1O2. The dominant peak at 2eV is the
structure of valence band of I1O2 which corresponds the t2g states. Broad structure from 4 eV
to 12 eV arise from O2p states and sharp structure around 19 eV is due to O2s state. The
Structure of O2p state in spectra of Sample B, Sample C and Sample D is gradually smeared
out from B to D compared with that of Sample A. These change may be related to the change
of crystalline structure.from crystalized structure to amorphous state.
Sample D
c
3
n
Sample C
<
(0
C
©
Sample B
c
Sample A
-3
0
1 0
0
1 0
Binding Energy (eV)
Fig. 6, XPS spectra of selected samples in valence-band region.
-------
2863
Sample D
Sample D
Sample C
Sample C
Sample B
Sample B
Sample
Sample A
/ \ Sample A
—' 1 '•
-530
Binding Energy (eV)
-75 -70 -65 -60 -55
Binding Energy (eV)
-540
-520
Fig. 7. Ir4/spectra for selected samples. Fig. 8.0Is spectra for selected samples.
Fig. 7 is the XPS spectra of h4f state and Fig. 8 is the XPS spectra of OIs State,respectively.
Ir4f spectra consist; of two peaks which are split by spin orbital interaction. The peak
position of each peak i s not changed among these samples. The peak position and shape of
OIs state are not changed among Sample A to D either. These results indicate that oxidi zation
state of each sample corresponds to that of IrC^.
CONCLUSION
Electrochromic performance of RF sputtered samples depend on the preparation conditions.
XRD analysis shows that the sample with low growth rate and high substrate temperature is
crystalized, while the sample with high growth rate and low substrate temperature has the
amorphous like structures. XPS measurements for those samples shows that the electronic
structure of each sample corresponds with I1O2 state. From these results we estimate that the
difference of electrochromic behavior mainly depend on the difference of crystalline structures.
REFERENCES
Gottesfeld, S., J. D. E. Mclntyre (1979). Electrochromism in anodic iridium oxide films. J.
Electrochem. Soc., 126.742-750
Kang, K. S. and J. L. Shay. (1983). Blue Sputtered Iridium Oxide Films (Blue SIROFs).
J. Electrochem. Soc., 130. 766-769
Lampert, C. M. (1984). Electrochromic Materials and Devices for Energy Efficient Windows.
Solar Energy Mater., H, 1-27
Shay, J. L. G. B., and L. M. Schiavone (1978). Electrochromism of anodic iridium oxide
films on transparent substrates. Appl. Phys. Lett., 22, 942-944
Schiavone, W. C. Dautremont-Smith, G. Beni, and J. L. Shay (1979). Electrochromic iridium
oxide films prepared by reactive sputtering. Apple. Phys. Lett., 25., 823-825
Sato, Y, K. O., T. Kobayashi, H. Wakabayashi, and H. Yamanaka (1987). Electrochromism
in Iridium Oxide Films Prepared by Thermal Oxidation of Iridium-Carbon Composite Film.
J. Electrochem. Soc., 134. 570-575
Wertheim, G. K„ and H. J. Guggenheim (1980). Conduction-electron screening in metallic
oxides: Ii02. Phys. Rev., B22. 4680-4683
-------
2864
THE INFLUENCE OF THE GEOMETRICAL DESIGN PARAMETERS
ON THE PASSIVE COOLING AND HEATING OF BUILDINGS
Professor Edna Shaviv and Isaac G. Capeluto
Faculty of Architecture & "fcwn Planning
Technion, Israel Institute of Technology, Haifa, Israel.
ABSTRACT
In this work a thorough analysis was carried out to find how sensitive is the thermal performance of
the building to its geometry. We concentrate mainly on four design parameters: building
proportions, orientation, shading and area of walls and windows. From the analysis that was
performed, we developed rules of thumb for the schematic design of buildings in hot, humid
Mediterranean climate. These rules can help the architect during the early design stages to figure
out the geometry of energy efficient buildings.
KEYWORDS
Building's proportions, Building's orientation, Energy Conservation, Load calculations, Modeling,
Passive heating, Passive cooling, Simulation, Sizing, Total Energy, Windows.
INTRODUCTION
During the early design stages the architect's main effort is to determine the building geometry. As
energy simulation tools require a detailed input of the building, which is unknown at the schematic
stage, they are left most of the time to advanced design stages. Often architects use,at the early
design stages,only very rough rules of thumb to guide them how to design an energy conscious
building.
The questions that arise are:
a. How sensitive is the building's thermal performance to the geometrical design parameters
such as: the building's proportions, orientation, shading, and area of walls and windows?
h Knowing the relative importance of the various geometrical design parameters, what is the
best building form that should be recommended?
The sensitivity analysis of the building's thermal performance to changes in the various
geometrical design parameters is carried out using a simulation model developed by Shaviv and
Shaviv (1977, 1978a,b). The influence of each parameter on the passive cooling and heating of
standard heavy constructed residential buildinga^that are usually Duilt by the Israeli ministry of
housing) was performed (see Figure 1). In this paper we present the results for the upper floor
apartment with four external walls. The thermal behavior of the building is checked during winter
and summer. The energy consumption of the building for heating and cooling is examined. The
results are summarized in easy to comprehend and implement graphics format.
upper floor
middle floor
first floor
a. Plan b. Section
Fig. l.The standard residential building, a. Plans of schematic apartment
with two, three, or four external walls, b. The building section.
The analysis of the influence of the building's geometry on its thermal performance was done
many years ago by Olgyay (1963). The results he got lead to the conclusion,, that it is important to
-------
2865
took for the energy efficient building geometry. Yet, his study was based on manual calculations
that include many approximations, like the steady state assumption, while ours i s based on an
hourly dynamic simulation. Our results are compared with the recommendations of Olgyay for hot
humid climatic zone.
THE BUILDING S PROPORTIONS
In his design guidelines for hot-humid region, Olgyay stated that: "Strong radiation effects on the
E and W sides should dictate the shape of buildings to a slender elongation. The optimum shape is
1:1.7, but up to 1:3.0 on the E-W axis is acceptable" (Olgyay, 1963, page 173 ). As Olgyay's
guidelines are still commonly applied, we would like to verify this design guideline. Using an
hourly simulation model instead of the Sol-air temperature and steady state assumptions that he
applied, may lead to more accurate results.
We investigate in this section the effects of the proportions on the thermal behavior and the penalty
for deviation from the optimum value. The first case is a non solar building with a fixed size window
on each elevation. The area of each window is 4% of the floor area. We changed the proportions of
the building from 2:1 up to 1:2, without changing its floor area (see Figure. 2a).
The results are:
a. Changing the building's proportions has very insignificant effects in the standard and
improved buildings.
b. In a poorly insulated building, like a building with concrete walls, the square proportion (1:1),
with minimum envelope's area, has a small advantage on oblong buildings. However, concrete
walls are beyond the recommended standards.
The second case is a non solar standard building with a variable size window on each elevation. We
kept the total area of windows equal to 16% of tne floor area as in previous case; however, this time
we changed the window's size so that a fixed relation of a window to aeievation is maintained (awindow
and aeievation are the area of window and the area of elevation^ respectively). The results are
presented in Figure 2b.
The results show:
a. When the windows have shutters, and therefore are well shaded in summer and insulated at
night, there is a little influence of the building's proportions on its thermal behavior.
b. When the windows are not shaded in summer and insulated at night, the effect of the building's
proportions can be large (a change of 1000 Kwh /year)
TEL AVIV • PROPORTION
TEL AVIV ¦ PROPORTION
without shutters
standard
concrete
with shatters
mproved
H Cooling
E3 Healing
Coding
Heating
i- »- w
I/) OJ

N m <¦
CM IT) »—
TEL AVIV • PROPORTIONS
Variable % ol south glazing

Heating

standard

1:2
' <•) o> w ® 'ejo>K>ig
»- »- OJ CVJ
a. b. c.
Fig. 2. The effects of the building's proportions on its thermal behavior.
The third case is a solar building with variable size windows on the south elevation and fixed size
windows on the northern, eastern and western elevations equals to 2% of the floor area (as is
recommended for standard buildings). The southern window changes between 3% of the floor area
and up to the maximum area that covers the whole elevation. A good night insulation and shading
devices in summer is assumed (See Figure 2c).
The results are:
a. For the same size of a southern window, the proportions ofthe building have a limited influence
on its thermal performance.
b. For the optimum size of a southern window, the proportions of the standard building have a
small influence on its thermal performance (less than 300 Kwh/year). This is because the
-------
2866
optimum size of a southern window in this climatic zone is 18% of floor area. Such a window
can be located in the building that is elongated in the E-W axis (2:1) or square (1:1). The
maximum area for a southern window on the N-S elongated building (1:2) is only 15% of the
floor area, which is the area of the whole elevation.
The small influence of the building's proportions should be compared with the great influence of
increasing the area of the southern windows that gives a change of 1300 Kwh/year (see Figure 2c).
Based on the above results we can conclude that for the hot-humid regions there is no need to define
very precise building proportions, in contrast to Olgyay's suggestions. The architect has the
freedom to design any building's proportions he may wish, as long as the building is not elongated
too much along the N-S axis (not below 1:2). A very narrow south elevation leaves no space for the
recommended southern windows.
EXTERNAL WALLS AREA
We found already that the building's proportions has little influence on the thermal performance.
The next question that rises is, therefore, "Is the building's geometry completely free? Can we have
lareer external walls without changing the building's energy demands?" Figure 3 shows that the
building's thermal performance is sensitive to the area of its external walls, even when we ixnjjrove
its insulation. In this figure we presented the energy demands of the same apartment unit having 2
to 6 external walls each with an area of 25 m2. (A mid apartment in an apartments' block has two
such walls, a corner apartment has three, while a private house has four. More than the area of four
external walls is found in buildings with zigzagging walls.)
The results are:
a. The increase in walls area increases significantly the energy required for heating and
decreases the energy required for cooling,
a The total yearly energy demand increases with the increase in the area of the external walls.
The conclusion is that free design may be allowed in this climatic zone, as long as we don't add too
much area to the external walls.
kwh
6000
5000
4000
3000
2000
1000
0
Fig. 3. The influence of the external walls area on the thermal behavior of the
building.
THE BUILDING S ORIENTATION
Olgyay stated in his design guidelines for hot-humid region that: "Sol-air orientation is balanced
at 5° E of South, with somewhat small deviation from it (10°) to remain desirable." (Olgyay,1963,
page 173)
We investigate here to what extent is the thermal behavior of the building sensitive to its
orientation, and how important is it to design an energy conscious building according to the above
suggested orientation? The orientation of the building is defined here as the orientation of the
elevation on which the largest windows are located. From design point of view it means that the
living room and probably more rooms face this orientation. We refer to this orientation as the main
one.
We started by assuming a large window (protected by shutters for summer shading) with an area of
10% of the floor area on the main elevation. The building is turned to eight different orientations
(Figure 4a).
TEL AVIV - WAIL AREA
standard
improved
W'coSii
Heating
ssr
-------
2867
The results are the following:
a. The orientation of the building has a great influence on the total yearly energy consumption of
the building. The energy saving is close to 1500 Kwh/year, which is about 50% of the total
yearly consumption of a standard insulated building, and 75% of an improved building. This is
due to the great decrease in heating demands. Let us mention that this saving costs no money,
as the cost of the building does not change with the orientation.
b. As we assume the windows to be well shaded in summer, there is a little influence of the
building's orientation on its cooling demands.
c. The behavior of the improved building is very similar to the behavior of the standard building
in regard to the orientation.
kwh TEL AVIV - ORIENTATION
m proved
(10% SW)
T2f Cooling'
Nj*aS»
standard
(10% SW)
5000-
TEL AVIV • SOUTH+-20 d*g
improved
standard
(8% SW)
(14% SW)
~"'Cooling
Heating
Fig. 4. The influence of the orientation on the thermal behavior of a building
with a large window on the main elevation, a. A non solar house, b.
A solar house with a suggested area of southern glazing.
Next we examined a solar house with a suggested southern glazing area of 14% (of the floor area) in
the standard building, and 8% in the improved one. We turned the main elevation in this
experiment from 20 degrees west of the south (S+20) and up to 20 degrees east of south (S-20). The
results are presented in Figure 4b.
The results show:
a. The orientation of the standard solar building in the above range has a negligible influence on
the total yearly energy consumption of the building (less than 200 Kwh/year). We can consider
this whole range of orientations to be preferable.
b. In the improved solar building the sensitivity of its thermal performance to the orientation is
more significant (about 500 Kwh/year). Here the best orientation is 10 degrees west of south
and up to 20 degrees east of south, which gives a change of less than 300 Kwh/year.
To conclude: Olgyay's recommendation of 5 degrees east of south (+-10) gives the range of 5 degrees
west of the south and up to 15 degrees east of it, agrees with present remits. But this range can be
enlarged with very little lost of energy, by 5 degrees in both directions.
SHADING
Wall Shading
Wall shading may be obtained by neighboring buildings, trees, or even by the geometry of the
building itself, as self shading. The last one is the most difficult to evaluate. The reason for the
difficulty is that we have to determine first the building's geometry to its last detail and then have
to introduce these details to the computer, which may be a tedious job. The question that arises is "Is
this input really necessary? Is the building's thermal performance sensitive to the wall shading?"
The answer is No," as we can see from Figure 5a. This means that the geometry of the building
may be free, and we don't have to worry about self shading.
Window Siimmpr Rhariinpr
The influence of window summer shading on the thermal performance of the building is presented
in Figure 4b. Although this design parameter may be a geometrical one, we suppose that window
summer shading is obtained by external shutters, which is a common practice in the hot-humid
Mediterranean climate. In figure 5b, the building has 10% southern windows (solar building),while
the remaining windows are identical to those in the basic case. We assume that the snutters
-------
2868
provide a shading coefficient of 90 to 50% for both direct and diffuse radiation, and a shading
coefficient of 50 to 10% for direct radiation, while the diffuse radiation remaiitf50% to allow the
daylighting. The Figure shows that summer performance is very sensitive to the shading
coefficient. Clearly a very good window shading is recommended in the hot humid climate of Tel
Aviv. We endorse a shading coefficient of 10% to 30% for direct and 50% for diffuse radiation. At
this range the thermal performance of both buildings changes by 300 Kwh only.
By comparing the results presented in Figures 5a and 5b, we may conclude that the building's
geometry that caused self shading is important only when the shade falls on windows.
kwh
6000
5000
TEL AVIV - WALL SHADING
TEL AVIV - WINDOW SHADING
4000
3000
2000"
1000-
standard
&
improved
VA Cooling
I
m
%
a e s 8 s • a s
8 k 8 8 5 8 k
%Dif©ct-diffuse radiation
a.
? ? ?
» 8 2
b.
standard
m
mproved
% 1 Heating
—
II
1 % $
2 % M 2 8 £
" %DirecWitfuse radiation
Fig. 5. The influence of shading on the thermal behavior of the building,
a. Wall shading, b. Window summer shading.
THE AREA AND ORIENTATION OF WINDOWS
We next investigate the preferred window size on each elevation. A standard insulated building
with variable size of windows between 2 to 10 % of floor area is examined. We assumed the use of
very good summer shading devices, as is a common practice in Tel Aviv. The results are shown in
Figure 6.
TBI AVIV - * NORTH H1N0MK
M CoiHM St-* >Mm» (l%!
"P5f
TIL KIN ¦ % SOUTH WIDOWS
k«* C
-------
2869
The results for the eastern and western windows are surprising. The interpretation might be that
any window size is adequate. However, one should be aware that we assumed the use of very good
summer shading devices, which lower the demands for cooling. Large windows on the west and east
orientations are not recommended. Similarly, small windows on the northern orientation are
advantageous. Southern windows in the standard building can be enlarged to 18% of the floor area,
and to 12% in the improved one.
SUMMARY AND CONCLUSIONS
In Figure 7 we show the changes in the total energy demand of the building due to the
variation of each geometrical design parameter along the whole possible range of values that
can be assigned to them. The recommended range, for which the total yearly energy change is
below 300 Kwh/year (which is less than 10% of the energy consumption of a well sealed and
shaded standard insulated building) is presented as well
TEL AVIV - Changes in total energy demands
Window summer S C. •»30tot0
Window summer S.C. -OOtolO
wall shading —any
wall shading -any
East & West windows
East & West windows
North windows
North windows
South g1a2ing:improved
South glazing improved
South glazingistandard —'
South glazing:standard
-omell
—any
-4to8%
-2to10%
-8to12%
-2to12%
l4to18%
2to18%
Orientation -S-20 to S+10
Orientation many
wall area:improved » 3wto4w
wall arearimproved - 2wto6w
wall areastandard - 3wto4w
wall area standard - 2wto6w
Proportions
Proportions
—2 1 to 1 2
-2:1 to 1 2
Q
Recomn
Possible
ended range
range
2000
Fig. 7. The alteration in the building's energy demands due to the change of the
geometrical design parameter throughout the whole possible range of values that
can be assigned to them, and the recommended range for each geometrical design
parameter, for which the total yearly energy change is about 300 Kwh/year.
Looking at the possible range for the geometrical design parameters, one can see that the thermal
performance of the building is very sensitive to changes in the windows summer shading and the
total area of external walls. Next in importance are the changes in the size of southern windows
and the building's orientation. The size of the northern windows has a noticeable influence as well.
When the building's insulation is improved, the influence of the total area of external walls, and. the
size of the southern windows decreases, but the influence of the window^ summer shading and the
building's orientations remain the same. On the other hand, the variations in the building's
proportion, wall shading, size or eastern and western windows, have very little influence on tne
thermal performance of the building, provided that the windows are well shaded in summer.
ACKNOWLEDGMENT
The research was supported by the Israeli Ministry of Housing and Infra-structure, Grant no 022-
474, and partially by tne S. Neaman Institute for Advanced Studies in Science and Technology.
REFERENCES
Olgyay, V., 1963. "Design with Climate - Bioclimatic Approach to Architectural Regionalism."
Princeton University Press. Princeton, New Jersey.
Shaviv, E., G. Shaviv, 1977. "A Model for Predicting Thermal Performance of Buildings*. WP
ASDM-8, Faculty of Architecture & TP, Technion. Haifa
Shaviv, E., G. Shaviv, 1978a. "Modelling the Thermal Performance of Buildings". Building &
Environment, Vol. 13, Pergamon Press Ltd., England, (pp. 95-108).
Shaviv, E., G. Shaviv, 1978b. "Designing of Buildings for Minimal Energy Consumption". CAD
Journal, Vol 10, No. 4, IPC Business FVess, GB, (pp. 239-247).
-------
2870
COOLING CEILING-POND IN HOT-HUMID CLIMATE
L. Sandoval, J. Pineda, R. Castaneda and
L. Sanchez
Facultad de Arquitectura, Universidad
de Colima, Campus Coquimat l=in.
Coquimat lcin, Col. 28100
MEXICO
ABSTRACT
In this work the results of experimentation with a ceiling-pond
in a hot humid climate to cool a room are presented. Two
identical rooms were constructed in order to measure the
performance of the room with and a room without the ceiling pond.
The experiment took place in Coquimatlan, Colima, a town located
in a hot humid region, nine kilometers from Colima city, Mexico.
Maximum temperature average is about 32.4 C, minimum temperature
average is about 18.6 C, mean temperature average is 24.6 C, and
the mean relative humidity is 67/C. There are 140.21 half cloudy
days, 97.73 completely cloudy days, 127.17 clear days, and 4.04
foggy days.
A series of four experiments were carried out testing different
conditions for the ceiling-pond. For the last experiment the pond
was covered with a floating insulation for 24 hours. To cool
water during the night, it was impelled through a sprinkler in
order to avoid the problem of covering and uncovering the pond.
KEYWORDS
Ceiling-pond; roof-pond; hot-humid climate; ceilig-pond
insulation; roof-pond insulation.
INTRODUCTION
Two modules, 1.50x1.50x0.94 m, simulating real single space rooms
were constructed (see fig. 1>. Both of the modules were built in
the same manner and were oriented in the same direction and
maintained similar spatial relations with their surroundings.
Similar materials were used for both modules: adobe walls,
concrete flat horizontal roofs with a brick ring around the roofs
as water containers, and ground floors.
A series of four experiments have been carried out in order to
test four ways to cool the interior of a testing room while
comparing it with a "control-room" which does not have a cooling
system, keeping all other conditions as similar as possible to
those of the testing room.
-------
2871
t-io. 1 The module. Two modules like this one were constructed.
The First Experiment
For the first experiment we kept the doors of both rooms open and
water in the ceiling-pond was kept without any insulation for 24
hours. Figure 2 shows a graph of temperatures at various times
of the day for the interior of the rooms, outdoors, and the water
on the roof of the testing room. Measurements were taken for
three consecutive days and the graph shows results for the second
day only.
We can see that under those conditions there existed a difference
of about three and a half centigrade degrees between the exterior
and the interior of the room with the ceiling-pond during the
hottest part of the day, while there was a difference of about
two degrees between the exterior and the interior of the room
without a ceiling pond. But it is most important to note that the
temperature in the climatized room was uncomfortable for about
only three hours, while the other room was uncomfortable for more
than four hours with a higher temperature (see fig. 2)
The Second Experiment.
In the second experiment we kept the same conditions as for the
first, except that the doors were closed during the day.
Measurements were taken for four consecutive days. In this case
the difference between the outdoor and the indoor temperatures in
both rooms increased
-------
2872
LU
LU
G_
S
LU
A...
/4"*' its"
/ /•'
lf^ A' .
J .X R
H « r^'
32.
30.
\
\
26.
24.
16.
lot
M-
hlQ. 2
for the
wa ksr.
1 1 1 1 1 1 1 1 I 1 1 1 I
s-nn qi'mi 1 no i|n pn 12< P:0 133 140 M"* l">,n l*>-3 1^:0 17-0 17 3 18-0 18;"* 19:0
TIME (HRS)
Results of the first experiment.A is for the outdoor,D is
testing room, O is for the control room and + is for the
temperature was uncomfortable for more than six hours; the
climatized room was comfortable at all times, although it reached
the Givoni's limit for people of developing countries on the
hottest day according to Givoni. However, people of developed
countries would have been uncomfortable for several hours; zero
for the best day and five hours for the worst one (see fig. 3).
O
LU
cc
o-
S
LU
F i Q.
?8
36.
3-1
¦>2
30
24
LL
-•(') '
18.
i
/-*
.n/
£ ..-K
7K'
A—

<
..n
¦ ,• i-j

.0"

•' 50 3 20 '3.20 10:il': 12.0 12.3 li.O 133 14 0 14 3 150 15.3 16.0 16 4 170 173 13.0 133 190 19 3
TIME (HRS)
Results of the second exDeriment.
-------
2873
The Third Experiment
The third experiment was made under the same conditions as the
second, (that is. one room with a normal roof, one room with a
ceiling pond, and both rooms with doors closed during the day and
open at night). However, we did change one condition: the pond
was insulated with a floating surface insulation during the day.
Under these conditions the temperature in the climatized room was
completely within the comfort zone, not taking into account
the humidity factor (see fig. 4). It must be noted that the
pattern of temperatures for the control-room remained similar in
all of the experiments.
UJ
cC
3
UJ
£U
?6.
34.
•>?
•JL. .
?0.
28.
26.
24.
22
20.
13
16®/
14.

-A
A o o
/

X
J*
a
r

/
I
I
V
1 r CT-S'
I - cv-0
/¦
I 1 I K
BSr13'
D P CH D UH~£ J
rl 1 1 ¥

1 1 k
H—h
6:207:208 209:3510.211 212.012 313.013.314:014:315:015:316 016:417.117:313:018.319.019:320 020:321:0
TIME (HRS)
Fig. 4 Results of the third experiment.
With the first three experiments we had proved that it is possible
to maintain daytime indoor temperatures well within the comfort
zone, according to Givoni. Then we had to prove that it was not
necessary to expose the entire surface of the pond to the
atmosphere in order to cool the water, since that would
necessitate the use of complicated systems to move the cover.
We had previously made some experiments with two similar small
containers, one of which was uncovered during the night while the
other one was covered with a floating insulation for 24 hours,
which also used a pump and a sprinkler during the night only.
Resulting in the temperature of the water in the covered
container at least as low as the minimun temperature of the day.
The uncovered container was able to reach temperatures up to 2
degrees lower than the minimum of the day.
We then did the fourth experiment with the modules.
The Fourth Experiment
-------
2874
For this fourth experiment we kept the floating insulation in*
place for 24 hours and used a pump during the night to
recirculate the water used for cooling (see fig 5). In fig. 6 we
can see that we got a similar pattern of temperatures to those of
the third experiment, that is; we kept the temperature well
within the comfort zone. In fig. 7 we have shown t^e results of
Fig. 5 The testing room for the fourth experiment.
the fourth experiment using the comfort diagram of Givoni to
include the measurements of the relative humidity. It can be seen
that almost the 24 hours of the day the conditions are of
comfort.
CONCLUSIONS
According to the experiments thus described, it seems that the
system is very promising for satisfying the cooling requirements
in hot-humid climates as^such in Colima, (in conduction with very
small dehumidifying systems for the most humid days of the year).
We still have to make experiments for the worst month of the year
(Nay), to estimate the dehumidifing needs under its conditions.
REFERENCES
Givoni, Baruch, (1990). Comfort Diagrams and design guidelines
for hot climates. Memorias del I Encuentro Nacional de Piseno v
Medio ftmbiente (in press), Colima, Colima, Mexico.
-------
2875
UJ
Q£
3
LU
Ou
.'A
-------
2876
Reflective Venetian Blind
A Multipurpose Element For Passive Solar Heating
S. Medved, P. Novak
Faculty for Mechanical Engineering, University of Ljubljana
ABSTRACT
Reflective blinds between two window glasses for solar radiation transmission
into a room were developed. Heat transfer and solar radiation transmission
ofawindow with "between" reflective blinds was determined. Also, the lamellas'
geometry was optimized. Through the simulation and the experiments, some
differences of heat transfer were discovered. For example, the heat transfer
of a window with a closed blind is 40% lower in comparison to the normal, no
night, thermally insulated window with ordinary no-reflectibe between-the-
glass blinds. The optimization of the lamella construction was made for the
solar radiation transmission into the room and was improved. With the adapted
program LOS-A25, the analyses of an annual heat savings were made. The com-
puter program for daylighting calculations or rooms with windows having reflec-
tive blinds was developed.
KEYWORDS
Reflective blind, heat transfer coefficient, transmittance of solar radiation,
passive solar heating, daylighting
INTRODUCTION
Windows are the thermally weakest element of a building, and have a major
influence on the rational use of energy and the thermal comfort of the occupants.
Efforts towards the construction of smart windows which give the optimum solution
to all problems are therefore natural. In Yugoslavia, more than 60% of windows
of all buildings are fitted with conventional double glazing. In developing a
new reflective Venetian blind between window panes, we tried to achieve a smart
window which simultaneously retained the advantages of a conventional blind
between the panes and made use of contemporary technology. These new jealousies
are characterized by a high reflective layer on both sides of the lamella. The
layer is as reflective for solar radiation as for thermal
radiation. (Fig. 1.) The geometry of the blind is reflected
into the depth of the room and accumulated in the ceiling to the maximum extent.
THE WINDOW AS AN ELEMENT ITSELF
Thermal Characteristics
To establish the thermal characteristics of the
Window with REflective Blind (WIREB), a
mathematical model of heat transfer was
1 reflectance (to)
111
i
thermal ( 2.6 * 26 um )
Fig. 1 Reflective blind-geometry and optical
dates
-------
2877
developed. The mathematical model is based on
solving the energy balance equation in a quasi
stadi condition for each element of the WIREB
(inner and outer glass, lamellas). The simulation
results revealed that: - open blind had no evi-
dent influence on thermal transfer through the
window; - with semi-open blind, there is only a
small reduction of thermal transfer (the dif-
ference U between open and semi-open blind -
= 60° - is 0.3 W/m K with reflective blind
and 0.15 W/m K with ordinary blind painted
white). The cause of this was examined by
means of convective flow visualization in a win-
dow, with and without blind. Some typical ex-
amples of stream lines are shown in Figure 3.
Figure 3a shows specific primary convective
flows a few mm from the window glass on the
hot and the cool side. With an open blind win-
dow (Fig. 3b), there is still a primary convective
flow but it is disturbed. A secondary flow ap-
pears inside the cavity between two lamellas,
which intensifies the convective heat transfer. In
these experiments, heat transfer reduction only
occurred through a reduction of radiation heat
transfer with the window with blind.
2.7
2.4
2.1
^ 1-8-1
3 1-2H
.9
.6-
.3-
F^3cz=0.9
Ilez=0.5
ez = 0.1
* H MM E3
EC
o 10 20 30 4-0 SO 60 90 90
tilt angle of the lammclas
Fig.2: Heat transmision coefficient of the win-
dow with blind
A B C
Fig.3: Stream lines in the window with blind
The influence of lamella distance was also
studied. Reducing the lamella distance (from
h = l to h = 1/2) caused a reduction of the heat
transfer coefficient by 0.1 W/m K, with open
blind. The difference is negligible with semi-
open blind. A greater reduction of the heat
transfer coefficient was found with closed blind.
Convective heat transfer in this case was deter-
mined in two ways in our model: - in the first
case, the closed blind were assumed to separate
the two half spaces (Fig. 2, model 1); - in the
second case, the closed blind had no influence
on convective flows and acted only as a radia-
tion shield (Fig. 2, model 2). This model is
recommended in other similar analyses in the
literature.
The mathematical model was tested on a
prototype WIREB on a standard hot-chamber
test loop. The comparison between the model
and the experiment, with open and semi-open
blind, showed differences of less than 4%. It was
found that open blind had no influence on the
heat transfer, since the measured heat transfer
coefficient of the WIREB with open blind and
WIREB without blind was the same (U
measure = 2.68 W/m K).
Differences between the heat transfer coeffi-
cients found by use of the models and by experi-
ment with a WIREB were found to be
significant. The simulated U value of the
WIREB using model 1 was 0.92 W/m K and
W/m K using model 2. Since the results of the
simulation of model 2 are similar to those
measured experimentally (U closed measure 0
1.69 W/m K, all analyses of the closed blind
were made by model 2. The stream lines con-
firmed these conclusions (Fig. 3c). It is obvious
that the characteristics of the WIREB with
closed blind are very close to those of a triple
window. The difference in the heat transfer
coefficients on the two models demonstrates
the possibility of improving WIREB thermal
characteristics by separating the two half spaces
in the window. In this case, the thermal charac-
teristics of a WIREB would be as high as those
of a high technology thermal window.
Solar characteristics of WIREB
Analyses of solar transmission for a WIREB
were made separately for direct and diffuse
radiation. Using simulation a, it was found that
the transmission of diffuse radiation was
highest with the lamella at an angle O = -15° to
-30° (Fig. 4). This angle is optimum regardless
of the lamella distance, but the highest trans-
-------
2878
mission was found at a lamella distance h = 1 (yodif
= 84.5%), or even higher than with a window
without blind (pdif = 82.5%). A ray-tracing
technique was applied to establish transmission
of direct solar radiation. By the use of a profile
angle which records the geometric relation be-
tween the window and the sun, the problem can
be solved two-dimensionally. The average value
of gravity functions of all restored rays repre-
sents the direct solar radiation transmission of
the WIREB. All absorbed and reflected flows,
as well as the direction of ricochet of solar rays,
can be determined by computer program. This
data can be used in the analysis of distribution
of direct solar energy into the room. Figure 5
shows examples of the graphic output of the
computer program.
south facing window with blind on an average
day during the heating season for Ljubljana city.
It is clear that the average energy transmission
of the window (also called the effective U value)
in a buildingin Ljubljana in the heating season
is 0.36 W/nrK for WIREB, 0.65 W/mTC for a
window with white painted blind and 0.93
W/m K for a window with black painted blind.
It is also clear that the new design of blind has
an advantage over classic inter-pane blind.
We concluded that although the WIREB is not
energy self-sufficient in the heating season,
since its g value is above zero, it is nevertheless
very comparable to well thermally insulated ex-
ternal walls.
wlndov
E22p*-o. IS
[ lpto^O. 62
BSBpvO. 82
-60 -4-5 -30 - 15
blind tilt (*))
Fig. 4: Transmitance of diffuse solar radiation of
the window with blind
tM ».< t
iu •: H.f i
(m *: <3.1 x
(m w:M.« x
tM»: M.7 x
(m •: cm x
Im w:*41
Fig.5: Ray-racing graf of sun rays
Annual energy balance of a window with blind
It is possible to assess the energy balance of a
window with blind by analyzing thermal and
solar flows. These can be established by using
the computer programs mentioned as a balance
of thermal and solar transmission. Figure 6
presents an example of hourly analyses for a
h ¦ 25mm, ! ¦ 18mm, d ¦ 3mm, ^ -¦ -5o, (foom - 20oC, tocation: Ljubljana
reflect ive^S-0.82,/9lR=-0.9),whiieps =0.0.62,/>IR-0.5).bl»d^>s-0.15,piR-0.1)
ref Uctl
wMta blind
black blind

jor f«b mor okt
«ec year
betting season
Fig.6: Annual energy balance of a window with
blind
THE WIREB AS A ROOM ELEMENT
The influence of WIREB on space heating ener-
gy
The thermal characteristics of a window greatly
influence the thermal comfort of the occupants.
Studies have shown that the characteristics of a
window have to be considered in conjunction
with the thermal characteristics of the apart-
ment. We used a LOS A25 computer code in our
studies, based on the original NBSLD computer
code. The modifications to the original program
enable the inclusion of night insulation of the
window, together with different distributions of
the solar radiation and solar transmissions of
the window as in the original program. A series
of ten-day measurements of meteorological
parameters was also used to check a built-in
program module for the calculation of solar
radiation on any orientation of surface. Average
-------
2879
differences in solar radiation are less than 3%,
except in cases of very low daily irradiation,
when the difference is greater. However, even
in this case, the absolute error is low (0.1
kWh/m day). The proper working of the pro-
gram with additional modules was checked by
experiment made on test cells at the ZRMK
institute in Ljubljana. Two by two cells were
provided with equal windows, one with reflec-
tive blind and the other without any blind (Win-
dow area =1/5 and and 1/7 floor area). It was
found that the WIREB used less energy for
heating the cells in the test period (19-29 April,
1989). The difference was 16% in the cell with
the largest window and 3% in the cell with the
smallest window.
Air temperature differences during the day in-
side the cell, surface temperature and con-
sumed energy for heating each of the cells were
compared to the computer simulation on the
LOS A25 program. A comparison between ex-
perimental and simulated results is shown in
Figure 7. The values were found to be similar
and the program was therefore used with the
additional modules to simulate apartments' an-
nual energy consumption in a multi-storey
building.
Test period 19.-29.*pril 1989
A window witb blind B window without blind C measurements D simulations
< AS
Rj-r
BS>^
BK
Fig.7: Measured and simulated temperatures in
cells
The influence of night insulation and the in-
fluence of various solar radiation distributions
on energy consumption was separately
analyzed with the simulation. It was found that
up to 14% of energy can be saved in a heating
season with WIREB compared to classic win-
dows. Savings are greatest with the largest win-
dows and greatest lamella interval. The
differences are greater in winter when the lower
values of heat transfer coefficient of the
WIREB are most significant. Differences are
lower in spring and autumn, when solar trans-
mission is less. It was also found that the annual
energy consumption for the selected room with
WIREB, with blindopen at night, is slightly
higher than such a room with an ordinary win-
dow the same size. We concluded that lower
diffuse transmission (for establishing lamella
angles appropriate for direct solar radiation
reflection) can be replaced by more effective
solar radiation distribution, but that the main
energy savings are shown by lower heat transfer
values.
Daylighting a room with WIREB
The luminous efficacy (lm/W) of daylight as
given in the literature /2/ was used in the
daylighting analyses. Luminous efficacy enables
daylighting data to be generated when only solar
radiation (W/m ) is known. A computer code
for establishing differences between the
daylight in rooms with classic windows and
rooms with WIREB was developed. With the
computer code, illumination with direct solar
radiation (using a ray-tracing model), illumina-
tion with diffuse solar radiation ana inner sur-
face reflection were separately evaluated. The
simulated results of the floor illumination of a
room with a south window are presented as an
example, with the following geometric
parameters: 1 = 25mm, h = 0.721, = 0°, profile
angle = 60°, angle of sun = 60°, window/room area
= 1/7, depth/breadth of room = 2/1. It was found
that the more uniform the illumination through
the depth of the room the better, and the best
results were obtained with lamella angles of
= +5° to -5° (daylighting factors (DF) in the
second half of the room are higher by up 50%).
A window - 1/5 A floor, h»25mm, l«25mm, ps=0.82
m N »
600 -
40C
200 -
ar -10 865:
Iz -5 87%
E23z +5 90S
cn 2 0 88%
I Iwindowimy.
jon f©fc> mor aor okt n
heatmg season
year
Fig.8: Mounthly consumption of heatig energy
-------
2880
Room with window Room with WIREB
100% 100%
0%
H
H
/6/ IEA Comparation of Six Simulation Codes - report
1987
/7/ J.Zupan£i£: LOS-AO Computer Code
/8/ P. Brunello: Advances in Solar Energy Technology,
Vol4. ISES 1988
/9/ R.M.Lebens: Passive Solar Heating Design, Applied
Science Publishers Ltd, 1980
/10/ A. Watt: Fundamentals of Three-Dimensional Com-
puter Graphics, Addison-Wesley Publishers Ltd., 1989
Fig.9: DF for different rooms
CONCLUSION
The reflective blind developed,with their good
thermal characteristics and simplicity, are a
cheap and effective element of energy savings.
They also contribute to better solar energy use
and correspond to Yugoslav standards (JUS
UJ5.610), which permit window blind within a
window or between the panes of a window only
where it improves solar energy use. We suggest
that our work continue with an optimization of
the geometry of the lamelias and the adjustment
mechanisms, to achieve an attractive window.
REFERENCE
/l/ S. Medved: Distribucija sonfne energije v prostoru s
pomoJjo refleksijskih zaluzij, Master work, Faculty for
Mechanical Engineering, Ljubljana 1990
/2/ J.P.Littcrair: The luminous efficacy of daylight - a
review, Lighting Research & Technology Vol. 17, No.4,
1985
/3/ S.Rheault & E.Bilgen : Heat Transfer Analysis in an
Automated Venetian Blind Window System, Journal of
Solar Energy Engineering, Vol 111, Nol, 1988
/4/ R.Siegel, J.R.Howell: Thermal Radiation Heat Trans-
fer, McGraw-Hill Book Company, 199
/5/ BJonson: Heat Transfer Through Windows, Swedish
Council for Building Research, D13:1985
-------
2881
DEVELOPMENT AND ANALYSIS OF BOTH PASSIVE AND HYBRID SOLAR COMPONENTS
F. Scamoni, I. Meroni, C. Pollastro and P. Tirloni
National Council of Research
ICITE - Institute of Building Technology
Via Lombardia, 49 - 20098 - San Giuliano - Milano - Italy
ABSTRACT
The earliest activities of research and development in the field concerning the application of passive and
hybrid solar technologies in buildings, have been undertaken in Italy since 1976; so far, the research in
this sector has passed through several stages of evolution up to the accomplishment of solutions that
are oriented either towards a semi-industrialized or an industrialized building production. In this sense
ICITE, the Institute for Building Technology of the Italian National Council of Research, has actively
contributed to the search of the possibilities that could be produced by the application of these products.
This paper, after a short summary about the evolution of the systems, shows up some components
(belonging both to the group of opaque systems and to the group of transparent systems) that, among
the many tested components provided by the building industry, have given meaningful results as for the
energy contribution and in view of their suitability to be integrated in the building. The experimentation
has been carried out by using an experimental set-up, conceived and realized by ICITE, with the clear
aim of characterizing and comparing the various systems and components with one another under real
working conditions.
KEYWORDS
Passive system, hybrid system, solar energy, building, test-cells, prefabricated component.
INTRODUCTION
Nowadays, one of the major goals of the building industry and architecture consits of controlling the
energy consumption with a view to the exploitation of the chances given by solar energy. Within this
context, the use of passive and hybrid solar systems for both dwelling and industrial buildings becomes
particularly significant. As we know, in passive solar systems the transfer of solar energy from the
absorber to the storage and then to the environment, takes place naturally, without the intervention of
mechanical means (pumps, fans etc.). On the contrary, these systems are called either active or hybrid.
The characterization of the thermal behaviour of these systems is one of the focal points of the research
carried out in all over the world that led to the definition of a method, known as Test-Cells, considered
to be a principle able to give meaningful answers upon the thermophysical performance of these
systems in real climatic conditions. This method consists of assessing a component through the
modification and the dynamics of the physical parameters of an environment towards which the
component itself acts as an energy-flux mediator It makes it possible to optimize and to compare these
systems for their integration in different building typology.
EXPERIMENTAL SET-UP
The principle chosen by ICITE focused on the aspect of use of the tested objects also for what concerns
the choice of the testing methods. Since the Test Cells principle was considered correct, it was decided
to use an experimental set-up suitably simulating the characteristics of a conventional building
environment. As a result of this situation, a series of main building criteria of the experimental set-up
were established; these criteria made it possible to realize 2,80 x 2,80 x 5 m heavy cells whose
characteristics simulate the conditions of useful thermal mass and thermal transmittance that can be
found in the most spread national building technologies.
-------
2882
This allowed the internal environment thermal characteristics dynamics to be drawn nearer to the
building's ones, without using sophisticated equipments or special air-conditioning devices. As
concerns ICITE's realization, the mediation tasks were assigned to the envelope, by simply getting its
basic thermal characteristics.
EXPERIMENTAL ANALYSIS
After monitoring the cells and the systems, the experimentations have been realized by introducing the
parameter Cg*, that is the energy needed to keep the test cell, on which the systems are fitted, at a
temperature level as similar as possible to the temperature level of a similar cell fitted with a reference
system. The energy contribution is measured over periods of time ranging from 1 to 5 days during
which the air temperature inside the cells as well as the energy consumptions have been recorded
(energy consumptions are considered as the electric power supplied to heat the cells, to the devices of
the different systems and to the instruments introduced inside the cells).
The energy contribution is expressed as follows:
Cg* = Q/(h • v • AT)
where:
- Cg* = energy contribution (W/m^°C);
- Q = energy absorbed during the measurement (Wh);
- h = duration of the measurements (hours);
- v = volume of the cell (m);
- AT = average difference between internal and external temperature during the measured period
in question (°C).
SOME SOLUTIONS AND EXPERIENCES
After setting up the method, by experimenting on systems with different configurations, a series of
components have been tested, which Italian firms as well as ICITE have planned and realized according
to the idea of high integration in the building. Hereafter a short summary is reported of the solutions
experimented by ICITE, that have given the most meaningful results as for the energy contribution and
in view of their suitability to be integrated in the building.
Wall-chimnev building system with radiating and ventilating false-ceiling
Fig. 1. Cross-section of the system.
The tested system, mainly consists of:
- an insulated closing wall;
- a plain-sheet, dark-coloured absorber treated in order to be resistant to heat and to solar radiation;
-------
2883
- a transparent covering with its relative frame;
- motor-operated valves and shutters;
- window with its relative shutter integrated in the system.
The operation of the system can be programmed in advance at fixed times; a special device stops hot air
entering the test cell whenever the temperature inside the cell itself exceeds the temperature value set
according to an environment thermostat. When the temperature difference (AT), measured between two
sensors, respectively set in the upper and in the lower part of the interspace, exceeds the value set
according to a differential thermostat, and the air temperature inside the test cell is lower than the one
that had been set:
- a fan, placed in the channel of the false ceiling, is switched on;
- the relative shutter is opened, allowing the hot air to pass from the absorber to the test cell through
the false ceiling.
When the difference of temperature measured between the two sensors is lower than the temperature set
according to a differential thermostat, the fan remains switched off and the shutter is kept in the closed
position. The system has been tested with a non-obscured window.
Experimental results
As regards the energy contribution, under conditions of good average daytime solar radiation, the tested
system allows a 30% saving compared to the reference. From diagrams of fig.3 it is possible to remark
a gain of the system also when the solar radiation is not particularly high.
W/rrv
LEGENDA
1 = Solar radiation
2 * Temp, inside the cell vith
the tested component
3 = Temp, inside the cell with
the reference component
4 - External temperature
10 12 14 16 18 20 22 24
Hours
Fig. 2. Parameters concerning a typical day.
~ Cell with reference component
~ Cell with wall-chimney building system
0,8
0,6
0,4
0,2
0
100
300
200
400
Solar radiation W/m^
Fig. 3. Energy contribution related to the solar radiation during the whole experimental period.
-------
2884
Wall-chimnev building system
n
IL
V4
n
Side without
thermo-ventilated unit
Side with
thermo-ventilated unit
Fig. 4. Cross-section of the system
The system mainly consists of:
- a solar absorber;
- a window;
- a concrete panel;
- a set of valves for the integration of the facade with the heating system;
- a heating system consisting of a 2500 W thermo-ventilated system.
Experimental results
The system has been operated according to four operational modes.
1) Non-working solar collector: this condition occurred during night-time or under conditions of poor
solar radiation when, as shown in fig.5, the room-temperature exceeded the temperature of the
absorbing-slab. In this case, it is easy to notice that the thermo-ventilated system goes into operation
(see the curve of the temperature of the cell fitted with the test component and the curve of the
temperature at the bottom of the chimney).
2) Passive heating: this condition was noticed during periods of good solar radiation when, meanwhile
the room temperature exceeded the set value, and the absorbing-slab temperature exceeded the room
temperature but not the limit one. In this case, as can be noticed following the trend of the room
temperature and the curve of the temperature at the bottom of the chimney, the thermo-ventilated system
goes out of operation, and the valve V3 opens, as shown by the temperature increase at the top of the
chimney. The natural convective flow is, however, rather slow, and, as a consequence, the thermal
exchange with the absorbing slab is weak (see curve of the absorbing-slab temperature).
3) In series heating: the chimney pre-heated air is conveyed into the thermo-ventilated system. This
condition took place during short times in the morning (between 9.00 and 10.00), when the room
temperature was less than the set value, and when the absorbing-slab temperature exceeded the room
temperature.
4) Solar chimney cooling: this condition took place when the absorbing-slab temperature exceeded the
limit one fixed at 65 ° C. The diagram of fig.7 shows that, with a sufficient solar radiation the heating
system does not work and the system is able to keep the room temperature at a value of 20° C without
any intervention of the thermo-ventilated unit; this is also due to the direct gain through the glazing.
-------
2885
W/rrv
2 900 "I
700 -
i—r
14 16 18
Hours
LEGENDA
1 = Solar radiation
2 = Temp, inside the cell with
the reference component
3 = Temp, inside the cell with
the tested component
4 = Air temp, inside the cavity
5 = Temp, of the absorber
6 = Temp, on the bottom of the
chimney
7 « Temp, on the top of the
chimney
8 = External temperature
Fig. 5. Parameters concerning a typical day.
Dynamic insulation prefabricated system
VN JV- > ¦ 1 ^
J
Fig.6. The systems under test.
This system was studied and designed by ICITE for building climatization; the prototypewas built and it
is still under test. The system particurarly designed to be integrated into a south-facing front of the
building, substantially comprises two modular parts: a prefabricated, multilayer opaque component, and
an energy management module wich forms a plant-engineering part of the system. The prefabricated
component basically consists of:
- an internal bearing layer made from lightened concrete;
- an insulating layer made from polystyrene foam sheets;
- an air-gap where the flow of the heated air occurs; provided in this interspace air conveying
means acting to channel the air flow so as avoid lateral recirculation of air,
- an external facing ancored to the bearing indoor layer made of composite material.
The wall has louvers formed in the lower part passing through both the bearing layer and the insulation
and, thus, communicating with the air space to ensure air recirculation. The energy management
module, is a modular component capable of providing management of heat flows through the dynamic
wall-system; such a module causes a bidirectional air flow between the internal and the external
environment and through the air-space of the prefabricated element to permit selection of
indoor-outdoor, outdoor-indoor, outdoor-outdoor cycles of flow. This module is a fully independent
unit in respect of energy demand and a fully automated, unit in respect to its operation. Energy
independence is provided by a properly dimensioned photovoltaic cells plant complete with storage
batteries. Automation is provided by an electronic control unit interfacing with heat sensing means to
-------
2886
decide operation of electrical parts according to preset programs. In the inside of its frame are located all
of the devices designed to provide ventilation, air flow conveyance and direction, as well as the means
for storing the energy that is necessary to operate said devices. More particularly, a hood is provided
which serves the purposes of collecting the air flow from the interspace, making it even and directing
the same either out- or in-wardly also by means of an axial fan acting to force the air flow either
towards the indoor- or towards the outdoor environment. The inwardly and outwardly directed open
ends of conduit are closed by respective gates which can be actuated for closing and opening them by
electrical pulses. The system can work according to two modes of operation that are preselected
depending on the season, as schematized in fig.7; namely, for gaining heat in winter and for removing
heat in summer. The winter operation mode is for the purpose of utilizing heating of air mass existing in
the interspace to transfer warmth to indoor environment. When temperature of air in interspace attains a
set value T* the energy management module acts to deflect air flow from the outside towards the inside
to admit warm air to the indoor environment through the fan which permits increased removal of the
heat from the interspace. Air from the outside will mix in part with internal change air which goes out
through louvers in the lower part of the wall, thereby to permit recovery of heat of the changed air
(fig.7a). In the case of temperature of air in the interspace being less than T* (night condition or day
condition with no sun), the energy management module will keep gates closed avoiding heat losses to
the outside (fig.7b). The louvers of the wall provide for normal ricirculation of air. In the summer
operation mode three dynamics of operation are provided:
- should the difference between the indoor temperature and the outdoor temperature be less than
2°C, the module will keep the outer gate open and the system will work as a normally ventilated
wall with, in addition, the possibility of air renewal (fig 7d);
- should the difference between indoor temperature and outdoor temperature be more than 2°C but
less than 4°C , then the module will open both the gates to permit a natural heat flow to occur
from indoor environment towards the outside;
- should the difference between indoor temperature and outdoor temperature be more than 4°C the
module, in addition to open gates will cause actuation of fan so as to force air withdrawal from
indoor environment until an indoor temperature is attained which is 2°C higher, at maximum, than
outdoor temperature (fig.7c).
E
a):winter with sun b):winter with no sun c):summer with internal d):ventilaied wall
overheating
Fig.7. Cross-section and modes of operation of the opaque prefabricated system.
CONCLUSIONS
The experimental methodology used, allowed to obtain all the parameters needed for the assessment of
the systems, some of which were probably too complex and did not lead to a better functional and
energetic performance. Nevertheless the outcomes of the researches witnessed the higher level of the
systems employed in the hybrid pattern (passive + active); this fact will undoubtedly contribute to
develop technologies involving the use of solar sources to fullfill high energy requirements of
buildings. A further development could arise from the latest systems which provide for the cooling as
well as for the heating, thus improving the overall comfort.
-------
3.12 Passive Strategies and Materials III
-------
Intentionally Blank Page
-------
2889
SOLAR PROCESSING OP SILICONE GLAZING
Arnold A. Valdez* and Cedtic G. Currin**
~Peoples Alternative Energy Services
Route 1, Box 3-A, San Luis, CO 81152, USA
**Currin Corporation, PO Box 1191, Midland, MI 48641, USA
ABSTRACT
Flexible and durable glazing can be made by coating glass fabric
with a silicone resin in a low cost coating tower using solar
process energy. This glazing and the processing are appropriate
for remote high altitude developing regions where glass is not
available as a glazing material for solar energy equipment. This
coating process was demonstrated, producing translucent glazing
1.0 meter wide. The projected glazing cost is about $8-10/m2.
KEY WORDS
Glazing; silicone glazing; glass; solar process heating; remote
solar applications; developing countries; silicone resin; curing;
coated fabrics.
NEED FOR GLAZING MATERIALS
The use of solar energy for meeting many basic needs of the
world's poor populations is widely recognized as a tool for
encouraging sustainable self-reliance. However, the lack of
glazing, a key material for constructing solar energy equipment,
critically limits solar energy applications in many communities.
Although very common in industrialized countries, glazing mater-
ials, particularly sheet glass, are expensive and scarce in
many remote areas; over 500 million people in developing nations
lack access to large glass sheets. Since glass is not made in
most two-thirds world countries, many solar energy applications
have been inhibited. For example, more than 200 million people in
Africa live in remote regions where imported glass has to be
transported hundreds of kilometers over poor roads resulting in
exorbitant costs.and considerable breakage.
A common second choice for glazing, polyethylene film, is widely
available in developing nations. Although the cost is low, this
film is inferior to glass, particularly in its resistance to
weathering. Polyethylene film degrades rapidly from intense
ultraviolet radiation at altitudes above 1500 meters where the
useful life of polyethylene film is typically less than one year.
An estimated 250 million people live at these high elevations,
many in rural communities.
-------
2890
CONCEPT OF SILICONE GLAZING
In the late 1970s many sought alternatives to glass as glazing
for solar energy equipment. This prompted consideration of
silicone resin coated glass cloth, a new material then under
development by Dow Corning Corporation for other applications.
This flexible material is made from two non-petroleum based
products: the glass fibers in the supporting fabric, which imparts
physical strength, and the silicone resin, which forms a durable
translucent film with optical properties similar to glass. The
coated fabric is made for large tensioned structures, such as
roofs for stadiums, by a capital-intensive industrial process.
A thinner translucent silicone coated glass fabric, called "sili-
cone glazing," can be produced using a simple, low cost process on
a small scale. For the past ten years, a core of grassroots
appropriate technologists sought to simplify the basic commercial
coating process so that silicone glazing could be produced in
remote developing regions. The initial demonstration of the
simplified coating process by Rojas, on a small batch basis, was
reported by Shepard (1981). Subsequent efforts were placed on
developing equipment and a process for continuous production of
silicone glazing, using only solar energy for the process heat.
SOLAR COATING PROCESS
Silicone glazing is made by coating and curing (crosslinking)
clear silicone resin on open weave glass fabric. The lenoweave
glass fabric, typically 0.1 mm thick with about 2.5 x 2.5 mm
openings, imparts dimensional stability and high tensile strength
to the silicone glazing. Glass fabric is relatively inexpensive,
widely available and is easy to transport in 1.0 meter wide rolls.
A liquid silicone resin was selected for coating the glass fabric
since these resins have high solar energy transmission, are
exceptionally resistant to weathering, are flexible over a wide
temperature range, and are chemically compatible with glass
fabrics. The resistance to ultraviolet radiation is particularly
important when the silicone glazing is used at high altitudes.
The selected silicone, Dow Corning® 1-2577 Conformal Coating, is
cured at temperatures as low as 80°C, has good coating character-
istics, yields a hard coating surface and can be used safely.
The continuous coating process, shown in Fig. 1, involves slowly
passing the glass fabric through a bath of the silicone resin and
then vertically up through a solar heated chamber to cure the
resin on the fabric. The equipment is called a "coating tower."
DEVELOPMENT OF CONTINUOUS SOLAR COATING PROCESS
The continuous coating process was developed in 1986-7 as a joint
project of Peoples Alternative Energy Services of San Luis,
Colorado, and Ghost Ranch Conference Center near Abiquiu, Mew
Mexico, USA. The process developed at Ghost Ranch (36°H latitude,
2000 meter elevation) uses only solar energy and low cost coating
equipment that is largely made from readily available materials.
The continuous coating process was demonstrated in a sturdy 4x4
meter adobe building about 5.5 meters high shown in Fig. 2.
-------
2891
COATKD h'ARRIC
(GLAZING)
Fig. 1. Cross section of solar coating process.
The well insulated, vertical, 2.4 meter high solar heated curing
chamber was built into the south wall facing the sun. A manually
adjusted 2 m^ area heliostat reflects additional solar energy into
this curing chamber. The 1.0 meter wide glass fabric unwinds from
a roll, passes through a dip pan filled with the liquid silicone
resin and then goes up through the vertical curing chamber. The
coated and cured fabric (silicone glazing) is cooled as it leaves
the top of the chamber; it is then wound on a take-up roll. Using
a photovoltaic-charged battery, a small motor powers the glazing
transport system. A roof ventilator removes the solvent vapors.
Fig. 2. Structure for coating silicone glazing.
-------
2892
The resin solution concentration, about 70 wt% in toluene, and the
speed of coating, typically 75 mm/minute, control the glazing
thickness. The chamber temperature, 80 to 120°C, and the coating
speed control the resin cure. Dampers on the chamber and manual
adjustments of the heliostat control the cure chamber temperature.
This initial coating facility was constructed using native mater-
ials, about $2600 of purchased materials, and nearly 2000 hours of
labor of many volunteers. The major items that would likely have
to be specifically imported for construction of this coating tower
in a developing nation are the insulated glass for the curing
chamber and the photovoltaic panel for the drive motor. Native
and scrap materials, including adobe blocks for walls and parts
from junked cars, were used as the primary materials for building
the coating tower and the fabric drive mechanism.
The initial coating experiments were made using Type 1610 leno-
weave glass fabric 0.3 meter wide. In August, 1987, the first
short length of 1.0 meter wide fabric was coated to a thickness of
about 0.25 mm. During the subsequent three years, over 50 m2 of
silicone glazing were produced, most frequently for demonstrating
the coating process. No difficulties were encountered, and three
non-technical people were individually trained for coating tower
operation. Over 25 m2 of silicone glazing was produced for the
1989 University of Colorado Solar Hogan Demonstration Project.
Major concerns in developing the continuous coating process were
the simultaneous adjustments of the resin solution, coating
speed and curing temperature. Coating speeds of 4.0 to 4.5
meters/hour and maximum curing chamber temperature of 100 to
120°C were routinely achieved. On days when the sky was mostly
clear, the coating was continuous for about 5 hours starting
at about 11:00 am after the initial heating of the cure chamber
for 2 hours. Cloud cover for up to 15 minutes did not seriously
affect the coating process due to the thermal mass in the cure
chamber. The coating process, including heliostat adjustment
every 15 to 20 minutes, requires only one pe.rson. At 36°N lati-
tude, the heliostat provides an estimated 30 to 40% of the energy
for curing the resin. Table 1 lists the processing conditions.
TABLE 1. Process Conditions for Coating Silicone Glazing
Chamber height,
meters
2.4
Equipment cost,
$ 4,000
Curing section,
meters
1.5
Depreciation*,
$/year 800
Cure time,
minutes
15
Maintenance,
$/year 400
Coating speed,
m/h
4.5

Coating time,
h/day
5.0

Glaz ing,
m2/day
22.5

Operation,
days/year
200

Glazing,
m2/year
4,500

* 5 year depreciation, since further process advancements may
warrant major changes in the coating tower within 5 years.
-------
2893
CHARACTERISTICS OF SILICONE GLAZING
Van Wert (1981) showed the translucent silicone glazing has a
solar energy transmission of 85 to 90%. Accelerated weather tests
conducted at an outdoor test facility in Arizona, USA, (Zerlaut,
1975) showed that this glazing may withstand 50 years of weather-
ing in arid climates. An early specimen of commercially coated
silicone glazing that was installed on a 4 m2 south wall solar
collector in New Mexico, USA, at 2100 meter altitude, was still
serviceable after 10 years (Stickney, 1990). Silicone glazing can
be rolled up, hauled over the worst terrain, easily installed in
solar energy equipment, and repaired using silicone adhesives.
Protection from forming a crease is necessary, however, to
maintain the tensile and tear strength of the glazing.
COST OF SILICONE GLAZING
The total cost for silicone glazing is about $8-10/m2 when this
process is conducted in regions where skilled labor rates are
$1.00/hour or less and solar radiation is abundant. This cost is
competitive with sheet glass in most cities where glass is avail-
able. The raw materials are the major expense, about $6/m2 of
glazing, primarily for the silicone resin. Table 2 lists process
data and economic estimates for producing silicone glazing.
TABLE 2. Estimated Cost of
Glazing from
Solar Coating
Process
Item Quantity
Base
Annual Cost*
Cost/m2
Silicone resin 0.25kg/m^
$22.00/kg
$
24750
$
5.50
Solvent 0.2 L/m2
$ 0.50/L

450

0.10
Glass fabric 1.2m2/m2
$ 0.50/m2

2700

0.60
[Materials sub-total:

$
27900
$
6. 20]
Energy None

0.00
Personnel 1
$8.00/day

1600

0.36
Equipment
$6.00/day

1200

0.26
[Manufacturing direct costs:

$
30700
$
6.82]
Overhead
$7. 00/day

1400

0.31
Financial $15,000
15% interest

2250

0.50
[Total manufacturing costs

$
36330
$
7.57]
Materials transportation
$1.00/kg

2010

0.45
Import duties
10.0%

2790

0.62
Total cost including importation charges
$
39150
$
8.70
Based on minimal resin waste and selling of all glazing.
BENEFITS FROM SILICONE GLAZING TECHNOLOGY
Solar processed silicone glazing is particularly appropriate for
remote high mountainous regions. It provides large area durable
glazing that is critical for most solar-thermal systems such as
water heaters, trombe walls for space heating, solar ovens, food
-------
2894
dryers and greenhouses. The coating process provides skilled
employment and business potential with a minimum need for capital.
The coating process and materials, when used with care, do not
harm people or the environment. It demonstrates the use of solar
energy in a manufacturing process, thus eliminating the need for
imported fuels or other conventional energy resources. Both raw
materials are much easier to transport than fragile sheet glass.
Regardless of the application, silicone glazing provides a way
to achieve greater self-reliance for those who never have had
access to glazing or solar energy equipment.
Furthermore, the basic solar coating process may be suitable for
specialized coating of other fabrics for roofing, water catch-
ments, water storage and agricultural applications. Substantial
further development might yield a process for producing paper.
Development of new products would be based on the skills developed
in coating silicone glazing and on recognition of special regional
needs for flexible coated materials.
Table 3 lists factors that are critical for the appropriate use of
this coating development.
TABLE 3. Requirements for Coating Silicone Glazing
1. Recognized solar energy needs in region with altitude
over 1500 meters and no sheet glass available
2. Local market for 4000 m^/year of silicone glazing
3. Customers with ability to pay $9/m2 for durable glazing
4. 18 MJ/m2 daily insolation for at least 200 days/year
5. Innovative entrepreneur with minimum primary level education
6. Resources for tower construction and working capital
7. Patience during tower construction, process start-up and
introduction of silicone glazing to market
8. Rapid communications channel to USA technology source
ACKNOWLEDGEMENTS
This development was nearly entirely a volunteer effort of dozens
of committed people. The many special efforts of Aubrey Owen,
Maria Valdez, Ralph Boone and Ed Schmidt were particularly
critical to this project. The substantial support by Ghost Ranch
Conference Center in providing the site, materials and support
services and by Dow Corning Corporation in providing the silicone
resin and glass fabric are also gratefully acknowledged.
REFERENCES
Shepard, M., and Z.C. Rojas (1981). Solar glazing for developing
countries. Sunpaper, 6^ (12), 21-25 (Dec).
Van Wert, B., and C.G. Currin (1981). Silicone-glass cloth for
solar glazing. J. Coated Fabrics, 11 (2), 92-99 (Oct).
Zerlaut, G.A. (1975). Accelerated outdoor weathering employing
natural sunshine. Proceedings 21.st Annual Meeting, Inst, of
Environ. Sci. Anaheim, CA. 153-159.
Stickney, B. (1990). Private communication.
-------
2895
TRANSPARENT HEAT INSULATING COATINGS OF Ag-SnOa ON A GIASS
Yu Shan-qing, Zhang Xiao-ping, Zhou Guo-ping and Ma Min~wei
SHANGHAI INSTITUTE OF ENERGY, SHANGHAI ACADEMY OF SCIENCE
777-29 Laohumin Rd. SHANGHAI 200237, CHINA
ABSTRACT
The transparent heat insulating coating of Ag-SnO* systems coated on a glass has been
studied. The relationship between theraal current and temperature through the windows,
as well as the efficiency of thermal insulation in the different constitutions of the
windows for the heating process were tested.
KEYWORDS
Ag-SnOa thin film; transparent heat Insulating coating; spectral selective coating;
spectral transmittance.
INTRODUCTION
The transparent heat insulating coatings(THIC) are spectral selective materials with
a better optical transmission in the visible wavelength and a high reflectance in the
infrared range. According to the need of the location or season, there are two mate -
rials: the solar control film and the heat mirror, they are important energy conser-
vation materials of windows in the buildings (Ge Xin-shi, Jun Bao and Yu Shan-qing,
1 980; Yu Shan-qing, 1990 ).
THIC have been systematically studied in SHANGHAI INSTITUTE OF ENERGY, it was found
that some D M unsymmetrical film systems consisted of SiO/Me, SnOi 'Me, CeOj/Me, ZnS'Me
and CdS>Me shown the higher trans (mission, the wider width of the pass-band, the
better spectral seIectivity(Yu Shan-qing and colleagues, 1988,1989,1990).Recently, the
optical properties of Ag-Sn02 film systems have been detailly studied: it was observed
that the dielectric layer of SnOj can increase the transmittance for whether unsymme-
trical film systems of SnOj/Ag, or symmetrical film systems of SnOj Ag/SnOj.The thin
film consisted of Ag-SnOj have maintaned a sat is factary stability under thermal
air condition(Yu Shan-qing and colleagues 1991). In this paper, the optical transmi-
ssion, thermal reflection and heat insulation of the coatings of Ag-Sn02 on a glass
will be reported.
OPTICAL PROPERTIES OF THIC
As mentioned above, the important parameters in evaluating performance of THIC from an
energy standpoint are the visible light transmittance (Ty) and the infrared reflec-
tance (Rjfl), which can be given by the expressions(Berining> 19 83 ):
j;
T(^)V(^)S(X) dX
# (1)
j*V(*OS(X)dX
-------
2896
j# [1-R(^ 6. )] sin 20.de. dX
)" W
-------
2897
ture is 64 Z! during 6 hours the average growing rate of teiperature is 7.0eOhr.
(4) For the window consisted of a THIC coated glass (the glass towards outside)and
a glass with lOmn air layer between thea. (G/F+G). the equilibrium temperature after
6 hours heating is 41.5"C. and the average growing rate of temperature is 3.30C''hr.
aor
Tue (tir)
TIm Uw)
Fig. 2. Heating rate of four arrangements of windows.
The scheme of heat transfer in four arrangement of windows were shown in Fig.3. The
formulae of the thermal current through windows in unit interval are as follows:
i. A single glass (G)
Tin T out T j
Qq, = = —J
Rc,i» +Rt,o«t +
' Ji»*.A
2. A THIC coated glass (G/F)
T/n ~ T *Mt

_!£_ + __L_
kj .A h«,.«t.A
T »nt

in +Rj^ +Rt,«t
h», in • A
+ *U-
kJ -A

kf.A

3. Double glass (G+G)
T/* ~ T#i/t ~ T0lft
----- +A- + -"
-------
2898
kf — heat conducting coefficient of glass. W/m'.K;
kf — heat conducting coefficient of film. W/m'.K;
k« — heat conducting coefficient of air. W/m'.K;
Vf — the thickness of glass, mi
Wf - the thickness of film, m;
VI4 — the thickness of air layer between two pieces of glass, mi
A — the area of the window. ma.
$U.t
1
J tout "ta
Ca) (c)
(a)

Fig. 3. Scheme of heat transfer and theraal network of
four arrangements of windows.
Owing to the film is thiner and its heat conducting coefficient is high, so heat con-
ducting resistance of film is less. i.e..
«f •
z Q
kj. A
therefore.
Q*= Q
From here it can be shown that the function of THIC is mainly depending on this high
infrared reflection to reflect the long wavelength energy of the incidence radiation.
The calculated formulas of heat transfer through the windows are as follows:
the windows for G or G/Ft
Q = 0. 1 308< T;» -T„* >
the windows for 2G or G/F+Gi
Q = 0. 083 ( T;„ -T..t )
The calculated results of the theraal current through the windows of four arrangements
under different time are summari. ed in table 1 and Fig.4..
Table I The difference of teaperature and thermal current through the

window in
four arrangements

A **
Time (hr.)
1
2
3
4
5
6

A T
35
43.5
46
47
47
47
G
Q
-------
2899
—rfr z!j — A"' -£r
Fig. 4. The difference of temperature and thermal current
through the windows in four arrangements
In order to evaluate heat insulating efficiency of THIC coated on glass. the following
formulars can be used.
1. The heat insulating efficiency of single glass with a coating comparing with a bare
glasst
=
2. The heat insulating efficiency of double glass comparing with a glass:
% = x 100* (8)
Qq
3. The heat insulating efficiency of single glass with a coating comparing with double
g lassi
1» =
—1 Hl x ,00*.
(9)
2$
4. The heat insulating efficiency of double glass with a coating coaparing with
double glasst

(10)
The calculated results are shown in table 2.
Table 2 Heat insulating efficiency of four arrangements of window.
No.
IfliKKhr. )
1
2
3
4
5
6

1
1,
60
54
52
51
49
47
52
2
h
49
49
46
46
45
43
46
3
7»
21
10
12
9
8
6
11
4
7«
(1
56
56
55
55
54
56
The experimental and coaputed results mentioned above ; show that the window cons-
tructed by THIC coated glass and a glass exhibits the best thermal insulating efficien-
-------
2900
cy. The average thermal increasing rate for side of low temperature is the smallest/
only 3.3°Ohr. The average thermal insulating efficiency( 74) is the highest. Contrast
with double glass* it s increase is up to 56*. considered comprehensively with techni-
cal and cost parameters/ The optimized insulating window consisted of a glass with a
layer of THIC. It s thermal average increasing rate is 4. 2'C/hr/ near the rate of dou-
ble glass with a THIC film between them. The thermal insulating efficiency of one coated
glass is 52* higher than one glass and is 11* higher than double glass. The test shows
that the thermal insulating efficiency of one glass coated with a THIC is better than
double glass window.
CONCLUSION
1. Ag-SnOj film shows better spectral selective property. The average trarismission in
visible wavelength range is about 55.4*. The peak transmission located on 0.55um is
about 75*. The film exhibits better heat reflection/ it's infrared reflectance is
higher than 90* in 2.5um wavelength and longer.
2. The heating test of four different arrangement windows shows that the glass coated
with Ag-SnO^ film will largely decrease heating velocity of the window/ it is only half
of uncoated glass window.
3. The calculated result of the relationship between heat current and temperature is as
f ol lows:
for G and G/F window/
Q = 0. 1 308 C T;„ - T.„< )
for 2G and G/F+G window.
Q = 0. 083 C Ti, - T.at ).
4. The window arrangement with a coated glass and glass (G/F+G) exhibits the best
thermal insulating efficiency/ but considering technical and cost parameters, the
coated glass window (G/F) is better. It s thermal insulating efficiency is 52* higher
than a bare glass (G) and is 11* higher than the double glass (2G).
REFERENCES
Berining H. ( 1983 ) Applied Optics 22 (24). 41 27-4141.
Ge Xin-shi. Jun Bao and Yu Shan-qing (1980) The Spectral-Selective Coatings For
Utilization of Solar Energy Academic Publishing House. Beijing.
Yu Shan-qing/ Hu Dun-ping/ Zhang Xiao-ping and Ma Min-wei ( 1 988 )Energy Technique 4/
30-31.
Yu Shan-qing. Zhang Xiao-ping and Ma Min-wei (1989) Materials Science Progress 3 (4)
353-359.
Yu Shan-qing/ Zhang Xiao-ping and Ma Min-wei (1990) Materials Science Progress 4 (1)
52-58.
Yu Shan-qing/ Zhang Xiao-ping and Ma Min-wei (1990) Clean * Safe Energy Forever Vol.3.
223 2-2236. ed by T.Horgome et.al. Porgamon.
Yu Shan-qing (1990) Energy Technique 3/ 18-27.
Yu Shan-qing/Zhang Xiao-ping and Ma Min-wei/ (1991) J. Inorganic Materials 6(1) 91-96.
Yu Shan-qing/ Zhou Guo-ping/ Zhang Xiao-ping and Ma Min-wei The optical properties of
Ag-SnOj film system" Thin Fila Science and Technology (To be published).
Zhang Xiao-ping/ Yu Shan-qing and Ma Min-wei "The ZnS/Me Heat Mirror Systems" Solar
Energy Materials ( To be published).
-------
2901
DESIGN OF SOLAR ENERGY HEATING WITH OPTICAL FIBERS
Li Bao jun, Xiong Xiaogeng, Wang Jingyu,
Yaeg Shilong and Jiang Xiangshan
Physical Dept., Shenyang Architectural Engineering
College, Shenyang, 110015, P.R. China
ABSTRACT
This paper contends that concentrated solar enregy is directly transported
with optical fibers in rooms without the light-thermal conversion process.
This makes full use of solar energy and lowers the effect of the climate
changes on efficiency of solar energy heating especially the cold areas.
Thus, it brings about a break-through both in the theory and in the economy
of the design of the solar houses.
KEYWORDS
Optical fiber; solar energy heating; numerical aperture; equation of light:
helical light.
INTRODUCTION
Active solar houses first convect solar energy into thermal energy with
some type of large solar energy collectors, and then the thermal energy
is transfered into a room by means of flowing water or air. Because solar
energy collectors have such high cost, they are not used widely. Although
passive solar energy houses are cheaper than the active solar homes, they
depend only upon the absorption and storage of southwalls and therefore
the thermal efficiency is low, especially in cold areas. In view of the
above-mentioned facts, we use optical fibers to transfer solar energy.
-------
2902
FEASIBILITY ANALYSIS
1) Solar energy ranges from 0.3 to 3 micron light wave length. Optical
glass fibers can transfer sunlight in the range of 0.5 to 2 micron light
wave length. Thus they can almost transfer solar energy entirely.
2) Optical glass fibers are pliable and tough, heat-resistant, light in
weight, easy to bundle. 3) The mechanical strength of the optical glass
fiber is 1.1x105 kg/cm2, so it is the right material to use. 4) Optical
glass fiber can receive light beams which are perpendicular to the optical
fiber axes. Thus receiving angle is great. 5) Numerical aperture of an
optical glass fiber is easily made to equal 1. Thus fibers can eliminate
transmission light and have good ability to focus light. 6) Optical glass
fibers are low in cost for the cheap materials and simple to manufacture.
7) The change of outside temperature has almost no effect on the heat
transfer of the optical glass fibers.
ENERGY TRANSFER SYSTEM
The solar energy transfer system with optical glass system consists of
three parts. Focus and collection part, light transfer part, radiator
part, shown as the following figures:
(D= =
(1) Focus and collection (2) Light transfer (3) Radiator
part part part
Fig. 1. Energy transfer system
The focus and collection part is made of the glass convex lens 15cm in
diameter with semispherical reflective focussing mirror. The former makes
sunlight focus to increase the density of the incident solar energy. The
latter increases solar energy transferred by the optical fibers.
The solar energy transfer part consists of the hollow glass fibers 10mm
in diameter and 4m in length. The external surfaces of the glass fibers
are coated with a material whose refraction index is less than glass's,
1.6x104 A0 in thick. The coat produces total reflection at interface
between glass and coat to reduce energy loss in the process of the light
transfer.
-------
2903
The radiator part is taper hollow glass fibers without coat at external
surfaces, so as to make solar energy transmit from the fibers.
PROCESS OF THE ENREGY TRANSMISSION
Sunlight is transmitted by optical glass fibers in the helical, whose
light equation can be written by
p(r)"3s~ = grad
s =
1
grad
9
(1)
(2)
The spread velocity of the light along fiber axis is given by
(]* I - p )1/2
»(x) =2=^=- (3)
.2 VU&l
, chromatic dispersion is equal to zero.
When n'(r)=
1+Ur)'
HEATING SCHEMATIC DIAGRAM
The heating schematic diagram, by means of the optical glass fibers, is
shown in the Fig. 2.
Semispherical solar energy collector is made of the focussing light head
formed by 200 strands of the optical glass fibers. This is because solar
energy can be fully collected in this way when sun is at different position
without the tracking system.
*7777
under
Sround
/////// f77/
(1) Semispherical solar energy
collector
(2) Optical glass fibers
(3) Light radiators and heating
device
Fig. 2 Heating schematic diagram
-------
2904
Focussing ends of every optical fiber are actually the same as convex lens,
which can condense solar energy. Then solar energy is directly introduced
into the room by the optical glass fibers, so that we can use them for
heating.
CONCLUSION
Solar energy heating with optical
about 100 dollar/m2, but the cost
The average temperature can reach
ranges from -20° to -18°C. Solar
has remarkable economic benefit.
glass fibers rises cost of the house
rise can be recouped in five years.
18°C., when outdoor average temperature
energy heating with optical glass fibers
NOMENCLATURE
S Poynting's vector
<-f Path function
n — Refraction index
v(r) Velocity of the light in optical glass fiber
Dielectric content
Magnetic conductivity in vacuum.
REFERENCES
Gloge, D. (1971). Weakly guiding fibers, flppl. Opt., Vol. 10, No. 10,
2200-2300.
Kapany, N. S. (1967). Fiber optics. Academic Press Inc., (New York).
Kawakami, S., and J. Nishizawa (1967). Kinitics of an optical wave
packet in a lens-like medium. J. Appl. Phys., Vol. 38, No.12,
pp. 4801-4820.
-------
2905
DEVELOPMENT OF NIGHT INSULATION DEVICES
IN PASSIVE SOLAR SYSTEM
Myongho Lee, Eon Ku Rhee, Bong-Gu Chun, Jungha Hwang
Department of Architecture
College of Engineering, Chung Ang University, Seoul, Korea
ABSTRACT
The research has been conducted to develop guidelines in designing night insu-
lation devices for passive solar system in order to maximize thermal efficiency
After an extensive literature search being conducted to investigate the state-
of-art of night insulation development, three scale models were constructed;
each representing Direct Gain, Trombe Wall and Attached Sunspace. Each model
contained two identical thermal rooms; one installed with night insulation de-
vice for experiment, and the other without night insulation for control. The
thermal performance of the models was monitored with various insulation devices
At the same time, a simulation has been conducted to model thermal performance
of passive solar system and especially of night insulation devices using compu-
ter programs. It was found that night insulation devices could improve thermal
performance of passive solar systems considerably and still could be economical
KEYWORDS
Passive solar system, Night insulation, Scale model, Computer simulation,
Direct gain, Trombe wall, Attached sunspace.
INTRODUCTION
A passive solar building requires a large area of south glazing in order to
maximize solar collection. However, this glazed area becomes the greatest
source of heat loss at night, for glass is the poorest thermal insulating
material. If the area is not properly night-insulated, not only the energy
saving effect of the system declines due to the deteriorated thermal perfor-
mance but the thermal environment will'be aggravated because of the large fluc-
tuation of indoor temperature. Therefore, it is essential to install an effec-
tive night insulation device in order to improve thermal performance as well as
to maintain a comfortable environment in passive solar systems.
The objective of the research is to develop guidelines in designing night insu-
lation devices for passive solar systems in order to maximize thermal efficiency.
Since the thermal performance of night insulation devices depends on many factors
such as thermal property of insulation materials, composition and joint of various
materials, position and operation of devices related to heat transfer character-
istics of various passive solar systems, etc., all these aspects were carefully
analyzed in order to evaluate the performance systematically.
-------
2906
EXPERIMENT
Construction of Test Models
In order to analyze the thermal performance of night insulation devices in
passive solar systems, three scale models were constructed; each representing
Direct Gain system, Trombe Wall system, and Attached Sunspace system. Each
model contained two identical thermal rooms; one installed with a night insula-
tion device, and the other without night insulation. The size of a thermal room
is a scale-down of a typical bedroom in a proportion of one fourth in area.
All the envelops of the model except south glazingwere insulated with 100mm
glass wool (K=0.29 Kcal/m h°C) and finished in white color to minimize heat
transmission. The direct gain model had 100mm thick (0.5B) red bricks placed
on the floor as thermal mass. In the Trombe wall and attached sunspace models,
the thermal storage wall of a layer of bricks (210mm thick, 1.0B) was built.
Night Insulation Devices
Various night insulation devices were selected after an extensive investigation
of a number of evaluation items. Table 1 shows the priority of evaluation items
considered for the selection.
TABLE 1 Priority of Evaluation Items for Selecting
Night Insulation Devices
Priority
System
1
2
3
4
5
6
7
Direct Gain
Thermal
Performance
Air
Tightness
Aesthetics
Operation/
Maintenance
Workability
Economy
Durability
Trombe Wal1
Workability
Operation/
Maintenance
Thermal
Performance
Air Economy
Tightness
Aesthetics
Durabi1ity
Durability
Attached
Sun space
Thermal
Performance
Operation/
Maintenance
Workability
Air
Tightness
Aesthetics
Economy
Durability
TABLE 2 Night Insulation Devices Selected for Experiment
System
Installtion
Inaction
Insulation Deviced Selected
Availability
Direct Gain
Inside
Quilt
Self-Produced
Outside
Slatted Shutter
Factory Manufactured,
Field Assembled
Trombe WA11
Inside
Roll Shade
Factory Manufactured,
Field Assembled

Outside
Sliding Shutter
Self-Produced
Attached
Sun Space
Inside
Quilt
Self-Produced
Outside
Roll Shade
Factory Manufactured,
Field Assembled
The night insulation devices which were selected for the experiment are listed
in Table 2.
-------
2907
Experiment-1
Experiment-1 examined the thermal performance of night insulation devices in-
stalled outside the glazing for direct gain and trombe wall systems. 'Sliding
Shutter1 was used for the trombe wall and 'Slatted Shutter' for the direct gain.
The outdoor temperature varied from -4.8°C to 5.2 t during the test period, and
the amount of average maximum insolation was 1.25 MJ.
'Slatted Shutter' installed outside the glazing was not very effective in im-
proving thermal performance of direct gain system. The average indoor temper-
ature of the experiment room during the test period was 13.9°C, only 0.7"C higher
than the control room where the average temperature was 13.2°C. At the same
time, the night insulation device contributed little in preventing the over-
heating problem. However, the average temperature at night rose 1.2°C higher
compared to the control room, indicating that the night insulation device had
some effect in reducing heat loss at night. The poor performance of 'Slatted
Shutter' was mainly due to inherent deficiencies of the device such as thermal
bridging and infiltration between the slats.
'Sliding Shutter' installed outside the glazing improved the thermal performance
of the Trombe wall system considerably. The average indoor temperature of the
experiment room was 3.0 C higher than the control room. During the night, the
magnitude of improvement was even higher, as the average temperature was 3.9°C
above the average of the control room.
Experiment-2
Experiment-2 tested the thermal performance of night insulation devices installed
inside the glazing for direct gain and Trombe wall systems. 'Quilt' was used
for the direct gain and 'Roll Shade' for the Trombe wall. The outdoor air tem-
perature varied from -2.4°C to 6.7c'C during the test period, and the amount of
average maximum insolation was 1.23 MJ.
'Quilt' installed inside the glazing was not very effective either in improving
the thermal performance of direct gain system, but slightly better than 'Slatted
Shutter' installed outside the glazing. The average indoor temperature of the
experiment room during the test period was 17.8°C, 1.0°C higher than the control
room (average 16.8°C). However, the overall thermal environment was somewhat
improved, as the average maximum temperature was slightly decreased (29.1°C
28.7 °C) and the average minimum temperature was increased (5.6 t 6.6 °C).
'Roll Shade' installed inside the glazing for the Trombe wall system increased
the average indoor temperature 1.8 C higher than the control room. It was also
effective in improving the thermal environment by increasing both the average
minimum temperature (3.93C -> 6.It) and the average maximum temperature (10.9°C
12.0 °C).
Experiment-3
'Roll Shade' was installed outside the glazing for the attached sunspace system
in Experiment-3. The outdoor air temperature varied from -2.7 °C to 8.2°C and
the amount of average maximum insolation was 1.68 MJ during the test period.
The average indoor temperature of the experiment room was 13.3°C, 0.9 t higher
than the control room. The average maximum temperature was decreased 0.5°C
(18.5°C -> 18.0CC) and the average minimum temperature was increased 2.1°C
(6.2 "C -» 8.3 °C).
-------
2908
EXPERI-MENT-1
EXPERIMENT-2
EXPERIMENT-3 EXPERIMENT-4
Fig. 1. Picture of scale model experiment
-------
2909
Although the indoor thermal environment was improved by installing 'Roll Shade'
outside the glazing of attached sunspace, it tended to promote the occurrence
of overheating inside the sunspace area.
Experiment-4
'Quilt' was installed inside the glazing for the attached sunspace system in
Experiment-4. The outdoor air temperature varied from 1.2 t to 15.2°C and the
amount of average maximum insolation was 2.11 MJ during the relatively warm test
peri od.
The average indoor air temperature of the experiment room was 16.9°C, 0.8 t
higher than the control room. The indoor thermal environment was quiteoimproved,
as the average maximum temperature was decreased 1.5°C (22.6°C -> 21.1 C) and
the average minimum was increased 2.2°C (10.lt -> 12.3°C). However, the problem
of overheating inside the sunspace area still existed.
COMPUTER SIMULATION
A computer simulation was carried out to analyze the thermal performance
of night insulation devices and to compare the result with the findings from
the model tests. 'Method 5000' program was used to simulate the energy load of
the models, and TTC (Thermal Time Constant) method was used to predict tempera-
ture variation in the models. Table 3 summarizes the result of 'Method 5000'
simulation.
TABLE 3 Simulation of Energy Performance of Passive Solar
Night Insulation Devices
(Unit: k\H)

Direct Gain
System
Trombe Wal1
System
Attached
SunSpace
Night
Insulation
Heating Load without Solar Gain(QnK)
670
523
592
Useful Solar Gain(SQ)
390
272
279
Annual Auxiliary Heating Load(2Qaux)
280
251
313

QUILT(inside)
220 (-21 .'1%)
-
-
With
Night
Insulation
Annual
Auxi1iary
Heating
Load
SQaux
— Qng—2Q
QUILT(inside)
-
-
238(-23.9%)
ROLL SHADE(inside)
-
194(-22.7%)
-
ROLL SHADE(outside)
-
-
225(-28.1%)

SLATTED SHUntH(outside)
227(-18.9X)
-
-

SLIDING SHUTOR(outside)
-
174(—30.7%)
-
'Quilt' installed inside the glazing for the direct gain system can reduce annual
heating energy consumption by 21.4% compared to the system without night insula-
tion. On the other hand, 'Slatted Shutter' outside the glazing can save 18.9%
of energy consumption. Since the simulation did not take account of the effect
of thermal bridging and excessive infiltration from the night insulation devices,
the overall performance may result in overestimation of energy saving. Vet, both
the model test and the simulation results prove that a simple and low-priced
insulation device inside the glazing is more effective than an expensive and
-------
2910
complicated outside-the-glazing insulation device for the direct gain system.
'Roll Shade' installed inside the glazing for the Trombe wall system has 22.1%
energy saving effect, and 'Sliding Shutter' outside the glazing reduces 30.7% of
energy consumption. This result is relatively consistent with the test result
which suggested that the night insulation devices perform the most for the Trombe
wall system.
For the attached sunspace system, 'Quilt' installed inside the glazing can save
23.9% of energy consumption, while 'Roll Shade' outside the glazing reduces
28.1% of energy consumption compared to the system without night insulation.
Meanwhile, the simulation of indoor air temperature for the test models using
TTC program shows that the temperature variation inside the models coincides
with the prediction of temperature variation by TTC algorithm. Therefore, it is
possible to predict the thermal performance of passive solar night insulation
devices through the TTC program.
CONCLUSION
The results of the research 'are summarized as follows.
1. The night insulation devices are helpful in improving the performance of
Trombe wall system and attached sunspace system. However, they are not
effective in direct gain system, where the control of thermal mass is more
isable.
2. For Trombe wall systems attached sunspace systems^the night insulation
devices installed outside the glazing are more effective than those installed
inside the glazing. For the direct gain systems however, the devices installed
inside the glazing performed slightly better.
3. The effect of night insulation devices can achieve the most in Trombe
wall systems attached sunspace systems direct gain systems,in descending
order.
4. It is possible to predict the thermal performance of passive solar night
insulation devices through the TTC program.
REFERENCES
Givoni, G. (1976). Man, Climate and Architecture. Applied Science Publishers.
Langdon, W. K. (19807"! Movable Insulation. Rodale Press
Mazria, E. (1979). Passive Solar Energy Handbook. Rodale Press.
Paul, J. K. (1979). Passive Solar Energy Design and Materials. Noyes Data
Corporation.
Pelanne, C. M. (1984). Journal of Thermal Insulation, Vol. 7. Technomic
Publishing.
Rhee, E. K. (1984). The Application of Passive Solar Trombe Wall System.
Engineering Journal, Vol. 20. Chung Ang University
Shurcliff, W. A. (1980). Thermal Shutters and Shades, Brick House Publishing.
-------
2911
COMPARATIVE ANALYSIS OF PACKED FLOW PASSAGE SOLAR AIR HEATERS
*
C. Choudhury, H.P. Garg and J. prakash
Centre of Energy Studies, Indian Institute of Technology
Hauz Khas, New Delhi - 110 016, India.
*Ramjas College, University of Delhi, Delhi - 110 007, India.
ABSTRACT
The primary concern of the investigations reported in the present
paper is the performance analysis and design optimization of
packed bed air heaters with single plate (non-selective absorber
for non-absorbing packing and glass cover for black painted
absorbing packing)above the air flow passage. The results of the
packed bed air heaters are compared with that of the
conventional bare plate air heater without packing in the air
channel.
KEY WORDS
Air heater; packed air duct; efficiency; pumping power.
INTRODUCTION
Several theoretical and experimental studies in the past have
been attempted to investigate the effectiveness of packings of
different materials, shapes and sizes in the flow passage of air
heating solar collectors and it has been established that packed
flow passage air heaters, due to larger surface area and tortuous
path through the bed which ensures very rapid heat exchange to
the flowing air, do provide much higher performance efficiency as
compared to their conventional air heater counterparts. A major
disadvantage of using packing in the collectors, however, is the
very high pressure drop experienced by the flowing air and hence
the high fan running cost of the system. The pressure drop can be
minimized by the proper choice of the design and operational
parameters of the system. Attempts have been made in the
direction (Choudhury and Garg, 1990) for the optimization of the
design and operational paremeters for solar air heaters with non-
absorbing packings (in conjunction with an absorber and a glass
cover) and absorbing (black-coated) packings (along with two
glass covers) in the air channel. In the present paper, for
economical running of the system, results of similar studies on
the packed bed air heaters with non-absorbing packing beneath the
absorber (without cover) and absorbing packing beneath a single
-------
2912
glass cover are reported. A parallel study on the conventional
bare plate air heater without packing in the air channel is also
carrried out in order to compare its performance efficiency and
pumping power with those of the packed bed air heaters.
Schematic views of the three air heater configurations
investigated in the present paper are illustrated in Fig. 1 (a),
(b) and (c). In the investigations, the heat transfer is
considered to be steady state and equal ambient temperatures are
assumed at the front and rear of the collector. In addition, the
inlet air temperature is assumed to be the same as^ the ambient
temperature. The steady state energy balance equations for
different components of the air heaters are :
THEORETICAL ANALYSIS
(block <£Ct»dl
— — - C#»»'
Potkinq I u«
Abterbt*
Kcxfc plot*
Both plot*
Bo ^)
hrp2 + Ub
W dy
Type III
IOC, + hrp1 (Tp-T/| ) = hc1f (T1 -Tf) + Ut (T-| -Ta); (8)
Itl«Cp = hrp-| (Tp-T1 ) + hcpf(Tp-Tf) + hrp2(Tp"T2)! (9)
hrp2 (Tp - T2) = hc2f (T2 - Tf) + Ub (T2 - Ta); (10)
MfC-p dTf
—— = h^fCT^Tf) + hcpf(Tp-Tf) + hc2f(T2-Tf) (11)
¥ dy
-------
2913
These equations for the three air heater types with different
design and operational variables were solved for the outlet air
temperature T- for solar flux of 900 W/m2, ambient temperature of
300 K and wind speed of 1.5 m/s, and the efficiency in each case
was obtained by using the relation
mfCf (T0 -
L, (12)
The relations used for computations of the heat transfer
coefficients for various parameters of the conventional and
packed bed air heaters are summerized by Choudhury & Garg (1990).
The pumping power P was obtained by using the relation P=mf Ap^Of
where /.p, the pressure drop in case of empty air flow passage is
Ap = (2G2/of) (L/Drt) (0.059 ReH-0*2) (13)
and that in case of packed air flow passage is
AP = (G2/of) (L/Dp) [d-C)/e3] [150[((1-S)/Ref) + 1.75]] (14)
RESULTS AND DISCUSSION
The effect of the superficial specific mass flow rate of air on
the computed values of the performance efficiency and pumping
power for the air heaters without and with packings in the flow
channel are presented in Fig. 2. A comparative analysis of the
curves reveal that for a fixed value of the air flow rate, the
efficiency of the packed bed air heaters (types II and III) is
much higher than that of the conventional bare plate air heater
(type I), its value being increasing with increasing rate of air
flow for all the air heater types. Typically, at 100 kg/hm2 of
air flow rate, the difference in efficiency between types I and
II air heaters is 31$ and that between types I and III is 44$
(for ring) to 45$ (for sphere). In determining the efficiency,
the shape of the packings has almost no effect in case of type II
and least effect in case of type III air heater, the efficiency
for sphere shaped packing being little higher due to its lower
porosity as compared to the ring shaped packing. The lower
efficiency of type II air heater, as compared to type III is
clearly due to the much higher rate of heat loss from absorber to
the surrounding in the former. Moreover, the rate of heat
transfer from the packing to the flowing air in type III is much
more higher than the indirect heat transfer (from absorber to
packing and from packing to air) as is the case in type II.
As discussed earlier, although the efficiency of the air heaters
increase with increase in air flow rate, the increase in m also
results in an increase in fan running cost of the system. In
fixing m^ for cost-effective operation of the collectors, the
chief consideration should, therefore, be the pumping power .
This fact suggests that the m value should not be increased
beyond 100 kg/hm2 as higher air flow rates result in very high
pumping power. All further results, therefore, are presented
only for flow rates 50 kg/hm2 and 100 kg/hm .
-------
2914
The dependence of the efficiency and the pumping power on the
length of the three air heater types are presented in Fig. 3 for
the above mentioned values of the superficial specific mass flow
rates of air. The curves indicate that for a fixed length, the
efficiency is the highest for the type III air heater with
spheres in the air channel. In case of type II air heater, the
shape of the packing material is observed to have no influence on
its performance. As one expects, the efficiency and the pumping
power values increase with increase in length for the types I and
II air heaters. However, in case of type III, the efficiency
decreases very slowly with increase in length beyond a particular
value for both the spheres and rings as packing. A comparison of
the efficiency and the pumping power for different collector
lengths suggests that at the specified duct depth value; a cost
effective design configuration corresponds to a length of 1 m for
sphere shaped packing and 2 m for ring shaped packing in case of
type III air heater for both the air flow rates and the packing
diameters. Similar analysis for the type II air heater suggests
the optimum length to be 2 m for sphere shaped packing. However,
for rings in the air channel of type II collector, the pumping
power puts no constraints on the length of the air heater for
cost-effective operation of the system. Same is the case for the
conventional bare plate (type I) air heater. However, suitable
lengths for these air heaters should be chosen depending on the
requirement of the increment of the air temperature flowing
through the system.
Figure 4 illustrates the dependence of the performance efficiency
and the pumping power on the duct depth of the air heaters. As
per expectations, the efficiency of the collectors type I and II
decrease with increase in duct depth. However, the situation is
not quite the same for the collector type III. Here, the
efficiency is first observed to increase and then decrease with
increase in Z. Since the pumping power decreases with increase
in duct depth and since its individual value at different duct
•0
Lit a SOlQ/hm?
I- At lOOkfftim?
Op* O.Mm
1 • 0.1m
m(k9/hm2)
Fig. 2. Effect of superficial
specific mass flow rate on
efficiency and pumping power
Fig. 3. Effect of duct
length on efficiency and
pumping power
-------
2915
depth is too low to be accounted for (except for spheres at a
duct depth of 5 cm at 100 kg/hm2 of air flow rate) a comparative
analysis of the curves for ^ and Ap for the design and
operational parameters specified indicate that for higher
effective energy gain of the system, there exists no practical
limit to how small the duct depth Z can be made (within the
specified range). For the type II air heater with spheres as
n • SOka/hw'
!.* • 100 kgih mi*
Tjrpt I
0
SOkg/hm*
l.m « tOOkgffcm'
"JfP* B
•• Tnx in
Fig. 4. Effect of duct depth
on efficiency and pumping
power
Fig. 5. Effect of packing
diameter on efficiency and
pumping power
packings at 100 kg/hm2 rate of air flow, due to the very rapid
decrease of efficiency and not so rapid decrease of pumping power
with increase in duct depth, the air heaters must be designed
with Z not less than 10 cm. Similarly, for the type III air
heater with spheres, which is observed to have the highest
performance, for the most cost-effective design, the duct depth
should be 15-20 cm for smaller diameter packings and 10-15 cm for
larger diameter packings.
Figure 5 shows the typical efficiency values as function of
packing diameters for 50 kg/hm2 and 100 kg/hm2 of air flow rates
respectively for the types II and III air heaters. Although
smaller diameter packings result in better performance of the
systems, use of too small spheres or rings in the air flow path
must be avoided as they impede the air flow and necessitate a
higher fan running cost. Considering this fact and comparing the
and P values in these figures, approximately 3-5 cm diameter
packings are recommended for use in the collectors. In addition,
due to the very low pumping power for rings, which is a
consequence of its higher porosity value, ring shaped packings
should be preferred over spheres for higher effective energy
gain of the system.
CONCLUSION
Since the primary concern throughout the design study of the
packed bed collector is the minimization of operating cost &
maximization of efficiency, the efficiency and the pumping power
values of the collectors for different design and operational
-------
2916
conditions are computed and compared to work out the optimum
operating condition of the collectors. For the various
parameters, the performance of type III air heater i,s observed to
be much higher than that of type II. At 100 kg/hm2 of air flow
rate which results in moderate pumping power, the optimized
design corresponds to a collector length of 1-2 m, duct depth of
10-15 cm and packing diameter of 3-5 cm. In addition, due to the
very high value of pumping power and not so high value of
efficiency for spheres as compared to those for rings, the ring
shaped packings should be preferred over the spheres for cost-
effective operation of the packed bed air heaters.
NOMENCLATURE
C specific heat (Wh/kg°C)
D diameter (m)
G superficial mass velocity of air (kg/hm2)
h heat transfer coefficient (W/m2oC)
I incident insolation (W/m2)
L length of air heater (m)
m superficial specific mass flow rate (kg/hm2)
M superficial mass flow rate (kg/h)
P pumping power (W/m2)
Ap pressure drop (Pa)
Re Reynolds number
T temperature (°K)
U heat loss coefficient (W/m2oC)
W width of air heater (W)
y coordinate along collector length
c£ solar absorptance
<£ void fraction or porosity
rl efficiency
T. transmittance
Suffix
1 first plate (absorber in types I & II, cover in Type III)
2 second plate (back plate)
a ambient
b back
c conductive-convective
f fluid (air)
H hydraulic
i insulation/inlet
o outlet
p packing
r radiative
s sky
t top
w wind
REFERENCES
1. C. Choudhury and H. P. Garg (1990). A study on the performance
of the air heating collectors with packed air flow passage,
communicated to Solar Energy.
-------
2917
SOLAR PERMEABILITY OF URBAN TREES IN THE DRY TEMPERATE
CLIMATES OF WESTERN ARGENTINA
A.C. deCerutti, C. deRosa, J.L. Cortegoso and A. Ravetto
Laboratorio de Ambiente Humano yVivienda (LAHV) CRIGYT-ME
Casilla de Correo 131 5500 Mendoza ARGENTINA
ABSTRACT
The arid region of central Western Argentina presents wide variations between
the extreme seasons and the possibility to utilize natural energy resources to
provide thermal comfort and achieve energy savings in the built environment.
From an urban and energy planning viewpoint the potential for passive condi-
tioning of the living spaces is a function not only of the urban layout of the
buildings but also by the usual massive presence of trees as a dominant feature
of the urban landscape.
The goal of this study is to quantify tine reduction of the solar radiation in
summer, through the foliage of four commonly used deciduous tree species in the
city of Mendoza. The results presented include values of global and diffuse solarrad-
iation, expressed as percentages of global radiation on a fully insolated horizon-
tal surface. Maximum and minimum values of crown's permeability range from 27.04
to 5.47%. In the case of diffuse radiation reductions, they were found to be closdy
correlated with:fcfre solid angle of the sky dome masked by the croims.
KEYWORDS
Urban climate; tree crowns; permeability; solar radiation; warm season.
INTRODUCTION
In arid regions, the annual climatic cycle presents wide variations between
extreme seasons with particular significance in the case of ambient temperature
not only yearly but on a daily basis as well. The quantity and intensity of sol-
ar radiation is also important, du6 to the usually high percentage of clear days.
These climatic features on the one hand are. strongly negative conditions but , on
the other arenatural energy resorces particulary fit to be utilized for thermal
conditioning of inhabitable spaces, thus reducing conventional energy budgets.
Argentina, a country vhich features extensive dry regions with temperate - cold
climates, presents also a population distribution vhich is predominantly urban.
According to the figures yield by the last National Population and Housing Survey
(1980); 82.67% of the total population lives in urban settlements of more than
a thousand people. This implies that any successful attempt to undertake an
important energy substitution program on the residential sector should be im-
plemented mainly on the urban building stock. From the point of view of city and
energy planning, it must be considered that the potential for natural thermal
conditioning of spaces in urban buildings is not only a function o£ building
design and technology but also of the city's volumetric pattern including non-
-------
2918
architectural elements, such as trees. The city of Mendoza, on the arid plains
of central-western Argentina, is in this sense of particular interest, due to the
massive presence of trees as a dominant feature of its urban landscape. Fig. 1.
Fig. 1. Aerial view of the typical urban tissue of Mendoza
This reality allows for the favorable modification on certain parameters of the
urban climate, particulary a significant reduction in the urban heat island
effect. It also acts as a conditioner of aplicability of climatic design strat-
egies in the aspects of conservation, passive solar heating (1) and natural cooling
In the case of energy conservation during the warm season, the main benefit is
the provision of shade on building surfaces, thus noticeably reducing heat gain
in the interior spaces. When the potential for passive solar heating is considered
it is of prime interest to assess values of the solar radiation reduction on collec-
ting surfaces, due to the presence of tree crowns, even when they are of decid-
uous species. This knowledge will allow a proper assessment of summer comfort
and the solar potential of the existing building stockias./w£ll:asthe optimization,
at the urban design level, of the relationship of trees and buildings in order
to obtain maximum environmental benefits, when new residential complexes are
dealt with.
METHODS AND MATERIALS
The goal of the present study is to quantify the reduction of solar radiation
in summer, due to the effect of shading of tree canopies. The four most commonly
used deciduous species in the city were analyzed, according to the following
method:
Selection of Species
In order to determine the most usual tree species within the urban milieu of the
Capital District of the city of Mendoza, a survey was performed over a,sample
group of 37 blocks, which were selected, according to the following criteria:
-------
2919
Different population densities, and their correlation with representative buil-
ding densities. As a result three sample groups of different building densities
were defined:
Group 1: Low building density (mean value 1.96 m3/m2): 11 city blocks.
Group 2: Medium building density (mean value 2.94 m3/m2): 17 city blocks.
Group 3: High building density (mean value 7,48 m3/m2): 9 city blocks.
For the selected sample blocks, trees were surveyed along,every street around
each block. The method utilized to determine the most usu^l species within the
urban milieu was to take into account all the species with quantities higher than
1% of the total number of trees in the area considered. Folir deciduous tree
species were thus determined: Mulberry tree (morus alba), Plane tree (platanus
acerifolia), European ash (fraxinius excelsior) and China tree (melia azedarach) .
Table 1.
TABLE 1 Representativity of selected tree species in the
urban milieu of Mendoza.
N°
DENSITY
SPECIES
LOW
MEAN
HIGH
SAMPLES
Quant
%
Quant
%
Quant
%
Quant
Z
1
Mulberry tree
240
3 2,29
171
2 0.60
251
5 3.10
662
32.71
2
Plaie tree
60
191
8.32
222
26.75
159
33.61
441
21.80
3
European ash
26.49
224
27.00
9
1,90
424
20.95
4
China tree
42
5.82
12
1.44
1
63
0.21
13.31
55
2.72
5
Others
188
27.08
201
24.21
452
21.82
Totals
721
100
830
100
473
100
2024
100
Design of the Method
Taking into account that the goal of the study was to assess values of the perme*
ability of tree canopies to solar radiation for different species, different
times of the year and hours of the day, a method was developed which would yield,
within the limitations due to the available instruments, precise and reliable
results, representative of the situation over the total floor area shaded by the
tree crown.
The method is based on the comparative analysis of simultaneous readings of solar
radiation under open sky and filtered through the tree's canopy, utilizing two
KIPP & ZONEN - CM5 pyranometers, duly calibrated and with the following configura-
tion:
A reference-solarimeter is placed at floor level under open sky, measuring sol-
ar radiation on a horizontal surface.
A second solarimeter is set, also at floor level under the shade of the tree
crown. Each reading taken by this instrument is set at the cross points of a
grid set on the floor, moving along each axis, with a total of 25 points of mea-
surement, over a total square grid of 2.0 meters on each side. At every reading
point the instrument is set until the signal becames totally stable
Over both instruments a translucent acrilic diffiuser of 0.4m on each side, is set
horizontally,.at a vertical distance of .07m over the solarimeter's dome. This
device allows for the homogeneization of the impinging radiation over the in-
struments sensor, thus reducing the sampling time and avoiding significant varia-
-------
2920
tions in the measurement conditions. In order to adjust the method, tests were
made to measure the . amount of diffuse and reflected radiation getting to the
sensor through the gap between the diffuser's edge and the earth surface. Since
the instrument was in all cases set on grass covered soil, only from 9 to 11% of
the total radiation recorded by the instrument reached the sensor through this
gap, and it was generally reduced in intensity in the same proportion as the rad-
iation impinging on the diffuser. Only small errors, not larger than 5%, were
recorded at the edges of the area shaded by the tree's canopy, due to this effect.
Fig. 2.
Fig. 2. Configuration of the measurements method
When diffuse radiation was evaluated through measurements , the diffuser was replaced
on either solarimeter, by a properly set black cardboard disk wich blocked the
direct beam on the instruments dome.
Selection of samples for monitoring
The sample trees to be monitored were selected primarily on their condition of
isolation so as to avoid, within the greatest possible measure, the inter-
ference of elements of the immediate environment. Other criteria for selection
were: unrestricted access to perform the measurements around the tree and the
features of the tree itself: age, health, shape and dimensions,close, as much
as possible, to the average of the species. As a consequence of the above men-
tioned conditions,a variable number of grids were laid over the area shaded by
the canopy. Their distribution was also a^function of the area ' s shape. In every
case it was a goal to define three distinct zones within the area: close to the
trunk, intermediate and close to t!he top. The goal of this pattern was to obtain
results that, in a way, could be easily processed and allow:ed ,at the same time
a comparative evaluation for different species.
Measurements of solar radiation
Global and diffuse radiation were measured simmultaneously on clear days under
open sky and over the area shaded by the crown following the annual cycle,and during
periods which are relatedtin each case,with the foliation periods of the species
considered. At each point of measurement two readings were taken: one two hours
-------
2921
before and the other at solar noon. No readings were taken during the afternoon
hours considering the situation of symmetry with the first reading. The data
acquisition system used was a John Fluke 2240, data logger. An instrument controll-
er ordered the information read, ich was then filed in diskettes.
A synthesis of the processed results is presented at this point, covering the
recorded summer period (December 89 - Rbruary 1990), so-lar global and diffuse
radiation are shown, for both instruments: under open sky and under the tree's
canopy. The latter are expressed as percentages of the former. Mean values were
computed from the two readings (mid morning and solar noon). Tables 2 and 3.
TABLE 2 Measurements of global radiation during the
summer period.
PLATANUS ACER IFOLIA
(Plltano)
Distribution
of grid
sL &
Gi

\
:"G]
k-i..'
j
\
f"G]
U.J
/
Area
Grid

0
b.48

A
1
7,38
7,70

2
7.25

0
8,33

3
1

8,33

0 ;
13.29

C
1

13,29

9.77
MELIA AZEDARACH
(Parafso)
Distribution
of grid
Area
Grid

Mnan
(%)

0
19,79

A
1

19.79

B
1

0
31.92

C
1

31,92

t
25.85
FRAXINIUS EXCELSIOR (Fresno europeo)
Distribution
of grid
Grid
Part
«)
30.58
37.97
41.98
Mean
(«
30,58
Total
«)
37.97 36,84
41,98
K0RUS ALBA (Morera)
Distribution
of grid
Area
Grid
Part
(X)
Mean
(%)
Total
<%)
iA
|b
i...
c
.—.
A
0
33.70
33,80
31,37
/'IqJ\ '
1
40,78

2
27.16
¦ Hi /i
B
0
26.02
30 ,73
1
31,79
2
34,40
C
0
29,60
29,60
1

1 i 0 : 2 ;
2

TABLE 3 Measurements of diffuse radiation during the
summer period.
PLATANUS ACERIFOLIA
Distribution
ot grid
Grid
Area
52.58
54,53
49.01
58,94-
58.94
60.02
69,08
69,08

FRAXINIUS EXCELSIOR
(Fresno europeo)

Distribution
of grid
Area
Grid
"a?
Mean
«)
Total
(X).

0
76.19

A
1

76.19

A
'

V—
B
1
78,93
78,93
79.63
i?-
4

/
•V
A

!
0

\r
c
1

83,78

1
'0
2

2
83.78

-------
2922
MEL IA AZEDARACH
(Paraiso)
Distribution
of grid
e /
J5j
\~z9s;
a 7

Part
GO
78,08
1 i
88.45
2 1
(Morera)
MORUS ALBA
Distribution
of grid
Total
Grid
Area
85.18
91.08
78.76
85.31
89.55
89.55
92.49
92,49
RESULTS
The results obtained by measurements during the warm season identify
definite tendencies in the patterns of solar global radiation filtered through
the crowns of the different species studied, related with their distribution over
the various zones of the canopy itself. A similar behaviour can be noticed in
the cases of: European ash, China trees and |>lane trees. In all of them.a consist-
ent variation is observed with increasing permeability at the top of the crown,
due to the thinning of the filtering elements in that area. This feature is
clearly apparent in all individuals studied, with ranges of variation between
12„13% for the maximum and 5,59% for the minimum. In the case of the Mulberry
tree, the variation gradient is inverse due, in this case, to a greater density
of leaves at the top of the crown, .. in quantitative terms is expressed in
a reduction of the permeability at the top of A,2%. Fig. 3.
B-T

1
1 "

-J 3
J 2

j4
a e e
Areas shaded by canopy
1. Mulberry tree.
2. China tree.
3. European ash.
4. Plapc. tree.
Fig. 3. Permeability to global radiation in summer
In the case of diffiise radiation,the curves plotted with the measured values,
a direct relationship can be observed between the obtained figures and the solid
angle of sky-dome masked by the tree crown. It is to say that, with greater area
of unobstruted dome a larger value of diffuse radiation is obviously recorded by
the instrument. Thus, a series of curves of growing values of diffuse radiation
can be plotted as the reading points get further away from the tree's trunk .
The variation ranges from 17.04% in the case of the Plane tree .to a minimum of
7,44% for the Mulberry tree. Fig. 4.
-------
2923
too i
ABC
Areas shaded by canopy
1. Mulberry tree. 3. European ash.
2. China tree. A. Plane-tree.
Fig. 4. Permeability to diffuse radiation in
CONCLUSIONS
The conclusions arrived at by comparing the results obtained for the four species
analyzed can be summarized as follows:
From the view point of the urban environment, the massive presence of trees is of
invaluable help in warm-dry regions to provide more suitable conditions for human
comfort, either for inside and out-doors environmnets. Effective control of solar
radiation impinging on vertical and horizontal man-made hard surfaces, minimizes
the absortion and storage of heat.
Careful attention must be paid to the selection of the tree species to fulfill
this role,particulary in the aspect of foliage density,ttjemain conditioner of the
permeability of crowns to the solar radiation. In summer the Plane
tree presents itself as the best choice among the four species analysed with a
mean permeability over the whole crown of 9.77%; the European ash is at the opp-
osite extreme of the permeability range, with a mean value of 16.19%.
In later stages of the work, the results obtained for summer and winter periods
will be confronted, in order to develop proposals that could combine maximum protec-
tion from solar radiation in summer with greater access to the resource in winter.
REFERENCES
(1) Canton de Cerutti, M.A. - de Rosa, C. - CorteRoso. J.L."Bloqueo de la radia-
cion solar a traves de la copa de especies caducifolias". International Congress
of Energy, Environmental and Technological Innovation. Caracas, Venezuela.
October 22-26, 1989.
(2) Canton de Cerutti, M.A. — de Rosa, C. — Cortegoso, J.L. "Incidencia del ar-
bolado en el potencial bioclimStico de entornos urbanos en zonas aridas'.'ASADES'
88. Annual Meeting of the Argentine Solar Energy Association. Salta, Argenti-
na. October 25-28, 1988.
-------
Intentionally Blank Page
-------
3.13 Transparent Insulation I
-------
Intentionally Blank Page
-------
2927
THE SPACE HEATING CONCEPT OF THE SELF-SUFFICIENT SOLAR HOUSE FREIBURG
W. Stahl, W.-S. Wilke
Fraunhofer-Institut fur Solare Energiesysteme
Oltmannsstr. 22
W-7800 Freiburg, FRG
ABSTRACT
Even on overcast winter days, the solar radiation flux on vertical house walls exceeds the conductive
heat flux outwards through the walls which is caused by the temperature difference between indoors
and outdoors.
On transparently insulated house walls the insolation even on such days can be utilized to
compensate for the conductive heat losses and to serve as an effective direct solar space heating
system. The Self-Sufficient Solar House (SSSH) which is going to be built in Freiburg is heated with
a solar fraction of more than 90% by such a wall with Transparent Insulation (TI). The space
heating demand of the SSSH is reduced to 300 kWh/a ( 2 kWh/m2 heated living area) auxiliary
heating energy.
The optimal ground-plan for the SSSH was found to be a sector of a circle. This shape gives the best
ratio between north oriented walls with energy losses and the south oriented TI wall. The energetic
optimization of the SSSH was done by computer with the simulation program HAUSSIM. This
program has been developed to simulate the new TI wall heating system.
The TI wall heating system is combined with other measures to minimize heating energy demand.
Windows with a U-value of 0.6 W/(m2K), opaque wall insulation with U < 0.2 W/(m2K) and an air
ventilation system with highly efficient heat recovery are the main components.
KEYWORDS
Transparent Insulation, Solar House, Computer Simulation, Self-Sufficient, Space Heating, Solar
Collector
INTRODUCTION
The utilization of the solar radiation intercepted by house walls was introduced by Trombe several
decades ago. His concept was the convective air heat transfer into the house. Our TI wall systems
are based on a passive heat flux through the house wall. The efficiency of TI walls for long term
averages is given by
T) = Ta/(1 + /Ln/Aw) (Goetzberger, 1985).
ra is the product of light transmission and absorptivity. A being the heat loss coefficient of the TI
and the wall respectively. From this equation it can be seen that the efficiency increases as the heat
transmission of the original wall increases and as the heat transmission of the translucent insulation
decreases. The low U-values of TI materials, less than 1 W/(m2K), combined with a diffuse solar
transmittance over 70%, are the main features which distinguish them from the Trombe concept.
The concept of the SSSH (Goetzberger, 1987) was developed after different demonstration projects
with TI walls showed promising results (Voss, 1989, Braun, 1991). If it is possible to supply almost
the entire heating demand with solar energy, then it makes sense to consider a totally Self-Sufficient
Solar House. The general outline of the project is presented elsewhere (Goetzberger, 1991).
-------
2928
TRANSPARENT INSULATION, SPACE HEATING DEMAND AND BUILDING
LAYOUT
Conventional opaque insulation only decreases the energy loss through the wall; the utilization of
solar radiation is not possible. To reduce the space heating demand (SHD), a small surface to
volume ratio should be obtained. The constraints on the architecture of low-energy buildings are
noticeable, most of them having a more or less cubic shape.
Buildings with TT walls must comply with other conditions to minimize the SHD. Physical properties
of the wall, the TI material, the construction of the facade and the solar radiation intensity are
decisive for the overall energy balance through the wall. In the northern hemisphere, a very long
building with a tranparently insulated south wall would have the lowest SHD, although the surface to
volume ratio is very high. With today's TI materials the energy balance of north walls is not positive.
It is conceivable that research will lead to materials with further increased efficiencies. Then the
predominantly diffuse radiation on a north oriented wall will be sufficient to compensate for the
heat losses. In that case, the architecture of a building would be totally released from constraints to
minimize the surface to volume ratio because of the SHD.
SIMULATION PROGRAM
Drawing on the successful development of TI materials, the simulation approach established itself
from the beginning of the project as an important foundation for the planning of the SSSH. The
multi-zone simulation program HAUSSIM has been developed; it is able to numerically simulate the
non-stationary temperature response of TI walls using a differential procedure. The program offers
the option of controlling the heat flux by moving components, e.g. the roller blind. The program runs
with daily or hourly meteorological data sets (e.g. Test Reference Years). From these values, a daily
temperature profile is determined by linear interpolation, and the daily radiation fluctuation on
vertical surfaces with the required orientation is calculated from different correlation and diffuse
radiation models. The solar radiation which enters the rooms through the light openings is
distributed homogeneously over all indoor surfaces and is absorbed there. The room air
temperature within a zone is calculated using combined heat transfer co-efficients i.e. radiation
exchange is not explicitly calculated. A detailed program description can be found in (Wilke, 1991).
As part of the IEA Task VIII, various building simulation programs were compared. The building
example defined by Task VIII was subsequently simulated with HAUSSIM and the results are in
very good agreement with those from the other programs.
BASIC SIMULATION RESULTS
The calculations are made with TRY weather data for Freiburg with a latitude of 48° N. To allow for
an increased probability of foggy days at the building site of the SSSH, weather data for December is
replaced by measurements made at the Institute. The TI wall is simulated with a roller blind to
control energy gains and to reduce heat loss during the night. Parameters of the simulated building
are given in Table 1.
-------
2929
TABLE 1. Simulation Parameters

thickness
heat conductivity
spec, heat

[cm]
[W/(mK)]
[J/(gK)]
Wall:
30
1.8
• 2200
Roof and ceilings:
20
1.8
2200
Opaque insulation:
20
0.04

TI material:
10
0.08

TI material: diffuse energy transmittance 76%
TI facade: U-value with raised blind 0.7 W/(m2K)
U-value with closed blind 0.5 W/(m2K)
Air exchange: 0.7/h
Air ventilation heat recovery system, efficiency 0.9
Internal heat sources: 6.3 kWh/d
The yearly SHD was first calculated for a cubic windowless building, ground area 10m * 10m and 5m
high. The flat roof and floor have opaque insulation. The results for opaque and transparently
insulated facades are given in Table 2..
TABLE2. Calculated SHD for a Windowless Building in Freiburg. Ground Area 10m • 10m, 5m hiph.
Buildine Volume 500nv\
SHD [kWh/a]
Opaque Insulation on all vertical surfaces
3500
TI west or east facade
1700
TI south facade
600
TI east, south and west facade
440
TI south-east and south-west facade
410
With opaque insulation on all walls the SHD is in the range of the best low-energy buildings. The TI
facades decrease the SHD further. The effect of the TI south facade is remarkable, the decrease of
SHD being more than 80%. The effect of the south-east and south-west orientation is interesting.
Due to the high insulation standard of the building, the heating period is reduced to the months
from November to February. Solar radiation during these months is mainly incident with surface
azimuth angles between -45° and + 45°. Therefore the two TI facades result in a lower SHD than the
three facades orientated east, south and west.
The dominating energy gains on the south oriented TI wall makes it obviously advantageous to
increase this surface area. For constant building volume of 500m3 and height of 5m, but different
ground-plans, the SHD are given in Table 3.
-------
2930
UBXE3. SHD for an Increasing Area of ftuth Oriented TT Wall (building height 5m and mnctart
: building volume 500mJV__ ~
TI east, south and west facade SHD [kWh/a]
ground area 10m * 10m (from Tab.2) 440
(south oriented TI wall 50m2)
ground area 12.5m" 8m 360
(south oriented TI wall 62.5m2)
ground area 16m • 6.25m 290
(south oriented TI wall 80m2)
The SHD decreases for an increasing area of south oriented TT wall. The increase of course is
limited by the interior architecture. In an elongated building, satisfying configurations for the rooms
are almost impossible to achieve.
A compromise was found in a ground-plan equivalent to a sector of a circle. With a radius of 9.85m
and an opening angle of 148° a better south orientation is achieved than with a semi-circle. The SHD
for this ground-plan shown in Fig.l is 250 kWh/a.
Fig.l. Ground-plan of a building with minimized SHD.
DETAILS
Fig.2. Ground-plan of the SSSH.
In cooperation with the architects the energetic considerations were combined with a pleasing
design for the SSSH. The circular shape is constructed as a polygon with 22 segments, each lm wide
(Fig.2). The TI wall also includes windows with a U-value of 0.6 W/(m2K) and an energy
transmittance of 0.42. The construction of the TI wall is shown in Fig. 3. The system to control
-------
2931
energy gains and to decrease thermal losses during the night consists of a roller blind, which is pulled
between the TI material and the glass cover from the basement to the top of the facade. The drive is
mounted on the flat roof; only one motor is needed. The blind is made from two layers, an outer
highly reflecting fabric is combined with an aluminized material for high IR reflection. The low iron
content glass cover is self-supporting, so that thermal contact to the wall is avoided.
BRICKWORK
TI MATERIAL
BLIND
GLASS COVER
Fig.3. Cross section of th TI wall of the SSSH
The air ventilation system of the SSSH is in operation from November to the end of February. To
get low air velocities, the air ducts have overdimensioned cross-sections. The whole system can thus
operate with only two 40 W fans. Heat is recovered with two stacks of thin aluminum plates. One
stack takes up the energy of the exhaust air and the other heats the fresh air. Every 30 sec the air
streams are exchanged. The expected efficiency of the heat recovery is 90%. A prototype version has
been tested. All the rooms have separate inlet and outlet ducts. The residents can vary the
ventilation according to the use of the rooms. A basic air exchange is guaranteed by a control
system.
The two-storey building will be used to demonstrate the advances in solar technology to the public.
On the first floor, demonstration and lecture rooms are available. The ground floor will contain two
apartments. All the technical equipment will be in the basement of the building. The basement is
insulated from the ground with glass foam.
Simulation calculations have been carried out for the design of the SSSH. Parameters of Table 1 are
valid. The building is divided into three zones: south orientated rooms, north orientated rooms and
corridors, and the north orientated staircase.
26
o 24
22
Max.
Min.
20
18
SEP OCT NOV DEC JAN FEB MAR APR MAY
Figure 4. Indoor air temperature in the south oriented rooms of the SSSH.
The indoor air temperature in the south orientated rooms is shown in Fig.4. This calculation is
started in September with an overall temperature of 20°C. The temperature rises due to the solar
gains. At 24°C the simulation program assumes cooling by natural ventilation through the windows.
The air ventilation system starts to operate in November when the temperature begins to decrease.
-------
2932
From this time on the windows are closed. In January, the air temperature reaches 18°C. The indoor
surface temperature of the TI wall is high, 20 to 26°C is typical. Agreeable comfort temperatures
result. The heating demand is calculated if air temperatures go below 18°C. The overall SHD for this
case is 300 kWh/a. This energy is provided by the hydrogen/oxygen-system (Heinzel, 1991). A
catalytic burner is installed in the ventilation system to heat the fresh air. An auxiliary heat source
may be installed in the bathrooms.
CONCLUSIONS
The TI wall of the SSSH utilizes the solar energy in winter time with a high efficiency. A blind
system controls the energy gains and protects the house against overheating in summer. The SHD is
reduced by a factor of 10 compared to the best low energy buildings. Therefore the need for
seasonal storage is avoided. The low SHD is the key for self-sufficiency. A disadvantage is the factA
that the TI wall only operates during the heating season. TI walls liberate architecture from the
constraints of low surface to volume ratios.
REFERENCES
Braun, P.O. and J. Schmid, E. Bollin, W. Stahl, J. Vahldiek, K. Voss, A. Wagner (1991). Transparent
Insulation Materials - Demonstration Projects and future Prospects, to be published in the Proc.
of this conference.
Goetzberger, A. and J. Schmid (1985). Review of Components for Passive Solar Energy Utilization.
Int. J. Solar Energy, 1985. Vol. 3, pp. 309-328.
Goetzberger, A. and W. Stahl (1991), The Self-Sufficient Solar House Freiburg, to be published in
the Proc. of this conference.
Heinzel, A. and K. Ledjeff (1991). The Self-Sufficient Solar House: Hybrid Energy Storage System,
to be published in the Proc. of this conference.
Voss, K., editor (1989), Transparent Insulation Technology for Solar Energy Conversion, Report
Fraunhofer-Institut fiir Solare Energiesysteme.
Wilke, W.-S. (1991).'Transparente Warmedammaterialien in der Architektur - Anwendungen,
thermisches Sytemverhalten und optimale Raumklimakonditionierung, Dissertation FhG-ISE,
Freiburg
-------
2933
TRANSLUCENT INSULATION FOR PASSIVE SOLAR ENERGY APPLICATION
H.A.L. van DIJK
TNO Building and Construction Research
P.O.box 29
2600 AA Delft
The Netherlands
ABSTRACT
A number of studies are being carried out at TNO on the effect of translucent insulation
materials (TIM) for application in walls (TIM-walls).
For the Dutch situation the feasibility of translucently insulated walls has been investigated.
The possible energy savings of current wall designs are estimated on 700 to 1200 MJ per m2
facade. During the summertime the overheating may be a serious problem. Natural ventilation
of the airgap is one of the possibilities to reduce the overheating. A preliminary test in one of
the PASSYS testcells at TNO showed promising results.
Further and more detailed research and development has been started, like building simulati-
ons with TIM-walls and a second TIM-wall test in the PASSYS testfacility.
KEYWORDS
Translucent insulation materials, passive solar components, computer simulations, PASSYS
testcells, parameter identification.
RESEARCH OBJECTIVES
In the research carried out at TNO with respect to passive solar energy systems the following
goals can be recognized.
quantitative determination of the system characteristics for product information;
full scale testing of new system concepts before applying them in practice, to
decrease the risk of costly flaws;
system optimization;
system modelling;
development of simplified design tools for energy savings calculations.
TRANSLUCENTLY INSULATED WALL PERFORMANCE; THEORY
A wall provided with exterior translucent insulation (TIM-wall; example fig.3) combines a
high thermal resistance with good optical properties to enable the utilisation of the solar radia-
tion on the facade.
Solar radiation is absorbed by the (black painted ) wall. The major part of the heat flux is
-------
2934
Solar radiation is absorbed by the (black painted ) wall. The major part of the heat flux is
towards the room, since the insulation reduces the heat flux to the ambient.
The heat capacity of the wall levels the fluctuations of the solar radiation and delays the heat
delivery to the room to a time where the heat demand is larger than during the daylight hours
(for instance in the evening). The summer situation is also important.
Measures should be taken to avoid overheating.
A number of Tl-materials is investigated and in particular materials with a honeycomb struc-
ture and the silica aerogels are promising (van Dijk, 1990).
In table 1 characteristics of some Tl-materials are presented.
TABLE 1 Properties of Tl-materials
Material
d
R10°C
T dif
9 dif

(m)
(m2.K/W)
(-)
(-)
PC honeycomb
0.10
1.10
0.75
0.82
PMMA foam
0.015
0.48
0.55
0.57
aerogel granules
between 2 PMMA plates
0.020
+0.006
1.17
0.37
0.42
monolythic aerogel
between 2 glaspanes 1 atm
0.1 atm
0.014
+0.006
0.84
1.92
0.55
0.55
0.6
0.6
double glazing low-E
coating argon filled

0.77
0.5
0.6
Ideal

00
1.0
1.0
R: thermal resistance; r: direct solar radiation transmittance; g: total solar energy
transmission coefficient
The influence of the design and orientation of the TIM-wall on the instantaneous and mean
net heat gain and temperature distribution in the material has been calculated with a dynamic
model of the TIM-wall. Figure 3 shows the calculated monthly net heat gain for a chosen
"reference TIM-wall", orientation South (see fig.3.). Figure 2 shows the effect of TIM
thickness on the mean net gain over the mid-winter period. Apparantly the 10 cm honeycomb
is not yet the (energetic) optimum. There exists an optimum thickness, because for infinite
thickness the net heat gain will decrease again to zero (zero heat loss, zero solar transmissi-
on).
In fig. 1 one can see high heat gains during the summer period. Possible solutions to prevent
this are shading devices, overhangs, blinds etc. Unfortunately a number of these solutions
have to be mechanicly controlled. A cheap solution is the opening of ventilation slots during
the summer season.
Calculations have shown that natural convection in the airgap between the wall and the TI
material can reduce the unwanted heat penetration in the house by two-third s.
-------
2935
net heat gain
(monthly mean)
,J
WW
123456789 10 11 12
MONTH
OJ
E
(D
7
2 4 6
10 1? 14 16
TIM thickness [cm)
Fig. 1. The monthly net heat gain;
"reference" TIM-wall, South
Fig. 2. The effect of TIM-
thickness on the midwinter
net heat gain.
ENERGY SAVING POTENTIAL
Translucent insulation applied in daylight openings could lead to an annual energy saving of
500-700 MJ per m2, compared to double glazing.
Calculations also indicate that for the Netherlands the possible energy savings of TIM applied
on a South wall will be about 700 - 1200 MJ per m2 TIM-wall of current design, dependent
on the heat demand of the house and compared to opaque insulation.
A TIM-wall has the advantage over a daylight opening that the heat gain can be better
distributed and shifted in time.
In practice, the heat gain depends of course on the properties of the TIM-wall, the dynamics
of the house and the occupants behaviour. For retrofit situations one should add the energy
saving by the thermal resistance of the TIM.
For the future perspective one should realize that this TIM-wall even in mid-winter still yields
a net heat gain, while other energy saving measures usually show progressively increasing
costs when they approach the zero heat loss level.
The preliminary conclusions are that TIM-walls are quite promising; for large scale appli-
cations, however, a cost reduction is necessary. Present high costs are mainly due to the com-
plexity of the overall design which is related to the vulnerability of the material.
TESTCELL EXPERIMENT
In order to prevent the introduction of TIM-walls in the building,which leads to
disappointing experiences, more information is needed on the performance of TIM-walls
under realistic conditions.
During 1989 a measurement was improvised on a prototype TIM-wall in the PASSYS
testfacility at TNO (figure 3).
The PASSYS testcells are part of the European network of identical passive solar testfacilities
-------
2936
which are under development at 12 testsites spread over the EEC countries (see also other
presentation).
This test showed clearly the positive effect of the TIM-wall, but also revealed some short
comings in the design and in the set-up of the test. Major design shortcomings were heavy
condensation between the panes and air leakage through the cavity (stack effect!).
This facade (see fig. 3) had the following elements:
* limestone wall of 21 cm;
* airgap of 6 cm;
* clear glass thickness 8 mm;
* PC honeycomb, thickness 10 cm;
* clear glass thickness 8 mm;
The frame material is wood.
Figure 5 gives the measured heatflux through the facade.
Figure 5 also shows the model output after identification of the TIM-wall parameters.

INDOOR
Fig. 3. Construction of
the prototype TIM-wall.
Fig. 4. Two of the PASSYS testcells
at TNO; to the right: testcell with
prototype TIM-wall.
Fig. 5. Measured and
identified TIM-wall
net heat gain during
PASSYS test.
250 260 270
me (DAY-MjKes^. 1989)
calculated —
-------
2937
SECOND TIM-WALL TEST AT TNO
Within the framework of follow-up studies, in February 1991 a test was started on a more
'commercial' type of TIM-wall in one of the PASSYS outdoor testcells at TNO. Due to the
coinciding interests of various PASSYS participants a common choice was made for a design
from the Fraunhofer Institut for Solar Energy Systems in Freiburg. The main elements of the
TIM-wall are: glass cover, roller blind, 10 cm honeycomb, glass, 0.2 m thick wall, see figure
6.
The test will lead to reliable information on the heat loss, heat gain and heat accumulation
properties of the wall. The intention is to continue the test at TNO over a long period, to
study the long-term effects like: durability, seasonal effects, and such.
First results will be presented at the Congress.
TNO has as an addition to the outdoor testcells developed a movable coldbox which can be
placed in front of the wall under test in order to measure directly the thermal transmission
properties under well controlled (cold and dark) outdoor conditions. If already available, first
results on a testcell/coldbox test on the TIM-wall will be presented too.
cross section. vertical . B - 9
T1 - wall element
roller blind unit
;? ju
! J
mounting
-TIM-

\ / / s > />
H
Kvall J
! ["mounting!

\ \ air filter
\ frame sealing
Fig. 6. View on the TIM-wall selected for the second test.
-------
2938
FURTHER THEORETICAL STUDIES
Moreover, in order to prevent that the introduction of TIM-walls in the building practice leads
to disappointing experiences, more information is needed on the effect of TTM-walls on the
heat balance of the building. Does the heat gain coincide with the heat demand of the
building, or does it lead to overheating problems.
Therefore, in December 1990 a follow-up study started to investigate the effect of dimension,
design, and orientation of a TIM-wall on the energy consumption and thermal comfort in
dwellings.
The study is carried out with a detailed building simulation model and should lead to
simplified design guidelines.
As far as the effect on energy consumption is concerned.it is the intention that this results in
an extention of the correlation method 'TCM-heat' developed at TNO with a TIM-wall
option. TCM-heat calculates the monthly mean heat demand; a key concept in this method is
the 'utilization factor' of the gains from sun and internal sources.
At the time of writing this contribution results from the current test and calculations were not
yet available.
Results will be presented orally at the Congress.
CONCLUSIONS
The first theoretical and experimental results on the application of translucent insulation for
passive solar walls are promising. The research on the TT-wall concept will be continued; with
the emphasis both on the long and short term goals. To mention are material development,
material modelling and testing, guidelines for Tl-wall application; testcell experiments and
demonstration projects.
ACKNOWLEDGEMENT
¦Many of the activities reported here have been carried out with the financial support of
NOVEM (the Netherlands Management Office for Energy and Environment).
REFERENCES
Dijk, H.A.L. van, et al. (1990)
Translucente thermische isolatie in gevels; een verkenning; TNO report, Febr. 1990.
(Translucent thermal insulation in facades; a feasibility study).
-------
2939
THE WORLD'S LARGEST DEMONSTRATION OF TRANSPARENT INSULATION FOR BUILDINGS
J.W. Iwidell
Energy Studies Unit, University of Strathclyde, Glasgow, Scotland
ABSTRACT
The new 376-student residences at the University of Strathclyde have been
designed using low-energy principles of insulation and heat recovery, and with
1000 square meters of transparent insulation on the south facade. The building
is a European Commission energy derronstration, first occupied in November 1989
and since monitored by the Energy Studies Unit. Early results are reported,
indicating the success of the principles, but also mentioning difficulties of
practical implementation.
INTRODUCTION
The principles of Transparent Insulation Material, TIM, are new becoming known
for passive solar building (e.g. Jesch 1989). Fig 2 outlines the
characteristics of the TIM system which we are reporting, namely the LEGIS
system of the Fraunhofer Institute for Solar Energy, Freiburg (FIFSE 1989).
Solar "short wave" radiation, both direct and diffuse, passes through the
single glazing of the outer surface, across an approximately 120 mm gap and
then through the TIM to be absorbed on the dark painted south wall of the
building. Heat loss from the outer surface of the wall by convection and long
wave radiation is greatly reduced by the presence of the TIM and outer
glazing, so the absorbed heat passes to the interior of the building. To
increase the insulating properties of the LEGIS construction at night, and to
prevent overheating in the day, a lew emissivity blind should be pulled closed
in the gap between the outer glass and the TIM.
The initiative for the project came from the Kaiser Bautechnik company of
Germany. Buildings in Scotland have long heating seasons from
September/October to May/June in south Scotland, and for nearly the entire
year in the northern Scottish islands. Therefore the economic benefit of solar
heating at these higher latitudes may be maximized, as first pointed out by
Kerr MacGregor (1981). The high proportion of diffuse solar conditions
increases the challenge for passive solar gain.
Hie site chosen is on a steeply sided east-west ridge in the centre of the
city near the Cathedral. There is little solar obstruction to the south. The
developers employed Glasgow architects of the Kennedy Partnership to design
the 5 and 6 floor blocks, Fig 1, in two east west lines along the top and
middle of the ridge (Edwards 1990). The Unit was then instrumental in
obtaining a significant grant from the Energy Directorate of the European
Commission for the construction and monitoring. Construction took place in
late 1988, with handing over commencing in November '89 with immediate student
occupation and the beginning of the installation of the on-line monitoring
system. During the first winter there was considerable condensation as the
building dried out and many building snags to rectify. Due to both technical
and other difficulties the LEGIS wall blinds have, to date of March '91, not
been installed.
Ihe Energy Studies Unit monitors energy performance, with details and seme
early results published in Fbrrest^ et al (1990 a,b). Detailed monitoring (100
-------
2940
air, water, wall and air duct temperatures; 6 wall heat fluxes; water flew; 28
electric power outlets; 1 gas consumption; 16 window and door openings;
recorded video tapes for blind positions; 6 solar radiation fluxes; wind speed
and direction; psychrometer) includes (i) one occupied flat of 4 students,
(ii) an unoccupied "manikin" room, and (iii) outside environment. Occupancy
and wall moisture sensors are being installed.
BUILDING OUTLINE
Study/bedrooms are to the south; the corridor, shewer rooms and toilets are
central; self-catering kitchens are to the north; corrnon rooms either cross
from north to south, or are predominantly to the north. LEGIS TIM components
are on due-south wall sections, with none on the southeast, southwest, west,
east and north walls. Window area is minimized to the north. Note that the
glazed stairwells are outside the insulated envelope. There are 9 stairways
for entrance to the apartments in each of 18 vertical sections. The sections
are arranged in 4 buildings. Each section has its own gas boiler for domestic
hot water, and its cwn electricity meter. Apart from our monitoring, there is
no further metering, and students pay a comprehensive rent that includes
electricity and hot water.
Apartments are (i) 4 student with one set of service rooms, and (ii) 8 student
with two sets of service rooms. East, north and west walls have 150 mm
rockwool insulation (twice the UK standard) with a brick outer layer and dense
block inner having a total thermal transmittance (U value) of 0.22 W/(m2 K).
All floors have 100 mm styrafoam insulation. North and south corner windows
are triple glazed; other windows are double glazed; and most windows have
internal blinds. The study/bedrooms have lew-emissivity roller blinds
positioned by electric motor drives; control is (i) by the central control
unit of the building that can open and close the blinds by time of day and by
a temperature/solar radiation algorithm, and (ii) by the occupants using a
switch at the window. Kitchen and the main common room windows have internal
adjustable Venetian blinds under manual control only.
Air-to-air heat recovery from forced ventilation is a feature of the design.
Extraction is from the toilet, shewer and from above the cooker; these ducts
then join, at the compact heat recovery unit placed iumediately above the
cooker, before the air is expelled. Incoming air passes through the heat
recovery unit to the cotnron room and corridor. The ventilation is on
continuous lew power (nominally 0.5 ac/h), which increases for about 10
minutes when triggered by the toilet or shewer light, or by the cooker
ventilation switch. Heat recovery efficiency is 70 to 80%. There is no forced
ventilation with the study/bedrooms.
Domestic hot water for each section of apartments is continuously circulated
from gas boilers in the roof space. The roof structure is aligned to accept
solar collectors for water heating which may be fitted later. Fixed lighting
is with lew-energy compact fluorescent lamps. The cookers are electric.
Back-up heating is by: (a) 750 W electric convective wall heaters in the
cannon rooms with individual thermostatic control, (b) 200 W electric "heated
plate" heaters in the study/bedrooms without thermostats and placed beneath
the study work-surfaces, themselves placed below the window. The latter
heaters can be activated at any time by the occupants, but are disabled at
regular preset times by a central control system, usually every 4 hours.
THE SCUIH TIM WALL CONSTRUCTION
The TIM used at Strathclyde is of honeycomb construction, consisting of open
apertures, each 3 rrm x 3 mm, to square section horizontal "tubes" 100 mm
long. The material is manufactured from extruded polycarbonate in slabs of 5
apertures height (15 nm) and 700 mm width. The slabs are later cut by the
-------
2941
manufacturer to give the 15 x 700 x 100 rim sections subsequently stacked as a
vertical pile up the wall. The TIM sections are held in an aluminium frame,
backed to the north by a large vertical sheet of transparent polycarbonate
held loosely against the TIM in the frame, and open to the south. Horizontal
plastic bars with a "T" section are placed in the TIM pile at intervals of
about 1000 mti to give fixings for lateral stability.
The LEGIS system wall blinds, each 700 mm wide and 12 to 15 meters high, cover
one vertical column of TIM to the complete height of the building. Each is
pulled upwards from the wrapped position at ground level by an electric drive
and a falling counter weight. The counter weight is a metal plate of about 10
iim thickness moving in channels between the blind and the TIM. The south
structural walls are made with solid heavy blocks of density around 2100
kg/cubic meter and painted dark blue on the outside for solar radiation
collection.
EXPERIENCE
The failure to yet fit the LEGIS blinds, and the presence of building faults
still under contention, precludes the presentation of detailed results at this
stage (March 1991). However certain conclusions can be made.
(a) Fig.3 demonstrates the heat transfer through a south wall section covered
by a TIM LEGIS module, but without the blind which has not yet been fitted.
The room was occupied. Heat into the building is defined as negative. Note
the approximately 3 hour delay between the peak of heat flux on the exterior
and interior wall surfaces, and the further 1.5 hour delay for peak internal
wall temperature. Considering averaged heat flux over periods of weeks, the
LEGIS covered south walls have negative U values.
(b) During the conventional heating season internal temperatures have always
been acceptable (above 20 * C, and often around 23 * C in winter) in the
kitchen/common rooms, due to the high insulation, casual gains and ventilation
heat recovery. Nevertheless students set the small back-up heater thermostats
to higher temperature, about 23 to 25aC; they are not paying extra for the
heat and have no self interest in economising.
(c) The continuously operating ventilation fans create low amplitude noise
that is annoying. Changes of fan speed, especially at night, are particularly
noticeable.
(d) In the surrmer of 1990, without the LEGIS blinds in place, the buildings
severely overheated on at least 3 periods of several days each. Interior
temperatures rose to 35 ° C, despite windcws being wide open to the outside
ambient air at about 28°C.
(e) The study/bedrooms have no forced ventilation, and air movement with the
corridor is prevented by the tightly fitting, self closing doors required by
the fire prevention regulations. Each such room is therefore thermally
isolated from the rest of the building apart from conduction through the heavy
block walls. Such an internal thermal break should not be part of passive
solar design. In practice unoccupied room temperatures without heating are
around 17 " C in winter, which is much higher than normal in Glasgcw. However
the contrast with the 4 to 6#C higher canton room temperatures gives the
perception of discomfort. When occupied, the study/bedrooms with the LEGIS
insulation reach 19 to 21 C from 17 C in a reasonable period. Ventilation is
by a slot ventilator at the top of rrost windows, and by opening the windows.
The latter action leads to heat loss and possible entry of rain onto the study
work surface; both results are unsatisfactory.
(f) Condensation has occurred on the inside surface of the outer glazings c£
-------
2942
triple glazed bedroom windows. Such windows have sealed doable glazing on the
inside, a space for the blind, and thai another glazed frame. Ventilation has
not always been satisfactory from the blind space to the outside in order to
prevent or clear condensation. The fault is being corrected.
(g) Within 4 months of fitting, dark staining was seen to have built 15) as a
very thin deposit within the TIM honeycomb. Analysis shewed the black deposit
to contain carbon as well as building dust. The staining is likely to have
resulted from city traffic air pollution. The staining is very thin, and we
have yet to determine the decrease in solar radiation transmittance.
(h) The dark staining and the occasional condensation indicate the need to
redesign the nodule construction. We shall conduct air movement and air
infiltration studies to investigate the effects.
(i) Overall we have no doubt that the TIM principles of both solar gain and
thermal insulation cire successful. When combined in a lew energy building,
interiors will be comfortable, with zero or small heating demand above casual
gains. Blinds to prevent overheating are essential in the Strathclyde
buildings.
REFERENCES
FIFSE (1989) "Transparent insulation technology for solar energy conversion",
Fraunhofer Institut fur Solare Energiesysteme, Freiburg, Germany.
Forrest R.J., Stuart C.F. and Twidell J.W. (1990) "Transparent insulation -
preliminary results from monitoring the passive solar residences at
Strathclyde", Proc. North Sun, A. Sayigh (Ed), Pergarrcn Press, to be
published.
Forrest R.J., Twidell J.W. and Stuart C.F (1990) "Transparent insulation -
student residences in Glasgcw", Proc. North Sun, ibid.
Jesch L. (Ed) (1989), "Proc. 3rd Workshop on Transparent Insulation
Technology", Franklin Co., Birmingham, UK. See also the Proceedings of the
previous Workshops.
MacGreqor A.W.K. (1981) "Solar heat for the Highlands and Islands of
Scotland", Proc. Energy for Rural and Island Ccrmunities, J.W. Twidell Ed.,
Pergamon Press, Oxford, UK. [Kerr MacGregor has later papers on this thane].
South facade of one
of the 5 residence
buildings.
-------
2943
8L
E
Transparent Insulation Material
Approx 20°C
Glass
Room interior
Automatic blind, Insulates at -
night, In winter, preventa
overheating by day in summer
Windows are triple-glazed and also have electric-powered blinds
Transparent Insulation Material
Direct
sunshine
Diffuse
sunshine
Opaque wall of heavy building
blocks which slowly conduct
and atore the heat
Room interior

10 cm Approx 20 cm
Solar heat can only pass into the building
Radiation from both direct and diffuse sunshine is absorbed into the wall as heat
3 mm
Channels
Backed by transparent polycarbonate sheet
Side
Front
Transparent Insulation Material - open channels
of transparent polycarbonate material
Fig, 2 • Diagrams of the Transparent Insulation Material and the LEGis wall
-------
2944
or \
C. ID
DAILY AVERAGE IMSOL.AT'i -
a 011*1 WKti(.At COU'H. IOO.-1
i"!
|:J lin
I i
DAILY AVERAGE AIR TEMPERATURE
ElfTCWH. :990
27* 275 276 277 278 279 2BO 701 202 283 ?H4 28S 236 ?H/ 280 289 290 291 292 293
DAILY HEAT FLUX
INTERNAJ. WAU. 13f>0
¦M
S, v. M

X
V
V
/
/
/
/
$
El
0

J7 lid!
: M ,V<> ?no 201 VM2 283 2U4 ?*'> ?nf. 2u - ~VJ : j<> 7-1'
Coy
daily average room air temperaiure
1990
« 275 27G 277 278 279 200 28' 282 283 204 2BS 286 207 28fl 289 290 291 ?*? T9J :
Julian Day
Fig.3,(a) Daily solar radiation/(W/m2) (c) external temperature/C
(b) interior south wall heat flux/(W/m2) (d) internal tenperature/C
For a study/bedroom over several days in October 1990.
-------
2945
THE POTENTIAL UTILIZATION OF TIM IN
PASSIVE SOLAR BUILDINGS IN CHINA
*
H.X. Yang , B.J. Brinkworth and R.H. Marshall
Division of Mechanical Engineering and Energy Studies,
University of Wales College of Cardiff, PO Box 917,
Cardiff, CF2 1XH, U.K.
ABSTRACT
This paper describes some simulation results for the potential utilization of
transparent insulation material (TIM) in passive solar buildings in China. As
northern and central China have a very long winter and clear weather and
southern China has a shorter winter, but a not very low temperature climate,
the potential of applying TIM to new passive solar buildings or to existing
houses is very great.
The measured meteorological data are processed for six different locations
over China. They are input Into the TIMWAL simulation model, developed for
simulating TIM passive solar buildings. Results show that the increment of
indoor air temperature could reach 5 to 10 degrees, depending the location of
the building considered.
KEYWORDS
TIM; transparent Insulation material; China; simulation; passive solar.
INTRODUCTION
Coal 1s the main heating energy source for buildings 1n northern and central
China 1n winter. Severe air pollution 1s caused. In southern China, most of
the buildings have no heating apparatus because of energy shortage, even
though Its ambient air temperature 1s quite low in Jan. and Dec.. In order
to save energy, heating usually starts late 1n winter and early 1n spring,
e.gvthe real heating season is from mid-Nov.to mid-March in the Beijing and
Tianjin areas, although the room temperature is below the human body
* This author is from Tianjin University, Tianjin, P.R. China, who 1s now
studying at University of Wales College of Cardiff as a PhD student.
-------
2946
requirement outside that season. The air pollution and energy shortage for
building heating in winter are challenging the people and its environment.
Transparent insulation technology for passive solar utilization in buildings
is coming closer to the aims of market penetration . It is an outstanding
new approach due to the huge potential for saving conventional heating energy.
It has advantages of being easy to retrofit to existing buildings and being
capable of preventing absorbed solar energy loss to the environment. The
possibility of applying this kind of material to the buildings in China is
therefore promising.
In northern China, especially north-west China, the temperature is quite low,
but the solar radiation flux and the sunshine hours are suitable for passive
solar utilization. In southern China, although the sunshine hours are not as
long as in northern China, the temperature is higher. The TIM passive solar
building perhaps is the solution to heating problem in winter. It is
therefore necessary to analyze China's meteorological data and run dynamic
computer simulations in order to provide reliable predictions of the thermal
performance of TIM passive solar utilization in China and to obtain the basic
information for further developing it throughout this country.
METEOROLOGICAL DATA ANALYSIS
The measured meteorological data for six locations were analyzed, representing
the weather situation throughout China: Haerbin in far north, Tulufan in
Northwest, Beijing in Northeast, Xian in central west, Shanghai in southeast
and Guangzhou in far south. The original data are the monthly average values,
measured during 1961-1977. The solar radiation on horizontal surface is
changed on to vertical surface and the sunshine hours are calculated by the
equation from Duffie (1980) [1]. The monthly average daily total solar
radiation is obtained by the method of Liu (1960) [3]. As show in Fig. 1,
most areas need space heating for at least five months and the far northern
area even needs a six months heating season. In the central and southern
areas, i.e., Xian and Shanghai areas, the ambient temperature is higher and
Table 1. Monthly average daily total solar radiation on horizontal
surface and monthly average sunshine
hours in
Dec. and
Jan.

location
daily radiation
sunshine
station
latitude
altitude
(kJ/sq
.m day)
hours

(degree)
(m)
Dec.
Jan.
Dec.
Jan.
Haerbin
45.41
171.7
5162.3
7477.6
144.3
161.0
Tulufan
42.58
34.5
6464.4
7649.3
159.2
202.0
Beijing
39.48
31.2
7950.7
9202.6
193.4
209.2
Xian
34.15
396.9
7385.5
8034.5
134. 1
144.0
Shanghai
31.10
4.5
8026.1
8507.6
139.2
153.5
Guangzhou
23.08
6.3
10458.6
9985.5
158.6
148.3
-------
2947
Temp.(degree)
30
20,
-20
—X— Beijing
Guanzhou
Haerbln
H— Tulufan
Shanghai
-30
~B- Xlan
-40
Sep.
Nov.
Dec.
Jan.
Feb.
March
April
May
Month
Fig.1 Monthly average environmental
temperatures at different locations
I (W/sq.m)
800
— Haerbln
Xlan
H- Tulufan
Shanghai
700
600
500
400
300
200
100
7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00
Hours
Fig.2-Average hourly total radiation on
vertical surface facing south in January
the sunshine hours are longer, which can be seen from Table 1.
Therefore, the characteristics of TIM passive solar utilization suggest the
possibility of using solar energy for space heating throughout the whole
winter in these areas. Although the ambient temperature in northern China is
quite low in winter, its clear sky and higher solar radiation flux make it
possible that the TIM passive solar heating can be used in early winter and
early spring, especially in Tulufan and Beijing areas, as shown in Fig. 2.
It can extend the heating season from four months to five months to satisfy
people's needs, while in deep winter, the solar energy can be utilized in
combination with other kinds of energy to save conventional energy.
-------
2948
MODEL BRIEF DESCRIPTION AND SIMULATION RESULTS
Recent results from Goetzberger (1984) [2] and Sick (1989) [5] and others
showed promising results for the TIM passive solar buildings from their
simulation analysis. In order to further investigate this new approach, a
detailed investigation of TIM passive solar buildings led to an updated
modified simulation program, called TIMWA1. It has considered the local
building constructure in China.The Trombe-Wall simulation methods indicated by
Chanessian (1978) [4] and Tasdemirolu (1983) [6] are used. The differential
equations for the TIM and its absorber wall are solved by finite difference
methods. The conductive heat transfer through the TIM under solar radiation
r7] is considered. The calculations are repeated hour-by-hour for the whole
heating season. The temperature distribution in the wall and the temperature
of indoor air at the end of each hour can be calculated.
For the calculation of possible TIM passive solar buildings in China, values
of 18 degrees indoor air temperature, double glazing U value of 2.28 W/sq.m
degree, 270 mm brick wall and south-oriented TIM wall were chosen. There is
no special insulation for other walls, which represents the existing situation
of the buildings in China.
Part of the output from TIMWAL is listed in Fig. 3. In the Beijing and
Tianjin areas, the TIM wall can Increase the indoor air temperature by about 5
degrees. This can make it possible to use only solar energy for space
heating in Nov. and March in this area, which will extend the heating season
by about 20%. Even in deep winter, solar energy can save conventional
energy, as shown in Fig. 4. Solar energy can increase the indoor air
temperature up to 10 degrees in Xian and Tulufan in Jan., where there exists
the best solar radiation and sunshine in deep winter.
From Fig. 4, it is also seen that the indoor air temperature can be kept above
7 degrees in Jan. in the Shanghai area, so that most of the heating energy can
Temperature (deg.)
20
40
60
0
Indoor air
ambient air
8 12 16 20 24 4 8 12 16 20 24 4 8 12 16 20 24 4 8 12 16 20 24 4
Hours
Fig.3 Temperature profiles of monthly
days in Beijing area from Oct. to Jan..
-------
2949
20
15
10
5
0
-5
-10
-15
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1 2 3 4 5
Hours
Fig.4 Temperature profiles of a monthly
average day in three locations in Jan.
be supplied by solar energy with the utilization of TIM. It may change the
heating situation in southern China, where most buildings have no heating
systems. Solar energy may play an important role here to solve the problems
of the conventional energy shortage.
ENERGY SAVINGS
A summary for part of the heating season at the above six locations is given
in Table 2. It is assumed that the room temperature is kept at 18 degrees.
Table 2, T IM passive solar building performance summary for Nov.
and Jan. (Monthly average dally summary, unit: kJ/sq m)

incident solar
location month
solar energy
percentage
energy 0n TIM
transmitted
of solar
wal 1
to indoor air
heating (X)
11321 7925 30
11727 8795 53
13933 10589 100
11957 8250 48
13452 9955 100
11133 7793 53
8880 6749 100
10329 7702 60
10010 7508 100
10438 7202 79
Temperature (deg.)
—Tulufan(amblent)
—h- Tulufan(lndoor)
Shanghal(amblent)
-B- Shanghal(lndoor)
-K- Xlan(amblent)
-0- Xlan(lndoor)
-B—~ ~ ~ ~ ~ D—G--
, 0—

Haerbin
Tulufan
Beijing
Xian
Shanghai
Jan.
Nov.
Nov.
Jan.
Nov.
Jan.
Nov.
Jan.
Nov.
Jan.
-------
2950
It is shown that solar energy can totally satisfy the heating needs in Nov.
and March in most areas except northern China, while the percentage of solar
heating is still promising in other months of the heating season.
CONCLUSIONS
Simulation indicates that north-west and southern areas have the most
potential for using TIM passive solar heating systems for
building throughout the winter, while northern China, especially Beijing and
Tianjin areas can use TIM walls to increase indoor temperature in early
winter and early spring, so that the heating season can be extended from four
months to five months.
For the large area of southern China, which has no heating systems in winter,
TIM passive solar building is a possible solution.
Economics will be one of the main problems facing TIM utilization in China.
In general, it will be more practical if TIM can be produced in China
economically.
ACKNOWLEDGMENTS
The author wishes to express his thanks to Prof. Wang Rongguang, University of
Tianjin, Tianjin, P..R.China , for her contribution of Chinese meteorological
data.
REFERENCES
1. Duffie, John A. and William A. Beckman (1980). Solar Engineering of Thermal
Processes, John Wiley & Sons, New York.
2. Goetzberger, A.., J. Schmaid and V. Wittwer (1984). Transparent insulation
system for passive solar energy utilization in buildings, Int. J. Solar
Energy, Vol.2, pp.289-308
3. L1u Benjamin Y.H. and Richard C. Jordan (1960). The interrelationship and
characteristic distribution of diffuse and total solar radiation, Solar
Energy, Vol.4, pp1-19.
4. Ohanessian, P.. and W.W.S. Charters (1978). Thermal simulation of a passive
solar houfo using a trombe-michel wall structure, Solar Energy, Vol.20,
PP275-281.
5. Sick, F., W.S. Wilke and J.P.. Kummer(1989). Simulation tools for predictic
of the thermal behavior of transparently insulated buildings, 2nd European
Conference Architecture, Paris.
6. Tasdemlrolu, E. (1983). The performance results of trombe-wall passive
systems under Asgean Sea climatic conditions, Solar Energy, Vol.30,
pp181-189.
7. Cross, B. and H.X. Yang (1990). Thermal conduction measurement of
transparent honeycomb insulation material under solar irradiation, 1st World
Renewable Energy Congress, Reading, UK.
-------
2951
VALUATION OF TRANSPARENT INSULATION DEVICES
Karin Jahn, Dirk Christoffers
Institut fur Solarenergieforschung (ISFH)
Sokelantstr.4, D-3000 Hannover
ABSTRACT
To valuate basic types of transparent insulating components and to
optimize the solar transmission and the heat transfer coefficient,
several test elements were built. A size of 20 cm x 30 cm was
chosen and wood was selected as frame material. The details of the
constructions were changed gradually so that the influence of each
measure could be ekamined. We started from a very simple element -
10 cm transparent insulation embedded in two glass panes. To
simulate the possible mounting at houses, spaces between the
insulation and the surfaces of the elements were introduced.
Afterwards foils were situated in the spaces to improve the thermal
insulation. The global transmission and the heat transfer
coefficient were determined at all stages of the development.
To reduce the weight of the components the transparent insulation
can be fixed to the cover pane by an adhesive and the other glass
pane can be substituted by a plastic foil. Therefore the
transmission and the aging under radiation of several adhesives
were investigated.
On the basis of the results of the experiments and with respect to
the coasts and the handling of the components two constructions are
selected.
KEYWORDS
Transparent insulation, component design, U-value, transmission
spectra, solar transmission, adhesive, aging, irradiation.
INTRODUCTION
For the application of transparent insulations at houses the design
of components is a critical problem. After the characterization of
a multitude of transparent insulations (Platzer 87, Wittwer 89)
several authors reported on the construction of efficient
transparent insulationd devices (Goetzberger 84, Boy 89, Bollin 89,
Wagner 89) . Most of these components were designed for tests at
houses and overheating was found to be a severe problem (Boy 86).
We decided to observe the basic construction principles in a first
step. In a later step the problems of overheating and shading will
be treated in connection with out-door tests.
-------
2952
RESULTS AND DISCUSSION
Element Construction
For the application at walls transparent insulation components have
to fulfill many demands. They must be resistant against rain,
humidity, high temperatures, solar radiation and dust and they must
work reliably for many years. Besides these requirements, it must
be possible to incorporate our components into the wooden framework
at our experimental houses and have frames easy to handle so they
can be exchanged after out door tests.
TABLE 1 Transparent Insulation Devices
Version
Structure
T»
U- value
(W/mlK)
no.O
glass /12 cm space / glass
0.81
3.90
no.1
glass / TIM / glass
0.69
1.13
no.2
glass / TIM / 2 cm space / glass
0.77
1.10
no.3
glass / TIM / PMMA-foll / 2 cm space / glass
0.73
1.02
no.4
glass / TIM / Hostaflon-foll / 2 cm space / glass
0.71
1.02
no.5
glass / TIM / foil / 2 cm space
0.77
1.05
no.6
glass / 2 cm space / TIM / 2 cm space / glass
0.75
1.06
no.7
glass / 2 cm space / TIM / PMMA-foll / 2 cm space / glass
0.69
1.00
no.8
glass / 2 cm space /PMMA-foll / TIM / PMMA-foll / 2 cm space / glass
0.67
0.94
no.9
glass / 2 cm sp. /Hostaflon-foll / TIM / Hostaflon-foll / 2 cm sp. / glass
0.63
0.94
Because of the wooden framework at the houses the frames of the
test elements, 20 cm x 30 cm, are made of wood. Also PVC and
aluminium will be tested. The main features of all constructions
are listed in Tab. 1. A 4 mm cover pane guarantees the protection
against the weather. For all components, execpt no.l, glass with a
very low iron content and a solar transmission t = 0.90 is used.
The transparent insulation is a 10 cm thick Polycarbonate(PC)-
capillar structure with 5 mm X 5 mm quadratic cells. Other
insulations will also be examined.
In the simple version no. 1 the transparent insulation is directly
attached to the cover pane and the rear pane. The samples no. 2 -
no. 4 are designed with 2 cm distance to the rear. This is a
characteristic distance between the transparent insulation and the
wall. The space is necessary to avoid direct contact with the hot
wall - more than 100°C have been observed (Boy 86). These
temperatures could destroy the PC-capillars. The decoupling also
reduces the heat losses via conduction. In no. 3 and no. 4 an
additional foil is attached to the back of the insulation. It acts
as a further convection suppression, but mainly protects the
insulation against pollution by dust and condensation. Besides, a
foil tightly attached to the insulation can probably substitute the
rear pane. This would drastically reduce weight and costs.
-------
2953
The third type shows 2 cm space between the insulation and the
front pane and the insulation and the rear pane. This corresponds
to a system with a shading device in front of the insulation and a
distance between insulation and wall. Again one or two foils are
introduced to improve the suppression of convection. Two different
foils, PMMA and Hostaflon were tested. The PMMA foil has a solar
transmission of r = 0.89, the Hostaflon foil of 0.86. The latter
shows a higher amount of scattering but it is stable up to 170 °C
and it is resistant against UV-light. These features are important
for passiv solar applications as well as for transparent
insulations in active solar components. Also the solution of the
next problem, the adhesive connection of the transparent insulation
with the front pane, is of interest in this field of research.
Adhesives
Especially when the rear pane is substituted by a foil, a tight
connection between the front pane and the transparent insulation is
necessary for the stability of the component. This can be realized
by an adhesive connection. For our application an adhesive has to
have a high transmission in the range of the solar spectrum. Its
transparency and colour must not change under solar radiation, also
because of the optical impression of the components at walls.
TABLE 2 Adhesive for Aging Test
Sample
Adhesive
Treatment
Operable
Temperature Range
Materials
Appearance
Aging under
Solar Radiation
AD 1
siloxane
room temperature
hardening at

aluminium, ceramics
glass, wood, plastics
transparent, but
milky
no change
AD 2
aery late
UV-llght
250 - 400 nm
-55°C / 125°C
polycarbonate
clear,
high transparency
turned yellow
AD 3
acrylate
UV-llght
250 - 400 nm
-40°C /100°C
Glass, wood,
plastics
transparent with
bubbles
turned
light yellow
AD 4
acrylate
UV-llght
250 - 400 nm
-40°C / 125°C
plastics
light yellow with
streaks
turned
yellow
AD 5
polyurethane
hardening at
room temperature

Makrolon, plexiglass
PVC, wood
transparent, but
milky
no change
AD 6
polyurethane
hardening at
room temperature

plexiglass,
wood
clear with
bubbles
turned
light yellow
AD 7
siloxane
baking: 5 mln. at
90 "C
hardening: 5 mln.
at 170-200°C

polyester
aluminium, copper,
polyamlde,
silicon rubber
clear,
high transparency
no change
Seven different adhesives, listed in Tab. 2 were selected and
applied on glass. They were irradiated for 80 hours with a spectrum
similar to the solar spectrum and an intensity of 1.8 kW/m2 . Before
irradiation, after 30 hours, 55 hours and 80 hours the transmission
spectra between 300 nm and 2500 nm were measured and compared with
the optical impression. The solar transmission ts before
irradiation, calculated from the transmission spectra with an
AM 1.5-spectrum, was higher than 0.85 for all samples.
The adhesive AD 2 showed the most drastic changes. Before
irradiation it was clear and absolutely transparent. Already after
30 hours it had turned yellow-brown. The transmission of AD 2 is
-------
2954
£ 0.8
before Irradiation after 30 hours after 60 hours
300 400
500 600 700
wavelength (nm)
BOO 900
Fig. 1 Transmission spectra of AD 2 before and after 30 h and
80 h of irradiation.
displayed in Fig. 1 for several degrees of aging. The reduction of
the transmission in the blue-green region corresponds to the
observed variation of the colour. AD 4, that is also hardened with
UV-light, is optimized for fibres and was yellow from the
beginning. The adhesives AD 3 and AD 6 changed their colours only
slightly. But they exhibit a high amount of bubbles. The remaining
adhesives did not age during 80 hours of irradiation. But AD 1 and
AD 2 are not clear and look milky. Concerning the aging test, the
structure and the transmission AD 7, would be the appropriate
adhesive. Unfortunately it requires high temperatures for
hardening. Thus it can only be used for capillar structures of
temperatures resistant materials like Hostaflon.
The present results propose AD 1, AD 3 and AD 5, eventually AD 6,
as materials for the connection of PC-capillars with glass. But the
aging test will be continued and other adhesives will be included.
Valuation of TIM Devices
For all devices the transmission spectra and the heat transfer
coefficient were determined. The results for the solar transmission
rs and the U-value are listed in Tab. 1. The comparison of the heat
transfer coefficient of the empty frame no. 0 and the sample no. 2
directly illustrates the influence of the transparent insulation:
The U-value is reduced from 3.9 W/m2K to 1.10 W/m2K. The comparison
of no. 1 and no. 2 shows that a space of 2 cm between the
insulation and the rear pane has nearly no influence. The
introduction of one foil results in a reduction of about 0.1 W/m2K.
Thus the U-value of the components with two foils, version no. 8
and no. 9, drops from 1.10 W/(m2K) to 0.94 W/ (m2K) . These were
expected to show the lowest U-value as they contain many elements
which suppress the convection, i.e. the insulation itself and two
foils. Additionally the decoupling of the glass panes and the
insulation minimizes the thermal conduction. A large device, 100 cm
x 100 cm, is expected to show a similar U-value because the heat
-------
2955
1
0,8-
c
o
CO
E 0,6
CO
c
<0
0,4 -
device no.6 device no.7 device no.8
0,2
500
1.000 1.500 2.000
wavelength (nm)
2.500
Fig. 2: Transmission spectra of the devices no. 6 - no. 8
transfer coefficient was measured in the center of the element and
this part dominates in large scale devices.
Although the thermal measurement favours the designs no. 8 and no. 9
one has to regard the transmission in the range of the solar
spectrum. Fig. 2 shows the transmission spectra for the
constructions no. 6 - no. 8 with the same dimensions but different
numbers of foils. As expected the transmission drops and rs
decreases about 8% for two foils. For a comparison of the different
constructions the values of the solar transmission are sufficient.
They were calculated from the transmission spectra with an AM 1.5
solar spectrum. The constructions with the lowest amount of
material exhibits the highest transmission. Sample no.l has a
relatively low transmission because simple glass that contains more
iron was used.
The introduction of foils reduces the transmission about 2% -6% per
foil, depending on the material. The difference between PMMA foils
and Hostaflon foils are small. But despite the higher stability
against temperature the Hostaflon can only be favoured for
applications where temperature above 120 °C are reached and the
high price is acceptable. PMMA or PC foils seam to be a better
solution in buildings.
A further aspect is the weight and the stability of the samples.
The lightest sample is no. 5. It also has a relatively low U-value,
1.05 W/m2K, and a high solar transmission of 0.77. Therefore it was
chosen to be tested in an out door test on the solar houses. But to
reach a sufficient stability this construction demands an adhesive
connection of the cover pane and the transparent insulation. The
adhesives AD 1 and AD 6 were chosen to be tested in this component.
Also the version no. 1 will be taken for the first test because of
its simple structure. But it will be improved by glass with higher
solar transmission.
Regarding only the heat transfer coefficient and the solar
transmission, a choice of the best device is difficult. One useful
further information is the angular dependence of the solar
transmission. To deliver reliable results this measurement requires
a special form of samples which will be built in the near future.
-------
2956
CONCLUSION
The measurement of the heat transfer coefficient and the
transmission spectra of nine test elements of basic constructions
for transparent insulation devices showed that a U-value of 1.13
W/(m2K) can be realized with simple structures. Spaces between the
cover panes and the insulation material and the introduction of
foils reduce the U-value to 0.94 W/(m2K) . But simultaneously the
solar transmission drops. Two constructions were selected for out-
door tests regarding the transmission, the heat transfer
coefficient and the weight of the devices. The weight of the
components can be reduced maintaining high solar transmission and
a relatively low U-value when the rear pane is substituted by a
foil. To reach a sufficient stability of this component an
adhesive's contact of the cover pane and the transparent insulation
material is necessary. An aging test demonstrated that only a few
adhesives are suitable for this application.
The experiments showed that further investigations for the
Valuation of transparent insulation devices are necessary.
Therefore the measurement of the angular dependence of the solar
transmission of test elements and the incorporation of other
transparent insulations and frame materials will proceed the
scaling up of the devices and the out door tests.
ACKNOWLEDGEMENT
At this place we want to thank Mrs. Jacobi for her engaged support
in building the test elements and her help to lay-out the paper.
REFERENCES
Bollin, E. (1989). One year experience with a solar house with
transparent wall insulation. Proc. of the 2nd European
Conference on Architecture (Paris). 618-620.
Boy, E. (1986). Experimental investigations on the thermal
behaviour of transparent insulated walls. Proc. of the 1st
Int. Workshop on Transparent Insulation Materials
(Freiburg). 42-45.
Boy, E., Munding, M. (1989). Transparent thermal insulation
tested in practice: interim results gained in two years of
testing. Proc. of the 3rd Int. Workshop on Transparent
Insulation Materials (Titsee). 94-97.
Goetzberger, A., Schmid, J., Wittwer, V. (1984). Transparent
insulation systems for passive solar energy utilisation in
buildings. Int. J. Solar Energy. 2, 289-308.
Platzer, W. (1987). Solar transmission of transparent insulation
materials. Solar Energy Materials. 16, 275-287.
Wagner, A. (1989). Renovation of a one-familiy house with
transparent insulation. Proc. of the 2nd European
Conference on Architecture (Paris). 337-339.
Wittwer, V., Platzer, W. (1989). Proc. of the 3rd Int. Workshop on
Transparent Insulation Materials (Titsee). 33-36.
-------
3.14 Transparent Insulation II
-------
Intentionally Blank Page
-------
2959
TRANSPARENT INSULATION MATERIAL
DEMONSTRATION PROJECTS AND FUTURE PROSPECTS
P. O. Braun, J. Schmid, E. Bollin, W. Stahl, J. Vahldiek, K. Voss, A. Wagner
Fraunhofer-Institute for Solar Energy Systems, Freiburg, F.R. Germany
ABSTRACT
In this paper, realized demonstration projects with Transparent Insulation Materials (TIM) and
applications in the field of solar hot water systems are presented. Further applications of TIM,
especially for daylighting purposes, were developed.
KEYWORDS
Transparent insulation materials (TIM); TIM-applications; solar space heating; daylighting systems;
process heat collector; integrated storage collector
INTRODUCTION
For the effective conversion of solar irradiation into thermal energy, Transparent Insulation
Materials (TIM) are required (e.g. in solar collectors the glass cover is used as a TIM cover). For
the passive solar heating of buildings, the glazings have this function, too. According to the level of
solar irradiation, the temperature level of the heat produced and the temperature difference to the
ambient, the transparency and a high thermal insulation lead to good thermal efficiency. For glass
covers, the TIM performance may be improved by using more than one sheet but the improvements
are limited by the fact that the reflection losses, which reduce the transparency, may eliminate the
improvements of better thermal insulation. The development of new TIM-structures
(Goetzberger,1984) have led to a substantial improvement in the quality of solar energy conversion
in general. As well as increasing the efficiency of conventional solar thermal systems, developments
in TIM can lead to a series of totally new kinds of solar energy utilization. The development,
optimization, production and application of TIM has produced many impressive results in its short
history and there is still a great potential for further improvements and new applications in the
future.
SPACE HEATING BY TRANSPARENTLY INSULATED WALLS
The new concept of external insulation of massive house walls with TIM combines the advantages of
conventional opaque insulations and of solar collector systems. The dark coloured wall surface acts
like the absorber of a collector by converting the solar radiation transmitted through a TIM into
useful heat. The transmission losses of the wall can be dramatically reduced by TIM and parts of the
solar gains can be stored in the wall itself. The contribution of these gains for heating purposes is
time-shifted, dependent on the wall's material properties (conductivity, heat capacity). The inner
wall surface acts like a large low temperature wall heater. The system provides passive solar heating
without any active heat distribution device. Figure 1 shows the application of transparent insulation
on massive house walls.
Of the large variety of TIMs (Wittwer, 1989) the PolyCarbonate HoneyComb structure (PCHC),
the PolyCarbonate (PC) or PolvMethvlMethAcrylat (PMMA) capillary structure are used most
frequently. Materials now available reach heat loss coefficients of 0.7W/m2K with typical diffuse
transmission values of more than 0.7. Aerogel materials are still subject to research. Although they
-------
2960
are inorganic materials and have superior thermal properties, these properties do not outweigh their
low solar transmission so far.
To prevent overheating in summer and to improve the thermal insulation at night, roller blinds with
layers of very low emissivity and high reflection on both sides have been developed. Other shading
systems like switchable glazings are still under development.
overheating protection
transparent insulation material
absorber
house wall
lemperct'jre profile
protective
arrgop
glczing
Fig. 1. Principle of transparent insulation of massive house walls
The most important aspect of using TT space heating systems is the reduction of the heating energy
demand by increasing the mean wall temperature. During a heating season, the total energy gains
from a transparently insulated wall are larger than the losses. Summarizing our experience from
measurements and simulation calculations, the contribution of TI to the heating energy demand
amounts 100 to 200 kWh/m2. Results from some eight-family houses in Freiburg, Sonnenackerweg
(Fig. 2), owned by a local public housing organization support these values. Three of these identical
houses built in the late 50's were retrofitted with several insulation technologies, as shown in Table
1. Using 120 m2 of transparent insulation as a totally passive solar system, the heating energy
consumption of the Tl-house was reduced by 80 % compared to the previous value. And, there is
still a reduction of nearly 40 % of the consumption of the well-insulated house no. 14/16.
TABLE 1 Properties of the applied Insulation Systems. Freiburg-
Sonnenackerweg
House
Insulation Technology/U-value
Heating Energy Consumption
No.
Orientation
Simulation
Measurement

south
north

W/m2K
W/m2K
kWh/m2
kWh/m2
old
none/1.0
none/1.0
225

6/8
opaque/0.3
opaque/0.3
100
-
14/16
opaque/0.2
opaque/0.2
77
77.5
10/12
transparent
opaque/0.2
45
47.5
-------
2961
Fig. 2. Solar space heating application using TIM, Freiburg-
Sonnenackerweg
DAYLIGHTING WITH TRANSPARENT INSULATION MATERIALS
Advanced TIM structure (Fig. 3) can be designed to incorporate the thermal insulation properties
similar to well-insulated conventional walls as well as other benefits. Replacing a wall with a TIM-
element leads to the following features (Schmid, 1989):
The light is of diffuse nature which prevents glare. If the ratio between the TIM-element area and
the room area is optimized, full daylight illumination can be obtained during longer periods of time.
Because of the large apertures' area, illumination at a higher level is obtained. The integrated roller
blind allows control of the incoming light flux. During the summer, the influx of solar energy can be
adjusted to the minimum requirements for lighting purposes. Undesired room heating is reduced to
a level which could not be obtained by artifical lighting methods. A further reduction of heat influx is
obtained by using a second roller-shade, which reflects the near infrared part of the solar spectrum
leading to a luminous efficacy of 150 - 200 lm/W. During the winter, the conventional way to reduce
glare is to use external shading elements. Both the desired reduction of light and the incoming solar
heating energy are reduced in the same way. Partially absorbing interior shade separates light and
heat influx in such a way that the undesired amount of light is converted into heat on the absorbing
shade.
The large translucent TIM area is of multiple benefit:
"Ilie larger area produces more light which is more uniformly distributed than light from a
window. The transparency of TIM-structure elements for sunlight is further improved by the
elimination of absorption on the horizontal part of the wall surface, which has to be conside-
red in case of glazings.
Due to TIM's good thermal insulation properties (0.3 - 0.5 W/m2K), the inside surface
temperature is nearly at room temperature.
Because of this property, no draft or other thermal uncomfort exists and the working area inside a
building can be extended directly to the TIM wall. This is a further improvement in daylight utili-
zation.
Including the first experiments of wall fitted TIM in 1982 /1/ the Institute is now involved in 22
projects with TI space heating and daylighting applications. Incorporating the advantages of TIM, it
-------
2962
is now planned to build an energetically self-sufficent house in Freiburg, which will obtain its entire
energy from the solar systems on its facade.
Daylight element with good insulation properties leading to high
thermal comfort
TIM-STRUCTURES FOR COLLECTORS
Standard flat plate collectors can be tremendously improved by using highly effective transparent
insulation materials as the front cover (Rommel, 1987). With a 10 cm honeycomb structure made of
polycarbonate, the front losses can be reduced to values of 0.9 W/m2K at T = 100 °C.
100
1=800 W/m
vacuum tube
£ 70
flat plate with
50
sel. flat plate
150
100
50
200
0
Fig. 4.
Efficiency characteristics of different collectors
-------
2963
A reasonable criterion for the comparison of process heat collectors is the measurement of their
efficiencies at an irradiation of 800 W/m2 and a temperature difference of T = 100 °C. Under these
conditions, a conventional flat plate collector with a selective absorber and single glazing has an
efficiency of approximately 20 %, whereas the efficiency of vacuum tube collectors can reach about
50 %. When a 10 cm TIM-cover is installed in the flat plate collector, efficiency values in the same
range than for the vacuum tube type are measured.
Up to now, integrated storage collectors with single or double glazings could be used successfully
only in warm climates. In Central European winter weather conditions, freezing can occur in the
collectors if simple glazings are used. These difficulties can be overcome by using transparent
insulation materials as front covers of integrated storage collectors (Goetzberger/Rommel, 1987;
Schmidt, 1988, 1990).
An integrated storage collector consists of a storage tank whose black surface acts as an absorber for
solar radiation. The tank includes the mains water under standard pressure conditions. The aperture
of the collector is covered with a layer of transparent insulation material to reduce thermal losses
and prevent freezing. The advantages of an integrated storage collector with TIM over a conventio-
nal (solar) DHW-system are
compact design, only one device for tank and collector
reduced aperture because of high efficiency
no heat exchanger, no pump, no control system
no antifreeze fluid
no risk of antifreeze fluid penetration to mains water
These advantages may help to cut down the cost of solar domestic hot water systems.
NEW APPLICATIONS FOR STORAGE COLLECTORS
instantaneous
heater
roof beam
opcque backside
insulation
Fig. 5. Cross section and system integration (examples)
-------
2964
CONCLUSION AND OUTLOOK
The experimental, theoretical and practical work during the last few years has led to an essential
understanding of transparent insulation materials. Materials now available reach heat loss coeffi-
cients of 0.8 W/m2K with diffuse transmission values of more than 0.7. The theoretical models have
been validated by experimental set-ups. Such models and further experiments will be used for
further improvement of the materials.
The initial applications have shown that well-known thermal solar systems can be significantly
improved by the use of TIM and completely new concepts are made possible. Applications of TIM
can lead to the following advantages and aspects:
new architectural possibilities due to the facades' new service
adaptation of the building to changing climatic condictions
increase of thermal comfort due to increased wall temperatures
new daylighting possibilities due to the excellent thermal insulation properties of TIM
process heat collectors with efficiency values in the same range as those for the vacuum tube
type
integrated storage collectors with no risk of freezing under central European weather
conditions
ACKNOWLEDGEMENT
The authors would like to express their gratitude to the German Ministry for Research and
Technolgy which has continuously supported this new technology to the described state of develop-
ment. Many colleagues, who are unnamed here, were involved in obtaining the present results.
Special thanks are due to their excellent work.
REFERENCES
Goetzberger, A., J. Schmid, V. Wittwer (1984). Int. J. Sol. Energy. 2. 289-308.
Goetzberger, A., Rommel, M. (1987). Solar Energy. 39. 211 - 219.
Rommel, M., V. Wittwer (1987). Proc. ISES Solar World Congress Hamburg. 641-645.
Schmid, J., W.-S. Wilke, F. Sick (1989). Proc. 2nd Europ. Conf. Arch.. Paris. 243-246.
Schmidt, C., A. Goetzberger (1990). Solar Energy. 45. 93-100.
Schmidt, C., A. Goetzberger, J. Schmid (1988). Solar Energy. 41. 487-494.
Wittwer, V., W. Platzer f 19891. Proc. 3rd Int. Workshop Transparent Insulation. 33-36.
-------
2965
TRANSPARENTLY INSULATED WALLS FOR BUILDINGS OF LOW LATITUDE
A PILOT PROJECT IN SHANGHAI, CHINA
# ** **
E. Bollin , Qian Shang Yuan , Ling Zhi Guang
*Fraunhofer Institut fur Solare Energiesysteme (ISE)
Oltmannsstr.22, D 7800 Freiburg, FRG
** Shanghai Institute of Energy Research (SIER)
777-29 Lao Humin Road,Shanghai, China
ABSTRACT
At Shanghai two test-houses were built in 1990 to demonstrate transparently insulated walls (TI
walls) as an efficient passive solar component for space heating under the climatic conditions of a
site at low latitude. The facade construction was appropriate to the Chinese conditions and built
with materials, available at the site. An extensive monitoring will help to analyse performance and
solar gains of the TI wall.
KEYWORDS
Transparently Insulated Walls (TI walls), passive solar space heating, monitoring, simulation test,
appropriate technology.
INTRODUCTION
During the past years TI walls have been tested and demonstrated to be an efficient component in
passive solar design in Central Europe (Goetzberger, 1984; Bollin, 1989).
Installed in front of a massive external wall of a residential house, between 100 and 200 kWh/m2-
facade area could be obtained under Central European climatic conditions. Up to 70% of the
annual space heating requirements could be saved by TI walls. Due to the high labor costs, the
investment costs of a transparently insulated facade produced in FRG are between 600 and 1200
DM/m2.
In Central Europe these transparently insulated buildings still need an additional centralized or
decentralized space heating system during the winter.
In China, south of the Yangtse-River, residential houses have no space heating system. During the
winter season the room temperature drops below 10°C. The buildings are constructed with massive
bricks.
Under these conditions transparent insulation has a high potential for application. TI walls can
increase the indoor living comfort by increasing the room temperatures of such Chinese
residential houses without any pollution of the densely populated environment and without
increasing the needs for non-renewable energy sources.
Supported by the Science and Technology Commission of the Shanghai Municipality and the
German Ministry for Research and Technology, the Shanghai Institute of Energy Research
(SIER) and the Fraunhofer Institut fur Solare Energiesysteme (ISE) built two test houses in
Shanghai in 1990. The two houses (TIM House, Reference House) have identical floor plans but
use different insulation techniques. The so-called TIM House (TIM = Transparent Insulation
-------
2966
Material) has transparently insulated external walls and additionally an improved insulation
standard. The Reference House is designed in accordance with the typical Shanghai house, which
has poor insulation values.
This paper will report about the building design, the TI wall design, the results of a dynamic
computer simulation test and the extensive monitoring system, installed in March 1991 in
Shanghai.
_ 6
T>
tNI
£ i
0
30
„ 20
o
— 10
JAN FEB MAR APR HAY JUN JUL AUG SEP OCT NOV OEC
I i Shanghai I I Freiburg
Ta - monthly mean ambient temperature
I - monthly mean value of the daily global
insolation on a horizontal surface
Fig. 1. The climatic conditions of Shanghai and Freiburg.
CLIMATIC CONDITIONS
Shanghai is located in the east of the mainland of China with 31°3' north latitude, 121° longitude. It
has a mild marine climate as it is a coastal city of the East China Sea. Figure 1 shows a comparison
between the climatic conditions of Shanghai in China and Freiburg i.Brsg. (48° north latitude) in
the FRG.
During the winter season Shanghai has more global insolation on a horizontal plane and higher
ambient temperatures than Freiburg. This indicates that in winter TI walls could perform at least
as well in Shanghai as in several pilot projects in Freiburg. Due to the lower temperature
differences between indoor and outdoor conditions, the thickness of the TIM can be reduced in
Shanghai without decreasing the TI wall performance.
The high ambient temperatures in Shanghai in summer as well as the high humidity values of the
Shanghai marine climate have to be carefully considered when designing the TI wall in Shanghai
BUILDING DESIGN
Two one storey buildings were designed and built in front of the main building of the SIER. The
two buildings have a flat roof and a floor plan in the shape of a rectangle (7.5 m x 9.3 m) with the
corners oriented to the south, west, east and north.
1
1-
r
Fi r

•q r
; ri
-1
n
-j
l

111 f
1

-i
-i
-|
-j
1
-|
_ r
1 i
¦ l

]\
-------
2967
Figure 2 shows the floor plan of the two test houses. Two bedrooms are located in the south-east.
The kitchen, the bathroom and a living room are located in the north-west part of the buildings.
During the performance tests, the buildings will be used as SIER offices.
Due to the higher insolation values of the south oriented external walls, the south-east and south-
west orientations are especially suited for TI walls. To maintain the higher room temperatures in
the TIM House during winter, the overall U-value of this house should be decreased compared
with the Reference House. Therefore, the TIM House has double glazed windows with plastic
frames. The northern external walls and the roof of the TIM House were built with sponge
concrete (U = 0.6 W/m2K), whereas massive bricks of 24 cm thickness and a density of 1.8 t/m3
were used for the walls behind the transparently insulated facade.
Tw
Tg
H
Tw
©
H Tr
Q
Tw ©
Tw Q
~J
Tg Tw S ;
Tr E
H^E
©E
Tr
©
Vr
Tr H
0
Tw
I
H
Ta
Tw
S I
Tw
Tw _
© Tr
H
Tr E
H_E
©
©
15
Tr ©
x ©
Tr H
©TW

TIM HOUSE
REFERENCE HOUSE
1 -living room
2 -kitchen
3 -bathroom
4 -bedroom 1
5 -bedroom 2
6 -hall
Tr -room temperature H -humidity
Tw-wall temperature Q -heatflux
Tg -temperature of I -insolation
the glazing S -solar cell
Ta -ambient temperature E -current meter
Fig. 2. Floorplan of the test houses in Shanghai.
In addition, an underground duct and two auxiliary fans were installed in the TIM House as a new
attempt of air conditioning, where cold air is ventilated from the shaded north facade into the
south-east zones of the building in summer.
In both buildings, passive solar heating is the only space heating system used.
SIMULATION TEST
With the Transient System Simulation Program TRNSYS (Klein,1988), the thermal behaviour of
the projected test houses at Shanghai has been simulated at the ISE. The TRNSYS "weather
generator" (type 54) allows the generation of hourly weather data on the basis of monthly data sets
using correlations described by Duffie (1980) and Klein (1988).
For the simulation test, the buildings in Shanghai were divided into four different zones, where
zone 1 stands for room 1, zone 2 for room 4, zone 3 for room 5 and zone 4 for room 2, 3 and 6. To
show the influence of various kinds of insulation techniques on the monthly mean zone
temperatures, several versions of the test house were simulated.
-------
2968
Figure 3 shows the values of the monthly mean zone temperatures in January and July for the
following cases:
case 1: Reference House with no insulation, single glazed windows, no shading device, air change
rate of 1 for the winter season;
case 2: TTM House with 5 cm TIM on the south-west and south-east external walls, no insulation
on the north-east and north-west external walls and on the roof, double glazed windows,
shading device for windows and TI walls, air change rate like case 1;
case 3: TIM House like in case 2, but with 5 cm Styrofoam on the north-west and north-east
external walls;
case 4: TIM House like in case 2, but the roof insulated with 5 cm Styrofoam;
case 5: TIM House like case 2, but with 5 cm Styrofoam insulation on the northern external walls
and the roof;
T1 °C)

ZONE 1

T!0
C]

ZONE
4

30-j

30~|

20-

—

—
Jul
20-

—
Jul
10-

-
Jan
10-

t—

w.
Jon

2
3

4
5

cose

T lcC)

ZONE 2

T leC)

ZONE 3

30q

30n

20-

—

i—
Jul
20-

—

f—
Jul
10-

Jon
10-

Jon

1
2
3

4
5

1
2

4
5

cose

N
case 1 = Reference House
case 2 = TIM House, TIM on SE, SW walls
case 3 = case 2, but 5 cm Styrofoam on NW, NE walls
case 4 = case 2, but 5 cm Styrofoam on the roof
case 5 = case 3 + case 4
Fig. 3. Calculated monthly mean zone temperatures of the Shanghai test buildings in July and
January.
The monthly mean ambient temperature at Shanghai is 3.9°C in January and 28.2°C in July.
Comparing the results for case 1 and 5, the monthly mean zone temperature of zone 2 could be
increased by 6.9 K in January and decreased dy 2.1 K in July. This is due to the TIM and the
improved insulation values. The results shown in Fig. 3 also demonstrate that a TI wall only
performs well if the overall U-value of the unheated building is considerably, and if thermal short-
circuits are avoided.
Comparing the results of zone 4 and zone 2 for case 5, the effects of the passive solar gains of the
TI walls to the room temperature can be seen. In January the temperature in zone 2 is 3.5 K above
the temperature of zone 4, whereas in July the monthly mean value of the temperature in zone 4 is
0.3 K below the temperature in zone 2.
Due to higher transmission losses of the 10 cm honeycombs and the reduced significance of the U-
value under Shanghai climate, 5 cm TIM honeycomb structure performs nearly as well as 10 cm
structure in the simulation test.
The two test houses at Shanghai were designed in accordance with the results of simulation case 1
and case 5. The calculated temperature differences between case 1 and case 5 are the goals of this
project.
-------
2969
TIM FACADE DESIGN
A further aim of the project presented here is that local materials and labor should be employed to
build the TT walls. The German side only provides know-how and the TIM for the facade
construction.
The TIM of 5 cm thickness was fabricated at Kaiser Bautechnik in Dusseldorf, FRG. It is made of
polycarbonate and has a honeycomb structure with an energy transmittance of 87 % and a U-value
of 1.39 W/m2K at 50 K temperature difference
Due to its low thermal conductivity, wood is proven as a suitable material for frame constructions
of the transparently insulated facade in the German pilot projects. In Shanghai wood is available
and a wooden frame construction can be built by carpenters at the site. For overheating protection
white Venetian blinds, available at the Shanghai market, were selected for integration into the
transparently insulated facade.
A
B
C
D
A - wooden frame E - foam rubber
B - glazing F - Polyester film
C - Venetian blind G - Honeycombs
D - cord H - wall
Fig. 4. Section of the transparently insulated facade, built in Shanghai.
Figure 4 shows a section of the transparently insulated facade with a wooden frame, built and
installed in Shanghai. As front cover, 3 mm float glazing was used. The Venetian blind can be
moved with the help of a cord from outside. To maintain the Venetian blind, the upper part of the
front cover can be opened. The backside of this facade module is covered by 50 /im Polyester film,
to prevent convection losses in the TIM at high temperatures. When fixing the modules onto the
black painted wall, an air gap of about 1.5 cm is adjusted between the TIM and the wall surface.
This gap is sealed by a strip of foam rubber, fixed on the wooden frame of the facade module.
-------
2970
The transparently insulated facade has an energy transmittance of about 60% in total. The TI wall
has an U-value of 0.75 W/m2K in total.
The facade modules were prefabricated in the labs of SIER, to protect the TIM against dust at the
site, which would reduce the transmission coefficient of TIM tremendously.
The cost of a transparently insulated facade produced in Shanghai including the 5 cm TIM could
be reduced to about 100 DM/m2.
To test the material and the facade's design under Shanghai weather conditions, two facade
modules were tested over 8 weeks in September and October at the site. After this test, the sealing
of the front glazing and the handling of the Venetian blind could be improved.
In December 1990 the transparently insulated facade, with a total area of about 40 m2, was
installed and the construction work for both houses was completed.
MONITORING
The aim of the monitoring program is to show the influence of TI walls on the room temperature
under the climatic conditions of Shanghai.
Therefore the indoor and outdoor climate is measured by temperature and humidity sensors. The
global solar radiation is measured on the horizontal and the vertical south-east and south-west
area. Because the heating comfort inside a transparently insulated building is very much
influenced by the surface temperature of the external walls, several temperature sensors were
mounted into the plaster of the external walls.
Additionally, the performance of the TI walls under a marine climate at a low latitude is of high
interest. Therefore the temperature profile inside the transparently insulated walls as well as the
heat flux through these walls is measured at the south-east and the south-west oriented external
walls of the TIM House.
The location of the sensors for the monitoring of the thermal behavior of the TIM House and the
Reference House is shown on the floor plan in Fig. 2.
In total 63 data measurements will be automatically recorded every minute and the mean values
will be stored onto flexible discs every 15 minutes.
The monitoring system was set into operation in March 1991 and the first results will be reported
at this conference. The monitoring period will last from March 1991 to April 1992.
ACKNOWLEDGEMENT
The authors would like to thank all the colleagues at ISE and SIER as well as the architect
involved, Mr.Weng Ruiping, for their kind cooperation and support of the presented project.
Special thanks are due to F. Sick and J. Kummer for running the simulation tests at ISE and to
Kaiser Bautechnik in Diisseldorf for providing the TIM free of charge.
REFERENCES
Bollin.E. (1989). One Year Experience with a Solar House with
TWI. Proc. of the 2nd Europ. Conf. on Architecture. Paris
Duffie,J.A. and W.A. Beckmann (1980). Solar Engineering of Thermal
Processes. John Wiley & Sons.
Goetzberger,A., and J. Schmid (1984). Transparent Insulation
Systems for Passive Solar Energy Utilisation in Buildings.
Int. J. Solar Energy. 2.
Klein, S.A. and others (1988). A Transient System Simulation
Program. Engineering Exp. Station Report. 38-12. Solar Energy
Laboratory, University of Wisconsin, Madison.
-------
Solar Gains by Transparent Thermal Insulation
Comparative Measurements on Test Cells over one Heating Season
Matjai ZUPAN, Joie BOSTJANCiC
Institute for Testing and Research in Materials and Structures
YU-61000 Ljubljana, DimiCeva 12, Jugoslavia
ABSTRACT
A test facility with four identical test cells was designed and built to make
comparative tests of solar energy efficiency on different types of walls and passive solar
elements. Measurements were made on a wali with 6 cm capillary structure TIM, compared
to opaque wall, once without and once with thermal insulation, and a wall with window.
These three types represent the most commonly used walls in Slovenia. The measurements
were performed from October, 16,1989 to May, 29,1990.
The transparent insulated walls prove to be a good promise for reduction of heating
energy.
KEYWORDS
Passive use of solar energy, transparent thermal insulation, heating energy reduction, test
cells
AIM OF THE TEST
About 30% of total energy consumption in Slovenia (Yugoslavia) is used for heating in
winter period. This energy should be reduced by any means, the consequences of big
consumption are high environmental pollution, high costs etc. One of the possibilities to
reduce energy consumption is use of solar energy.
The quantity of solar energy in the country during the heating period varies with location.
Temperature inversion and, as a consequence, smog and fog appear quite often in the basins,
where over one half of Slovene population is situated. Although the amount of incident solar
energy is thus reduced, it still is of great interest and value. In the places above the inversion
and near the sea a much higher amount of incident solar energy is available and can be used
for passive heating beyond any doubt.
In the last years the development of transparent thermal insulations in the world and our own
measurements showed some very promising results.
-------
2972
The aim of our research was to evaluate the energy losses through the unit of south
orientated walls of different construction over whole heating season. To reach this goal we
decided to measure the amount of energy needed to maintain temperature in the test cells
with different types of south walls above 20 °C. The comparison was made between the wall
with TTM and walls commonly in use in the country. This quantitative information will
be of help to the architect while making the decision whether the transparent thermal
insulation is to be used or not.
TEST FACILITY AND PRELIMINARY MEASUREMENTS
In order to measure the energy balance of different walls,a test facility with four identical test
cells was prepared, tested and calibrated (Zupan, Sijanec, 1989; BoStjan£i£, Zupan, 1990)).
Data acquisition and control system was designed to measure the data and to control the
temperatures in the cells. Temperatures were measured on different points inside the cells,
on the test walls and in the surrounding. Beside that heat flows and solar energy were
measured - global, diffuse and energy on south orientated vertical surface.
In the cells with 5 adiabatic and one active wall the auxiliary heating system maintains the
temperatures. It is turned on whenever the average temperature in the cell drops under 19.8
°C and is turned off when it risesabove 20.2 °C. The energy needed to maintain these
conditions is measured and presented as the most significant result of the test.
The first series of measurements were made on a wall with poly-metil meta-acrylic (PMMA)
foam as a transparent insulation material (TIM). Measurements of light transmittance and
durability were made on different materials, too (Zupan, 1989). On the basis of the results of
different measurements and observations we decided that TIM with honeycomb or capillary
structure made of polycarbonate is to be used for further tests.
TEST WALLS
For the second series of measurements four test walls were prepared. We decided on the
constructions, commonly used in our part of the country which should be compared to the
wall thermally insulated with TIM.
I. The first wall is made of 38 cm full brick, representing a type of wall, built in the past in the
country. There were some ideas that any thermal insulation on a south wall has a negative
effect on the overall energy consumption that had to be checked.
II. The second wall is a 19 cm hollow brick wall thermally insulated with 8 cm polystirene,
representing a type of construction, used recently.
III. The third wall is made of identical bricks as the second, but insulated with 6 cm of
polycarbonate capillary structure TIM and covered by 4 mm glass.
IV. The fourth test wall is extremely thermally insulated, with double glazed window of about
25 % wall surface.
CLIMATIC CONDITIONS
The measurements started on October, 16, 1989 and continued over the heating season till
May, 29,1990. Partial results from the beginning to the December, 4, were presented
-------
2973
already. (Zupan, BoStjan£i£, 1990).
The place where the measurements were performed is Ljubljana, (46° N, 14° 30' E, 300 m
altitude) situated in a basin with many days of temperature inversion with smog and fog
as its consequences.
The climatic data of the measuring period were:
- External temperature was between -10.9 and 29.33 °C
- Average outdoor temperature was 6.23 °C
- Solar energy on the south orientated wall was between 0.03 and 7.06 kWh/m2day
- Average solar energy on the south orientated wall was 2.25 kWh/m2day
The graphs below show the average outdoor temperature and average daily incident solar
energy per month.
Average monthly outdoor temperature
from Dec., 7 1989 to May, 29,1990
o
Jan
Feb
May
Dec
Mar
Apr
Month
Average daily solar energy
from Dec., 7, 1989 to May, 29, 1990

O
V>
global
fvertical
diffuse
Dec
Jan
Mar
Feb
Apr
May
Fig. 1.
-------
2974
RESULTS
The total consumed energy, used for heating the cells with test walls, during the heating
season 1989/90 was:
139.3 kWh - 38 cm brick wall without thermal insulation
83.4 kWh -19 cm brick wall with 8 cm polystirene
64.5 kWh - thermaly insulated wall with window
35.4 kWh -19 cm brick wall with transparent thermal insulation.
The graph below shows the energy consumption of the cells per month.
Average daily heating demand
from Dec., 7, 1989 to May, 29, 1990
full brick
opaque insulation
mSLf window
Y TIM
12 1 2 3 4 5
month
Fig. 2.
The results of the performed measurements show, that in the presented climatic conditions
the wall with TIM is very suitable for the passive use of solar energy. The energy
consumption depends primarily on the availabel solar energy.
The temperatures measured on the wall with TIM show that the wall must have additional
protection against overheating on the extremely sunny days. On the sunny days the
temperatures on the black painted wall behind TIM raised up to 110 °C, tohi-ch can be
harmful for the material. The inside of the wall was heated up to 37 °C, but the cells
can not exchange energy with other spaces as in a real house, where these temperatures
would be considerably lower.
The graphs below show the daily values of consumed heating energy for each cell together
with average daily outdoor temperature and daily sums of incident solar energy on the south
wall. The x-axis shows days of the measurement starting at December, 7, 1989.
-------
2975
Averoge do1'/ CLtdoo' te.Tpe.'oture
Avero$c dcily sole.' enegy on the south vcrticcl surface
4 L:
j_. -i-U -
V:
lift \
Vi/vM"
, I
¦Wi i
wo 1 ao
> f k!'
Hi"- U'"
hi
w
ill:
O 30
ISO 1JO 1UC

. Auxiliary lectira dc.T0.1d - c'o ly vo'ues
Ce.l without thermol msu:o'icn

Xj ^
ilA
SO 90 1 20 ISO
Auxiliicrv heot;no demand - dc'ly vciucs
tell w'th 3 cm polysiirenc
y
r
Vu

Auxiliicry heating demond - dcily values
Cell with window
o 30
Auxi'liofV heoti.ng demond - dcily vo'ues
CeM wi'.h t.'onsoorer.t insulotion
111
iO eo c
wo 1SO 100
CONCLUSIONS
Fig. 3.
The results on the four test walls pointed out that the wall with transparent thermal
insulation had only about 45% energy losses compared to the wall with opaque thermal
insulation. The efficiency for the solar gains is very high and it shows promise for the
region where the measurements were performed (the basin) and even better for the regions
with low temperatures and a clear sky (the mountains).
The solar energy enters the cell with a delay of about 5 hours, so this type of wall should
be combined with windows for the direct solar energy gains. The windows should,however,
have night protection.
-------
2976
The material to continue the tests with is capillary or honeycombe structure polycarbonate.
On the basis of these results a real object with about 20 m2 of TIM was designed and
constructed for further investigations.
REFERENCES
1. Zupan, M., M. Sijanec (1989), Computer Modelling of Test Cells for Mobile Solar Test
Object, Proc. Clima 2000. Sarajevo, 133-138
2. BoStjancic, J., M. Zupan (1990), Mobile Solar Test Object, Proc. 1st World Renewable
Energy Congress. Reading, 2367-2371
3. Zupan, M. (1989), Comparative Measurements of Energy Efficiency on Different Walls,
Including PMMA Foam Covered Wall, Proc. 3rd Workshop on Transparent Insulation
Materials. Freiburg,
4. Zupan, M., J. BoStjan£i£ (1990), Solar Gains by Transparent Thermal Insulation -
Comparative Measurements on Test Cells, Proc. 1st World Renewable Energy Congress.
Reading, 2387-2391
-------
2977
PERFORMANCE CRITERIA FOR TRANSPARENT INSULATION MATERIALS
IN BUILDINGS
G. Brou wer
Van Heugten Consulting Engineers
P.O. Box 305
6500 AH Nijmegen, the Netherlands
ABSTRACT
In Task X of the IE A Solar Heating and Cooling Programme, titled Solar Materials Research and
Development, the focal point of concern is to investigate and to improve material efficiency with
respect to energy performance and durability of solar buildings and solar systems. The scope of the
cooperative Subtask A research described here is the establishment of performance criteria for new
materials; a part of this research covers "transparent insulation materials in buildings, especially for
dwellings". The research provides firstly a means of material selection for various solar applications
and locations and secondly a methodology to estimate the energy efficiency. Tools for material
selection were given using climate data, operating and boundary conditions, simulation programs and
related material properties. A database of material properties directed to the application fields solar
energy was performed as a support to designers and manufacturers for selecting materials. Besides ,a
methodology is presented, developed in cooperation with the Fraunhofer Institute FRG and the TNO
institute for Applied Physics NL on the energy benefits of transparent insulation.
The results, the ways and means to improve system performance and material durability respond to
requests to design and to operate efficiently with respect to environmental quality and material
conservation in the future.
KEYWORDS
Material selection; transparent insulation; passive solar heating; simulation; material properties; energy
benefits; thermal performance.
INTRODUCTION
In the research of IEA Task X the following objectives were stated :
- Investigate how new materials or assemblies of materials can improve system efficiency and expand
the application of solar systems to a wider variety of needs.
- Estimate and evaluate the energy benefits of using new materials.
- Determine the necessary and quartistnecriteria for the properties of advanced materials which
possess greater system and component performance.
Taking into account, this,the study on performance criteria for new solar materials,provides a means
of selection of wall and window glazing materials for various locations and solar applications and a
methodology to estimate the energy benefits of, in particular transparent insulation in Passive Solar
Dwellings.
-------
2978
LOCATIONS AND CLIMATE FACTORS
The criteria to classify locations or climate types of research on solar energy applications have
concentrated on the ambient temperature and the solar radiation. They primarily affect the solar
material performance. Because of the considered application category in this part of IEA research, viz
: Passive Solar Heating and in relation to the participating countries in IEA research, the following
climate regions from Trewartha's classification (1961) were considered : Subtropical, Temperate
Continental (= Temperate Semi-Arid), Temperate Oceanic and Boreal.
Climate factors affect on the one hand heat gain from solar devices and on the other hand they
influence degradation and, consequently, decrease the heat gain. The use of solar radiation in an
optimal way for typical applications is the principal feature of such a system. On behalf of
performance and degradation aspects the most important climatic factors are : solar radiation,
temperature, air pollution, Ultra Violet radiation and precipitation (rain, snow, hail).
SOLAR ENERGY APPLICATIONS AND CONFIGURATIONS
The realisation of passive solar buildings exists within a context of requirements that increase in
scope with the application (e.g. building) for which the system is designed, to the site, to the region
and its climate, up to the scale of national energy goals. Climate conditions may exert a strong
influence on thermal performance, designs, components and configurations, but also on material
selection. Energy conservation links the environment (climate) directly with the energy demand,
which is in turn related to the application. Two main aspects are of importance in relation to
application and solar system: the amount and the simultaneous occurrence of energy demand and
ambient energy load over a year. The thermal performance and the economic feasibility determine
whether realisation in a specific application field is justified or not. E.g. buildings with a low heat
demand due to high internal heat loads require more solar shading as solar utilisation. The annual
local climate conditions (solar radiation), the specific annual heat or cooling demand,and the
governmental rules and permissions (health, safety, fire resistance) restrict a specific solar energy
application and are fundamental first evaluation criteria before selecting devices and materials in a
more technical way.
The selection on technical grounds takes place by considering the operating conditions, the boundary
conditions with respect to the engineering properties and the thermal performance during life time.
They should be estimated from a parameter study and a cost/gain optimisation process.
MATERIALS AND SELECTION
An important attribute of a solar energy device is the energy benefit for the consumer.
It reflects firstly the optimisation of the design of a solar energy device and secondly the application
of the most benefiting material. The search for more promising and improved materials and the
increasing level of required performance and quality strengthen the need of accurate knowledge of the
applicable materials. Questions arise about the methodology and the accuracy of measurements for
characterizing and evaluating them. As such improvement of the energy benefits of materials concerns
clear specifications of requirements for the many different application environments. To organize the
materials and its specifications a classification scheme is needed. It is briefly shown in Fig.l.
The Database Solar Materials itself deals with the following : physical background, incl. the
composition and kind of material; optical and thermal properties; mechanical properties and durability.
The selection of glazing materials strongly emphasizes the properties of solar transmittance and heat
loss coefficient. For some transparent insulation materials,these properties were presented in the
graph of Fig. 2. Very promising devices which have a high solar transmittance as well as a low U-
value are evacuated aerogel, honeycomb (Arel) and capillary (Okapane) constructions.
-------
2979
Collector,
window
and wall
glazing
f ">
glasses

S N
polymers
f N
coatings
S N
composite
glazing
systems
1
l l l
fioa!
low iron
and
reflective
energy
absorbing
thcrmo
plastics
Ihermo sets
low
emittancc
glazing
optical
switchable
films
solar control
films
double glass
panes
transparent
insulation
Fie. 1. Classification of materials
Translucent and Transparent Insulation Materials
(A
<5
"O
Are!
(50-100 mm)
>
Ultramid
HIT
\ 16(ev.)
b
Glass
*• Glassfclt (21-40 mm)
©kapanePMMA (12-40 mm)
ex (10-30 mm)
ux PC (10-60-100 mm)
Iwall
- j^..-.^y<..Isocryl (16-80 mm)
1 2
U- Value, W/m2.K
Fig. 2 .Solar Transmittance (diffuse) and U- Value
Notes: - aerogel in mm thickness
- aerogel, Glassfclt, Glassbulbs,
Isopalux, Isonex, P-film sheet:
between glass or polymer.
- ev. = evacuated
- Arel
-------
2980
PERFORMANCE OF TRANSPARENT INSULATION ON SOLAR-RADIATED
WALLS IN DWELLINGS
The aim of passive solar design is to minimize the amount of auxiliary heating without impairing
comfort or convenience. The solar savings is the extent to which the solar design has reduced the
auxiliary heat requirement of the solar building relative to a building without this solar device. The
IEA study describes the effect of these materials on the energy benefits when used in dwellings. On
behalf of a rough first estimate of the energy benefits of glazing for heating purposes, without taking
into account the building effects, the basic formula can be used:
T| solar = g - U x Ti-Tamh = (solar efficiency factor) (1)
S
g = fraction of incident solar radiation (at the average angle of incidence) that is directly transmitted
by the glazing and indirectly (absorbed within the glazing) conducted into the building.
U = thermal transmittance coefficient (W/m2K)
Ti = average indoor temperature (°C)
Tamb = average ambient temperature (°C)
S = average solar radiation (incident) (W/m2)
Considering the 5 winter months (November to March)jthe following approach was derived for
locations with Temperate Continental climate.
Ti solar = g - U x 0,25. (2)
This approach of how the combinations of solar transmittance and U-value of transparent insulation
materials in different locations act is also presented in Fig. 3. on the scale on the right hand side (only
for comparison purpose). However, in thermal performance calculations also the absorbed energy
within the glazing and conducted into the building has to be included. If in this graph only the solar
transmittance is usedjhe net solar utilisation is then too pessimistic. The consideration of the building
effects was implied in a case study carried out by some participants and coordinated bij W. Platzer
(1990). The new simulation code of Fraunhofer Institute "SIMHAUS" was used. Effects of solar
energy capture against energy conservation were considered with respect to heating and cooling.
The optimisation of transparent materials in a particular application requires the availability of a very
detailed simulation code for phenomena as:
a) the thermal behaviour of the material and its device, viz.: the angle modifier, the spectral
energy distribution of the reflectance, absorptance and transmittance, the amount of air
ventilation and shading in the air cavity behind the material.
b) the energy demand, viz.: the judgement of thermal comfort criteria with respect to air and
radiant temperatures, overheating, temperature control.
c) the ambient climate on a hourly basis viz.: temperature, direct and diffuse solar radiation (with
spectral distribution).
The used elaborate model SIMHAUS was refined for the aim of this IEA research. From the
results of a number of situations with the program SIMHAUS Platzer (1990) derived regression
equations. They predict annual heating and cooling loads based on envelope (including TIM) and
interior loads for dwellings, viz:
The gain utilisation amounts to 1-e (-i,3l/(SLR+0,177i) (3)
while SLR = solar-load ratio.
By combination of the regression equations and the solar efficiency factor, the energy benefits of
using solar walls in dwellings can be easily calculated. Then the g- and U-values have to be replaced
by equivalent values for the total TIM-Wall, while figure 3 can be used.
-------
2981
Windows and Transparent Insulation Materials
0.60
-0.20
losses
lb =\0
Solar gain
factor
South orientation
Location :
Ti-Tamb
S
averaged
5 months
fni?K/W)
Madison (USA)
Edmonton (Ca)
Denver (USA)
Toronto (Ca)
Trier (Ge)
Rapperswil (Swi)
De Dilt (Nl)
Copenhagen (De)
Messina (It)
0,19
0,20
0,09
0,18
0,27
0,26
0,28
0,32
0,08 (4 months)
In formula: Solar Gain Factor = g - U x Ti - Tamb
S
Fig. 3. Example of potential energy saving of windows and transparent insulation
for solar heating as a function of net solar transmittance , U-value and climate,
overheating is not taken into account (only for comparison purposes).
-------
2982
CONCLUSIONS AND RECOMMENDATIONS
The selection of materials for solar energy applications and new energy conservation techniques
complies with the obtainment of higher efficiencies. New materials and material assemblies are
required which possess a higher averaged energy performance during life time. Two main aspects for
a material as such are distinguished:
a. thermal properties of materials in operating conditions, according to the maximum energy
benefits that it brings to the application
b. degradation of materials during life time, which in turn do deteriorate the annual energy
performance of the assembly in its application.
Material selection
A means of material selection for various solar applications throughout the world is as follows :
Define climate data and criteria; select application and device; select simulation tool or calculation
method, establish operating conditions; consult database of solar materials; establish energy
performance and overheating; concern degradation effects. A database on material properties is
available with most recent data.
Methodology for material performance estimation on transparent insulation materials in dwellings
The impact of the thermal properties of transparent insulation materials on the thermal performance of
solar radiated, south oriented, walls for heating is as follows : The maximum energy gain will
generally be achieved at the highest amount of g-Ux (Ti-Tamb)/S, where (Ti-Tamb)/S depends on
the geographical location of the facility. See Fig. 3.
A correlation equation has been developed : "gain utilisation versus solar load ratio" with an excellent
correspondence with calculated monthly gains. For the purpose of IEA,Task X,this utilisation
function is satisfactory as a general approach of the impact of TIM on the heating demand in
dwellings. In addition(a guideline to estimate the energy benefits of transparent insulation the energy
saving for heating or the excess indoor temperature specific for dwellings was derived in the IEA
research using this correlation equation and the solar efficiency factor of figure 3.
With equivalent design criteria,the energy benefits per m2 south-oriented solar wall area in dwellings
amount for West-European localities to 30-40% of location Denver. Solar walls in Madison benefit
per m2 about 50% of Denver.
Tentative predicates for applying materials in solar technology with respect to energy and
environment are needed.
This information contributes to the greater efficiency of use of materials with the potential of
improving energy performance and environmental quality.
REFERENCES
Brouwer, G. (1991). Solar materials research and development. Performance criteria for new solar
materials (Draft). IEA Solar Heating and Cooling. Task X. Van Heugten, Nijmegen, NL.
Brouwer, G. (1991). Solar materials research and development. Database Solar Materials (Draft
Working Document). IEA Solar Heating and Cooling. Task X. Van Heugten, Nijmegen, NL.
Dijk, H.A.L. van (1990) Translucente thermische isolatie in pevels: een verkennine (in Dutch).
TNO - Bouw, Delft, NL.
Platzer, W.J. (1990) Case Study Transparent Insulation. IEA Solar Heating and Cooling. Task X.
Fraunhofer-Institute, Freiburg. Ge.
Trewartha, G.T. (1961). The Earth's problem climates. University Wisconsin Press, Madison. US.
-------
2983
HONEYCOMB PASSIVE WATER HEATING SYSTEMS
N.D. Kaushika
Centre for Energy Studies, Indian
Institute of Technology, Delhi
New Delhi - 110 016, INDIA
ABSTRACT
Transparent honeycomb insulated passive water heating systems using ground
(concrete/sand) as solar collector storage media are investigated. It is
pointed out that a ground based water channel placed beneath a honeycomb
insulated slab (thickness 0.3 m) of concrete sand material can continuously
supply hot water (AT = 60 C) at solar conversion efficiency of 401.
KEYWORDS
Honeycomb; Transparent insulation, Thermal Storage; water heating systems.
INTRODUCTION
The use of honeycomb device to reduce heat losses through the cover system
of the flat plate collector is well known (Hollands, 1965). In recent years
transparent honeycomb structures have been considered for application in
solar ponds and other water storage systems. Experimental results of small
water storage cube with transparently insulated walls and top surface
(Rommel etal, 1987) and of cuboid transparently insulated at the surface
(Kaushika and Banerjee, 1983) have been reported. Simulation calculations
on the thermal performance of honeycomb solar ponds (Sharma and Kaushika,
1987) and large storage tanks (Rommel etal, 1987) have shown encouraging
results. However, containment of large storage volumes of water and the
temperature stratification therein is an outstanding problem in such systems.
It is, therefore, of interest to investigate other solar collector/storage
media for above application. The ground (sand/concrete) having equally
large thermal capacity and insulation characteristics as water is investigated
in the scope of present article.
ANALYSIS AND RESULTS
A simple configuration of the proposed system is given in Fig. 1. It
consists of a network of pipes buried in top region of the ground whose
surface is blackened and covered with an air filled honeycomb. In steady
state the temperature distribution in regions I (o
-------
2984
fSUN J
/T\
x = 0
* =1,
- infinite medium
T
L Honeycomb
X.
Blackened surface
0 (t)
Fig. 1 Schematic diagram of honeycomb ground collector/storage
system
The corresponding boundary conditions are
at x = o
= 0* -COext S - Qv.
-K ^T| I = ^ WeiJJ
a 30 | DC - o
or
c53=- l^C-r-O
Uu - Q\_/ QTLX--OJ - Ta)
OCfc. 3C. =. \
-r,
OLrvx.cL~
— k. \ — -Wr c>"T";
*-Ji_^sTTc*.-o-) -. (3^1
"S^c- | X. -
- - (&J
- -tSj
—t£j
-------
2985
also as x —> oo j is finite (7)
For operational conditions corresponding to heat retrieval at constant
temperature :
Tj (x = 1T) = T2 (x = lj) - T0 (8)
solving eqs. (1-7) we get
Q(t) = - kA1 (9)
Tj (x = o) = To - 1^ (10)
where
v^-W.S - Uu(Ta -TA;
A, = -

If the air filled in the honeycomb is assumed to be stabilized to be
nonconvective, the heat loss from the absorber to the surfacecpver (Q^) is
due to conduction and radiation, and the heat loss from the surface cover to
the ambient air (Q ) is due to wind forced convection and radiation. So we
have ' C
f-t > &V khl -
I - fc"c-W-u-A~r/s][t H- Cuu -S-i / < j]
AT - C.T0-T\)
C.N >
Numerical computations have been made for the efficiency of heat collection,
Values of various thermophysical parameters related to lexan honeycomb
are taken from Sharma and Kaushik (1987). Thermal conductivity of ground
is taken to be .519 W/mK. Efficiency ( "n ) is found to decrease with the
increase in AT and for a given collection temperature the efficiency
decreases with increase of depth (1.) of plane of heat retrieval and is
maximum when 1. - o. The variation of corresponding to 1. = o is
illustrated in Fig. 2(a, b).
-------
2986
T0 = SO C
0.7
601
70 C
IEXAN HONEYCOMB
0.3
ao
0.05
0.10
L (m )
06
OA
LEXAN
T0 = 80 t
02
015
010
005
L(m)
( b )
Fig. 2 Variations of solar conversion efficiency with honeycomb depth and
AT (T - TJ.
o A
-------
2987
The slab (o <. x
-------
Intentionally Blank Page
-------
3.15 Convection and Mass
-------
Intentionally Blank Page
-------
2991
HEAT TRANSFER BY CONVECTION AND THERMAL RADIATION IN A SMALL
SCALE TROMBE WALL SIMULATION ENCLOSURE
L.T. James and W.A. Gross
Department of Mechanical Engineering
University of New Mexico
Albuquerque, New Mexico, U.S.A.
ABSTRACT
A simulated Trombe wall with high surface emissivity significantly enhances cross
enclosure heat transfer in a one-third scale model. Data from prior enclosure
experiments do not include direct comparison of Trombe wall heat input for low
emissivity (0.2) and high emissivity (0.9). We report combined convection and
radiant heat transfer experiments and analyses which confirm that radiant heat
transfer is negligible for all emissivities less than 0.2, but is a significant
portion (10%) of cross enclosure heat transfer for emissivities greater than 0.2.
For emissivities greater than 0.7 on the hot wall, molecular gas absorption
becomes significant (10%). Experiments span the range of end wall temperatures
of 5-25C and Rayleigh numbers of 109- 1012. High viscosity gaseous Freon-12 and
Freon-22 in the enclosure scales the Rayleigh number to match full size building
conditions. We recommend that future experimental models and numerical studies
include radiative heat transfer when realistic wall surfaces are planned.
KEY WORDS
Enclosure; Convection; Radiant Heat Transfer; Trombe wall; emissivity.
INTRODUCTION
Radiant heat exchange between walls in buildings and dwellings promotes more
uniform heat distribution. An objective of this research was to determine when
radiant heat exchange becomes significant, because thermal radiation is neglected
in most experimental and numerical analyses of heat transfer across enclosed
spaces. Scientists at Los Alamos National Laboratories (White, 1985, 1986; Wray
and colleagues , 1979; Yamaguchi, 1984) conducted a series of convection dominant
heat transfer experiments to investigate strong natural convection heat flow in
unevenly heated passive solar buildings. An objective of the small scale
experiments was to verify that the results correlated with large scale models and
therefore provide a method of modeling and predicting the performance of passive
solar designs prior to construction. Follow-on experiments (James, 1990) in the
same small scale enclosure have exhibited convection dominated heat transfer at
low value wall emissivity (0.1) and combined convection and thermal radiation at
a high value heating wall emissivity (0.9).
The enclosure is a one-third scale model with a simulated unvented Trombe wall at
one end of a room. These experiments include a two room model with a connecting
doorway but the enclosure is easily converted to other more complex
configurations. Three dimensional grids of thermocouples inside the enclosure and
in the wall insulation provide data for measuring inside temperature profiles and
-------
2992
wall heat loss. We conducted experiments covering the range of end wall
temperature differences of 5-2!& and convective heat transfer in the range of
Rayleigh numbers of 109-1012. These values are pertinent to models of full scale
buildings although a direct comparison to full scale data is not currently
available.
The experimental enclosure is an insulated, airtight rectangular box made of 1mm
stainless steel. White (1986) and Weber (1979) describe the enclosure in detail.
Figure 1 is a cross section of the enclosure showing the insulation and the open
enclosure configuration. A 'Styrofoam' partition inserted in the center provides
the two room enclosure. The inner space of the enclosure where experimental
building sections are constructed has dimensions of height, length, and width of
0.86m, 1.12m, and 2.34m with aspect ratios of H/L - 0.4 and H/W = 0.77. This size
accommodates models of one-third to one-fifth of full scale. The end walls
consist of a heated wall which simulates the inside of a sun facing Trombe wall,
a cooled wall which simulates heat loss, and heavily insulated (at least R-40,
English Units) side walls, roof and floor. The heated wall is moveable for
modeling of aspect ratios, H/L, of 0.5-5.
The heated wall contains an assembly of 10 heater tapes sandwiched between two
aluminum plates and can generate 250U. Individual heater tapes receive power
through a rheostat allowing adjustment of localized heat flux to simulate uniform
heat flux or an isothermal wall. A cooled wall at the other end of the enclosure
has copper tubing for cooling water from a controlled bath (+ or -0.01'C). The
inlet and outlet temperatures and the heater tape voltages are monitored and sent
to the data acquisition system to provide heat input and heatloss data to the
computer.
During an experiment the enclosure is sealed and filled with a high viscosity gas
(Freon-12 or Freon-22) to scale the convection characteristics to the smaller
geometric size. For each of the four tests, we found that it took about twenty-
four hours for the enclosure temperatures to stabilize whenever we changed the
steady heat input and removal rates. This was due to slow changes in heat storage
in the insulation.
DESCRIPTION OF THE SMALL SCALE MODEL

"Styrof
Plywood •
Vermiculite
Fiberglass
Heater Plate
Cooling
Plate
DIMENSIONS = CM
Fig. 1. Longitudinal section of small scale enclosure.
-------
2993
All electrical leads feed through vertical U-shaped traps containing mineral oil.
The power to the heater tapes and the thermocouple leads passes through these
traps, which permit a slight pressure above ambient to be maintained inside the
enclosure. The three dimensional temperature profiles come from the one hundred
forty thermocouples made of teflon coated copper-constantan forty gauge wire.
Thermocouples are positioned throughout the enclosure and embedded in the walls.
Heater tape voltages, thermocouple output, and cooling water temperature changes
and flow rates are continually monitored to assure steady conditions. All sensors
feed directly to analog-to-digital conversion circuits so that each data set is
averaged, processed, tabulated and plotted in minimal time.
The overall heat transfer measurement accuracy is relative to the resolution of
small changes in temperature differences and varies from 5-20% depending on the
end wall temperature difference and the inlet/outlet temperature difference in the
cooling water. Major errors are from measurement precision, calibration accuracy,
and electronic drift.
THE FOUR EXPERIMENTS
Two experiments with only one room demonstrated the natural convection patterns
established from a single heated wall at low and high emissivities. Two
experiments with a partition separating adjoining rooms demonstrated the
convection patterns established through a doorway connecting the heated room and
an unheated room. The single room experiment with low emissivity wall provided
the baseline for comparison with published experimental data. The hot and cold
wall emissivities were 0.02 (for aluminum tape) and the side walls, top and floor
emissivity was 0.1 (for polished stainless steel). A central partition with
doorway height ratio of 75% divided the enclosure to provide a second set of data
at low emissivity. Flat black paint on the hot wall changed the emissivity to
greater than 0.9 for the third and fourth sets of data at high emissivity for the
same configurations as the first and second experiments.
OBSERVATIONS AND ANALYSIS
The experiments with high hot wall emissivity (0.9) exhibit different temperature
profiles and heat flow patterns as well as increased heat transfer efficiency.
Analytical heat balance calculations for the enclosure using experimental values
and several points between high and low emissivity help to determine limits for
emissivity significance.
Emissivity Effects
As expected, enhanced emissivity increases radiant heat transfer between walls.
This also results in increased side wall temperatures and a slight increase in the
bulk gas temperature. The higher side wall temperatures appear to cause more
convection heating and therefore the higher bulk temperature. Two additional
effects include increased side wall conduction losses through the insulation due
to the higher temperature gradient, and an overall more uniform or even
temperature distribution in the enclosure. There is a smaller horizontal
temperature gradient across the enclosure in the gas and a reduction in vertical
temperature gradients.
In the two room experiment, the partition wall temperature increased as expected
due to the increase in radiant heat exchange. For comparable heat input at the
hot wall, a slightly lower hot wall temperature resulted for the case of high
emissivity, and there was less heat loss from the backside of the hot wall. These
results clearly indicate increased hot wall thermal emission.
-------
2994
It is not possible to measure separately convective and radiative heat transfer
in the experiments. A heat balance calculation using experimental values of wall
temperatures and heat input, permits analysis of an enclosure with thermally
radiating walls and conductive heat input and removal. Predictions show that, for
all wall emissivities below 0.2, the cold wall absorbs only 7% by radiation of the
emitted energy of the hot wall and less than 1% of the total heat transfer across
the enclosure. Calculations and the experiments show that, for low emissivity
(less than 0.2) walls, radiant heat transfer may be neglected. For emissivity
values greater than 0.2, up to 0.9 on walls of the enclosure, 10-30% of the cross
enclosure heat transfer is due to radiation. The experimental values of wall
temperatures and hot wall emissivity of 0.9 lead to a prediction that 45% of the
electrical heat input is converted to radiant heat from the hot wall. Radiant
heat transfer should be considered in calculations or experiments with
emissivities which are typical of building inside walls. At higher emissivities
on the hot wall the heat transfer due to thermal radiation increases and the
theoretical gas absorption becomes significant. Approximately 60% of the total
thermal radiation from the hot wall falls in the absorption bands of Freon-22.
We calculate that 7% of the total energy transfer across the enclosure can be
absorbed by the gas and contribute to the convective heat transfer. For wall
emissivities above 0.7 in enclosures with molecular gas absorption, the gas
interaction should be included in the analysis.
Single Room Enclosure
The baseline experiment was a single room (aspect ratios H/L-0.5 and H/W-0.78)
with low emissivity hot and cold walls at the ends of the long length as shown in
Fig. 1. Measured temperature profiles indicate that a thermally stratified core
region formed in the enclosure. Heat transfer was apparently confined to
turbulent boundary layers along the walls, ceiling and floor, excluding the
quiescent core. The temperature profile shows a temperature inversion near the
hot wall which persists over the full range of hot to cold wall temperature
difference (5-20C). Olson and colleagues (1985) and also Bejan and colleagues
(1981) report similar experimental results and Markatos and Pericleous (1984) as
well as Paolucci and Chenoweth (1988) report similar results for numerical
studies.
The single room enclosure with high emissivity hot wall had a significant
difference in the heat transfer from hot to cold wall compared to the low
emissivity experiment. The heat transfer correlation is
Nu - C X Ran (1)
where the Nusselt number, Nu, is based on the enclosure length between hot and
cold wall, C is a proportional constant, the Rayleigh number, Ra, is based on the
height of the hot wall and cold wall where most of the convection heat transfer
occurs, and n is a correlation exponent. The Nusselt number was 30% higher for
the high emissivity experiment for comparable heat input. We obtained lower
values for the correlation exponent than normally expected for turbulent flow
(0.25 or less for laminar flow, 0.33 or greater for turbulent flow). We also note
that numerous results reported in the literature indicate similar values for
gaseous experiments, but higher values of n for water experiments. The higher
value of n — 0.33 is typically associated with vertical flat plate data and not
for enclosures. There must be some enclosure effect which inhibits higher heat
transfer for turbulent flow compared to laminar flow.
-------
2995
Two Room Enclosure with Connecting Doorway
A stagnation zone existed above the doorway in the room including the hot wall and
there was a temprature inversion above the doorway in the room with the cold wall,
as well as a core of thermally stratified gas in each room, as seen in Fig. 2.
Vertical temperature gradients in the cores decrease in slope as the hot to cold
wall temperature difference increased. Heat transfer was not confined to the
boundary layers but existed in specific flow patterns through the rooms and
doorway. Since these experiments didn't include flow visualization, the
temperature profiles can only suggest heat transfer paths. For the high emissivity
hot wall data set, the temperature profiles tend toward vertical linearity and
become horizontal across the room. The data suggest a more uniform temperature
and heat distribution for high emissivity and enhanced radiant heat transfer.
£
o
o
X
0
H
1
r
Fig. 2. Isotherms in two room enclosure.
CONCLUSIONS
The experiments simulated a full scale room with an unvented Trombe wall heat
input. The geometric scale was one-third to one-fifth of full scale and the
temperature differences across the enclosure (5-25C) were the same as actual
buildings. The effects with low. wall emissivity are comparable to previously
reported results. The experimental results, and analytical predictions indicate
that wall emissivity becomes significant for emissivities greater than 0.2. The
experimentally observed effects of enhanced hot wall emissivity (0.9) are improved
heat transfer across the enclosure and more uniform heat distribution inside the
enclosure.
There are three regions where the heat transfer is sufficiently different to
warrant different calculation methods: (1) For emissivities of all walls less than
0.2, radiant heat transfer calculations can be neglected; (2) For emissivities
greater than 0.2 on any wall, radiant heat transfer is likely to be significant;
(3) For emissivities greater than 0.7 radiant emission may be strong enough to
heat the enclosed gas due to molecular absorption. These results support a
recommendation for emissivity considerations in future predictions of building
heat distribution, and indicate a need for data from which qualitative thermal
radiation effects can be estimated for non-ported Trombe wall designs.
-------
2996
The experiments also revealed temperature reversals over an internal doorway,
thermal stratification in rooms and and even stagnation zones as well as quiescent
cores within the models. Overall experimental results parallel reported data well
enough to support use of small scale models to analyze new building designs or
heat distribution schemes.
REFERENCES
Bejan, A., A. Al-Homoud, and J. Imberger (1981). Experimental study of high
Rayleigh number convection in a horizontal cavity with different end
temperatures. J. Fluid Mechanics. 109. 283-299.
James, L. T. (1990). Convection and Thermal Radiation in a Small Scale
Enclosure. Masters Thesis, University of New Mexico, Albuquer'que, NM.
Markatos, N. C., and K. A. Pericleous (1984). Laminar and turbulent
convection in an enclosed cavity. Int. J. Heat & Mass Transfer. 27. 755- 772.
Paolucci, S., and D. R. Chenoweth (1988). Natural convection in shallow
enclosures with differentially heated end walls. Trans. ASME J. Heat
Transfer. 110. 625-634.
Weber, D. D., and W. 0. Wray (1979). LASL similarity studies: Part I Hot
zone/cold zone: A quantitative study of natural heat distribution mechanisms
in passive solar buildings. 4th National Passive Solar Conference, Kansas
City, MO, Oct. 3-5.
Weber, D. D., W. 0. Wray, and R. J. Kearney (1979). LASL similarity studies:
Part II Similitude modeling of interzone heat transfer by natural convection.
4th National Passive Solar Conference, Kansas City, M0, Oct. 3- 5.
Weber, D. D., and R. J. Kearney (1980). Natural convective heat transfer
through an aperature in passive solar heated buildings. 5th National Passive
Solar Conference, Amherst, MA, Oct. 19-26.
White, M. D. , C. B. Winn, G. F. Jones, and J. D. Balcomb (1985). The
influence of geometry on natural convection in buildings. 10th National
Passive Solar Conference, Raleigh, NC, Oct. 15-20.
White, M. D. (1986). Partitioned Enclosure Heat Transfer with Real and
Boussinesa Fluids. Ph. D Dissertation, Colorado State University, Fort
Collins, CO.
Yamaguchi, K. (1984). Experimental study of natural convection heat transfer
through an aperature in passive solar heated buildings. 9th National Passive
Solar Conference, Columbus, OH, Sept. 24-26.
-------
2997
AN EXPERIMENTAL AND NUMERICAL STUDY OF PASSIVE SOLAR
VENTILATION IN BUILDINGS
G. S. Barozzi1, M. S. Imbabi2, E. Nobile1, and A. C. M. Sousa3
(1) Istihito di Fisica Tecnica, Universita' degli Studi di Trieste, Italy; (2) Department of Engineering, King's College,
University of Aberdeen, United Kingdom; (3) Department of Mechanical Engineering University of New Brunswick,
Fredericton, N.B., Canada.
ABSTRACT
A new bioclimatic building, featuring an integrated solar chimney roof, provided the ideal
prototype for design and construction of a l/12th scale model building and its micro-climate.
Laborarory tests and Computational Fluid Dynamics (CFD) simulations of the model allowed
the building's performance to be assessed, and together added valuable physical insights into
the performance of the proposed passive solar ventilation system. Experimental results also
provided the necessary feedback to develop the numerical model, offering a worthwhile
alternative to expensive, time consuming full-scale tests.
KEYWORDS
Building, CFD, Passive Solar, Scale Model, Ventilation
INTRODUCTION
Solar chimneys are free convection devices, in which air flow is predominantly driven by the
buoyancy of hot air. A prototype bioclimatic building was recently constructed in Nigeria
(Costa and co-workers, 1986), as part of a field experiment in passive solar architecture. The
roof was nominally designed to allow hot air to rise and exit through the apex, drawing cool
outdoor air through the interior of the building in a continuous process. It was initially
estimated that this arrangement could alone displace sufficient quantities of air to effectively
cool the occupants. Rigorous analysis was not carried out a priori because of the complexity of
the problem. To evaluate the performance of the solar chimney, a l/12th scale model of the
Nigerian prototype was built and tested under simulated climatic conditions (Imbabi, 1990a).
In parallel with this, a 2D CFD simulation of the scale model was performed, with the objective
of investigating the free convection mechanism.
Scale Model Review
Building energy is governed by the three mechanisms of heat transfer, namely conduction,
radiation and convection. Perfect similitude between a prototype and model, in the case of
convection, requires the dimensionless governing equations of mass and momentum to be
identical in both instances. This implies that similitude cannot be achieved if the same fluid is
used by both the model and the prototype. For this reason, the use of scale building models
-------
2998
has been questioned, and some contend that full-scale models must always be used (Moog,
1981). Such notions offer little comfort to the modeller!
Parczewski and Renzi (1963) suggested that for turbulent forced convection similar velocity
profiles (ie, kinematic similitude) can occur, even in the absence of dynamic similitude. By
extension, the same may be true for buoyancy dominated flows above a critical value of the
Rayleigh number. Attempts to validate this hypothesis for enclosures (Parczewski and Renzi,
1963; Nakahara and co-workers, 1975) have not proved decisive, however. That real buildings
are open to and interact freely with the outside world adds further complication to the
problem!
The task which a physical model is expected to perform is an influential factor. In an earlier
study (Imbabi, 1980), dimensional analysis was used to derive the thermal properties for a
l/7th scale model building and its Heating, Ventilation, and Air Conditioning (HVAC) plant.
The objective was to test selected HVAC control systems, and good correlation with numerical
simulation results was later established (Imbabi, 1990b). This demonstrated that partial
similitude, in which geometry and boundary conditions are properly scaled, can yield meaning-
ful results when applied correctly.
Numerical Model Review
Whittle (1986) reviewed Finite Volume techniques for the simulation of air movement and
convective heat transfer in buildings, reporting a number of successful 2D andJD applications.
However, because of the complexity of CFD, and the compelling problems related to the
correct definition of boundary conditions (Wiltshire and Wright, 1990), it was recommended
that CFD codes should not be used as black boxes, and that balanced appraisal of simulation
output requires critical engineering judgement.
The use of CFD in the current study differs from earlier applications in three essential areas.
To start with, predictions were restricted to simulating the response of a laboratory built and
tested model building. This allowed the question of boundary conditions to be addressed with
much greater confidence. Secondly, the building was simulated as part of the outside world, a
reality frequently ignored. Finally, the subject of the investigation is passive solar ventilation,
and not merely HVAC. A full description of the numerical procedure is reported elsewhere
(Nobile and co-workers, 1989), and only selected features will be highlighted where appropri-
ate.
THE PROTOTYPE AND MODEL BUILDINGS
An experimental bio-climatic building, recently completed in Ife, Nigeria (7.5° N, 4.6° E), was
used as the prototype for the tests. It consists of a single room, 5.5 x 3.5 x 3.0 m (L x W x H) in
size, with a corrugated metal roof and fibre-board ceiling. The roof differs from the norm in
possessing an integral solar chimney, to promote ventilation. A l/12th scale replica of the
building was designed on the basis of a detailed parametric study (Imbabi, 1990a), and where
possible, the original materials of construction were employed. This permitted the use of
ordinary building materials, conveniently disposing of the need to scale densities and specific
heat capacities (Imbabi, 1980). The model was built as illustrated in Fig.l.
-------
2999
-•j |"- 0.60
Solar Chinney
0.11
Wood Truss
0.03
Jr-on Roof
0.13
F/B Ceil.
0.25
Cone. Fir
0.11
0.29
0.1L
Solar Chinney
0.13
0.03
Iron Roof
O.U
Panel
0.25
Colurin
Floor-
¦VF*—5*"?—
0.01
ELCVATICN CRDSS-SECTIDN
Fig. 1: The model building details.
THE EXPERIMENTAL PROCEDURE
It was considered prudent to initially model only the steady-state response of the model. Scale
model time (roughly proportional to the length scale) was used to good effect, dramatically
shortening the duration of tests. Consequently, the need to replicate ambient outdoor condi-
tions was banished, and a large laboratory chamber was judged adequate to represent the
model's micro-climate. Climatic factors were limited to the outdoor ambient temperature and
solar radiation. The latter was confined to vertical insolation, to ensure maximum roof
exposure and symmetry. An array of solar lamps was used to deliver a pre-determined radiative
flux onto the roof of the model building. In the absence of wind movement, and with the sun
directly overhead, this represented the climatic extreme, where ventilation was most needed.
Two window (or inlet) positions were exam-
ined. Measurements of temperature and air vel-
ocity were confined to one vertical section, nor-
mal to the length axis of the building, to enable
comparisons between experimental and nu-
merical results. A BURR-BROWN series 20000
PC data-acquisition system was used. Under
steady-state conditions, data streams from 12
miniature PRT's (6 fixed to the underside of the
roof and 6 movable sensors inside the model
building) were automatically averaged, con-
verted and written to an output data file. DAN-
TEC's 54N50 low velocity anemometer was
employed for air velocity measurements at ref-
erence sections, including inlet and outlet. Ex-
perimental air velocities were augmented with
laser enhanced visualisations of the flow in the
model. The test chamber, model building, and
solar simulator are shown in Fig. 2.
Fig. 2: The test chamber.
-------
3000
RESULTS
Space limitations dictate that only one set of results, obtained for one inlet geometry, are
presented. The section studied, and the adopted inlet and outlet boundary conditions are
depicted in Fig.3.
Computed velocity vectors and temperatures
(normalised values) are displayed in Figs.4
(a) and (b) respectively. Enhanced smoke
traces of the flow through the inlet and in the
roof-ceiling cavity are reproduced in Figs.5
(a) and (b). The visualised flow patterns
agree well with numerical predictions in both
instances (note the cavity vortex).
Finally, in Figs.6 (a), (b) and (c), numerical
vs experimental temperature and velocity
profiles are compared at the centre-line,
inlet, and outlet sections respectively. Des-
pite 3D effects and possible localised tur-
bulence occuring in the physical model, the
comparisons are encouraging.
p=t=0v/bx=0
i=exp.
p=t=dv/bx=0
t-exp ^
Fig.3: Schematic of the section studied.
(b)
Max. Vel.: 20.5 cm/i
(a)
Fig.4: Geometry #7 (a) velocity vector plot, and (b) temperature map.
-------
3001
(a) (b)
Fig.5: Geometry #1 flow visualisations (a) at the inlet, and (b) in the roof-ceiling cavity.
• Expennentol
— Nunpriccl
Experinentol
Nunpncot
Cxpermenta
NunerlCQ I
•6000 -4000 -2000 0 2000
N.D. Velocity
0 200 400 600 800 1000
N.D. Velocity
N.D. "Temperature
Fig.6: Numerical vs experimental (a) centre-line temperature profiles, (b) inlet velocity profiles,
and (c) outlet velocity profiles.
CONCLUSIONS
The study clearly demonstrates that a passive solar design can be used to ventilate the interior
of a building. The results also suggest that the performance of the system can be improved, by
further modifying the solar chimney roof and the inlet window geometries. Implicit to the
above statement is that partial similitude can provide fairly accurate, meaningful results. It is
an accepted fact that indiscriminate use of small-scale models can lead to complete similitude
failure - hence the importance of defining from the outset the objectives of a modelling
exercise. The ability to fully quantify the scaling errors associated with partial similitude is a
task for the future.
-------
3002
As a benchmark for advanced numerical simulation codes, the use of scale models, as opposed
to full-scale buildings, has the advantages of speed, economy, and flexibility. This has been
particularity true in the present study, where the building form and the ventilation system are
both quite unique.
ACKNOWLEDGEMENT
This research was carried out under the auspices of the International Centre for Theoretical
Physics (ICTP) ItalianLabs programme, which sponsored one of the authors (M. S. Imbabi) while
on sabbatical leave at the University of Trieste. Additional funding was also provided byaMURST
(Italy) 40% grant.
REFERENCES
Costa, R., P. Piva, O. Barduzzi, I. Golubovic', and M. Diaz (1986). Architettura bioclimatica nell' area caldo-umida
dci territori Yoruba (Nigeria), In Energia e Ambiente Costruito, Tradizioni*. <• Tnnnva7innp Udine.
Imbabi, M. S. (1980). Control systems in buildings. M.Sc. Dissertation, Brunei University (U.K.).
Imbabi, M. S. (1990a). A general procedure for the small-scale modelling of buildings, Int. Jnl. of Energy Research.
14, 311-321.
Imbabi, M. S. (1990b). Computer validation of scale model tests for building simulation, Int. Jnl. of Energy
Research. 14.727-736.
Moog, W. (1981). Room flow tests in a reduced scale, ASHRAE Transactions. 87.1162-1181.
Parczewski, K. I., and P. N. Renzi (1963). Scale model studies of temperature distributions in internally heated
enclosures, ASHRAE Journal. 5.60-68.
Nakahara, N., T. Goto, Y. Miyagawa, M. Kobayashi, T. Yasuda, and S. Ito (1975). Verification of similarity theory
on space air distribution and study on occupied zone air-conditioning for large spaces through scaled model
experiments and actual measurements, ASHRAE Int. Best Paper (New York*).
Whittle, G. E. (1986). Computation of air movement and convective heat transfer within buildings, Int. Jnl. of
Amhip.nf Energy. 7. 151- 164.
Wiltshire, T. J., and A. J. Wright (1990). Advances in building energy simulation in the U.K. - the Science and
Engineering Councils programme, Energy and Buildings. 10.175-183.
Nobile, E., T. Russo, and G. S. Barozzi (1989). An efficient parallel algorithm for the numerical solution of
Navier-Stokes equations using FORTRAN structured multiprogramming, In C. A. Brebbia and A. Peters (Eds.),
Applications of Supercomputers in Enpinp.p.rinp: Fluid Flow and Stress Analysts Applications. Southampton,
U.K., 3-14.
-------
3003
THE BLACK GLOBE THERMOMETER FOR INDOOR/OUTDOOR
MEAN RADIANT TEMPERATURE MEASUREMENT
T. Lewis Thompson, George V. Mignon, and Nader V. Chalfoun
The Environmental Research Laboratory
The University of Arizona
Tucson, Arizona
ABSTRACT
This paper discusses the performance of globe thermometers of
varying diameters over a range of air velocities, including the
free convection regime, and points out precautions which should be
taken in mounting the instruments for indoor or outdoor
measurements. When used in outside applications, the air velocity
may vary substantially relative to indoor conditions, and the
accuracy of the instrument at the velocity extremes becomes
important. Under outdoor conditions, the sensors used should be
accurate to at least *0.5 f. Small diameter globes, say two inches,
can be used for low velocity (indoor) applications. If the globe
is to be mounted near a wall, a suitable shield between the globe
and wall may be required.
KEYWORDS
Human comfort; comfort instrumentation; thermal analysis; infrared
radiation; heat transfer; free convection; forced convection.
INTRODUCTION
An experimental program is currently in progress at the
Environmental Research Laboratory to extend the human comfort
analysis used for indoor spaces to outdoor settings. This has led
to experiments involving the measurement of the mean radiant
temperature (MRT) in both indoor and outdoor locations with black
globe thermometers.
The simplicity and low cost of the black globe thermometer often
makes it the instrument of choice for determining the mean radiant
temperature (MRT). The globe thermometer is usually a flat black,
hollow sphere six inches (15 cm) in diameter with a temperature
sensor in the center. The sensor has usually been a thermometer,
thermocouple, or thermistor.
MEAN RADIANT TEMPERATURE
The mean radiant temperature is defined (ASHRAE, 1977) as: "The
uniform surface temperature of an imaginary black enclosure with
which man (also assumed a blackbody) exchanges the same heat by
-------
3004
radiation as in the actual environment". While the globe
thermometer is spherical, it is considered, in most cases, a good
approximation to a sedentary person (ASHRAE, 1977). The spherical
shape also allows significant simplifications in the computation
of configuration factors (Feingold and Gupta, 1970); while for the
human body configuration factors must be determined with
numerical, or graphical techniques, or with photography (Fanger,
1970).
The mean radiant temperature is determined by measuring the globe
temperature, air temperature, and air velocity. An energy balance
around the sphere at equilibrium gives:
hc{Tc-Ta) = ea{TlR-T*c) = hr[TUR~Tc) (1)
where Ta, Tq, and T^r are the absolute air, globe, and mean
radiant temperatures, respectively; e is the emittance of the
globe; a is the Stefan-Boltzmann constant; and hr is the radiative
heat transfer coefficient (Humpreys, 1977).
The equation given by ASHRAE (1977) for a six inch black globe is:
^-['/•c + c/k(7c-tJ]"4 (2)
where V is the air velocity, and C is 1.03xl08fpm-1/2 or
2.47xl08 (m/s)-1/2. This expression is only valid for a six inch
globe, and does not apply for very low air velocities, when free
convection prevails.
Equation (1) may be applied to other sphere diameters over a wider
air velocity range by using the heat transfer expressions:
[for forced convection (Humpreys, 1977), 102 < Nrs < 105]
N,,aJ = 0.32Nlf (3)
[for free convection (Cess, 1973), 10^ < Npa < 109]
NNu.H-0A9NlR? (4)
where N>ju is the Nusselt number, N^e is the Reynolds number, and
Npa is the Rayleigh number. At low velocities (Nre <6000) , when
there is some uncertainty in choosing either equation (3) or (4),
the convective heat transfer coefficient can be calculated as
(Churchill, 1977):
A'A.„ = (^Nt,/ + A/'L,)"3 (5)
When equations (1), (3), (4) and (5) are used to compute the MRT
for a standard six inch black globe thermometer, there is good
agreement with the results of equation (2) over the range 10 < V <
300 fpm (0.05 < V < 1.5 m/s). Outside that range, equation (2)
appears to give slightly high results; however, the difference is
less than 1°F (0.6°C) over the range 5 < V < 1000 fpm (0.03 < V <
5 m/s) . These figures are based on a 10 F difference in the mean
radiant and air temperatures, near room temperature.
MRT ERROR ANALYSIS
Using equations (3)-(5), the mean radiant temperature can be
expressed as a function of the air and globe temperatures, and the
air velocity, or:
-------
3005
TUR-f{Ta,Tc.V)
(6)
and the variance of the MRT is given by (Volk, 1958):
o2(T,
fdTK
V dTa
o2(T,
&T i
3T c
a2fT,
dT
3V
"* 1 a2(K)
(7)
Equation (7) can be evaluated numerically by finite difference
techniques.
When the air and globe temperatures are measured with calibrated
thermocouples or thermistors, the error in the measurements are on
the order of *1 f or cr(7"„)-cr(rc)~o.s f. The precision of air velocity
instruments varies widely; however, an error as low as ±20fpm can
be achieved without great difficulty (
-------
3006
A radiation shield, such as that shown in Fig. 2, can be used to
reduce radiation exchange between the globe and the wall. While a
simple sheet of polished metal is often sufficient to shield the
globe from near-by objects, a polished metal cylindrical assembly
is required for the wall. A wall mounted globe would require a
very large flat plate behind it for shielding, because the sphere
would be able to "see" the entire wall.
\
Radiation Shield
Polished Metal Inside —
Support. 'V
Stainless Stsel
or Plastic y
/
Globe Thermometer
2 in. Diameter (Nominal)
•— 6 in. —
T
r
Wall
Fig. 2. Radiation shield for black globe thermometer.
The globe thermometer must extend beyond the sides of the shield
for convective heat exchange with the room air. An air velocity
sensor should be located near the globe.
An example will be helpful to illustrate the steps required to
determine the dimensions of the radiation shield. Consider a globe
thermometer located at a height of one meter in the center of a
room 10m W x 10m L x 2.5m H (33'x33'x8'), as shown in Fig. 3.
Walls
0.055 (all)
^ceiling 0,37
10 m —
F = 0.41
floor
,10 m
T
2.5 m
Fig. 3. Globe thermometer in center of room, one meter from floor.
In this example, the surfaces in the room are considered black to
long-wave radiation, and the MRT measured by the globe will be
determined by the surrounding surface temperatures and angle
factors (Feingold and Gupta, 1970):
TUR=[£FsaT:y/A (8)
where Fsn is the angle factor between the sphere and surface "n",
and Tn is the absolute temperature of the n"1 surface.
-------
3007
The angle factors for a centrally located globe are also given in
Fig. 3; note that the ceiling and floor dominate, and the
influences of the four walls are equal. If the temperature of one
wall was 80 F, and the remaining surfaces were at 70 F, the globe
thermometer would detect a MRT of 70.7 F.
If the globe is mounted near a wall, as illustrated in Fig. 4, the
angle factors are completely changed, and the wall near the globe
now dominates. Assuming the globe is located at a height of one
meter, centered, and six inches from the 80°F wall of the previous
example, the globe will read a value of 74.7 F if the other
surfaces are at 70 F.
F ..= 0.036
wall
Fwai r001?
^ceiling

r.jf.. - ~
;;;;
F„ = 0.24
floor
Wall
F .,= 0.036
wall
Fig. 4. Globe thermometer mounted with its center six inches from
one wall, at a one meter height.
No amount of radiation shielding can restore the relative values
of the angle factors for the various surfaces. However, it is
possible to reduce the influence of the wall on which the globe is
mounted.
2 inch globe,
6 in. from wall
74
U-
Q- 72
MRT
room center
Shield Radius, ft
Fig. 5. Effect of radiation shield dimensions on globe thermometer
readings.
-------
3008
The shield analysis has been performed for a two inch diameter
sphere, as shown in Fig. 2. The interior of the shield is assumed
to have an emittance of 0.1 (bright metal, such as stainless
steel) , and the emittance of the other surfaces is taken as one
(black). As Fig. 5 indicates, a unit with a 5 in. high cylindrical
portion will give the best results for a given radius; however, it
may interfere with air flow past the globe. A globe shielded by a
four inch high cylinder with a one foot radius will read
approximately 1 F higher than the centrally located globe, and may
be the best compromise. Some prototype testing will be required
for a final selection.
CONCLUSION
When globe thermometers are used to determine the mean radiant
temperature under conditions where the air velocity may vary
substantially, the sensors used should be accurate to at least
iO.s /•'. Small diameter globes, say two inches, can be used for low
velocity (indoor) applications. If the globe is to be mounted
near/on a wall, a suitable shield between the globe and wall is
required.
ACKNOWLEDGEMENT
This work was sponsored primarily by the Energy Office, Arizona
Department of Commerce, Phoenix, Arizona, as part of the Arizona
Solar Oasis Project. For reprints of this paper from the
Environmental Research Laboatory, request ERL#91-2.
REFERENCES
ASHRAE Handbook and Product Directory. 1977 Fundamentals.
American Society of Heating, Refrigerating, and Air
Conditioning Engineers, Inc. New York, p. 13.10.
Cess, R. D. (1973). Free-convection boundary-layer heat transfer.
In W. M. Rohsenhow and Hartnett, J. P., Handbook of Heat
Transfer. Section 6. McGraw-Hill Book Company, New York, p.
6-15.
Churchill, S. W. (1977). A comprehensive correlating equation for
laminar, assisting, forced and free convection. A.I.Ch.E J1.
23, 10-16.
Fanger, P. 0. (1970). Thermal Comfort. Danish Technical Press,
Copenhagen.
Feingold, A., and Gupta, K. G. (1970). New analytical approach to
the evaluation of configuration factors in radiation from
spheres and infinitely long cylinders. Journal of Heat
Transfer. Feb.. 69-76.
Humphreys, M. A. (1977). The optimum diameter for a globe
thermometer for indoor use. Annals of Occupational Hygiene. 20.
135-140.
National Semiconductor Corporation (1988). Linear Databook 2.
Santa Clara, CA. p. 6-5.
Volk, W. (1958). Applied Statistics for Engineers. McGraw-Hill
Book Company, New York, pp. 141-145.
-------
3009
MEASURING AIRBORNE HEAT FLOWS IN PASSIVE SOLAR BUILDINGS
WITH TRACER GAS METHODS
F.D. Heidt, R. Rabenstein
Dept. of Physics, University of Siegen, D-5900 Siegen, Germany
ABSTRACT
Thermal analysis of passive solar buildings has to consider not only solar intensities and heat flows
due to conduction and radiation but also external and internal air exchange. This will be demonstrated
for a two-zone model of a building with attached green house. It will be demonstrated how airborne
heat flows can be succesfully measured using tracer gas methods known from indoor air quality
analysis.
KEYWORDS
Passive solar buildings; heat flow measurements; internal air exchange; tracer gas methods.
PROBLEM STATEMENT
Thermal analysis of solar buildings is based on calculations of solar and internal heat gains as well
as of heat flows due to heat conduction and radiation exchange. The role of convection is usually
restricted to the heat transport between solid surfaces and air. The air exchange of a building with its
external environment is characterized by a global air exchange rate, whose value is mostly guessed
from experience only. Internal air exchange between different zones of a naturally ventilated solar
building is usually not taken into account at all, or it is considered to be large enough to assume a
homogeneous air temperature in the whole building.
This procedure is not satisfactory, since the external air exchange, which has a major influence on
the heat balance, depends on the meteorological conditions, the air tightness of the building envelope,
and the occupants'behavior. Also, the internal air exchange between the zones of the building is of
considerable importance for thermal comfort and the possibility of the utilization of solar gains.
However, a calculation of both kinds of air exchange from first principles is generally difficult and
belongs to the area of "Computational Fluid Dynamics" (CFD). Simpler zonal calculation methods
require often unknown boundary conditions and parameters. Therefore, a detailed investigation of the
air exchange and the corresponding advective heat transport in passive solar buildings requires
measurements.
-------
3010
TRACER GAS METHODS
Principle of Operation
The application of tracer substances is a general concept to investigate transport processes. In the
field of air infiltration and ventilation research tracer gasses,are used toflottte flows of air, moisture
or contaminants inside buildings or between indoors and outside. Depending on the objective of
investigation, buildings are treated either as single zones or as multi-cell systems. Tracer gas methods
are based on the injection and mixing of one or more suitable gasses into the different zones of a
building and on the measurement of the time history of the gas concentrations in these zones. It is
possible to calculate the concentrations from the mass balance in the zones, if the air flows between
the zones are known. In reverse, it is also possible junder certain circumstances, to calculate the air
flows from concentration measurements. There exist several methods, which have been reviewed and
compared in the literature (Charlesworth, 1988; Heidt, 1990; Sherman, 1988).
Measuring System
The above-mentioned types of measurements require special facilities for gas-handling, gas-analysis,
data acquisition and evaluation. At the University of Siegen a mobile measuring system MULTI-CAT
has been developed which performs such measurements. MULTI-CAT stands for "multichannel
concentration analysis of tracers" and designates the whole measuring system which consists of the
instrumentation as well as of software developments for measurement control and data evaluation.
Figure 1 shows the block diagram of MULTI-CAT subsystems with corresponding electrical and
pneumatic connections. Up to eight building zones are interconnected with the injection/sampling unit
which' operates under the control of a personal computer. The injection unit is supplied with tracer
gas (N20 or SF6). Samples of the air in each zone are fed into the IR gas-analyzer for concentration
measurements. Concentration data are processed by the personal computer and measurement reports
are generated automatically. MULTI-CAT provides a software-driven man-machine-interface which
guides the user through the whole measurement procedure. A more detailed description is given in
(Rabenstein, 1990).
IR-gas-analyser
tracer gas supply
control unit
personal
computer
building
zones
injection and
sampling unit
Fig. 1. Schematic diagram of MULTI-CAT measuring subsystems
pneumatic, electrical connection
ILLUSTRATIVE EXAMPLE
Two-Zone Model
A two-zone model is the simplest version of the more general case of buildings with multiple-cell
structure. On the other hand, this simple type of model seems to be fairly appropriate for many
practical situations, such as greenhouse and main body of a building, staircase and ap; artments,
basement and first floor, and others. Figure 2 shows the elevation of a house with attached sunspace
and the nomenclature for the air flow scheme of a corresponding two-zone model.
-------
3011

\/'Z//'/'Z//

/////'///////.
Z//'/'Z////'Z////'X////.

zone 0

F10 «.
1
- F20

zone 1
zone 2
Fm

f02

* ^12™

south
north
Fig. 2. Typical example and air flow scheme for a two zone building model
Our physical analysis for this model is based on the assumption of well mixed zones with volumes
V, and V2 and with steady-state temperatures T; in each zone i (i-0,1,2) . Mass balances of tracer
gas in both zones yield two differential equations for the time dependent concentrations c, and c2:
c,(0 = /nq(f) + f12c2(t) with: fn
Cj(0 =/21c1(0 +/22c2(f) with: /21
^10+^12

F T
f - 21 1
J12 j/ rp
F T
12 2
il x x _ _
yy 722 "
2 il
^ t2
^20+^21
(1)
Differential equations (1) are linear of first order with constant coefficients. Their solutions are well-
known. They depend on the initial conditions for cI and c2 and can be found in mathematical
textbooks, e.g. (Boas, 1966). Mass balances of the air in both zones result in two additional
algebraic equations (Roulet, 1989)
^10 +
•^20 +
*21 =
TI
F —
*01 T
•"o
T1
F —
02 T
10
Tl
F —
ru T
T2
F —
12 rp
(2)
This analysis is applied to the case of an illustrative example which is depicted in Fig. 3. It com-
prises the plan view, construction dimensions and most of the parameters relevant for thermal balance
calculations. As will be demonstrated below, the as yet unknown interzonal air exchange rates FI2
and F2I are of considerable importance for the heating demand in this type of buildings.
Measurement Procedure
In order to determine the internal air flow rates F12 and F2I for the considered case,we can conduct
tracer gas measurements according to the following pulse-decay technique: Initially a certain amount
of tracer gas is injected into zone 1 and well mixed. Then the concentration values are measured
every time step alternately in both zones. After some period of time a second pulse of tracer gas is
injected into zone 2 and well mixed again. Alternate concentration measurements in both zones are
continued. Under ideal conditions this operating procedure would result in concentration profiles as
they are portrayed in Fig. 4 (left). At first, concentration in zone 1 decreases whereas it increases in
zone 2. After the second pulse,the concentration in zone 2 steps up and decreases in the sequel. The
concentration in zone 1 increases at first and ends in a decay profile. In reality there is no ideal
mixing and flow rates may not be steady. Therefore, measured concentration profiles will show
evident fluctuations. This scattering of concentrations is demonstrated on the right of Fig. 4.
-------
3012
H- 2n -h n
V
zone 2
V,
A =
25
n2
M =
50
n3
v2 =
200
m3
10m
U-values in Wrn2K"'

fenestration
3.0
exterior wall
0.5
interior wall
1.5
ceiling
0.3
floor
0.3
fenestration per-

centage in zone 2
10%
Fig. 3.
Plan view, constructional dimensions and thermal parameters of the studied case
Moreover, the higher concentration level after the injection of the second pulse into zone 2 may not
always be reached as fast as it is plotted in Fig. 4. In this case, time intervals for the evaluations of
concentration profiles must be shifted accordingly.
zone 1
zone 1
zone 2,
zone 2
0.2 0.4 0.6 0.0
time in h
Fig. 4. Concentration plots for pulse-decay method in a two-zone building
left: theoretical scheme, right: measured (markers) and approximated (lines) values
Data Evaluation
Solutions of equations (1) are linear combinations of two exponential functions whose coefficients and
exponents depend on initial concentrations, zone temperatures and flow rates. The employed pulse-
decay technique ensures that these flow rates can be determined under fairly stable numerical
conditions. To this end, for both time intervals measured time series of concentration data are
approximated by superpositions of two exponential functions. The approximation algorithm is based
on a Marquardt-Levenberg routine (Press and others, 1988) and takes account of the fact that only
4 flow rates and 4 initial conditions (two for each time interval) are the determining factors for all
coefficients and exponents. Assuming, for the first run, that all temperatures (T0,T„TJ are equal,
results are as follows (all values in m3/h):
Fl0 = 48.9 ( 44.0), Fn = 284.8 (289.7), F0I = 32.2 ( 29.5)
F20 = 98.3 (103.4), F21 = 301.5 (296.4), = 115.0 (107.2)
-------
3013
The corresponding concentration profiles are given by the lines on the right of Fig. 4. The ap-
proximative mean external air change rate (F0I + F02)/(Vt + VJ for both zones is about 0.6 h"1.
Values in parenthesis are final results corrected for the non-isothermal case with zone temperatures
T0 = 0 °C, T, = 25 °C, T2 = 20 °C.
DISCUSSION OF RESULTS
Estimation of Associated Heat Flows
Steady-state heat balance equations have been established for both zones in order to estimate contri-
butions of internal air exchange to the total heating demand in the illustrative example. Denominating
temperature differences between the zones and outdoors by ATt and AT2, respectively, yields:
zone 1: ^A^ + E(TX-T^) = Al
zone 2: L2AT2 - E(TrT2) = H
E is the heat exchange coefficient between zone 1 and zone 2 which depends on the interzonal air ex-
change rate F = (Fn + F2I)/2.
E = 37.5 — + F' 0.33 — (4)
* miK
Lt, L2 are loss coefficients describing heat flows from zones 1 and 2 to the outside. According to the
numbers in Fig. 3, L, = 100 W/K and L2 ~ 120 W/K are obtained. A (25 m2) is solar aperture, H
heating demand for zone 2, T0 (0 °C) outdoor temperature, AT2 = T2 (20 °C) required indoor
temperature for zone 2 and / (125 Wm 2) is solar intensity behind glass panes for typical winter
conditions in Germany. Resolving equations (3) for AT, and H results in:
Al + EAT,
AT, = J, - T0 =
Li E (5)
Al - L. AT.
H = L.AT. - E
12 L, + E
Inserting E from (4) and the quoted values for L,, L2, A, AT2,1 into equations (5) gives:
T, = 28.3 °C, H = 2092 W (for F = 0 m3/h)
TI = 24.8 °C, H = 1750 W (for F = 300 m3/h)
The calculations for this exemplary case thus show that the required heating rate without interzonal
air exchange is about 20% higher than that with an internal air exchange of 300 m3/h. The interzonal
airborne heatflow from zone 1 into zone 2 amounts to about 480 W or 27% of required heating rate.
Error Analysis
Values of air flow rates FiS depend on approximation results for concentration profiles which are
subjected to some experimental scattering. Practical experience shows that this scattering is always
in the range of some few ppm around mean profiles. This behavior was modelled with Monte-Carlo
simulations of these concentration curves. "Exact" concentration plots (for F10 = F01 = 30 m3/h,
F20 — Fg2 = 120 m'/h, F12 = F2I = 300 m3/h) were corrupted with additional Gaussian random noise
of 2 ppm standard deviation. For 500 "numerical experiments" the resulting concentration profiles
were analyzed and the corresponding air flow rates are presented in a histogram (Fig. 5). It proves
that the original data of flow rates are reproduced as mean values. Analyzing the standard deviations
of the mean interzonal air exchange F = (F12 + F2I)/2 and the mean external air infiltration
Q = F01 + F02 shows that both are well below 10% of the mean value.
-------
3014
120
100
BO
60
40
20

l\ :
¦ /
t
I :
0
200
inter
250 300 350 400
zonal air exchange F In m*3/h
1
100 120 140 160 180 200
external air Infiltration Q In m*3/h
Fig. 5. Histograms of Monte-Carlo simulated air flow rates illustrating accuracy limits of
the measurement procedure
CONCLUSION
Interzonal air exchange has been identified as an important parameter, not only for indoor air quality
but also for the realisation of solar gains in passive solar buildings. Two-zone models are appropriate
for many practical cases. A pulse-decay tracer gas technique can be applied successfully to in-
vestigate such situations. Errors in determining interzonal air exchange and external air infiltration
seem to be lower than 10 %.
ACKNOWLEDGEMENT
The authors gratefully acknowledge the intensive work of Josef Eble, who performed the program-
ming and testing of the approximation algorithm.
REFERENCES
Boas, M. L. (1966). Mathematical Methods in the Physical Sciences. John Wiley & Sons, New York.
Charlesworth, P. S. (1988). Air Exchange Rate and Airtightness Measurement Techniques -
An Applications Guide. Air Infiltration and Ventilation Centre, Coventry, Great Britain.
Heidt, F. D., R. Rabenstein, and G. Schepers (1990). Comparison of Tracer-Gas Methods for
Measuring Air Flows in Two Zone Buildings. Proceedings 11th Air Infiltration and
Ventilation Conference, Belgirate, Italy.
Heidt, F. D., and R. Rabenstein (1990). Die Messung des externen und interzonalen Luftaustauschs.
Proceedings 7. Internationales Sonnenforum, Frankfurt/Main, Germany, 450 - 455.
Press, W. H., B. P. Flannery, S. A. Teukolsky, and W. T. Vetterling (1988). Numerical Recipes.
Cambrigde University Press, Cambridge.
Roulet, C. A., and R. Compagnon (1989). Multizone Tracer Gas Infiltration Measurements -
Interpretation Algorithms for Non-Isothermal Cases. Building and Environment, 24, 221-227.
Rabenstein, R., and F. D. Heidt (1990). The Man-Machine Interface for the Air Exchange
Measurement System MULTI-CAT. Proceedings 11th Air Infiltration and Ventilation
Conference, Belgirate, Italy.
Sherman, M. (1989). Air Infiltration Measurements Techniques. Proceedings 10th Air Infiltration
and Ventilation Conference, Espoo, Finland. 63 - 88.
-------
3015
DEVELOPMENT OF DESIGN STRATEGIES FOR THERMAL MASS
IN PASSIVE SOLAR DIRECT GAIN SYSTEM
Myongho Lee, Eon Ku Rhee, Gooksup Song
Department of Architecture
College of Engineering, Chung Ang University, Seoul, Korea
ABSTRACT
This experimental study has been performed to suggest the optimum arrangement
of thermal mass in Direct Gain Passive Solar System. The data were obtained
from two test models. The one was designed with concentrated thermal mass and
the other with dispersed thermal mass. The average indoor temperature of dis-
persed thermal mass model was about 2°C higher than the concentrated thermal
mass model. It can be suggested that the dispersed thermal mass is more effec-
tive in terms of energy performance.
KEYWORDS
Passive solar system; Direct gain; Thermal mass; Test model; Concentrated ther-
mal mass; Dispersed thermal mass.
INTRODUCTION
The study aims to provide basic design strategies for thermal mass in passive
solar direct gain system by investigating thermal performance of two different
methods in thermal mass arrangement - concentrated thermal mass and dispersed
thermal mass.
Two test models were constructed for the experiment. The one was designed with
the concentrated thermal mass where heat-storing red bricks (160mm THK) were
placed only on the sun-striking floor. The other model was designed with the
dispersed thermal mass where the same number of bricks were placed spread over
the floor and the walls (60mm).
The concentrated thermal mass model represents a room/space designed as direct
gain system where the walls consist of light-weight partitions. The dispersed
thermal mass model, on the other hand, may stand for the system where the walls
are made of high heat capacity materials such as bricks or concrete.
CONSTRUCTION OF TEST MODELS
Two identical test models were constructed for the experiment. The composition
and thermal characteristics of the model are listed in Table 1.
-------
3016
TABLE 1 Construction of Test Models
Component
Composition
Thermal
Thickness
Thermal
Thermal

Conductivity
(mm)
Resistance
Transmittance

(Kcal/mWC)

(m2h°C/Kcal)
(Kcal/m2hcC)

Exterior Air Film

0.035

i- Floor
Polyvnod
0.150
15
0.100

Wall
Polystrene Foam
0.025
150
6.000
0.157
L Roof
Polyvcod
0.150
15
0.100

Interior Air Film

0.139

Glazing
Pair Glass

0.260
3.846
(3nw]+6mm+3mm)

EXPERIMENT
Common red bricks were used as thermal mass for the experiment. Total number
of bricks used in both models were the same - 134 bricks each. The experiment
was conducted during the 10 day period between February 18th and 28th of 1990.
The models were located on the roof floor of Engineering Building, Chunq Anq
University.
Every 10-minute data were collected using c-c thermocouples and actinometers
which were connected to the data logger. The data of the final three days of
experiment were used for analysis. The weather data during the period are shown
in Fig. 1.
mtdoor temp
Fig. 1. Weather condition during the analysis period
RESULTS OF EXPERIMENT
Variation of Indoor Temperature
The indoor temperature difference between the upper and the lower parts of the
concentrated thermal mass model was minimal during the experiment period, because
-------
3017
The indoor air temperature of the concentrated model depends on the convection
air flow from floor surface. On the other hand, in the dispersed thermal mass
model the lower part of the room was about 4 - 5°C higher than the upper part
during the morning hours was about the same in the afternoon. The temperature
difference in the morning occurs, as the lower part is influenced by the warm
floor surface while the upper part is still affected by the cool adjacent wall
surfaces. However, the temperature diffence between the walls and the floor
becomes minimal in the afternoon, so does the indoor air temperature difference
between the upper and the lower parts of the model.
The average indoor temperature of the dispersed mass model at night was about
3 - 4°C higher than that of the concentrated model, since the heat source of
the former is both the floor and walls while heat source of the latter is only
the floor. Overall, the average daily indoor temperature of the dispersed model
was about 2°C higher than the concentrated model.
p
50
30 -
20 -
A.
10 -
-10
0
6
12
18
0
6
12
18
0
6
12
18
outdoor tenp. i— lover part -—upper part
dispersed dispersed
o —>lower part. upper part
concentrated concentrated
Fig. 2. Variation of indoor temperature
TABLE Z Distribution of Indoor Temperature
(unit- *C)
Symbol
Description
Feb. 25th
Feb. 26th
Feb. 27 tli
max
min
avg
max
rain
avg
max
min
avg

outdoor temp.
2.5
-3.1
-0.6
3.3
-5.9
-1.3
6.8
-4.0
1.2

lesser part(concentrated)
.33.6
5.5
17.6
30.0
9.2
17.9
31.4
8.8
19.4

upper part(concentrated)
28.8
6.1
16.3
26.4
9.1
17.2
28.7
9.3
18.6

laJsr part(dispersed)
31.3
5.6
15.7
27.7
7.3
15.7
30.3
7.6
17.5
o
upper part(dispersed)
31.7
5.4
15.5
27.9
6.9
15.4
30.8
7.0
17.1
-------
3018
Temperature Variation in Concentrated Model
The temperature of the front part surface of thermal mass reacted very sensiti-
vely to the incoming solar radiation. It rose the most during the daytime and
its diurnal swing was much larger than that of the rear part surface. The rea-
son is that the front part mass is directly exposed to the insolation (thus can
be named 'direct thermal mass'), while the rear part is influenced by convection
and conduction from the front part mass (thus 'secondary thermal mass').
There was about 3-4 hours of time lag between the direct thermal mass and the
secondary thermal mass in the concentrated model in terms of heat transfer.
50
r
a.
30 -
20 -
10 -
-10
12
18
0
6
12
18
6
0
0
12
18
outdoor temp. front part front part
(surface) (bottom)
— »- rear part — — rear part
(surface) (bottom)
Fig. 3. Temperature variation in concentrated model
TABLE 3 Temperature Distribution in Concentrated Model
(unit- °C)
Symbol
Description
Feb. 25th
Feb. 26th
Feb.27th
max
min
avg
max
min
avg
max
min
avg '

front part(surface)
37.7
5.4
18.5
31.7
8.4
18.3
34.4
8.3
20.0

front part(bottom)
28.0
4.7
14.2
23.9
8.9
16.8
27.4
8.3
17.6

rear part(surface)
20.1
6.5
12.7
18.9
9.7
14.2
20.8
9.2
15.2

rear part(bottom)
17.6
6.0
10.5
16.4
9.3
13.3
18.7
8.9
13.7
Temperature Variation in Dispersed Model
The heat flow characteristics in the dispersed thermal mass model happened to
be more complicated than the concentrated model. The most temperature swing
appeared at the front surface of the floor where the solar radiation directly
fell. The second most temperature swing occurred in the air temperature at the
lower part of the space, because it is affected primarily by the temperature of
-------
3019
the front surface floor. The reason for the air temperature at the lower part
being 3 - 4 X higher than the upper part during the morning hours is that the
latter is influenced by the surrounding walls which are heated not much by the
insolation but mainly by the radiation and convection from the floor surface.
However, the air temperature at the lower and upper parts of the model became
almost identical during the afternoon when the walls became warm and started to
reradiate. During the daytime, the temperature at the rear part of the floor
surface became the lowest, because the area is away from the direct solar influ-
ence. However, the area became warmer than the front part of the floor surface
between evening and sunrise, for the front part loses a large amount of heat
through the south glazing.
In the dispersed thermal mass model, the front part of the floor surface tended
to be overheated during the daytime, while it became the coldest part at riight.
-10 "j i ii i r 11 inTpi i ri |) 11 n| ti i n |i rrnjn n 11111111111 n| 111111 mnpi r
^ 12 18 0 B 12 18 0 6 12 18
- front part
(bottom;
outdoor temp. front part
„ . (surface)
° rcar part wal1
^sur^ (bottom) (back part)
—x— wall
(surface)
Fig. 4. Temperature variation in dispersed model
TABLE 4 Temperature Distribution in Dispersed Model
(unit: "Q)
Symbol
Description
Feb. 25th
Feb. 26th
Feb.27th
max
min
avg
max
min
avg
max
min
avg

front part(surface)
41.3
2.6
18.3
32.8
6.8
16.6
35.8
5.4
19.0

front part( bottom)
38.2
2.7
16.9
31.3
6.6
16.0
32.6
4.6
17.7

rear part(surface)
20.1
6.2
12.5
19.5
10.3
14.9
21.1
8.6
15.6

rear part(bottom)
20.1
4.7
11.3
18.0
8.2
14.0
20.9
8.6
15.3
—X—
wal1(surface)
25.0
7.0
15.3
22.5
11.2
17.6
24.9
10.8
18.3
—A—
wall(back part)
23.5
6.6
13.7
21.1
11.3
16.6
23.9
9.8
17.1
-------
3020
Concentrated Thermal Mass Dispersed Thermal' Mass
Fig. 5. Picture of test models

Plan
©: location of sensor
Section
Fig. 6. Drawing of dispersed model
CONCLUSION
The results of the study are summarized as follows.
1. The system with dispersed thermal mass performed better than the system
with concentrated thermal mass in terms of energy efficiency, for the average
indoor temperature of the former was about 2°C higher than the latter, thus
suggesting that the interior partitions of a direct gain building should be
made of materials with high thermal capacity rather than light weight partitions.
2. The temperature difference between the upper and lower parts of the concen-
trated thermal mass model was minimal, whil-e in the dispersed model the lower
part of the room was 3 - 4°C higher than the upper part during the morning hours
and about the same in the afternoon. Therefore, the direct gain system with dis-
persed thermal mass is more appropriate for '1iving-on-the-floor' lifestyle
popular in Korea.
3. The floor surface temperature reached SS'C in the concentrated thermal mass
model and 41°C in the dispersed model. To avoid overheating, the use of phase
change materials as thermal mass should be considered.
-------
3021
AIR MOVEMENT RESPONSE OF THE BUILDING
ON THE VARIOUS DISTURBANCES
N. KaZic", P. Novak"
* Mech.Eng.Dep., University of Titograd, Titograd, YU
" Mech.Eng.Dep., University of Ljubljana, Ljubljana, YU
ABSTRACT
The movement in building is controlled by external (pressure field around the object), as
well as by internal (geometry, topology etc. inside the building) factors. The experimental
investigation of this phenomenon is very hard task, so the numerical simulation technic
was applied. Air flow field description in a multi room building requires a huge number
of data to be used. In order to reduce it, the global air movement response (AMR) of the
building on steady and dynamic disturbances is introduced. To demonstrate this approach
AMR of a multi room buildings on steady and dynamic wind effects is presented.
KEYWORDS
Multi room building; air movement response; infiltration rate; simulation model; unsteady
flow; dynamic wind effect.
INTRODUCTION
From many points of view, prediction of air movement and infiltration rate in building is
important. Energy performances of the building as well as the fire propagation and the
transport of smoke and other pollutants are in close relation to this phenomenon The
energy performances of the building are strongly dependent on convective energy transfer
mechanism, controlled by air movement in the building. Especially, this is the case in the
passive solar architecture building design.
In order to find out the air movement in a building, the pressure field has to be determined.
The pressure field depend on many factors: external (wind, outside air temperature, to-
pography around the building) and internal (inside air temperature field, geometry and
topology of building, operation of openable windows and doors, air supply by installation
etc).
While the air temperature and many other factors can be treated as static disturbance, the
dynamic nature of the wind flow generates the dynamic effects [Malinowski, 1971; Nakasaki,
1987]. These dynamic effects caused by wind velocity fluctuations produce the pressure
fluctuations on the boundary surfaces of a buildings . While low frequency variations in the
pressure field (0.1 - 1.0 Hz) induce flow due to the air compressibility in the internal space,
higher frequency variations will affect only the turbulent diffusion. These dynamic effects
are strongly dependent upon the object configuration. The "low frequency" case has to be
considered with much more care. For very slow variations (~ 1. Hz), the problem could
-------
3022
be reduced to the steady flow, but for faster changes it has to be treated as unsteady flow
problem.
The air movement response (hereinafter AMR) of building is a very complex phenomenon,
especially in the case of multi-room building. According to definition, it represents the air
flow response of each part of the building on external and internal disturbances. Bearing
in mind the complex structure and interaction between any two parts of the building, the
description of AMR requires a great number of information. In that way, the main difficulty
in a defining AMR for the object with complex structure is how to define AMR to be a
global characteristic of the building. One acceptable solution is to define AMR as the global
air mass exchange between the whole building and outside space. In other words, by this
definition, AMR represents the integral air flow rate through envelope of a building. As the
flow through outside boundary surface is a part of the air movement within the building,
this problem has to be solved firstly. The solution is based on the simulation process of the
air movement in building, enabling the calculation of AMR.
Generally, the mathematical model of fluid flow in building is described by partial dif-
ferential equations with complex boundary conditions. The problem can be significantly
reduced, if instead of the detail picture, the "global" stream picture of the fluid flow is
sufficient.
The global approach to the fluid flow problem considers the problem on the room level, i.e.
it does not "look at" what happens inside the room. The analysis is concentrated on the
flow through the boundary surfaces of each room.
In previous works [Ka2ic and Novak 1989; KasSic 1990], the dynamic model of air movement
in multi-room building has been presented. In order to make the problem solvable, some
simplified assumptions were introduced :
a. The building has the same flow characteristics in the steady and unsteady flow case.
b. The wind fluctuations are of a low frequency, but fast enough compared with heat
transfer phenomena in building.
c. According to global approach, the pressure field is uniform within individual parts of
the building.
d. The air temperature change in opening is negligible .
e. The mixing process in the room is perfect.
f. The mapping function to maps the wind velocity to outside pressure field of each outside
surface is known.
Bearing in mind above mentioned, the mathematical description of the unsteady air flow
building was defined by following expressions:
Integral continuity equation
MATHEMATICAL MODEL
Integral energy equation
-------
3023
Bernouilli's equation
m = CAg\Ap\N sign(Ap)
State equation of the ideal gas
- = R6, u = cv 9, i = cp6
e
where:
p [Pa] - pressure, g[kg/m3] - density, 6 [C] - temperature, u[J/kg], U [J] - internal en-
ergy, i [J/kgj - enthalpy, c„,cp[J/kgK] - heat capacity, w [m/s] - velocity of the air;
A [m2] - surface area, V [m3] - volume , MI, MO[kg/s] - input and output mass flow rate by
installation, ° [W] - total energy flux keeps the space in desired state in steady state regime;
m [kg/s] - mass flow rate through opening, C,N - flow coefficients for the opening accord-
ing to Bernoulli's expression, z [m] - vertical coordinate, t [s]-time.
The boundary conditions are defined by mapping function which by definition map wind
velocity to pressure field around the building [Kusuda, 1976].
The whole problem is now reduced to the simultaneous solution of the above equations.
After obtaining the trajectory of the pressure and air temperature fields, the mass flow rate
for each opening is calculated by Bernoulli's equation . Finally, the global AMR could be
obtained by integration of mass flow rate over all outside boundary surfaces, .
AMR OF THE MULTI ROOM BUILDING EXAMPLES
The wind and outside air temperature affect simultaneously the flow regime in build-
ings.On the other side, both these influences are controlled by outside natural phenomena ,
and are out of our control. Using the simulation model, the AMR of a building as a function
of these influences could be found.
a.Steady Wind Flow
In a few numerical experiments, the multi-room building with 4 floors and 72-foom (Fig.
l) was exposed to steady wind flow. The wind velocity and direction as well as outside air
temperature, were changed during this calculation.
In order to obtain the net air mass flow rate exchange between the object and the outside
space, the integration process on the envelope of the object was proceeded. The integral
mass flow rate obtained in this way is, by definition, the global AMR of the examined
building. The result, shown on Fig. 2 is function of the wind direction, wind velocity and
air temperature difference.
The superposition of the wind and buoyancy forces (caused by air temperature difference)
are clearly expressed. In the beginning, when the wind velocity is low, the infiltration
mechanism is mostly controlled by buoyancy forces. When the wind velocity increases,
starting from zero, the influence of these forces diminish; after some critical value, the wind
velocity becomes the main control factor in infiltration process. This critical point coincides
with the separation point of two curves corresponding to different wind direction (full and
dashed line).
Besides, it is obvious from Fig. 2, that the decreasing infiltration rate coincides with
increasing of wind velocity. It is logical and expected results.
-------
3024
m
IP Eg
S rfis5S
f"

ilfciiiaui*
asiitflK!]
X
DIRW=0
Fig. 1 The 72 room multi-story building exposed to steady wind flow
ma [kg/8!" Global air infiltration rate:
enviroment - building
DIRW=315
0.8"
6{ [C] - inside air temperature
»o [C] - outside air temparature
W [m/s]
DIRW=0
0.4 • -
W [m/s]
0.2"
W [m/s] - wind velocity
10
Fig. 2 The global air infiltration rate of the 72- Poom
building exposed to steady wind flow
-------
30? S
b.Dvnamic Wind Flow
In order to analyze the dynamic wind effect on AMR of the building, the 16-story multi
room building (Fig. 3) was exposed to steady wind flow with impulse disturbance in the
stream (Fig. 4).
N
Wind velocity [m/s]
W=ll m/s
Wind
W=6 m/s
0.5
t [sl - time
Fig. 3 The 16-story building exposed
Fig. 4 The impulse
to disturbed wind flow disturbance in the wind stream
In order to exclude the influence of the buoyancy forces on infiltration rate, the air tem-
perature of the inside and outside space of the building in these numerical experiments was
taken to be the same (20 C) . The mapping function wind to pressure field around the
building was same as in the case of the steady wind flow. Not having better model, this
approximation is accepted as satisfactory solution. Using the same numerical technique as
in the steady flow case, the trajectory of the air state vectors in the building(pressure and
temperature) was obtained. Knowing that, the air mass and energy flow rate through each
opening in the building was calculated in every moment.
In the same way, as in the case of the steady wind flow, the integral effect of the wind on
the object through AMR was obtained . In this case, the integral mass flow rate " in" and
"out" of the building is shown (Fig. 5a).
The shape of the curves clearly shows the dynamic integral response of the building
on the impulse disturbance in the wind stream. It is clear that the accumulation effect
plays important role in that process. The difference between mass flow rate "out" and "in"
(Fig. 5a) results in a "net" integral mass flow curve (Fig. 5b). Bearing in mind the shape of
the wind impulse disturbance, this curve could be analysed in this light: the greater "out"
then "in" mass flow rate (while wind velocity is increasing), and opposite, while velocity is
decreasing, means that the building's surfaces which are located on the leeward side, are
much more sensitive to the influence of the wind flow than other surfaces.
-------
3026

ArhcfrQj!
46
yf-
K
mass flow "out"
Jiliiii
-mass flow
^a.
A
6i = 20 °C
e0 = 20°c
!k
tmcir
b.
V
[7
t fs] - time n
f
/
met [kg/8] - Global air infiltration rate: enviroment - building
Aiheb = mfj? + - "net" global air infiltration rate
Fig. 5 The global air infiltration rate of the 16-story
building caused by impulse disturbance in a wind stream
CONCLUSION
Air movement response of the building on various disturbances represents a very com-
plex phenomenon, which description requires a lot of information. This is especially case
when the object is multi-room building with complex geometry. On the other side, the
definition of the AMR through the global air mass flow exchange between the building and
the outher space, is very often a satisfactory solution. The AMR of building on steady
wind flow and on air temperature difference, shows the global characteristic of the building
related to this disturbances. On the other side, AMR on dynamic wind effect describes the
infiltration characteristic of the building with respect to dynamic disturbances in a wind
stream. Although this approach is promising, more research work has to be done.
ACKNOWLEDGMENT
The autors would like to express their gratitude to JRC/Sys.Eng.Institute/ N.N.Energy
Sector, Ispra (Va)-Italy where this wark partly has been done.
REFERENCES
Ka£ic,N., and Novak, P. (1989). A dynamic model of air movement in a building.
Proceedings 2. World Congress Clima 2000. Sarajevo, Yugoslavia.
Ka2ic, N. (1990). Simulation model od,air movement in the multi room building.
Report EUR 12584 EN. Commission of the European Communities, Ispra, Italy.
Kusuda, T. (1976). NBSLD. The computer program for heating and cooling loads
in buildings. NBS, Washington , USA
Malinowski,H.K.(1971). Wind effect on the air movement inside buildings. Proceedings
3. Int.Conference on wind effect on building structures. Tokio, Japan.
Nakasaki,M.(1987). Influence of turbulent wind on ventilation. Preprint for Indoor-
Air Conference. Berlin.
-------
3.16 Comfort
-------
Intentionally Blank Page
-------
3029
"MRT°" UPDATE: A STUDY ON THE THERMAL EFFECT
OF DRY AND WET PAVING AND LANDSCAPE MATERIALS ON RESTORING
HUMAN THERMAL COMFORT CONDITIONS AT OUTDOOR SPACES.
N. V. Chalfoun, T. L. Thompson, and M. R. Yoklic.
The Environmental Research Laboratory, The University of Arizona,
2601 E. Airport Drive, Tucson, Arizona 85706, U.S.A.
ABSTRACT
A computer program "MRT' predicts the Mean Radiant Temperature for use in evaluating
human thermal comfort conditions in outdoor spaces (Yoklic and colleagues, 1990). The
thermal condition of outdoor space is often affected by man-made environmental modifiers
such as paving materials, landscape materials, quality of surrounding buildings, and
different shading conditions. The predicted Mean Radiant Temperature (MRT) indicates
the thermal stress acting on human skin due to radiation which affects comfort conditions
(Fanger, 1970). Previous applications of the "MRT' to urban design projects centered
around studies on the effect of different paving materials and shading conditions. These
effects where analyzed under different natural environmental forces which the "MRT"
program manipulates in order to predict the surface temperature, radiant energy, and the
Mean Radiant Temperature. In this paper, an update on the "MRT' program is presented
where the effect of wet surfaces on the space thermal condition is taken into consideration.
These surfaces are represented by pools, ponds, wet grass and shrubs, plant material, and
freshly watered concrete pavers. A case study on the combined effect of the surface paving
materials, shading landscape materials, and nearby canal water on the Mean Radiant
Temperature as well as on the radiant energy and temperatures of the surfaces is
presented.
KEYWORDS
Human Thermal Comfort, Microclimate, Outdoor Spaces, Paving Materials, Landscape
Materials, Shading, Mean Radiant Temperature
INTRODUCTION
The "MRT' program is an analytical tool which provides opportunities for improving
overall community energy efficiency, especially in arid regions, through careful planning
and attention to the open space between buildings. The program manipulates the physical
characteristics of a given environmental setting such as: dry-bulb and wet-bulb air
temperatures, atmospheric pressure, radiation, wind speeds, reflectivity, absorptivity and
emissivity of surrounding natural and man-made surfaces. These parameters are analyzed
by the program in order to predict the Mean Radiant Temperature1 acting on . human
skin and affecting thermal comfort sensation. The effect of nearby water bodies such
as ponds and pools involves the humidity and cooling through the latent heat of
vaporization. An application is conducted for the Scottsdale Canal project to study the
effect of the canal water in modifying the predicted mean radiant temperature.
'The Mean Radiant Temperature is the uniform surface temperature of an imaginary black
enclosure with which man (also assumed as a black body) exchanges the same heat by
radiation as in the actual environment (ASHRAE, 1981).
-------
3030
MRT ENERGY BALANCE EQUATION
In the "MRT' version 1.0, the simple energy balance for massive dry surfaces is presented
in the following equation:
a*M = (1)
or:
t a1s+Hctg + Z^rlsky '9ground (2)
h+zh
Where ts and t are the absolute surface and air temperatures respectively; t, is the sky
temperature (Sellers, 1982); qs is the solar radiation; h is the convective neat transfer
coefficient; hr is the radiative heat transfer coefficient calculated from:
h = <4V> „ 4o(-f)3 (3)
M*)
o is Stefan-Boltzman constant; t is the average between ts and tsky; while a and t are
the solar absorptance and long-wave emittance for the surface; qground is the net heat gain
by conduction to the ground.
For dry massive surfaces with low thermal conductivity and where conduction into the
mass of the material can be neglected, qground is assumed to be zero (Fig. 1).
air
wind
surface

Fig. 1. Energy Balance for a Dry Surface Assuming No Conduction
Heat Loss to the Surface
RECENT PROGRAM MODIFICATIONS:
Development of the current "MRT" version 2.0 program are presented in the two items
below. The temperature prediction for wet surfaces in consideration of latent heat of
vaporization. Also, a generalized method for calculation of coefficient of heat transmission
by convection.
-------
3031
Temperature of Wet Surfaces:
Calculation of the temperature of wet surfaces such as ponds and pools must include the
effect of the latent heat of vaporization term involving humidities. The energy balance
around a wet surface can be represented by the following simplified equation:
an alternate form of this expression is:
(4)
Sh
(5)
or:
t=-
Shc
~Shr
("7T>*K
(6)
Where s is the slope of a line on an enthalpy-temperature diagram for air between the
ambient air and saturated air at the surface temperature or:

(7)
Terms in equations (3) through (6) are: A = heat of vaporization of water; Ha and Hs are
the humidities of air for the ambient air and air saturated at ts] ks is the mass transfer
coefficient (h JC); twh is the wet bulb temperature of the ambient air; and Cs is the "humid
heat" of air.
CONCRETE
(ROUGH)
REFLECTIVITY
ABSORPTIVITY
EMIT7ANCE
PAVING MATERIALS
0.4
0.6
0.97
a
GRANITE2
0.45
0.55
0.92
3
BRICKS1
(FIRECLAY)
0.45
0.55
0.92

LANDSCAPE MATERIALS
GRAaS'
(DRY)
0.32
0.68
0.9
SHRUBS1
0.3
0.7
0.88
WATER1
0.06
0.94
0.96
From Kreith, F. and Krcider, J. F. (1978) Principles of Solar Engineering. MacGraw-Hill Book Co., Washington,
London.
O
From Wolf, W.L. and Zissis, GJ. (1978) The Infrared HandBook, Office of Naval Research, Department of the Navy,
Washington, DC.
Fig. 2. Dry Surface Paving Materials and Wet Surfaces Land Scape Materials
Used by MRT for Thermal Comfort Prediction of Outdoor Spaces
-------
3032
Convective Heat Transfer Coefficient:
The previous MRT version 1.0 used ASHRAE pre-determined convective heat transfer
coefficients from materials' surfaces. These coefficients were derived for three specific wind
speeds of 3, 7.5, and 15 mph as shown in Table 1 below:
Table 1, "ASHRAE" Pre-Determined Convective Heat Transfer Coefficients:
Wind
Speeds
[mph]
Surface Outside Film Heat
Transfer Coefficient by Convection
[ BTU / hr.ft2 ,°F ]
Equivalent Surface
Film Resistance
[ hr.ft2 .°F/BTU ]
Still Air
1.63
R=0.6l
3.0
2.64
R=0.33
7.5
4.00
R=0.25
15.0
6.00
R=0.17
A more generalized method which accounts for the any wind speed case is now in use.
The new method first calculates both the natural and the forced convection coefficients and
then computes a combined coefficient them as shown in the following equations:
The forced convective heat transfer coefficient is (Lof, 1982):
V^=°-23FW (8)
Where Vw is the wind speed in miles per hour [mph]. This expression was derived from
swimming pool data, but should hold for large, flat areas near the soil surface.
The natural convection heat transfer coefficient is:
- f9)
V
-------
3033
Shading
Trees -£,u
More recently, the program has been employed to analyze the effect that adjacent water
bodies have on the thermal sensation in an outdoor space. In a master planning project
for the City of Scottsdale, Arizona the
"MRT' program was employed to
represent the potential effect a local
canal and shading trees could have on
pedestrian comfort. Figure 3
illustrates the proposed pedestrian
walkway along the canal. The role of
the canal in improving pedestrian
comfort is depicted in fig. 4 below.
These results show the mean radiant
temperature of the concrete with and
without the effect of canal water as a
modifier. Also illustrated is the
potential positive effect if shading is
provided. In these graphs an
estimated mean radiant temperature
for the paving and water is presented
in both a full sun and a 75% shade
condition. A most significant
improvement is gained by shading. Although the effect of water on modifying the
predicted MRT is marginal, its potential on restoring human comfort is substantial due to
cooling by evaporation in zones of lower humidities such as Scottsdale, Arizona.
MRT-',
2®
Concrete Paver
Canal
Water
Fig. 3. Schematic of a shaded Pedestrian
Walkway Near a Body of Water
JUNE
40 SHADE
MEAN RADIANT TEMPERATURE
I CONCRETE 0COMC/UATER
8 Ml 11 Ml 2 PR
SCOTTSDALE CANAL phoemix,
MASTER PLAN S«-*r. Jane
JUNE MEAN RADIANT TEMPERATURE
fSy- SHADE
¦ concrete ~comc/uater Boater
F ltft)
0 AH 11 Ml Z PM
SCOTTSDALE CANAL phoemix.
MASTER PLAN Sumnnr. Jvrm
Fig. 4. Mean Radiant Temperatures for the Scottsdale Canal Project
ENSUING WORK
For massive objects influencing the thermal environment the surface temperatures must be
estimated to calculate the mean radiant temperature. If there is significant heat conduction
into the material beneath the surface, the peak outer temperature will be reduced. The
effect of this conduction on the MRT is currently under investigation.
The program is also being further developed to include derivation of appropriate Human
shape factors to be used for the study of combining different surface temperatures and solar
radiation to determine an average mean radiant temperature. This requires consideration
of the areas of the surfaces contributing to the thermal condition of the target
environmental setting as well as the human shapes (view factors) receiving it. Radiation
exchange factors (angle factors) for a standing person and common outdoor objects are
being evaluated, including factors for rectangular and circular ponds, walkways, walls, and
buildings.
-------
3034
CONCLUSION
The thermal condition of a given environmental setting is predicted by the use of the MRT
program. These conditions are often affected by man-made modifiers such as paving
materials, landscape materials, quality of surrounding buildings, and different shading
conditions of the space. In predicting the thermal conditions of outdoor spaces, the
proximity to water bodies or the existence of man made fountains, ponds and pools are
important modifiers. Fountains are assumed to be at the wet-bulb temperature, while the
surface temperature of ponds and lakes are calculated based on an energy balance around
them, with mass transfer of water vapor from the surface included. A case study on the
effect of the proximity of a pedestrian walkway to a canal water is considered for the
Scottsdale Master Plan Project in Scottsdale Arizona. It is noticeable that the predicted
MRT in the month of June as affected by the body of water of the canal are about 8 to 13
degrees below those for the dry case. The effect of both the canal water and an assumed
75% shading condition from trees is also considered. Here a MRT reduction of about 49
to 75 degrees is achieved relative to the normal dry environment and unshaded case of the
walkway.
The "MRT' program provided opportunities for analyzing and improving conditions of
outdoor spaces through careful planning and attention to the environmental forces acting
on that space. Using the physical and thermal characteristics of a given environment such
as: dry-bulb and wet-bulb air temperatures, atmospheric pressure, radiation, wind speeds,
long-wave reflectance, absorptance and emittance of surrounding natural and man-made
surfaces, a Mean Radiant Temperature (MRT) is predicted by the program under different
shading coefficients. The predicted (MRT) indicates the thermal stress acting on human
skin due to radiation which affects his comfort levels.
"MRT' provides one analytical tool for use in the evaluation of outdoor spaces.
ACKNOWLEDGEMENTS
The development of this program is part of the Arizona Solar Oasis project. Funding was
provided by the State of Arizona, Energy Office, Department of Commerce Contract No.
889-014.
For reprints of this paper from the Environmental Research Laboratory, request ERL #91-6
REFERENCES
ASHRAE (1981). Handbook of Fundamentals. The American Society of Heating,
Refrigerating, and Air-Conditioning Engineers, Atlanta, GA.
Churchill, S. W. (1977). A Comprehensive Correlating Equation for Laminar, Assisting,
Forced and Free Convection. A.I.CH.E Jl.23. 10-16.
Fanger, P. O. (1970). Thermal Comfort, Danish Technical Press, Copenhagen, p. 174.
Lof, G. O. G. and Lof, L. G. A. "Performance of Solar Swimming Pool Heater-
Transparent Cover Type." Proc.1977 Ann. Meeting. AS/ISES. Orlando, FL.
Sellers, W. D. (1982), Physical Climatology. University of Chicago Press, Chicago, U.S.A.
Yoklic, M., Chalfoun, N., Thompson, L. and Kent, K. (1990). "MRT: An Original
Computer Program for Predicting Thermal Conditions of OutDoor Spaces", Proc. of 1st
World Renewable Energy Congress, Pergamon Press, New York, Vol 4, pp 2244-2249
-------
3035
PREDICTING LOCAL THERMAL SENSATION IN A BUILDING
A. CORDIER, M. GALEOU, F. MONCHOUX, F. THELLIER
Laboratoire Energie Solaire Et Thermique de 1'Habitat
UNXVERSITE PAUL SABATIER - 118, route de Narbonne
310 62 Toulouse Cedex - FRANCE
ABSTRACT
Thermal comfort, one of the aims of house design, is not definable
only by the physical parameters currently used: air temperature and
humidity, for example. A software, including a human body model
with its physiological reactions, has been built up and previously
described. We added a new program, capable of translating the
predicted physiological data, like skin temperatures, into thermal
sensations. This program is tested by comparison with experimental
results. An example of the use of the complete software is given.
KEYWORDS
Thermal comfort, Building, Thermal sensation, Modelling.
INTRODUCTION
Simulation studies are more and more used to improve thermal effi-
ciency of buildings. However, house designers must never forget
that their first aim is to provide comfort to the inhabitants. The
problem is that most simulation tools are only capable of
calculating inside air temperature.
Physical data, like dry or wet bulb temperature, wall temperature,
air velocity, etc., are far from being sufficient to predict
comfort level, because the body of the inhabitant is a complex
thermal machine. Physiological reactions (shivering, sweat
emission, vasomotricity) are enhanced with the environmental shifts
to maintain the central temperature to a constant value and have to
be taken into account. Moreover, thermal comfort is not only the
result of physical data and physiological reactions, but of
psychological (and even sociological) interpretations.
A software predicting physiological data skin temperature for
each part of the body, sweat emission and evaporation, and so on,
has been supplemented with a program dedicated to the translation
of those data into thermal sensation of the occupant slightly
warm, too cool, for example.
COMPUTING PHYSIOLOGICAL DATA
The first step is to compute physiological data of a body in a
house with a variable climate and equipped with standard heating
or cooling system. To carry out such a computation, we modified the
model written by Stolwijk and Hardy (1971) for NASA and implemented
-------
3036
it in TRNSYS, a general software for house simulations from
University of Wisconsin.
Many modifications were necessary in order to implement
the Stolwijk model. Furthermore, we improved it through a more
accurate computation of thermal exchanges between body and envi-
ronment, and we added clothes to the original model of the nude
body (Cordier and others, 1987; Thellier, 1989) .
Unfortunately, the use of the original TRNSYS Types modelling the
components of the room is not possible, because infra-red
exchanges, which are very important for the body heat balance, need
for their calculation the knowledge of the temperature of every
surface of the room (Cordier and others, 1989) . So, we had to write
new models of wall, window, heating floor and a Type computing
infra-red exchanges between up to 30 surfaces (a minimum of 28
surfaces is required to represent a man in a simple room with a
window).
The validity of our model has been tested by comparison with
experimental results, which were found close enough to computed
results, in the usual range of temperature, to consider the model
as a good representation of reality. The modified Stolwijk
model, called MARCL, gives access to the temperatures of each part
of the body (head, trunk, arms, hands, legs, feet), along with the
mean skin and central temperatures, amount of sweat evaporated or
dripping, skin wetness, intensity of the physiological reactions
(used to detect whether the body remains at the thermal neutrality
or not). All these parameters can be used to predict how an
inhabitant feels his environment. This new step, a translation of
MARCL outputs into thermal sensation, has been done, grounded on a
large litterature review (Gal6ou, Grivel and Candas, 1989) .
DETERMINING THERMAL SENSATION
Most of the publications giving correspondence between
physiological data (skin temperature, etc.) and thermal sensation
must be used with the greatest attention, because experiments are
not fully described and local skin temperatures are not measured in
the same way and place. Furthermore, the terms of the scales used
to characterize the thermal sensation of the subject differ from
one investigator to the other, including more or less psychological
content; for example, the thermal neutrality can be noted as
"comfortable" (Bedford's scale, Bedford, 1936) or "neither cold nor
warm" (ASHRAE scale, Nevins and others, 1966) . Even if for
practical reasons we consider that these two scales give similar
results, we realize that, in practice, a warmth vote is not a
physical quantity like a temperature and consequently the
measurement of thermal sensation is far from being straightforward.
Nevertheless, useful data can be found in the literature, provided
their context is correctly defined. Though we limited our purpose
to comfort in houses, many in a medium climate, we had to review a
very large amount of publications to build up reliable tables
giving sensation versus skin temperatures.
Skin temperatures, measured as well as computed, are always local
feet, head, etc., but evoked sensations can be local or global.
Local thermal sensation can be deduced from local temperature
through convenient tables, but to go from local temperature to
global sensation requires the calculation of a mean skin
temperature, taking into account the local temperature, area and
sensitivity of each part of the body. Here also, discrepancies
exist between the different authors.
-------
3037
The synthesis of the local thermal sensations in order to obtain a
global thermal sensation (just like the calculation of mean skin
temperature from local ones) is an interesting, but completely
unknown, subject ; it requires psychological and neurophysiological
knowledge about the integration possibilities of the human brain,
which is not available today.
Mean skin
temperature
(computed)
Local skin temperatures
(measured or computed)
local sensation
Global sensation
Tables
Tables
Calculation
Fig.l. Determination of thermal sensation
Two parameters have a leading influence on the passage from
temperature to sensation activity and clothes. Activity can be
characterized by metabolism, expressed in Watts per square meter of
body surface, or in MET (1 MET = 58 W/m2) . We only consider two
possibilities light activity or heavy work, the limit being a
metabolism of 2 MET (3.2 km/h walking). Clothing insulation unit is
Clo (1 Clo = 0,155 m2K/W) . We separate light summer clothes (< 0.6
Clo) from winter clothes. So, four cases are under study. Each case
has its own table giving thermal sensation from temperatures.
According to the case, significant physiological or physical
parameters are not the same local skin temperatures, for
instance, have not any influence on the thermal sensation when
metabolism is high. Table 1 shows which parameters are important, w
is the skin wetness; Tin* internal temperature ; Tsjc mean skin
temperature ; local skin temperatures are successively ; T^e • head;
Ttr: trunk; Tar: arms; Tha: hands; Tie: legs and Tfe: feet.
X&2LE 1 Significant Physiological parameters
Case

Variables
w
^in
Tsk
^he
Tt r
Tar
Tha
Tie
Tfe
Met<2
all
clothincr
X
X
X
X
X
X
X
X
X
Met>2
all
clothing
X
X
X

In addition to the physiological parameters, the physical data
characterizing thermal environment can play a role on the
determination of thermal sensation The parameters which have to
be taken into account in each case are air temperature and
gradient, radiative and dew-point temperatures, air velocity,
vertical and horizontal radiative asymetries.
AN EXAMPLE
We apply our program to an experiment done by Grivel, Hoeft and
Candas, (1989) for which ambient and physiological parameters, as
well as sensation notations, are at our disposal. We did not use
these experimental results to fill in our tables "sensation versus
temperatures" .
The experiments were carried out in a climatic chamber where
thermal parameters could be accurately controlled and measured. Six
-------
3038
nude men, 3 at a time, were sitting at rest in the chamber for 3
hours.
Following variables were recorded air temperature (Tair), air
velocity (Vair), dew point temperature (Tdp), rectal temperature , 4
local temperatures (chest, upper arms, thigh and calf). Mean skin
temperature (Tskm) was then calculated.
After a neutral pre-period, the subjects were exposed during 160
min. to air and walls temperature variations as shown on Fig.2. The
other variables having the following values Meta = 1,08 MET,
Tair = "walls' ^air = 0.25 m/s, T^p = 5 °C.
On figure 3, the average experimental mean skin temperature (Tskm)
and the correspondant computet, value (TsjcC) are shown, versus time.
The two dashed lines are the limits of the maximum inter-individual
differences.
38.8
36,8
34,8
32,8
30,8
28,8
26,8
24,8
22,8
20,8
18,8
! !
. Tair CC)

.LI

ti
me (n
iin)
~L
1 6 0 1 6 3 2 4 8 64 80 96 112 128144 16C
.2. Ambient temperature
Fig
variation versus time
Tsk (°C)
time (mm)
•16 0 16 32
Fig.3. Mean
128 144 160
skin temperature variation
versus time: experimental (thin line)
and computed results Tskc (thick line)
During the experiment, subjects answered questions about their
thermal sensation, according to the scales given in Table 2.
TABLE 2 Different -Sensation Scalea
it is :
7. much too warm
6. too warm
5. comfortably warm
[H
7.
6.
is :
hot
warm
slightly warm
as a whole, my skin
7. hot
6. warm
5. slightly warm
30
20
10
comfortable
5. neither cold nor warm 4. neither cold nor warm
3. comfortably cool
2. too cool
1. much too cool
SCALE 1
4. slightly cool
3. cool
2. cold
1. very cold
SCALE 2
3. slightly cool
2. cool
1. cold
SCALE 3
-10
-20
-30
MODEL
Subjects beeing nude and at rest, boundaries between the different
possible sensation notations are shown on Table 3.
XAELE 3 Correspondence temperature-sensation £&£ a maa—at
° c
cold
cool
slightl]
cool
neut ral
ilightly
warm
warm
hot
Tsk
30,
0
31
7
33
2
34,
4
34
6
35
3
The
31
5
32,
1
32
7
35,
1
36,
4
36
8
Ttr
29,
0
32,
8
33
4
34
0
35,
5
36
4
Tar
25,
9
28
6
31
1
33,
2
35
0
37
0
Th a
21,
8
28,
5
30
o
34
0
35,
3
37
5
Tie
24,
3
28,
4
30
1
33,
0
34,
0
35
2
Tfe
21,
3
25,
9
28
6
31,
6
33,
8
37
2
-------
3039
In figures 4 and 5, the average sensation declared by subjects with
the first scale of Table 2 (Smi) is compared with the thermal
sensation predicted from measured skin temperature (Smc, Fig.4) and
the computed ones (See, Fig.5).
7
Sml
Smc
6
"20
Sm 1
5
-10
4
Smc
3
-10
2
"-20
t, mm
-30
16 0 16 32 48 64 80 96 112 128 144 160
7
Sml
See
6
-20
Sml
5
-10
4
See
--10
3
2
--20
t, min
-30
16 0 16 32 48 64 80 96 112 128 144 160
Fig.4. Predicted sensation from Fig.5. Predicted sensation from
measured temperatures (Smc) and calculated temperatures (Sco) and
actual sensation 1 (Smi) . actual sensation 1 (Smi) .
When Fig.4 shows semi computed sensation (computed from measured
Iflcn), Fig.5 shows fully computed sensation (computed from values of
Ttlcc which are calculated from ambient parameters) . The agreement
between these curves and actual sensation curve seems to be better
in Fig.5. The reason is that declared sensation is slightly diead of
the measured temperature when computed Tgkc is also slightly ahead of
actual Tskm. More studies are required to say if this favourable
effect is only due to luck! Nevertheless, it obviously appears
that we can now predict, with an acceptable fitness, thermal
sensation of a man in a given environment.
CONCLUSION AND PERSPECTIVES
Many things have to be done to improve our model. Table 3 has been
built up from a wide review of published results, but a better
knowledge of experimental conditions leading to these results is
necessary to be sure of its validity. Furthermore, the comparison
of Fig.4 and 5 are done with actual sensation expressed in the
scale 1 (Table 2). Other comparisons are possible. Fig.6 and 7
shows the comparison with scale 2 and scale 3 (Table 2). Many
others scales exist; the choice between all the actual sensation
scales has to be done.
The example given above is only concerned with mean skin
temperature and thermal sensation. The influence of local
temperatures and physical parameters on the global thermal
sensation is a complex problem, needing the help of psychologists
and sociologists. This interesting extension of our work is too far
from our competences and out of our planning realm.
-------
3040
Sm2

Scc

Sm2 /

Scc

t, min
•30
-0
-10
|--20
30
-16 0 16 32 48 64 80 96 112 128 144 160
Fig.6. Predicted sensation from
calculated temperatures (Scc)
and actual sensation 2 (Sm2)
7
6
5
4 "
3 -
2
Sm3

^ See

Sm3 /

Scc \\

l, min
•30
-0
--10
--20
30
-16 0 16 32 48 64 80 96 112 128 144 160
Fig.7. Predicted sensation from
calculated temperatures (Sco) and
actual sensation 3 (Sm3) .
ACKNOWLEDGMENT
This work has been supported by Plan Contruction, Electricity de
France, Gaz de France and the A.F.M.E. (French Agency for Energy
Management). The authors thank F. Grivel and V. Candas from the
Laboratoire de Physiologie et Psychologie Environnementales of
Strasbourg, France, for their help and the experimental data.
REFERENCES
Bedford T. (1936) . The warmth factor in comfort at work-a
physiological study of heating regulatory ventilation. Msd. Res
Xnd. Health Res . Board. (London) ¦ 76 (report) .
Cordier A., F.Monchoux, G.Serin and F.Thellier (1987). Simulation
of an occupied building for comfort analysis. ISES. 4. 3275-3279.
Cordier A., F.Monchoux, F.Thellier and D.Zely. (1989). Cold surface
influence on thermal comfort. ISES. 2. 1094-1098.
Gal£ou M., V.Candas and F.Grivel (1989). Le confort thermique
aspects physiologiques et psychosensoriels. Etude biblio-
graphique. CNRS/INRS, LPPE Strasbourg, (report).
Grivel F., A.Hoeft and V.Candas (1989) . Body temperatures during
warm and cool periodic transients influence of clothing.
Advances in indus Erno . and safety. 1.
Nevins R.G., F.H.Rohles, W.Springer and A.M. Feyerhern, (1966).
Temperature- humidity chart for thermal comfort of seated
persons. ASHRAE Trans.. 72: 283-291.
Stolwijk J.A.J. and J.D.Hardy (1971). Control of body tempe-rature.
Hand. of physio - reaction to envir arrents. 45-68.
Thellier F. (1989). Mod61isation du comportement thermique de
1' homme et de son habitat. Une approche de l'6tude du confort.
Th6se de l'U.P.S., Energ^tique n° 510, Toulouse.
-------
3041
COPING WITH DISCOMFORTS
Judith H. Heerwagen*, Joel Loveland**, and Richard Diamond***
* Center for Planning and Design, University of Washington, Seattle, Washington
**Department of Architecture, University of Washington, Seattle, Washington
** "Lawrence Berkeley Laboratory, Berkeley, California
ABSTRACT
The research described in this paper investigates how occupants of six office buildings cope with
different kinds of ambient environmental discomforts and problems. Discomforts and problems
are also assessed in terms of occupants' subjective assessments of their bothersomeness and how
much they interfere with work.
KEY WORDS
Discomfort; environmental coping; behavioral coping; psychological coping; success of coping;
bothersomeness of discomforts and problems; interference from discomforts and problems.
INTRODUCTION
Although many researchers have explored issues of comfort and environmental satisfaction, there
is a dearth of information on how people cope with environmental discomforts and problems. The
coping process is an important aspect of the person-environment interface and is likely to have an
impact on occupants' work performance, stress reduction, health, and overall satisfaction with the
environment. Decades of research on stress and coping have focused primarily on high level
stressors in the psycho-social environment (Lazarus and Folkman, 1984; Lazarus, 1966) or on
specific occupational and ambient stressors such as noise, air pollution, and crowding
(Evans, 1982; Zimring, 1982; Campbell, 1983; Cohen, Evans, Stokols, and Krontz, 1986).
Relatively little attention has been paid to the work environment, particularly to the ways in which
people resolve the many discomforts and irritations that may surface as they go about their work.
The research described in this paper is an investigation of how building occupants cope with the
multitude of ambient "hassles" that occur regularly during the workday - such as temperatures
that are too hot or cold, stuffy air, glare from windows, reflections on their computer screen.
Before the research is described more fully, it is helpful to look more carefully at coping processes.
In general, there are two basic coping approaches an individual can use when confronting difficult
or stressful situations and problems: (1) he/she can alter or in some way directly address the
situation or problem; or (2) he/she can manage emotional or cognitive responses to the situation
(Lazarus, 1966; White, 1974). Coping behaviors in the first approach include attempts to change
the situation itself (environmental coping) or attempts to adjust to the situation by changing one's
behavior (behavioral coping or accommodation). The kinds of coping strategies used in a given
situation are a function of individual differences in personality or experience as well as
characteristics of the situation (White, 1974; Coelho, Hamburg, and Adams, 1974; Pearlin and
Schooler, 1978)). Research on stress indicates that people tend to use a number of different
coping approaches rather than just one (Lazarus, 1966; Baum, Singer and Baum, 1983).
One of the problems in environmental coping research is to identify the features of environments
that tend to elicit different coping processes. It is generally considered "more adaptive" and
-------
3042
"healthier" for people to exert control over the environment (that is, to engage in environmental
coping) when the opportunity exists to do so (Cohen, Evans, Stokols, and Krontz, 1986).
Environmental conditions that are uncontrollable are likely to lead to more accommodation and
to emotion-focused coping processes (Lazarus and Cohen, 1977). Furthermore, uncontrollable
environmental demands are frequently associated with negative affect, performance decrements
(especially on complex tasks), negative social behavior, decreased motivation and a sense of
helplessness (Baum, Singer, and Baum, 1982; Cohen, 1980; Seligman, 1975). There is some
indication also that negative effects are not limited to high intensity stressors, but may also exist
for multiple, low-level "mini-stressors" that can't be controlled. There is increasing evidence,
however, that negative effects can be reduced when instrumental control over the environment is
provided (Evans and Cohen, 1987).
A critical aspect of coping, and one which has not received as much attention, is the extent to
which coping actually improves one's situation. The general sense is that coping is adaptive and
beneficial (Lazarus, 1966). However, when coping efforts fail to modify the situation, the person
may experience an even greater sense of helplessness (Seligman, 1975). Although relatively little
research has addressed the outcomes of coping processes, there is some indication that even
apparently successful coping may have negative side effects produced by increased effort to deal
with the situation or constricted attentional processes (Cohen, 1980; Evans and Cohen, 1987).
Negative outcomes include cognitive fatigue, reduced social interactions, lack of attention to social
cues, negative moods, and a general reduction in coping capacity for future environmental
demands.
METHODS
The primary study instrument was a 27-page Workspace Satisfaction Survey that assessed
occupants responses to the thermal, lighting, acoustical, and air quality aspects of the environment
as well as general office features and aesthetics. Occupants were asked to identify how they coped
with discomforts and problems, how successful their coping efforts were in terms of producing a
more satisfactory situation, and how much the problem bothered them and interfered with their
work. A total of 213 occupants in six buildings in the Pacific Northwest filled out the survey. The
surveys were hand delivered and each was identified with a specific workspace to facilitate spatial
analysis. Each workspace was identified by location (perimeter versus interior), degree of
enclosure, type of workspace, orientation (if on the perimeter), and the building floor. The
response rate for the surveys ranged from a high of 100% in one building to 40% in the building
with the lowest rate of return. All buildings in the study were part of the Bonneville Power
Administration's Energy Edge Program.
Discomforts and problems were assessed for the thermal environment, air quality, lighting, noise,
and computer work. For each of the problems encountered, we asked occupants to note how they
tended to respond. The coping measure was a check list of items that included environmental
actions (e.g., close/open a window, close/open blinds), behavioral actions (e.g., adjust clothes,
drink something hot/cold, operate a space heater or fan, close the door), and
psychological/emotional processes (e.g., "I ignore the situation and concentrate harder on my
work", "I just put up with it-there is nothing I can do").
RESULTS
Data presented in Tables 1 through 4 show how occupants responded to the heat or cold
discomfort, sun warmth and brightness, air quality problems, and noise problems. The data
indicate that occupants respond differently to these different environmental problems. One of the
key factors may be the ease with which they can make changes in the environment or their own
behavior that will produce a more satisfactory situation.
As can be seen in Table 1, the vast majority of responses to warm or cold discomfort were to
adjust clothing, drink something hot/cold, or talk to a staff person responsible for dealing with
thermal problems. Those who had simple environmental controls in their workspace (such as
-------
3043
thermostats, heaters, or windows) were able to use these devices. In situations such those involving
noise (especially from phones or conversations), it may be more difficult to change the
environment or one's own behavior in ways that resolve the problem. See Table 2. Many people
work in partitioned spaces which do not allow them to close the door to shut out unwanted noises.
Thus, the best way to cope may be to engage in psychological processes, such as trying to ignore
the situation or just putting up with it. Although people could ask their coworkers to be more
quiet, this action may have a high social cost.
Table 1. Coping with Warm and Cold Discomfort
COLD DISCOMFORT
Changes In the Environment Percent Using
Use a space heater 27
Adjust thermostat 17
Close drapes 7
Close a door or window 5
Changes In Behavior
Adjust clothing 77
Drink something hot 48
Contact staff person 34
Talk to co-workers about the problem 26
Walk around to warm up 17
Move to a more comfortable space 12
Emotional/Psychological Processes
Just put up with it; there's nothing I can do 23
Try to ignore the problem
and concentrate harder on work 19
WARM DISCOMFORT
Changes In the Environment
Close the drapes 32
Open a door or window 16
Adjust thermostat 13
Changes In Behavior
Adjust clothing 59
Drink something cold 43
Contact staff person 32
Talk to coworkers about the problem 24
Go outdoors for a while 2 0
Move to a more comfortable space 11
Emotional/Psychological Processes
Just put up with it;there's nothing I can do 29
Try to Ignore the problem and
concentrate harder on work 2 2
Table 2. Coping with Office Noise
Changes In the Environment Percent Using
Close the door 24
Changes In Behavior
Ask my coworkers to be quiet 15
Work somewhere else 6
Wear ear plugs 6
Tak to my boss about the problem 3
Psychological/Emotional Processes
Ignore the problem
and concentrate harder on work 59
Just put up with H;there's nothing I can do 45
Table 3. Coping with Air Quality
Changes In the Environment Percent Using
Open a door or window 27
Use a fan 13
Changes In Behavior
Go outdoors for a while 43
Talk to coworkers about the problem 2 2
Contact staff person 18
Adjust clothing 10
Move to a more comfortable location 10
Psychologlcal/Emotlnal Processes
Just put up with it; there's nothing I can do 35
Ignore the problem
and concentrate harder on work 22
Talk myself into getting used to it 18
Behavioral and psychological coping were also selected frequently for air quality problems (see
Table 3). Interestingly, 43% of the occupants who experience air quality problems said they go
outdoors for a while. The only situation in which environmental coping was the most common
coping response was for excessive sun warmth or brightness (see Table 4). Closing the blinds was
the easiest way to resolve the problem. (It is worth noting that psychological coping predominated
in the one building which did not have operable window blinds.)
The data also indicate that simple coping processes (such as drinking something hot or cold) are
used more frequently than complicated or costly processes (such as leaving one's workspace).
However, even though going outdoors for a while or working elsewhere were not used as
frequently as other coping processes, from 10% to 20% of the occupants engaged in these
behaviors across situations.
-------
3044
Table 4. Coping with Sun Warmth and Brightness
SUN WARMTH
Changes In the Environment Percent Using
Close drapes 72
Open a door or window 8
Adjust thermostat 7
Changes In Behavior
Drink something cold 36
Move to a more comfortable location 14
Go outdoors for a while 12
Work elsewhere in the building 5
Psychological/Emotional Processes
Try to ignore the problem
and concentrate harder on work 15
Just put up with it; there's nothing I can do 13
SUN BRIGHTNESS
Changes In the Environment
Close the drapes 77
Changes In Behavior
Move to a more comfortable location 11
Work elsewhere in the building 6
Psychological /EmotlonalProcesses
Just put up with it; there's nothing I can do 12
Try to ignore the problem
and concentrate harder on work 7
The Costs of Coping
Although coping behaviors are generally expected to improve one's situation, there may be times
when coping is ineffective or detrimental to one's well-being. This may be expected, for instance,
when one tries to ignore an uncomfortable or disturbing situation. The coping item "I try to ignore
the problem and concentrate harder on my work" suggests an effortful endeavor that may, in fact,
be very difficult to carry out For instance, people who ignore sun brightness were more
dissatisfied with the lighting (Chi Square = 25.97, p=.002), had significantly more headaches (Chi
Square = 9.8, p=.04), more sore throats (Chi Square = 22.4, p=.002), and more nose irritation
(Chi Square=9.97, p=.04) than occupants who did not choose this coping behavior in response to
excessive sun brightness in their workspace. Similar results occur for coping with sun warmth and
noise.
The coping data also show that a number of occupants engage in avoidance or escape behaviors
(such as going outdoors, working elsewhere in the building, or walking around). This was
especially true for problems with thermal discomfort and air quality. If sufficient numbers of
people engage in these behaviors on a frequent basis, the end result could be lowered efficiency
and productivity on their jobs , which could lead to substantial personal and organizational costs.
Does Copinp Work?
Occupants were asked to rate how frequently their coping actions produced a more satisfactory
condition. As can be seen in Table 5, responses varied across situations. Coping strategies were
least successful for air quality, ventilation, and noise problems, and most successful for sun
problems.
Table 5. Ratings of Frequency of Coping Success
Percent Percent
Never/Rarely Usually/Always
Warm cfiscomfort
18
49
Cold discomfort
15
55
Too Kttle ventilation
33
45
Air quatty problems
48
32
Noise problems
32
43
Sun warmth
8
86
Sun brightness
11
76
Reflections on

computer
8
74
-------
3045
Table 6. Ratings of Bothersomeness /Interference of Discomforts
Percent
Bothered
Percent
Interferes
Usually /Always Mod./Very Much
Warm discomfort
Cold discomfort
Air quaity problems
Too ittle ventilation
Noise
Sun brightness
Sun warmth
Computer
45
45
46
49
46
52
37
32
37
25
78
34
40
26
reflections
54
67
The Impact of Environmental Discomforts
Another way of assessing environmental discomforts and problems is to look at how much they
bother people and interfere with their work. As shown in Table 6, almost half of the occupants
were bothered "moderately" or "very much" by all of the discomforts except sun warmth.
Interestingly, there was more variance in response to the question of how much the problem or
discomfort interferes with occupants' work. Having too little ventilation or experiencing
reflections on the computer screen created the most interference; thermal discomfort and air
quality problems interfered the least. The reasons for these differences are not apparent at this
time.
Cooinp in Different Settings
Because occupants of private offices frequently can operate drapes, open or close windows,
operate lighting, and rearrange furniture, they should be more likely than occupants of partitioned
or open workspaces to engage in coping behaviors aimed at changing the environment and less
likely to engage in psychological coping. To test this prediction, we combined all of the coping
responses for each kind of discomfort or problem into one group. The group was then categorized
according to whether the coping actions were environmental, behavioral, or psychological.
Analysis of variance conducted on the mean number of coping behaviors used by occupants in the
different workspace types indicates that occupants in private offices used an average of 3.25
environmental coping actions, compared to a mean of 2.34 for occupants of partitioned
workspaces and 1.9 for those in totally open work areas. The differences among the groups were
significant and in the predicted direction (F=4.32, p=.01). Analysis of psychological coping shows
that occupants of partitioned workspaces used an average of 3.28 psychological coping actions as
compared to 1.9 for those in private offices and 2.13 for occupants of totally open workspaces
(F=6.82, p=.001). There were no differences among the workspace groups in behavioral coping.
DISCUSSION
Because it is generally considered "more adaptive" and "healthier" for people to exert control over
the environment when opportunities exist (Cohen, Evans, Stokols, and Krontz, 1986), we would
expect to find that environmental coping strategies will be used whenever possible. The data
presented here suggest that such opportunities to change environmental conditions to suit
momentary needs are not widespread. In fact, the only situation for which environmental coping
was the most frequently used option was for excessive sun warmth and brightness. The building
environment for most occupants does not allow much opportunity for control of temperature,
noise, ventilation, or air quality conditions.
-------
3046
The data also suggest that although behavioral and psychological coping processes are widespread,
they are not very successful in alleviating discomforts and problems. For instance, psychological
coping predominates for noise, ventilation, and air quality problems; yet, the frequency of coping
success is the lowest for these problems. Furthermore, occupants frequent use of going outdoors
when they experience air quality problems may offer a temporary respite, but this also will not
resolve the situation. The discomfort is likely to re-emerge when they return to their workspaces.
The time and effort invested in leaving may be costly when its ineffectiveness in improving the
situation is taken into consideration.
The issue of control over the environment warrants much more attention from designers and
building engineers because of the potential negative outcomes associated with uncontrollable
problems and discomforts. Studies on control show that negative effects can be ameliorated when
some form of instrumental control is provided, even when it is not used. Knowing that potential
control is available provides a powerful psychological reinforcement (Evans and Cohen, 1987).
Although the "take-home" message of the research on environmental control seems to suggest that
building occupants should be given maximum opportunities for control, the issue is much more
complex and warrants further investigation. If occupants lave too many choices and are required
to make frequent adjustments in their workspace, this may actually lead to a sense of frustration
and stress (Fleming, Baum, and Singer, 1984; Evans and Cohen, 1987). The environment can
seem "out of control" in these cases. Future research should address such issues as how much and
what kind of control is important to occupants. The research presented here suggests that one
important consideration may be the effect of discomforts on work performance or other outcomes
such as affect, health, and social functioning.
Another critical aspect of coping is the extent to which it actually improves one's situation. When
coping efforts fail to modify the situation, a person may experience increased stress and an
increased sense of helplessness (Seligman, 1975). Although relatively little research has addressed
the outcomes of coping processes, there is some indication that even apparently successful coping
(e.g., it appears to solve the problem), can have after-effects. That is, the coping process itself can
produce cognitive fatigue, negative social interactions, negative affect, inattentiveness, and a
general reduction in coping capacity (Evans and Cohen, 1987). Whether or not these findings
apply to the everyday demands and discomforts one experiences in the work environment is not
known. The data presented here suggest that ambient discomforts are bothersome and frequently
interfere with work. Furthermore, since occupants' coping behaviors do not readily resolve many
of the situations, it is not unreasonable to expect that negative outcomes may indeed exist.
ACKNOWLEDGEMENTS
The research described in this paper was funded by the U.S. Department of Energy and the
Bonneville Power Administrtion (#DE-AC06-89RL11659). The principal investigator is Judith
Heerwagen and co-investigator is Joel Loveland. Rick Diamond of the Lawrence Berkely
Laboratory worked with the principal investigators on the development of the survey and data
analysis. Other members of the research team include graduate assistants John Barnes, Mary
Schaughnessey, Julian Somers, Nancy Quense, Carreen Press, and Ivan Suen.
REFERENCES
Baum, A., Singer, J.E., and Baum, C. 1982. Stress and the environment. In G.W. Evans (Ed.)
Environmental Stress. New York: Cambridge University Press.
Campbell, J. 1983. Ambient stressors. Environment and Behavior. 15:355-380.
Coelho, G.V., Hamburg, D.A., and Adams, J.E. (Eds.) 1974. Coping and Adaptation. New York:
Basic Books.
Cohen, S. 1980. Aftereffects of stress on human performance and social behavior: a review of
research and theory. Psychological Bulletin. 88:82-108.
-------
3047
Cohen, S., Evans, G.W., Stokols, D. and Krontz, D.S. 1986. Behavior. Health, and Environment:
Stress. New York: Plenum.
Evans, G. (Ed.) 1982. Environmental Stress. New York: Cambridge University Press.
Evans, G. and Cohen, S. 1987. Environmental Stress. In D. Stokols and I. Altman (Eds.) The
Handbook of Environmental Psychology. Vol.1. New York: Wiley.
Fleming, R., Baum, A., and Singer, J. 1984. Toward an integrative approach to the study of stress
Journal of Personality and Social Psychology. 46(4): 939-949.
Lazarus, R.S. 1966. Psychological Stress and the Coping Process. New York: McGraw-Hill.
Lazarus, R.S. and Folkman, S. 1984. Stress. Appraisal and Coping. New York: Springer
Lazarus, R.S., and Cohen, S. 1977. Environmental stress. In J. Wohlwill and I. Altman (Eds.)
Human Behavior and the Environment. New York: Plenum.
Pearlin, L.I. and Schooler, C. 1978. The Structure of Coping. Journal of Health and Social
Behavior. 19:2-21.
Seligman, M.E.P., 1975. Helplessness. San Fransisco: Freeman.
White, R.W. 1974. Strategies of adaptation: an attempt at systematic description. In G.V.Coelho,
D.A. Hamburg, and J.E. Adams (Eds.) Coping and Adaptation. New York: Basic Books.
Zimring, C. 1982. The built environment as a source of psychological stress: impacts of buildings
and cities on satisfaction and behavior. In G. Evans (Ed.) Environmental Stress. New York:
Cambridge University Press.
-------
3048
ARCHITECTURAL DESIGN METHODOLOGY
FOR ENVIRONMENTAL COMFORT
Prof. Calderaro *, Arch, A- Ciolf1 *#
Faculty of Architecture University "La Sapienza" of Rome
via A. Gransci, 53 Rome, Italy
* Responsible of Bioc'l inat ic Arch itecture Committee
«¦* Secretary of Biocl i.mat ic Architecture Commit tee
ISES Italian Section
ABSTRACT
The aim of the study has beer, to elaborate a simplified design methodology to
individualize the thermo-phys ical and geometries character if»t ics of the
building to control the internal comfort in all climatic conditions with
particular regard to passive coolings problems.
KEYWORDS
BiocliMatic design; Bioclimat ic control strategy; Methodology comf ort control;
Summer comfort method; Simplified method.
INTRODUCTION
fhroughout,the research is going to individualize a biocl imatic design
methodology that make possible, on the ground of external ambient
characteristics, adequate comfort conditions of internal ambient situationsjpar-
thiafy passive cooling problems.
Such metodology regards the possibility to individualize the necessary
characteristic;?; for the buildings to obtain adequate comfort conditions for the
interncil ambient.
To elaborate such method we have adopted an hourly mathematic simulation model
carried out in preceding works < Calderaro, 1987). This simulation model has
been used for- the elaboration of simplified passive cooling calculation method.
It controls the energetic behavior and internal comfort degree for several
climatic conditions and various type® of residential building.
The comfort index examined in the calculation mode] has been proposed by Fanger
in the I970. These index PMV were used in the 1384 internet ional standard for
estimation of the moderate thermal environments. For human thermal sensation a
scale of seven values uas adopted:
-------
3049
+ ?i very uarn
+ 2 uiarn
+ 1 light warm
0 comfort
+ 1 light cold
f Z cold
+ 3 very cold
The simulation model elaborated fcunded on the Ceehique of finite differences,
enables the evaluation of the hourly, energetic and thermohycjrometrio behaviour
of both traditional and biocl mat ic buildings with various types of passive
solar heating and cooling systems.
This model, already has been validated according to experimental results uith
the prototype of a complex bioclimatic greenhouse uith natural heating and
cooling systems (Caldsraro, Nov. 1988). Using this simulation model a
simplified methodology has been developed for an easy approach to calculation.
Therefore it has been defined that a number of passive system conf igurat ions, of
which the dimensional, operative (overhangs, variable resistences) and thermal
capacity characteristics have been varied. There are several simulations
carried out in different climatic condition:", for all configurations.
Through the results of this analysis, the elaboration of a progressive design
method for the comfort control is used uith regard to the PMV and PPD indexes.
Uith this method it has been possible to individualize the best design solutions
for the various climatic, conditions.
Several simulations have been carried out for different systems configurations
through dimensional ratics related to different reference conditions for the
occupants (metabolic rate and external uorl< , thermal resistance of clothing,
relative interior air velocity).
In particular, the methodology proposed has individual izecj
according to the external climatic characteristics and building types, the best
solution relative,tp the following aspects:
- sun sheltering for solar radiation
- thermal capacity of the rooms;
- heat transmission and thermal capacity for external walls and roofs;
Through results obtained by simulation ue have reached the following
concluslon:
- the remarkable values of temperature and PMV are due to effects of solar
radiation principally;
- the variation of relative humidity generally remairB in the comfort range;
- high values of internal temperature, in some cases, even if shielding devices
are present, are due' to the high values of external air temperature in those
climatic; conditions.
- the presence of shielding for all systems remarkably reducesthe temperature
and PMV values.
For the study there were carried out, through the mentioned simulation model,
several simulations for some ambients that have different shape, thermophysic
-------
3050
chcirao t cr: r> t i cs ( i nrsul at i ngeirid thermal capacity), in different ftalian climatic
conditions and relative simmer conditions.
Analyzing the obtained results;
-¦ it's aluays necessary in order to limit . of the ambient';, to fit
in elements of external obstructions (sun sheltering) on the. external glass
• surfaces for total protection of direct solar radiation and partial ¦
diffuse solar radiation;
- for the considered conditions the thermal insulating assumed & limited role;
- uas very affecting the influence of thermal capacity of the ambient;
- relative. to climatic characteristics of external environment very important
the entity of the incident solar radiation and the state of the temperature
of external air.
And then it's going on to individualize the correlation throungh the different
parameters (geometric arid thermophysics of the buildings and climatic for
external enviroment) and the PMV.
The adopted parameters are:
- total daily solar radiation incident on the horizontal surface (Kcal/h sqm);
- the difference through the max external temperature and an internal
temperature of reference (
-------
7G00
6000
5000
4000
3000
Q2 0.4 0.6 08 1.0 1.2 14
PMV (/)
-------
3052
I'he graphs'tfbtained offer the possibility to individualize the initial stage
of the design,^the characteristics of the enviroment able to achieve the
cva) u- at ion of the condition of internal ambient (PMV).
In the fallowing stager, of the research well amplify the field of parameters
of "the building and the externa}, environment in order to yield more complete
methodology.
REFERENCES
V.Calderaro, A simulation model for the hourly evaluation of the
t hermohygrome tr :i c , energetic, lighting and acoustics behaviour of buildings,
International Congress of International Solar Energy Society - Proceeding
'Jo!'_-r Mori d Congress 1Q07, Hamburg, 13-U3 September I9G7,
V.Calderaro, Simulation model validation of bioclimatic greenhouses, Proceeding
Thirteenth National Passive Solar Conference dell'ASES (finerican Solar Energy
Society) -- MIT - Boston - 21 june 1980;
V.Calderaro, Passive solar architecture Handbook of design, ed. Kappa. Rome,
Italy, 1981.
V.Calderaro, H. Ciolfi, Evaluation of thermchigrometric behaviour of building
with simplified calculation method for the passive cooling, Proceeding
Conference 1SES on Bioclimatic Architecture Energy Saving and Environmental
Quality, Rome Italy, 9/10 novemher 1980; e qualita ambiental e , Roma,
novembre 1 988 *,
V.Calderaro, A, Ciolfi» Evaluation Methodology of environmental comfort in
bioclimatjc architeciure, Proceeding International Conference of J SES Italian
Section on "Evolution of external perimetral components in bioclimatic
architecture" , Milan, Italy, 5-6 April 1990;
V.Calderaro, A general model for cooling design, Proceeding Workshop on Passive
Cooling organized of CXE (Commission Of the European Communities of the Joint
Research Center of Ispra, 'I - A april I lJ 90,
V. Ca.lderar o , E. EJarbera. lie •, ign methodology and performance verification of
the buildings envelope with the use of operational graphs, Proceeding
International. Conference of ISES Italian Section on "Evolution of external
perimetral components in bioclimatic architecture", Milan, Italy, 5-6 April
j S H (');
V.Calderaro, Bioclimatic innovative components and relative simplified
calculation methods Proceeding Intornat)ona.l Conference of ISES Italian
Section on "Evolution of external perimetral components in bioclimatic
architecture" , Milan, Italy, IV S April 1 99 <2>.
-------
3053
NEW INITIATIVES + SOLAR DESIGN = HEALTHY BUILDING DESIGN
Bion D. Howard*
"The Alliance to Save Energy, Program Manager - Building Science
1725 K. St. NW Suite 914, Washington, DC 20006 USA
ABSTRACT
This paper provides an overview of how passive and active solar residential design participates
in the provision of energy efficient, healthy and marketable indoor environments. Consid-
eration of solar building design provides another useful asset in fighting indoor air pollution
for healthy, energy efficient dwellings. Indoor air quality (IAQ) and other environmental
challenges inside buildings can either be perceived as threats or as opportunities to learn,
grow and profit as a designer and/or builder. Proper attention to alleviating these problems
may also reduce the likelihood of litigation related to fitness of use of the structure.
Research in indoor air quality is reviewed for designers and builders of energy efficient
homes. The author suggests thirteen general design and construction guidelines for maximum
energy efficiency with good indoor air quality. These basic IAQ enhancement measures add
only marginally to the cost of a high quality home. The improved product can then be
marketed effectively to energy and environmentally conscious consumers.
INTRODUCTION
Leading indoor environmental experts are leaning towards new initiatives in the delivery
mechanism for healthful houses from the indoor micro-environmental scale right up to the
global environmental perspective, that include both passive and active solar design. If our
housing is to become a better reflection on our personal character and regard for our planet,
then certain conditions need to be met and guidelines employed.
Design parameters, material specifications, economics and technology decision-making operate
in a complex fashion leading to the delivery of new housing. These factors may either help
or hinder the building occupants, by providing either a relatively safe or an un-healthy indoor
environment. We have become adept at measuring numerous indoor environmental technical
parameters, but builders and homeowners aren't research chemists. Once contamination is
encountered, mitigation is always more expensive than sound planning for its exclusion. The
occupant must take a proactive role in maintaining good indoor environmental quality
following delivery of the well designed, constructed and properly finished quality home.
Problem Assessment -- Unhealthy Buildings?
U.S. Building "stock" represents serious energy utilization and environmental challenges and
opportunities, such as:
Energy use = 36% to 40% of total primary energy
Energy "bill" = $165 to $200 billion / year
Potential Energy Savings = $85 - $100 billion / year
Pollutant emissions = total 500 million tons/year (90% C02)
Potential emission reduction = 220 - 250 million tons/year
-------
3054
Providing only a one-percent reduction in residential energy use would result in a $2 billion
savings, and about a 5 million-ton reduction in estimated energy related emissions, in the first
year following implementation (Howard and Sheinkopf 1991).
The building stock in the United States alone is of staggering proportions. According to EPRI
there are about 157 billion square feet (over 16 billion M2) of floor space in residential
buildings. If this area is simply multiplied by 8.25 feet (analogue of average ceiling height)
the resulting indoor conditioned volume equals about 1.3 trillion cubic feet (42 billion Ms).
If this volume obtained an air change rate of only 8-times per day, about the recommended
ASHRAE Standard level 0.35 Air-changes/hour (ASHRAE 1989), this means our buildings
exchange about 7.2 billion cubic feet of air per minute with the outdoors. Despite all of this
air flow (about 90 cfm per dwelling) health physicists warn that indoor pollution levels can
be much higher than outdoor levels. This continuous low-level pollution may be gradually
eroding our citizen's physical health by causing multiple chemical sensitivity disorders which
also take their toll on job productivity, and emotional well-being (Ashford and Miller 1989).
Various complex indoor environmental problems are documented in many studies by EPA,
NIOSH, CPSC and the medical research community. A great deal more detailed research
on commercial building indoor environments has been done, due to the potentially severe
negative impact on worker productivity. Residential indoor air quality (IAQ) guidebooks
have been developed (Nisson 1989), general builder guidelines proposed (Environmental
Health Watch 1990), comprehensive builder workshops provided (NAHB-RC 1989), and
industry-based voluntary programs planned (Howard 1990).
In residential buildings indoor air problems and other environmental difficulties can include,
but are not limited to, the following ten key general conditions:
elevated source concentrations due to ill-advised selection of construction products,
pollutants infiltrating from the building exterior including radon gas in soils,
ineffective rates of ventilation, and/or air mixing efficiency in rooms or zones,
spillage and backdrafting of combustion by-products from unvented fuel appliances,
excessive indoor moisture generation and/or, poor drainage, standing-water or
intermittent moisture sources near foundations, roofs, walls, doors and windows,
incorrect, insufficient or non-existent appliance and mechanical system maintenance,
improper or insufficient sanitation, including poor or excessive hygienic practices,
occupant importation of cleaning chemicals, dry-cleaned clothing, solvents, fuel gases,
hobby materials, household pesticides, tobacco smoking, and pet-care supplies,
improper wood-stove, fireplace and/or portable fueled unit heater design or use, and
unvented ethnic culinary practices using odorous foodstuffs, oils and spices.
Builders should mainly be responsible for reducing toxic exposures by good design and the
specification of less polluting (lower emission rate) materials, adhesives, caulks, carpeting,
cabinets, paint, finishes, etc. The housing consumer needs to be informed, in simple, easy to
understand terms, of the proper means to operate the home to avoid increasing the likelihood
of poor indoor air quality or other serious indoor environmental problems following delivery.
Solar "Resources" and Systems
Why can solar design help, and how much is there? Solar energy provides about 15,000 times
World primary energy use per year. At one time, all fossil fuel resources were "solar" through
photosynthetic biomass production in the primeval rain-forests and oceans. Modern society
essentially only taps 1% or less of this prodigious resource now. Renewable energy sources
including hydro-electric and biomass utilization are already providing four-times the world's
-------
3055
current nuclear electrical generating capacity.
Solar designed buildings already save U.S. consumers about $140 million per year, and are
also helping to reduce related power plant emissions. There are now over 300,000 passive
solar homes and 1.2 million solar hot water systems now in operation in the U.S. According
to the Solar Energy Industries Association (Sklar 1990, SEIA 1990) U.S. solar hot water
heaters now save about $240 million per year. However, this levd of adoption pales before
the Japanese, who now have over 1.5 million DHW systems in metropolitan Tokyo, alone.
Photovoltaic (PV) systems are also emerging as an important component in the future of
home power systems. The south facing unshaded U.S. single family residential roofs, equaling
just 8% of the total U.S. roof surface area, could account for an $8 billion per year electric
power resource yield if it were used for PV array sites (Howard and Sheinkopf 1991).
A wide array of monitored performance results is available documenting active and passive
solar building performance. Typical residential passive designs have provided 35-40%
auxiliary energy consumption reductions (Howard and Saunders 1989). Solar domestic hot
water systems have also been field tested with good results, in the 50-90% savings range
(Howard 1982, Sklar 1990). Architecturally designed ventilation systems have proven effect-
ive in reducing energy use in hot climates, and help maintain indoor comfort (FSEC 1987).
Groups like the National Association of Home Builders (NAHB) have endorsed passive solar
utilization in their voluntary guidelines and programs (Balcomb 1990). Even building codes
organizations like CABO (Council of American Building Officials) have discussed inclusion
of conservative credits or trade-offs for energy efficient passive solar techniques in their
Model Energy Code (CABO 1989).
The American Society of Heating Refrigerating and Air-conditioning Engineers (ASHRAE)
has published both passive and active solar design manuals (ASHRAE 1984,1990). ASHRAE
encourages the use of passive solar designs and solar low-temperature water heating in the
soon to be released Residential Standard 90.2, and in the revised Commercial Standard 90.1.
Work is continuing on whole-building design implications of daylighting, thermal mass,
insulation and HVAC control interaction that have potential for improving IAQ. Automated
"smart" controls will take advantage of ("learn") the thermo-optical properties of solar
buildings, and further integrate monitoring and control functions improving system
performance. Other important emerging areas of potential value to improving residential
IAQ include:
CONSTRUCTION MATERIALS WITH APPLICATION 'TAILORED"
THERMAL AND MOISTURE PROPERTIES,
SOLAR PRE-HEAT SYSTEMS FOR OUTSIDE VENTILATION AIR,
HYBRID SYSTEMS - ACTIVE CHARGE / PASSIVE DISCHARGE
SUCH AS THE "AIR-CORE" APPROACH (Howard 1986), and
SOLAR-REGENERATED DESICCANT COOLING.
DISCUSSION
John Spears, a consultant to EPA and the housing industry, summarized residential IAQ
challenges and solutions in the Healthy House Catalog (Environmental Health Watch 1990).
He described crucial stages in house construction and discussed potential problems, the
pollutants in question, and recommended appropriate solutions. This information is
summarized and updated (Table 1) derived from Spears' "Builder's Guide" recommendations.
-------
3056
TABLE 1.
Healthy House Construction (adapted from Environmental Health Watch 1990. bv author)
Process Step Problems Pollutants Proposed Solution
Site Selection Moisture, Soil Gas, Radon, methane^
Water and air- pesticides, VOC ,
pollution (various), lead, CO, COz,
Insufficient sunlight microbes, toxic
Poor solar access, wastes
Site investigations,
isolate foundation from soil,
test and filter water,
well drained land,
avoid certain fertilizers,
review solar orientation.
Site Prep.
Foundation
Building
Mechanical
Plumbing
Poor drainage, Pesticides, micro-
termites, methane, organisms, off-gas
pesticides, cracking from woody waste
of foundation walls.
Concrete additives, Radon and decay
radon, other soil products, mold,
gas, insects and mildew, various
pests, moisture, pollutants enter,
basement too "dark"
Uncontrolled air- Hydrocarbons, CO,
leakage, leaks from COz, VOC, Penta,
attached garage,
off-gasing from
construction and
finishing products,
treated lumber.
Arsenate comp's,
aerosols, NOz,
asbestos, soil gas,
sawdust.
Adjust grading for drainage,
clean backfill, no waste in
backfill,alternative pest control,
retain solar friendly trees, shrubs.
4" (10.1 cm) sub-slab gravel
under slab, radon passive
vent stack, good waterproofing
(not "damproofing"), avoid
heavy soil termite treatments,
use "daylit" basement design.
Reduce air-leaks, provide
controlled mechanical ventilation,
seal-off and vent garage,
use low formaldehyde emission
manufactured wood products, vent
house during painting, choose
safe finish compounds and materials,
utilize well insulated passive design.
Excess air-flow or
insufficient flow,
backdrafting and
spillage of comb,
by-products, excess
or insufficient humid-
ity, particulate and
dust, AHU depressure
Combustion gases, Ventilation system, provide fresh air
SRP3, mildew, moid, intake, sealed combustion appliances,
bioaerosols, radon, insulate ducts on exterior, proper
register & return location, avoid
oversizing H/AC system air-handler,
maintain 30-50% rel. humidity, pre-
condition fresh air using solar heat,
consider hydronic under-floor heating.
mineral / cellulose
(insulation) fibers.
Pollution entering
from well water
or from fabrication
Lead, organics, Avoid lead in solder, test water,
particulate, install whole house water filter,
microbes consider plastic piping, use solar
for domestic hot water.
Appliances Allergic reactions, Combustion prod's. Whole house vacuum cleaner,
headaches, odor,
moisture
microbes, mold,
SRP, dust, bio-
aerosols.
Notes: * VOC denotes Volatile Organic Compounds
@ SRP denotes Suspended Respirable Particulate
vented combustion equipment or
fuel-switch to solar water heater with
off-peak electric back-up, give
occupants information on proper
maintenance procedures.
-------
3057
RECOMMENDATIONS
Using the information provided in Table 1 and the recommendations presented here, with
study of the original documents cited in this paper, prepares builders and designers for the
minimal application of good indoor air quality practices. Building professionals should:
• maximize economical envelope insulation levels, so auxiliary system(s) may be downsized,
• reduce air-leakage using proper sealing techniques and blower door test verification,
• employ passive solar heating, daylighting, and architectural cooling design,
• use high-tech glazings to boost indoor glass surface temperatures avoiding condensation,
• consider mass-materials for floors and eliminate / reduce wall-to-wall carpeting,
• seal and ventilate soil gas and moisture entry routes at the insulated foundation,
• carefully select construction and finish materials to minimize pollutant off-gasing,
• provide a flow-rate, and over-ride controllable mechanical ventilation system,
• carefully design, size, install, seal, exterior insulate and test H/AC forced-air ducting,
• consider alternative heating systems such as solar "Air-Core" or hydronic radiant floors,
• preheat winter ventilation air using a sunspace, vented Trombe' wall, or solar air-panel,
• design floor-plans to encourage interior air-circulation for ventilation effectiveness,
• deliver the completed building with an "operators manual" boosting occupant awareness,
and fully explain the operation of all relevant features to the consumer.
CONCLUSION
Analysis by the author, based on over seven years of studying the relationship between home
energy efficiency and indoor environmental quality, indicates the guidelines provided in this
paper are basically correct, though certainly not all inclusive.
Designing residential buildings to include solar technologies can reduce purchased energy use,
provide better indoor thermal comfort, boost indoor air circulation, lessen need for large
mechanical systems, lower the potential for mildew and mold, daylight otherwise darkened
spaces, and provide fuel-switching options; all reducing the potential for indoor air quality
problems.
Readers are cautioned, however, that good indoor environmental quality in buildings is a
complex and systematic achievement. Bits-and-pieces approaches are likely to produce
uncertain results or problems. Comprehensive design for energy efficient, healthy, safe and
marketable solar homes is an art of applied building science. Good information now exists
to help builders and designers make this vital transition to providing healthy housing.
DISCLAIMER
The author, and the Alliance to Save Energy, provides no expressed or implied warranty
through the application or misuse of these fundamental technical recommendations. These
basic guidelines are provided strictly for purposes of discussion on the myriad issues facing
energy efficient builders in the 1990's regarding the indoor environment. These IAQ
recommendations shall not be used to abridge any state or local building codes.
REFERENCES
. ASHRAE Standard 62-1989 "Ventilation for Acceptable Indoor Air Quality." ASHRAE
Publications, 1791 Tullie Circle NE, Atlanta, Georgia.
-------
3058
. ASHRAE. 1984. "Passive Solar Heating Design."
. ASHRAE. 1990. "Active Solar Energy System Design and Engineering."
. Environmental Health Watch. 1990. Healthy House Catalog, cf. J.W. Spears, pg. 21-28.
Avail. Housing Resource Center, Cleveland, Ohio.
. FSEC. 1987. Cooling With Ventilation. Florida Solar Energy Center, Cape Canaveral
. Model Energy Code 1989 (90R) Council of American Building Officials (CABO),
Falls Church, Virginia.
. SELA. 1990. "Annual Survey of Low and Medium Temperature Solar Collector
Production." Solar Energy Industries Association, Washington, DC.
Ashford, N.A and C.S. Miller. 1989. "Chemical Sensitivity: A Report to the New Jersey State
Department of Health." New Jersey State Department of Health, Newark, New Jersey.
Balcomb, J.D., C. Ely, and PSIC. 1990. "Passive Solar Builder Guidelines Program" SERI and
the Passive Solar Industries Council (PSIC), Washington, DC
Howard, B.D. and K.G. Sheinkopf. 1991. "Solar Building Options for Demand Side
Management." Proc. 1991 ISES World Congress. Boulder, Colorado.
Howard, B.D. 1990. "Healthful Residential Indoor Environments - Builder Program
Development." pg 289-294. Proc. of 5th International Conference on Indoor Air Quality and
Climate. Toronto (Avail: Indoor Air Tech., PO Box 22038, Ottawa, CANADA K1V 0W2.)
Howard, B.D. and D.H. Saunders. 1989. "Building Performance Monitoring - Thermal
Envelope Perspective." (invited1) 4th - Building Thermal Envelopes Conference.
ASHRAE/DOE/BTECC. (Avail: ASHRAE, Atlanta, GA)
NAHB Research Center (Howard, ed.) 1989. "Residential Energy Efficiency and Indoor Air
Quality." Proc. Better Buildings Conference. February 1989, New York-SEO, Albany.
Howard, B.D. 1986. "Air Core Systems for Passive and Hybrid Energy - Conserving Buildings."
ASHRAE Transactions. Volume 92. Part 2. Atlanta, GA
Howard, B.D. 1982. "Monitored Performance of 18 Solar Hot Water Systems in the NSDN."
Proc. of ASES Annual Meeting. Houston, Texas.
Nisson, J.D. 1989. "Indoor Air Quality in Homes." Energy Design Assoc., Cutter Information
Corp., New York, NY
Sklar, S. (Pers. Comm.) "Cost effectiveness of solar hot water systems." September 1990.
SEIA, Washington, DC.
-------
3059
THE HELIOTHERAPISTS
R.A.Hobday
Databuild, 4 Venture Way, Aston Science Park, Birmingham,
United Kingdom
ABSTRACT
The work of some of the European practitioners and advocates of heliotherapy, who
re-established the use of sunlight as a therapeutic agent after many centuries
of neglect, is described. Their influence on building design and public health
is also discussed.
KEYWORDS
Heliotherapy; sunlight; health; building design; schools; hospitals; direct gain.
INTRODUCTION
In recent years the therapeutic benefits of sunlight have received rather less
attention than the hazards of excessive exposure. However, for many thousands of
years solar radiation has been used to help prevent, and cure, a wide range of
afflictions. Sunlight was prescribed by the early physicians of Greece and Rome,
but it was only in the nineteenth century that the health benefits of exposure
began to be investigated scientifically.
THE RE-EMERGENCE OF HELIOTHERAPY
In 1903, three significant events in the history of the 'sun-cure' occurred.
Firstly, evidence came to light which led to the rediscovery of the great Health
Temple at Kos (Caton, 1906) . This had been built on the site where Hippocrates,
the 'Father of Medicine', is believed to have practised sunlight therapy.
In the same year, 1903, the Nobel Prize for Medicine and Physiology was awarded
to the Dane, Dr. Niels Finsen (1860-1904), who was the first physician of the
modern era both to study and exploit the use of sunlight in clinical treatment.
Finsen was awarded the Nobel Prize for his success in treating lupus vulgaris,
a highly disfiguring form of tuberculosis affecting the skin (Sourkes, 1966;
Diffey, 1982). The diseased tissues of his patients were exposed to concentrated
ultra-violet radiation derived from the sun or artificial light sources; the
results were considered to be superior to those achieved by surgery and Finsen,
the founder of modern phototherapy, received many patients from around the world
at his Medical Light Institute in Copenhagen. Ten years earlier Finsen had come
to public attention when he reported a method of treating smallpox which
consisted of isolating patients in red-lighted rooms in order to protect them
from ultra-violet 'chemical' rays at the other end of the spectrum. This
-------
3060
procedure was successful in reducing the scarring, suppuration and secondary
fever associated with smallpox (Finsen, 1901). However, it was Finsen's cure for
tuberculosis which brought about renewed interest in the therapeutic properties
of light both in Europe and the United States (Bloch, 1990).
The third significant event of 1903 was that Dr. Auguste Rollier (1874-1954)
began the treatment of tuberculosis with sunlight at his clinic at Leysin,
Switzerland. This was claimed to be the first clinic devoted exclusively to the
systematic treatment of external, non-pulmonary 'surgical' tuberculosis. Whilst
numerous Alpine sanatoria had been built to provide general treatment for
pulmonary tuberculosis, prior to the heliotherapy of Rollier, the only remedy
available for tuberculosis of the bones and joints was surgery. Rollier
considered tuberculosis to be a disease in which the individual power of
resistance was decisive and that surgery did not alter the constitution of the
patient and could not prevent relapses or delayed manifestations of the disease.
Rollier's approach to heliotherapy, in which sunlight was administered to the
entire surface of the body, was largely empirical, given the paucity of knowledge
concerning the action of light at this time. It required great skill on behalf
of the physician and was very closely tailored to the needs of each individual
under treatment. Exposure to the sun was very gradual and was preceded by a
period of acclimatization to the Alpine conditions of Leysin. Heliotherapy, as
practised by Rollier, combined exposure to the sun with cool conditions and open
air; Rollier recommended that in the summer months the treatment should be given
between 6 am. and 9 am. - and at lower altitudes even earlier still - if the
best results were to be achieved.
The 'heat' of the sun was avoided except where the patients needed to be warmed,
such as cases of chronic rheumatism. Gradual exposure to cold air was important
because it raised and maintained high metabolic activity in patients. This, in
turn, was considered to improve their general health and enabled them to resist
and overcome infection. It was also argued that cool conditions and low winter
and summer sun reduced the risk of over-exposure to the sun's rays. Rollier
considered exposure to the sun at temperatures greater than 18°C to be a 'hot-air
bath' and not a sun-bath (Rollier, 1927).
Diseases which were prevalent in the heavily polluted industrial cities of Europe
at this time, notably tuberculosis and rickets, were responsive to the sun-cure
practised by Rollier. He also reported success in treating a number of non-
tubercular conditions including burns, varicose ulcers, osteomyelitis, war
wounds, lacerations of the skin, septic abscesses and fractures. He published
several books on heliotherapy including 'La Cure de Soleil' in 1914 and 'Quarante
Ans d'H61ioth£rapie' in 1944.
In Britain Dr.C.W.Saleeby (1878-1940), whilst not a heliotherapist, was a
formidable advocate of the importance of sunlight to health; he drew public
attention to the need for smoke abatement in cities and to the hygienic
properties of sunlight. During his campaign for what he termed 'helio-hygiene'
Saleeby had visited New York and a number of other cities in North America where
sanitary regulations forbidding the production of coal smoke had resulted in a
significant decrease in the incidence of tuberculosis.
It was in America that the work of the English physician Dr. Theobald Palm (1848-
1928) was drawn to Saleeby's attention. Palm had shown in 1890 that the chief
factor in the causation of rickets is the deprivation of sunlight (Powers and
colleagues, 1922), but his work had largely been ignored in Britain.
In 1924 the Sunlight League, established under the chairmanship of Dr. Saleeby
-------
3061
to promote the health benefits of sunlight, adopted the following proposals which
were derived from the work of Palm:
The recording of sunshine in the streets and alleys of our
smokey cities, as well as at health resorts; using means to
indicate the chemical activity of the sun's rays rather
than its heat.
The removal of rickety children from large towns to
Sanatoria in sunlit places.
The systematic use of sunbaths as a preventative and
therapeutic measure in rickets and other diseases.
The education of the public to the appreciation of sunlight
as a means of health; teaching the nation that sunlight is
Nature's universal disinfectant, as well as a stimulant and
tonic.
Such knowledge will also stimulate efforts for the
abatement of smoke and the multiplication of open spaces,
especially as playgrounds for the children of the poor.
A twenty year campaign by Saleeby for effective legislation to control air
pollution was rewarded with the less than satisfactory Public Health (Smoke
Abatement) Act, 1926. However, an enquiry into the scientific basis of the sun-
cure was undertaken largely due to his efforts; a Committee on Biological Action
of Light was appointed by the Medical Research Council in 1922 (Saleeby, 1929).
Saleeby was also Rollier's champion and was responsible for the publication of
the English translation of 'La Cure de Soleil'. This appeared in 1923 under the
title 'Heliotherapy' and gives descriptions of the methods used by Rollier when
treating tubercular and non-tubercular patients and the results he obtained.
Included in this remarkable book are Forewords by Saleeby and by Sir Henry
Gauvain (1878-1945) who used sunlight therapy at two sister institutions in
Hampshire; one at Alton and the other on the south coast at Hayling Island.
Gauvain, a leading orthopaedic surgeon, approached heliotherapy somewhat
differently from Rollier in that he combined exposure to sunlight and fresh air
with sea bathing. Patients who were very weakly, or who could not respond
effectively to the strong stimulus of the seaside, started their treatment with
gradual and careful exposure for a long period inland. He also made use of
techniques and equipment developed by Finsen, and one of the many awards he
received in recognition of his work was Honourary Membership of the Copenhagen
Medical Society (Anon, 1945). Gauvain believed that the south of England was a
particularly suitable location for the treatment of tuberculosis because of the
variety of weather available and the accessibility of the sea. Whilst Gauvain did
not regard heliotherapy to be a cure for surgical tuberculosis, he considered it
to be of great value in all cases where the natural resistance of the patient to
disease needed to be improved (Gauvain, 1933). In an article to The Times of the
11th May 1922 he outlined his approach to heliotherapy and argued for its wider
adoption:
'Sunlight stimulates and enlivens; it is of help in almost
all conditions. It is the greatest of all natural tonics -
like good champagne, it invigorates and stimulates;
indulged in to excess it intoxicates and poisons. In
suitably graduated doses, which vary greatly in individual
cases, cumulative and favourable effects are produced which
are not accompanied by those unfortunate sequels associated
with many drugs. Sunlight and fresh air are such valuable
therapeutic agents that the extravagant claims made by
-------
3062
enthusiasts are not only wholly unnecessary, but actually
may defeat their object. Statements such as "the sun will
cure all forms of surgical tuberculosis" are exaggerations,
are contrary to fact, and, if persisted in, will tend to
bring into disrepute a method of treatment of the highest
value, but possessing definite limitations.
In suitably selected cases, graduated and skilled exposure
to sunlight aids and accelerates the cure of surgical
tuberculosis, both by the beneficent local reaction it
produces and by its effect on the general well-being. So
also will many other conditions respond, and it is for the
general recognition and wide application of this potent
natural therapy that I plead.'
The British public were also made aware of the therapeutic value of open-air
treatment and sunlight by Sir Leonard Hill (1866-1952) who, from 1914 to 1930,
was Director of Applied Physiology at the National Institute for Medical
Research. He investigated the influence of environmental temperature, humidity
and air movement on health and published widely on the subject (Douglas, 1953).
Of particular importance to Hill, and for that matter Rollier and Gauvain, were
the beneficial effects of sunlight, fresh air and exercise to children (Hill,
1925) . In the years before World War II it was generally accepted that these
three factors were essential to the healthy development of the young.
BUILDINGS AND SUNLIGHT
In the 1920s and 30s the renewed appreciation of the importance of the
therapeutic and sanitary properties of sunlight brought about, in part, by the
work of Finsen, Rollier, Saleeby, Gauvain and others had a marked influence on
the design of public buildings; many schools and hospitals were oriented to
accept sunlight in an attempt to improve the health of pupils and patients alike.
Sir Leonard Hill was a member of the Royal Institute of British Architects Joint
Committee whose report 'The Orientation of Buildings' (RIBA, 1933) was the first
publication to give architects authoritative guidance as to the use of sunlight
in buildings. Gauvain, an authority on hospital planning, wrote the Foreword to
this document; the health benefits of sunlight were stressed in the Introduction
which begins as follows:
'During the last few years an extraordinary and even
revolutionary change has taken place in all countries in
the general appreciation, both by the medical profession
and by the general public of the values of fresh air and
light, particularly sunshine.
The treatment of some diseases by exposure of the skin to
the action of light, natural or artificial, has in a
marvellously short space of time leapt from the obscure
position of a somewhat contemptuously neglected specific to
the status of one of the most valuable and even invaluable
weapons in the medical armoury.'
Sunbathing was also addressed:
'We have seen sunbathing, regarded not long since as a mere
fad practised only by eccentrics, develop - almost within
-------
3063
the space of months - first into a fashionable cult, and
later into a recognised necessity for which facilities are
now being provided by municipal authorities.
The results of these medical triumphs have been far
reaching - and now they directly affect the work of the
architect.'
HELIOTHERAPY'S DECLINE
Although still used to treat arthritic and psoriatic complaints in a number of
countries, sunlight therapy has declined in popularity in Western Europe since
the 1940s. The incidence of tuberculosis has, in general, decreased and medical
practise has changed. Schools are no longer built to exploit the health benefits
of fresh air and sunlight, and hospitals have ceased to be built with the open-
air balconies and solaria favoured by Gauvain and his fellow heliotherapists.
Sunbathing is undertaken largely for cosmetic reasons and the health benefits of
exposure during the early hours of the day are not widely appreciated.
However, the public appears to be at risk from inadequate exposure to the sun;
the residents of modern cities spend a large proportion of the daylight hours
indoors, or in transit, and are deprived of the biologically active component of
sunlight (Kime, 1980; Downing, 1988). There is also growing concern for the
health of occupants of modern buildings; particularly those in energy-efficient
offices with low levels of natural light and little fresh air. The author
speculates that greater understanding of such issues may lead architects to
design buildings which are naturally ventilated, and which exploit the
therapeutic and hygienic properties of sunlight. According to Butti and Perlin
(1980), the Romans considered the solar heating of buildings to be healthier than
artificial heat; and the notion that the heating and illumination of buildings
with sunlight confers physiological and psychological benefits on occupants is
gaining currency once again (Venolia, 1988; Samuels, 1990). It is for these
reasons that direct-gain passive solar design, which appears to offer the
benefits of both energy thrift and occupant health, may increase in popularity
in the years ahead.
CONCLUSIONS
Whilst the health risks of exposure
concern, the work of Rollier, Gauvain
importance of sunlight in building
overlooked. A re-evaluation of the
overdue.
to the sun's rays are a cause for serious
and other heliotherapists suggests that the
design and public health should not be
work of the pioneers of heliotherapy is
ACKNOWLEDGEMENTS
The author would like to express his gratitude to the many librarians who
assisted with this research, in particular, Mrs. M. Bougas at the Royal College
of Surgeons of England; Mr. R.J. Moore of the National Institute of Medical
Research; Miss S. Gold of the Wellcome Institute for the History of Medicine; and
-------
3064
the staff of Aston University Library.
REFERENCES
Anon, (1945). Obituary: Sir Henry Gauvain. Brit. Med. J. . Feb 3, 167.
Bloch, H., (1990). Solartheology, heliotherapy, phototherapy, and biologic
effects: a historical overview. J. Natl. Med. Assoc.. 82, No.7, 517-521.
Butti, K. , and J. Perlin (1980). A Golden Thread - 2500 Years of Solar
Architecture and Technology. Marion Boyars, London.
Caton, R. (1906). Hippocrates and the newly discovered health temple at Cos.
Proceedings of the Roval Inst. Gt. Brit.. 18, 258-268.
Diffey, B.L. (1982). Ultraviolet Radiation in Medicine - Medical Physics
Handbooks: 11. Adam Hilger Ltd., Bristol.
Douglas, C.G. (1953). Leonard Erskine Hill. Obit. Notices Fellows of Roval Soc..
.8, 431-443. Downing, D. (1988). Day Light Robbery. Arrow Books, London.
Finsen, N.R. (1901). Phototherapy. Edward Arnold, London.
Gauvain, H.J. (1922). The sun cure - ailing children's new hope. The Times. 11th
May, 17-18.
Gauvain, H.J., (1933). The Hastings Popular Lecture on sun, air and sea bathing
in health and disease. Supplement to Brit. Med. J.. Feb 25th, 59-61
Hill, L. (1925). Sunshine and Open Air - Their Influence on Health, with special
reference to the Alpine Climate. Edward Arnold, London.
Kime, Z.R. (1980). Sunlight Could Save Your Life. World Health Publications,
Penryn, California.
Powers, G.F., E.A. Park, P.G. Shipley, E.V. McCollum, and N. Simraonds (1922). The
prevention of the development of rickets in rats. J. Am. Med. Assoc. . 78, No.3,
159-165.
RIBA (1933) . The Orientation of Buildings - Being the Report of the RIBA Joint
Committee on the Orientation of Buildings. Royal Institute of British
Architects, London.
Rollier, A. (1914). La Cure de Soleil. Payot, Lausanne.
Rollier, A. (1927). Heliotherapy - with Special Consideration of Surgical
Tuberculosis. Second Edition, Oxford Medical Publications, London.
Rollier, A. (1944). Ouarante Ans d'H61ioth6rapie. University of Lausanne,
Lausanne.
Saleeby, C.W. (1929). Sunlight and Health. Fifth Edition, Nisbet and Co., London.
Samuels, R., (1990). Solar efficient architecture and quality of life: the role
of daylight and sunlight in ecological and psychological well-being. In, A.A.M.
Sayigh (Ed.), Proc. 1st World Renewable Energy Congress. Reading, UK, Pergamon
Press, London, pp. 2653-2659.
Sourkes, T.L. (1966). Nobel Prizewinners in Medicine 1901-1965. Abelard-Schuman,
New York, pp. 16-19.
Venolia, C. (1988), Health, buildings and the sun. SunWorld, 12, 2, 48-51.
-------
3.17 Passive Cooling I
-------
Intentionally Blank Page
-------
3067
MODELLING A PASSIVE EVAPORATIVE COOLING TOWER
B. Givoni, UCLA
Graduate School of Architecture and Urban Planning,
UCLA, Los Angeles, USA
ABSTRACT
An experimental mathematical model has been developed,
calculating the performance of a passive evaporative cooling
tower as developed by Cunningham and Thompson (1986). The model
calculates the hourly tower exit air temperatures, flow rate and
the speed of the air exiting the tower.
INTRODUCTION
A passive evaporative air cooling system has been developed and
tested by Cunningham and Thompson (1986) in Tucson, Arizona. The
system consisted of a down-draft passive evaporative cooling
tower, attached to a building. At its top vertical wetted
cellulose pads, 10 cm. thick (CELdek), were installed. A plywood
"X" baffle at the top of the tower catches the wind and directs
it downward. Water is distributed at the top of the pads and is
collected at the bottom by a sump and recirculated by a pump.
A solar chimney was built at the opposite side of the building,
to enhance the cooled air flow through the cooling tower and the
building. Outdoor air flows down the cooling tower through the
building to the attic and then out through the solar chimney.
During the daytime the air is heated in the attic, which thus
serves as part of the solar chimney.
A description of this system and experimental data (in a tabular
form) of its performance during a test lasting two days (August
22-23, 1985) was published in Cunningham and Thompson (1986) . The
measurements included the ambient conditions (DBT, WBT and wind
speed), tower exit air speed and temperature, indoor air
temperature, solar chimney data, etc.
The performance of the system was very impressive. At 4 p.m. on
the second day, when the outdoor temperature reached a maximum
at 40.6 °C and the WBT 21.6 was °C, the tower exit air temperature
was' 23.9 °C and the indoor temperature only 24.6 °C.
Analysis of the test results of Cunningham and Thompson, Givoni
(1991) has demonstrated that in reality the solar chimney did
not affect the performance of the cooling tower and that the air
flow rate was practically independent of the ambient wind speed.
Thus it seemed possible to develop a simple mathematical model
of this cooling system and the indoor temperatures of a building
cooled by it (Givoni 1991). A short description of the model,
-------
3068
only the part calculating the performance of the tower is
presented in this paper.
ANALYSIS OF THE EXPERIMENTAL DATA
The Tower's Exit Air Temperature.
The temperature drop of the air exiting from the tower was first
expressed as a function of the WBT depression (DBT-WBT). A linear
relationship, with a slope of 0.87 of the regression line, was
determined. Consequently the temperature of the exit air (Texit)
was computed initially by the formula:
Texit = DBT - 0.87 * (DBT - WBT); °C (i)
This formula applies to an evaporative cooling system with pads
similar to those described above. This very simple formula
provides good correlation between the measured and computed
temperatures (standard error of estimate 0.2 °C). It would be
useful in predicting the temperature of the exit air in cases
when local wind speed data is not available.
Effect of the Wind Speed
To check the extent of the effect of the wind speed on the exit
air temperature, the deviations of the temperatures calculated
by formula 1 from the measured ones were expressed as a function
of the ambient wind speed. With increasing wind speed the
computed temperatures were progressively lower than the measured
ones (negative values of the deviations). The observed effect of
the wind speed (WS) on the deviation of the predicted from the
measured exit air temperature (DelWind), and the needed
correction could be approximated by the formula:
DelWind = - 0.3 + 0.27 * WS; (C)
Two hypotheses may explain this observed effect of the wind:
J'!116 "ind may increase the air flow rate and speed through the
pads and consequently the air would have less time to be cooled
£ if Pads* This. hypothesis was checked by analyzing the
direct effect of the wind speed on the tower's exit air speed and
was not supported by the experimental data (see later on)
b) The wind blowing past the pads, creates turbulence at the
Sthr» T the paas- drying th"» „hU*
rying the vapor away, and thus reducing the effective
tower At6 thisllnsta°aethe air Hflowin
-------
3069
temperature, including the wind effect, would therefore be:
Texit = DBT - 0.87* (DBT-WBT) - 0.3 + 0.27*WS; °C (3)
Figure 1 shows the diurnal patterns of the measured exit air
temperatures and that computed by the initial and the modified
formulae. The agreement between the measured temperatures and
temperatures computed by the expanded formula is a little better
I standard error of estimate: 0.1 °C) than with the temperatures
computed by formula (1), although the difference is quite small.
The expanded formula can be used where data on the local wind
speed (near the tower) are available.

\ \

<>»
/

k /

o 10 30 90 40
VIND VELOCITY M/S
MEASURED Texit 0 SHORT FORMULA x FULL FORMULA
Fig. 1: Exit Temperature (°c)
Effect of Heat Gain bv the Tower
The flowing air is cooled and acquires its lowest temperature
while passing through the wetted pads. While flowing down, it
gains heat by conduction through the Tower's walls and its
temperature rises. In the case of the Tucson tower, the
calculated elevation of the flowing air temperature was less than
0.2 °K. With a reasonable thermal resistance of the tower's
envelope and the flow rates which are generated during the hot
hours of the day, it can be assumed that the air temperature does
not change significantly while flowing down the tower.
Consequently the tower's air temperature is practically
determined only by the ambient DBT and WBT and the type and
thickness of the pads (the length of the air flow through them).
Limiting the model to pads of the same type as those used in the
study of Cunningham and Thompson (1986) enables the use of the
formulae derived from the experimental data for estimating the
exit air temperature as general ones for a tower with this type
of pads.
THE TOWER'S EXIT AIR SPEED
Effect of WBT Depression.
Theoretically, the thermosyphonic flow and the exit air speed in
an evaporative cooling tower are proportional to the square root
of the temperature difference between the air columns, outdoors
-------
3070
and within the tower. This difference, in turn, is proportional
to the WBT depression times the square root of the tower's height
(in this case 7.6 meters). The exit air speed is the flow rate
divided by the tower's cross section area (in the present case
3.2 m2) .
Experimental formulae were derived on the basis of the
experimental data, expressing the flow rate and the exit air
speeds first as a function of the WBT depression. The formulae
thus obtained were:
Flow = 0.03*Aevap * sqrt(H*(DBT-WBT)); (4)
ExitSpeed = Flow / Atower; (m/sec) (5)
where:
Flow = The tower's air flow rate; (mA3/sec)
ExitSpeed = Tower's exit air speed; (m/sec)
Aevap = Area of the wetted pads; (m2)
Atower = Cross area of the tower; (m2)
Figure 2 shows the measured exit air speeds and the speeds
calculated by formula # 5.
J
II
11
II
12
I
I
U
1
¦ KAsiifD — mm
Fig. 2: Exit Air Speed (m/s)
Effect of the Solar Chimney on the Exit Air Speed.
The temperature elevation of the air while flowing through the
solar chimney was used as a measure of its "activity". To see the
effect of the solar chimney on the flow rate, the deviation of
the exit air speed computed by formula # 5 from the measured
speeds, was expressed as a function of the temperature elevation
in the chimney. The range of these deviations was very small
(from -0.1 to +0.1 m/s.), without a clear direction. It can,
therefore, be suggested that the effect of the chimney on the
flow rate can practically be disregarded.
-------
3071
Effect of the Wind on the Exit Air Speed.
The deviation of the exit air speed computed by formula # 5, from
the measured speeds, was expressed as a function of the measured
wind speed. There was no significant effect of the wind on the
exit air speed (the regression "r" value was -0.03). Consequently
also the effect of the wind on the tower exit air speed can be
disregarded. Thus the flow rate and the exit air speed can be
considered as a function of the temperature depression and the
tower's design details alone, as is expressed by formula # 5.
REFERENCES
Cunningham, W. A. and T.L. Thompson (1986): "Passive Cooling with
Natural Draft Cooling Towers in Combination with Solar
Chimneys". Proc. Passive and Low Energy Architecture (PLEA^.
Pecs, Hungary.
Cunningham, W. A., G. V. Mignon and T. L. Thompson (1987):
"Establish Feasibility for Providing Passive Cooling with Solar
Updraft and Evaporative Downdraft Chimneys". The Environmental
Research Lab. University of Arizona, Tucson.
Givoni, B. (1991): "A Simple Model of a Passive Evaporative
Cooling Tower". Submitted to Solar Energy.
-------
3072
ERRORS AND ASSETS IN OUTDOOR MIST COOLING
Jeffrey Cook, Regents Professor
College of Architecture and Environmental Design
Arizona State University
Tempe, Arizona 85287-1605
USA
ABSTRACT
The recent technical availability and the popular applications of outdoor evaporative cooling devices
are reviewed and measured on site in the hot dry desert of the Phoenix metropolitan area. The
building type of public hostelry includes restaurants and cafes that provide seating and service
outdoors. Sample manufacturers claims are examined and shown as being unreasonable.
Effectivness is further compromised by installation. The net measured cooling effect was between
3° and 6° F (2° to 4°C). However, there are positive psychological as well as visual effects that are
reinforced by the variability aspects of the experience. Although this elementary technology has
considerable potential it is not well demonstrated by installations examined.
KEYWORDS
Evaporative cooling; fog, artificial; microclimatic modification; mist cooling; outdoor cooling;
passive cooling.
INTRODUCTION
While passive, hybrid and active strategies for interior space cooling have a broad range of
demonstrated applications, the passive cooling techniques of outdoor spaces have had little
attention until recently. In the hot dry conditions of desert summers the use of artificial fog or the
mechanical misting of water has become prevalent within the last five years to condition outdoor
spaces.
Visible watery vapor suspended in the air has several names. In physics true "fog" droplets have a
diameter of 5 to 50 microns. "Mist" is larger, between 50 and 100 microns, although this is also
the dimension of some natural low lying fog. Typically natural clouds are in the middle of the
"fog" definition of 10 to 20 microns. Raindrops are 400 to 1,000 microns in diameter. Droplets
below 5 microns disappear almost instantly into water vapor which is invisible.
Misting systems are assembled from components commercially available nationally by installers in
desert cities such as Phoenix. One of several manufacturers of the critical hardware is Bete.
Another is Mee Industries, Inc. of El Monte, California who advertise themselves as "The Fog
People" and have trademarked "Fog Scape" for "environmental designs, animal and people
cooling, special effects and tropical effects." However, the industry and the technique is usually
referred to as "misting."
Sample Manufacturer Claims
"Imagine" "Imagine and dream about the nicest, most ideal environment you can. Imagine the
ability to enhance your current environment to achieve that dream. Imagine a Garden of Eden.
Imagine fog".
-------
3073
"The engineers at did. Then they invented a machine that makes it all possible. A machine
that can change blistering desert atmosphere to that of a pleasant seaside resort. A machine that can
cool a backyard patio, or a tennis court, or a whole golf course. A machine that can cool a hot
greenhouse and make plants grow like never before, or put roots on cuttings, or make seeds
germinate. A machine that can create a cloud that drifts and swirls around patrons and plants in an
arboretum rain forest. A machine that can blanket hundreds of acres of farm field to protect crops
from freezing on cold nights".
"Outdoor Air Conditioning" "On a hot summer day, outdoor air temperature can be cooled to the
low seventies - even though the temperature normally would have soared into the hundreds".
"The fog machine atomizes water into droplets so tiny they stay suspended in air. On hot days
the droplets totally evaporate. This cools the surrounding air by as much as 40 degrees. And fog
cleans the air and eliminates smog and dust".
"Resorts, public parks, and private residences use fog to cool patios, poolsides, outdoor eating
areas, tennis courts, and even golf courses".
"Special Effects" " fog droplets are so small they stay suspended in air just like nature's
clouds. This can create remarkable effects in landscapes, amusement parks, fountains, laser light
shows, and discos. fog is used extensively in the motion picture industry".
"Many variations of fog effects can be created. The system can be controlled to create a thick,
billowy cloud, constantly moving and changing shape; or to create a thin, wispy fog. It can be
used both indoors and out".
"Freeze Protection" "On frosty winter evenings fog droplets will not evaporate, but stay
suspended in air, forming a dense low lying fog. This fog can be used to blanket crops and
prevent frost damage. The fog blanket acts just like a wool blanket. It conserves heat, and keeps
crops warm".
"Industrial Applications" "Fog is used in many industrial processes for precise control of humidity
and temperature".
"Fog is used for static electricity control, air washing, pollution control, chemical spraying, dust
suppression, and for humidifying storage facilities. Fog is used for concrete during and other
curing processes".
"Fog is also used to cool large shop areas. Fog is much more energy efficient and cost effective
than conventional methods".
"Greenhouse Climate Control" "There are thousands of fog systems in use in commercial
greenhouses all over the world. Fog is used to cool, humidify, and reduce stress to plants. Fog is
used to propagate cuttings and to germinate seeds. Fog cooling is far more effective than other
cooling methods".
"The key factor in greenhouse use is the very tiny droplet size of fog. If the droplets are too
big, as with micromist systems, then plants get wet and disease occurs. The system generates
a true fog droplet, five times smaller than high pressure mist systems".
"Growers using fog report remarkable increases in plant growth and quality".
"Fog is also used in mushroom operations, to achieve the ideal humidity necessary to produce
quality mushrooms".
Observations about Manufacturers
Various manufacturers make competing claims about droplet size, operating pressure, efficiency,
controls, wetting, maintenance and the appearance of a fog or mist or smoke or steam as well as
visibility and invisibility. Although the physics of misting is well known, the mechanical design of
-------
3074
systems is typically by the manufacturer or the installer since virtually every outdoor location is
unique. Among the first design questions is whether the system is primarily for cooling or for
visual effect.
Manufacturer advertisements typically do not include methods of calculation, or certified
measurements made by independent testing organizations. They do make a few quantified claims.
For instance, to get an evaporative cooling effect "as much as 40° F', or "to get to the low 70's (F°)
from the 100's" (F°).
The difference between dry bulb and wet bulb air temperature, called the wet bulb depression, is
the design basis for calculating fan driven evaporative coolers (see table 1). The expectation is
typically that such mechanical wetting of air is between 70% and 90% efficient. With a 40° wet
bulb depression this would bring the temperature down for the cooler output between 28° and 36°
below dry bulb. It would be exceptional to get a mechanical evaporative cooler with a 40° cooling
advantage. Similar standards apply to misting systems. Although physically possible it would
require complete saturation, a difficult feat without precipitation, and especially in the partially
uncontrolled environment of outdoors.
Dry Bulb (F°) Relative Humidity Wet Bulb (F°) Wet Bulb Depression
110° 10% 67° 43°
105° 14% 68° 37°
105° 8% 65° 40°
100° 10% 63° 37°
Table 1. The Cooling Potential and Wet Bulb Depression as shown on the Psychrometric chart
Part of this difficulty with an evaporative misting system is because misting nozzles are point
sources. Thus, though they operate at high efficiency in atomizing water particles into the air, they
do not present a continuous moistened surface to the dry air to be cooled. Thus in hot dry
locations the pads of evaporative coolers are high performers. However, in more humid locations
mechanical misting can be more efficient since pressured nozzles can force additional water vapor
into the air in contrast to moist pads which do not project droplets into the air. Unfortunately,
typical details of the installation of misting nozzles examined tend to dissipate rather than
concentrate the remarkable performance of the nozzles.
SYSTEMS
To attempt microclimatic modification thereby cooling the outdoors for human uses, misting
systems typically consist of horizontal pipes of water under higher pressure than city line pressure
(300 to 1000 psi rather than 50 psi.). Some systems modulate performance by varying the
pressure such as between 300 and 500 psi - below 300 psi one gets droplets that are too large and
thus precipitate. Other systems are modulated by adding or reducing the number of nozzles by
using solonoid valves on clusters of nozzles. Both manual and thermostat controls are used.
Pipes are mounted outdoors along building surfaces, or on the edges of canvas and light metal
canopies at a height of 6' 8" (2 meters) or higher. The pipes have specially designed spray nozzles
mounted every 8" to 24" (200 to 600 mm) that provide a continuous mist when the system is
operating. The spray nozzles atomize water vapor into droplets as small as 1/10 the diameter of a
human hair. In absolutely still air the artificial fog will descend and dissipate by vaproizing into
water vapor and thereby cooling by evaporation. In the slightest breeze, which is sometimes
induced by the cooling effect of the mist itself, the fog may move in any direction and will dissipate
much more quickly. The fog has the appearance of smoke from a rapidly burning flame.
(Occasionally fire departments are alerted.)
For atmospheric effects and for the humidification of vegetation the misting system may be
installed close to the ground, hidden in planting or arranged off of large tree trunks, such as in the
indoor tropical conservatory of the Myriad Gartens at Oldahoma City, or the atrium of the Mirage
-------
3075
Hotel on the Strip at Las Vegas. Misting systems may also be placed in or near fountains to
exaggerate waterfall effects and are often used for special effects in movies. Misting systems are in
abundance in various Seville pavilions of Expo '92 both for cooling and visual effects.
STANDARDS
Questions about water use, energy needed and cooling effect usually bring ambiguous answers, in
part because outdoor conditions are largely open ended and ill defined. According to one industry
leader a typical rule is that 30 nozzles will deliver one gallon per minute at 1000 psi and require a
3/4 horsepower pump that consumes 3/4 kilowatt of electricity. Each nozzle cools a net of 14,000
cu. ft. (378 m2) of air by 1 degree (F) in 1 minute; 14,000 cu. ft. is 20' x 20' x 35' (378m2 is 6 x
6 x 10.5 meters) (Theoretical is about 16,000 cu.ft). Typically in such a system with nozzles on a
line every 2 feet (60 cm) the cone shaped plumes will overlap within 3 feet (90 cm).
LOCATION
This investigation concerns the use of misting to temper outdoor conditions for the hospitality
industry in a hot desert climate. Tempe, a city adjacent to Phoenix, has a characteristic desert
conditions at an elevation of 1100 feet.
Observations have been made on days when conditions were consistent. Temperature
measurements have been taken in the shade at table top height 30" (750 mm) at restaurants and
cafes that are identified by their tradenames (see Table 2). All are located within a block of each
other. The same instruments were moved from one location to another. Installations are by
different manufacturers. The term "ambient" refers to outdoor dry bulb temperature nearby which
is presumably not affected by the local conditioning of the artificial fog. In addition to the
continuously operating misting systems over the table top locations, conditions were also observed
near a large fountain with a five minute preprogrammed alternating fountain cycle that operated
continuously in the courtyard of the Coffee Plantation where misting is a one minute part of the
clock fixed cycle.
OBSERVATIONS
Location Large Fountain Coffee Plantation Chili's Asphalt Parking Lot
on 5 minute cycle
Time 1:50 pm 2:05 pm 2:20 pm 2:35 pm
Ave. in Fog 92° F (33.3° Q 90° (32.2° C) 92°(33°C)
Lowest 74° F (23.3° Q 81° (27.2° C) 88° (31.1° C)
Ambient 95°F(35.0°C) 95° (35.0° C) 95° (35.0° C) 98° (36.7° C)
Ave. Advantage 3°F(1.7°C) 5°(2.8°C) 3°( 1.7° C) 6°( 3.4° C)
Table 2. Dry Bulb Temperatures and Average Cooling Advantage 25 Sept 1990
Although the measurements reported were made on an almost calm day there was sufficient air
motion to changeably and continuously move away the effect of the cooling mist and its
temperature advantage. These were characteristic conditions at all locations examined. A
thermocouple thermometer with an instant read-out was used with continuous readings to provide
averaged values. Only sensible temperatures were measured. The "lowest" reading represents the
maximum temperature advantage caused by misting.
It can be seen that the average cooling advantage was between 3° F and 6° F (2° and 4° C)but that
there could be brief experiences at a table of up to 14° F (7.8° C) of cooling advantage. Near the
large fountain for a brief period the cooling advantage could beuptoa21°F(11.7°Q temperature
drop within a fog cloud. These larger cooling effects represent the level of continuous cooling that
are the standard expectation when provided by a fan driven evaporative cooler under similar but
enclosed conditions. It should be noted that historically various drive-in fast food locations in the
Phoenix area often provided outdoor seating covered by a roof with evaporative coolers.
-------
3076
ANALYSES
In general, misting installations appear to be casual and incidental, and not designed to provide
maximum cooling benefit. The intended microclimatic advantages are constantly changing, making
scientific measurement difficult but also providing constantly changing human conditions. In use,
the mist and its immediately cooled adjacent air is allowed to drift away even if there is almost no
air motion. When there is die slightest breeze the cooling advantage of the mist is immediately lost.
This could be modified by providing some environmental enclosure whether temporary such as the
use of canvas or plastic fabric, or more permanent such as rigid walling or skirting materials. One
restaurant owner was well aware that these aspects could improve his system performance.
In addition the misting nozzles have lineal patterns that do not coincide with seating patterns. They
also do not provide a complete coverage over the outdoor area that is intended to be used. Thus
doubling or quadrupling the nozzles would multiply effectiveness even more, because dilution and
drifting would be less perceived. Typically corners of canopies have the greatest variability.
Either mist becomes so concentrated it starts to precipitate or pool, or the effects are too thin and
dissipate.
The effect of these misting installations is not the continuous delivery of outdoor cooling, but an
occasional or intermittent human experience. This does have the advantage of allowing the human
subjects to dry out, to avoid wet hair, skin or clothing and to relish the variation of an outdoor
experience that would otherwise be continuously hot, dry and well above the comfort zone. In
general the fog is "dry" until it vaporizes which is the moment when latent cooling is provided.
Simultaneously humans in a hot climate are continuously perspiring. The drift of mist conditioned
air away from the immediate seating area can also serve as a lure to attract pedestrians by when
sidewalks or plazas nearby get the benefit. Thus what may appear to be errors in design or
installation may still provide certain advantages to the business.
From a corrective point of view the use of architectural containment to restrain the drift of
conditioned air, as well as the placement of the mist nozzles and the use of a more concentrated
density could add greatly to effectiveness at little or no additional cost
Other Effects: Reports from newspaper interviews add experiential and psychological dimensions
to the evaporative misting systems:
1. "This is a fun new technology. There is a lot of interest, surprise and disbelief.
2. "People think they're going to get wet, but they don't."
3. "Some customers only want to sit outside 'after we have finished water our plants."'
In the management of a restaurant where controls are manual, regular customers take the liberty of
requesting the system to be turned on or off. Occasionally the smell of chlorine is experienced
because of the excessive quantity in Tempe water. This is not attractive for food or beverage
service.
The owner of one of the restaurants requested that his landlord install a system by his observation
that a number of local fashionable hostelries already had such systems, and that it would extend the
outdoor season. After one year he is "semi-satisfied." Obviously it did not totally match his
expectations.
CONCLUSIONS
The installations examined are not optimal by any standards. In general it may be concluded that
outdoor cooling by misting is primarily a visual and an aesthetic experience. It therefore represents
an economic opportunity as a business investment, or incentive for outdoor activities including the
hospitality industry. In spite of the impressive theoretical potentials as well as manufacturer claims
for outdoor cooling through evaporation, the installations examined did not produce fundamental
and dependable comfort of major dimensions, but did produce a significant psychological
reinforcement to the suggestive visual and thermal effects.
-------
3077
THE IMPACT OF MOISTURE UP TAKE FROM EVAPORATIVE COOLERS ON
THE COOLING LOADS OF RESIDENTIAL BUILDINGS
Hofu Wu, Arch.D., AIA
Department of Architecture, California State Polytechnic University
Pomona, California 91768-4048, USA
ABSTRACT
The moisture issue from direct and two-stage evaporative cooling systems is being investigated
using an existing desert residence. The monitoring includes moisture absorbtion by various home
furnishing materials and the energy consumptions of evaporative cooler and air-conditioning unit.
Alternating operations of a conventional heat-pump A.C. unit and an evaporative cooling system
were conducted to study the actual hourly energy use. The cooling load fluctuations due to the
moisture content of building material are compared between the use of evaporative coolers and the
A.C. unit. The results show insignificant load increase for the switching from evaporative coolers
to heat-pump A.C. system. The moisture absorbed by furnishings do not pose significant
problems in the cooling load for the residence in hot, arid climate.
KEYWORDS
Passive cooling; Evaporative cooling; Moisture up-take; Cooling energy loads; Residential
Building; Moisture absorbtion and desorbtion; Home furnishings.
INTRODUCTION
In arid climate, the use of water evaporation to ameliorate the comfort in living spaces has been
widely used. The process of evaporative cooling takes the advantages of waters high heat of
fusion and the rather abundant and low cost of water resources. Therefore the energy needed for
cooling in an evaporative cooler is far less than the conventional compressor A. C. unit. Recent
developments of two-stage and indirect evaporative cooling with add-on heat exchanger provide
greater promises for indoor comfort of this cooling process. But still the big disadvantage of
evaporative cooling systems is the high humidity content of the supply air. Concerns have been
raised that.ihe use of nighttime natural ventilation may bring in excessive moisture and increase
the cooling latent load for the next day. Likewise, the moisture up-take by furnishings from using
evaporative cooling systems may also inadvertently increase the following day's cooling energy
loads. Most commonly used of home furnishing materials such as: carpet, fabric drapes, furniture,
and wood, etc. are potential absorbers for moisture in the air.
During most of the warm, dry summer days, the large quantity of air exchange from the
evaporative cooler does not present difficulty in maintaining human comfort. The moist supply air
may be more welcomed by the occupants. But later in the summer, due to the "Monsoon" season,
the high level of moisture in the air can impair the efficiency of evaporative cooling systems. This
added humidity from evaporative cooler can cause discomfort for the occupants. Several auxiliary
systems such as: ceiling fans or auxiliary fans may extend the comfort zone by providing higher
air flow rate around the human body, although there is a limitation to the maximum air speed
acceptable. In addition to the issue of comfort, buildingwith evaporative cooling systems tends to
encourage the formation of fungi, mold and algae which could deteriorate the quality of the living
-------
3078
space. When house is furnished with both evaporative cooler and conventional A.C. system, the
user can alternate the operation of both systems in order to reduce the cooling energy cost and to
provide adequate indoor comfort. During this time, the only possible usage of evaporative cooler
will be in the evening and early morning hours, and the rest of the day will rely on the compressor
cooling system. From the utility company's point of view, this situation does not improve the sale
of electricity but only to increase peak demand loads on the power generation.
MONITORING SETUP
This research monitored the magnitude of energy demand and energy consumption of an existing
house and the indoor and outdoor climatic conditions along with the changes of moisture content in
some home furnishing sampling materials. Modified psychrometers (with both DB, and WB)
made out of thermal couples with distilled water, wick and fans were used to record the
environmental conditions. Weight cells were used to measure the changes in weight of home
furnishing samples. Watt transducers were used to record the energy usage of both heat-pump
A.C. unit and the direct and two-stage evaporative cooling system.
MOISTURE ABSORBED BY FURNISHINGS
Although the majority of the moisture supplied through an evaporative cooler is being exhausted to
the outside, the remaining water vapor may be absorbed by the house furnishings. In turn, the
absorbed moisture would be detrimental to the cooling energy load if the occupant switched back
and forth between an evaporative cooler and a conventional refrigeration A.C. unit. Weight cells
were installed to monitor and record the moisture content of several furnishing samples inside a test
house in Tempe, Arizona along with the nearby climatic conditions. Figure 1 shows the amount of
water moisture absorbed by the carpet sample along with specific modes in system operation. One
can observe the extraction of moisture from the carpet sample through those days when the heat-
pump were in operation. The increase of moisture content in the sample was obvious when the
direct evaporative cooler was used. Due to the small increment of moisture content in the sample
by the hourly interval, only daily weight changes were plotted. The contrary result on Oct. 12th
where the reversed extraction of moisture happened maybe due to the weather being cool. In that
day the direct evaporative cooling system was only in operation for a few hours and the moisture
inside the sample was drying out (evaporation by itself) during off period.
I) o
S -0.1
Oct. 4 Oct. 5 Oct. 6 Oct 7 Oct 8 Oct. 9 Oct. 10 Oct. 11 Oct. 12 Oct. 13 Oct. 14 Oct. 15 Oct. 16
Date
E E
& 2.
E *
as
Fig. 1. Amount of water absorbed by the carpet sample during monitoring period.
-------
3079
OPERATION BETWEEN EVAPORATIVE COOLERS AND A. C. UNIT
Typically, homeowners will operate the direct evaporative cooling system at the time when the
outdoor air temperature is low and the humidity is acceptable. This will ensure the indoor thermal
condition to be comfortable. Then they would only use the conventional A.C. system during the
peak demand hours when the outdoor condition turns to the worst. Unless a multi-stage
evaporative cooler is used, there are times that the direct evaporative cooling system can not fulfill
the comfort need during peak demand hours. For this study, four different days were operated
with four different operating schedules and their energy impacts were recorded for comparison.
COOLING ENERGY LOADS WITH VARIOUS OPERATION MODES
Alternate between Direct Evaporative Cooler and Heat-Pump A.C. Unit
Figure 2 and 3 show the hourly cooling loads of the test house when the cooling system was
alternating between the direct evaporative cooling system and the heat-pump AC unit. In Figure 2,
the evaporative cooler was running during the nighttime and early morning hours (from 10:00 pm
till 10:00 am). But the heat-pump unit was operating between 10:00 am and 10:00 pm. In Figure
3, the reversed operating schedule was used, where the direct evaporative cooler was operating
during the daytime. The peak energy demand of the heat-pump (4.7 KW, in Fig. 2) was four
times higher than the peak demand of the direct evaporative cooler (1.2 KW, in Fig 3). The daily
consumption in Figure 2 was 44 kW h, and for the daily energy consumption in Figure 3 was 30
kwh. The daily energy consumption showed a difference of 46%. This operating schedule was
in accordance with the utility company's window of peak hours. Examining the daily moisture
absorbtion in Figure 1, one can observe the migration of moisture through the furnishings on the
day when the evaporative cooling system was in operation and the extraction of moisture on the
day when the heat-pump system was in operation.
5000
4000
J000
2000 -"
Evas
H.P
liiii.
0 ; 2 3 4 5 6 7 8 9 '0 I 112 13 14 15 !6 17 !
Hour
j _ I _
2021 2223
Fig. 2. Hourly energy profile using direct evaporative cooler in off-peak hours and heat-pump in
peak hours.
-------
3080
2000
1500
1000
500
I Direct
0 H P
0 I 2 3 4 5 6 7 8 9 10111213141516171819 20 21 22 23
Hour
Fig. 3. Hourly energy profile using heat-pump in off-peak hours and direct evaporative cooler in
peak hours.
Alternate between Two-Stage Evaporative Cooler and Heat-Pump A.C. Unit
Figure4 and 5 show the hourly cooling loads of the test house when the cooling system was
alternating between the two-stage evaporative cooling system and the heat-pump system.
Whenever the heat-pump is in operation, the hourly demand went up to 2.5 kW, but for the two-
stage cooler the demand was less than 500 Watts. The daily energy consumption showed a 29% in
difference. The alternating operation of these two cooling systems were in accordance with the
utilities window of peak hours. There was a dramatic change in moisture content of the carpet
sample between these two days.
CONCLUSIONS AND RECOMMENDATIONS
The use of direct or two-stage evaporative cooling systems may unduly penalize the energy load
requirement, when switching from an evaporative cooler to a heat-pump system. Though the
recorded data did show higher energy load for operating heat-pump system right after the use of
evaporative cooling system, the peak load did not occurr until nighttime. In comparing the
direct and two-stage evaporative coolers, the amount of water moisture input from direct system is
much greater than a two-stage evaporative cooling system. However, the use of evaporative cooler
at night did not pose a great deal of energy penalty on a typical heat pump system's load.
Certainly, if one uses evaporative coolers during the day and Switches to heat-pump at night (or
during off-peak hours), then the peak demand load would be lowered, and the energy benefits can
be easily ascertained. Unfortunately, this type of operation usually can not satisfy the indoor
thermal comfort during peak summer days.
Moisture adsorbed by furnishings did not pose significant penalty on conventional A.C. system
due to the self evaporation in off hours. The trace of moisture was insignificant from the use of
evaporative coolers and the imposed load penalty was not conclusive. Future
investigation on the quantity of moisture input to the space, and documentation on the increase of
energy required for the removal of moisture by conventional air conditioning is needed. This
inconclusive statement should be validated again during peak "Monsoon" season.
-------
3081
5000
4000
3000
2000
1000
¦ 2-Stage
0 HP.
0 I 2 3 4 5 6 7 8 9 IC I II 2 I 3 U 15 I 6 1 / 16 19 20 21 22 23
Hour
Fig. 4. Hourly energy profile using two-stage evaporative cooler in off-peak hours and heat-pump
in peak hours.
2000
1500
1000
500
¦ 2-Stage
E3 H.P.
0 I 2 3 4 5 6 7 8 9 1CI11213I415I61718IS2C212223
Hour
Fig. 5. Hourly energy profile using heat-pump in off-peak hours and two-stage evaporative cooler
in peak hours.
-------
3082
ACKNOWLEDGEMENT
The author would like to express special appreciation to Mr. Marvin Gorelick form U.S. DOE for
his support and foresighted vision on the research direction. Appreciation also due to graduate
assistants; Jung-Ho Huh and Dale Johnson for their patience and diligent work during the
monitoring and reporting periods.
REFERENCES
Wu, H. & Yellott, J. I. (1987). "Investigation of A Plate-Type Indirect Evaporative Cooling
System for Residences in Hot and Arid Climates," ASHRAE Transaction, V.93, Pt.l, NY 87-12-
2.
Wu, H., (1989). "Performance Monitoring of A Two-Stage Evaporative Cooler", ASHRAE
Transaction, V.95, Pt.l, Chicago.
Byrne, S.J. and Ritschard, R. L. (1985). "A Parametric Analysis of Thermal Mass in Residential
Buildings", ASHRAE/DOE/BTECC. Thermal Performance of the Exterior Envelopes of Buildings
III, Clearwater Beach, Florida.
West, D. (1984). "Evaluation: Indirect/Direct Evaporative Cooling for Residential Use," Salt
River Project, Arizona.
ASHRAE (1989). ASHRAE handbook - 1989 fundamentals, Atlanta, American Society of
Heating, Refrigeration and Air-Conditioning Engineers.
-------
3083
PASSIVE QOOLING OF AN ADMINISTRATIVE BUILDING IN SION
(SWITZERLAND^
C.Angay (1), O.Guisan (2), B.Lachal (2), R.Meldem (3), W.Weber (2)
1) Etat du Valais, D6partement des Travaux Publics, Service des BStiments, CH-1950 Sion
2) CUEPE, University of Geneva, CH-1231 Conches
3) Meldem Energie SA, Consulting Engineers, CH-1920 Martigny
Center for Building Science, Lawrence Berkeley Laboratory, Berkeley, CA 94720
ABSTRACT.
This study presents the analysis of a passive cooling plant installed in an office building. This paper
discusses its performance and technology. A comparison with a traditional air conditioning system
highlights the energy conservation potential and the economical benefit of the passive cooling system.
KEYWORDS.
Passive Cooling; Administrative Building; Energy Conservation; Technology; Modeling.
INTRODUCTION. (1)
The main purpose of this study was to examine the technological, economic and energy saving
advantages and disadvantages of a passive cooling system installed in an administrative building in
comparison with a traditional air conditioning system. This building, located in the City of Sion which lies in
the Rh6ne Valley in the Swiss Alps, was built in the 19th century. In 1989 it was renovated and converted
into offices. In summer, the outside air temperature often reaches 30°C and, although the building was
properly insulated, the top floor, consisting of garrets under a slated roof, could reach more than 30°C. It
was therefore decided to install a cooling system for this floor. The air used to cool the floor was brought
from the wide and hollow cellar of the building. Another purpose of this study was to gain a better
understanding of passive cooling systems well suited to the Swiss climate in order to help engineers and
architects design more energy efficient buildings.
PLANT DESCRIPTION.
Figure 2.1 shows a cross section of the
building with the cooling system. The
characteristics of the plant are summarized in
Table 2.1 below. The data from the hottest office
located in the west side was gathered to study
the performance of the system.
An important parameter of this kind of plant
was the Radon concentration, which was
measured before the decision was made to cool
the building. The concentration was found to be
below the maximum admitted value. New
measurements were taken after the system's
installation in the cellar and in offices of the top
floor. Contrary to our expectation, Radon
concentration in the offices was higher (51
Bq/m3) than the cellar's concentration (24
Bq/m3)-
(?)
Fig 2.1 Cross section of the building.
-------
3084
TABLE 1. Main Characteristics of the Plant
Cooling Source: Cellar
Depth : 10 [m]. underground
Volume : 300 [m3]
Walls+Ground Area: 500 [m2l
Temperature Range : 18-20 [°C]
Air Distribution Network
Fan:
Prated output: 900 [W], 1600 [m3/h]
Pmeasured : 400 [W], 1000 [m3/h]
Ducts :
Before fan : l=8 [m]
After fan : l=23 [m]
steel, not insulated
Cooled Area : 200 rr#
West office data :
Wall insulation : 0.10 [m]
Roof insulation : 0.16 [m]
Ground area : 17.5 [m*l
Volume: 44.5 [m3]
Windows : double pane
Skylight window area: 1.1 fm2]
Shell window area : 0.4 [m^l
Heat capacity storage : 1500 [Wh/K]
Changing air rate : 5 [1/h]
Internal loads : 78 [W] (1 PC on 24IV24)
1 -2 employees
Monitored 24h/24h
MONITORING (3)
During the summer of 1989, temperatures and other data were recorded in order to characterize
the building without the cooling system. The passive cooling plant was completed during the spring of
1990. We began monitoring the building to evaluate the cooling performance in July 1990. The data
consisted of hourly measurements on 38 channels including humidity, solar radiation, air flow rate,
electrical power, temperature and many other discrete measurements.
RESULTS (A)
The two figures below show the measured performance of the passive cooling plant. Figure 4.1
shows the air temperature shapes during the two hottest weeks of 1990 as well as the homogeneity of the
inside air temperatures while cooling was on. For a weekend the cooling was turned off to confirm its
effects and an immediate stratification of the temperatures appeared. The difference in temperature
between the floor and ceiling was due to the solar gain, as one can see by comparing the solar radiation
curve with the temperature curves. Figure 4.2 shows the mean air temperature frequency in the offices
during 1989 (without cooling) and during the hottest summer week in 1990 (with approximately 70 hours
without cooling, cf fig 4.1). In the summer of 1989, which was cooler than summer 1990, the west office
was unoccupied for most of the month of August.
30.000
25 000

Toff*
fa

h
J\
A

\ !

If
i\
\i
"A

fl#*t
V
vl

V
coot
NG O
"I

V
\l

•*
I

A
l\
A

A
'i
71
1
f\
/1
I
I
I
I
I
I
t\
l\
iliil
il
&• 7- a- 9- 10- II- 12- 13- 14- IS- 16- 17- 18 19
Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug Aug
Comfortabtal-
Fig 4.1 Temperatures in the west office and climate Fig 4.2 Frequency of the mean temperature in the
characteristics. west office in 1989 and 1990.
-------
3085
The comfort improvement is evident. While testing and manipulating the cooling system it became
obvious that its performance could be improved and that some of the design solutions commonly used in
ventilation, and applied here, were not suitable. We modeled the whole system to determine which
parameters were the most influential on the system's efficiency.
We divided the cooling plant into three subsystems that were modeled and checked independently.
Then, we composed a very simple but reliable dynamic model for the entire system which gives hourly the
mean air temperature in the west office.
Analysis of the Office
The first subsystem we studied was the west office. For this we used a simplified method solving the
power balance on one node only. This approach required us to determine the physics of the office very
precisely. Used parameters were :
Qref : Cooling power [W]
Gs : Solar gain transposed on the roof [W/m2]
Aeff : Effective area receiving solar radiation [m2]
Qint : Internal loads [W]
Ub : Heat transfer coefficient of the office [W/m2K]
Tint : Inside temperature [°C]
Text : Outside temperature [°C]
Cb : Heat capacity of the office [Wh/K]
The power balance of the office can be expressed as follows :
Qref + Gs'Aeff + Qint - Ub'(Tint-Text) = Cb*dTint/dt (5.1)
We first determine the parameters of the west office, then neglecting the capacity effect in (5.1), we obtain
the well-known H/M diagram, given in Figure 5.1 and applied in daily value.
Qref + Gs*Aeff + Qint - Ub*(Tint-Text) = 0 (5.2), or
H =Ub -M'Aeff (5.3), with
M = Gs/(Tint-Text) represents a meteorological factor.
H = (Qref+Qint)/(Tint-Text) represents the cool (or heat) demand.
The good correlation allows us to use a linear regression. The slope of the line closely represents with
a precision of 95% the effective capture area of solar radiation: 0.31 [m2] < Aeff < 0.46 [m2] with skylight
window blind half closed. The vertical axis intercept represents the U value of the building with a high
imprecision interval: 1.9 [W/K] < Ub < 27 [W/K].
The equivalent solar radiation capture area was determined for the whole office. It is interesting to
determine how this value is divided between the roof, the shell window and the roof window, which are
the solar gains hunters. The influence of those elements obtained through power balances on each of
them are given in Table 5.1 in equivalent areas in [m2]:
MODELING OF THE SYSTEM
(5)
Aeff - 0.39 m2
Ub = 14.0W/K-
H (W/K)
600
Figure 5.1 H/M Diaqram of the office.
-------
3086
TABLE 2. Equivalent Solar Radiation Area of Each Important Solar Capturing

Roof
Roof window
Shell window
Total
Blind opened
0.1
0.4
0.1
0.6
Blind half opened
0.1
0.2
0.1
0.4
Blind closed
0.1
0.1
0.1
0.3
We are now able to build a simplified dynamic model for the office's temperature from the equation (5.1)
with good results.
Tint(h+1) = Tint(h) + (Qref(h) +Gs(h)*Aeff + Qint(h) - Ub*(Tint(h)-Text(h))}/Cb (5.4)
Analysis of the Cellar
An accurate and detailed description of the heat transfer in the cellar is very complex and was not
the purpose of this study. The measurements showed that the exchange area was large enough to
smooth the daily outside air temperature fluctuations. We therefore described the outside air cooling
process as a mix of two air flows, one at the storage temperature and one at the daily averaged outside air
temperature. Averaging each day's data, it was possible to determine a mixing factor which characterizes
the heat transfer in the cellar.
Mf = (T ec-T ext)/(Ts-T ext) (5.5)
where Mf is the mixing factor, which is particular to this cooling source, Tec the temperature of the air
exiting the cellar, Ts the storage temperature and Text the outside air temperature. The temperature
difference between the model and the measurements is less than 0.5°C.
The Air Distribution Network
The last element we modeled was the air duct distribution network with the fan. The modeling of
this part was long and was specific to the design of the plant. We investigated this part carefully to
determine the causes of the losses. The significant sources of losses were the fan, the heat transfer
through the duct walls, the air rubbing on the wall and the infiltrations. Their importance depended on the
mass flow rate. Possible improvements for the air distribution network are discussed in the next section.
DISCUSSION (6)
This section describes some technological conclusions drawn from studying the passive cooling
system and compares it with a traditional air conditioning system applied to this building.
Technology
The feasibility of this passive cooling system depends first on the availability of a cooling source
and a sufficient exchange area and second on an efficient use of the cool air. When the first condition
above is achieved, designers must optimize the air distribution system. In this particular case, because of
the design options chosen, improvements up to 50% could be achieved. These options were :
1) Suppression of air infiltrations in the depression zone, which can be achieved without extra costs, if
executed carefully.
2) Fan engine located out of the duct to avoid cooling the motor with the cool air.
3) An accurate fan rating with optimal efficiency in the range of utilization.
4) Insulated ducts.
The cooling is most efficient when the losses are minimized. This occurs in the studied case, not for the
highest air flow, which corresponds to a fan supply voltage of 380 [V], but for 260 [V], Figure 6.1 shows
the mean inside air temperature frequency in the office as a function of the voltage of the fan and August
1990 climate. This figure shows that optimal cooling conditions depend not only on the cool air flow but
also on losses relative to it. While the optimal cooling is achieved with a fan voltage of 260 [V], the most
energy efficient cooling occurs at a voltage of 160 [V], We used Figure 6.1 to explain to the employees
how to use the control system more efficiently. Instead of running the system with a voltage of 380 [V] as
they previously did, they now use it with a voltage of 100 [V], and raise it to 260 [V] depending on the
inside air temperature.
-------
3087
36
34
32
/- 100 v
30
28
380 V
/ 280 V
26
24
22
40 60 eo
% OF THE TIME WHERE THE TEMPERATURE IS OVER
0
10
20
30
70
80
100
90
Fig 6.1 West office mean simulated temperatures frequency with different voltages, August 1990 climate.
Energy Conservation
The secondary purpose of this work was to check the energy consumption of a passive cooling
system and to compare it with the consumption of a mechanical cooling system. We rated a traditional
cooling plant for the 4th floor, simulated its performance and compared it to the passive cooling plant's
performance. The assumptions for the modeling of the mechanical cooling plant were:
1) A COP of 2.2 for the air conditioning unit.
2) The system turned on if the blown air temperature was less than the office temperature and the air
temperature of the office was over 25°C.
3) Due to the architecture, no air recycling was possible.
The results of the comparison between both systems are shown in both Table 6.1 and Figure 6.2.
Fan Voltaqe [VI
0
100
160
260
380 I A/C
Mean Temperature [°C1
31.3
25.9
25.4
24.7
24.8 S 24.9
Mean Temperature of the 5% hottest hours [°C1
34.6
28.1
27.3
26.5
26.6
26.7
Mean Electrical Power Required [W/m2l
0
0.4
0.7
1
2
4
Monthly Electrical Enerqy Consumption fkWh/m2l
0
0.3
0.5
0.7
1.5
3
Table 6.1 Comparison of the air temperature in the west office and the electrical power required to reach
this temperature as a function of the fan voltage, August 1990.
passive coolingj-
i§ mean temperature in the
west office
O electrical power per m*
19.0 I ¦"""I"""' i "in ) ¦¦ t Hffl.
fan voltage —~380 V 260 V 160 V 100 V
Fig. 6.2 Comparison between the passive cooling and the air conditioning plant, August 1990.
-------
3088
Costs
The first cost of this passive cooling system was $ 7,000 and those of a traditional air conditioning
system were approximated to be $ 14,000. The passive cooling plant is 50% cheaper than the traditional
air conditioning system. The operation costs are proportional to the consumed energy and are 4 times
higher for the traditional cooling plant than for the passive cooling system. These numbers, coupled with
the comfort provided by the passive cooling system, show that there are efficient alternatives to heavy
energy consumer systems.
CONCLUSIONS
In order to avoid overheating in buildings, architectural measures and diminution of unnecessary
internal loads must be taken into account. The behavior of the occupants and the way they manage
overheating sources is also very important. If, despite the arguments above, cooling is still necessary,
designers must know that passive cooling plants can fulfill these needs with relatively small investments
and operating costs. In some cases loads are so great that only mechanical cooling can assure appropriate
comfort conditions. But, as shown through the example of this building, many offices buildings could have
a passive cooling plant successfully installed with lower costs, fewer negative environmental effects and
the same comfort level as provided by a traditional cooling system. This study shows that it is possible to
reach the same, and even better, comfort conditions by using a passive cooling system requiring half of
the investment and less than 25% of the operating costs of a traditional air conditfoning°plant.
ACKNOWLEDGMENTS
Work for this paper was highly facilitated by the financial support of the "D6partement de I'Energie
de I'Etat du Valais, Sion". We also want to thank for their kindness the collaborators of both the "Service
des Bdtiments" and the "D6partement de I'lnstruction PubHque".
REFERENCE
R.Meldem, B.Lachal, C.Angay, W.Weber, O.Guisan, Rapport final: Climatisation passive du bdtiment
Aymon £ Sbn, to be published May 1991.
-------
3089
CAPACniVE AND RESISTIVE INSULATION IN ARCHITECTURE FOR NAIROBI- KENYA
Daniel K.Irurah and John S. Reynolds,
Department of Architecture,
University of Oregon
Eugene, OR 97403
U. S. A.
ABSTRACT
This study carries out a comparative analysis on the effect of high thermal mass construction technol-
ogy using either natural stone, bricks or concrete blocks, and lightweight construction technology
using frame and infill of similar materials as the previous technology but at smaller thicknesses, for
Nairobi's climate. The analysis evaluates the internal surface temperature cycle for external walls of
different orientations and thicknesses in response to outdoor surface temperatures and based on ther-
mophysical properties of time lag and decrement factor of wall or roof components. The rhythm of
internal surface temperature regime for external walls is compared with typical indoor air temperature
regimes and comfort needs to assess periods of heat gain or loss from indoor air as well as the need
for, and optimum location of resistive insulation on external walls.
KEYWORDS
Capacitive insulation, sandwiched resistive insulation, time lag, phase shift, decrement factor, sol air
temperature, lightweight construction, high mass construction.
1.0 INTRODUCTION
The diurnal temperature range for Nairobi averages about 12 C° . This is about four times the range
between the mean temperature of the hottest and the coldest months- the annual range- which is about
3 C° . A bioclimatic analysis of Nairobi's climate done by Irurah and Reynolds (1990) indicates a
need for space heating throughout the day during the cool season and between late evening to mid
morning during the warm season, with radiant space heating ranging between 100 to 800 W/m^ (30
to 50 Btu/hft^). The bioclimatic analysis also indicated that there is no major requirement for cooling
either by natural or mechanical ventilation if the resultant building maintains or enhances the natural
thermal conditions because temperature and humidity rarely rise beyond the comfort levels.
These observations indicate that regulation of the diurnal temperature cycle, rather than the seasonal
cycle, is more critical in passive thermal control in Nairobi. The use of thermal mass and conventional
insulation- Capacitive and Resistive Insulation, respectively, as termed by Koenigsberger and co-
authors (1973), is therefore a highly appropriate strategy for passive thermal control in Nairobi.
This study carries out a comparative analysis on the effect of high thermal mass cosntruction technol-
ogy using either natural stone, bricks or concrete blocks, and lightweight construction technology
using frame and infill of similar materials as the previous technology but at smaller thicknesses. The
analysis evaluates the internal surface temperature cycle for external walls of different orientations and
thicknesses in response to outdoor surface temperatures and based on thermophysical properties of
time lag and decrement factor of wall or roof components used. The rhythm of internal surface tem-
perature regime for external walls is compared with typical indoor air temperature regimes and com-
fort needs to assess periods of heat gain or loss from indoor air as well as the need for, and optimum
location of resistive insulation on external walls.
The study indicates that capacitive insulation is optimum on the west and south facades while sand-
-------
3090
wiched resistive insulation with lightweight construction would be ideal on the east and north facades.
2.0 EXTERNAL AND INTERNAL SURFACE TEMPERATURE CYCLE
2.1 West Facade
The internal surface temperature cycle of west facing wall through a 24 hour period is established for
February 21 and July 21- die hottest and coldest months respectively- on the simplifying assumption
that internal surface temperature is solely influenced by external surface temperature and thermophysi-
cal properties of the wall components. Interaction between internal surface temperature and indoor air
temperature cycles is then considered in section 2.3
Four wall thicknesses are considered in the analysis; SO, 150,200, and 250 mm. Thicknesses of 50 -
150 mm approximately correspond to the frame-infill lightweight construction, while thicknesses of
150 - 250 mm corresponds to high mass construction. Glazing as an infill in lightweight construction
is not considered because its thermal storage is negligible. However the influence of glazing on indoor
air temperature cycle and the heat gain or loss behavior of the building is considered in assessing the
need for resistive insulation.
Fig. 1(a) and (b) show the outdoor air temperature and the external surface temperature cycles for
walls of different orientations and horizontal surfaces like flat roofs. Fig 2(a) and (b) show the out-
door air temperature and internal surface temperature cycle for a west facing wall of natural stone,
brick or concrete blocks (the materials have similar thermal physical properties of density, absorptiv-
ity, emissivity and surface conductance). Sol air temperature is used instead of outdoor air tempera-
ture for the period in which the wall is exposed to direct solar radiation- approximately from 12.00 to
18.00 Hrs. for the west facade- and such that;
Tsa = T0 +1 x a Eq. (1)
fo
where Tsa = Sol air temperature (°C),
To = Outdoor air temperature (°C),
I = Intensity of Solar Radiation on vertical
Surface (W/m^),
a = Absorptivity of Radiation by Surface
(0.65 used for materials considered)
fQ = Surface Conductance (W/m^ C°).
Internal surface temperature is determined using the equation;
Tsi = mTso + u (Ti - mTso) Eq- (2)
where Tsi = Internal surface temperature (°C),
mTso = Mean external surface temperature
including sol air temperature where
applicable (°C),
u = Decrement factor for the wall thickness
considered. This is the proportion by
which external surface temperature is
reduced as the heat transfers through
the wall, relative to the mean external
surface temperature.
Tl = External surface temperature "1" hours
before, where "1" is the time lag in Hrs.
for the wall thickness considered;
including sol air temperature where
applicable. (°C)
-------
3091
Horizontal Surface
K
W. Facade

S. Facade
E. Facade
Outdoor Temperature _
00 00
Time (Hrs.)
12.00
24.00 00 00
Horizontal Surface
N. Facade
E. Facade
12.00
24.00
Time (Hrs.)
Comfort Zone
—o—o— Mean Outdoor Temperature
(a) February — Warm (b) July — Cold
Fig. 1 Diurnal and sol air temperature curves for walls and and roof in Nairobi.
2.2 Daily and Seasonal Analysis
A comparative analysis of internal surface temperature regime and thermal comfort requirements
shows that a strategic design would aim at shifting the peak time of internal surface temperature
(phase shift) for all facades further away from the period of high solar radiation (9.00 to 17.00 Hrs.)
as the ambient air temperature is within or close to comfort zone for that period. The design would
also aim at dampening the relatively high sol air temperature on roofs, the east and west facades
(decrement effect). Fig. 2 (a) and (b) indicate that the two objectives can be realized for west facades
through wall thicknesses of over 200 mm. High mass construction usually falls within this thickness
range. The time lag of over 6.5 Hrs. means that peaking in internal surface temperature is delayed till
about midnight when outdoor air temperature would have fallen well below comfort level. A decre-
ment factor of 0.3 or less generates a more stable internal surface temperature regime. Given adequate
proportions of such surface areas in relation to the floor area of the building, there would be a stabiliz-
ing effect on indoor air temperature regime.
Wall thicknesses of less than 200 mm appear to be poorly suited for the above strategy as they only
achieve a phase shift of less than 6.0 Hrs and dampening effect of more than 0.5. Lightweight con-
struction falls within this range of thickness and performance.
2.3 Effect on Indoor Air Temperature
In Eq. (2) it is assumed that indoor surface temperature of an external wall is solely dependent on ex-
ternal surface temperature and thermopysical properties of the wall components. This is only true if
the indoor air temperature is maintained at a level close to the mean outdcwr air temperature. In build-
ings with no mechanical heating or cooling, indoor air temperature would be higher during the day
and lower during the night following fluctuations in heat output from internal sources like lighting,
user metabolism and cooking, and direct solar radiation through unshaded glazing. Fig. 3(a) and (b)
indicate typical indoor air temperature regimes for commercial and residential buildings (Febniary and
July respectively) based on typical use patterns for each category and assuming no direct solar radia-
tion or mechanical thermal control.
-------
3092
Time Lag (150mm)
i
Time Lag (50mm)- v
150mm
(feM
50mm
*
v
50mm (July^;
150mm (July)
Time Lag (250mm)
Time Lag (200mm> —v
250mm (Feb.)
200mm
V'yTT \
' 200mm (Jul y)\
>250mm (July)
12.00
00.00
Time (Hrs.)
¦«— External Surface Temperature (Feb.)
(a) 50 and 150 mm wall thicknesses
24.00 00.00 12.00
24.00
Time (Hrs.)
External Surface Temperature (July)
(b) 200 and 250 mm wall thicknesses
Fig. 2 External and internal surface temperature for west facade at various thicknesses
A comparison of temperature regimes in Fig. 2 (a) and Fig.3 (a) and (b) shows that daytime internal
surface temperature regime for lightweight construction closely matches that of indoor air temperature
regime for wall thicknesses of 50 to 150 mm. During the warm day hours the building rapidly gains
heat through such external walls and roof, and during the cool night hours the building rapidly looses
heat to the colder outdoor air. While resistive insulation may be used to reduce the rate of heat gain or
loss in such a building, opportunity for thermal storage and phase shift through time lag and the stabi-
lization of indoor air temperature through the decrement effect would be lost
The regimes in Fig. 2(b) and Fig. 3(a) and (b) show a shift in regime in internal surface temperature
of high thermal mass and indoor air temperature regime. The internal surface temperature is low when
indoor air temperature is high- during the day hours- and vice versa during the night hours. The
thermal mass would be gaining heat from indoor air during the day thus moderating the rate of internal
heat gain and thus indoor air temperature. During the cool nights the internal surface temperature of
the mass gradually gets higher than the internal air temperature. This allows for dissipation of heat to
the indoor air and hence an upward moderation of indoor air temperature. This would be a highly op-
timum strategy during the warm season. During the cool season, outdoor air temperature remains be-
low the comfort level almost throughout the day and hence the retention of internally generated heat
inside the building becomes desirable for indoor thermal comfort.
This suggests a need for seasonal internal resistive insulation on external walls in order to reduce the
rate of heat loss from indoor air to the large thermal mass and subsequently to outdoor air during the
cool nights and the cold season. Fig. 4 shows the effect of resistive insulation on the decrement factor
if placed on the internal or external surface of the wall asumwing heat flow from outside to inside.
When direction of heat flow is reversed (from inside to outside) an internal resistive insulation would
be most optimum.
-------
3093
ortZone
Comfort Zone
Mean Outdoor Temperature

^Residential
i i
Commercial
loli
Residential
mmercial
24.00 00.00
(b) July
00.00
Time (Hrs.)
(a) February
Fig. 3 Typical indoor air temperature.
24.00
Time (Hrs.)
la ted Internally
i 1
sulated Externally
200
Thickness (mm)
Fig. 4 Effect of internal and external insulation en decrement factor
Source: Markus and Morris (1980) p. 323.
Cost and technical difficulties of protecting the internal or external insulation may necessitate the use
of sandwiched resistive insulation. Air space in walls would be particularly optimum. Givoni (1976)
observes that providing an air space on thick walls can increase time lag values to more than 8.0 Hrs.
and reduce the decrement factor values to less than 0.1 without increasing wall thicknesses beyond
those of high mass construction range analysed in this study. Walls with an air space provide an
advantage in that they regulate heat flow in either direction. This is of particular significance for
the persistently cool nights through out the year and for the cool days and nights of the cold season.
-------
3094
3.0 OTHER FACADES AND ROOFS
A similar analysis of the east facade and roofs indicates that time lags of over 10.00 Hrs. would be
necessary to achieve an optimum phase shift The thicknesses of walls would then go beyond 300
mm and may prove to be uneconomical in terms of space lost as well as financial cost For the east fa-
cades and roofs therefore, a combination of lightweight construction with resistive insulation seems to
be an appropriate strategy.
As observed in section 2.3, the sandwiched resistive insulation is likely to be the most feasible alter-
native. Sandwiched resistive insulation between roof cover and die ceiling is not a common practice in
Nairobi,although it appears to be highly desirable. The use of aerated hollow concrete blocks with an
air cavity seems to be highly effective on the east facade.
Irurah and Reynolds also observed that the east facade could best serve as space heater through direct
solar radiation for the early to mid morning period,given appropriately designed glazed surfaces and
sunshading, while the west facade would best serve as thermal mass to provide heat in the cool night
hours. Such design strategy heavily depends on strategically optimised solar access on both facades.
As Fig.l (a) and (b) indicates, the external surface temperature on the north and south facades is not
very critical. This is because of the latitudinal location of Nairobi close to the equator (1°, 14' S).
These facades could be designed in such a way as to supplement the west and the east facades. The
south facade would best be treated as the west facade as it receives high levels of insolation during the
hot hours of the warm season. The north facade receives its highest levels of insolation in the warmer
hours of the cool season. Such facades could therefore be used to supplement the east facade in space
heating through user controlled solar radiation. Use of double glazing would minimize the rate of heat
loss during the cool night hours or the cool season.
4.0 CONCLUSION.
A comparative analysis of high mass and lightweight construction technologies has been done to eval-
uate their performance in passive thermal control in Nairobi's climate using time lag and decrement
factor approach. Phase shift of internal surface temperature regime in relation to external surface and
outdoor air temperature regimes has been used as an indicator of appropriateness of the two tech-
nologies. The study indicates that high thermal mass— capacitive insulation— of brickwork, natural
stone or concrete blocks, 200 to 250 mm thick is most appropriate on west facades. Integration of
sandwiched resistive insulation, like air space, is also desirable. Lightweight construction with sand-
wiched resistive insulation is most appropriate for east facade and roofs where it serves the double
purpose of dampening the impact of high daytime external surface temperatures and minimizes heat
loss from the building during the cool night hours and the cold season.
The north and south facades are relatively weak facades in passive thermal control due to the equato-
rial location of Nairobi. The south facade could be best designed to supplement the role of the west
facade while the north facade could be best designed to supplement the role of the east facade.
5.0 ACKNOWLEDGEMENTS
The authors gratefully acknowledge the Institute for International Education for the Fulbright grant
which supports Daniel Irurah's studies at the University of Oregon.
6.0 REFERENCES
Givoni, B. (1976). Man. Climate and Architecture. Applied Science Publishers Ltd., London.
Irurah, D. K., and J. S. Reynolds. (1990). Architecture and Solar Energy in Inter-Tropical
Regions. In S. M. Burley and M.J. Coleman (Ed.), 15th Passive Solar Conference
Proceedings. American Solar Energy Society, Boulder, Colorado.
Koenigsberger, H. and co-authors, (1973). Manual of Tropical Housing Part 1: Climatic Design.
Commonwealth Printing Press Ltd., Hong Kong.
Maikus, T. A., and E.N. Morris. (1980).Buildings. Climate and Energy. Pitman Publishing Ltd.,
London.
-------
3095
THERMAL PERFORMANCE OF A ROOM COUPLED TO AN
EVAPORATIVE COOLING TOWER
M. S. Sodha, R. L. Sawhney, J. Kaur and
S. P. Singh
Canter of Energy Studies and Research
Devi Ahilya Vishwavidhyalaya
Indore- 452 001 (INDIA)
ABSTRACT
The thermal performance of a building fitted with an evaporative
cooling tower has been evaluated in terms of discomfort index for
hot-dry climate of Jodhpur. The effect of various evaporativeIy
cooling parameters (height and cross-sectional area of the tower,
packing factor, area of the pads, resistance offered to air flow)
and local wind conditions on the thermal performance of the
building has been analysed in terms of discomfort index. It is
found that for given other parameters of the tower and wind
conditions, there is an optimum height of the tower for which
thermal discomfort condition in the building is minimum. The
optimum values of the tower height for comfort conditions in the
building for various other tower parameters have been obtained.
KEYWORDS
Passive cooling! cooling tower; discomfort index; evaporative
cooling; Stack effect
INTRODUCTION
Wind towers (Baud-Gears) have been used in hot arid climates for
natural ventilation and passive cooling of buildings. Cunnigham
and Thomson (1986) suggested that the performance of these towers
can be improved by providing evaporative cooling pads at the top
of the tower. In this study, the thermal performance of a non
air-conditioned building coupled to the evaporative cooling tower
has been analysed. Tropical summer index (TS1), (Sharma and Ali,
1986) has been used as the comfort index for Indian tropical
climatic conditions. The effectiveness of the tower is measured
in terms of "discomfort index" of the building (in degree hours)
as defind below.
The analysis is made for one-room building of size 12m x 6m x
2.7m, provided with two windows (3.6m1 each), in north and south
wall and two doors (2.6mi each), one in east wall and other in
north wall. For periodic nature of the environment parameters and
finite heat capacity of the walls, floor and roof, modified
admittance method of Sodha and co-workers (1986) has been used to
calculate conductive gain through walls, roof and floor.
Considering conductive, convective and radiative gains through
all the components of the building and air flow through cooling
-------
3096
tower to the room, the hourly values of the room air temperature
are calculated to obtain discomfort index for the entire summer
season.
ANALYSIS
When the ambient air passes through the evaporative pads, the
temperature T^ of the air coming out of the pads is given by
(Sawhney and co-workers, 1987)
T - T
t w
T - T
a w
= exp (-h_ A F / m )
r D p p a.
where Ap is the pad area, Fj> i3 the packing factor,
mass flow rate of the air through pads, and h^
transfer coefficient and is given by (Berman, 1963).
i s
(5.50 + 1
,65 v )
P
10
-3
Kg/m Sec
(1)
is the
the mass
(2)
where vp is the velocity of the air passing through the pads.
The velocity of air, v^. inside the tower, which depends upon
pressure difference (Wind effect) and density variations
effect), can be calculated as
-------
3097
TSI = 0.308 T„,t + 0.745 T„ - 2.06 v 1/2 + 0.841 (6)
RW R r
where is the wet bulb temperature of the room air and can be
calculated from the known values of Tj and obtained above.
and Vy are the dry bulb temperature and velocity of the room air
respectively. The discomfort index (Dl) in the room for any
period can be calculated from the relation
Per i od
(Dl) . = f (TSI - 27.5) dt (7)
period )
RESULTS AND DISCUSSION
The thermal performance the a building is evaluated for the hot-
dry climate of Jodhpur. The climatic data of this region, for
each of the required months of the cooling season (April-
September) is taken from the handbook of solar radiation data for
India (Mani and Rangarajan, 1980).
Fig. 1 shows the hourly variation of T4 , Tt , TR, Tw, TS1, and TR
(temperature of the same building without tower) for given tower
parameters. It is seen from this figure that the tower is quite
effective in cooling the air sinking through the tower and in
reducing the room air temperature in the comfort range. And for
the humid air supplied by the tower TSI values at any hour are
higher than the room air temperatures.
u
40
36
32
23
24
20
16
0
Fig. 1. Hourly variation of Tft , Tt , TR , Tw, TS I and TR
for At= lma , Ap = 2.5m*, Fp = 300, f=2.0
Fig. 2 shows the variation of monthly Dl for each month of the
cooling season (Deshmukh and co-workers, 1990) for different
combinations of tower parameters.
For given tower parameters, the seasonal DI can be obtained by
-------
3098
the area under the curve for the
shows the variation of seasonal
of A .p and At- It is seen
corresponding parameters. Fig. 3
DI with Fp for different values
that seasonal DI decreases

Fp
o
o
2

Fp
100. At
2

Fp
300. At
1 C nn2

Fp
200, At
20m?
*1
— — Fp
100. At
1 0m>
1
Fp
200. A.
1 5m
t 2000
MONTH
Fig. 2. Monthly variation of DI for different
values of Fp and At for f=2.0, h=6.0
20000
16000
150 200
VALUE OF Fp
Fig. 3. Variation of seasonal DI with Ff for different
values of Ap and A^ (f=2.0, h=6.0>
assymptotically with Fp . For a given value of tower parameters
and packing factor the DI is higher for smaller values of A-p and
the difference between the seasonal DI for different values of
pad area decreases with increasing value of Fp .
The effect of tower height on the thermal performance of the
building for various combinations of other tower parameters is
shown In Fig. U. It Is seen that seasonal DI first decreases with
height, attains a minimum value and then increases with height.
-------
3099
This suggests that for given other parameters of the tower and
the building (to which the tower is coupled) there is an optimum
height of the tower beyond which any increase in the height of
tower will not give any additional advantage in creating comfort
condition in the building. The optimum heights and residual
discomfort index for other tower parameters (f and F ) are shown
in the Table 1.
TABLE 1. Optimum Heights(m) & Residual Discomfort lndices(deg.hrs)
(1/f) values
F 0.2 0.3
0. 4
0.5
0,6
0. 7
0. 8
0.9
1.0
50 14369 14397
14387
14388
14387
14390
14387
14406
14440
C15.5) **(6.5)
(4.0)
(2.5)
(2.0)
(1.5)
(1.0)
(1.0)
(0.5)
100 9021
9022
9021
9022
9021
9031
9034
9021
(18.5)
(9.5)
(6.0)
(4.0)
(3. 0)
(2. 5)
(2.0)
(1.5)
150
6347
6347
6347
6347
6348
6347
6347

(15.5)
(10.0)
(7.0)
(5.0)
(4.0)
(3. 0)
(2.5)
200

4677
4677
4678
4678
4677
4677

(16.5)
(11.5)
(8.5)
(6.5)
(5.0)
(4.0)
250

3537
3537
3537
3537
3537

(17.0)
(12.0)
(9.5)
(7.5)
(6. 0)
300

2777
2777
2777

(15.0)
(1.5)
(9. 5)
350

2203
2203

(15.5)(
12.5)
*»Bracketted values are
opt imum
hei ght
s, here
Ap-2.5m f At—
lm1
15000
13000-
x 10000
o 7000
1000
0
21
16
20
8
12
I
t€IGHT (m)
Fig. 4.
Variation of seasonal D1 with h for different
values of Fp and f (At*l«a, Ap*2.5wl)
-------
3100
900C-
o 60 0C -
Fp : 300. At 1 5 tri
VALUE CP K
Fig. 5. Variation of seasonal Dl with k for different
values of F-p and At when Ap=2.5ral, f=2.0, h=6.0
The effect of winds on the thermal performance of the building is
shown in Fig. 5. It is seen from this figure that for the heights
of tower considered here (h=6m), winds in the region do not help
in improving the thermal comfort conditions in the building.
ACKNOWLEDGEMENTS
Financial support of Department of Non-conventiona1 Energy
Sources, Govt, of India, for carrying out this work is thankfully
acknow1 edged.
REFERENCES
Berman L. D. (1963). Evaporative air conditioning. The Industrial
Press, NY, USA.
Cunningham, W. A. and Thompson (1966). Passive cooling with
natural draft cooling towers in combination with solar
chimneys. PLEA Conference. Pecks.Hungary.
Mani, A. and S. Rangarajan (1980). Handbook of So 1ar Radiation
Data For India. Allied Publishers Pvt. Ltd., New Delhi, India.
Sawhney R. L., S. P. Singh, N. K. Bansal, and M. S. Sodha (1967).
Optimization of an evaporative cooler for space cooling. 1nt.
J. Housing Science And 1ts AppIications. 11(3). 225-231.
Sharma, M. R. and S. Ali (1966). Tropical Summer Index -a sudy
of thermal comfort on Indian subjects. Bui Iding And
Environment. 21(1). 11-24.
Singh, S. P., R. L. Sawhney, N. K, Bansal, and M. S. Sodha
(1967). Sizing of an evaporative cooler for thermal comfort
inside a room. Housing Science. 11(2). 141-146.
Sodha M. S., N.K. Bansal, P. K, Bansal, A. Kumar and M. A. S.
Malik (1986). Solar Passive Bui 1 ding. Pergamon Press, Oxford,
UK.
-------
3.18 Passive Cooling II
-------
Intentionally Blank Page
-------
3103
THE FLORIDA CRACKER STYLE:
A PASSIVE COOLING COMPROMISE
I. Melody, L. Maxwell and R. Vieira
Florida Solar Energy Center
Cape Canaveral, Florida
ABSTRACT
This paper discusses the passive cooling design elements of the Cracker
architectural style, which was typical in areas of the southeastern coastal U.S.
before the advent of mechanical air conditioning. It also presents three modern
Cracker Style home designs, which were winners of a design competition conducted
by the Florida Solar Energy Center. The homes combine a high level of energy
efficiency in hot, humid climates along with passive solar cooling concepts
typical of the Florida Cracker Style.
KEYWORDS
Passive solar cooling, vernacular architecture, home design.
BACKGROUND
No practical passive design solution has yet been implemented that can provide
complete comfort without air conditioning in hot, humid climates. The middle
region of the Florida peninsula illustrates the climate problem. Annual relative
humidity averages 75%. Annual average temperature is 72 degrees F, and summer
season temperatures average 80 degrees F. The region experiences 3226 cooling
degree days and only 733 heating degree days annually. Except during storms,
summer wind speeds are less than 10 miles per hour, and wind direction is highly
variable. Diurnal and seasonal temperature changes are slight, offering very
small heat sinks.
While passive solar design concepts may not yet provide total cooling in this
warm, humid climate, they can significantly reduce air conditioning loads,
increase comfort levels and provide air conditioning energy savings. In addition,
such design concepts can extend the length of seasonal swing periods (spring and
fall) during which air conditioning is not needed.
THE CRACKER STYLE
Before the advent of mechanical air conditioning, Florida settlers exploited
natural phenomena for thermal comfort. They surrounded their single-story homes
with wide-roofed verandas to shade walls and windows. Windows were designed to
open fully to admit even slight breezes; however, main window openings were
structurally protected from summer sun and seasonal rains. High interior ceilings
allowed for air stratification and frequently were topped with cupolas,
clerestories or gable windows. These allowed for daylighting without direct solar
gain and helped vent interior heat.
-------
3104
Homes designed with these concepts were said to follow "the Cracker Style" --a
descriptive if inelegant term that evolved from Florida settlers' main occupation:
cattle ranching. These ranchers used the "crack" of a whip to round up cattle,
and the onomatopoeic word became attached to the people and their home styles.
CRACKER STYLE DESIGN COMPETITION
The Florida Solar Energy Center conducted a design competition in 1990 under the
theme of "The Florida Cracker Style." It was the second in a series of annual
competitions funded by the Florida Governor's Energy Office. The goal of the
competition was two-fold: 1) Produce home designs that incorporate the Florida
vernacular style for both aesthetics and natural cooling, and 2) ensure that the
home designs offer a high level of energy efficiency even when they are
mechanically air conditioned. Three home categories were established based on
size and room requirements: 1000-square foot, two-bedroom, one-bath; 1750-square-
foot, three-bedroom, two-bath; and 2500 - square - foot, four-bedroom, two-bath.
The most important design requirements were those that related to passive cooling
and energy efficiency. They were as follows:
o Designs to be orientation-specific or independent of solar orientation,
o Permanent design features to shade all window areas from direct sun from May
through August,
o Ample natural ventilation for cooling in each room,
o Mechanical air conditioning in addition to passive cooling,
o Air handler and ducts in conditioned spaces,
o Washer, dryer and water heater in unconditioned areas,
o Shaded outdoor living areas.
o No knee walls between attic and conditioned space,
o No skylights or fireplaces.
o Attic ventilation via soffit, ridge, strip and/or gable vents,
o Radiant barriers in attic space over outdoor living areas and conditioned
rooms.
o Appropriate levels of insulation in floors, walls and attics,
o Ceiling fans in all indoor living and sleeping areas.
In addition, each submittal required completion of a energy-efficiency calculation
form based on the Florida Model Energy Code for Buildings. Required for all new
Florida homes, the calculation produces an energy performance index, which must
be lower than 100.
COMPETITION WINNERS
Of the 81 submissions, the following three designs merited Grand Awards.
Construction drawings for the winning designs are available from the Florida Solar
Energy Center.
The Cross Creek
Designed by William Wagner, AIA, of Gainesville, FL, the Cross Creek (Fig. 1)
combines energy efficiency and versatility. The EPI for this design was
calculated to be 43 points. With an entrance that opens to the east, the home
suits lots that face east or south. A mirror image of the design will also fit
west-facing sites. The judges noted that this design presented good ventilation
strategy and excellent shading, particularly for an east-facing home.
-------
3105
Fig. 1. The Cross Creek, by William Wagner, AIA, Gainesville, FL.
Figure 2 presents the Cross Creek floor plan. Verandas on two sides bring light
and fresh air into the living space while protecting walls and windows from direct
sun. Outdoor dining is a pleasant option on the screened porch adjacent to the
dining room. Windows on both sides of the kitchen cool occupants with natural
breezes. Windows on opposite walls of the master bedroom also provide maximum
ventilation during sleeping hours.
AA A
lining
Fig. 2. Floor plan for The Cross Creek.
-------
3106
The Suwannee
Designed by architect Ronald Haase of Gainesville, FL, the Suwannee is a two-story
Cracker Style (Fig. 3). Although it is designed to face north, it will also be
energy efficient if sited to the south. Judges noted that the design offers very
good natural ventilation along with excellent shading of the first floor.
Fig. 3. The Suwannee, By Ronald Haase, Gainesville, FL.
In winter, a wood stove helps heat the home, while windows that open to a wrap-
around veranda provide ventilation during summer. Even when air conditioning is
a must, the 48 point EPI rating of this home should help keep utility bills to a
minimum. The versatile back porch expands usable living space far beyond the 1700
square feet under air (Fig. 4) . A covered walkway provides protection from
frequent summer storms and anchors the detached garage to the home.
Fig. 4. Floor plan for The Suwannee.
-------
3107
The St. Johns
The St. Johns, designed by architect John Hall of Ormond Beach, FL, provides 1076
square feet of conditioned floor area. Like the other competition winners, it
combines excellent solar shading of the structure and good ventilation potential
through much of the building (Fig. 5). It has an EPI rating of 40.

Fig. 5. The St. Johns by John Hall, Ormond Beach, FL.
This design provides even more passively cooled outdoor living space than is at
first evident (Fig. 6). By incorporating the optional back garage door, the
garage becomes a breezeway workshop or childrens' play room.
r
v
L"Ti
A"
Living room
II
Y
DTt
V
Dining room
rp
y HV,
J1
Fi
n
r
aIoI
N
r
Fig. 6. Floor plan for the St. Johns.
-------
3108
CONCLUSION
While mechanical air conditioning is necessary in hot, humid climates, passive
solar design can reduce air conditioning loads, increase comfort, and extend the
comfort season. Design concepts developed for natural cooling often have their
roots in regional history. Such historical design elements can successfully be
adapted to improve the energy efficiency of modern homes, as is demonstrated by
the winners of the Florida Solar Energy Center's competition based on the Florida
Cracker Style.
ACKNOWLEDGEMENTS
Design competition judges: Walter Mauder, AIA; Fran Arnold; Lyle Fugleberg, AIA;
Gordon Mock. Cost estimator, Gary Cook. Contract monitor, Daryl O'Conner.
REFERENCES
Fairey, P. (1981). Passive Cooling and Human Comfort. Florida Solar Energy
Center.
Maxwell, L. (1990). 1990 Energy Efficient Florida Home Design Competition.
Florida Solar Energy Center.
Vieira, R. and K. Sheinkopf. (1988). Energy-Efficient Florida Home Building.
Florida Solar Energy Center.
-------
3109
AN ARCHITECT'S APPROACH TO PREDICTING THERMAL PERFORMANCE OF HOUSING IN
WARM REGIONS WITH REFERENCE TO NORTH AFRICA
Mohamed B. Gadi and Ian C. Ward
Building Science Unit
School of Architectural Studies
University of Sheffield
Sheffield S10 2TN, UK
ABSTRACT
The thermal approach presented in this paper was based on the results of a
thermal simulation for a typical North African house design. The simulation
was carried out after a survey of climatic data for eleven locations in the
region and a series of wind tunnel tests on a model scale of the same house.
The wind tunnel tests were aimed at measurement of wind pressures across the
model which were used to predict the hourly ventilation rates through the
building according to the local people's daily life patterns. The climatic and
ventilation data, together with the physical properties of common local building
materials and the expected daily internal heat gains, were then fed into a
validated computer programme to calculate the mean internal environmental
temperatures and the daily total solar heat gains through the building
components. The solar heat gains, per a square metre of component area, and
other relevant data, presented in graphical and tabulated forms, can be used
within any standard thermal model, such as the CIBSE model, to predict the mean
indoor temperatures for any North African or similar house form. A variety of
building orientations and local materials can be tested under summer or winter
conditions and according to calm or mean wind speeds. The work introduced in
this paper is part of a continuous research programme into a more general and
practical approach to design of housing in warm regions.
KEYWORDS
Design tool; thermal performance; natural ventilation; occupancy patterns; North
African housing.
INTRODUCTION
In many developing countries, housing design is still considered by most
building designers as a process of producing plans and elevations that can be
converted into cheap dwellings for people to live in —a process in which the
least or sometimes no regard is given to the influence of local climate on
buildings' thermal performance. One of the reasons behind this is the
increasing population in addition to economic difficulties. As a result, these
countries could not afford to build well equipped and staffed building research
centres where better and practical design standards could be developed to suit
the local environmental, social and economic conditions. This has led to
insufficient experience among the majority of building designers of the means
-------
3110
for utilizing potential resources through buildings' thermal design.
On the other hand, a number of thermal design methods have been produced in
developed parts of the world but, due to their complexities and the amount of
data and time they require, they are mostly dependent on computer facilities,
(Biondi and co-workers, 1988). In countries where computers and technical
software are rarely available for public use, such design tools could not be
helpful. Yet what architects in these countries need, at least until
developing countries overcome their economic obstacles and join the
industrialized world, is exactly what they can use. In other words, they need
a straightforward approach related to their specific environment. However, a
simple design tool of this type would assist architects to understand or
realize some aspects of thermal design and therefore introduce them to more
comprehensive design methods which would enable them to search for better
solutions to improve building s* thermal performance in their regions (Gadi and
co-workers, 1990a, 1990b).
THE CLIMATE OF NORTH AFRICA
The region of North Africa can be classified as a warm region, and its local
climate is influenced by three natural elements; the Mediterranean Sea, the
Great Sahara Desert and the Atlas Mountains, Fig. 1. Analysis of fifty years
of weather records for two groups of population centres (seven coastal and four
inland) has shown small variations in ambient temperatures and available solar
irradiance within each group, mainly due to differences in altitude and
distance from sea, (Gadi, 1989)-
Medit. Sea
mmm
W5
in
NORTH AFRICA
Fig. 1. Climatic zones and data locations.
I . Coastal zone. II . Inland (desert) zone. Ill. Mountainous zone.
1 . Melila. 2 . Algiers. 3 . Tunis. A.Tripoli. 5. Bengazi. 6.Sallum.
7. Port-saeed. 8.1nsalah. 9.Sabha. 10.El-kufra. 11. Aswan.
-------
3111
THERMAL SIMULATION
The present thermal simulation was carried out on a typical North African house
design using a validated computer programme based on the standard thermal
admittance procedure, (Sattler, 1986). Due to the small differences in ambient
temperatures and solar irradiance values within each of the two groups of
locations, mentioned earlier, only two average sets of climatic data were
considered, each representing a certain climatic zone in the region.
In order to simplify the approach, two very common local structural systems
were considered. A heavyweight structure consisting of a reinforced concrete
roof and sand-limestone walls. Load distribution per square metre of floor
area for this system was 1740 Kg/m2 . The second structural system was a light
to medium weight with a load distribution of 510 Kg/m2. It is made of a
hollow-clay block roof and hollow concrete block walls.
The hourly casual heat gains were estimated according to the daily life cycle
of the local residents and assuming an average family size of five persons.
The daily life cycle was considered to start with the breakfast period (700
w/h), then follows the house work with one person (140 w/h) . Around noon is the
time for preparing the traditional North African main meal (2380 w/h) followed
by a period of dining and relaxing for the whole family (575 w/h). The second,
lighter meal begins after sunset (1190 w/h) then a relaxation period (1075 w/h)
and finally bedtime (350 w/h).
In a region, like North Africa, with a generally warm climate, natural
ventilation plays a significant role in designing to achieve better indoor
environment, (Gadi and co-worker, 1990c). It is essential, for a more
realistic thermal assessment, to employ the actual ventilation rates related to
the building type and its users. There are several techniques for predicting
natural ventilation rates through buildings, of which model scale was found to
be more appropriate. In the present work, the same house was modelled to a
proper scale and tested inside a boundary layer wind tunnel. The model had
eleven cells and movable doors and windows which were opened and closed
throughout the experiment according to four main daily life patterns of the
local inhabitants, (Gadi and co-worker, 1990d)- Ventilation rates were
predicted from the wind pressure measurements which were taken inside and
outside the building model while it was mounted on the wind tunnel floor within
a simulated local wind environment.
A SIMPLE APPROACH
During the preliminary stage of housing design, it would be helpful for the
designer to be able to predict the indoor temperature of a particular dwelling
with a certain structure, envelope and orientation.
In the CIBSE Guide, (CIBSE, 1975) it is explained that the mean internal
environmental temperature can be calculated from the following equation:
Qt = !2AgUg + Cv'(tei " W + (tei " W (1)
where
Oj. = mean total solar and casual heat gain w
^A r area of glazed exposed surface m2
U ® thermal transmittance of glazing w/m2oC
= area of exposed opaque fabric m2
-------
3112
Uj, = thermal transmittance of exposed opaque fabric w/m2oC
C = ventilation loss w/°C
v
t . = mean internal environmental temperature °C
t = mean sol-air temperature °C
t = mean outdoor air temperature °C
ao
1/C = (1/0.33 N.V) + (1/4.8 A) (2)
A = total area of surfaces bounding the enclosure m2
N = number of air changes per hour Ac/h
V = volume of internal spaces m3
From a range of thermal simulations, t -values were calculated for both
climatic zones, Table 1. Equation (1) can be used to calculate the mean
internal environmental temperature for any similar building form, structural
systems and climatic conditions. The mean total solar heat gains through any
building component, and for any orientation,can be obtained from Fig. 2 for calm
and mean wind speeds. The mean casual heat gain was found to be around 726
w/h. The ventilation heat transfer can be obtained from Fig. 3 or equation (2).
Thermal properties of the building materials are presented in Table 2.
Table . 1 ¦ Environmental Data for Coastal (C) and Inland(I) Zones
z

Mean Sol-air
Temperature
teo

o
n

Mid January

Mid
July

tao, c
N,Ac/h
e

N
E
S
W
R
N
E
S
W
R
Jan.
July
Jan.
July
C
Calm
16
27
26
19
21
27
46
30
38
42
13
25
1
2
Windy
14
19
•19
15
17
25
28
27
29
31
8
10
I
Calm
15
15
22
16
21
33
53
36
46
54
11
31
1
2
Windy
13
12
17
14
16
31
35
32
36
38
6
8
Table.2.Thermal Properties of Building Materials
Material
S-heat
J/?: kg
U-value
W/m t
Limestone
0.25
m
900
5.0
Hollow Concrete Blocks
0.25
m
100
2.8
Hollow Clay Blocks
0.20
m
960
2.2
Reinforced Concrete Slab
0.15
m
840
9.3
Glazing

1840
4.7
-------
3113
A 00
300
200H
100

"/'y
/ /
y /
/
+ tt.
-9 V
coastal
mean
\ J *
\V/"
a*
zone,
wind.
' *-»t\ / ,
0^ y
2s

inland
mean
zone,
wind.

v~'~«
/ /
/ /
/ /
/

/ ~
'
\ \
\ \
\\
\ \
\\
' \
_D q-.-g"

nV
N V

O-
• «
0
© _
e n
90
0
0° 90° 0° 90° 0° 0* 90" 0'
NWSE N N W S E N
[* a -f b4-c4-d+-+-94-b-+- c •+- d H
900-
700-
>»
O
*U
5 300-
a
100-
M
coastal
calm
Inland
calm
zone
wind
zone
wind
500-
Fig. 2. Daily total solar heat gain
summer winter
x - heavyweight roof **-*"**
* wall —®—
mediumweight roof
wall
single glazing -b—«-
building orientation
-------
3114
10 987654 3 2 1 N,Ac/h
V.m3 400
300
15
20
-100
•200
600 400 200 0 200
Fig. 3. Ventilation heat transfer , w/c
400 A,m4
REFERENCES
Biondi, P., Salui, F. and Sciuto, S. (1988). Simplified Design Tools for
Mediterranean Climates. Proceedings of the sixth international PLEA
Conference, Porto, Portugal, Pergair.on Press.
CIBSE, (1975). Summertime Temperatures in Buildings. CIBSE Guide, Section A8,
The Chartered Institute of Building Services and Engineering, London, UK.
Gadi, Mohamed B. (1989). An Investigation into the Ventilation and Thermal
Performance of Contemporary Housing in North Africa. Proceedings of the 2nd
European Conference on Architecture. UNESCO, Paris, France, Kluwer Academic
Publishers.
Gadi, Mohamed B. and Ward, Ian C. (1990a). Passive Ventilation Performance
of a Composite Domed Roof within a Suburban Wind Environment. Proceedings of
the 8th PLEA Conference, Halifax, Canada.
Gadi, Mohamed B. and Ward Ian C. (1990b). Passive Utilization of Wind and
Solar Energy within a Composite Domed Structure for Promoting Natural
Ventilation in Warm Regions. Proceedings of the 1st World Renewable Energy
Congress, Vol. 4, Reading, UK, Pergamon Press.
Gadi, Mohamed B. and Murta, Kenneth H. (1990c). Passive Environmental
Aspects of Vernacular and Contemporary Housing in North Africa. Proceedings
of 8th PLEA Conference, Halifax, Canada.
Gadi, Mohamed B. and Ward, Ian C. (1990d). The Effects of Typical
Architectural Features and Occupancy Patterns on the Natural Ventilation
Performance of North African Housing. Proceedings of 1st World Renewable
Energy Congress, Vol. 4, Reading, UK, Pergamon Press.
Sattler, M.A. (1986). A Computer Programme for the Thermal Design of
Unconditioned Buildings. Internal Report, BS81, Department of Building
Science, University of Sheffield, Sheffield, UK.
Acknowledgement is made for the financial help provided by the Committee of
Vice-Chancellors and Principals of UK Universities as part of an 0RS Award to
the author, the technical assistance from the Building Science Unit of the
School of Architectural Studies at Sheffield University and the research data
supplied by Dr. Mohamed Muntasser, President of the International Energy
Foundation, and Dr. Mohamed Usta, Director of the Engineering Science Research
Centre in Tripoli.
ACKNOWLEDGEMENT
-------
3115
ON THE POTENTIAL AND EEKECTiVKNESS
OK PASSIVE NIGHT VENTILATION COOLING
Daniel Keuermann and Wendy Hawthorne"-*
Center for Energy and Environmental Physics
Jacob Blaustein Institute tor Desert Research
Ben-Gurion University of the Negev
Sde Boqer Campus, 84993 Israel
""currently with Alien Associates. Toronto, Ontario, Canada,
"supported by a grant from the Sally Berg Kam.il y Fund.
ABSTRACT
A simple approach to the evaluation of passive cooling by night
ventilation is proposed. The approach is motivated by the desire
to obtain tools that can be used in the initial stages of build-
ing design. A simple thermal network with lumped capacities is
employed to describe the thermal behavior of the building. Krom
the analytic solution to the governing equations of the thermal
network the effectiveness of night ventilation cooling is deter-
mined as a function of building parameters and ambient condi-
tions. The distribution of the frequency of occurence of the
ambient conditions is extracted from hourly meteorological dal.a.
facilitating the study of the long-term performance ol the night
ventilation cooling strategy. The present study is restricted to
non-aircondit.ioned buildings.
KEYWORDS
Night ventilation, cooling efficiency, potential cooling, passive
cooling, long-term performance.
INTRODUCTION
The field of ventilation cooling has been reviewed in depth by
Chandra (1989) and Clark (1989). There is apparently a need tor-
simple design tools that can guide engineers and architects in
their decisions regarding a structure early on in the design
stage. NiLes (198b) proposed a simple model for the prediction
of temperatures in a night ventilated building for a design day.
In this study we propose a link between the simple model predic-
tions for a given day and the long-term seasonal performance; ol
the night ventilation strategy.
Accurate predict ion of natural vent ilation is nearly impossible,
even with sophisticated computer simulations, as air i low pat-
terns in and around the building are quite unpredictable. In
addition, human behavior, the kind ol windows, design >>i and
partitions in the building, the immediate environment and the
microclimate all inlluence the actual performance ol natural
-------
3116
ventilation control strategies to a degree that is difficult to
assess. It seems therefore not unreasonable to study the effects
of night ventilation using simple models that include the physics
of the problem only to the first order. Based on the relative
success in employing lumped capacity models for studying passive-
ly heated solar buildings (Gordon and Zarmi, 1981; f'euermann,
Gordon and Zarmi, 1985), a similar approach to night ventilation
cooling was adopted. This was done with the understanding that
the models cannot accurately describe the thermal performance of
a building but that they permit the study of the building perfor-
mance sensitivity to heat transfer coefficients, to the location
and the distribution of thermal mass in the building and to
ventilation rates. While in the case of passive solar heating
the effect of climate was introduced via the distribution of the
daily solar to load ratio (the ratio of daily solar collectible
energy to daily building heating load), we use the distribution
of a daily potential cooling ratio, to be defined later on, which
is determined from hourly meteorological data. In the present
study we consider buildings without any air conditioning backup,
and ignore effects of humidity.
ambient
room
mass
W\A
R=l/U,
R= 1/U,
mass-ai r
R=l/U. ,
inf
fig. 1. Thermal network approximation for building.
THK MKTHOl)
Thermal Network
The simple thermal network depicted in Figure 1 describes a
building with insignificant thermal mass in the envelope but with
mass in the floor or partition wails. The mass .is assumed to
exchange heat at a rate of with the room air only.
Heat transfer through the external wails is designated by U,„.
Air infiltration is translated into a wail heat transfer coeffi-
cient Ui„, and is shown in parallel to the wall heat loss coeffi-
cient. Its value changes from day to night.
Other thermal networks could be used to describe buildings (Kabl,
1988), e.g., for the description of thermal mass in the envelope.
The approach however remains the same. If we approximate the
ambient conditions by a sinusoidal input function of ambient
temperatures, the solution to the governing equation describing
the thermal network is analytical. The boundary conditions of
the equations are formulated such that the solution satisfies the
quasi-steady-state constraint that the initial temperature of the
mass must equal its final temperature after a full cycle. The
night ventilation strategy entails increasing the air int'iJ.tra-
-------
3117
tion rate at night, i.e. during hours where the ambient tempera-
ture is below the room temperature. The effectiveness of the
strategy is expressed by the improvement of comfort which can be
measured by the reduction in degree hours of room temperature
(Troom) above a given comfort temperature (TcomT), namely:
CE = {I(T.oom-T=on)T)-w- I(T.oom-TcomT)-r)v>/X;(lVoom-TCom,)n;nV (1)
where ' + and 'w nv' or 'no nv' indicate summation of positive
values only, and with or without night ventilation, respectively.
CE (cooling efficiency) is a function of the ratio of the day to
night infiltration rate, of the mass in the structure, and the
heat transfer coefficients. The latter parameters may be lumped
into two time scales: the response times of the mass to changes
in the ambient temperature for daytime and nighttime, respective-
ly (Niles, 1986). CE is unity when the room temperature remains
below the comfort temperature after night ventilation is used.
Fig. 2 illustrates the thermal behavior of a building with para-
meters given in Table 1. Heat transfer coefficients are given in
per square meter of mass exposed to room air; doors and windows
are included in an average wall heat transfer coefficient.
TABLE 1. Parameters for Single Koom Bui iding.
geometry:
mass:
infiltration:
heat transfer:
length=4 m, width=4 m, heiight=2.5 m
in the floor, 0.08 m concrete
density = 2000 kg/m3
heat capacity = 834 J/kg/K
day - 0.5 air changes per hour (ACH)
night - 30 ail' changes per hour (ACH)
air density = 1.2 kg/m:3
air heat capacity = 1012 J/kg/K
exterior walls, roof: U„= l . 75 W/m-'-mass/K
(wall and roof nominal coefficient = 0.5 W/m^/K)
infiltration: U,„,= 0.42 W/bi2 -mass/K (0.5 ACH day)
U,of= 25.3 W/m2-mass/K (30 ACH night)
mass to air: 3.7 W/m^/K (day)
i r = 4.8 W/m^/K (night)
(Chandra and Kerestecioglu. 1984)
Capacity for Potential Cooling of a Climate
We define the potential for cooling for a given day by a poten-
tial cooling ratio, PCR, as follows:
(Tcomf ~ T»mb , m t n ) / (1 «
-------
3118
32 t
ambient
28
room no nv
O
! room w.
4)
mass no nv
comfort
mass w. nv
20
18
16
Time of day (hours)
Fig- 2. Extent of discomfort without night ventilation (no nv)
indicated by the hatched area and with night ventilation (w. nv)
- dotted area. Building parameters as shown in Table 1.
temperature following a sine function, our above analysis using
this function as input to the governing equations would be exact
and represent the seasonal performance of the (idealized) buil-
ding. Our strategy in dealing with a real climate is to calcu-
late the frequency densty, g(PCR). The cooling elficency is
calculated as a function of PGR, and the relationship between
T«„,b(t) and PCR is given by
l«»nb(t) = Tco,,,, + < A > * (1-2 * FOR) + *sin(w*t) (3)
where w = 2*pi/24 (1/hours), and = seasonal average daily
amplitude. The seasonal cooling efficiency, CK..„on)is obtained
by summing the results while weighing them by the frequency of
their occurrence:
••
CEm.a.on = w/9
-------
3119
ACH=30 /
PCR -
Fig. 3. Cooling efficiency, CE, as a function of potential
cooling ratio, PCK, for varying nighttime infiltration rates.
(/>
c
a>
a
c
a)
3
c
a>
AAA /
i itiin 111 ii 11
m
i i i i i i i i
0.2 0.4 0.6 0.8
1.2 1.4 1.6 1.8
PCR (-)
Fig. 4. Frequency density of PCR for two comfort temperatures,
Tcomr, for May to October, Fresno, California.
-------
3120
lation than with. In Figure 4 the frequency distribution ot PCR
is plotted for two different comfort temperatures tor Fresno, Ca
for which the cooling season is from May t:o October. In Table 2
the seasonal performance, calculated from eqn. (4), is compared
to an hour by hour simulation using the same thermal network and
hourly ambient temperatures of the typical meteorological year,
increased night ventilation was assumed whenever the ambient
temperature dropped below the room temperature.
TABLE 2. Comparison of seasonal cooling efficiencies based on
simulation and proposed method. Building parameters as in Table
1, for Fresno, California, cooling season, (day: ACH = (Kb)
--
night:
Tcom*
ACH = 2
eq.(4) simulation
ACH = 8
eq.(4) simulation
ACH = 30
eq.(4) simulation
22 C
2b C
0.123 0.120
0.1b3 0.14/
0.2b3 0.249
0.304 0.296
0 . 33b 0 . 330
0.4U/ 0.42U
DISCUSSION
The purpose of this study has been to propose a simple approach
for the development of design tools and the long-term prediction
of the thermal performance of night ventilation cooling. Such
tools may help architects make decisions in the early stages of
building design. The approach entails the approximation ol the
building by simple thermal networks, the calculation of cooling
efficiencies based on sinusoidal ambient temperatures, and the
estimation of the seasonal performance from climate dependent:
frequency distributions ot potential cooling ratios. The method
has been applied to buildings without air conditioning backup.
The results have been compared, with reasonable success, to a tew
hour-by-hour computer simulations. The effect ot inclusion ol
solar or internal heat gain has yet to be investigated. Fur ther
studies are required and comparison with experimental data are
needed to permit better judgement of the suggested method.
REFFJRKNCES
Chandra, S., and A. A. Kerestecioglu (1984). Heat transler in
naturally ventilated rooms: Data from full-scale measurements.
ASHRAE Transactions 90 (IB): 211-22b.
Chandra, S. (1989). In J. Cook (Ed.), Passive Cooling, Mir Press,
Cambridge, Ma. Chap. 2. , pp. 42-80.
Clark, G. (1989). In J. Cook (Ed.), Fassiye Cooling, MIT Press,
Cambridge, Ma. Chap. b.b, pp. 392-419.
Feuermann, D., J. M. Gordon, and Y. iarmi (198b). On massive
envelopes in passively heated solar buildings: An analytic
sensitivity study. Sol. Energy, 3b, pp. 271-279.
Gordon. J. M., and Y. Zarmi (19H1). Analytical model lor passive-
ly heated solar houses - 1. Theory. Sol. Energy, 27, pp, 331-42.
Niles, Philip W. B. (198b). Equations to predict the temperature
inside ol a night ventilation cooled building on a design day.
11th Nat. Passive Solar Conference, Boulder Co.. pp. idb-JJl.
Rabl, Ari (1988). Parameter estimation in buildings: Methods tor
dynamic analysis of measured energy use. ASME Journal ol Solar
Energy Engineering, 110, 1, pp. b2-bb.
-------
3121
A PASSIVE SOLAR HOUSE IN A HOT HUMID CLIMATE DESIGNED TO BE
COOLER BOTH WINTER AND SUMMER
L.M. Holder III
4202 Spicewood Springs Road
STE 214
Austin, Texas 78759
2
Nestle Ridge is a 2800 ft house located along a bluff overlooking Barton
Creek in Austin, Texas. The clients requested the project be designed to be
cooler than normal in the winter and feel cooler in the summer.
Fig. 1.
-------
3122
The program required careful location of all glazing elements. The main
collector areas of the house were located higher in the space than normal.
On cold sunny days this warmed air as well as the waste heat from the house
can be captured at high points in the house and recycled to the lower area
using a heat recovery fan and duct system. On warmer winter days the heat
can be vented out. This is especially significant since winter in Austin
includes temperatures in the 80's even in January. During the summer,
because of orientation and shading of the glass, there is almost no direct
gain and radiation due to sun on glass surfaces.
Glass areas on the north and south with selected areas on the east allow
daylighting into every area of the house. In addition to providing superior
quality lighting, this strategy minimizes interior loads from artificial lights.
Selected operable sections of those windows provide for natural ventilation
and night flushing. As air is drawn in through low windows on the south it
is exhausted through high windows on the north. The opening area and it's
location is designed to induce an adequate ventilation rate even when there
is no wind, a condition which occurs often in the summer in Austin.
mm
SoxLth. E levcLi-ion
-------
3123
Since the house is on a steeply sloping site, there is a large area under the
house being used to precool air in the spring and fall before being ducted
into the house. This air is naturally drawn from under the house using the
low pressures developed inside the house from the air movement
characteristics. Moving through dampers in the center of the house, this air
movement provides a source of cooler air during the swing seasons.
To minimize loads from the relentless "Texas Sun", the roof is multilayered.
Continuous ridge and vented roof edge allow air to move below the metal
roof and above the radiant barrier. The radiant energy is trapped in this
layer, heats the air to temperatures of 140-160 degrees , and the converted
energy is carried out by convection. The radiant barrier separates the upper
ventilation layer from the lower ventilation layer. Continuous soffit vents
draw air into this area to be vented out through the ridge vent. Peak
temperatures of 90-100 degrees are normal for this lower area and minimize
transfer through the insulation below during the cooling season.
JZ=i CZL
West Elevg.ti.on
9omt». 1 /¦*" - f-tT
Providing supplemental heat in the living area and bedroom during the
winter, is a fireplace with a heat exchange firebox. The core of the masonry
around the fireplace is filled with crushed limestone to provide mass in the
center of the house for temperature stabilization year round.
-------
3124
Just below the sheetrock, 1/2" urethane insulation covers the interior
surface of the exterior wall of the house. This effectively isolates the
sheetrock from the dynamic loads of the daily temperature swings and
allows the wall surfaces to track the temperature of the interior. This
combined with the low "e" glass and no direct gain on the glass in the
cooling season moderates mean radiant temperatures of all surfaces, allowing
human comfort at higher temperatures in the summer and lower
temperatures in the winter.
Careful combination of elements of orientation, shell materials, and skin
detailing contribute to the performance and comfort characteristics of this
house over a large range of climatic conditions with little dependence on
mechanical equipment.
East ELexia.ti.o-n.
VV - f-CT
Fig. 6.
-------
3125
MODULATED SOLAR SHIELDING OF BUILDINGS:
A solar radiation control strategy for low energy buildings in hot dry and
semi-arid climates
G S Yakubu and S Sharpies
School of Architectural Studies, University of Sheffield
Sheffield S10 2TN, United Kingdom
ABSTRACT
This paper presents some of the results from a series of research studies of modulated solar
shielding in the context of solar radiation control in hot-dry and semi-arid climates, carried
out primarily at the University of Sheffield which has been partly reported elsewhere (Yakubu
and Sharpies, 1990a, 1990b). Solar shielding refers to the solar protection of the entire or large
parts of the building's external fabric and not just those elements which directly transmit solar
radiation. The concept of solar shielding was conceived from the interplay of the climatic and
environmental factors of hot dry and semi-arid lands, which are not fully addressed by formal
shading techniques. The research was undertaken with particular reference to the hot semi-
arid climate of northern Nigeria and included full scale measurements, laboratory tests and
thermal modelling.
Measured and simulated results portrayed not only a significant agreement but also indicated
that solar shielding could have a higher solar protection efficiency than shading devices, in the
reference climate. The results also portrayed that the overall effectiveness of the technique
is dependent upon the extent to which the shielding system is conceived as a modulated
one. Furthermore it was found that the efficiency of the system cannot be optimised to its
full potential as a result of conflicts between the needs for maximising the prevention of solar
thermal gain of the building and minimising the side effects on daylighting and view during
the day time.
In this paper, some of the thermal implications of the technique are presented.
KEYWORDS
Solar radiation control; shading devices and techniques; modulated solar shielding; hot-dry
and semi-arid climates.
INTRODUCTION
Of all the factors and elements affecting the built environment in any part of the world,
climate seems to be the most significant. Depending on the climatic location, buildings
either need to be heated or cooled to maintain comfortable indoor conditions. In the hot dry
and semi arid climates the ambient heat is very intense and passive cooling and the reduction
of building over-heating are the major design constraints. Therefore there is every need to
develop climatic control techniques that can effectively cope with the thermal stress of the
climate. Architectural forms developed from such control techniques should enable buildings
to respond not only to the climate as put forward by the exponents of bioclimatic design such
-------
3126
as Olgay and Olgay (1957), Olgay (1963), Koenigsbergcr and colleagues (1973), Givoni (1976),
Evans (1980) and others, but also to the traditional and socio-cultural idiosyncrasies of the
people as postulated by Rapoport (1969) and Rudofsky (1964) among others.
The responsiveness of buildings to the stress of the climate is very crucial in a hot dry/semi-
arid climate which is so severe that life is hinged on the fringe of survival. Much of the
discomfort is due to the heat from the scorching solar radiation. Daytime temperatures of 49°
C or more (far higher than skin temperature) have been recorded while on the other hand
night time temperatures could be as low as 10 °C. The air is hot and dry with relative humidity
sometimes as low as 10% (Golany, 1980). Consequently a systematic control of solar thermal
radiation in buildings is the most important strategy for designing passive and low-energy
buildings in this climatic belt.
SOLAR RADIATION CONTROL IN BUILDINGS
The basic approach to the control of solar radiation in buildings has been the adoption of
the concept of shading and/or sun control strategy which, in earnest, should mainly be used
in higher latitude climates for which they were conceptually designed and for which they are
suitable. This concept is based on the use of shading devices to protect glazed areas of the
building fabric against direct solar rays.
One of the most prominent features of all shading devices is that they have been designed
and optimised with a fundamental underlying consideration of having a capability to give
maximum shading during the hot summer and conversely let in the maximum sunlight in
winter. Olgay and Olgay (1957) emphasized this when they suggested that
"...the effectiveness of a shading device depends on the proportionate success
with which it covers a given surface during the overheated period without inter-
cepting the sun's energy during the under-heated times".
In hot dry and semi arid climates however, there is a constant need to keep the sun out,
throughout the year, which, among other reasons, makes the use of the above shading concept
inapproriate for the climate. In light of the fact that there is no universal architecture, form
or typology that meets the needs of all climates it is suggested that a shading philosophy must
be so evolved as to reflect regional climatic diversity. In whatever perspective the inclements
of the climates are viewed, the case of hot-dry/semi-arid climates is especially precarious.
Some of the hazards include the dust storms, the sheer stress resulting from the high diurnal
temperature range, the dry air and the sharp contrast between indoors and outdoors. Fur-
thermore, the need for protection against glare and diffuse/multiple intereflected irradiances,
especially from the ground, calls for an alternative solar protection strategy for hot dry and
semi-arid climates. An effective shading or solar protection methodology should also address
these additional needs.
The unglazed part of the fabric of contemporary buildings, constructed with modern materials
such as concrete and concrete block do absorb relatively more heat than the adobe fabric of
the traditional buildings. Thus there is need for some form of thermal barrier between the
fabric and the harsh outdoors. Creating this thermal barrier would call for an alternative
shading strategy that would not in application necessarily be on the immediate facade of the
building structure, a principle upon which the use of shading devices is based. Quoting Marcel
Breuer, Olgay and Olgay (1957) emphasised this, saying:
"The sun control device has to be on the outside of the building, an element of
the facade, an element of architecture..."
-------
3127
Most contemporary research on shading deals with variations of the mode of use/design of
different solar protecting devices based on this concept and the effects of these variations on
parameters of indoor thermal comfort.
Modulated Solar Shielding
Taking inspiration from the interplay of climatic and environmental factors of hot dry and
semi-arid climates, and the direct experience of living and working in the climate, the concept
of modulated solar shielding was conceived. Solar shielding refers to the solar protection
of the entire or large parts of the building's external fabric (and not just those elements which
directly transmit solar radiation), by an impermeable or semi-pcrmeable structural membrane
located away from the building fabric by a usable space. The structural membrane is referred
to as the shielding system, while the boundary space between the shielding system and the
building fabric is called the shielded space.
Thus the concept is based on the shielding of the building fabric, (as opposed to mere shading
of the glazed openings), by a controllable or adjustable system separated away from the
parent structure by a usable space. This space thus becomes a semi outdoor space, creating
a buffer zone between indoors and outdoors. The shielding structure itself becomes a semi-
porous secondary shell acting either as a selective filter or an excluder depending on the
weather conditions and the time of day. Hence the technique combines the advantages of both
the 'exclusive' mode in which the control system aims to exclude the effect of the external
environment upon internal conditions, and the 'selective' mode which depends on the selective
admission of substantial elements of the external environment into the building (see Hawkes,
1982). The degree to which the system can be effective would depend to a large extent on the
design of the system as a modulated one, with suitable materials. The object of the research
was to study the application of this concept as a passive solar radiation control strategy for
the hot dry/semi-arid climatic belt, especially with respect to the reduction of solar heat gain.
MEASUREMENTS AND SIMULATION
Fieldwork was undertaken to measure and collate data on dry bulb temperature in a full
scale residential building (Fig. 1) which incorporated some degree of solar shielding of its
fabric, in a hot semi-arid climate. It was carried out in Kano, a historic and commercial
city in northern Nigeria. Most of the external facades of the sleeping areas in the reference
building were protected from direct solar radiation by means of thin louvre systems which
thus created a buffer zone between the inner living spaces and the ambient. Hourly dry bulb
temperature values were measured by means of copper-constantan thermocouples connected
to a data logger which can automatically scan and record data from over 60 sensor channels.
The temperature values were sampled every ten minutes and the average recorded every hour
for sixty days.
Then the building was simulated using an existing thermal model (SERI-RES Version 1.2)
to enable a parametric study of the shielding technique to be made. In order to provide a
base case, the building was initially simulated without any form of shading (NOSII). Then the
building was simulated with shading provided by concrete overhangs and sidefins (OVH) and
then with shading provided by a balcony (BALC). Finally it was simulated with the shielding
option. Further studies of the performance of the shield by variations of the mode of shielding
were examined vis-a-vis:
(a) Partial shielding of the facades only (PSS);
(b) Full shielding of the facades (FSS);
(c) Complete shielding of the building including the roof (CSB).
-------
3128
RESULTS
The following are some of the results from the thermal simulation studies. Full details of the
study can be obtained from Yakubu (1990).
The use of overhang and sidefins for shading did not have any appreciable decrease in the
overall solar radiation transmitted inwards (Figs. 2 and 3).
Balcony shading was only slightly better than the use of ordinary overhang and sidefins.
The shielding option was found to be the most effective in reducing solar thermal gain. The
solar shielding technique was first simulated with wooden louvres inclined at angle 30° to the
horizontal. As the louvre angles were increased at 15° intervals to 45°, 60° and 75°, and for
87° the following were observed:
(i) Increase in the louvre angle which correspondingly decreases the aperture in the shielding
system induced a gradual reduction of the inwardly transmitted solar radiation (Fig. 4).
(ii) A decrease in the aperture of the shielding system resulted in a gradual decrease in the
indoor air temperature due to the combined effect of the decrease in solar heat gain and the
corresponding decrease in the airflow. The level of temperature variation according to louvre
angle is dependent on whether the building is partially shielded on the sides, fully shielded
on all sides or completely shielded on all sides and the roof. At 75° louvre angle, the partial
shielding of the building sides reduced the solar transmitted gain by 21 MJ (50.9%) while up
to 95% reduction can be obtained by completely shielding the facade of the building. However
these huge reductions are not translated into a proportionate reduction of the temperature
profiles due to several factors, namely the impairment of longwave radiative/convective heat
loss of the building by the shield, and the longwave thermal transfer from the shield and the
roof inwards into the living spaces.
A more vivid picture of the daytime conditions are presented by the mean hourly values of
temperature (Fig. 5). It shows that shielding has a greater potential to reduce thermal dis-
comfort during the day. Another feature of the shielding option is the fact that the temperature
range in this option is the lowest, implying a lower indoor thermal stress.
The effect of partially shielding the building facade is dependent upon the proportion of the
facade that is shielded, its orientation, the sizes of the openings on the unshielded parts and
their shading characteristics.
CONCLUSIONS
Whilst shielding has been shown to be relatively more effective in reducing indoor thermal
gain, its efficiency is a function of not only the area of the building's external surface that is
shielded, but also of the angle of inclination of the louvre slats of the shielding system. When
the external surfaces of the rooms are partially shielded, there is a substantial reduction in
solar thermal gain which is even higher when the whole sides are shielded. Shielding is fully
optimised when the external surfaces exposed to the sun, including the roof, is shielded.
Whilst huge reductions (over 90%) of the inwardly transmitted solar radiation can be obtained
from solar shielding technique, it does not give a relatively corresponding high reduction
of indoor temperatures. This is partly because the shielding system's solar thermal gain is
partially re-radiated as longwave radiation, coupled by convective transfer, inwards into the
living spaces. This effect must be minimised by appropriate choice of materials, colour, and
adequate ventilation of the shielded buffer spaces.
-------
3129
REFERENCES
Evans, J. M. (1980). Housing, Climate, and Comfort. Architectural Press, London.
Givoni, B. (1976). Man, Climate and Architecture. Applied Science Publishers, Barking.
Golany, G. (1980). Housing in arid lands: Design and planning. Architectural Press, London.
Hawkes, D. (1982). Building shape and energy use. In Hawkes, D. and J. Owers (Eds)
Architecture of energy. Construction Press, Essex.
Koenigsberger, 0. H., T. G. Ingersol, A. Mayhew, and S. V. Szokolay (1973). Manual of
tropical housing and building: parti - climatic design. Longman, London.
Olgay, A. (1963). Design with climate. Princeton University Press, New Jersey.
Olgay, V. and A. Olgay (1957). Solar control and shading devices. Princeton University Press,
New Jersey.
Rapoport. A. (1969). House form and culture. Prentice Hall, Englewood Cliff, New Jersey.
Rudofsky, B. (19G4). Architecture without Architects. The Museum of Modern Arts, New
York.
Yakubu, G.S. (1990). Modulated Solar Shielding Of Buildings: A study of a solar radiation
control strategy for low energy buildings in hot dry and semi-arid climates. PhD Thesis,
University of Sheffield, Sheffield.
Yakubu, G.S. and S. Sharpies (1990a). The implications of pressure coefficient profiles for
passive ventilation cooling. Proc. World Renewable Energy Congress, Reading, UK.
Yakubu, G.S. and S. Sharpies (1990b). The impact of room orientation on indoor thermal
conditions. Proc. Passive and Low Energy Conference, Halifax, Canada.

-------
3130
Solar Gain (MJ)
Fig 2: Mean Daily Solar Gain (MJ)
For Different Solar Control Options
SR-NE
SR-SE SR-SW
Zones
IFSS E5S PSS ~ BALC
Temperature ( C)
J OVH ~ NOSH
SR-NW
Fig 3: Mean Maximum Air Temperatures
for Different Solar Control Options
SR-SE SR-SW
Zones
3 FSS CD PSS ^BALC ~ OVH ~ NOSH
Solar Qaln (MJ)
Fig 4: Mean Daily Transmitted Solar
Gain For Different Louvre Angles
(Full Shielding Option)
J
SR-SE
SR-SW
SR-NW
Zones
3 75 CZ3 60 ^ 46 ~ 30
Temperature ( C)
Hour
Fig 5: Mean Hourly Temperature for
Different Solar Control Options
-------
3131
VALIDATION OF VIDEO IMAGE CAPTURE/INTERPRETATION METHOD
TO APPROXIMATE SOLAR RADIATION REDUCTION
DUE TO TREE SHADING IN THE SUMMER
John L. Motloch, Associate Director
Center for Urban Affairs, Texas A&M University
Kyoo D. Song, Doctoral Student
Department of Architecture, Texas A&M University
ABSTRACT
This paper presents the procedure to validate the video image processing method for approximating
solar radiation transmissivity through tree crowns. Video image capture hardware and digital image
processing software were employed to capture and store tree crown images; then the computer
program CROWN, developed as part of this research, was used to predict the transmissivities of tree
crowns. In addition, field measurements were conducted on five locally common tree species under
locally typical summer sky conditions to examine the reliability of the video image capture/
interpretation method. Asa result, the video image capture/interpretation method proved to be reliable
and efficient in collecting tree crown images and predicting reductions of direct solar radiation through
tree crowns.
KEYWORDS
Solar radiation; tree shading; tree transmissivity; video image; digital image processing; radiation
measurement.
INTRODUCTION
Estimates of tree transmissivities are required to evaluate the influence of tree shading on building
energy use. A number of researchers have employed and tested (Heisler, 1983) photographic methods
for estimating tree density or transmissivity. Heisler (1982) and Wagner and Heisler (1986) projected
slides of crown images onto translucent dot grid paper and counted dots coinciding with tree parts.
Shiler and Greenberg (1979) used videoscanning and digitizing techniques on 35 mm black and white
negative film to find the ratio of occluded area to total area within the outline of tree forms. The
photographic interpretation of tree crown densities facilitated the convenient collection of data, and
proved to provide reasonably reliable estimates of solar radiation reductions for leafless trees under
clear skies, but has not been accepted as providing reliability in estimating radiation transmissivities
in leafed (summer) conditions. This method also requires people to count the number of grids or dots,
and demands excessive time in collecting performance data on a large sample and range of tree species
of differing sizes and leaf conditions, in different climates, and introduces subjectivity in interpreting
images. The videoscanning method requires time for film processing which might be critical in the
research procedure. The goal of this research is to establish the computerized method of interpreting
tree crown transmissivities.
METHODS
Selection of Trees
Five different tree species which represent different summer crown densities were selected. These
trees were located near the electric power source for measurement instruments. The selected trees
included Live Oak, Crape myrtle, Pine tree, Post oak, and Mesquite. Figures 1 through 5 show these
trees.
Video Image Capture/Interpretation
Tree crown images were captured and interpreted using hardware and software shown in Table 1. Tree
crown images were captured just before sunrise, when the trees were silhouetted against the sky, as
-------
3132
TABLE 1 Video Image Capture/Interpretation Hardware and Software

Item
Use
Hardware
8 mm video camera
TARGA-16 board
A 386 AT class PC
Capturing tree crown images
Digitizing crown images
Digitizing and interpreting
Software
TIPS
CROWN
Digitizing crown images
Interpreting dizitized images
Fig. 3. Pine tree
(Texas A&M Univ. Campus)
Fig. 2. Crape myrtle
(Texas A&M Univ. Campus)
Fig. 1. Live oak
(Texas A&M Univ. Campus)
Fie. 4. Post oak
Fig. 5. Mesquite
(College Station Central Park) (College Station Central Park)
Fig. 6. Mesquite Image 1
(South)
Wagner and Heisler (1986) suggested, to obtain sharp contrasts between tree crown image and sky.
Tree crown images were captured while maintaining the tilt angle of the video camera consistent with
the sun angle in the summer season (June) which represented the average solar altitude angle from 9
A.M. through 3 P.M. for the south wall and 3 P.M. to sunset for the west wall. To represent the average
solar altitude angle for the south wall, the video camera was tilted 80° above the horizon, and it was
tilted 49° to represent the average solar altitude angle for the west wall. For each wall direction, three
directional views of the crown area (taken at approximately 120° azimuthal angles to one another)
were recorded by the video camera. The crown areas which could be recorded by the video camera
varied in accordance with the tilt angles of the video camera and the sizes of the tree crowns. For the
west wall, most of the crown area could be captured while aligning with the solar altitude angle. For
the south wall, however, as the video camera was tilted 80° above the horizon, the video camera needed
to be located just underneath the tree crowns. In this case, since it was not possible to capture the entire
crown areas, several partial images were captured from arbitrary portions of each tree crown. In this
manner, a total 6 images.(3 south wall + 3 west wall) were recorded for each of the five trees. The RGB
video images recorded by the video camera were digitized and stored in disk files using the TARGA-
16 board and the TIPS imaging software installed in the PC. The digitized image files were processed
as discussed in the authors' previous paper (Motloch and Song, 1990). One of the crown images
interpreted by the CROWN program is shown in Fig. 6.
Solar Radiation Measurement
The reliability of the video image capture/interpretation method to estimate tree transmissivity was
examined by comparing its results with those obtained from field measurement on the same trees under
-------
3133
TABLE 2 Instruments for Solar Radiation Measurement
#
Model
Use
1
2
3
4
5
6
7
8
9
Eppley PSP
Eppley NIP
Qualimetrics 3020
Qualimetrics 3020
Qualimetrics 3020
Qualimetrics 3020
Qualimetrics 3020
QDL-2000
TANDY 1000
Itho on a horizontal surface in the open
Idno on a normal surface in the open
Itho on a vertical surface in the open
Itvs on a vertical surface in the tree shade
Itvs on a vertical surface in the tree shade
Itvs on a vertical surface in the tree shade
Isvs on a vertical surface in the tree shade
Data logger
Laptop computer for data recording
TABLE 3 Measurement Dates and Sky Conditions
Date
Tree
Sky Condition
6/08/90
Mesquite
Clear
6/11/90
Crape myrtle
Clear
6/12/90
Post oak
Partly cloudy
6/15/90
Live oak
Partly cloudy
6/20/90
Pine tree
Clear
summer sky conditions. In the field, six pyranometers, a pyrheliometer and a data-logger we connected
to a lap-top computer were used to record solar radiation intensities. Table 2 shows the instruments
for field measurement. To measure radiation on vertical walls, pyranometers were mounted vertically
on wood panels which were supported by steel frames. Three Qualimetrics 3020 pyranometers were
used to record total incident radiation on south and west facing walls within the shadow pattern of tree
crowns (Itvs). The fourth pyranometer, also a Qualimetrics 3020, was shielded from direct radiation
using a shadow band to record diffuse sky radiation on the vertical walls within the tree shadow (Isvs).
The shadow band, measuring 3 inches in width and 25 inches in radius, was made of metal sheet
painted in non-reflective black paint. In the open space, the Eppley PSP pyranometer was placed
horizontally to record unshaded total radiation on a horizontal plane (Itho). An additional Qualimetrics
3020 pyranometer was installed vertically to record unshaded total radiation on a vertical plane (Itvs).
In addition, the Eppley NIP pyrheliometer mounted on a solar tracker measured direct beam radiation
on a normal plane (Idno). The schematic diagram of the instrumentation and the actual arrangement
of instruments are presented in Fig. 7 and Fig. 8, respectively. Through the measurements, in addition
to solar radiation data, the measurement times (in solar time) were recorded every minute from 9:00
A.M. to 3:00 P.M. for the south walls and from 3:00 P.M. to 4:00 P.M. for the west walls. Measurement
times were later used in the lab to calculate sun angles. Table 3 shows the dates and sky conditions for
the measurement on each tree. The portions of solar radiation transmitted through tree crowns were
calculated by the following methods.
The ratio of total radiation in the tree shade to that in the open was calculated by:
ItvRatio = Ave(Itvs) / Itvo
where Ave(Itvs) = Average of total radiation on vertical surfaces in the tree shade
Itvo = Total radiation on a vertical surface in the open
The ratio of direct beam radiation in the tree shade to that in the open was determined by:
IdvRatio = Ave(Idvs) / Idvo
where Ave(Idvs) = Average of direct radiation on vertical surfaces in the tree shade
Idvo = Direct beam radiation on vertical surfaces in the open
= Idno Cos
-------
3134
In this case, Ave(Idvs) was calculated by:
Ave(Idvs) = Ave(Itvs - Isvs)
where Isvs = Diffuse radiation on a vertical surface in the tree shade measured using the
pyranometer shielded by the shadow band.
P1=Eppley PSP
P2=Eppley NIP
P3=Qualimetrics 3020
P4=Qualimetrics 3020
P5=Qualimetrics 3020
P6=Qualimetrics 3020
P7=Quolimetrics 3020
with shadow band
|e!*3i r
Logger I—M L
Computer
Tree
Power
Fig. 7. Schematic diagram of instrument arrangement
Fig. 8. Actual arrangement of instruments
RESULTS
The final results of the transmissivities obtained from the visual image capture/interpretation method
and field measurements are shown in Table 4. V alues indicate average radiation transmissivities of tree
crowns for south walls and west walls.
-------
3135
TABLE 4 Average Transmissivities determined bv the Visual Image
Capture/Interpretation Method and Field Measurement
Tree
Wall
Average Transmissivities (%)
video image capture/interpretation
Measurement
Image 1 Image 2 Image 3
Average
Idv Itv
Live oak
South
Wr-st
12.82 11.75 11.29
10.47 11.98 13.45
11.95
11.96
16.3 55.5
10.5 18.8
Crape myrtle
South
West
20.19 17.73 14.73
13 75 10 89 13 13
17.55
12 47
15.2 56.1
9 5 75.2.
Pine tree
South
West
10.62 8.10 8.55
2.52 2.76 4.93
9.09
3.40
7.2 40.2
3.7 17.5
Post oak
South
West
15.59 17.70 11.74
5 15 3 78 3 48
14.70
4 14
16.7 45.4
6 7 75 1
Mesquite
South
West
15.03 48.09 46.41
4.18 2.78 9.50
36.51
5.49
29.8 60.9
5.9 18.1
As shown in Fig. 9 and Fig. 10, both total and direct radiation measured in the field showed lower
transmissivities on west walls because of the path of incoming radiation in relation to the tree
branching and leaf distribution patterns. The results obtained by the video image capture/interpreta-
tion method closely approximated average measured transmissivities of direct beam radiation on both
south and west walls.
Fig. 9.
LO CM PT PO MQ LO CM PT PO MQ
TREE SPECIES TREE SPECIES
Video Me as. Idv Meas. ttv
Average transmissivities on south wall Fig. 10. Average transmissivities on west wall
CONCLUSIONS
Several conclusions can be drawn from the procedures to validate the reliability of the video image
capture/interpretation method by comparing its results with the actual field measurements of solar
radiation on vertical surfaces.
1) The number of the crown images to be captured from a given tree may depend on the level of
accuracy required to estimate the reduction of solar radiation through the tree crown. If the highest
accuracy, say, hourly radiation, is required to estimate solar radiation reductions due to a specific tree,
the crown images may be captured by changing the tilt angles of the video camera to simulate hourly
sun altitude angles. However, this procedure may require a lot of time to gather performance data for
many different tree species.
-------
3136
2) Field measurements confirmed that magnitudes of radiation reductions varied with different
densities and shapes of tree crowns and orientations of the receiving walls. In general, radiation
reduction due to tree shading on west walls were much greater than those on south walls. This was due
to the path of incoming radiation in relation to the tree branching and leaf distribution patterns.
3) The video image capture/interpretation method proved to be efficient and reliable in estimating
reductions of direct beam radiation through tree crowns.
4) However, the video image capture/interpretation method was not reliable in estimating reduction
of diffuse radiation from the sky and reflected radiation from the ground, because this method could
take only visible light into account. Another approach to overcome this limitation will increase the
reliability of this method in predicting reductions in total daily radiation.
5) Even though reductions of diffuse sky radiation and ground-reflected radiation were not addressed
in the current research, they may be determined by considering geometric relations between trees and
wall surfaces. As reductions of sky diffuse radiation and reflected radiation from the ground depend
on the size, shape and density of the tree crown and the distance between tree and wall surface, they
may be estimated by calculating the density and form factors of the tree crown and the shadow on the
ground seen from the wall surface. Further detailed methodology should be developed in further
research.
6) The reliability in predicting radiation reductions on south walls should increase with greater
numbers of images within the crown to predict crown density more accurately .
7) Since direct beam radiation is a major portion of the total solar radiation from clear summer skies
under which summer cooling loads reach their peak, the video image capture/interpretation method
is considered to be a more efficient vehicle to gather performance data than previous photographic
methods.
ACKNOWLEDGEMENT
This research was supported by funding through the Center for Energy and Mineral Resources of
Texas A&M University, College Station, Texas, U.S.A.
REFERENCES
Heisler, Gordon M. (1983). Measurements of Solar Radiation on Vertical Surfaces in the Shade of
Individual Trees. Forest-Atmosphere Interaction. Proc. of Forest Env. Meas. Conf.. Oct 23-28.
Heisler, Gordon M. (1982). Reduction of Solar Radiation by Tree Crowns. Progress in Solar Energy.
American Section of the International Solar Energy Society, Inc.
Motloch, John L. and Kyoo D. Song (1990). Approximating Tree Transmissivities Using Visual and
Image Capture/Interpretation Methods. 15th National Passive Solar Conferences. Austin, Texas,
March 19-22. pp. 335-339.
Schiler, Marc and Donald P. Greenberg (1979). Computer Simulation of Foliage Shading in Building
Energy Loads. Proceedings of the 16th Annual Design Automation Conference. New York, Institute
of Electrical and Electronics Engineering.
Wagner, J. Alan and Gordon M. Heisler. (1986). Rating Winter Crown Density of Deciduous Trees:
A Photographic Procedure. Landscape Journal. Vol. 5. No. 1.
-------
3.19 Passive Computer Analysis I
-------
Intentionally Blank Page
-------
3139
THIRD GENERATION OF A THERMAL SIMULATION PROGRAM: tsbi3
J. E. Christensen and K. Johnsen
Danish Building Research Institute
DK-2970 Horsholm, Denmark
ABSTRACT
This paper describes a new thermal analysis program tsbi3 for PC's which has
been developed by the Danish Building Research Institute for commercial use. The
program makes an hourly analysis of indoor climate, energy consumption, etc.^
using weather data from the Test Reference Year. The program is user-friendly
and flexible, and operated by a combined window-/menusystem interface using
keyboard and mouse.
KEYWORDS
Energy Analysis of Buildings; Thermal Simulation Program; Energy Consumption;
Indoor Climate; tsbi3; Atria;
Wecther date:
Dry bulb temperoture
Humidity rotio
Direct normol radiation
Diffuse radiation
Wind speed
Ficiive zone 1
Outdoor air
Panel
_ , cooling
Surfaces *
Subsurfaces:
Constructions
-layer
- me'erio's
Wirdows
-clings
-rromes
Doors
Shodows
Local
ventilation
system
Equipment
Fictive zone 3
% Ficiive^ i Trons
% zone 2; * miss.c
% Ground
£ig. 1. Key diagram for tsbi3: Schematic drawing of room model, showing
the program's modules of calculation.
Preceding page blank
-------
3140
INTRODUCTION
The computer program tsbi3 (.thermal simulation of buildings and installations)
is a new thermal analysis program. The program has been developed by the Danish
Building Research Institute for research and for commercial use by consulting
engineers in the field of heating and air-conditioning. The program can be used
for analysing indoor climate, energy consumption, passive solar energy, auto-
matic control functions, etc._| in connection with the planning and design of
buildings, energy-conservation measures, renovation of buildings, and heating
and air-conditioning systems.
Experience from introduction courses held in cooperation with the Danish
Association of Heating and Ventilating Engineers shows that it takes only eight
hours to learn how to use the program. That makes tsbi3 a very flexible tool for
design engineers. The program is supplied with library facilities to make easy
and quick choises of the desired materials, components and systems.
The program is particulary suitable in connection with the planning of major
building projects and retrofitting. Here, the consequences of alternative
solutions can be calculated rapidly, thereby facilitating comparison and choice
of optimum design, and the results can be used as documentation for the client
and the authorities - for example, where air-conditioning plant with automatic
cooling is desired.
THE MODEL
The instationary heat conduction through the walls is done by an implicit
difference method in which the model itself determines the number of nodes
necessary for each construction. Both heat flow by transmission and by air-
mixing between zones are allowed for.
As outdoor climate data is used the Test Reference Year (TRY), including the
following parameters: Outside dry bulb temperature, windspeed, humidity, direct-
and global radiation, and the cloud cover index. The climate data is described
in (Andersen, 1982). However, the program can also use other climate data - for
example, foreign meteorological data. The user's manual for tsbi3 includes a
special chapter on the layout of meteorological data for the program.
When the program is running, the simulation is achieved by hour-by-hour ana-
lyses with time-step levels for the regulation from 30 minutes and on down to
just a few minutes.
The building model used in tsbi3 is defined as a hierachy of objects with
attributes and relations between objects of different levels.
The model is divided in 5 levels:
Level 1: Building level containing data about the building site,
weather data, etc. Possible relations to surroundings
(buildings etc.) casting shadows.
Level 2: Zone level containing objects that define the zones
(rooms) of the building.
Level 3: Subsystem level containing objects of surfaces (with
relations to level 2 defining which zones the surface
is facing), objects of schedules for equipment,
appliances, people and HVAC-systems.
Level 4: Part level containing objects of subsurfaces, windows
and doors, objects of technical systems and behaviour,
to be used when scheduled.
Level 5: Detail level containing for subsurfaces layers of
different materials, and for windows relations to
schedules of shading devices and shutters. All with
relations to external libraries of materials.
-------
3141
USER INTERFACE
tsbi3 is user-friendly and flexible, and operated by a combined window-/menu-
system interface using keyboard and mouse. The menus in the program are built
up in a way that they overlap each other with the outer level to the left going
to the right, fig. 2.
Everywhere in the program the user can enter functionkeys in order to get on-
line help. In addition the program immediately gives warnings and errors, if the
entered parameter is not likely/legal according to the other parameters.
By entering a function the user can get a general view for the submenu, where
the user is. This can be a general view of zones, surfaces, instalations,
etc, fig. 3. By entering another function the program can control the project in
order to find errors and missing data. These facilities help the user to create
a correct building model and the general view.
— t s b l* 3 A02 — Copyright (C) 1991 Danish Building Research Institute
OFFICE —
new
get
edit
simu
resu
save
save
DOS
real
f icti
centr
optio
comme
cat en
1
2one
Corridor
South office
North office
Help FH

New F2

Corridor

Copy F3

6.30
Net area (m*)
DeleteF4

6.84
Gross area
Cm')
BeforeFS

3.08
Net height
(m)
After F6

19.4
•Volume
Cm3>
SelectF7

SURFACES ~

list F8
. S
Heating/cooling ~

Check F9

Solar distribut ~

IntroFlO

titut, Indeklimateknik, 2970 Hersholm
Fig. 2. Example of the menu in tsbi3 with the outer level to the left
going to the right.
INPUT
The global input description of tsbi3 consists of the following data: part of
building geometry, surfaces (walls, roofs, floors, windows, doors, materials of
walls, topology of the building site, equipment (electrical appliances, ligh-
ting, etc.) and HVAC-systems and user profiles (people, infiltration, venting,
etc), weather data.
The number of data, the user has to enter in order to define a building model
will typically ie between 100 and 250, according to the complexity of the building
and the installations. Most of the data can normally be found directly in the
building project material.
In connection with tsbi3 there have been developed standard libraries, where
the different materials, constructions, glasses, frames, type of people, sche-
dules and dayprofiles are defined. In addition to this the user can enter local
data for the specific building.
-------
3142
BuiIdingmodel:
OFFICE

ZONES...
net
gross
height
solar
surfaces
Name
area
m2
m
to-air lost
Corridor
6.30
6.84
3.08
0.20
0.05
Corr-wall 2-3

Coor-wall 1-2

Floor C

CeiIing C

Internal Beam
South office
15.40
17.10
3.08
0.20
0.05
Ext-walI S

Coor-wall 1-2

Side-walls S

Floor $

Ceiting S

Internal mass S
North office
15.40
17.10
3.08
0.20
0.05
Ext-watl N

Corr-wall 2-3

Side-walls N

Floor N

CeiIing N

Internal mass N
Fig. 3. General view of Che zones in the office model.
SIMULATION
The simulation of temperature and energy consumption is done by using a multi-
zone building model. The instationary heat conduction through the walls is
calculated by an implicit method in which the model itself determines the number
of nodes necessary for each construction. The heat balance for each room
includes the internal heat gain from people, electricity, lighting together with
the infiltration, window opening and the HVAC-systems. The heat balance for the
loads and HVAC-systems is done simultaneously.
OUTPUT
The output from the program provides indoor temperature and humidity, energy
balances, heating and cooling demands, internal loads, solar radiation, infil-
tration, venting by windows, ventilation, heat loss by transmission, electric
lighting, shading conditions, solar irradiation on the facades etc. All the
results can be written at different levels on hourly, daily, weekly, monthly or
yearly basis - either on the screen, the printer or a file for later processing.
It is possible to choose between tabular or graphical form. Fig. 4 shows a
graphical form of the outdoor and indoor temperature in three zones.
Beyond this the programme can outline the actual/present building model in
three detail levels, so that the user very quickly can make sure that the data
is correctly typed. The model documentation includes a number of helping quanti-
tites or intermediate results.
-------
3143
3D
Indoor tmp,C
Corrldor
20

1
Hour
10
16
20
0
4
B
12
24
Pm-1, OFFICE, Tuesday 5.6 19S0
Fig. 4. Graphical form of the outdoor temperature together with the in-
door temperatures in the three zones: South office, north
office and corridor.
GLASS BUILDINGS / ATRIAS
Based on experience with simulation of glass buildings using the two American
programmes BLAST and SUNCODE a simulation model has been set up in tsbi3 for
atrias. This make the program suited for simulation of glass buildings / atrias.
Infiltration and ventilation by opening windows are simulated using a
constant air change plus a contribution dependent on temperature and windspeed
that makes it well suited for the' calculation of high rooms with large thermal
impetus, e.g. atriums. A special module has been developed to calculate the air
change in atriums. Here the user can state the actual opening spaces for inlet
and outlet, and difference of height between them. This calculation module has
already been used by Raundahl & Jorgensen A/S, consulting engineers, in connec-
tion with the projecting of a glass pyramid at Viby Square, Arhus, Denmark.
Measurements carried out there in the summer 1990 have shown good accordance
with the calculations.
USER'S GUIDE
The user's guide (Johnsen, 1991) gives thorough instruction in the use of the
program in the most suitable sequence for the daily user. The manual starts with
a test example requiring active use of the keyboard, which thus allows the user
to familarize himself with the program's main functions and facilities.
-------
3144
To make the test example entirely specific, it starts with a description of
the building, the rooms, the installations, and the design assumptions, chosen
for the example. The user is then led through the example with the help of
instructions that describe all procedures step by step and show all the screen
pictures the user will see on his computer.
RUN TIME AND IMPLEMENTATION
The time used for simulation of a year for a one zone model on a PC/AT is from
10 minutes depending of complexity of the model.
tsbi3 is implemented on IBM-PC with MS-DOS (or possible OS/2), using Micro-
soft C (ANSI C). The program uses dynamic memory allocation, which secures the
optimum use of the computer capacity.
INTEGRATION OF BUILDING PERFORMANCE EVALUATION TOOLS
tsbi3 will take part in the EEC's R&D work programme of the COMBINE project,
which is a part of the JOULE-programme. COMBINE will perform a first step
towards future intelligent integrated building design systems. The COMBINE
programme will emphasize the integration of energy performed related aspects
into the building design process.
DAYLIGHT ANALYSIS USING SUPERLITE
Two links have been developed in IEA - Task XI between the daylight program
SUPERLITE and thermal analysis programs. tsbi3 has been incoorporated in this
daylight/energy analysis system. This is described in the paper "Daylight
analysis using SUPERLITE, SUPERLINK and tsbi3" in categori PS-3 at this
conference.
CONTACT ADDRESS
The computer program tsbi3 is available on diskettes together with weather data
from the reference year TRY and other auxiliary programs. Further information
about the program can be obtained at:
DANISH BUILDING RESEARCH INSTITUTE
Kjeld Johnsen, Karl Grau or Jargen Erik Christensen
Postbox 119
DK-2970 Horsholm, Denmark
Phone +45 42 86 55 33 Fax: +45 42 86 75 35
REFERENCES
Andersen, B. and co-workers (1982). Meteorological data for heating.
ventilation, sanitary and energy engineering. Danish reference year TRY.
SBI-Report 135. Danish Building Research Institute
Johnsen, K., K. Grau and J. E. Christensen (1991). User's Guide to the program
tsb!3: Thermal Simulation of Building and Instalations. Danish Building
Research Institute.
-------
3145
NUMERICAL SIMULATION AND ANALYSIS OF THE
THERMAL PERFORMANCE OF BUILDINGS
D. Christoffers, K. Jahn
Institut fur Solarenergieforschung (ISFH)
SokelantstraBe 5, D-3000 Hannover
Germany
ABSTRACT
In the context of developing energy efficient buildings the computer
proves to be a valuable tool. A building energy analysis computer
program (VITRUV) is presented whose main features are: a one zone
model with passive solar capabilities, an easy interactive entry,
default data sets including meteorological data and graphical
presentation of monthly energy needs. Computer runs allow testing
several alternative designs, facilitate sensitivity analyses,
parametric studies and the generalization of measured data. Often it
is not possible to determine the whole set of parameters needed for
a simulation. This may be due to the lack of sensors or to the
complexity of the problem. However, missing parameters (i.e. heat
transfer coefficients, storage capacities) can be extracted from the
experiments. Parameter identification methods are discussed.
KEYWORDS
Energy efficient buildings, passive solar design, thermal
performance, testmethods, simulation, parameter identification.
INTRODUCTION
The Institute of Solar Energy Research Hameln/Emmerthal has built
two experimental solar houses for test and demonstration of active
and passive energy saving measures. One house is used for
experiments, the other as a reference. Both houses are built
following basic principles of low energy architecture.
Computer simulations can be a valuable tool for energy calculations
and optimizations during successive design steps. This paper deals
with numerical simulations as a tool for
- planning and optimizing the system,
- analysing the experiments.
To estimate the heating energy demand as well as the thermal comfort
of the houses, calculations in accordance with the German
regulations DIN 4701 were carried out. They were completed by
several computer simulations, performed by means of a program
(VITRUV) developed in our institute (Schreitmuller, 1988) .
-------
3146
Different from DIN 4701, VITRUV also considers solar gains from
windows, transparent insulations (solar walls) and conservatories
(Christoffers and Schreitmiiller, 1990) .
The erection of experimental buildings allows measurement of the actual
impact of energy saving measures. A performance analysis gives
information about the heat loss coefficient by transmission (U-
value), the ventilation rate, the solar heat gain factor (g-value),
heat distribution, loading and uaLoading of storage capacities,
thermal convenience or cost effectiveness. Another important aim is
to validate simulation programs. Experiments and simulations should
be iterative steps, combining monitored and simulated results. The
main thermal parameters of the building can be determined by means
of parameter identification methods. With validated programs a
generalization of the results can be achieved.
SIMULATION PROGRAM
VITRUV is a one zone model with passive solar capabilities, an easy
interactive entry, default data sets including meteorological data
and graphical presentation of monthly energy needs. A program run
consists of the following steps:
- interactive compilation of an input file
- generation of a synthetical climate
- estimation of the irradiation on the envelope
- calculation of energy flows
- output of data as files or graphs
- list of possible variations
Input file
An appropriate solar design needs accurate knowledge of likely
meteorological data at the particular site. VITRUV uses monthly
values of mean temperature and of solar radiation to generate a
synthetic climate e.g. hourly values of temperature, of direct
beam sunlight and diffuse light. Monthly values are available for
more than 350 cities all over the world.
Energy load and comfort of a building are mainly influenced by the
outside shell. The appropriate parameters can be easily specified by
the user, eg. areas of walls, roof and windows, material constants
(U-value, transparency) or orientation of the building.
The thermal behaviour of inner rooms is affected by furniture,
carpets, curtains etc. Their modelling can not be so detailed in
simplified programs. VITRUV consideres only one zone with a rough
estimation of effective thermal capacities. The ventilation rate
must be estimated by the user. Day/night temperatures are to be
specified.
The program has passive solar capabilities. Direct gains,
conservatories and transparent insulations (solar walls) can be
added to the layout. The shading (percentage of irradiation blocked
from the envelope) is considered.
-------
3147
The creation of the input file can be simplified using default data
sets, there are three kinds of houses to choose from. Input files
can be stored or printed.
The program calculates monthly energy needs and the number of days,
on which indoor temperatures exceed the comfort level. The data are
presented either in form of data files or graphically (Fig. 1).
Fig. 1. Monthly energy needs of a single family
house. Living area 180 m2, volume 432 m , thermal
capacity 17.0 kWh/K, internal loads 10 kWh/day,
minimum temperatures 20 °C / 16 °C (day/night).
Site Hameln (global irradiation 953.4 kWh/m2year,
mean temperature 8.7 °C) . Variations: reduction of
U-values of the outer walls from 0.4 W/m2K to 0.25
W/m2K (second bar), additional reduction of the
ventilation rate from 0.7 Vol/h to 0.5 Vol/h
(third bar).
PERFORMANCE ANALYSIS
Monitoring
The following types of data are to be recorded: climatic data,
temperatures, heat flows, air change rates, additional heating and
occupant effects. The data acquisition is the basis for an
assessment of the passive solar features realized in this project.
Heart of the system is a Helios interface from Philips/Fluke. Helios
gathers data from the sensors and generates control signals for
peripheral devices (i.e. heating, ventilation).
Testmethods
Output
4.000
£
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year
The different test evaluation techniques can roughly be categorized
into three groups (van Dijk, 1990): integrated absolute approach,
transient comparative approach and parameter identification. The
-------
3148
first one is a sort of statistical analysis, - where test data are
averaged over a large time interval. The second permits the analysis
of the instantaneous differences in measured heat fluxes, but two
buildings are needed which are identical except for the components
under investigation. This paper confines to the third method, the
parameter identification method.
40
35
30
25
Plate 1
Plate 2
20
15
2.500
3.000 3.500
0
500
1.000
1,500
!.000
TIme/s
Fig. 2. Temperature versus time measured at the
two plates of the U-value meter
300
250
| 200
.Fly* t.mMMure0
150
Flu* 2 measured
"Fiorteafeutawcr
Flu* 2 calculated
3.000 3.500
1.500
2.000
2.500
500
1.000
Tlmo/8
Fig. 3. Heat fluxes from the hot plate to the
sample (flux 1) and from the sample to the cold
plate (flux 2) in comparison to the calculated
curves. The sample is a floor tile.
The parameter identification algorithm needs a mathematical model
for the test object and its environment. At the beginning the values
of the model parameters, for example thermal conductivities and
capacitances, are unknown. The problem is to compute those estimates
of the parameters which will give the best fit between the
experimental data and the calculated ones.
The most common techniques are the steepest descent method and the
Taylor series method. A discussion is given by Cools, Giguel and
Neirac (1986). While the steepest descent method has a slow
convergence after the first few iterations, the Taylor series method
-------
3149
often runs aground because of divergence of the successive iterates.
Marqard (1963) uses a combination of the above described methods,
trying to benefit of their advantages, while avoiding their
drawbacks. In this paper the Simplex method (Nelder and Mead, 1965)
is presented, which has a wide application, since it does not use
derivatives and can be programmed very easily.
The parameter identification procedure shall be illustrated with a
simple example, i.e. the determination of the thermal conductivity
and the heat capacity of a floor tile. For a measurement, the sample
is placed between two copper plates. The upper plate is heated
electrically, the temperature of the lower is controlled by a
thermostate. There are two heat flux meters between the plates and
the sample. A personel computer controls the temperature settings
and records the data from the temperature sensors (Fig. 2) and the
heat flux meters (Fig. 3).
The sample is homogeneous, the length and the breadth are large
compared to the thickness, so the dimensions perpendicular to the
direction of the heat flux can be considered as infinite. Heat
losses over the lateral surfaces can be neglected. With these
assumptions the mathematical model consists of the one dimensional
differential equation:
With a = A/pc, A, representing the thermal conductivity, p the
density and c the specific heat capacity. T, t and x are standing
for temperature, time and position, respectively. The numerical
solution can be found using finite differences. In dividing the xt-
plane into a network, different increments on x and t are chosen:
At=k and Ax=h. The transformation of (1) in two difference equations
yields:
from which we can calculate the temperatures T(x,t+k) and the heat
flux densities q(x,t+k), once we know the boundary conditions, i.e.
dT/dt = a d2T/dx2
(1)
T(x,t+k)=T(x,t)+[q(x-h,t+k)-q(x,t+k)]k/(cph)
q(x,t+k)=[T(x,t)-T(x+h,t)]X/h
(2)
(3)
1,4
2"
E 1,3
1,17

993
—! 1 I 1 I I I
950 960 970 980 990 1000
Heat capacitance / (J/kg K)

Fig. 4. Parameter identification with the Simplex
method. See text for explanations.
-------
3150
the start temperature of the whole arrangement and the time courses
of the plate temperatures.
The iterative steps of the simplex method are depicted in Fig. 4. It
is assumed that the respective differences (sum of squares) between
measured and calculated heat flux densities are plotted over the em-
plane. The problem is to find the deepest point within this
landscape. Three pairs of estimates form a triangle to start from,
the so-called simplex (shaded). The simplex moves towards the
minimum by three basic operations: reflection, expansion and
contraction. The vertix with the highest value of the sum of sqares
is replaced at each iterative step. The algorithm stqps when the sum
of sqares is under a given value.
Identified parameters can be used by design tools like VITRUV. With
the help of simulation programs, it is possible to generalize the
data extracted from field experiments for broader applications.
CONCLUSIONS
Simplified design tools as VITRUV are considered to be much more
adapted to the user than the detailed models. When models are simple
many runs can be made, allowing the user to test several
alternatives. They offer an assessment of the energy performance of
a building, without a particular knowledge in energy calculations.
Simulation programs must be validated by experiments. These programs
enable us to generalize the results achieved in test buildings.
Experiments and simulations should be iterative steps, combining
monitored and simulated data. The results obtained from heat flux
measurements and the investigations of other authors show that the
main thermal parameters of buildings can probably be determined by
means of parameter identification methods.
REFERENCES
CHRISTOFFERS, D. and K.R. SCHREITMULLER (1990). Vitruv - a user
friendly thermal design tool for architects. In: Field monitoring
for a purpose. Volume II. IEA Workshop, Chalmers University of
Technology, Gothenburg, Sweden, 49-52.
COOLS, C., GICQUEL, R. and F.P. NEIRAC (1989). Identification of
building reduced models. Application to the characterisation of
passive solar components. Int. J. Solar Energy. Vol. 7. 127-158.
Marquard, D.W. (1963). An algorithm for least squares estimation of
nonlinear parameters. SIAM Journal. Vol. 11. 431-441.
Nelder, J.A. and R. Mead (1965). A simplex method for function
minimization. Comp. J.. Vol. 7. 308-313.
Schreitmiiller, K.R. (1988). User-friendly design tools.for solar
energy applications. In: Proceedings of the 6th International
Solar Forum, Vol.1. DGS-Sonnenenergie Verlags-GmbH, Miinchen, 93-
100.
van Dijk, H.A.L. (1990). The CEC PASSYS outdoor testfacilities;
developement of a common testmethod. In: Fieldmonitoring for a
purpose. Volume II. IEA Workshop, Chalmers University of
Technology, Gothenburg, Sweden, 49-52.
-------
315 L
PASCAUD
Passive Solar in Computer Aided Urban Design
Ton Trijssenaar1
Rijksweg 436
2071 CT Santpoort
Holland
ABSTRACT
PASCAUD is the result of the study Passive Solar in Computer Aided Urban Design,
sponsored by the European Commission's DG XII and the Netherland's Agency
for Energy and Environment. It is an application program that enables a seamless inte-
gration of solar criteria into the computer aided design process of high density housing
projects.
The paper is an abstract of the Final Report of the PASCAUD study. It gives a global
description of the usage of the software, and it summarises both the background of the
project as well as the user oriented aspects of the software development process.
KEYWORDS
Large scale high density housing projects; automated passive solar design tools;
computer aided urban design; user oriented software development.
INTRODUCTION
One of the most substantial results of last decade's energy research is the integration of
passive solar concepts in the architectural design of houses and buildings. Yet, to facili-
tate its large scale application, solar architecture will in turn have to be integrated in the
regular design practice of towreand cities. Since 1981, the municipality of
Haarlemmermeer, Holland, pioneered this urban approach by applying the solar
concepts on some of its large scale housing projects. Based on these experiences it
started in 1985 to help develop the study PASCAUD, Passive Solar in Computer Aided
Urban Design. The resulting PASCAUD-software is an application program within a
1 PASCAUD is based on work done by the project-group PASCAUD consisting of:
B. Welschen, Inpark, Dillenburgsingel 69, Leidschendam (contractor);
E. Sonnemans, Van Heugten Consultants, St.Annastraat 145, Nijmegen (sub-contractor energy);
H. Pluckel (chairman) and
B. Jansen, Municipality Haarlemmermeer, Nieuweweg 65, Hoofddorp (sub-contractor urban design); and
A. Trijssenaar (secretary), Province of North-Holland's Energy Office, Houtplein 33, Haarlem (project
coordination).
The study has been carried out under contract with the Commission of the European Communities, D G
XII, and the Netherland's Agency for Energy and Environment. The contribution of the Province of North-
Holland has been in addition to the contract.
-------
3152
residing CAD-package. This combination enables a seamless integration of solar
criteria into the computer aided design process of high density housing projects. Given
the Intergraph CAD-systems, that Haarlemmermeer had either in use or in order prepa-
ration in 1985, the current PASCAUD version works under UNIX within Intergraph's
CAD-package MicroStation. As a result of the PASCAUD-study though, the software
design structure is such that most of the program can eventually be converted to other
CAD-systems.
SUNVILLE
In Haarlemmermeer, passive solar is indissolubly linked with extra insulation & draught-
control, efficient heating & ventilation and optimal solar access. A feasibility study on the
large scale use of this concept was initiated in 1981.2 This was followed (Hensen
1987) by an experimental 'Energy Park' of 56 houses in Haarlemmermeer's principal
town, Hoofddorp ("Mainville"). A subsequent demonstration project of 275 houses was
commissioned in 1984 in "Overbos 8", one of Hoofddorp's townjbxtensions.3 The
evaluation of this "Sunville" demonstrated an overall energy conservation of 63%
against marginal extra investments.
The south orientation of more than 70% of the houses proedto lead to substantial extra
savings, even in these very well insulated houses. The net solar contribution of the
energy balance is 39% in south houses and 21% in west houses. As a result, south
houses use 20% less auxiliary heating than west houses do. Additionally, the south
houses score significantly higher in thermal comfort, both in winter and in summer
situations. Furthermore, the 5000 m2 south facing roof area of the project can support
future options like solar power from photovoltaic cells.
USER ORIENTED SOFTWARE DEVELOPMENT
The need for the PASCAUD-study stems from the complication that solar criteria posed
on the urban design process of the Overbos 8 project. It showed that solar architecture
asks for iterative loops between the designs of both the urban designer and the
architect. Unlike the conventional top-down process, architectural details like glazing,
roof shapes and shadow from other buildings are to be considered at the urban design
level, rather than at the architectural level only. Consequently, the design process of
Overbos 8, though very successful in energy terms, demonstrated a necessity for CAD-
support in subsequent large scale passive solar projects. But, while the need for the
PASCAUD-study stems from the the complications that solar posed on the urban de-
sign process, its feasibility was largely based on the (Intergraph) computer systems, the
municipality had either in use, or in order preparation in 1985. And even this sophisti-
cated machinery could not have met the requirements of the urban design unit without
the availability of software development tools that enable the close involvement of the
end-user.
Extending the solar projects to a solar CAD-system would normally mean that computer
requirements start to take over from what was thought to be required in the field. More
often than not this will ignore the carefully built knowledge-base of the end-user. Only
recently have methods and techniquesslipped through that enable the involvement of
the end-user. (Turner 1988) Bringing this "user oriented development" to its full poten-
tial,though, are the automated software development tools, based on the standard
2 Passieve zonne-energie in de Haarlemmermeer. NOVEM, Utrecht 1982 (in Dutch)
3 Overbos 8. European Commission's Monitor Series Issue Nr.39, ECD, London 1989
-------
3153
methods and techniques.4 Largely because of the use of these tools, PASCAUD now
not only meets the specifications of one of Europe's more complex urban design
environments. In pursuing user oriented methods, techniques and tools, a process is
explored that enhances the reusability of energy software modules. The ESPRIT-
project of the EC regards this as analogous to component reusability in the
manufacturing industry and of both cost-cutting and quality enhancing value.
GRAPHICAL NICHE
On a CAD-system, the graphical objects of a site-plan drawing are essentially "icons"
that represent hidden layers of information. And while the layers underneath are being
filled with details as the design moves on, the icons, or objects make only subtle
changes. In this way the urban designer can concentrate on the spatial distribution of
the objects without having the drawing being cluttered with details from other
disciplines.
The primary objective of the PASCAUD-study was to create a solar energy niche on the
site-plan drawing, rather than having to fall back on a separate energy layer, or worse,
on an energy expert. The solutions, incorporated in PASCAUD are the "Shadow
Contours Module"; a graphical representation of the sun's shadow contours, and the
"Energy Report Module"; an ability to print out an energy report of a design alternative
just by pointing and clicking on the desired area of the site-plan drawing.
Customary to CAD, all houses and buildings on the drawing will have been copied from
a library of generic objects, complete with all energy properties attached to it. To
enhance modifications of a library's energy data, PASCAUD provides two "Data Entry
Modules", a graphical one and a menu driven one. This will furthermore simplify a
hitherto tedious and error prone task.
SHADOW CONTOURS MODULE
Simply by pointing and clicking on the desired area, PASCAUD attaches shadow
contours to the objects of the site-plan drawing. The sun's position that the shadow
represents can be selected for any plane, latitude, time and day of the year, either by
default settings, or by overwriting the last used values that appear on the screen. The
objects can than be rearranged on the drawing, using the attached shadow contours as
a yardstick for the effects that sunlight may have on a design alternative. It enables the
urban designer to consider qualitative aspects of sunlight much more easily than was
previously possible. It also enables an approximation of the quantitative effects that the
sun might have on a design alternative. This narrows down the need for a printed
energy report to only the most finalized design alternatives.
In a design's presentation stage, the software of the residing CAD-system can be used
to turn the shaded objects into a colourful "hidden- line" drawing, giving a realistic view
of the sunlight's effects, rather than an "artist impression" like most CAD programs do.
ENERGY REPORT MODULE
The energy report will show information of all the houses of a selected area ranging
from window orientations to space heating needs and from total energy use to total
generating capacity that has to be installed for the area. To get this on file requires the
4 BLUES Software Development Tools, Interprogram B.V., Wildenborch 3, Diemen, Holland.
-------
3154
urban designer not much more than to point and click on the desired area of the site-
plan drawing. But this user-friendliness comes with a price, for it takes the computer
fifteen minutes of background processing for up to fifty houses, to five hours for a project
of 200 houses. This still gives an advantage over previous conditions where energy
calculations were virtually inaccessible to the urban designer. However, as most solar
energy targets will be met by visual approximation with the shadow contours module,
the urban designer will hardly need an energy report other than for exploring "generic"
allotments and for final comparisons of project alternatives.
The relative infrequent use of the energy module implied that it could be made versatile
enough to facilitate both, the global urban design analysis, as well as the detailed archi-
tectural analysis. This approach also opened the way for future extensions to even
more comprehensive energy calculations.
ENERGY CALCULATIONS
The PASCAUD-study accentuates a distinction between urban design, which mainly
influences the exposure of buildings to sunlight, and architectural design, which mainly
influences the actual use of the received sunlight. So from a top-down, urban de-
signer's point of view, PASCAUD would only be required to output the solar irradiation
values that the buildings receive, leaving calculations of actual fuel use to the architect.
To accommodate all involved disciplines though, the irradiation values are to be further
processed by an energy model to trace the effect that changes of irradiation might have
on the energy bill. Yet it remains a prime requirement for PASCAUD to accurately simu-
late the supply of solar energy to the buildings, regardless of how the architect uses the
energy information, or whatever energy model is going to be attached to it later on.
In order to react to changes of solar irradiation, the energy calculation models need in-
put on the amount of collected solar energy. This is calculated by a solar irradiation
model.
After some unsuccessful efforts with built-in irradiation programs of established energy
models, PASCAUD now has an interface with the dedicated irradiation model SIBE
(Van der Voorden 1985, 1988). This in turn has been coupled with the correlation
model TCM-Heat (Van Dijk 1987)5. The output of this is finally put in an energy report.
The advantage of this arrangement is the versatility and accurateness of the dedicated
irradiation model. The disadvantage is a considerable response time instead of the
preferiBdinstant response to changes. This meant that the energy calculations had to
be assigned to batch processing, which make things even longer. Such response times
will be acceptable though, since most of the urban-solar criteria will be met with the
shadow contour module. And since land-use has higher cost constraints than fuel use,
most of the design cycling will go between the shadow module and a separate "Land-
Use Module", working in cooperation with PASCAUD. This leaves the use of the energy
module for experimental site layouts and for the final stages of a design alternative.
GRAPHICAL DATA ENTRY
The implemented irradiation model makes use of 'data points' for which the solar irra-
diation is calculated. In PASCAUD, each of these points represents a part of a window
and in order to provide the model with the location of these points, a graphical data en-
try module have been developed. This module builds upon the 3D drawing of the
5 Weather data: CEC publication Test Reference Years TRY, Weather Data Sets for Computer Simulation
of Solar Energy Systems and Energy Consumption in Buildings, CEC DG XII, 1985.
-------
3155
house, consisting of closed polygons of floors, walls, roofs and windows, all on their
own level, of which 8 levels of the residing CAD-program are assigned to PASCAUD.
From the 3D drawing of one selected house, the module puts every line-point (vertex) of
the drawing in a sequential order, and it attaches energy properties to the roof, wall and
window elements of the drawing.
After initiation, a movable 3D reference point (star) appears in the centre of gravity of
the selected house. From this point, the module registers the sequential order of all the
vertices of all the windows. Each window polygon will then automatically move to the
outside to indicate its recognition as window. A sequence of prompts follows next in
which energy properties are being attached to each roof, wall and window. In this se-
quence, one element after another is illuminated on the drawing,while a prompt-win-
dow appears in which values are to be typed in on thermal conductivity, absorbtance
and transmittance. The designer is then prompted to indicate both the main entrance of
the house and its main window area. Having been prompted for a unique type-name
of the house, the information is then grouped as a 'cell', but not before a final check is
prompted on the correct level, colour, line thickness, etc. Finally, the cells become part
of the Project Library from which the urban designer can select the required house
types and distribute them over the site-plan drawing in as many copies as is necessary,
without having to enter any new data again.
In principle, PASCAUD is now ready for use in almost all the phases of the urban de-
sign process.The PASCAUD package is operational on a VAX/Intergraph combination,
working under VAX/VMS, with two interacting screens, of which one is a colour screen.
The system, that the municipality purchased in 1989,though, is an Intergraph (InterAct
3050) with a considerably improved performance, but working under its own version of
UNIX instead of VMS. The municipality has commissioned a conversion,however,to
make PASCAUD able to work under Intergraph's UNIX and its CAD-program
MicroStation.6
An interesting question in this respect, is how PASCAUD can be converted to work
under the MicroStation versions for both the PC and the Macintosh. Possibilities should
also be examined to integrate PASCAUD into CAD -packages of different vendors.
Thought should also be given to the idea of making the shadow contours module
available for use in all existing CAD-programs.
An urgent step, however, is the start of a well documented application of PASCAUD in
the development of the next passive solar project in Haarlemmermeer, as this
municipality moves on to its next town^xtention of 15000 houses.
6 Inquiries on the software can be obtained from:
Inpark. B.V. Tel: +31 70 111433. Fax: +31 70 111159
Dillenburgsingel 69
2263 HW Leidschendam
Holland
-------
3156
PASCAUD PROJECT DOCUMENTATION
- Project Plan PASCAUD, 1985
- Definition Study Report, 1986
- Global Design Report. June, 1987
- Technical Documentation, July 1989
- Users Manual, July 1989
- Final Report, September 1989
- Delivery Report, May 1991
REFERENCES
Hensen, J.L.M.(1987). "Energie Proeftuin": Results of an experiment on low energy
housing in the Netherlands. Proceedings European Conference on Architecture.
Munich.
Van Dijk, H.A.L. and Arkensteijn, C.A.M. (1987). Windows and Space Heating
Requirements; Parameter studies leading to a simplified calculation method. The
Netherlands National Report on Step 5 of lEA's Windows and fenestration
programme. TNO Institute of Applied Physics TPD, Delft.
Van der Voorden, M. (1985). Determining the rate of solar irradiation in the built
environment. Proceedings ISES Conference . Montreal.
Van der Voorden, M. (1988). Advantages of a non-ray tracing methodology for direct
solar irradiation calculation. Proceedings North-Sun Conference. Borlange.
Turner, W.M. (1988) System Development Methodology. North-Holland, Amsterdam
-------
3157
EVALUATION OF LEARNING MODELS FOR BUILDING ENERGY SIMULATION
L. Jankovic
Intelligent Buildings Unit, Birmingham School of Architecture,
Faculty of the Built Environment, Birmingham Polytechnic, Birmingham B42 2SU, UK
ABSTRACT
The paper explores a new approach to building energy modelling and simulation by means of
learning models. A Learning Model was developed and subjected to a number of numerical
experiments, using data from monitoring of real buildings. The Learning Model was compared
with a conventional model, and its accuracy was found to be higher or equal to that of a
conventional model, while computing time was shorter by a factor of 60.
Learning models can already be used for evaluation of building energy consumption, for building
energy rating on the basis of short term monitoring, and for predictive control applications.
Although in this early stage the use of learning models is limited to existing buildings, there are
indications that they can comprise an important part of a future learning design database.
The aim of this paper is to draw attention to this emerging new method for building energy
modelling, simulation and control, and to demonstrate the necessity for further research.
KEYWORDS
Building simulation; machine learning; energy fingerprints; simulation accuracy; learning models;
conventional models; building design; building control.
INTRODUCTION
The main obstacle to progress in the field of building simulation are conventional methods for
modelling comprehensive systems. These methods require that the system being modelled is
divided into elements, that the elements are described with equations based on fundamental
principles, and that the overall system model is assembled from the element models. When the
system comprehensiveness increases, the conventional modelling methods have a consequence of
increased requirements for computer processing power and for high performance machine
environment by conventional simulation models.
The number of calculations required to simulate a building using conventional methods can be
very large. For example, our annual simulations of environmental performance of a five storey
office building, consisting of ten zones, lasted approximately 8 hours on a SUN 3/60 workstation
running ESP simulation program by Clarke and McLean (1989). This clearly means that
conventional models have become largely inefficient.
Earlier attempts to produce simplified and accurate simulation models were not fully successful.
Virk and co-workers (1989) used results from monitoring of a test cell to establish the relationship
between inputs and outputs of a stochastic model. Penman (1990) developed a five parameter
model and established a relationship between inputs and outputs on the basis of short term
monitoring of a building. And while the work of the former required high computing power for
-------
3158
stochastic modelling, and the latter lacked flexibility because of the fixed parameter set, both
models did not recognise different heat transfer mechanisms through which solar gains, casual
gains, and heating system input drive building thermal performance. Both models were developed
without relating their performance to data from longer term monitoring of real buildings, and this
resulted with reports of problems associated with them.
This paper reports on development and evaluation of a different, and perhaps more successful
Learning Model. Origins of this work were reported by Jankovic and co-workers (1987), when
building thermal properties, such as effective thermal capacitance, effective conductance-area
product, and the time constant were determined from a series of dynamic heating tests of a
building, and were subsequently used in a simplified model. The main idea of this approach was to
use a simplified building heat balance, and through experimental approach determine effective
parameters which would make the simplified balance accurate. This idea was further developed by
Jankovic (1988), and Jankovic and Jesch (1989), while the implementation of the learning features
by Jankovic (1991) resulted with a self contained Learning Model. The model was given
flexibility, achieved by means of a variable number of parameters in the heat balance equation. It
became apparent that this Learning Model would not only be useful for simulation, but also for
control, and, in the future, for design.
The results of numerical experiments presented here show that the Learning Model can be more
accurate and approximately 60 times faster than a conventional model of the same building.
Simplicity of the Learning Model makes its operation possible even on low cost computers. Such
performance characteristics open new doors for building energy simulation and control, and for
research in learning design tools.
The Learning Model of building energy performance was developed on the basis of long term
experimental data. The driving functions for building environmental performance and the
building output had been monitored beforehand, and weather data and building room temperature
data had been acquired on an hourly basis. During the monitoring period of five months the
monitored building a three bedroom family house, had been unoccupied and unheated.
Actual building energy properties, such as overall conductance-area product, effective thermal
capacitance, effective solar aperture, time constant, and others, are always different from their
theoretic equivalents, as shown by Jankovic (1987, 1988). These actual values, if they can be
determined, make the performance of simplified heat balance simulation models more accurate
than that of general simulation models (Jankovic, 1989). And while special tests had to be carried
out in order to determine the actual building thermal properties before (Jankovic, 1987),
relationships between them can now be found through machine learning, on the basis of short term
monitoring, as reported by Jankovic (1991). These relationships are numerical expressions of the
essence of building energy performance, and hence the name 'building energy fingerprints'
(Jankovic, 1991). The actual building energy properties can then be calculated from the latter.
In the process of the Learning Model development, the building heat balance was expressed as:
LEARNING MODEL OF BUILDING ENERGY PERFORMANCE
Ceff * dT/dt = -UAf* (T-Ta) + Aeff * S
(1)
where
S
T
Ta
Aeff
UAeff
Ceff
dt
dT
dT/dt
Effective aperture of the building with respect to solar radiation;
Effective thermal capacitance of the building;
Time-step;
Room temperature differential;
Room temperature time derivative;
Solar radiation on the total window surface;
Room air temperature of the building;
External air temperature;
Overall conductance-area product for the building envelope;
-------
3159
For the purpose of determining building energy fingerprints, the building inputs and outputs,
known from monitoring over a period of time, were 'frozen' and considered as a series of
constants. Building thermal properties were unknown and the objective was to find them, as the
numerical values of these unknowns would fully define the relationship between the inputs and
outputs. In order to produce more appropriate expressions for achieving this, both sides of
equation (1) were divided by Ceff and the resulting equation
dT/dt = - UAeff/Ceff * (T-Ta) (2)
was subjected to the substitution
UAeff/Ceff = 1/kl
Aeff/Ceff= l/k2 (3)
which completed the process of rearranging equation (1) into
dT/dt = - 1/kl * (T-Ta) + l/k2 * S (4)
where kl and k2 were building energy fingerprints. The building room temperature was calculated
as
Tj = Tj-l + (dT/dt)j-i * dt (5)
where
j Monitoring time in hours;
Tj Predicted temperature of the building j hours after the beginning of monitoring.
Errors between the prediction and measurement were expressed as
j=N
S = LI (Tj-Tm j) I (6)
j=2
where
Tm,j Measured room air temperature of the building j hours after the beginning of
learning;
N Duration of the learning period.
As Tm,j in equation (6) was known from monitoring and Tj was a function of k] and k2, the error
S was also a function of kl and k2:
S=f(kl,k2) (7)
Learning of building energy fingerprints was achieved by means of minimisation of errors
expressed with equation (7). The minimisation was carried out by a modified multidimensional
minimisation routine based on the Downhill Simplex Method by Press and co-workers (1986). The
results of learning were building energy fingerprints kl and k2-
Equation (7), incorporating equations (4), (5) and (6), and the minimisation routine comprised the
Learning Model.
NUMERICAL EXPERIMENTS WITH THE LEARNING MODEL
The Learning Model was subjected to a series of numerical experiments using hourly data from
five months of monitoring of a three bedroom family house. The objective was to run parallel
simulations of the same house with the Learning model and a conventional model, and to
compare the results in the end. The task of the Learning Model was to find the values of building
energy fingerprints kl and k2 from equation (3), while searching for the minimum of equation (7),
and subsequently to simulate the building using the results of learning. The whole process was
driven with data from monitoring. The Learning Model software was written in C language, and
the experiments were carried out on an IBM PS/2 computer.
-------
3160
surface
. Measured internal air temperature
— Measured external air temperature h\
i; • «•
:'VV
r ¦ • • * yv1/. v* V-
\i\hill\llhhi\i \i\s \i\fh)\h^j\i
i 1 n
mhim/M \/)i \/\f y s!\!\!w
120 122 124 126 128 130 132 134 136 138 140 142 144 146 148 150
DAY
Fig. 1. Simulations of a family house with Learning Model and TRNSYS
1300
1200
1100 y
1000 §
• 900 -
• 800 o
¦ 700 §
• 600 q
• 500 £
This three dimensional minimisation problem has a scenery of a numerical valley surrounded by
numerical mountains, where the variables ki and k2 from equation (3) and S from equation (7)
correspond to X, Y, and Z coordinates of a Cartesian coordinate system. The values to be learned
are the coordinates of the minimum of function S. The minimisation routine searches for these
coordinates by rolling the geometric figure Simplex from the mountain top to the bottom of the
valley.
Comparison Between the Learning Model and Conventional Model
The comparison of results of parallel simulations with the Learning Model and with a
conventional model TRNSYS by Klein (1988) and co-workers is shown in Fig. 1, where better
performance of the Learning Model becomes apparent.
Figure 2 shows the comparison of accuracy of the two models when the learning period is equal to
the simulation period. It appears that the Learning Model outperforms the TRNSYS model during
all time steps. It can be seen from Fig. 2 that errors of temperature predictions by TRNSYS model
are less than or equal to loC during 46% of the simulation time, while errors of temperature
predictions by the Learning Model are less than or equal to loC during 86% of simulation time.
The situation changes if the learning period is considerably shorter than the simulation period.
Accuracy of the Learning Model is now reduced, but it gets higher as the length of the learning
period increases (Fig. 3).
It was also noticed that the values of building energy fingerprints kl and k2 changed with time.
Figure 4 shows discrete points corresponding to energy fingerprints obtained through the learning
process for every month of the five month period. This change seems to be consistent, and
exponential, as the curves resulting from the curve-fitting process show in Fig. 4. When the results
of curve fitting were used back in the Learning Model to change the values of building energy
fingerprints on a monthly basis, the performance of the Learning Model improved, so that it
became more accurate than TRNSYS model, even outside the learning period (Fig. 2).
Machine time for execution of the simulations by both models was also measured and compared.
And while the conventional model required approximately 2 hours to complete 5 months of hourly
simulations, the Learning Model required approximately 2 minutes to do the same. This indicated
that the Learning Model was approximately 60 times faster than the conventional model.
-------
3161
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

•*

1 1 1
/

T
RNSYS
J

f /

0.0 1.0 2.0 3.0
TEMPERATURE ERROR [C]
Fig. 2. Comparison of simulation accuracy inside the learning period
4.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0

^ """"

• ** ^'
~ *
**

~
/' z'

/•' /"
n *

TRNSYS
Learning Model (56 days of learning)
Learning Model (28 days of learning)
Learning Model (14 days of learning)

^ +'
¦'V

jT''

Learning period commenced in February [
Simulation period: 1st February to 30th June |

0.0 1.0 2.0 3.0
TEMPERATURE ERROR [C]
Fig. 3. Comparison of simulation accuracy outside the learning period
4.0
3400
3200
3000
2800
2600
2400
2200
2000
1800
1600
1400
1200
1000
800
600
400
200
0
a kl
k2
--kl =331.5 xe"a38xtime
k2 = 4192.1 x e "°-30 x time
Feb
Apr
MONTH
Fig. 4. Seasonal changes of building energy fingerprints
Jun
-------
3162
IMPLEMENTATION OF LEARNING MODELS IN BUILDING DESIGN TOOLS
If the Learning Model discussed here is not only restricted to building environmental properties
but is extended to building structure specification, geometry, and material specification, and if the
results of learning are handled by a database with learning capabilities, then such database would
become a powerful design tool, once it gains enough artificial experience.
Each building designed by such a database would be monitored continuously. Costs of monitoring
are expected to go down in the future, because of better communication systems and fibre optics
technology. The intelligent design database could therefore have a continuous feedback from the
buildings, throughout the building lifetime. This would make it possible to investigate
consequences of particular designs and would result with improvements of future designs. The
value of such a database would grow with time, as more artificial experience is gained from new
projects. This would dramatically change the way in which buildings are designed, it would
contribute to savings of investment, running and refurbishment costs of buildings, and it would
result with better buildings for building users.
CONCLUSIONS
Conventional simulation models are becoming increasingly inefficient, as they attempt to simulate
more complex buildings, and in doing so the simulation time can extend to several hours. For this
reason we tried a totally different approach to simulation, and produced a Learning Model.
The performance of the Learning Model was compared with that of a conventional simulation
model, and both were compared with data from long term monitoring of an existing building. The
Learning Model was found to be more accurate and almost 60 times faster.
At present, the use of learning models is limited to existing buildings, and is subject to availability
of monitored data necessary for the learning process. Despite this limitation, learning models can
already be used for prediction of building energy consumption, for building energy rating, and for
predictive control applications.
The simplicity, high speed and accuracy of the Learning Model makes a future development of a
learning building design database possible, indicating the necessity for further research, and
opening new doors within and beyond the boundaries of building simulation, control, and design.
REFERENCES
Clarke, J. A. and D. McLean. (1989). ESP - A Building and Plant Energy Simulation Program.
ABACUS Simulations Ltd & ESRU - University of Strathclyde, Glasgow.
Jankovic, L. (1991). Building Simulation Models Which Learn. Proceedings of CIBSE National
Conference, Canterbury, 7-9 April 1991. CIBSE, London.
Jankovic, L. and L. F. Jesch (1989). Evaluation of Mathematical Models on the Basis of
Monitored Results. Proceedings oflSES Solar World Congress Kobe 1989. Pergamon Press.
Jankovic, L. (1988). Solar Energy Monitoring, Control and Analysis in Buildings, Ph.D. Thesis,
University of Birmingham.
Jankovic, L„ T. W. Greeves, and L. F. Jesch (1987). Dynamic Heating Test Analysis for a
Building. Proceedings oflSES Solar World Congress Hamburg 1987. Pergamon Press.
Klein, S. A. et al. (1988). TRNSYS - A Transient System Simulation Program. Solar Energy
Laboratory, University of Wisconsin.
Penman, J. M. (1990). Second Order System Identification in the Thermal Response of a
Working School. Building and Environment. Vol 25, No. 2, pp. 105-110, Pergamon Press.
Press, W. H., B. P. Flannery, S. A. Teukolsky, and W. T. Vetterling (1986). Numerical Recipes.
Cambridge University Press.
Virk, G. S., D. L. Loveday, K. I. H. Alkadhimi, and J. M. Cheung (1989). Advanced Control
techniques for BEMS. Proceedings of the First International Congress on Condition
Monitoring and Diagnostic Engineering Management 89. Kogan Page, London.
-------
3163
TRANSPARENT INSULATION MODULE FOR TRNSYS
L. Fulop, L.F. Jesch and S. Gllanl
Solar Energy Laboratory
University of Birmingham
Birmingham B15 2TT, UK
ABSTRACT
The TRNSYS simulation program developed at the University of Wisconsin - Madison is a very
powerful tool for the simulation and analysis of many solar systems. Its modular structure makes it
also usable for completely new applications. One such new applications is Transparent Insulation
(Tl).
There are different types of Tl systems with different features. A program module developed in the
Solar Energy Laboratory at the University of Birmingham was written first of all for honeycomb type
Tl, but it is also usable for other type which can be characterized by the input data required by the
program module including the transmittance functions.
The reason why Tl requires a special module is that the incidence angle dependent transmittance
and absorptance are different from ordinary glazings. It also has an important wavelengths
dependence.
The program module can be used for simulating a solar collector, a mass wall or a daylighting
element. The inputs, outputs and parameters are organized in a way which is similar to the existing
glazing modules.
KEYWORDS
Transparent Insulation, Simulation, TRNSYS, Measuring, Comparison,
INTRODUCTION
There are already existing programs to simulate mass wall with Ti (3). Some of these programs are
neither for sale or for public use. Others cannot be run on a PC.
A TRNSYS based program gave satisfactory solutte. for the simulation of Tl systems.
The simplest way in TRNSYSisrto.uae algebraic operations between modules (4). The only
disadvantage of this method is that some of the parameters cannot be modified because they are
internal ones in the existing module. For example the thermal resistance of the Tl cannot be
calculated correctly. The modification of some existing module leads to the creation of a new module.
This new module uses some of the routines of the Thermal Storage Wall, TYPE36. The new routines
can be included in other modules such as collector or window as well.
-------
3164
DESCRIPTION OF THE MODULE
The new Tl wall module incorporates the finite difference method to determine the transient change
in the temperature at different nodes. Transient energy flows are calculated after calculating the new
temperatures of the nodes inside the wall. U-values of the honeycomb material are taken from
Chattha. To calculate the angle dependent transmittance of the honeycomb insulation, the
equation given by Chattha was used:
The conductive and radiative heat exchange are calculated correctly.
PARAMETERS USED BY THE MODULE
1 Units system 1=SI 2=British
2 Area of wall
3 Thickness of the wall
4 Thermal conductivity of the wall
5 Product of specific heat and density
6 Absorptance
7 Emittance of glazing
8 No. of glazing
9 Glazing thickness-extinction coefficient product
10 Refractive index
11 Heat transfer coefficient of opaque insulation inside
12 Heat storage capacity of opaque insulation inside, per square unit
INPUTS USED BY THE MODULE
1 Room temperature
2 Ambient temp.
3 Wind velocity
4 Heat transfer coefficient from glazing to ambient
5 Heat transfer coefficient from wall to air
6 Total radiation
7 Diffuse radiation
8 Angle of incidence
9 U-Value of transparent insulation
OUTPUTS OF THE MODULE
1 Energy conducted to room from wall
2 Change of internal energy of wall
3 Solar energy absorbed by the outer surface of wall
4 Energy lost from glazing to environment
5 Temperature of glazing
6 Node 1 temperature
5+i Node i temperature
DERIVATIVES USED BY THE MODULE
1 Initial temperature of 1 st node
i Initial temperature of ith node
-------
3165
EVALUATION OF THE MODULE
To evaluate the accuracy of the
module test simulation was used to
compare it with measured values. The
measurements had been done in
Birmingham, UK on an existing typical
house in 1990. The wall covered by Tl
is 225mm thick brickwork. The
measured data were the ambient and
the room air temperatures and the
surface temperatures on the inside
and outside surface of the wall. Heat
flux was also measured on the inside
surface of the wall, close to the
temperature sensor. The total solar
radiation was measured on horizontal
and on vertical surfaces coplanar with
the wall is nearly south facing.
Diffuse radiation was measured on a
horizontal surface only.
The computer model is concentrated
on the examined Tl module which is
using the measured data as input, with
the ambient and the room
temperature as well.
The comparisons of the measured
and simulated values are shown in
figures 1 to 5. Fig.1 shows the wall
temperatures on the external surface,
Fig.2 on the internal surface for a short
period in the end of April. The heat flux
on the internal surface for the same
period is illustrated in Fig.3. There are
small differences in the peaks arid
there is some phase shift but the
approximation of the measured
values by simulated values is good.
A stepwise parametric approximation
was used to achieve good
correspondence between measured
and simulated values for a long period
of time. The statistics are shown on
Fig.4 for temperatures and on Fig.5 for
heat flux represented by average
values. The simulated temperatures
are very close to the measured ones.
The average measured and simulated
temperatures on the internal surface
are 26.8°C and 26.6°C respectively.
On the external surface these values
are 33.4°C and 33.1 °C. The simulated
average heat flux is 14.5W/m and the
measured 14.4W/m . The accuracy is
99% in all three cases
C
60
50
40
30
20
10
o I-¦ Measured— Simulated -..Ambient
1
Days of the year
Fig. 1. External wall temperatures and ambient air
temperature
C
50
30 -
••• Measured—Simulated -* Room
Days of the year
Fig. 2. Internal wall temperatures and room temperature
W/m2
-10
-15
-20
-25
-30
-35
-40
— Simulated
Days of the year
Fig. 3. Heat flux through the internalsurfaro
-------
3166
W/m2
_ Outside
V/ A Simulated
Inside
l\ I Measured
Simulated
Measured
Fig. 4. Average wall temperatures Fig. 5. Average energy transferred to the
room
LIMITATIONS OF THE EVALUATION
Firstly the measurements have a limitation imposed by certain level of accuracy. Secondly the
simulation evidently cannot follow each small event by user behaviour in a house. The special
limitation is that a honeycomb structure Tl has very different properties against beam and diffuse
radiation. From measured vertical total and horizontal total and diffuse radiation the vertical beam
and diffuse components were calculated and this brought with it a certain limitation familiar to
TRNSYS users.
CONCLUSION
Despite the limitations of the evaluation the new TRNSYS module can be considered as a very
accurate simulation program for Tl. The main strength of the program is the good accuracy provided
by long-term statistics. The short-term approximation is not as good as the long-term one, but it is
also acceptable.
REFERENCES
Klein, A.A., et al.: TRNSYS, A Transient System Simulation Program, Solar Energy Laboratory,
University of Wisconsin - Madison 1988
ASHRAE Handbook - 1985 Fundamentals, American Society of Heating, Refrigeration and
Air-ConditioningEngineers, Inc., Atlanta, GA, 1985
Sick, F.: Transparent insulation - simulation programme, Proceedings of the 3rd workshop on
TIM, Titisee, Germany, 1989
Jankovic, L., Jesch, L.F.: Honeycomb transparent insulation modeled by a second generation
TRNSYS module, Proceedings of the 3rd workshop on TIM, Titisee, Germany, 1989
Chattha, J.A., Transparent Honeycomb Insulation in Theory and Practice, Ph.D. Thesis,
University of Birmingham, UK, 1990
Boy, E.: Comparison of measurements on transparent thermal Insulation systems with numerical
simulations of building envelopes, Proceedings of the 2nd workshop on TIM, Freiburg, Germany,
1988
-------
3167
AN EXTENSION OF THE TRNSYS MULTIZONE COMPONENT
FOR TRANSPARENT INSULATION APPLICATIONS
F. Sick, J. P. Kummer1
Fraunhofer-Institut fur Solarc Energicsystemc,
Oltmannsstr. 22, D-7800 Freiburg i. Br.
ABSTRACT
It is described how the multizone component ("TYPE 56") of the simulation program TRNSYS (Klein,
1988) has been extended in order to account for transparent or translucent massive or massless layers in the
envelope of the building. Any reasonable combination of transparent and opaque layers including
controllable shading devices is allowed. The user-friendliness of the "Building Input Description" (BID) has
been maintained. The mathematical model is briefly described. A comparison of measured and simulated
wall temperatures of a transparently insulated building in Freiburg is given.
KEYWORDS
TRNSYS; building simulation; traasparent insulation.
INTRODUCTION
Transparent insulation (TI) materials are obtaining increasing importance for the passive use of solar energy
in buildings (Goetzberger, 1989; Schmid, 1989). They can be used for both heating and daylighting
purposes. An accurate prediction of the thermal behaviour of a passive solar building during the design phase
is necessary in order to obtain low heating loads in the winter and avoid overheating problems during sunny
and warm periods. This prediction is usually done by detailed computer simulations. Conventional building
envelopes can be reasonably modeled by using constant properties, which is done in most building
simulation programs. Due to their transparent nature, TI materials have incidence angle-dependent optical
properties. Furthermore, their heat transfer coefficients are temperature-dependent. This,along with the
demand for a user-friendly and fast TI software.led to the development of a simplified model for the
description of TI walls (Kummer, 1989). The algorithm has been integrated into a modified TYPE 56
component and its Building Input Description (BID) program.
THE MODEL
Modeling of the transparent portions of a wall is accomplished with both detailed and simplified models. In
the BID preprocessor for Type 56, a detailed resistor network is used to model the transparent part of the
system. An example network is shown in Fig. 1.
1 Currently employed at Johnson Controls, Milwaukee, WI/USA
-------
3168
Glz
Air
TIM
Glz
Air
-v-
Outside
Tb
Rrad
Optical Layers
Rrad
Inside
's.o Hs,i
Massive
Layers
Fig. 1. Example for a detailed resistor network of the transparent part of a wall
All optical layers are assumed to he of negligible mass so that a steady state analysis can be
performed. Infrared radiation exchange between surfaces is modeled with a linearized radiation heat
transfer coefficient. The transmission, reflection, and absorption properties of the system and each
layer within the system are calculated using methods presented by Platzer (1988). A temperature and
solar radiation dependent heat transfer coefficient (1/RT) is determined for the optical system by
running a short internal simulation using the detailed network within the BID preprocessor. The
results from the preprocessor are stored in a data file and used by the Type 56 component for the
actual simulation. A simplified model of the walls and windows containing TI materials is used at this
point (Kummer, 1989). This model uses the resistor network shown in Fig. 2. Heat flows through the
massive layers of a wall (q and qs ^ are modeled with transfer functions as described in the
TRNSYS documentation. Trie principle of superposition is used in order to treat the heat transfer
due to the solar absorption terms S; and S0 separately from the heat transfer due to the temperature
difference across the optical layers calculated with RT. The terms S; and S represent the absorbed
solar energy reaching the respective surfaces by all heat transfer modes. They are calculated with
radiation properties determined by the BID preprocessor. The detailed model used in the
preprocessor BID was originally developed as a stand-alone TRNSYS component for modeling walls
consisting of TI materials. This model was validated with measured data from experiments and with
other detailed models at the ISE (Sick, 1989). The simplified model was developed and validated from
simulations with the detailed model.
Tb
AAA-
S;
•Optical Layers-
-*<>
q s.,
q s.i
Massive
Layers
T;
Fig. 2. Simplified resistor network of the example shown in Fig. 1
THE EXTENDED BUILDING DESCRIPTION
The BID program was modified to allow the definition of transparent layers of negligible and non-
negligible mass. These layers, which are referred to as transparent layers, can be used to construct
-------
3169
various walls and windows. The difference between walls and windows is that at least part of the wall
consists of non-negligible mass while windows are considered to be of negligible mass. The BID
program and Type 56 component can be used to model wall and window systems with multiple layers
of glazings, TIM, air-gaps, air-gaps with roller blinds, and a transparent medium with non-negligible
mass (e.g. a glass block). In the following paragraphs changes and additions to the description of the
BID program as it appears in the TRNSYS manual are presented. Readers who are not familiar with
the TRNSYS BID structure may omit this section and jump directly to the example described
afterwards. To be consistent with the TRNSYS manual, the same notation as there is used.
The TYPES section
There are 13 different TYPES that can be defined. The 12 TYPES from the old version are still
available, although many have been modified, and one new TYPE (ANGLES) has been added for use
in defining transparent layers. The ANGLE type defines a set of angles for which optical properties
are given when defining a transparent layer. This allows the user to determine the number of optical
property values that will be required in the layer definition. The forms for defining massive and
massless layers available in the old version have not changed. Additional layers can now be defined
which are useful in constructing transparent wall and window systems. New to this version is the use
of layers in defining windows. The new LAYER types are termed "transparent layers" to indicate that
they belong to transparent wall or window designs. As can be seen below, they are not necessarily
transparent. The 6 new transparent layers are:
1. The opaque absorber layer
This layer is similar to the massive layer defined in the old version, however additional
optical properties are required such as the absorptance for solar radiation. This layer is used
in a wall system with transparent massless layers which must have one such layer as the
outermost massive layer.
2. The transparent absorber layer
This layer is used for wall systems with transparent massive or massless layers. It represents
a layer with non-negligible mass that is partially transparent to solar radiation. Optical
properties are calculated from the index of refraction and the extinction coefficient.
3. The air gap/roller blind layer
This layer represents an air-gap with a controllable roller blind element in the center of the
gap. The user inputs the optical properties of the roller blind in its closed position. The
control signal is an input to TYPE 56.
4. The air gap layer (without roller blind)
This layer represents an air-gap without a roller blind. This layer type must be used in the
optical portion of a wall or window while the massless layer type available in the old version
is to be used between opaque massive layers.
5. The transparent insulation (TIM) layer
This layer represents a layer of transparent insulation material, e.g. honeycomb structures
or aerogel. The IR penetration depth is the most important new parameter for the
calculation of the equivalent conductivity (Platzer, 1988).
6. The glazing layer
This layer represents transparent materials that can be considered massless. This is
normally the case for standard glass panes.
The transparent WALL types and WINDOW types have a slightly different description from their
standard opaque types. New is the slope-dependent calculation of convection coefficients in air gaps
of transparent parts of walls. Orientations are described the same as in the old version. However, for
each orientation iipus for the total incident radiation, the incident beam radiation, and the incidence
angle of the beam radiation will be required.
The BUILDING and the OUTPUT sections
Buildings are described the same as in the old version with the exception of the window. In the past
an over-all value of the window transmittance was required. Separate values for the beam and diffuse
-------
3170
transmittances are in this new version required. For window systems defined with transparent layers
the optical properties are calculated.
Walls are also defined in the same manner as in the old version. "Optical" walls are determined within
the program by the presence of transparent layers in the wall defintion. Any walls made of air-gap/-
roller blind layers will require a control input for each roller blind during the simulation.
Some of the outputs are slightly modified and there is one additional output available: The absorber
temperature of a transparently insulated wall.
EXAMPLE
In order to check the new part of the program, the dynamic behaviour of the transparently insulated
wall is used as an indicator of the program performance. A comparison will be made between
predicted and measured surface temperatures of the opaque portion of a transparently insulated wall,
i.e. the absorber temperature and the inside wall surface temperature. A cross-section of the wall is
shown in Fig. 3.
frame
shading device
elagpane
transparent
insulation
transparent film |

v.\^\
absorber (black oaintl 4s-
.... • -V r

masonry I
i
"" •
J —l-J
F

Fig. 3. Cross-section of the investigated transparently insulated wall in Freiburg
The investigated building is an 8-family apartment building in Freiburg which was retrofitted with
transparent insulation (Vahldiek, 1990) and extensively monitored. For the purpose of high-quality
predictions of the thermal behaviour of the building by simulation, a simple parameter optimization
was carried out. The penetration depth of the TIM and the conductivity of the massive wall portion
were slightly adjusted. Figure 4 shows the absorber temperature and the inside wall temperature of a
south-east oriented TIM wall for a 3-day-period in February 1990. It appears that the simulated wall
is slightly quicker responding to the incident radiation than the real one. The differences in the
absorber temperatures around noon of the first and second day shown here can be explained with an
inaccuracy in the TRNSYS radiation processor. Figure 5 shows that the measured incident radiation
is lower at these times than the simulated incident radiation which is calculated from the global
horizontal radiation.
-------
3171
^ Absorber Temp., measured
' Absorber Temp., simulated

Inside Wall Temp., measured
Inside Wall Temp., simulated
~6 12
19. Feb. 1990
6 12 18
18. Feb. 1990
0
0
6 12 18
20.Feb. 1990
Fig. 4. Absorber temperature and inside surface temperature of a transparently insulated
south-east oriented wall in Freiburg: comparison of measured data
with the TRNSYS simulation results.
700-
measured solid
simulated dotted
600

.2 400
c 300
200-
100-
0
6
12
18
0
6
12
18
0
6
12
18
0
18. Feb. 1990 19. Feb. 1990 20.Feb.1990
Fig. 5. Incident radiation on the vertical south-east fassade for the period investigated as measured
(solid line) and calculated from the measured global horizontal radiation (dotted line).
-------
3172
Besides the proof of correct treatment of a transparently insulated wall, it was checked and validated
that the roller blind "mechanism" works accurately. The energy balances show good agreement with
the measured data. The new windows work significantly better than the old ones due to the use of
angle-dependent transmittance values. The conventional part of the TRNSYS TYPE 56 was checked
and used before by many users many times and seems to be a reliable tool for building energy
analysis. Equipped with TIM it is probably one of the most comprehensive, flexible, easy-to-use and
at the same time inexpensive programs for this purpose.
NOMENCLATURE
qs conductive heat flux into the wall at the outside surface of the outer massive wall layer
qs'j conductive heat flux from the wall at the inside surface of the inside massive wall layer
rad resistance for IR radiation heat transfer
RT temperature and absorbed solar radiation dependent resistance value of the transparent
layers
Sj total solar energy reaching the absorber
S total solar energy absorbed by the transparent layers that reaches the outside surface
temperature of the outside surface of the outer massive wall layer (absorber)
T temperature of the inside surface of the indside massive wall layer
Tb temperature at the outside surface of the outermost transparent layer
REFERENCES
Goetzberger, A. (1989). Transparent Insulation - A New Solar Energy Component. Proceedings 2nd
European Conference on Architecture (Paris).
Klein, S. A. and others (1988). TRNSYS - A Transient System Simulation Program. Engineering Exp.
Station Report. 38-12. University of Wisconsin, Madison.
Kummer, J. P. (1989). A Simplified Model for Transparently Insulated Walls. Proceedings 3rd
International Workshop on Transparent Insulation Technology (Titisee/Freiburg)
Platzer, W. (1988). Solare Transmission und Warmetransportmechanismen bei Transparenten
Warmedammaterialien. PhD-Thesis, Universitat Freiburg.
Schmid, J., W.-S. Wilke, and F. Sick (1989). New Concepts for Daylighting and Heating of Buildings.
Proceedings 2nd European Conference on Architecture (Paris).
Sick F., W.-S. Wilke, and J. P. Kummer (1989). Simulation Tools for the Prediction of the Thermal
Behaviour of Transparently Insulated Buildings. Proceedings 2nd European Conference on
Architecture (Paris).
Vahldiek, J., E. Bollin, and A. Wagner (1990). Altbausanierung mit transparenter Warmedammung -
Retrofitting with transparent insulation. Proceedings 7. Int. Sonnenforum. Frankfurt, Germany
-------