University of Southern California Dissertations and Theses

A parametric study of the thermal performance of green roofs in different climates through energy modeling

(USC Thesis Other)

A parametric study of the thermal performance of green roofs in different climates through energy modeling

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

A PARAMETRIC STUDY OF THE THERMAL PERFORMANCE OF GREEN ROOFS IN
DIFFERENT CLIMATES THROUGH ENERGY MODELING

By

Sananda Mukherjee

A Thesis Presented to the
FACULTY OF THE USC SCHOOL OF ARCHITECTURE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF BUILDING SCIENCE

August 2013

Copyright 2013 Sananda Mukherjee

ii
ACKNOWLEDGMENTS
“Gratitude is the memory of the heart.” - Jean Baptiste Massieu

I wish to express my profound gratitude to my thesis committee with Professor Pablo La Roche,
Ph.D., as my chair, and Professors Kyle Konis, Ph.D., and Joon-Ho Choi, Ph.D. whose counsel,
encouragement and wisdom culminated in the completion of my graduate thesis. Prof. La
R oche ’ s undy i ng pat i en ce , pr of o und c r i t i ci sm and unr e l en t i ng con f i d enc e i n m e m ot i v at ed m e t o
do my very best and I sincerely thank him for helping me through this journey. I would like to
thank Prof Konis for helping me to broaden my perspectives on the different aspects of my
research, sharpen my abilities of critical thinking and always having faith in me. I am grateful to
Prof. Choi for his time and patience. His hard working attitude always inspired me and kept me
g oi ng t hr oug h t hos e ‘ l ong ni g ht s’ w h en a l l t he d at a a nd spr e ads h ee t s s t a r t e d t o m ak e m e di z zy .
This thesis would not have been possible without the support and encouragement of my thesis
advisors Prof. Douglas Noble and Prof. Ilaria Mazzoleni. They are both exceptional mentors and
without their perseverance and constant motivation, this thesis would never get submitted on
time!
I would like to thank Jeffrey Landreth, whose technical support and guidance helped me through
my software related problems and simulation nightmares. He taught me a whole new dimension
of Microsoft Excel and changed the way I looked at spreadsheets. I sincerely thank him for
teaching me the art of data interpretation and analysis.
My special thanks go to Dr. Steven Sandifer whose kindness and cooperation at the nascent stage
of m y r es ea r ch al l ow ed m e t o ‘ g et t he bal l r ol l i ng ’ . H i s i nt er e st i n m y t hes i s and h i s wil l i ng nes s t o

iii
share information pertaining to the same instilled the confidence in me to take my research
forward. In addition, I would like to extend my gratitude to Prof. Tim Kohut and Amy Hackney
who kept encouraging me throughout my thesis journey, and took time off to attend my thesis
review sessions despite having no obligation to do so. I appreciate their support all the way.
I consider myself extremely lucky to be a part of the Chase L. Leavitt Graduate Building Science
Program that embraces individuals from all spectrums of life and different parts of the world. It
has been an honor to be under the tutelage of such talented and dedicated teachers, and be a part
of a class with such wonderful people! I learnt something about life from each friend that I made
in the MBS family and I am hoping that the wonderful memories will continue to form, as we
move on from here. In a way, each one of them has contributed to my thesis at some point, so I
thank them all.
I would like to thank the developers of Skype and Teamviewer who made communication over
geographical miles much easier. Be it a simple thesis discussion with my Chair or a midnight chat
with friends on the other side of the world, technology helped a lot to get me to the finish line!
Lastly, I extend my deep and abiding gratitude to my parents, whose forbearance has been sorely
taxed but whose support has been boundless. Without them, I would not have been the person that
I am today.

iv
Table of Contents

ACKNOWLEDGMENTS ............................................................................................................. ii
LIST OF FIGURES .................................................................................................................... viii
LIST OF TABLES ....................................................................................................................... xii
Chapter 1: Introduction to Green Roofs...................................................................................... 1
1.1. Introduction .................................................................................................................. 1
1.2. Evolution of green roofs as a roofing option ............................................................... 4
1.3. Green Roofs compared to other roofing types ............................................................. 4
1.4. Goals and Objectives ................................................................................................... 5
1.5. Research Hypothesis .................................................................................................... 6
1.6. Scope of work .............................................................................................................. 7
1.7. Chapter structure .......................................................................................................... 7
Chapter 2: Literature study on Green Roofs .............................................................................. 9
2.1. Understanding Green Roofs ......................................................................................... 9
2.1.1. How Green Roofs Work .............................................................................. 10
2.1.2. Green Roofs Benefits .................................................................................. 12
2.1.3. Green Roof Trends ...................................................................................... 12
2.2. Field experiments and case studies on the thermal performance of green roofs........ 15

v
2.3. Energy modeling and computer simulations of green roofs ...................................... 20
Chapter 3: Building Energy Performance Simulation ............................................................. 27
3.1. Building Energy Simulation Model ........................................................................... 27
3.2. Simulation Engine and Graphical User Interface ...................................................... 28
3.3. Software overview ..................................................................................................... 29
3.4. The Ecoroof Model in Design Builder/EnergyPlus ................................................... 30
3.4.1. Model Description ....................................................................................... 30
3.4.2. Limitations of the Ecoroof model ................................................................ 32
Chapter 4: Methodology .............................................................................................................. 34
4.1. Baseline model validation phase ................................................................................ 34
4.2. The simulation phase ................................................................................................. 36
4.2.1. Testing Parameter 1: Insulation ................................................................... 39
4.2.2. Testing Parameter 2: Vegetation Type (Leaf Area Index) .......................... 39
4.2.3. Testing Parameter 3: Growing Media or soil depth .................................... 40
4.2.4. The Climate Types ...................................................................................... 41
4.2.5. The Building Energy Model: Office Prototype ........................................... 42
4.3. The Analysis phase .................................................................................................... 43
Chapter 5: Green Roof model validation ................................................................................... 45
5.1. Simulation results of the field experiment ................................................................. 45
5.2. Analysis and validation of results .............................................................................. 45

vi
Chapter 6: Analysis and Data Interpretation ............................................................................ 48
6.1. Parametric Analysis of data for Phoenix.................................................................... 49
6.1.1. Effect of Insulation on EUI ......................................................................... 49
6.1.2. Effect of LAI on EUI ................................................................................... 51
6.1.3. Effect of Soil depth on EUI ......................................................................... 52
6.1.4. Reduction of EUI over base case ................................................................. 53
6.1.5. Effect on annual heating and cooling loads ................................................. 54
6.1.6. Effect on heat balance through the roof ....................................................... 56
6.2. Parametric Analysis of data for Los Angeles ............................................................. 58
6.2.1. Effect of Insulation on EUI ......................................................................... 58
6.2.2. Effect of LAI on EUI ................................................................................... 59
6.2.3. Effect of Soil depth on EUI ......................................................................... 60
6.2.4. Reduction of EUI over base case ................................................................. 61
6.2.5. Effect on annual heating and cooling loads ................................................. 62
6.2.6. Effect on heat balance through the roof ....................................................... 63
6.3. Parametric Analysis of data for Chicago ................................................................... 66
6.3.1. Effect of Insulation on EUI ......................................................................... 66
6.3.2. Effect of LAI on EUI ................................................................................... 67
6.3.3. Effect of Soil depth on EUI ......................................................................... 68
6.3.4. Reduction of EUI over base case ................................................................. 69

vii
6.3.5. Effect on annual heating and cooling loads ................................................. 70
6.3.6. Effect on heat balance through the roof ....................................................... 71
6.4. Summary: Comparison of the assemblies against different climates ......................... 73
6.4.1. Performance metric: EUI ............................................................................. 74
6.4.2. Performance metric: Annual Heating and Cooling Load ............................ 77
6.4.3. Performance metric: Annual Heat Balance through the roof ...................... 82
Chapter 7: Conclusion ................................................................................................................. 84
7. 1. Conclusions drawn from study ...................................................................... 84
7. 2. Limitations of study ....................................................................................... 87
7. 3. Future Work .................................................................................................. 88
Bibliography ................................................................................................................................. 92
Glossary ........................................................................................................................................ 96
APPENDIX A ............................................................................................................................... 97
APPENDIX B ............................................................................................................................... 98
APPENDIX C ............................................................................................................................... 99
APPENDIX D ............................................................................................................................. 106
APPENDIX E ............................................................................................................................. 111

viii
LIST OF FIGURES

Figure 1: In the Meadow Hills, model by Hundertwasser, 1989 ..................................................... 1
Figure 2: Graph showing electricity load with respect to temperature ............................................ 2
Figure 3: NASA Heat map showing the annual mean temperature change over the globe from
1950-2008 ........................................................................................................................................ 2
Figure 4: Temperature differences between different urban fabrics highlights the UHI effect ....... 3
Figure 5: Dark roofs contribute to the UHI...................................................................................... 3
Figure 6: Evolution of Green Roofs ................................................................................................. 4
Figure 7: Evapotranspiration and shading on a green roof- The main factors that distinguish it
from any other roofing type ............................................................................................................. 5
Figure 8: Design of successful green roof systems m i m i cs ea r t h ’ s na t u r al soi l l ay er s .................. 10
Figure 9: Vegetated roof heat flow ................................................................................................ 11
Figure 10: Energy balance of two different roofing types ............................................................. 12
Figure 11: Estimated growth of North American Green Roof Industry ........................................ 13
Figure 12: Top 10 North American Metro Regions- Green Roofs installed in 2011..................... 13
Figure 1 3 : T es t c el l ar r ang e m ent s f or L a Roche ’ s expe r i m ent s .................................................... 15
Figure 14: Details of the Smart Green roof test cell ...................................................................... 16
Fig ur e 15 : Psy chom et r i c c h ar t s s how i ng an ov er v i ew o f La R och e’ s e xp er i m ent r e sul t s ............. 16
Figure 16: Comparative membrane temperatures for black, white and green roofs ...................... 18
Figure 17: Daily temperature pattern (Sept 5) of different roofing types illustrating the overall
reduction in maximum temperature and suppression of the daily swing of the vegetated roof ..... 19

ix
Figure 18: Green roof module predictions as compared to measured soil surface temperatures for
green roofs at the University of Central Florida test site. The figure panels represent two weeks of
hourly data within each of four seasons ......................................................................................... 21
Figure 19: A screenshot of the online green roof energy calculator .............................................. 22
Figure 20: Calculated impact of green roof using tool- screenshot ............................................... 23
Figure 21: Percentage of effectiveness achieved by each type of green-roof over the Traditional
and the Un-Isolated roofs in Cairo-Egypt ...................................................................................... 26
Figure 22: EnergyPlus Simulation engine and its various GUIs.................................................... 28
Figure 23: Few energy modeling programs that are commonly used to run energy simulations .. 29
Figure 24: Screenshot of Ecoroof option within Design Builder ................................................... 31
Figure 25: Methodology flow chart ............................................................................................... 34
Fig ur e 26 : Sand i f er ’ s g r ee n r oof as s em bl y se t up ........................................................................... 35
Fig ur e 27 : Sand i f er ’ s g r ee n roof assembly modeled in Design Builder ........................................ 36
Figure 28: Test parameters and their subset variables ................................................................... 38
Figure 29: Plants with different LAI; LAI=1(on the left), LAI=5 (on the right). .......................... 40
Figure 30: Different soil depth variables ....................................................................................... 41
Figure 31: Heating and cooling degree day chart, highlighting the spectrum of climate types
covered in this study ...................................................................................................................... 42
Figure 32: Prototypical office building modeled in Design Builder- (a) Screenshot of the single
story building (b) Screenshot of the HVAC zoning diagram for the typical floor. ....................... 43
Figure 33: Measured versus simulated data scatter plot ................................................................ 46
Figure 34: Measured versus simulated data scatter plot after adjustment ..................................... 47
Figure 35: Effect of increasing insulation thickness on EUI ......................................................... 50

x
Figure 36: Effect of increasing LAI on EUI .................................................................................. 51
Figure 37: Effect of increasing soil depth on EUI ......................................................................... 52
Figure 38: Percentage reduction in EUI over base case for different green roof assemblies ......... 54
Figure 39: Reduction in annual heating and cooling loads over base case .................................... 55
Figure 40: Heat balance through the roof over the year................................................................. 56
Figure 41: Comparison of heat flux through roof for peak heating and cooling days ................... 57
Figure 42: Effect of increasing insulation thickness on EUI ......................................................... 58
Figure 43: Effect of increasing LAI on EUI .................................................................................. 59
Figure 44: Effect of increasing soil depth on EUI ......................................................................... 60
Figure 45: Reduction of EUI over base case .................................................................................. 61
Figure 46: Reduction in annual heating and cooling loads over base case .................................... 62
Figure 47: Heat balance through the roof over the year................................................................. 64
Figure 48: Comparison of Heat flux through the roof for peak heating and cooling days ............ 65
Figure 49: Effect of increasing insulation thickness on EUI ......................................................... 66
Figure 50: Effect of increasing LAI on EUI .................................................................................. 67
Figure 51: Effect of increasing soil depth on EUI ......................................................................... 68
Figure 52: Percentage reduction in EUI over base case for different green roof assemblies ......... 69
Figure 53: Reduction in annual heating and cooling loads over base case .................................... 70
Figure 54: Heat balance through the roof over the year................................................................. 72
Figure 55: Comparison of heat flux through roof for peak heating and cooling days ................... 73
Figure 56: Comparison of EUI Reduction of different assemblies over base case ........................ 75
Figure 57: Parametric comparison of EUI for different roof assemblies ....................................... 76
Figure 58: Parametric Comparison of Peak cooling load for different roof assemblies ................ 78

xi
Figure 59: Parametric comparison of annual cooling loads for different roof assemblies ............ 80
Figure 60: Parametric comparison of peak cooling loads for different roof assemblies ................ 81
Figure 61: Parametric Comparison of Peak heating load for different roof assemblies ................ 83
Figure 62: Home screen (screenshot) - Explanation about the tool ............................................... 89
Figure 63: User is first asked to select climate zone or type from map or by clicking on
appropriate button below ............................................................................................................... 90
Figure 64: User is then asked to select the desired variable. Upon selection, user is automatically
redirected to new window where the relevant charts or graphics are shown. ................................ 91

xii
LIST OF TABLES

Table 1: Exterior ambient temperature (Tamb), Average and Maximum interior temperatures (Ti)
for different roofing types- Nov to mid Feb .................................................................................. 17
Table 2: Mean and maximum indoor air temperature and heating, cooling, and total energy
demand in La Rochelle for different LAI levels ............................................................................ 24
Table 3: Mean and maximum indoor air temperature and heating, cooling, and total energy for
different insulation levels ............................................................................................................... 25
Table 4: Mean and maximum indoor air temperature and heating, cooling, and total energy
demand in Athens, La Rochelle and Stockholm ............................................................................ 26
Table 5: Ranges of data input for Ecoroof model .......................................................................... 33
Table 6: List of Parameters related to a green roof assembly ........................................................ 37
Table 7: Building Energy Model Inputs ........................................................................................ 43
Table 8: Summary of EUI reduction- best and worst case scenario .............................................. 74
Table 9: Summary of annual heating and cooling load reductions over base case ........................ 77
Table 10: Summary of heat flux through the roof- best and worst scenario .................................. 82
Table 11: Correlation between parameters and metrics ................................................................. 85
Table 12: Building operation schedules ......................................................................................... 98
Table 13: Comparison between Field measured data and simulated data ..................................... 99
Table 14: Correlation between data sets before and after adjustment.......................................... 106
Table 15: Comparison between different green roof assembly types and cool roof base case- for
Phoenix ........................................................................................................................................ 111

xiii
Table 16: Heat balance through the roof for different assembly types in Phoenix ...................... 112
Table 17: Heat balance through the roof on peak heating and cooling days in Phoenix ............. 112
Table 18: Comparison between different green roof assembly types and cool roof base case- for
Los Angeles ................................................................................................................................. 113
Table 19: Heat balance through the roof for different assembly types in Los Angeles ............... 114
Table 20: Heat balance through the roof on peak heating and cooling days in Los Angeles ...... 114
Table 21: Comparison between different green roof assembly types and cool roof base case- for
Chicago ........................................................................................................................................ 115
Table 22: Heat balance through the roof for different assembly types in Chicago ...................... 116
Table 23: Heat balance through the roof on peak heating and cooling days in Chicago ............. 116

xiv
ABSTRACT

In recent years, there has been great interest in the potential of green roofs as an alternative
roofing option to reduce the energy consumed by individual buildings as well as mitigate large
scale urban environmental problems such as the heat island effect. There is a widespread
recognition and a growing literature of measured data that suggest green roofs can reduce
building energy consumption. This thesis investigates the potential of green roofs in reducing the
building energy loads and focuses on how the different parameters of a green roof assembly
affect the thermal performance of a building.
A green roof assembly is modeled in Design Builder- a 3D graphical design modeling and energy
use simulation program (interface) that uses the EnergyPlus simulation engine, and the simulated
data set thus obtained is compared to field experiment data to validate the roof assembly model
on the basis of how accurately it simulates the behavior of a green roof. Then the software is used
to evaluate the thermal performance of several green roof assemblies under three different climate
types, looking at the whole building energy consumption. For the purpose of this parametric
simulation study, a prototypical single story small office building is considered and one parameter
of t h e g r ee n r oo f i s a l t e r ed f or ea ch s i m ul at i on r un i n or der t o und er s t an d i t s e f f ec t on bui l di ng ’ s
energy loads. These parameters include different insulation thicknesses, leaf area indices (LAI)
and growing medium or soil depth, each of which are tested under the three different climate
types. The energy use intensities (EUIs), the peak and annual heating and cooling loads resulting
from the use of these green roof assemblies are compared with each other and to a cool roof base
case to determine the energy load reductions, if any. The heat flux through the roof is also

xv
evaluated and compared. The simulation results are then organized and finally presented as a
decision support tool that would facilitate the adoption and appropriate utilization of green roof
technologies and make it possible to account for green roof benefits in energy codes and related
energy efficiency standards and rating systems such as LEED.
1

Chapter 1: Introduction to Green Roofs

Figure 1: In the Meadow Hills, model by Hundertwasser, 1989 (Hundertwasser)

"The true proportions in this world are the views to the stars and the views down to the surface of
the earth. Grass and vegetation in the city should grow on all horizontal spaces - that is to say,
wherever rain and snow falls vegetation should grow, on the roads and on the roofs. The
horizontal is the domain of nature and wherever vegetation grows on the horizontal level man is
off limits; he should not interfere. I mean taking away territories from nature, which human
beings have always done.”
-Friedensreich Hundertwasser, Austrian painter and architect

1.1. Introduction
In recent years, there has been great interest in the potential of different roofing options to reduce
the energy consumed by individual buildings as well as mitigate large scale urban environmental
problems like the Urban Heat Island (UHI) effect, particularly in urban areas where roofs
constitute a large percentage of the overall surface area. Dark surfaces in urban areas absorb more
heat from the sun than surrounding vegetated area and cause the UHI effect. Trapped heat in the
urban environment and lack of vegetation increase the air temperature up to 6-8 °F on hot

2
summer days (Pomerantz, et al. 2000). These phenomena increase demand for air conditioning
and energy.

Figure 2: Graph showing electricity load with respect to temperature (D. Sailor, Urban Heat
Islands, Opportunities and Challenges for Mitigation and Adaptation. 2002)
This UHI issue can be addressed at three different scales- building scale, urban scale and global
scale. The following Figure03, 04 and 05 elucidate the problem at the aforementioned three
levels.

Figure 3: NASA Heat map showing the annual mean temperature change over the globe from
1950-2008 (National Aeronautics and Space Administration, NASA)

3

Figure 4: Temperature differences between different urban fabrics highlights the UHI effect
(Shickman 2011)

Figure 5: Dark roofs contribute to the UHI
Vegetated or green roofs act as UHI mitigation tools, where water evaporation from the
vegetation as well as thermal mass and thermal resistance of the green roof contribute to reduce
indoor and outdoor temperatures in the building and the urban area depending on the type of
vegetation, depth and type of growing medium and local climate. This in turn would help to
reduce the cooling load of a building, resulting in reduced air cooling requirements and therefore
reduced energy consumption and associated output of atmospheric carbon.
This thesis investigates the potential of green roofs at a building scale and studies their effects on
the energy loads and heat balance through the roof of a prototypical small sized office building in
different climates.

4
1.2. Evolution of green roofs as a roofing option
With the increased desire for high-performance buildings and sustainable building products,
green roofs have become a growing roofing option. The green roof technology has a history that
pr eda t e s t h e m oder n e r a. Today ’ s g r ee n r oof s a r e m oder n v er s i ons of ce n t ur i es ol d r oo f i ng practices. From the hanging gardens of Babylon to the sod roofs in Iceland during the Viking era,
adaptations of earth-covered roofs have been around for many years. Understanding the
motivations behind green roof usage in history broadens our understanding of their current uses.

Figure 6: Evolution of Green Roofs

1.3. Green Roofs compared to other roofing types
There are several different roofing types that have been used in buildings and their reason for
selection is generally guided by location, building code, client requirement etc. Currently there
are very few code requirements globally pertaining to green roofs. However, different green
building rating systems like LEED have allotted up to 15 credits that are pertaining to green
roofs. This apart, there are various government incentive and tax benefits provided to the owner
that act as a selection driver over other roofing types, leaving aside the environmental benefits of
such a roof.
The difference between the green roof and a conventional (bare) roof of a building is both
qualitative and quantitative. The process of heat transfer into the vegetated roof is totally

5
different. The solar radiation, the external temperature and the relative humidity are reduced as
they pass through the foliage, which covers the roof. The plants for their biological functions,
such as photosynthesis, respiration, transpiration and evaporation absorb a significant proportion
of the solar radiation (Kruche, Althaus and Gabriel 1982). The remaining solar radiation
influences the thermal load and internal climate of the building. The reduced solar gains also
results in increased roof life as compared to the normal life expectancy of a c onv en t i on al ‘ nak ed’ roof which is around 15 to 20 years (International Green Roof Association, IGRA). The R-value
(i.e. measure of thermal resistance) of the assembly does not remain constant over time unlike in
the case of other roofing types. A green roof acts as a heat sink- it is an active energy device,
literally collecting, processing, and releasing energy according to its immediate need. This is
where the main challenge of thermal modeling of a green roof comes into play. The energy
balance of a green roof is discussed in further details in Chapter Two.

Figure 7: Evapotranspiration and shading on a green roof- The main factors that distinguish it
from any other roofing type

1.4. Goals and Objectives
Green roofs have become important to green building practices, as a response to mitigating the
negative impacts of an urban scale issue like Urban Heat Island (UHI), in addition to controlling

6
storm water drainage, increasing biodiversity, improving air quality etc. Yet fundamental
questions about their effectiveness in improving the thermal performance of buildings remain.
The three primary objectives of this research are as follows-
1) To determine if green roofs are a better alternative to cool roofs as a roofing option for
different climate types, in terms of thermal performance of a building.
2) To determine how the different parameters that comprise a green roof assembly affect the
thermal performance of the same.
3) To be able to simulate these effects so that we do not have to rely on case study data or
lessons learnt and therefore make informed decisions about the assembly selection.

With the help of a quantitative and physically-based building energy simulation tool that
represents the effects of green roof constructions, the process of assessing green roof benefits
becomes much quicker. The user group for such a type of simulation tool may be restricted to
those individuals possessing a considerable knowledge of the theory of green roofs and the
software. But the information gained through the simulation exercise may be made available to
the public in the form of a decision support tool. Therefore the ultimate goal of this research is to
develop such a decision support tool that would facilitate the adoption of green roof technologies
and make it possible to account for green roof benefits in energy codes and related energy
efficiency standards and rating systems such as LEED.

1.5. Research Hypothesis
The different parameters of a green roof assembly affect the thermal performance of a building,
and the extent to which is does that is largely defined by the climate type.

7
1.6. Scope of work
This thesis investigates the benefits of using a green roof and quantifies those benefits in terms of
energy load reductions through computer simulation and compares them to the base case of cool
roof, within the context of three different climate types.
For the study, a green roof assembly is first modeled in the Design Builder modeling environment
and then the accuracy of the model is ascertained by validating the computer simulated data
against field experiment data of such a model. The validation is for a single climate type- warm
and coastal, and it is assumed that the model will work with reasonable efficiency for other
climate types. The same roof model is then used for the parametric simulation studies whereby, in
each simulation run one parameter of the green roof is changed and its thermal performance is
recorded, under the same climatic conditions as the first. The parameters that are tested in this
study are limited to three in number- insulation, vegetation type and growing medium depth or
soil depth- and each of these have their own subset variables that are discussed in details in
Chapter Four. The thermal performance analysis is done on only one building type -a prototypical
one story small-sized office building- and it is assumed that the building function does not change
over time. The simulation exercise is carried out for three different climate types- hot-dry, warm-
coastal and cool-humid. The three cities that are selected to represent each of these climates are
Phoenix (Arizona), Los Angeles (California) and Chicago (Illinois) respectively.

1.7. Chapter structure
This thesis document covers the existing research done on green roofs, followed by the
simulation study, results and conclusions. Chapter Two deals with the literature and background

8
study on green roofs in general, to understand why green roofs are important and how they
function in terms of heat exchange. One part of this chapter focuses on the field experiments done
on green roofs and related research, while another part deals with the energy modeling and
simulation studies done on green roofs so far. Chapter Three talks about Building Energy
Performance Simulation. It includes an overview of the energy modeling software that is being
used for the study and discusses the limitations of the model. Chapter Four covers the
Methodology followed for the purpose of this study. It explains the different phases of approach,
and all the variables involved in each phase. Chapter Five comprises of the results obtained from
the validation study of modeled green roof assembly. In this chapter, the field data is compared to
the simulated data to determine the accuracy of the model. Chapter Six contains the data analysis
and some conclusions resulting from the parametric simulations that are carried out with different
green roof assemblies. An analysis of all the simulation data obtained from the parametric
exercise is included in this chapter, arranged by climate type. Under each climate type, all of the
performance metrics and how they are affected by the different test parameters are discussed in
detail. This is followed by an overall conclusion in Chapter Seven, which also contains the
limitations of the study and scope for future work. The raw simulation data can be found in the
Appendices (referred to within different sections of the document) at the end.

9

Chapter 2: Literature study on Green Roofs

There is a widespread recognition and a growing literature of measured data that suggest green
roofs can reduce building energy consumption. Roofs at any elevation are inherently stressful
environments, particularly for planting, because they are subject to excessive heat, accelerated
evapotranspiration and potentially damaging winds. Therefore before installing a green roof on a
building, significant amount of research must be done. Plenty of information and guide books are
now available on building green roofs and their maintenance. But this chapter primarily deals
with the study related to the thermal performance of green roofs.
Since the subject of green roofs is so vast, it was imperative to organize the background study
i nt o spe ci f i c dom ai ns t hat a r e r el ev an t t o t h e s co pe o f t he a u t hor ’ s r es ea r ch. Therefore under this
chapter, three distinct sections will be covered which are as follows-
1) Understanding green roofs: Why they are important and how they work (from a thermal
perspective)
2) Field experiments and case studies on the thermal performance of green roofs
3) Energy modeling and computer simulations of green roofs

2.1. Understanding Green Roofs
A green roof or ecoroof is the roof of a building that is partially or completely covered with
vegetation and a growing medium, planted over a water-proofing membrane. It may also include
additional layers such as root barrier and drainage and irrigation systems (Cantor 2008). Once a

10
green roof is defined, the functions and benefits of the technology through a review of literature
can provide a justification for their usage.

Figure 8: Design of successful green roof systems m i m i cs ea r t h ’ s na t u r al soi l l ay er s (Nagler
2008)

2.1.1. How Green Roofs Work

Hotter roofs and higher air temperatures means hotter buildings and/or more energy consumed by
air conditioning. It is important to understand how green roofs work in terms of transferring heat
through the roof assembly, and therefore how they are more efficient in keeping the roof cooler as
compared to a conventional roof. Each layer of the green roof deals with heat flow differently.
1. Foliage – This is the most complex because it uses all methods of heat control provided by
nature – convection, evaporation, conduction, solar reflectivity, radiative heat emission, thermal
mass. Plants can also utilize extra heat conserving strategies such as defoliation during winter.

11
2. Stem gap – The air trapped between the foliage and the top of the planting medium provides
limited conduction and the stem itself conducts a small amount of heat between the foliage and
the roots.
3. Medium – Planting media conducts heat and often has enough thermal mass that it needs to be
taken seriously. It can also cool through evaporation when adequately moist.
4. Drain layer – The amount of heat conduction through the drain layer depends on how wet it is.
A more significant role of the drain layer can be through mass transfer - the removal of heat from
saturated media by providing a drain path for water heated within the media.
5. Waterproofing – The membrane provides simple conduction and some thermal mass.
6. Insulation (if necessary) – Insulation slows heat conduction and has negligible thermal mass.
7. Roof deck – Again, the roof deck provides simple conduction but can have significant thermal
mass. (Wark 2010)

Figure 9: Vegetated roof heat flow (Wark 2010)

12
So the energy balance of a green roof differs from a traditional roof in the sense, in the latter there
is no evaporative cooling.

Figure 10: Energy balance of two different roofing types (Velasco 2011)

2.1.2. Green Roofs Benefits
Green roofs provide numerous benefits, however, it is worth remembering that a client will often
have their own priorities. By considering each benefit at the earliest stages of design it will be
possible to create a green roof which is multi-responsive (Groundworks Sheffield). The reaper of
the benefit can be the building owner, the community and/or the environment. Some of these
benefits include- extended roof life (up to 60 years), reduced heating and cooling costs due to
lower energy demands, storm water management by reducing storm water runoff, government
and municipality incentives, moderating the UHI effect, improving air quality, reducing smog,
reducing noise, increasing biodiversity, improving aesthetics in the public realm etc.

2.1.3. Green Roof Trends
Green roofs have been in the use since time immemorial, as mentioned earlier. But with passing
time and evolution of technology, green roof systems are becoming increasingly complex and

13
more efficient, in terms of improving building performance. The factors that influence the growth
of the green roof market can be placed into four categories: regulatory, environmental, aesthetic,
and economic. The interplay of these drivers gives us the following statistics represented in the
form of bar charts.

Figure 11: Estimated growth of North American Green Roof Industry (Green Roof for Healthy
Cities, GRHC May 2012)

Figure 12: Top 10 North American Metro Regions- Green Roofs installed in 2011
A Green Roof Industry Survey by the Green Roofs for Healthy Cities (GRHC) members found
the North American industry grew by 115% in 2011, up significantly from a 28.5% growth rate
recorded in 2010 (Green Roof for Healthy Cities, GRHC May 2012). According to Lux Research
analysts, green roof installations will rise 70% to 204 million square meters globally, but costs

14
and lack of validation will limit their rise (Lux Research Inc. 2012). Green roofs are seeing
tremendous growth in cities like Chicago, Portland, Manhattan and Washington DC, which has
set a goal of 20 percent green roof coverage by 2020 (Stutz 2010). One of the reasons for this
growth is the policies and tax incentives offered by the concerned legislations. There are federal
legislations like the US Green Building Council (USGBC), Environmental Protection Agency
(EPA) that set forth regulations engendering the use of green roof design and implementation to
become far more commonplace. The City of Chicago developed a Grant Program that offered up
to up to 50% of cost or $100,000 for green roof development of green roofs covering 50% or
more of a rooftop space. The city of Portland offers a Floor Area Ratio bonus in its building code.
Developers may build an extra 3 sq/ft per foot of green roof they construct without additional
permits. They also offer a grant reimbursement of up to $5 per sq/ft for reducing stormwater
infrastructure with a green roof (Plant Connection n.d.). Several other cities offer similar
incentives like grants, FAR bonus, expedited permits and tax reduction benefits, which in sum
have facilitated the growth of the green roof industry. Germany, widely considered a leader in
green roof research, technology, and usage, has had decades of experience with green roofs. An
estimated 10 percent of all flat roofs in Germany are rooftop gardens (Peck, et al. March 1999).
Cities like Stuttgart and Copenhagen have begun to mandate green roofs on most new
construction. A 2005 study calculated that if 75 percent of the flat roofs in Toronto were green, it
would save the city $37 million a year on storm water management, energy bills, and costs related
to urban heat island effects (Banting, Doshi, et al. 2005).

15
2.2. Field experiments and case studies on the thermal performance of
green roofs

Several research laboratories and universities have conducted studies on the thermal performance
of green roofs. A recent paper published by Professor Pablo La Roche of California State
Polytechnic University (Cal Poly), Pomona ,focuses on green roofs performing much better in
conjunction with other passive cooling techniques (night ventilation), in a hot dry climate. The
green roof testing was conducted at the Lyle Center for Regenerative Studies on the Cal Poly
Pomona campus situated in the hot/arid climate of Southern California. Prof. La Roche ran a
series of experiments using these test cells (and he continues to do so, in present day). The four
test cells that were compared include a traditional green roof with insulation underneath; a non-
insulated green roof; a green roof with an insulated plenum and a fan referred to as the variable
insulation green roof; and a test cell with a conventional insulated roof, as shown in Figure 13.

Figure 13: Test cell arrangements for La Ro che ’ s expe r i m ent s (La Roche, Carbonnier and
Halstead, Smart Green Roofs: Cooling with variable insulation 2012)

16

Figure 14: Details of the Smart Green roof test cell

(La Roche, Carbonnier and Halstead, Smart
Green Roofs: Cooling with variable insulation 2012)

The mass of the green roof acts as a heat sink, so they can contribute to cooling of spaces if the
mass of the soil is cooled by night ventilation, for example. His paper examines the passive
cooling potential of green roofs using a system that selectively couples and decouples the green
r oof ’ s m as s wi t h t he i nt er i o r env i r o nm ent t o i m pr ov e i ndoor t em per a t ur e s. Whe n t he f an i s ON the plenum is ventilated and when it is OFF the ceiling acts as an insulator. It would be
interesting to see whether mechanically ventilated buildings if provided with green roofs perform
bet t er t ha n i t w oul d, had i t bee n n at u r a l l y v ent i l at ed. L a Roche ’ s i dea of c om bi ni n g pas si v e
cooling techniques with green roofs will be applicable here.

Figure 15 : Psy chom et r i c c h ar t s s how i ng an ov er v i ew o f La R och e’ s e xp er i m ent r e sul t s (La Roche
2012)

17
In his Graduate thesis at the University of Southern California (USC) on ‘ P er f or m anc e o f R oof Materials High SRI, Low SRI, And Green Roof In California Climate Zone 8 Los Angeles,
C al i f or n i a ’ , M as t er of B u i l di ng Sci en ce s t ud ent Babak Zareiyan, focused on comparing the
performance of green roofs with other roofing types. In his study, he constructed four different
t es t ce l l s e a ch o f w hi ch di f f er ed onl y i n t he r oo f t y pe. H i s s e t up w a s s i m i l ar t o La R oche ’ s s e t up.
His experiments and analysis showed that an uninsulated green roof performs best, when
compared to the others (see Table 1).
Table 1: Exterior ambient temperature (Tamb), Average and Maximum interior temperatures (Ti)
for different roofing types- Nov to mid Feb (Zareiyan 2011)

Another study was conducted at Columbia University, Center of Climate Systems Research. This
research was done to compare the performance of three different roof surfaces on a real building.
Temperature data were analyzed from a 4 inch depth modular green roof, a nearly black roof, and
a hi g hl y r ef l e ct i v e whit e c o at i ng r oof m at er i a l a t t he C on Edis on “Le ar n i ng C ent e r ” ( T LC ) i n Long Island City (LIC), Queens, New York. This one year test resulted in an interesting
conclusion regarding heat flow through the roof layers. The study referred only to the heat
transfer through the roof. The data analysis and heat gain and loss calculations were intended to
find a way to save energy via design a more efficient roof for that specific climate. The study
showed that, although the green roof protected indoor environment and mitigated the urban heat
island effect, insulation has a great effect on the inside temperature. In terms of the urban heat
island effect, the green roof maintains the surface temperature lower than white roof (In summer,

18
white roof membrane temperature peaks were on average 30°F and green roof membrane
temperature peaks were on average 60°F cooler than black). Water runoff control, water and air
quality improvement, and building aesthetic values are other benefits of green roof while
impervious black or white roof based on the level of reflectivity just change urban albedo.

Figure 16: Comparative membrane temperatures for black, white and green roofs (Gaffin, et al.
2010)

Finding the heat flow for the winter and summer seasons and the relative energy benefits for the
three different roof assemblies was the main purpose of this report. Comparing the data recorded
by thermistors located above and below the roofs provided the results. The result proved that the
green roof is as effective as the most highly reflective material and that the surface temperature of
the green and white roof is fairly close together.
Dr. David J. Sailor, a Professor of Mechanical and Materials Engineering and Director of Green
Building Research Laboratory at Portland State University, studied the energy performance of
g r ee n r oo f s i n g r ea t de t a i l . I n one o f hi s pr es e nt a t i ons e nt i t l ed ‘ En er g y Per f o r m anc e of G r ee n
Roofs: Th e r ol e of t he r oo f i n a f f e ct i ng bui l di ng ene r g y and t h e u r ban at m osphe r i c env i r onm ent ’ ,
he explained the heat transfer mechanism in a green roof, discussed the green roof impacts on

19
roof temperatures and heat flux and also talked about impacts on whole bui l di ng e ner g y use .
cc or d i ng t o h i s r es e ar ch f i ndi ng s, a v er ag e r o of t op su r f ac e t em per a t ur e of a g r ee n r oof i s 3 - F
cool e r du r i ng a s um m er day and warm er by 1 F at n i g ht w hen c om par ed t o a st a ndar d r oo f . (D.
Sailor, EPA Heat Island Reduction Program Webcast 2010)
Dr. Steven Sandifer, in his PhD dissertation from University of California, Los Angeles (UCLA)
ent i t l ed ‘ U si ng t h e Land sc a pe f o r Pas si v e Coo l i ng and B i ocl i m at i c Co nt r ol : pp l i ca t i ons f o r higher density and larger sc al e ’ , pr es e nt s t he r e sul t s f r o m a s er i es of expe r i m ent s o n t he u se s of several less well studied elements of the landscape in the cooling of buildings; vines, landscape
ponds and vegetated roofs (Sandifer 2009). The experiments demonstrate that vegetated roofs,
among the other studied landscape elements, have the potential to reduce heat gain significantly
and perform well as bioclimatic elements in cooling strategies for buildings. Sandifer built typical
test roof modules and using thermocouples attached beneath each test module, he recorded the
surface temperature changes over time. The results from one typical day of study are shown in
Figure 17.

Figure 17: Daily temperature pattern (Sept 5) of different roofing types illustrating the overall
reduction in maximum temperature and suppression of the daily swing of the vegetated
roof (Sandifer 2009)

20
2.3. Energy modeling and computer simulations of green roofs
Research on green roofs has been covered through physical experiments and collection of real
building energy data. But energy modeling of green roofs is relatively unexplored, in comparison.
In earlier times, the industrial practices saw building energy modelers/ analysts simply model a
green roof as a conventional roof with a calculated cumulative R-value, when running a whole
building simulation. This approach might have been dictated by the absence of appropriate
simulation engines that took into account the dynamic behavior of vegetation. However, that
excuse no longer holds. There are many programs in the market currently that are capable of such
simulations. This literature study includes references to research done so far in the realm of
thermal modeling of green roofs using various energy modeling programs.
The challenges of modeling green roofs still remains unmet to an extent, although through the
help of various list reserves and research papers recently published, modeling of green roofs has
become relatively easy. One such paper is by Dr. David J. Sailor, who studies the energy
performance of green roofs through computational methods in great detail. I n hi s p ape r , ‘ A green
r oof m odel f or bui l d i ng ene r g y si m ul at i on p r og r am s’ Sai l or discusses the factors he considered
behind developing the algorithm for green roof (ecoroof) for EnergyPlus simulation engine. A
physically based model of the energy balance of a vegetated rooftop has been developed and
integrated into the EnergyPlus building energy simulation program. This green roof module
allows the energy modeler to explore green roof design options including growing media thermal
properties and depth, and vegetation characteristics such as plant type, height and leaf area index.
The model has been tested successfully using observations from a monitored green roof in
Florida. Figure 18 shows the comparative results from a section of his validation study. A
preliminary set of parametric tests has been conducted on prototypical 4000 m
2
office buildings in

21
Chicago, IL and Houston, TX. These tests focus on evaluating the role of growing media depth,
irrigation, and vegetation density (leaf area index) on both natural gas and electricity
consumption. Building energy consumption was found to vary significantly in response to
variations in these parameters. Further, this response depended significantly on building location
(climate). Hence, it is evident that the green roof simulation tool presented by him can serve a
valuable role in making informed green roof design decisions (D. J. Sailor 2008).

Figure 18: Green roof module predictions as compared to measured soil surface temperatures for
green roofs at the University of Central Florida test site. The figure panels represent
two weeks of hourly data within each of four seasons (D. J. Sailor 2008)

22
Sailor was also involved in the development of the online green roof energy calculator tool that is
free to use. The green roof energy calculator allows you to compare the annual energy
performance of a building with a vegetative green roof to the same building with either a dark
roof or a white roof. At the present time simulations are available for new construction (ASHRAE
90.1-2004) and old construction (pre-ASHRAE 90.1-2004) office and residential buildings driven
by typical precipitation and weather data. Representation of an irrigation schedule is optional.
This calculator was developed through a collaboration involving researchers like Sailor and staff
at Portland State University, University of Toronto, and Green Roofs for Healthy Cities. The
effort was funded by the US Green Building Council with additional financial and in-kind support
from University of Toronto, Portland State University, GRHC, and Environment Canada (Green
Roof Energy Calcultor 2012). sc r ee nsho t of t h e t oo l ’ s i nt e r f ac e i s sh own in Figure 19.

Figure 19: A screenshot of the online green roof energy calculator (Green Roof Energy Calcultor
2012)

23
This calculator expects the user to input certain data pertaining to the location, building and green
roof assembly. There is also an option to feed in the utility rates of the city, after which the
calculator generates a table showing the impact of the specified green roof. This includes the
annual energy savings compared to a dark and a white roof, the average sensible and latent heat
flux to the urban environment and the annual roof water balance, depending on whether the roof
is irrigated or not. An example of the impact of a green roof in Arizona, calculated by this tool is
shown in Figure 20. This is for a new office building in Phoenix, with a total roof area of 20,000
sq.ft. The green roof has a growing media depth of 4 inches, a leaf area index of 5, covers
approximately 100% of the total roof area, and is not irrigated.

Figure 20: Calculated impact of green roof using tool- screenshot

24
The research discussed so far mainly deals with the performance of green roofs and their
comparison to other roofing types like dark roofs and cool roofs. Some recent research has been
done on t he par am et r i c ef f e ct s o f t he g r ee n r oo f a ss em bl y on t he bui l d i ng ’ s energy loads through
simulation exercises. One such example is the study done by Issa Jaffal, Salah-Eddine
Ouldboukhitine and Rafik Belarbi, who published their findings in a research paper entitled ‘A
comprehensive study of the impact of green roofs on building energy performance’. A
mathematical model of green roof thermal behavior was developed and then coupled with the
building energy model in TRNSYS software by creating a new component. Then an evaluation of
the impact of a green roof on the energy performance of a single-family house with an area of 96
sq. m. was carried out. Detailed simulations were conducted for the house with conventional and
green roofs in a temperate French climate (La Rochelle). Mean and maximum indoor air
temperature and heating, cooling, and total energy demand in three different climates (Athens, La
Rochelle and Stockholm) was also studied later. It was observed that increasing the LAI reduces
the summer indoor air temperature and the cooling demand, but increases the heating demand.
Also, the impact of the variation of this parameter decreased at high levels (Table 2).

Table 2: Mean and maximum indoor air temperature and heating, cooling, and total energy
demand in La Rochelle for different LAI levels (Jaffal, Ouldboukhitine and Belarbi 2012)

25
When the summer indoor air temperature and the energy demand were evaluated for different
insulation depths in La Rochelle, the green roof was seen to reduce the mean and maximum
indoor air temperature by 6.5⁰ and 9.3 ⁰C, respectively, for the uninsulated roof. However, these
reductions were both less than 1.0 ⁰C in the case of the 30 cm insulated roof. Heating demand
reduction of 48% for the uninsulated green roof was observed. For thicknesses greater than 10
cm, the relative importance of this additional insulation, and thus the effect of the green roof on
the heating demand, became negligible. Also, impact of a green roof on the cooling demand
decreased as the insulation level increased (Table 3).

Table 3: Mean and maximum indoor air temperature and heating, cooling, and total energy for
different insulation levels (Jaffal, Ouldboukhitine and Belarbi 2012)

When the metrics of the temperate climate of La Rochelle were compared against a
Mediterranean climate of Athens and a cold climate of Stockholm, it was seen that the
effectiveness of green roofs also greatly depended on the climate. The enhancement of thermal
comfort and reduction of cooling demand were more effective when the climate is hot. However,
a significant reduction of the heating demand was observed in cold climates. For hot climates,
green roofs could increase the heating demand, but this increase was still minor when compared
with the cooling demand reduction. The total energy demand decreased with green roofs in hot,
temperate and cold climates (Table 4). This general improvement made them an energy-efficient
solution for a wide range of European climates.

26
Table 4: Mean and maximum indoor air temperature and heating, cooling, and total energy
demand in Athens, La Rochelle and Stockholm (Jaffal, Ouldboukhitine and Belarbi 2012)

Another parametric study of green roofs was carried out by a group of researchers in Cairo, Egypt
to determine the effectiveness of green roofs on reducing the energy consumption of a residential
building. This study was focused more on the impacts of various properties of soil like growing
media depth, thermal bulk properties of soil, plant height, stomatal conductance and soil moisture
conditions through irrigation. DesignBuilder was used as the simulation program to carry out the
study. The green roof assembly types were compared against the Egyptian code compliant
traditional roof and uninsulated roof. Soil conductivity proved to be the strongest parameter
affecting the energy savings of the building followed by soil depth. It was observed that as soil
conductivity decreased or soil depth increased, the energy consumption increased (Figure 21).

Figure 21: Percentage of effectiveness achieved by each type of green-roof over the Traditional
and the Un-Isolated roofs in Cairo-Egypt (Kamel, et al. 2012)

27
Chapter 3: Building Energy Performance Simulation

Computer simulation of building energy behavior is primarily useful for understanding the
operation of the building heating and cooling systems. Energy simulation results identify which
aspects of a building will use the most energy, and thus the design elements most in need of
improvement. Simulations, when coupled with preliminary costing information, can be the basis
for a cost-benefit analysis that identifies the green upgrades that provide cost-effective energy
savings and should be included in a building design. A careful analysis of energy simulations can
provide information that will allow informed decisions to be made during design, ensuring the
project budget is spent wisely (Enermodal Engineering 2008).

Computer simulations also offer
the advantage of standardizing the energy calculations (Kaplan and Caner 1992).

However, like
all computer programs, these energy simulation programs have very limited capabilities to
compensate for bad assumptions or sloppy input.

3.1. Building Energy Simulation Model
The two main components of a building energy simulation model are the building envelope and
its contents (floors, walls, ceiling, roof and occupants) and plant components (HVAC equipment
and other environmental control systems). Due to the complexity of a building model, computer
simulations can analyze the effects of different ECMs (Energy Conservation Measures) and their
complex interactions more efficiently, comprehensively and accurately than any other available
method. The building models can be subject to error. Modeling can be no more accurate than the

28
assumptions that lie behind both the proposed building and the baseline building models

(Kaplan
and Caner 1992).

3.2. Simulation Engine and Graphical User Interface
Most building energy simulation programs come with a graphical user interface (GUI) and an
actual simulation engine. The GUI is used to prepare simulation input files for the simulation
engine and to display the simulation results once the simulation is complete. The GUI determines
the ease of use of the simulation program while the simulation engine determines the reliability of
the simulations results.
There can be more than one GUI for the same simulation engine. The combination of the two is
critical, as a good GUI with a weak engine will not yield reliable results. Also, different GUIs are
known to handle different aspects of building design or operation with varying levels of
proficiency and accuracy.

Figure 22: EnergyPlus Simulation engine and its various GUIs

29
A wide variety of energy simulation software packages are available in the market today. Some
of the commonly used simulation engines are EnergyPlus, Blast, DOE-2 etc. and the front end
software like Design Builder, HEED, EnergyPro, eQuest, IES Pro, BEOpt are a few programs
that use these simulation engines.

Figure 23: Few energy modeling programs that are commonly used to run energy simulations

For this research, Design Builder, a 3D graphical design modeling and energy use simulation
program that uses the EnergyPlus simulation engine, is used.

3.3. Software overview
DesignBuilder is one of the most powerful and easy to use Graphical User Interface to
EnergyPlus available. It is a complete 3-D graphical design modeling and energy use simulation
program providing information on building energy consumption, CO2 emissions, occupant
comfort, daylighting effects, ASHRAE 90.1 and LEED compliance, and much more (Design
Builder 2012). It is a mature product which offers flexible geometry input and extensive material
libraries and load profiles. Also, Design Builder has quality control procedures which assure the
accuracy of the results as it has been tested under the comparative Standard Method BESTEST/

30
ASHRAE STD 140 (Reinhart and Ibarra 2009). The EnergyPlus simulation engine is well
i nt eg r at ed w i t hi n D e si g n B ui l d er ’ s e nv i r onm ent . The versions of the software used for the
purpose of this research are as follows-
 GUI: DesignBuilder v3.0.0.105
 Simulation Engine: EnergyPlus v7.0.0.036

3.4. The Ecoroof Model in Design Builder/EnergyPlus
In response to the need for green roof design tools, a computational model of the heat transfer
processes involved on a vegetated roof has been developed by Dr. David J. Sailor, as mentioned
earlier in Chapter Two. The Ecoroof (Green Roof) model, first introduced in EnergyPlus v2.0,
was developed at Portland State University, by Sailor and his students. It is based on the FASST
vegetation models developed by Frankenstein and Koenig for the US Army Corps of Engineers
(The Encyclopedic Reference to EnergyPlus Input and Output 2011). The model has been
integrated into the EnergyPlus building energy simulation program. In the following two sections
the model has been discussed in details.

3.4.1. Model Description
As implemented in E ner g y Plus t he g r ee n r oo f m odul e al l ow s t h e us er t o sp ec i f y “ ec or o of ” as t he
out e r l ay er of a r oo f t op co n st r u ct i on u si ng a “ Ma t e r i a l : R oof V eg et a t i on ” ob j ec t . T h e us e r ca n t h en
specify various aspects of the green roof construction including growing media depth, thermal
properties, plant canopy density, plant height, stomatal conductance (ability to transpire

31
moisture), and soil moisture conditions (including irrigation) (EnergyPlus Engineering Reference:
The Reference to EnergyPlus Calculations 2012).
The model accounts for:
• long wave and short wave radiative exchange within the plant canopy,
• plant canopy effects on convective heat transfer,
• evapotranspiration from the soil and plants, and
• heat conduction (and storage) in the soil layer

Figure 24: Screenshot of Ecoroof option within Design Builder
The model formulation includes the following:
• Simplified moisture balance that allows precipitation, irrigation, and moisture transport
between two soil layers (top and root zone).
• Soil and p l an t ca nopy ene r g y bal anc e bas ed on t he r m y C or ps of Eng i ne er s ’ F S ST vegetation models (Frankenstein and Koenig), drawing heavily from BATS (Dickenson
et al.) and SiB (Sellers et al.).

32
• Soil surface (T
g
) and foliage (T
f
) temperature equations are solved simultaneously each
time step, inverting the ConductionTransferFunction (CTF) algorithm to extract heat flux
information for the energy balance calculation.
The two simultaneous equations involved in the energy balance calculations (EnergyPlus
Engineering Reference: The Reference to EnergyPlus Calculations) are as follows-
Energy Budget in the foliage layer:

[

(

)

]

(

)

………… ……… ….( 1 )
Energy Budget in the soil layer:

(

)[

(

)

]

(

)

…….. ( 2)
The details of the parameterizations for each of the terms in these equations can be found in
Appendix A of this document.

3.4.2. Limitations of the Ecoroof model
For the simulation of this green roof model, certain parameters pertaining to the vegetation and
growing medium (soil) are required to be entered. The model has only been tested with the
ConductionTransferFunction (CTF) solution algorithm – a warning is issued when other solution
algorithm choices are selected. Also, for this algorithm to work, the parameters have specific
ranges for the data inputs. These ranges have been shown in Table 5.

33
Table 5: Ranges of data input for Ecoroof model (The Encyclopedic Reference to EnergyPlus
Input and Output 2011)
Parameter Data range Typical values
Height of plants 0.005m -1.0m
Leaf Area Index (LAI) 0.001- 5
Leaf Reflectivity 0.05- 0.5 0.18- 0.25
Leaf Emissivity 0.8 – 1.0
Minimum Stomatal Resistance 50s/m – 300s/m
Thickness (of layer) 0.05m – 0.7m 0.15m- 0.3m
Conductivity of Dry Soil 0.2W/(m-K) – 1.5W/(m-K) 0.3W/(m-K) - 0.5W/(m-K)
Density of Dry Soil 300 kg/m
3
- 2000 kg/m
3
400 kg/m
3
- 1000 kg/m
3

Specific Heat of Dry Soil >0.0 J/(kg-K)
Thermal Absorptance 0.0- 1.0 0.9- 0.98
Solar Absorptance 0.0- 1.0 0.6- 0.85
Visible Absorptance 0.5 – 1.0
Saturation Volumetric Moisture
Content of the Soil Layer
0.1- 0.5
Residual Volumetric Moisture
Content of the Soil Layer
0.01- 0.1
Initial Volumetric Moisture
Content of the Soil Layer
0.05- 0.5

34
Chapter 4: Methodology

The methodology that is followed to carry out this study is divided into three distinct phases, as
shown in the flow chart diagram below (Figure 25). They are described in detail in the following
sections.
MODEL SIMULATION PHASE ANALYSIS PHASE
VALIDATION PHASE

Figure 25: Methodology flow chart

4.1. Baseline model validation phase
This phase ’ s ai m is to see whether the green roof assembly can be modeled accurately enough to
emulate the real scenario. Upon validation, the same roof assembly is used on a hypothetical

35
building to test the different variables of the green roof. For the purpose of validation of the
baseline green roof assembly model a field experiment is first selected, against which the model
is tested. The chosen experiment was carried out by Dr. Steven Sandifer, who recorded his
findings as a part of his PhD dissertation on 'The use of landscape elements in passive cooling
strategies for buildings' at UCLA in 2008 and his research findings have been discussed earlier in
Chapter Two, under field experiments on the thermal performance of green roofs.
The same construction assembly that Sandifer used for his research is modeled in Design Builder
and the climate data recorded is collected from him in order to customize the climate data file that
will be used for the simulation. After editing the climate data to match the recorded climate data
of Los Angeles, the simulations are run to determine the thermal performance of the different
roofing types. The simulation results are then compared to the field experiment data that Sandifer
cited in his dissertation and the accuracy of the model is thus determined by observing the
deviation in simulation data when compared to field data. A benchmark of ± 10% difference
between the two data sets is assigned to determine whether the model is accurate in terms of
simulating the experimental green roof assembly mock up.

Figure 26 : Sand i f er ’ s g r ee n r oof as s em bl y se t up ( phot o cour t es y - Steven Sandifer)

36

Figure 27 : Sand i f er ’ s g r ee n r oof as s em bl y m odel ed i n D es i g n B ui l der

4.2. The simulation phase
Once the accuracy of the model is corroborated against field experiment data, this model is then
used for the future parametric simulations in the study, on a prototypical small sized office
building in the different climate zones. The aim of this phase is to determine how the different
parameters of the green roof assembly affect the thermal performance of the same in different
climate types, and to see how they perform against a standard cool roof. The performance metric
with which the comparison is made includes the percentage reduction in Energy Use Intensity
(EUI) of the building with a green roof over the one with a cool roof, in addition to annual and
peak heating and cooling loads and heat balance through the roof.
There are several parameters that may affect the performance of the green roof, like- insulation
and growing media thickness, plant type, roof slope, roof orientation, building footprint, building
height, shading, location or climate, building function etc. Table 6 shows a list of such parameters
and their subset variables. This list is by in no means exhaustive, but it gives an idea of the gamut
of variability in green roof design.

37
Table 6: List of Parameters related to a green roof assembly
Parameters Variable 1 Variable 2 Variable 3
Unit of measurement/
Performance metric
EUI Δ t (temperature
difference)
$ (cost)
Green roof orientation North South East
Thickness of growing
medium/ soil depth
2" 4" 6"
Thickness of insulation 0" 2" 4"
Thickness of drainage
medium
1" 2" 4"
Drainage system Drainage plates Drainage mats Granular drainage
Roof slope Flat roof Sloped roof
(15⁰, 30⁰)
Combined
Climate Hot humid Hot dry Cold dry
Shading Yes No Partial
Ratio of roof footprint 1:1 1:2 1:3
Building height 1 story 3 stories 5 stories
Floor to floor height 10' 12' 15'
Building function Residential Commercial Industrial
Vegetation type
(species)
Grass Sedum Short trees
Vegetation type (LAI) 0.5 1 5
Irrigation Yes (always) No (never) Smart (Schedule
based)
Test model (virtual) Dumpy house (box)-
test cell
Hypothetical
building
Existing building

In order to narrow down the parameters for parametric testing, a hierarchy of the aforementioned
parameters is made and the following three are selected-

38
• Insulation
• Type of vegetation (LAI)
• Growing media depth (or soil depth)
These parameters come with a subset of variables which are illustrated in Figure 28 and discussed
in details in the following sections. Each of these parameters is tested within the context of three
different climate types. A total of 24 different green roof assembly types are tested.

Figure 28: Test parameters and their subset variables

39
4.2.1. Testing Parameter 1: Insulation
Green roofs, due to the presence of vegetation and soil layers, act as a thermal mass themselves.
The green roof may or may not have additional insulation depending on the design of the
assembly. The insulation parameter in this study considers polystyrene insulation of varying
thicknesses. This insulation is placed under the waterproofing membrane and just above the
structural roof of the building. The four different variables to be investigated in this study include:
• ” t h i ck i ns ul a t i on: Wi t hou t i n sul at i on
• ” t h i ck i ns ul a t i o n
• 6” t h i ck i ns ul a t i on
• 8” t h i ck i ns ul a t i on

4.2.2. Testing Parameter 2: Vegetation Type (Leaf Area Index)
Green roofs can comprise of different vegetation depending on climate, geographic location and
desired aesthetics. Different plant types have different leaf area indices (LAI), which is broadly
defined as the amount of leaf area in a vegetation canopy per unit land area (Scurlock et al. 2001).
LAI is a critical parameter that is known to affect the heat fluxes between atmosphere and
vegetation. In the context of green roofs, it affects thermal performance of the roof assembly,
primarily on account of the amount of solar shading it provides to the roof surface. The
vegetation type parameter in this study includes the following two variables:

40
 LAI= 1
 LAI=5

Figure 29: Plants with different LAI; LAI=1(on the left), LAI=5 (on the right).

4.2.3. Testing Parameter 3: Growing Media or soil depth
The growing medium or soil depth of a green roof assembly is generally dictated by the plant
type. A green roof assembly with low vegetation will normally have a growing medium depth of
2 i nch es t o 6 i nch es . Su ch a n as s em bl y f al l s unde r t h e ‘ Exte ns i v e g r ee n r oof ’ t y pe ca t eg or y .
A green roof assembly that consists of a diversity of plants, including shrubs and trees, need a soil
dept h g r e at er t han 1 2 i n che s. Such an a s se m bl y f al l s u nder t he ‘ I nt ens i v e g r ee n r o of ’ t y pe
category. They generally entail a greater cost and require more maintenance. Green roof
assemblies with a soil depth ranging from 6 inches to 12 inches fall under the hybrid category of
‘ Sem i- I nt ens i v e g r ee n r oof s ’ . T hes e r o of s hav e a g r ea t e r di v er si t y of pl ant s t h an t h e e xt ens i v e
roof, but not the soil depth to support trees or larger shrubs. Plant material may include perennials
and small shrubs in addition to low-growing ground covers (DDC Cool and Green Roofing
Manual 2007).

41
For the purpose of testing the soil depth parameter, the vegetation type (LAI=5 or LAI=1) is
assumed to be able thrive at the different thicknesses of growing media being tested. The
following three variables are investigated:
• 3 ” t h i ck soi l ( ext ens i v e)
• 6” t h i ck soi l ( se m i -intensive)
• 12” t h i ck so i l ( i n t en si v e)

Figure 30: Different soil depth variables (Groundworks Sheffield)

4.2.4. The Climate Types
The parametric study is carried out for three different climate types. A typical city representing
each climate type is selected for the purpose of the simulation. They are as follows-
 Hot dry- Phoenix, Arizona (cooling dominated climate)
 Warm Coastal- Los Angeles, California (heating +cooling)
 Cool humid- Chicago, Illinois (heating dominated climate)

42

Figure 31: Heating and cooling degree day chart, highlighting the spectrum of climate types
covered in this study (National Climatic Data Centre 2002)

4.2.5. The Building Energy Model: Office Prototype
The parametric study is carried out on a prototypical small sized office building. The building is
modeled in Design Builder as a code compliant building with respect to California. References to
the following standards have been made, for modeling the building envelope and setting the
building schedules-
• Commercial Buildings Energy Consumption Survey (2003 CBECS)
• American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE
Standards 90.1-2004)
• Illuminating Engineering Society of North America (IESNA)
U.S. Department of Energy Commercial Reference Building Models of the National Building
Stock- a technical report by National Renewable Energy Laboratory (NREL) has largely been
referred to as well, for further building information and schedules. The operating schedules
derived from ASHRAE and modified by NREL, are included in Appendix B. Table 7 below
shows the building energy model (BEM) inputs.

Phoenix Los Angeles Chicago

43
Table 7: Building Energy Model Inputs
CATEGORY INPUT DATA
SOURCE
Orientation East-west (longer axis)
Area per floor 10,000 sq. ft.
Floor to floor height 13 ft. (9 ft. flr to ceiling + 4 ft. plenum)
Glazing fraction 0.33 2003 CBECS
Shading Interior blinds
Exterior walls Steel frame wall with insulation ASHRAE 90.1
Roof Cool roof- metal deck (basecase) ASHRAE 90.1
Green roof – metal deck (variable)
Floors Slab on grade floors, unheated+carpet ASHRAE 90.1
Windows Single pane, 0.25 SHGC ASHRAE 90.1
HVAC VAV with terminal reheat
Heating: Natural gas
Cooling: Electricity
2003 CBECS
SHW Fuel type: Natural gas
Lighting (LPD) 1 w/ft2 ASHRAE 90.1
Occupancy 0.005 people/ft2 ASHRAE 90.1

(a) (b)
Figure 32: Prototypical office building modeled in Design Builder- (a) Screenshot of the single
story building (b) Screenshot of the HVAC zoning diagram for the typical floor.

4.3. The Analysis phase
Once the simulations are complete and the energy loads on the building are computed, the data is
then analyzed to test the hypothesis- whether green roofs really reduce the energy loads on a

44
building, when compared to cool roofs. If yes, then by how much? In addition, which
combination of variables would result in the optimal thermal performance of the green roof?
The data analysis and conclusions resulting from the parametric simulations that are carried out
with different green roof assemblies are drawn and compared to base case. The analysis is
presented by climate type and under each climate type, all of the performance metrics and how
they are affected by the different test parameters are discussed in detail. All the important results
that can be drawn from the analysis are then neatly put into a final matrix that serves as the
guideline for green roofs and can be interpreted by anyone without going into the details of it.
This essentially includes the summary of the research in the form of comparison charts for each
performance metric.
An online green roof decision support tool may then be developed using these results and
summarized charts. The development of the website is left for future work, but an initial
conception and formulation of the tool is done by the author. Screenshots of the same are
included in Chapter 7 under future work.

45
Chapter 5: Green Roof model validation

5.1. Simulation results of the field experiment
As mentioned in Chapter Four under Methodology, in order to determine the accuracy of a green
r oof as s em bl y m odel , Sand i f e r ’ s r oo f a ss em bl y i s s e l e c t ed a nd m odel ed as pe r de s cr i ption in
Design Builder and the simulated data thus obtained is compared to his measured field data. The
performance metric that is considered here is the inner surface temperature of the roof assembly.
The simulation is carried out spanning a period of seven days (Aug 8- Aug 14), on an hourly
basis. The week selected for the test is a typical summer week for the location in which the actual
test is carried out.
The absolute percentage difference is calculated between the two data sets and a difference of ±
10% is kept as the benchmark for accuracy. It is seen that all of the simulated data falls within
the 10% difference of measured field data. The tables in Appendix C show the comparison of the
two data sets. The highlighted cells containing weather data indicate the final inputs in Design
Builder to generate the corresponding simulated data set. The following section deals with the
analysis of this data and the conclusions drawn from the comparison of the two data sets.

5.2. Analysis and validation of results
On comparing the results obtained through the simulations with the measured data, it is observed
that the absolute percentage difference over the week is below the 10% benchmark.

46
The simulated data is plotted against the measured data to generate a scatter plot diagram (Figure
33). This plot helps us to understand the amount of deviation between the two data sets and the
nature of deviation- whether it is positively or negatively correlated and whether the trend is
similar to the ideal case.

Figure 33: Measured versus simulated data scatter plot
It is observed that the simulated data is always higher than the measured data by around 3.9⁰ F. In
an ideal case, if the simulated data matched the measured data (i.e. S
g
= M
g
), all the points in the
above scatter plot would lie on the dotted line. But that is clearly not the case. Here S
g
= 1.06 M
g
.
Also, the trend line has a slope very similar to the ideal case, except it suffers from a lateral shift.
This shift may be attributed to one particular variable or to the fact that certain weather
parameters like solar radiation and humidity were not taken into account during the simulation of
R² = 0.6005
70.00
72.00
74.00
76.00
78.00
80.00
82.00
84.00
86.00
70.00 72.00 74.00 76.00 78.00 80.00 82.00 84.00 86.00
Simulated data (Sg)
Measured data (Mg)
S
g
=M
g
Trendline

47
the inner surface temperature of the green roof assembly. The cause for this lateral shift was
explored to an extent by the author but was ultimately left for future work due to time constraints.
However, it is interesting to note that when the simulated data set is multiplied by a factor
(inverse of the slope of the original plot), there is a substantially improved agreement between the
simulated and measured data sets. The slope of the trend line in Figure 33 is 1.06, so each
simulated result is multiplied by the inverse of that, i.e. 0.9434. The original and adjusted
correlation between the two data sets is shown in a table included in Appendix D.
The final scatter plot that is generated by plotting the adjusted simulated data against the original
measured data is illustrated in Figure 34. In this plot, the trend line tends to match the ideal case
more strongly.

Figure 34: Measured versus simulated data scatter plot after adjustment
Since the performance trend of the assembly has been established as reasonably close to the real
case, the same assembly is assumed fit to be used for further parametric testing.
70.00
72.00
74.00
76.00
78.00
80.00
82.00
84.00
86.00
70.00 72.00 74.00 76.00 78.00 80.00 82.00 84.00 86.00
Simulated data (Ts)
Measured data (Tm)
Sg' = Mg
Trendline

48
Chapter 6: Analysis and Data Interpretation

Once the baseline validation model is established to be emulating real life conditions with
reasonable accuracy, the same green roof assembly is used as the roof for the prototypical small
office building for further parametric testing. As mentioned earlier, this test is done for all the 24
different assembly types in 3 different climates, with one variable of a parameter being changed
with each simulation run with the purpose of understanding how that variable affects the different
thermal performance metrics that have been selected for this study. So once the building energy
model of the prototypical small office building is created, the parametric simulations are carried
out. The simulation results are presented in tables, included in Appendix E. They are arranged
according to climate type. Under each climate type, there are three tables:

 First table carries the annual simulated data related to each of the performance metrics
that were spoken about earlier, for all the 24 different green roof assemblies and a cool
roof assembly.
 Second table carries the simulated data for heat balance through the roof over the year,
for each month.
 Third table carries the simulated data for peak heating and cooling loads for the cool roof
assembly and the green roof assembly that has the lowest EUI among the 24 different
types.

49
The simulated data in these charts are further analyzed in this chapter. The nomenclature
followed here is the same as described below and remains consistent throughout the document.
L= Leaf Area Index (factor, unitless)
S= Soil depth (thickness in inches)
N= Insulation (thickness in inches)
Assembly type= LxSxNx
For e x am pl e, L5S3N i s a n as se m bl y w i t h L I =5, s oi l dept h = 3” and no i ns ul a t i o n. L1S 12N 8 i s
an a s se m bl y w i t h L I =1, s oi l d ept h= 6” and i nsu l a t i o n= 8”.
The analysis of the data derived through simulations is done according to climate type. The charts
and graphs in this chapter have all been derived from the data specified in the tables in Appendix
E. The thermal performance of the different assemblies is seen from the perspective of a single
parameter or metric in the following sections.

6.1. Parametric Analysis of data for Phoenix
6.1.1. Effect of Insulation on EUI
It is observed that when the insulation thickness is increased the EUI of the building increases.
T h i s i nc r ea se i s m or e s i g ni f i ca n t w hen t he i n sul at i on t h i ck nes s i s bum ped up t o 2 ” f r om ” t h i ck insulation. Increasing it further only increases the EUI by a small margin.

50

Figure 35: Effect of increasing insulation thickness on EUI
From Figure 35, the following observations can be made-
 Uninsulated assemblies have the lowest EUI.
 Differences in EUI between assemblies are more prominent when insulation is absent.
 As insulation thickness increases, the impact of increasing insulation reduces between the
different assemblies.
Whe n t he t h i ck nes s i s i nc r e as ed f r om ” t o 2”, t he EU I i nc r ea se s by 1.17 kBtu/sq.ft. on an
average, considering all assembly types. This is where the maximum impact of this parameter on
t hi s m et r i c occ ur s. Whe n i n sul a t i on t h i ck nes s i s i n cr e as ed f r om ” t o 8”, t he ov er al l EU I of t he
building increases by an average of 1.48 kBtu/sq.ft.

47.5
48
48.5
49
49.5
50
50.5
0" 2" 4" 8"
EUI (kBtu/ sq.ft)
Insulation thickness (in)
L5S3
L1S3
L5S6
L1S6
L5S12
L1S12
L= LAI
S= Soil depth (in)

51
6.1.2. Effect of LAI on EUI
The effect of increasing LAI is not that significant on the EUI of the building for this climate
type. The EUI of the building marginally increases with the increase in LAI.

Figure 36: Effect of increasing LAI on EUI
A few observations that can be made from Figure 36 are as follows-
 Largest impact of LAI on EUI is for assemblies with no insulation (greater slope) -
Lesser the impact when higher the insulation
 The general trend is applicable for all assembly types- marginal increase in EUI with
increase in LAI.
The LAI parameter has the greatest impact on the EUI of the building for S6N0 and S12N0
assembly types; the EUI increases by 0.3 kBtu/ sq.ft when LAI of the assembly is changed from 1
to 5. Otherwise all other assembly types show an average EUI increase of 0.13 kBtu/ sq. ft. Thus
this parameter has the least effect on the EUI performance metric, for this climate type.
47.5
48
48.5
49
49.5
50
50.5
1 5
EUI (kBtu/ sq.ft)
Leaf Area Index (LAI)
S3N0
S3N2
S3N4
S3N8
S6N0
S6N2
S6N4
S6N8
S12N0
S12N2
S12N4
S12N8

52
6.1.3. Effect of Soil depth on EUI
The effect of increasing soil depth is not that significant on the EUI of the building for this
climate type. The EUI of the building marginally increases with the increase in soil depth.

Figure 37: Effect of increasing soil depth on EUI
From Figure 37, the following observations can be made-
 Different assemblies follow similar trend when soil depth is increased, except for the
uninsulated green roof assemblies. While the general trend is a marginal increase in EUI
with increase in soil depth, the uninsulated assemblies show a drop in EUI first, when
i ncr ea s ed f r om 3” t h i ck soi l t o 6” t h i ck soi l , an d t h en i t shows an i nc r e as e l i k e t he ot he r assembly types.
 Uninsulated assemblies have the lowest EUI and greatest impact of increasing soil depth.
47.5
48
48.5
49
49.5
50
50.5
3" 6" 12"
EUI (kBtu/ sq.ft)
Soil depth (in)
L5N0
L5N2
L5N4
L5N8
L1N0
L1N2
L1N4
L1N8

53
Overall assemblies show minimal change in EUI with increasing soil depth, where the EUI
i ncr ea s es by .1 k B t u/ sq. f t on an av er ag e when soi l d ept h i s i nc r ea se d f r om 3” t o 12”. T h i s i s
taking all assembly types into consideration. The uninsulated assemblies however show a drop in
EUI by 0.37 kBtu/sq.f t w h e n soi l de pt h i s i nc r ea se d f r o m 3” t o 6” a nd t he n show a n i nc r ea se i n
EU I by . kB t u/ sq. f t w h en s o i l d ept h i s f u r t h er i nc r ea se d t o 1 2”. T h e L5N an d L1N ar e t h e
only two assembly types that show a noticeable impact on changing soil depth.

6.1.4. Reduction of EUI over base case
Since the thermal performance of the different green roof assemblies are compared to a cool roof
base case, the percentage reduction in EUI over the base case for each of these assembly types is
noted. It can be seen from Figure 38 that the uninsulated green roof assemblies have the largest
potential for reducing EUI for this climate type, irrespective of the other parameters (LAI and
Soil depth).

54

Figure 38: Percentage reduction in EUI over base case for different green roof assemblies
All the green roof assemblies show a reduction in EUI over the base case. The highest reduction
of 7.97% is caused by L1S6N0 and the lowest reduction of 4.16% is caused by L5S6N8 and
L5S12N8 assembly types. The average reduction is 5.17% when all assembly types are
considered and the maximum difference between EUI between the best and worst performing
assemblies is 1.98 kBtu/sq. ft.

6.1.5. Effect on annual heating and cooling loads
Green roofs are observed to be more effective in reducing the annual cooling loads than annual
heating loads for this climate type. Figure 39 shows the percentage of annual cooling and heating
load reductions over the cool roof base case. As can be seen from Figure 39, all the different
green roof assembly types show reductions in cooling loads but there are some assemblies which
show an increase in the annual heating loads.
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
Uninsulated 2" insulation 4" insulation 8" insulation
% EUI Reduction
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0
L5S3N2
L1S3N2
L5S6N2
L1S6N2
L5S12N2
L1S12N2
L5S3N4
L1S3N4
L5S6N4
L1S6N4
L5S12N4
L1S12N4
L5S3N8
L1S3N8
L5S6N8
L1S6N8
L5S12N8
L1S12N8

55

Figure 39: Reduction in annual heating and cooling loads over base case
Although assemblies like L5S3N0 and L1S3N0 show 21% reduction in annual cooling loads over
base case, they also suffer from an increase in annual heating loads by 30% and 26% respectively.
Generally heating costs are much greater than cooling costs for a building (Energy Information
Administration, EIA 1995), so if building operational costs are a driver for assembly selection,
one would be careful not to choose assembly types like L5S3N0 and L1S3N0. In such a case, it
would be a better option to go with assemblies like L1S3N8 and L1S12N8 which have a
relatively lower cooling load reduction potential than L5S3N0 and L1S3N0 but at least they do
not cause the heating loads to increase over base case. The former assemblies show an overall
heating-cooling load reduction of around 15% over base case. The average annual cooling load
reduction potential, when considering all assembly types, is 12.86% whereas the maximum
reduction is 21.37% (L5S3N0). The average annual heating load reduction potential, when
considering all assembly types, is -4.46% whereas the maximum reduction is 6.51% (L1S12N8).

-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
5%
10%
15%
20%
25%
L5S3N0
L5S6N0
L5S12N0
L1S3N0
L1S6N0
L1S12N0
L5S3N2
L5S6N2
L5S12N2
L1S3N2
L1S6N2
L1S12N2
L5S3N4
L5S6N4
L5S12N4
L1S3N4
L1S6N4
L1S12N4
L5S3N8
L5S6N8
L5S12N8
L1S3N8
L1S6N8
L1S12N8
Annual heating load reduction
Annual cooling load reduction

56
6.1.6. Effect on heat balance through the roof
Green roofs show a significant reduction in the heat fluctuation through the roof assembly, when
compared to a cool roof. This can be clearly seen in Figure 40 which shows the transfer of heat
through the roof over the period of a year, where the thick blue line indicates the cool roof base
case, the dotted lines are green roof assemblies with LAI=5 and the continuous lines are green
roof assemblies with LAI=1.

Figure 40: Heat balance through the roof over the year
While cool roof suffers from significant heat gains over the year, peaking at 1272 Btu/ sq.ft in
July, green roof assemblies suffer from an overall heat loss through the year. The hottest it gets is
in August, with an average of 366 Btu/ sq.ft heat loss. This reduction in heat balance through the
roof results in a longer life span of the green roof as compared to a cool roof, since greater heat
exchange through the roof means more structural damage over time. This phenomenon has a deep
-2500
-2000
-1500
-1000
-500
0
500
1000
1500
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Heat balance (Btu/ sq.ft)
Cool roof
L5S3N0
L1S3N0
L5S3N2
L1S3N2
L5S3N4
L1S3N4
L5S3N8
L1S3N8
L5S6N0
L1S6N0
L5S6N2
L1S6N2
L5S6N4
L1S6N4
L5S6N8
L1S6N8
L5S12N0
L1S12N0
L5S12N2
L1S12N2
L5S12N4
L1S12N4
L5S12N8
L1S12N8

57
impact on the life cycle cost analysis of such roofing types, and therefore can be another driver
for assembly selection.
A micro scale study of the heat balance through the roof is carried out to understand the thermal
performance in further detail. This exercise is done for the peak heating day (10
th
February) and
cooling day (23
rd
July) of that climate type with the assembly that has the lowest heat flux over
the year. In this case it is L1S12N8.

Figure 41: Comparison of heat flux through roof for peak heating and cooling days
From Figure 41, it is evident that not only are the peaks lower but also the diurnal heat fluctuation
through the green roof assembly is much lower than that of the cool roof. When the cool roof
starts to gain heat during the day, the green roof is still losing heat. A thermal lag of around 6 to 8
hours is observed. This would reflect on the energy costs, since the green roof assembly peaks at
an hour when the cooling/heating are not in peak demand and so the costs are lower.

-3
-2
-1
0
1
2
3
4
5
Heat balance (Btu/ sq.ft)
Time of day
Green roof-
Peak heating
Cool roof-
Peak heating
Green Roof-
Peak Cooling
Cool roof-
Peak Cooling

58
6.2. Parametric Analysis of data for Los Angeles
6.2.1. Effect of Insulation on EUI
It is observed that when the insulation thickness is increased the EUI of the building increases.
When the insulation thickness is increased up t o 2 ” f r o m ” the EUI of the building tends to
increase steeply and then with further increase in insulation thickness, the EUI marginally
decreases.

Figure 42: Effect of increasing insulation thickness on EUI
From Figure 42, the following observations can be made-
 Assemblies perform better without any insulation.
 Assemblies with same LAI have similar EUIs at higher insulation thicknesses.
 Differences in EUI between assemblies are more prominent when insulation is absent.
The insulation thickness parameter has nominal impact on the EUI of the building as compared to
the other parameters, for this climate type. An average increase in EUI of 0.82 kBtu/ sq.ft takes
39.4
39.6
39.8
40
40.2
40.4
40.6
40.8
41
41.2
0" 2" 4" 8"
EUI (kBtu/ sq.ft)
Insulation thickness (in)
L5S3
L1S3
L5S6
L1S6
L5S12
L1S12

59
pl ac e when i nsu l a t i o n t h i ck nes s i s i nc r ea se d f r om ” ( n o i nsu l a t i o n) t o 2” t h i ck insulation. This is
where the maximum impact of this parameter on this metric occurs. When insulation thickness is
increased from 2” t o 8 ”, t he ov er al l EU I of t he bu i l d i ng decreases by an average of 0.11
kBtu/sq.ft, when considering all assembly types, with the exception of L1S3 and L1S6 which
show an increase in EUI by 0.05 kBtu/sq.ft. When the effect of increasing insulation on EUI for
this climate is compared to that of Phoenix, it is observed that there is more variance between the
assemblies for Los Angeles.

6.2.2. Effect of LAI on EUI
The effect of increasing LAI is significant on the EUI of the building for this climate type. The
EUI of the building decreases with the increase in LAI.

Figure 43: Effect of increasing LAI on EUI
The following observations can be made from Figure 43-
39.4
39.6
39.8
40
40.2
40.4
40.6
40.8
41
41.2
1 5
EUI (kBtu/ sq.ft)
Leaf Area Index (LAI)
S3N0
S3N2
S3N4
S3N8
S6N0
S6N2
S6N4
S6N8
S12N0
S12N2
S12N4
S12N8

60
 Largest impact (greater slope) of LAI on EUI is for assemblies with greater insulation
thickness.
 The general trend is a decrease in EUI with an increase in LAI, with the exception of the
uninsulated assemblies- S6N0 and S12N0, which show a slight increase in EUI, as
opposed to a reduction.
The LAI parameter has the largest impact on the EUI performance metric of the building when
compared to the other parameters, for this climate type. An average reduction in EUI of 0.48
kBtu/ sq.ft can be achieved by increasing the LAI from 1 to 5, while a maximum reduction of
0.57 kBtu/ sq.ft occurs for S12N4 assembly type.

6.2.3. Effect of Soil depth on EUI
The EUI of the building marginally increases with the increase in soil depth, with the exception
of the uninsulated assemblies.

Figure 44: Effect of increasing soil depth on EUI
39.4
39.6
39.8
40
40.2
40.4
40.6
40.8
41
41.2
3" 6" 12"
EUI (kBtu/ sq.ft)
Soil depth (in)
L5N0
L5N2
L5N4
L5N8
L1N0
L1N2
L1N4
L1N8

61
From Figure 44, the following observations can be made-
 Assemblies with same LAI generally show similar impact on EUI when soil depth is
increased.
 The uninsulated assemblies L1N0 and L5N0 do not follow the general trend. In their case
EUI first decreases w hen so i l d ept h i s i nc r e as ed f r om 3” t o 6 ” a nd t he n i t s l i g ht l y increases from there on.
 The L5N8 assembly shows negligible change in EUI with increasing soil depth.
Overall assemblies show minimal change in EUI with increasing soil depth. The largest impact is
seen for L1N0, where the EUI decreases by 0.48 kBtu/ sq.ft when soil thickness is increased from
3” t o 1 2”. Otherwise, the other insulated assemblies show an average increase in EUI of 0.17
kBtu/ sq.ft.

6.2.4. Reduction of EUI over base case

Figure 45: Reduction of EUI over base case
0%
1%
2%
3%
4%
5%
6%
Uninsulated 2" insulation 4" insulation 8" insulation
% EUI Reduction
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0
L5S3N2
L1S3N2
L5S6N2
L1S6N2
L5S12N2
L1S12N2
L5S3N8
L1S3N8
L5S6N8
L1S6N8
L5S12N8
L1S12N8
L5S3N4
L1S3N4
L5S6N4
L1S6N4
L5S12N4
L1S12N4

62
All the different green roof assemblies perform better than the base case with respect to EUI as
they all show a reduction in EUI over the base case. The bar chart in Figure 45 clearly shows that
uninsulated assembly types cause the maximum reduction in the EUI of the building. From the
bar charts it can also be concluded that LAI plays a major role in the EUI reduction potential,
given a particular insulation thickness. The highest reduction of 5.34% is caused by L1S6N0 and
the lowest reduction of 1.68% is caused by L1S12N2 assembly type. The average reduction is
3.09% when all assembly types are considered and the maximum difference between EUI
between the best and worst performing assemblies is 1.53 kBtu/sq. ft.

6.2.5. Effect on annual heating and cooling loads
Green roofs are observed to be more effective in reducing the annual cooling loads than annual
heating loads for this climate type. Figure 46 shows the percentage of annual cooling and heating
load reductions over the cool roof base case.

Figure 46: Reduction in annual heating and cooling loads over base case
-15%
-10%
-5%
0%
5%
10%
15%
20%
25%
L5S3N0
L5S6N0
L5S12N0
L1S3N0
L1S6N0
L1S12N0
L5S3N2
L5S6N2
L5S12N2
L1S3N2
L1S6N2
L1S12N2
L5S3N4
L5S6N4
L5S12N4
L1S3N4
L1S6N4
L1S12N4
L5S3N8
L5S6N8
L5S12N8
L1S3N8
L1S6N8
L1S12N8
Annual Heating load reduction
Annual Cooling load reduction

63
As can be seen from Figure 46, L5assemblies are generally more effective in reducing cooling
loads than the L1 assembly types. Also, all the different green roof assembly types show
reductions in cooling loads but there are some assemblies which show an increase in the annual
heating loads like L5S3N0 which show 20.05% reduction in annual cooling loads over base case,
but they also suffer from an increase in annual heating loads by 8.86%. Keeping the building
operational costs in mind, it would be a better option to go with assemblies like L1S6N0 and
L1S12N0 which have a relatively lower cooling load reduction potential than L5S3N0 but at least
they do not cause the heating loads to increase over base case. The former assemblies show an
overall heating-cooling load reduction of around 22.25% over base case. The average annual
cooling load reduction potential, when considering all assembly types, is 11.27% whereas the
maximum reduction is 20.05% (L5S3N0). The average annual heating load reduction potential,
when considering all assembly types, is 2.19% whereas the maximum reduction is 9.06%
(L5S12N8).

6.2.6. Effect on heat balance through the roof
Green roofs show a significant reduction in the heat fluctuation through the roof assembly, when
compared to a cool roof. This can be clearly seen in Figure 47 which shows the transfer of heat
through the roof over the period of a year, where the thick blue line indicates the cool roof base
case, the dotted lines are green roof assemblies with LAI=5 and the continuous lines are green
roof assemblies with LAI=1.
While cool roof suffers from significant heat gains over the year, peaking at 577 Btu/ sq.ft in July,
green roof assemblies suffer from an overall heat loss through the year. The hottest it gets is in

64
July, with an average of 66.5 Btu/ sq.ft heat gain. There is more variance between the assemblies
in terms of heat loss in winter as opposed to the differences in heat gain during summer.

Figure 47: Heat balance through the roof over the year
As can be seen from Figure 47, while most of the assemblies have a lower heat balance than the
base case, some experience greater heat exchanges through the roof. Also, reduction in heat flux
through the green roof assembly results in a longer life span of the roof. Therefore, when life
cycle cost analysis of the roof is being considered, it is better to go with assemblies like L1S12N8
which has a lower heat flux rather than L5S3N0 which has a heat flux greater than the cool roof
base case. The overall heat fluctuation through the roof observed in this climate type is slightly
greater than that observed in Phoenix.
A micro scale study of the heat balance through the roof is carried out to understand the thermal
performance in further detail. This exercise is done for the peak heating day (6
th
January) and
-1400
-1200
-1000
-800
-600
-400
-200
0
200
400
600
800
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Heat balance (Btu/ sq.ft)
Cool roof
L5S3N0
L1S3N0
L5S3N2
L1S3N2
L5S3N4
L1S3N4
L5S3N8
L1S3N8
L5S6N0
L1S6N0
L5S6N2
L1S6N2
L5S6N4
L1S6N4
L5S6N8
L1S6N8
L5S12N0
L1S12N0
L5S12N2
L1S12N2
L5S12N4
L1S12N4
L5S12N8
L1S12N8

65
cooling day (6
th
September) of that climate type with the assembly that has the lowest heat flux
over the year. In this case it is L1S12N8.

Figure 48: Comparison of Heat flux through the roof for peak heating and cooling days
From Figure 48, it can be inferred that the pattern followed by the green roof is similar to what
was observed in the case of Phoenix. However, the only difference here is a shorter thermal lag of
around 3 hours. The heat flux during peak heating day is significantly reduced in comparison to
Phoenix.

-2
-1
0
1
2
3
Heat balance (Btu/ sq.ft)
Time of day
Green roof-
Peak heating
Cool roof-
Peak heating
Green Roof-
Peak Cooling
Cool roof-
Peak Cooling

66
6.3. Parametric Analysis of data for Chicago
6.3.1. Effect of Insulation on EUI
It is observed that when the insulation thickness is increased the EUI of the building decreases.
W hen t he i nsu l a t i on t h i ck n es s i s i ncr ea s ed up t o 2 ” f r o m ” the EUI of the building tends to
decrease steeply and then with further increase in insulation thickness, the EUI gradually
decreases.

Figure 49: Effect of increasing insulation thickness on EUI
This trend of decrease in EUI with increasing insulation thickness is unlike the trend seen for Los
Angeles or Phoenix. From Figure 49, a few more inferences can be drawn-
 Assemblies perform better (lower EUI) with greater insulation.
 Differences in EUI between assemblies are more prominent when insulation is absent,
and this difference gets minimized as the insulation thickness increases.
 At higher insulation thicknesses, effect of other parameters gets nullified.
52
53
54
55
56
57
58
0" 2" 4" 8"
EUI (kBtu/ sq.ft)
Insulation thickness (in)
L5S3
L1S3
L5S6
L1S6
L5S12
L1S12

67
The general trend seen here is almost the inverse of that occurring in Phoenix. As can be seen
f r om t he f i g ur e, at i nsu l a t i o n t hi ck nes s o f a r ound 8”, i t does no t m at t e r w ha t so i l d ept h o r L I t he
assembly has, because at that thickness all assemblies perform pretty much in the same way with
respect to EUI. When the thickness is i nc r ea se d f r om ” t o 2 ”, t h e EU I decreases by an average of
1.11 kBtu/sq.ft, considering all assembly types. This is where the maximum impact of this
par am et e r on t hi s m et r i c oc cur s. Whe n i n sul at i on t hi ck nes s i s i nc r ea se d f r om ” t o 8”, t h e ov er a l l EUI of the building decreases by an average of 2.33 kBtu/sq.ft., the maximum reduction of 3.55
kBtu/sq.ft being caused by assembly L1S3. The insulation thickness parameter has the greatest
impact on the EUI of the building as compared to the other parameters, for this climate type.

6.3.2. Effect of LAI on EUI
The effect of increasing LAI is not that significant on the EUI of the building for this climate
type. The EUI of the building is negligibly affected with the increase in LAI.

Figure 50: Effect of increasing LAI on EUI
52
53
54
55
56
57
58
1 5
EUI (kBtu/ sq.ft)
Leaf Area Index (LAI)
S3N0
S3N2
S3N4
S3N8
S6N0
S6N2
S6N4
S6N8
S12N0
S12N2
S12N4
S12N8

68
A few observations that can be made from Figure 50 are as follows-
 EUI of different assemblies almost remain constant when LAI is increased from 1 to 5,
with the exception of the uninsulated assemblies that show a slight change (decrease).
 The impact of this parameter on the metric is felt only for the uninsulated assembly types,
and that too is a nominal impact.
Unlike in the case of Los Angeles where LAI affects the thermal performance of the assembly, in
Chicago, the increase of LAI has negligible impact on the EUI of the building, with a nominal
impact caused by S3N0 assembly type that decreases EUI by 0.51 kBtu/ sq.ft when LAI is
increased from 1 to 5.

6.3.3. Effect of Soil depth on EUI
The EUI of the building marginally decreases with the increase in soil depth, with the exception
of the uninsulated assemblies.

Figure 51: Effect of increasing soil depth on EUI
From Figure 51, the following observations can be made-
52
53
54
55
56
57
58
3" 6" 12"
EUI (kBtu/ sq.ft)
Soil depth (in)
L5N0
L5N2
L5N4
L5N8
L1N0
L1N2
L1N4
L1N8

69
 Assemblies with same insulation thickness generally show similar impact on EUI when
soil depth is increased.
 The uninsulated assmeblies L5N0 and L1N0 show maximum impact of changing soil
depth.
 The impact of increasing soil depth on EUI reduces with increasing insulation thickness.
Overall assemblies show very little change in EUI with increasing soil depth. The largest impact
is seen for L1N0, where the EUI decreases by 2.18 kBtu/ sq.ft when soil thickness is increased
f r om 3” t o 1 2”. Ot her w i se , all other insulated assemblies show an average increase in EUI of 0.29
kBtu/ sq.ft.

6.3.4. Reduction of EUI over base case

Figure 52: Percentage reduction in EUI over base case for different green roof assemblies
It can be seen from Figure 52 that the insulated green roof assemblies have the largest potential
for reducing EUI for this climate type, irrespective of the other parameters (LAI and Soil depth).
-5%
-4%
-3%
-2%
-1%
0%
1%
2%
3%
Uninsulated 2" insulation 4" insulation 8" insulation
% EUI Reduction
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0
L5S3N0
L1S3N0
L5S6N0
L1S6N0
L5S12N0
L1S12N0

70
While some of the assemblies (mainly the uninsulated ones) have an EUI greater than the cool
roof base case, most of them perform better than base case with respect to EUI. The bar chart in
Figure 52 clearly shows that insulation thickness causes the maximum difference in the EUI
reduction potential of that assembly type, in this climate. The average EUI reduction over base
case when considering all the different assembly types is 0.38%. The L1S12N8 assembly type has
the highest percentage of EUI reduction of 2.40% over base case and L1S3N0 assembly type is
the worst performing one that shows an EUI increase of 4.25% over base case. The difference
between the best and worst performing assemblies in terms of EUI is 3.66 kBtu/ sq.ft.

6.3.5. Effect on annual heating and cooling loads
Green roofs are observed to be more effective in reducing the annual cooling loads than annual
heating loads for this climate type as well (Figure 53).

Figure 53: Reduction in annual heating and cooling loads over base case
Figure 53 shows the percentage of annual cooling and heating load reductions over the cool roof
base case. As can be seen from the bar chart, assemblies with lower insulation thickness and soil
-20%
-15%
-10%
-5%
0%
5%
10%
15%
20%
L5S3N0
L5S6N0
L5S12N0
L1S3N0
L1S6N0
L1S12N0
L5S3N2
L5S6N2
L5S12N2
L1S3N2
L1S6N2
L1S12N2
L5S3N4
L5S6N4
L5S12N4
L1S3N4
L1S6N4
L1S12N4
L5S3N8
L5S6N8
L5S12N8
L1S3N8
L1S6N8
L1S12N8
Annual Heating load reduction
Annual Cooling load reduction

71
depth show an increase in annual heating loads over the base case, although they reduce the
annual cooling loads by up to 17.71 % (assembly L5S6N0). Keeping the building operational
costs in mind, it would be a better option to go with assemblies like L5S12N8 and L1S12N8
which have a relatively lower cooling load reduction potential than L5S6N0 but at least they do
not cause the heating loads to increase over base case. The former assemblies show an overall
heating-cooling load reduction of around 10.29% over base case. The average annual cooling
load reduction potential, when considering all assembly types, is 9.39% whereas the maximum
reduction is 17.71% (L5S6N0). The average annual heating load reduction potential when
considering all assembly types is -3.01% whereas the maximum reduction is 3.99% (L1S12N8).

6.3.6. Effect on heat balance through the roof
Green roofs show a significant reduction in the heat fluctuation through the roof assembly, when
compared to a cool roof. This can be clearly seen in Figure 54 which shows the transfer of heat
through the roof over the period of a year, where the thick blue line indicates the cool roof base
case, the dotted lines are green roof assemblies with LAI=5 and the continuous lines are green
roof assemblies with LAI=1.
While cool roof suffers from significant heat gains over the year, peaking at 624 Btu/ sq.ft in July,
green roof assemblies suffer from an overall heat loss through the year. The hottest it gets is in
July, with an average of 32.6 Btu/ sq.ft heat gain.

72

Figure 54: Heat balance through the roof over the year
From the figure above, it can be inferred that there is more variance between the assemblies in
terms of heat loss in winter as opposed to the differences in heat gain during summer. Also,
reduction in heat flux through the green roof assembly results in a longer life span of the roof. So
it would be a better option to consider L5S12N8 assembly type which has a lower heat flux than
L1S3N0 which has a heat flux higher than the base case. Overall, the assemblies have greater
variance in heat loss during winter months than what was observed in the case of Phoenix or Los
Angeles.
A micro scale study of the heat balance through the roof is carried out to understand the thermal
performance in further detail. This exercise is done for the peak heating day (7
th
January) and
cooling day (16
th
July) of that climate type with the assembly that has the lowest heat flux over
the year. In this case it is L5S12N8.
-4000
-3500
-3000
-2500
-2000
-1500
-1000
-500
0
500
1000
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
Heat balance (Btu/ sq.ft)
Cool roof
L5S3N0
L1S3N0
L5S3N2
L1S3N2
L5S3N4
L1S3N4
L5S3N8
L1S3N8
L5S6N0
L1S6N0
L5S6N2
L1S6N2
L5S6N4
L1S6N4
L5S6N8
L1S6N8
L5S12N0
L1S12N0
L5S12N2
L1S12N2
L5S12N4
L1S12N4
L5S12N8
L1S12N8

73

Figure 55: Comparison of heat flux through roof for peak heating and cooling days
From Figure 55, it can be inferred that the pattern followed by the green roof is similar to what
was observed in the case of Phoenix or Los Angeles. The thermal lag here is also around 3 hours.
The heat flux during peak heating day is significantly greater in comparison to Phoenix or Los
Angeles and during peak cooling day, it is lower.

6.4. Summary: Comparison of the assemblies against different climates
The following figures summarize the above analyses and compare the results between different
climates. Two types of graphics have been selected to demonstrate the results- one is a matrix of
scatterplots that show the correlation between the assembly parameter and the performance
metric. The slope of the trend line indicates the sensitivity of that parameter to the performance
metric. The other graphic is a bar chart or line chart that highlights the assembly with the
strongest impact on that particular performance metric. This is shown for all three climate types.
-6
-5
-4
-3
-2
-1
0
1
2
3
4
Heat balance (Btu/ sq.ft)
Time of day
Green roof-
Peak heating
Cool roof-
Peak heating
Green Roof-
Peak Cooling
Cool roof-
Peak Cooling

74
6.4.1. Performance metric: EUI
Table 8 gives an idea of the thermal performance of the best and worst green roof assembly type,
with respect to EUI reduction over base case.
Table 8: Summary of EUI reduction- best and worst case scenario
Metric % Reduction in EUI over base case (kBtu/ sq.ft)
Phoenix
(Assembly type)
Los Angeles
(Assembly type)
Chicago
(Assembly type)
Green roof- Highest (best) 7.97
(L1S6N0)
5.34
(L1S6N0)
2.40
(L1S12N8)
Green roof- Lowest (worst) 4.16
(L5S6N8)
1.68
(L1S12N2)
-4.25
(L1S3N0)

The above data has been highlighted in the bar charts shown in Figure 56, and compared to the
other assembly types for each climate type. From the figure it can be inferred that Phoenix has the
highest reduction in EUI while Chicago has the lowest EUI reduction caused by green roofs. All
assemblies for Phoenix and Los Angeles perform better than base case, but there are 10 assembly
types that perform worse than base case for Chicago. The negative impact of these assemblies
over base case has been highlighted in the red dotted box.

75

Figure 56: Comparison of EUI Reduction of different assemblies over base case

76
It can be seen from the scatterplot matrix in Figure 57 that the strongest parameter (observing
slope of trend line) affecting the EUI of the building for Phoenix and Chicago is the insulation
thickness, although the nature of their correlation is different. For Phoenix, insulation thickness is
positively correlated to EUI while for Chicago it is the reverse. For Los Angeles, the LAI
parameter affects the EUI more than the other two parameters. All the green roof assemblies
perform better than base case for Phoenix and Los Angeles, but not for Chicago. Also, the
variance in EUI between the different green roof assemblies is more prominent in Chicago than
for any other climate type.

Figure 57: Parametric comparison of EUI for different roof assemblies

77
6.4.2. Performance metric: Annual Heating and Cooling Load
Table 9 gives an idea of the maximum heating and cooling load reduction potential of a green
roof assembly over base case.
Table 9: Summary of annual heating and cooling load reductions over base case
Metric % Reduction over base case
Phoenix
(Assembly type)
Los Angeles
(Assembly type)
Chicago
(Assembly type)
Maximum annual cooling load
reduction
21.37
(L5S3N0)
20.05
(L5S3N0)
17.71
(L5S6N0)
Maximum annual heating load
reduction
6.51
(L1S12N8)
9.06
(L5S12N8)
3.99
(L1S12N8)
Maximum overall annual
heating cooling load reduction
15.95
(L1S12N8)
22.55
(L1S6N0)
10.36
(L5S12N8)

The above data has been highlighted in Figure 58, which shows the percentage reduction of
annual heating and cooling loads over base case. The maximum reduction and average reductions
are mentioned on each chart for each climate type.

78
Figure 58: Parametric Comparison of Peak cooling load for different roof assemblies

79
Since green roofs are observed to be more effective in reducing the cooling loads of a building,
this phenomenon has been studied in a little more detail than the rest of the metrics. The effect of
green roof assemblies on the annual cooling loads gives an idea of the seasonal impact of the roof
assembly on the building. It can be seen from the scatterplot matrix in Figure 59 that the strongest
parameter (observing slope of trend line) affecting the annual cooling loads of the building is the
insulation thickness for all climate types. It is also interesting to note that a larger LAI results in
lower cooling loads for all climates. Theoretically, greater LAI means more shading on the roof,
which in turn implies reduced solar heat gains through the roof. This would logically result in
lower cooling loads that can be seen in Figure 58, but when we see the whole building energy
use, the impact of LAI gets weakened or overridden by other factors. The impact of the LAI
parameter on the annual cooling loads of the building is most strongly felt in Los Angeles, while
the impact of the soil depth parameter is most felt in Phoenix. Also, the variance in this metric
between the different green roof assemblies is more prominent in Phoenix than for any other
climate type. Overall, all the different green roof assembly types perform better than the cool roof
base case with respect to this performance metric, in all climate types.

80

Figure 59: Parametric comparison of annual cooling loads for different roof assemblies
The effect of the different green roof assemblies is also observed on a diurnal scale and the
correlation between the parameters is compared across the three different climate types. The peak
cooling day for each climate type is noted and the peak cooling load on that day is plotted to
generate the final scatterplot matrix (Figure 60).

81

Figure 60: Parametric comparison of peak cooling loads for different roof assemblies
From the figure above, an estimate of the effect of the different parameters on the peak cooling
loads can be obtained. The conclusions are not the same as what was observed in the case of
annual cooling loads. Here, the strongest parameter affecting the performance metric is the
insulation thickness for Phoenix, LAI for Los Angeles and soil depth for Chicago. It is interesting
to note that the insulation thickness has hardly any effect on this performance metric for Los

82
Angeles. Overall, just like in the case of annual cooling loads, all the assemblies perform better
than the base case for this metric too.

6.4.3. Performance metric: Annual Heat Balance through the roof
Table 10 gives an idea of the heat flux through the roof assembly over the year for different
roofing types- base case and best and worst green roof assembly types (in terms of highest and
lowest heat flux).
Table 10: Summary of heat flux through the roof- best and worst scenario
Metric Heat flux (Btu/ sq.ft)
Phoenix
(Assembly type)
Los Angeles
(Assembly type)
Chicago
(Assembly type)
Base case- Maximum 1696

913

1980

Green roof- Maximum (worst) 1296
(L5S3N0)
1360
(L5S3N0)
3945
(L1S3N0)
Green roof –Minimum (best) 98
(L1S12N8)
231
(L1S12N8)
677
(L5S12N8)

The best and worst cases are plotted in the charts shown in Figure 61 and are compared to the
base case cool roof. The reduction over base case is shown as a percentage, for the peak month of
July, for each climate type.

83

Figure 61: Parametric Comparison of Peak heating load for different roof assemblies

84
Chapter 7: Conclusion

7. 1. Conclusions drawn from study
The parametric study of the green roof assemblies shows that they have a positive thermal impact
on the building and have the potential to reduce the energy use intensity of the same, in
comparison to a cool roof. The extent to which it does that, however, depends on the climate type.
A few conclusions that can be drawn from this parametric study are discussed in the following
paragraphs.

Different parameters affect the thermal performance of the building to varying extents. The
matrix in Table 11 shows the correlation (positive or negative) between the different parameters
and performance metrics, for each climate type. The color coding indicates the extent to which
the metric is affected by changing the parameter- from strong to weak.

85
Table 11: Correlation between parameters and metrics

This basically summarizes the conclusions drawn from Figures 56, 58 and 59. The matrix tells the
reader whether the metric increases or decreases on increasing or decreasing the parameter and
the strength of impact that parameter has on that metric.

86
Insulation thickness is the parameter that affects the EUI of the building the most in Phoenix and
Chicago. It increases the EUI by 1.48 kBtu/ sq.ft in Phoenix while decreases the EUI by 2.33
k B t u/ sq.f t i n C hi ca g o w he n i nsu l a t i o n t h i ck nes s i s i n c r ea s ed f r om ” t o 8 ”. For L os ng el es , t he
LAI has the largest impact, where it reduces the EUI by up to 0.48 kBtu/ sq.ft when LAI is
increased from 1 to 5. The soil depth parameter affects the EUI nominally for Phoenix and Los
Angeles, but for Chicago the impact of changing the soil depth is a little stronger.

Overall, green roofs are more effective in reducing the cooling loads of a building. Some
assemblies can result in higher heating loads, so a careful selection of assembly type is important,
in order to keep the building operational costs low.

Significantly lower solar heat gains on the green roof are observed when compared to the cool
roof base case. The diurnal heat flux (swing) and the heat exchange through the roof assembly
over the year is also less than that of a cool roof, for most of the assemblies. For Phoenix, the heat
gain reduction can be up to 90% when compared to a cool roof, while in Los Angeles, the
reduction is up to 74% and in Chicago it is 65% reduction. Therefore the green roof has a much
longer life span than a cool roof, since structural damage of the roof due to heat flux is reduced
over time. This plays a big role in the life cycle cost analysis of a building, and so careful
selection of assembly type is important here as well, since some of these assemblies have a
greater heat flux than the cool roof, depending on the climate that is being considered.

In case of the green roof, a thermal lag is also noted. The green roof mainly loses heat during the
day and gains heat at night, as opposed to a cool roof which gains heat during the day (thereby
increasing cooling loads on the building when the peak demand for electricity is high) and loses

87
heat at night (thus increasing the heating load on the building). Due to the thermal lag provided
by the green roof, the latter peaks at a time when the electricity costs are not at peak, therefore
impacting the building operational costs over time.

There are several other parameters of the green roof that may be tested to determine the
corresponding impacts on the thermal performance of the same, although only three such
parameters are covered within the scope of this thesis. Also, the energy savings may be different
for different climates. These areas can be covered under future work.

7. 2. Limitations of study
The validation study done prior to the parametric testing to determine the accuracy of the green
roof assembly model has certain limitations that are important to note. Firstly, the performance
metric selected for the validation study was not one of the metrics chosen by the author to do the
parametric testing. The only data that was obtained by the author from the experimenter was the
inner surface temperature of the green roof membrane, hence the study was carried out using that
metric. When simulating the same, weather parameters like solar radiation, relative humidity etc.
were not taken into consideration because of the way EnergyPlus calculates the surface
temperature of an item. However, during the whole building energy simulation, these weather
parameters were considered. The validation results showed a lateral shift from the ideal case, and
this may be attributed to multiple reasons or just one variable. The possible reasons for this shift
has been discussed earlier, but has not been narrowed down to one cause.
For the parametric study, only three parameters were selected with a limit on the number of
subset variables each parameter came with. There are many more variables that may be tested but

88
were not due to limitation of time and scope of research. Also the number of climate types was
limited to three.
It was assumed that 100% of the building roof area was a green roof, which may not be the case
always. This research did not address the green roof coverage aspect due to limitation of the
green roof model integrated in DesignBuilder.

7. 3. Future Work
Some of the limitations of the study that are discussed in the above section can qualify as future
work in this area of research. This includes the number of parameters being tested and the
different climates under which the testing is done. A detailed list of parameters and their subsets
can be found enlisted in Table 6: List of Parameters related to a green roof assembly.
Another area to investigate is the use of different simulation softwares to see which one simulates
the behavior of green roofs more accurately and effectively. A comparative analysis of softwares
will benefit to highlight which program or simulation engine is preferable for such analyses.
The output of this research can be in the form of a decision support tool available to general
public via the internet. The tool, in the form of a website, will contain the findings of this research
and can expand in terms of content when further research on this topic is carried out with more
parameters and climate types, to make it more extensive. The tool has been conceived and
developed to a preliminary stage by the author to incorporate the current findings of this research
that are covered in this document. The website (currently hosted by Wix) can be assessed by
following this link-http://ihatefish87.wix.com/green-roof. A few screenshots with captions
explaining the tool are given below.

89

Figure 62: Home screen (screenshot) - Explanation about the tool

90

Figure 63: User is first asked to select climate zone or type from map or by clicking on
appropriate button below

91

Figure 64: User is then asked to select the desired variable. Upon selection, user is automatically
redirected to new window where the relevant charts or graphics are shown.

92
Bibliography
Banting, Doug, et al. "Report on the Environmental Benefits and Costs of Green Roof
Technology for the City of Toronto." Department of Architectural Science, Ryerson
University, 2005.
Cantor, Steven L. "Green Roofs in Sustainable Landscape Design." W.W. Norton & Company,
2008.
"DDC Cool and Green Roofing Manual." Manual, New York City Department of Design and
Construction, New York, 2007.
Design Builder. Design Builder. 2012. http://www.designbuilderusa.com/ (accessed November
2012).
Energy Information Administration, EIA. Energy Use and Cost of Office Buildings. 1995.
http://www.eia.gov/emeu/consumptionbriefs/cbecs/pbawebsite/office/office_howuseener
gy.htm (accessed 2013).
"EnergyPlus Engineering Reference: The Reference to EnergyPlus Calculations." EnergyPlus
Documentation, October 2012.
Enermodal Engineering. "Energy Modeling: at the heart of green building design." Enermodal
Engineering, July 2008.
Gaffin, S.R., C. Rosenzweig, J. Eichenbaum Pikser, R. Khanbilvardi, and T. Susca. A
Temperature and Seasonal Energy Analysis of Green, White, and Black Roofs. New
York: Center for Climate Systems Research, Columbia University, 2010.
Green Roof Energy Calcultor. Green Building Research Laboratory. 2012.
http://greenbuilding.pdx.edu/GR_CALC_v2/grcalc_v2.php#retain (accessed October
2012).
Green Roof for Healthy Cities, GRHC. Annual Green Roof Industry Survey for 2011. Survey,
GRHC, May 2012.
Groundworks Sheffield. Benefits of Green Roofs. http://www.greenroofguide.co.uk (accessed
November 2012).
Hundertwasser. n.d. http://www.hundertwasser.at/index_en.php (accessed February 20, 2013).
International Green Roof Association, IGRA.

93
Jaffal, Issa, Salah-Eddine Ouldboukhitine, and Rafik Belarbi. "A comprehensive study of the
impact of green roofs on building energy performance." Renewable Energy Renewable
Energy 43 (2012): 157-164.
Kamel, Basil, Sherine Wahba, Khaled Nassar, and Ahmed Abdelsalam. "Effectiveness of Green-
Roof on Reducing Energy Consumption through Simulation program for a Residential
Building: Cairo, Egypt." Construction Research Congress 2012. ASCE, 2012.
Kaplan , M, and P Caner. "Guidelines for Energy Simulation of Commercial Buildings." 1992.
Kaplan, M., and P. Caner. "Guidelines for Energy Simulation of Commercial Buildings." Report
for Bonneville Power Administration, 1992.
Kruche, P., D. Althaus, and I. Gabriel. "Okologisches Baun." Herausgegeben vom
Umweltbundesampt, Wiesbaden und Berlin Bauverlag. 1982.
La Roche, Pablo. "Cooling with Variable Insulation Green Roofs: Experimental Series in a Hot
Dry Climate." International Workshop on Environment and Energy, December 4-7.
Greenbelt, MD: Goddard Spce Flight Center, 2012.
La Roche, Pablo, Eric Carbonnier, and Christina Halstead. "Smart Green Roofs: Cooling with
variable insulation." PLEA2012 - 28th Conference on Passive and Low Energy
Architecture, November 7-9. Lima, Peru: PLEA, 2012.
Lux Research Inc. "Burgeoning Green Roofs and Green Walls Market to be Worth $7.7 Billion in
2017." The Wall Street Journal (Business Wire), October 2012.
Nagler, Erich. "Illustration: Design Means." Design Means: The Work of Erich Nagler.
Metropolis Magazine. 2008.
http://www.designmeans.com/work/illustration/green_roof.html (accessed October
2012).
National Aeronautics and Space Administration, NASA.
National Climatic Data Centre, NCDC. Monthly Station Normals of Temperature,
Precipitationand Heating and Cooling Degree Days 1971-2000. U.S. Department of
Commerce, 2002.
Peck, Steven W., Chris Callaghan, Monica E. Kuhn, and Brad Bass. "Greenbacks from Green
Roofs: Forging a New Industry in Canada." March 1999.
Plant Connection. Green Roof Legislation, Policies & Tax Incentives.
http://www.myplantconnection.com/green-roofs-legislation.php (accessed January 15,
2013).

94
Pomerantz, M., B. Pon, H. Akbari, and S.C. Chang. The Effect of Pavement Temperatures on Air
Temperatures in Large Cities. Heat Islands Group, Berkeley, CA: Lawrence Berkeley
National Laboratory, 2000.
Reinhart, Christopher, and Diego Ibarra. "Design Builder/EnergyPlus, Tutorial#1, Getting
Started." Building Performance Simulation for Designers- Energy. Harvard University-
Graduate School of Design, September 30, 2009.
Ryerson University. "Report on the Environmental Benefits and Costs of Green Roof Technology
for the City of Toronto." Department of Architectural Science, Ryerson University, 2005.
Sailor, D.J. "Energy Performance of Green Roofs: The role of the roof in affecting building
energy and the urban atmospheric environment." EPA Heat Island Reduction. June 3,
2010.
"Urban Heat Islands, Opportunities and Challenges for Mitigation and Adaptation." North
American Urban Heat Island Summit. Toronto, Canada: Data courtesy Entergy
Corporation, 2002. Sample Electric Load Data for New Orleans, LA (NOPSI, 1995),
May 1-4.
Sailor, David J. "A green roof model for building energy simulation programs." Energy and
Buildings 40 (8) (2008): 1466-1478.
Sandifer, Steven A. "Using the Landscape for Passive Cooling and Bioclimatic Control:
Applications for higher density and larger scale." PLEA2009 - 26th Conference on
Passive and Low Energy Architecture, June 22-24. Quebec City, Canada: PLEA, 2009.
Shickman, Kurt. "Introduction to Cool Roofs and Pavements." Cool Roofs and Pavements
Toolkit. 2011. http://www.coolrooftoolkit.org/knowledgebase/introduction-to-cool-roofs-
and-pavements/ (accessed December 2012).
Stutz, Bruce. Green Roofs are Starting To Sprout in American Cities. Online Article, Yale
Environement 360, 2010.
"The Encyclopedic Reference to EnergyPlus Input and Output." EnergyPlus Documentation.
October 13, 2011.
The Sprucery Garden Center. http://thesprucery.com/ (accessed January 2013).
US Environmental Protection Agency, EPA.
Velasco, Paulo Cesar Tabares. "Estimating Heat and Mass Transfer Processes in Green Roof
Systems: Current Modeling Capabilities and Limitations." ASRAE Energy Modeling
Conference, April 4. NREL, 2011.

95
Wark, Chris. "Cooler than Cool Roofs: How H e at D o e sn’ t Mov e T hr oug h a Gree n R oof ." greenroofs.com. May 4, 2010. http://www.greenroofs.com/content/energy_editor002.htm
(accessed October 2012).
Zareiyan, Babak. Performance of Roof Materials High SRI, Low SRI, And Green Roof In
California Climate Zone 8 Los Angeles, California. Masters Thesis, Los Angeles:
University of Southern California, 2011.

96
Glossary

BUILDING ENERGY LOADS are how much energy your building needs. These demands can
be pr ov i d ed by el e ct r i c i t y , f uel o r by pas s i v e m ea ns. buil di ng ’ s ene r g y l oads dep end on bo t h i t s
site and program. There are internal loads (activities and energy sources that demand energy
inside the building) and external loads or envelope loads (caused due to the weather).
1

ENERGY USE INTENSITY (EUI) is defined as a unit of measurement that describes a
bui l di ng ’ s ene r g y use . I t r e pr es ent s t h e e n er g y cons um ed by a bui l di ng r el a t i v e t o i t s si z e. A
bui l di ng ’ s EU I i s c a l cu l a t e d by t ak i ng t he t o t a l en er g y cons um ed i n o ne y ea r m ea sur ed i n ( k B t u) and dividing it by the total floor space of the building. Generally a lower EUI signifies a better
energy performance.
2

GREEN ROOF (also called eco- roof, living roof or vegetated roof) is the roof of a building that
is partially or completely covered with vegetation and a growing medium, planted over a water
proofing membrane. It may also include additional layers such as root barrier and drainage and
irrigation systems.
3

1
http://sustainabilityworkshop.autodesk.com/fundamentals/building-energy-loads
2
http://www.energystar.gov
3
Cantor, Steven L., Green Roofs in Sustainable Landscape Design. W. W. Norton & Company Inc., 2008.

97
APPENDIX A:
Nomenclature used in Energy Budget Equations
Source: (D. J. Sailor 2008)

F
f
net heat flux to foliage layer (W/m
2
)
F
g
net heat flux to ground surface (W/m
2
)
H
f
foliage sensible heat flux (W/m
2
)
H
g
ground sensible heat flux (W/m
2
)

total incoming short-wave radiation (W/m
2
)

total incoming long-wave radiation (W/m
2
)
L
f
foliage latent heat flux (W/m
2
)
L
g
ground latent heat flux (W/m
2
)
LAI leaf area index (m
2
/m
2
)
T
f
foliage temperature (Kelvin)
T
g
ground surface temperature (Kelvin)
α
f
albedo (short-wave reflectivity) of the canopy
α
g
albedo (short-wave reflectivity) of ground surface
ε
f
emissivity of canopy
ε
g
emissivity of the ground surface
ε
1
ε
g +
ε
f
- ε
g
ε
f

σ Stefan-Boltzmann constant (5.67 x 10
-8
W/m
2
K
4
)
σ
f
fractional vegetation coverage

98
APPENDIX B:
Schedules
Day of the week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Weekdays 0 0 0 0 0 0 0.1 0.2 0.95 0.95 0.95 0.95 0.5 0.95 0.95 0.95 0.95 0.7 0.4 0.4 0.1 0.1 0.05 0.05
SummerDesign 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.05 0.05
Sat 0 0 0 0 0 0 0.1 0.1 0.5 0.5 0.5 0.5 0.5 0.5 0.1 0.1 0.1 0 0 0 0 0 0 0
Other 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Weekdays, SummerDesign 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0
Sat, WinterDesign 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0
Other 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Weekdays 15.6 15.6 15.6 15.6 15.6 15.6 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 15.6 15.6
SummerDesign 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6
Sat 15.6 15.6 15.6 15.6 15.6 15.6 21 21 21 21 21 21 21 21 21 21 21 21 15.6 15.6 15.6 15.6 15.6 15.6
WinterDesign 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21
Other 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6 15.6
Weekdays, SummerDesign 26.7 26.7 26.7 26.7 26.7 26.7 24 24 24 24 24 24 24 24 24 24 24 24 24 24 24 24 26.7 26.7
Sat 26.7 26.7 26.7 26.7 26.7 26.7 24 24 24 24 24 24 24 24 24 24 24 24 26.7 26.7 26.7 26.7 26.7 26.7
WinterDesign 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7
Other 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7 26.7
Weekdays 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.9 0.9 0.9 0.9 0.8 0.9 0.9 0.9 0.9 0.8 0.6 0.6 0.5 0.5 0.4 0.4
Sat 0.3 0.3 0.3 0.3 0.3 0.3 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.35 0.35 0.35 0.3 0.3 0.3 0.3 0.3 0.3 0.3
SummerDesign 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
WinterDesign 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Other 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
Weekdays 0.05 0.05 0.05 0.05 0.05 0.1 0.1 0.3 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.7 0.5 0.5 0.3 0.3 0.1 0.05
Sat 0.05 0.05 0.05 0.05 0.05 0.05 0.1 0.1 0.5 0.5 0.5 0.5 0.5 0.5 0.15 0.15 0.15 0.05 0.05 0.05 0.05 0.05 0.05 0.05
SummerDesign 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
WinterDesign 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Other 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05
Weekdays, SummerDesign 0.05 0.05 0.05 0.05 0.05 0.08 0.07 0.19 0.35 0.38 0.39 0.47 0.57 0.54 0.34 0.33 0.44 0.26 0.21 0.15 0.17 0.08 0.05 0.05
Sat, WinterDesign 0.05 0.05 0.05 0.05 0.05 0.08 0.07 0.11 0.15 0.21 0.19 0.23 0.2 0.19 0.15 0.13 0.14 0.07 0.07 0.07 0.07 0.09 0.05 0.05
Sun, Hol, Other 0.04 0.04 0.04 0.04 0.04 0.07 0.04 0.04 0.04 0.04 0.04 0.06 0.06 0.09 0.06 0.04 0.04 0.04 0.04 0.04 0.04 0.07 0.04 0.04
Building SHW (DHW) Schedule
Building Occupancy Schedule
HVAC Operation Schedule
Heating Setpoint Schedule
Cooling Setpoint Schedule
Building Equipment Schedule
Building Light Schedule
Table 12: Building operation schedules

99
APPENDIX C:
Comparison of measured and simulated data
for baseline model validation

Table 13: Comparison between Field measured data and simulated data

100

Table 13: Comparison between Field measured data and simulated data (contd.)

101

Table 13: Comparison between Field measured data and simulated data (contd.)

102

Table 13: Comparison between Field measured data and simulated data (contd.)

103

Table 13: Comparison between Field measured data and simulated data (contd.)

104

Table 13: Comparison between Field measured data and simulated data (contd.)

105

Table 13: Comparison between Field measured data and simulated data (contd.)

106
APPENDIX D:
Table 14: Correlation between data sets before and after adjustment
Date and
Time
Measured
data
Simulated
data
Correlation
between
data sets
Adjusted
simulated data
Adjusted
correlation
between data sets
8-Aug Mg Sg C1= Sg/Mg Sg'= Sg*0.9434 C2= Sg'/Mg
0:00 72.93 80.42 1.10 75.87 1.04
1:00 72.82 79.40 1.09 74.91 1.03
2:00 72.46 78.59 1.08 74.14 1.02
3:00 72.12 77.76 1.08 73.36 1.02
4:00 71.89 76.90 1.07 72.55 1.01
5:00 71.60 76.04 1.06 71.74 1.00
6:00 71.35 75.54 1.06 71.26 1.00
7:00 71.22 75.54 1.06 71.26 1.00
8:00 71.96 75.58 1.05 71.30 0.99
9:00 72.90 75.78 1.04 71.49 0.98
10:00 73.67 76.13 1.03 71.82 0.97
11:00 74.26 76.60 1.03 72.26 0.97
12:00 74.82 77.10 1.03 72.74 0.97
13:00 75.47 77.90 1.03 73.49 0.97
14:00 75.97 78.98 1.04 74.51 0.98
15:00 76.39 80.11 1.05 75.58 0.99
16:00 76.82 81.16 1.06 76.57 1.00
17:00 77.07 82.04 1.06 77.40 1.00
18:00 76.91 82.63 1.07 77.95 1.01
19:00 76.32 82.53 1.08 77.86 1.02
20:00 75.74 82.43 1.09 77.76 1.03
21:00 75.33 82.25 1.09 77.59 1.03
22:00 74.98 81.90 1.09 77.26 1.03
23:00 74.59 81.37 1.09 76.76 1.03
9-Aug
0:00 74.23 81.53 1.10 76.92 1.04
1:00 73.98 80.48 1.09 75.92 1.03
2:00 73.71 79.67 1.08 75.16 1.02
3:00 73.31 78.84 1.08 74.38 1.01
4:00 73.00 77.98 1.07 73.57 1.01
5:00 72.68 77.11 1.06 72.75 1.00
6:00 72.45 76.60 1.06 72.26 1.00
7:00 72.32 76.60 1.06 72.26 1.00
8:00 73.17 76.63 1.05 72.29 0.99
9:00 74.41 76.83 1.03 72.48 0.97
10:00 75.34 77.18 1.02 72.81 0.97
11:00 76.28 77.65 1.02 73.25 0.96
12:00 76.93 78.17 1.02 73.75 0.96
13:00 77.45 78.97 1.02 74.50 0.96

107
Date and
Time
Measured
data
Simulated
data
Correlation
between
data sets
Adjusted
simulated data
Adjusted
correlation
between data sets
9-Aug contd. Mg Sg C1= Sg/Mg Sg'= Sg*0.9434 C2= Sg'/Mg
14:00 77.72 80.06 1.03 75.53 0.97
15:00 77.34 81.21 1.05 76.61 0.99
16:00 77.92 82.26 1.06 77.60 1.00
17:00 78.15 83.16 1.06 78.45 1.00
18:00 77.99 83.75 1.07 79.01 1.01
19:00 77.65 83.65 1.08 78.92 1.02
20:00 77.38 83.55 1.08 78.82 1.02
21:00 77.11 83.38 1.08 78.66 1.02
22:00 76.87 83.03 1.08 78.33 1.02
23:00 76.62 82.51 1.08 77.84 1.02
10-Aug
0:00 76.39 82.57 1.08 77.89 1.02
1:00 76.10 81.52 1.07 76.91 1.01
2:00 75.90 80.73 1.06 76.16 1.00
3:00 75.65 79.91 1.06 75.39 1.00
4:00 75.40 79.07 1.05 74.59 0.99
5:00 75.25 78.22 1.04 73.79 0.98
6:00 74.93 77.71 1.04 73.31 0.98
7:00 74.88 77.72 1.04 73.32 0.98
8:00 75.31 77.77 1.03 73.37 0.97
9:00 75.99 77.98 1.03 73.57 0.97
10:00 76.50 78.33 1.02 73.90 0.97
11:00 76.98 78.80 1.02 74.34 0.97
12:00 77.41 79.30 1.02 74.81 0.97
13:00 77.83 80.10 1.03 75.57 0.97
14:00 78.21 81.17 1.04 76.58 0.98
15:00 78.55 82.31 1.05 77.65 0.99
16:00 78.84 83.34 1.06 78.62 1.00
17:00 78.98 84.22 1.07 79.45 1.01
18:00 78.62 84.79 1.08 79.99 1.02
19:00 78.19 84.68 1.08 79.89 1.02
20:00 77.85 84.58 1.09 79.79 1.03
21:00 77.59 84.40 1.09 79.62 1.03
22:00 77.29 84.06 1.09 79.30 1.03
23:00 77.00 83.55 1.09 78.82 1.02
11-Aug
0:00 76.69 83.50 1.09 78.77 1.03
1:00 76.42 82.50 1.08 77.83 1.02
2:00 76.23 81.74 1.07 77.11 1.01
3:00 75.94 80.94 1.07 76.36 1.01
4:00 75.67 80.14 1.06 75.60 1.00
5:00 75.45 79.33 1.05 74.84 0.99

108
Date and
Time
Measured
data
Simulated
data
Correlation
between
data sets
Adjusted
simulated data
Adjusted
correlation
between data sets
11-Aug contd. Mg Sg C1= Sg/Mg Sg'= Sg*0.9434 C2= Sg'/Mg
6:00 75.15 78.84 1.05 74.38 0.99
7:00 74.97 78.88 1.05 74.42 0.99
8:00 75.29 78.96 1.05 74.49 0.99
9:00 75.74 79.18 1.05 74.70 0.99
10:00 76.32 79.53 1.04 75.03 0.98
11:00 76.93 79.99 1.04 75.46 0.98
12:00 77.50 80.48 1.04 75.92 0.98
13:00 78.21 81.25 1.04 76.65 0.98
14:00 78.67 82.28 1.05 77.62 0.99
15:00 79.18 83.37 1.05 78.65 0.99
16:00 79.68 84.37 1.06 79.59 1.00
17:00 79.81 85.20 1.07 80.38 1.01
18:00 79.66 85.73 1.08 80.88 1.02
19:00 79.21 85.59 1.08 80.75 1.02
20:00 78.80 85.48 1.08 80.64 1.02
21:00 78.40 85.29 1.09 80.46 1.03
22:00 78.12 84.94 1.09 80.13 1.03
23:00 77.88 84.43 1.08 79.65 1.02
12-Aug
0:00 77.45 83.38 1.08 78.66 1.02
1:00 77.13 82.36 1.07 77.70 1.01
2:00 76.73 81.57 1.06 76.95 1.00
3:00 76.41 80.75 1.06 76.18 1.00
4:00 76.08 79.91 1.05 75.39 0.99
5:00 75.74 79.07 1.04 74.59 0.98
6:00 75.40 78.53 1.04 74.08 0.98
7:00 75.31 78.55 1.04 74.10 0.98
8:00 75.87 78.60 1.04 74.15 0.98
9:00 76.80 78.82 1.03 74.36 0.97
10:00 77.63 79.19 1.02 74.71 0.96
11:00 78.42 79.66 1.02 75.15 0.96
12:00 78.96 80.17 1.02 75.63 0.96
13:00 79.54 80.97 1.02 76.39 0.96
14:00 80.04 82.04 1.02 77.40 0.97
15:00 80.33 83.17 1.04 78.46 0.98
16:00 80.73 84.20 1.04 79.43 0.98
17:00 80.94 85.05 1.05 80.24 0.99
18:00 80.71 85.60 1.06 80.75 1.00
19:00 80.08 85.48 1.07 80.64 1.01
20:00 79.56 85.38 1.07 80.55 1.01
21:00 78.82 85.20 1.08 80.38 1.02
22:00 78.40 84.84 1.08 80.04 1.02
23:00 77.90 84.33 1.08 79.56 1.02

109
Date and
Time
Measured
data
Simulated
data
Correlation
between
data sets
Adjusted
simulated data
Adjusted
correlation
between data sets
13-Aug. Mg Sg C1= Sg/Mg Sg'= Sg*0.9434 C2= Sg'/Mg
0:00 77.43 82.81 1.07 78.12 1.01
1:00 77.05 81.79 1.06 77.16 1.00
2:00 76.60 81.00 1.06 76.42 1.00
3:00 76.21 80.18 1.05 75.64 0.99
4:00 75.94 79.34 1.04 74.85 0.99
5:00 75.43 78.49 1.04 74.05 0.98
6:00 75.07 77.94 1.04 73.53 0.98
7:00 74.88 77.95 1.04 73.54 0.98
8:00 75.11 78.00 1.04 73.58 0.98
9:00 75.13 78.23 1.04 73.80 0.98
10:00 75.76 78.59 1.04 74.14 0.98
11:00 76.84 79.06 1.03 74.58 0.97
12:00 77.79 79.58 1.02 75.08 0.97
13:00 78.53 80.38 1.02 75.83 0.97
14:00 79.00 81.46 1.03 76.85 0.97
15:00 79.29 82.59 1.04 77.92 0.98
16:00 79.07 83.63 1.06 78.90 1.00
17:00 78.71 84.48 1.07 79.70 1.01
18:00 78.26 85.02 1.09 80.21 1.02
19:00 77.90 84.91 1.09 80.10 1.03
20:00 77.50 84.81 1.09 80.01 1.03
21:00 77.05 84.63 1.10 79.84 1.04
22:00 76.64 84.28 1.10 79.51 1.04
23:00 76.23 83.76 1.10 79.02 1.04
14-Aug
0:00 75.83 82.30 1.09 77.64 1.02
1:00 75.45 81.23 1.08 76.63 1.02
2:00 74.98 80.44 1.07 75.89 1.01
3:00 74.68 79.61 1.07 75.10 1.01
4:00 74.26 78.75 1.06 74.29 1.00
5:00 73.81 77.89 1.06 73.48 1.00
6:00 73.47 77.31 1.05 72.93 0.99
7:00 73.18 77.30 1.06 72.92 1.00
8:00 73.56 77.36 1.05 72.98 0.99
9:00 74.61 77.58 1.04 73.19 0.98
10:00 76.05 77.94 1.02 73.53 0.97
11:00 77.07 78.42 1.02 73.98 0.96
12:00 77.85 78.94 1.01 74.47 0.96
13:00 78.44 79.76 1.02 75.25 0.96
14:00 78.98 80.86 1.02 76.28 0.97
15:00 79.70 82.02 1.03 77.38 0.97
16:00 80.38 83.08 1.03 78.38 0.98
17:00 80.83 83.95 1.04 79.20 0.98
18:00 80.40 84.51 1.05 79.73 0.99
19:00 79.65 84.40 1.06 79.62 1.00

110
20:00 78.98 84.32 1.07 79.55 1.01
21:00 78.37 84.15 1.07 79.39 1.01
22:00 77.79 83.81 1.08 79.07 1.02
23:00 77.38 83.30 1.08 78.58 1.02
AVERAGE 76.90 80.80 1.06 76.23 1.00

111
APPENDIX E:
Simulated data sets for the prototype building

E.1. Simulation results for Phoenix:

Table 15: Comparison between different green roof assembly types and cool roof base case- for Phoenix

112
Table 16: Heat balance through the roof for different assembly types in Phoenix

Table 17: Heat balance through the roof on peak heating and cooling days in Phoenix

113

E.2. Simulation results for Los Angeles:

Table 18: Comparison between different green roof assembly types and cool roof base case- for Los Angeles

114
Table 19: Heat balance through the roof for different assembly types in Los Angeles

Table 20: Heat balance through the roof on peak heating and cooling days in Los Angeles

115

E.3. Simulation results for Chicago:

Table 21: Comparison between different green roof assembly types and cool roof base case- for Chicago

116
Table 22: Heat balance through the roof for different assembly types in Chicago

Table 23: Heat balance through the roof on peak heating and cooling days in Chicago

Abstract (if available)

Abstract In recent years, there has been great interest in the potential of green roofs as an alternative roofing option to reduce the energy consumed by individual buildings as well as mitigate large scale urban environmental problems such as the heat island effect. There is a widespread recognition and a growing literature of measured data that suggest green roofs can reduce building energy consumption. This thesis investigates the potential of green roofs in reducing the building energy loads and focuses on how the different parameters of a green roof assembly affect the thermal performance of a building. ❧ A green roof assembly is modeled in Design Builder- a 3D graphical design modeling and energy use simulation program (interface) that uses the EnergyPlus simulation engine, and the simulated data set thus obtained is compared to field experiment data to validate the roof assembly model on the basis of how accurately it simulates the behavior of a green roof. Then the software is used to evaluate the thermal performance of several green roof assemblies under three different climate types, looking at the whole building energy consumption. For the purpose of this parametric simulation study, a prototypical single story small office building is considered and one parameter of the green roof is altered for each simulation run in order to understand its effect on building’s energy loads. These parameters include different insulation thicknesses, leaf area indices (LAI) and growing medium or soil depth, each of which are tested under the three different climate types. The energy use intensities (EUIs), the peak and annual heating and cooling loads resulting from the use of these green roof assemblies are compared with each other and to a cool roof base case to determine the energy load reductions, if any. The heat flux through the roof is also evaluated and compared. The simulation results are then organized and finally presented as a decision support tool that would facilitate the adoption and appropriate utilization of green roof technologies and make it possible to account for green roof benefits in energy codes and related energy efficiency standards and rating systems such as LEED.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Green-roof design decision support: climate specific green roof design recommendations

PDF

Mitigating thermal bridging in ventilated rainscreen envelope construction: Methods to reduce thermal transfer in net-zero envelope optimization

PDF

Net zero energy building: the integration of design strategies and PVs for zero-energy consumption

PDF

Bridging performance gaps by occupancy and weather data-driven energy prediction modeling using neural networks

PDF

Energy use intensity estimation method based on building façade features by using regression models

PDF

A simplified building energy simulation tool: material and environmental properties effects on HVAC performance

PDF

Microclimate and building energy performance

PDF

Double skin façade in hot arid climates: computer simulations to find optimized energy and thermal performance of double skin façades

PDF

Pre-cast concrete envelopes in hot-humid climates: examining envelopes to reduce cooling load and electrical consumption

PDF

Economizer performance and verification: effect of human behavior on economizer efficacy and thermal comfort in southern California

PDF

Enhancing thermal comfort: air temperature control based on human facial skin temperature

PDF

Developing environmental controls using a data-driven approach for enhancing environmental comfort and energy performance

PDF

Building energy performance estimation approach: facade visual information-driven benchmark performance model

PDF

Energy performance of different building forms: HEED simulations of equivalent massing models in diverse building surface aspect ratios and locations in the US

PDF

Improving thermal comfort in residential spaces in the wet tropical climate zones of India using passive cooling techniques: a study using computational design methods

PDF

Energy savings by using dynamic environmental controls in the cavity of double skin facades

PDF

District energy systems: Studying building types at an urban scale to understand building energy consumption and waste energy generation

PDF

Energy simulation in existing buildings: calibrating the model for retrofit studies

PDF

Night flushing and thermal mass: maximizing natural ventilation for energy conservation through architectural features

PDF

Bridging the gap: a tool to support bim data transparency for interoperability with building energy performance software

Asset Metadata

Creator Mukherjee, Sananda (author)

Core Title A parametric study of the thermal performance of green roofs in different climates through energy modeling

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 07/03/2013

Defense Date 06/19/2013

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag building loads,DesignBuilder,Energy,energy modeling,green roof,OAI-PMH Harvest,parametric study,simulation,sustainability,thermal performance

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor La Roche, Pablo (committee chair), Choi, Joon-Ho (committee member), Konis, Kyle (committee member)

Creator Email sananda23@gmail.com,sanandam@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-284872

Unique identifier UC11293279

Identifier etd-MukherjeeS-1740.pdf (filename),usctheses-c3-284872 (legacy record id)

Legacy Identifier etd-MukherjeeS-1740.pdf

Dmrecord 284872

Document Type Thesis

Format application/pdf (imt)

Rights Mukherjee, Sananda

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA