04 Dec 10sg

National Decentralized Water Resources Capacity Development
Project (NDWRCP) Research Project
QUANTIFYING EVAPORATION AND

TRANSPIRATIONAL WATER LOSSES FROM
GREEN ROOFS AND GREEN ROOF MEDIA
CAPACITY FOR NEUTRALIZING ACID RAIN
by:
Dr. Robert Berghage
Dr. Al Jarrett
Dr. David Beattie
Dr. Kathleen Kelley
Dr. Shazia Husain
Farzaneh Rezai
Bret Long
Ayako Negassi
Robert Cameron
The Pennsylvania State University
Dr. William Hunt

North Carolina State University
2007
Legal Notice
This report was prepared as part of the National Decentralized Water Resources Capacity
Development Project (NDWRCDP) by the organization(s) named below as an account of work
funded by the United States Environmental Protection Agency (U.S. EPA) through Cooperative
Agreement No. X-830851 with the Water Environment Research Foundation (WERF). Neither
WERF nor EPA, members of WERF, the organization(s) named below, nor any person acting on
their behalf: (a) makes any warranty, express or implied, with respect to the use of any
information, apparatus, method, or process disclosed in this report or that such use may not
infringe on privately owned rights; or (b) assumes any liabilities with respect to the use of, or for
damages resulting from the use of, any information, apparatus, method, or process disclosed in
this report.
Copyright© April 2007 by the Pennsylvania State University. All rights reserved.
This Report was prepared by the Center for Green Roof Research at the Pennsylvania State
University.
The research on which this report is based was developed, in part, by the United States
Environmental Protection Agency (EPA) through Cooperative Agreement No. X-830851 with
the Water Environment Research Foundation (WERF). However, the views expressed in this
document are solely those of the author and neither EPA nor WERF endorses any products or
commercial services mentioned in this publication. This report is a publication of The Penn State
Center for Green Roof Research not WERF or EPA. Funds awarded under the Cooperative
Agreement cited above were not used for editorial services, reproduction, printing, or
distribution.
This document was reviewed by a panel of independent experts selected by the author. Mention
of trade names or commercial products does not constitute WERF nor EPA endorsement or
recommendations for use. Similarly, omission of products or trade names indicates nothing
concerning WERF's or EPA's positions regarding product effectiveness or applicability.
ii
ACKNOWLEDGMENTS
Report Preparation
Principal Investigator:
Robert Berghage, Ph.D.
The Department of Horticulture,
Albert Jarrett, Ph.D.
The Department of Agricultural and Biological Engineering,
Project Team:
David Beattie
Dr. Kathleen Kelley
Dr. Shazia Husain
Ayako Negassi
Robert Cameron1
Department of Horticuture,
Farzaneh Rezai
The Department of Agricultural and Biological Engineering,
Bret Long
The Environmental Pollution Control Program at Penn State, Harrisburg
iii
ABSTRACT AND BENEFITS
Abstract:
Green roofs are becoming increasingly common in North America, where they are being
promoted as a stormwater management BMP. Although green roofs have been used in Europe,
particularly Germany for over 30 years, the North American Industry is still relatively new and
installation, management, and performance standards are relatively poorly developed. Research
on the performance of North American green roofs has really only been done for the last decade,
and although much has been learned, there are still many unanswered questions. There is ample
evidence that green roofs can reduce stormwater runoff in Eastern North America by 40 – 60%,
but the relative contribution of the media and plants to this stormwater retention has not been
characterized or quantified. Further, for this retention and evapotranspiration to be of use to
stormwater engineers and developers, tools to predict the retention and detention of stormwater
on a green roof are needed. This project describes studies of the evaporation and
evapostranspiration of water from green roof modules planted with three common green roof
plant species. Green roof plants like sedum and delosperma used water quickly when it was
available and reduced their water use rate when they were drought stressed. This makes sedums
and delosperma ideal plants for green roof use. Plants contribute as much as 40% of the roof
capacity to retain stormwater depending on the frequency and intensity of the storm events. This
data and other runoff data from larger field study roofs provided the basis for models that
describe and predict the function of a green roof described in this report. In addition to
influencing the quantity of runoff, green roofs can also influence the quality of runoff. One of the
most consistently reported benefits of a green roof for runoff quality is the neutralization of acid
precipitation. It is clear however, that this is a finite property of the medium, controlled by the
potential buffering capacity of the medium. To maintain this capacity and hence the water
quality benefit, green roof maintenance should include periodic liming to replace the neutralized
media buffer. This project describes the buffer potential of two commercial green roof media,
and details a testing procedure. The testing procedure allows a green roof manager to estimate
when lime will be needed, and what the potential buffering capacity of a green roof media will
be. With the two media evaluated there were slight differences in total buffer potential, however
the differences were not great and the response to acid addition was similar for both media, with
both having sufficient buffering capacity to neutralize acid precipitation in Central Pennsylvania
for at least 10 years before liming would be required.
Benefits:
♦ The water use from green roof media of three common species of green roof plants was
evaluated. The results of these evaluations dispel two common misconceptions about
green roofs planted with sedums and other succulent species and support their use in
green roof applications. The three plants used water freely immediately following
irrigation and conserved water when media moisture was less available. This data
disproved the common misconception that these plants conserve water all the time
making them poorly suited for removing stored water from a green roof medium. The
data also indicate that the contribution of the plants to the stormwater management
function of a green roof can be considerable, up to 40% of the total function depending
on storm intensity and frequency. This result calls into question the reports that the media
iv
is really the only thing providing the stormwater management function in a green roof.
The results provide new evidence that these plant species are in fact, very well suited for
this use.
♦ The storm-based retention and detention simulation model was developed. Results have
been published and presented at several conferences. This spreadsheet-based model has
been distributed and is being used by a number of engineering and development firms to
plan the use of green roofs in their stormwater plans.
♦ A test method to quantify acid rain buffering capacity for green roof media was
developed and evaluated with two commercial media. The data suggest that common
commercial media can neutralize acid rain for 10-30 years depending on acid deposition
rates. The model developed can be used to determine when a roof should be tested to
determine lime requirements. The results have also been used to suggest the amount of
lime to add to a green roof media to raise the pH to a desired target. Leaching of cation
metals from a green roof suggests that with relatively clean acid rain the roof will not
reduce metal cation content in the leachate runoff compared to runoff from a non-greened
roof. The ability of media to retain cation metals was however fairly large and if
contaminated irrigation water is applied to a green roof the roof can adsorb some of these
ions. The accelerated acid aging test has been presented at the International Green Roof
congress in Basel, and the Green Roofs for Healthy Cities Conference in Boston as well
as several other workshops and seminar programs.
♦ Five short manuscripts detailing the benefits and uses of green roofs are presented. These
are being used as a part of the background for a college class in eco-roof technology.
Keywords: Green roofs, acid rain, green roof media, sedums, plant water use, green roof
stormwater retention, green roof benefits.
v
TABLE OF CONTENTS
Acknowledgments.......................................................................................................................... iii
Abstract and Benefits..................................................................................................................... iv
List of Figures .............................................................................................................................. viii
Executive Summary ...................................................................................................................... xii
1.0 Introduction.................................................................................................................... 1-1
1.1 Literature Cited .................................................................................................... 1-2
2.0 Green Roof Plant Water Use ........................................................................................ 2-1
2.1 Abstract ................................................................................................................ 2-1
2.2 Introduction.......................................................................................................... 2-1
2.3 Materials and Methods......................................................................................... 2-3
2.4 Results .................................................................................................................. 2-4
2.4.1 Species Responses ................................................................................... 2-4
2.4.2 Seasonal Effects ..................................................................................... 2-17
2.5 Discussion .......................................................................................................... 2-17
2.6 Summary and Conclusions ................................................................................ 2-19
2.7 Literature Cited .................................................................................................. 2-21
3.0 Annual and Individual-storm Green Roof Stormwater Response Models............... 3-1
3.1 Abstract ................................................................................................................ 3-1
3.2 Introduction.......................................................................................................... 3-2
3.3 Green Roof Hydrologic Response Models .......................................................... 3-3
3.3.1 Annual Green Roof Response (AGRR) Model ....................................... 3-4
3.3.2 Storm Green Roof Response (SGRR) Model .......................................... 3-5
3.4 Results and Discussion ........................................................................................ 3-5
3.4.1 Annual Green Roof Response (AGRR) Model ....................................... 3-5
3.4.2 Storm Green Roof Response (SGRR) Model .......................................... 3-7
4.0 Green Roof Capacity to Neutralize Acid Runoff ........................................................ 4-1
4.1 Abstract ................................................................................................................ 4-1
4.2 Introduction.......................................................................................................... 4-2
4.3 Materials and Methods......................................................................................... 4-3
4.3.1 Media Titration ........................................................................................ 4-3
4.3.2 Accelerated Media Aging Test ................................................................ 4-4
4.3.3 Simulated Acid Precipitation and Planted Green Roof Media ................ 4-4
4.3.4 Metals in Runoff and Media Cation Exchange........................................ 4-4
4.4 Results and Discussion ........................................................................................ 4-4
4.4.1 Media Titration: ....................................................................................... 4-4
4.4.2 Accelerated Aging Tests. ......................................................................... 4-5
4.4.3 Simulated Acid Precipitation and Planted Green Roof Media .............. 4-10
vi
4.4.4 Metals in Runoff and Media Cation Exchange...................................... 4-15
4.5 Discussion .......................................................................................................... 4-15
4.7 Appendix A........................................................................................................ 4-20
4.7.1 Green Roof Acid Rain Buffer Test ........................................................ 4-20
vii
5.0 Background Educational and Promotional Materials for Green Roofs: A Series of
Articles to Promote Understanding of the Benefits of Using Green Roofs............... 5-1
5.1 An Introduction to Green Roofs: What They Are, Where They Originated,
and Why They Are Still Relevant Today............................................................. 5-1
5.1.1 History...................................................................................................... 5-1
5.1.2 The Current State of Green Roof Technology......................................... 5-3
5.1.3 Literature Cited: ....................................................................................... 5-3
5.2 Green Roofs: How They May Be a Benefit to Businesses and Industry ............. 5-5
5.2.1 Improved Working Conditions ................................................................ 5-5
5.2.2 Increased Retail Sales .............................................................................. 5-6
5.2.3 Increased Occupancy Rates ..................................................................... 5-6
5.2.4 Public Relations ....................................................................................... 5-7
5.2.5 Literature Cited ........................................................................................ 5-8
5.3 Green Roofs: Longer Roof life and Reduced Energy Consumption Yields
Valuable Economic Benefits................................................................................ 5-9
5.3.1 Longer Roof Life ..................................................................................... 5-9
5.3.2 Insulation/Heating and Cooling Benefits............................................... 5-10
5.3.3 Literature Cited ...................................................................................... 5-12
5.4 Green Roofs: Reducing the Impact of Urbanization and What is Being
Done to Encourage Usage.................................................................................. 5-14
5.4.1 Stormwater Mitigation ........................................................................... 5-14
5.4.2 Biodiversity and Other Environmental Benefits.................................... 5-14
5.4.3 Law and Subsidies ................................................................................. 5-15
5.5 Green Roofs: Potential Benefits to Human Health and Well-being .................. 5-18
viii
LIST OF FIGURES
Chapter 2.0
Figure 1 Weighing lysimeters in PSU greenhouse. S. sexangulare planted modules are
shown in this photo. Note suspension cables connecting roof modules to
load cells ........................................................................................................... 2-3
Figure 2a Cumulative water loss (mm) from roof modules planted with S. spurium....... 2-6
2b Daily water loss (mm/day) rates from roof modules planted with S. spurium
Fitted daily water loss functions for unplanted modules [(Daily water loss
(mm/day)) = 1.94 x 0.852Day] and for planted modules [(Daily water loss
(mm/day)) = 3.52 x 0.849Day] ......................................................................... 2-6
2c Difference in water loss rate (mm/day) between planted (S. Spurium) and media
only roof modules. Average temperature during the measurement period was
27oC, with a minimum temperature of 11oC and a maximum of 40oC............. 2-6
Figure 3a Water loss on day 2 after irrigation for green roof modules planted with
S. spurium and unplanted (media only) ............................................................ 2-7
3b Hourly water loss rates on day 2 after irrigation............................................... 2-7
3c Difference in hourly water loss rates between planted and unplanted S. spurium
roof modules on day 2....................................................................................... 2-7
Figure 4a Water loss on day 10 after irrigation for green roof modules planted with S.
spurium and unplanted (media only) ................................................................ 2-8
4b Hourly water loss rates on day 10 after irrigation............................................. 2-8
4c Difference in hourly water loss rates between planted and unplanted
S. spurium roof modules on day 10. ................................................................. 2-8
Figure 5a Cumulative water loss(mm) from roof modules planted with D. nubigenum .. 2-9
5b Daily water loss (mm/day) rates from roof modules planted
with D. nubigenum............................................................................................ 2-9
5c Difference in water loss rate (mm/day) between planted (D. nubigenum) and
unplanted media only modules. Average temperature during the measurement
period was 26.8oC, with a minimum temperature of 13.3oC and a maximum
temperature of 41.1oC ....................................................................................... 2-9
Figure 6a Water loss on day 2 after irrigation for green roof modules planted with D.
nubigenum and unplanted modules................................................................. 2-10
6b Hourly water loss rates on day 2 after irrigation
D. nubigenum roof modules on day 2 after irrigation..................................... 2-10
Figure 7a Water loss on day 10 after irrigation for green roof modules planted with D.
nubigenum and unplanted (media only) modules ........................................... 2-11
ix
7b Hourly water loss rates on day 10 after irrigation........................................... 2-11
D. nubigenum roof modules on day 10 after irrigation................................... 2-11
Figure 8a Cumulative water loss (mm) from roof modules planted with S. sexangulare. 2-12
8b Daily water loss (mm/day) rates from roof modules planted with
S. sexangulare ................................................................................................. 2-12
8c Difference in water loss rate (mm/day) between planted (S. sexangulare) and
unplanted (media only) modules. Average Temperature during the measurement
period was 27oC with a minimum of 21oC and a maximum of 38oC ............. 2-12
sexangulare and unplanted (media only modules) ......................................... 2-13
9b Hourly water loss on day 1 after irrigation ..................................................... 2-13
S. sexangulare roof modules on day 1. ........................................................... 2-13
sexangulare and unplanted (media only)........................................................ 2-14
10b Hourly water loss on day 10 after irrigation....................................................... 2-14
S. sexangulare roof modules on day 10 ........................................................... 2-14
Figure 11 Cumulative water loss (mm) from roof modules planted with D. nubigenum.
Average temperature during the measurement period was 8.6oC, with a minimum
temperature of 4.7oC and a maximum temperature of 16.7oC........................ 2-15
Figure 12a Cumulative water loss (mm) from green roof modultes planted with D.
nubigenum....................................................................................................... 2-16
12b Daily water loss(mm/day) rates from roof modules planted with
D. nubigenum.................................................................................................. 2-16
12c Difference in water loss rate (mm/day) between planted and unplanted (media
only) modules. Average temperature during the measurement period was
18.6oC, with a minimum temperature of 15.1oC and a maximum temperature of
34.8oC.............................................................................................................. 2-16
Figure 13 Sedum sexangulare ......................................................................................... 2-17
Figure 14 Delosperma nubigenum .................................................................................. 2-17
Figure 15 Sedum spurium ............................................................................................... 2-18
Figure 16a Calculated water retention for a 12.7 mm rain (0.5”) occuring following various
dry periods for unplanted (media only) and vegetated (S. spurium) green roofs
and the relative contribution of S. spurium to total retention ......................... 2-20
16b Calculated retention from a 6.35 mm (0.25”) rain.......................................... 2-20
x
Chapter 3.0
Figure 1 Typical Green Roof Profile............................................................................... 3-2
Figure 2 Percent retention as a function of the precipitation return period..................... 3-5
Figure 3 Rainfall and green roof runoff for 1999, a typical year in State College.......... 3-6
Figure 4 Percentaage of annual rain retained as a function of the roof’s storage capacity .. 3-6
Figure 5 Observed and modeled results for the June 2, 2003 rain .................................. 3-7
Figure 6 Predicted and observed green roof runoff depths ............................................. 3-8
Figure 7 Rainfall and runoff rates and cumulative rainfall and runoff depths for a
2-year rain applied to our green roof after 5 days without rain in February..... 3-8
Figure 8 Rainfall and runoff rates and cumulative rainfall and runoff depths for a
100-year rain applied to our green roof after a 5-day dry period in July.......... 3-9
Chapter 4.0
Figure 1 pH of rain in 2005. (NADP, 2007) ................................................................... 4-2
Figure 2 pH of runoff from green roofs and non-green asphalt roofs at the Center
for Green Roof Research in Rock Springs PA.................................................. 4-3
Figure 3 Titration of a clay based medium with sulfuric acid......................................... 4-6
Figure 4 Green roof medium pH recovery after titration with a total of 0.013 meq
sulfuric acid / cc media ..................................................................................... 4-6
Figure 5 pH of a clay-based green roof medium before and after daily additions of
sulfuric acid....................................................................................................... 4-7
Figure 6 pH of a slate-based green roof medium before and after daily additions of
sulfuric acid....................................................................................................... 4-7
Figure 7 Comparison between clay and slate based medium pH before daily acid
additions............................................................................................................ 4-8
Figure 8 Comparison between clay and slate based medium pH after daily acid
additions............................................................................................................ 4-8
Figure 9 pH of a clay based medium before acid additions. Media was allowed to rest
(equilibrate) for 1,4, or 7 days between acid additions..................................... 4-9
Figure 10 pH of a slate based medium before acid additions. Media was allowed to rest
(equilibrate) for 1,4, or 7 days between acid additions..................................... 4-9
xi
Figure 11 pH of acid irrigation water leachate from pots of green roof media but no plants.
Pots were irrigated as needed (every 3-5 days) with water adjusted to
pH 4................................................................................................................. 4-10
Figure 12a/b Plants growing in green roof media irrigated with pH 4 adjusted water. Photos
taken after 5 months of acid irrigation............................................................ 4-11
Figure 13 Leachate from Delosperma in green roof media irrigated with pH 4 acid adjusted
irrigation water................................................................................................ 4-12
Figure 14 Leachate from Talinum in green roof media irrigated with pH 4 acid adjusted
Figure 15 Leachate from Artemisia in green roof media irrigated with pH 4 acid adjusted
Figure 16 Leachate from Agastache in green roof media irrigated with pH 4 acid
adjusted irrigation water ................................................................................. 4-13
Figure 17 Leachate from Potentilla in green roof media irrigated with pH 4 acid adjusted
Figure 18 Leachate from Dianthus in green roof media irrigated with pH 4 acid adjusted
Figure 19 Leachate from Sedum album in green roof media irrigated with pH 4 acid
Figure 20 Leachate from Sedum spurium in green roof media irrigated with pH 4 acid
Figure 21 Leachate from Sedum album in green roof media irrigated with pH 4 acid
Figure 22 Metal ions in runoff from green and non-green roofs at the Center for
Green Roof Research in Rock Springs, PA. Average of five sampled storms
with three roofs sampled per storm................................................................. 4-15
Figure 23 Hydrogen ion deposition in the U.S. in 2005.................................................. 4-16
xii
EXECUTIVE SUMMARY
Rooftop greening has been suggested as a method to reduce impacts of urbanization,
reducing the impervious surface within a developed zone. The stormwater benefits offered by
green roofs include not only direct retention of a portion of the rainfall, but also a delay in the
runoff peak and decrease in the peak rate of runoff from the site as well as potential
improvement in some runoff water quality parameters.
One of the major limitations to promoting this use of green roofs has been the lack of
accepted design tools or models to predict the effects of the green roofs, and a major limitation in
current models has been our lack of a good understanding of evaporation and evapotranspiration
(ET) by the green roof systems. Many of the plants commonly used on an extensive green roof
are drought tolerant succulent plant species. By their nature they conserve water, so it has been
hypothesized that they will, by conserving water, provide relatively little recharge in the medium
water storage capacity between rain events. This hypothesis, and data on medium water storage
and potential evaporation rates has led to the conclusion that an unplanted roof may be nearly as
effective as a planted roof for stormwater management. Data from this project demonstrate that
plants can and most likely do, contribute to the stormwater management function of a green roof
system. These plants use water relatively rapidly when it is available and conserve water when it
is not. When it rains the planted
roof loses 2x the amount of water Figure 1. Daily water loss (mm/day) rates from roof modules planted
lost by an equivalent unplanted with S. spurium.
roof, recharging the media water 3.5
storage potential much faster than Unplanted
could be achieved without the 3
Predicted
plants (Figure 1). If there is no rain 2.5
Planted
Predicted
for an extended period these plants
conserve the remaining moisture in 2
their tissues to survive where other 1.5

species might quickly reach the
permanent wilting point and die. 1
The relative contribution of the 0.5

b
plants to the stormwater function of
a green roof is greatest 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
(approximately 40% of the total Days after irrigation
stromwater retention response in
this study) in areas with relatively
frequent, relatively small rains. This essentially means that these plants are ideally suited for a
green roof in climates like that of the Northeastern U.S. The green roof stormwater models
developed and refined during this project predict the potential of a sedum green roof to retain and
detain 40-60% of the annual stormwater in the Northeast.
One of the major water quality benefits reported for green roofs has been their ability to
neutralize acid rain. Acid precipitation is known to cause a number of problems in urban runoff
including acidification of surface waters and potential acid leaching of metal ions from rooftop
flashings, downspouts and other exposed metals on a roof. To manage a green roof to maintain
xiii
the ability to neutralize acid rain over the long term one must understand both the exchange
capacity of the medium and the acid buffering capacity inherent in the green roof system. This
understanding allows a maintenance manager of a green roof to predict the frequency and effects
of routine procedures (liming) needed to maintain this capacity over the life span of the green
roof. An accelerated acid addition (aging) test was developed. The procedure involves the
addition of small aliquots of acid to a known volume of media. The media acid slurry is allowed
to rest for a period of at least 24hr following each acid addition to let the system come to
equilibrium with the media buffer. This slow titration should be continued to a stable end point
of pH 6.0 or less.With the two media evaluated the buffering potential was similar in the
desirable range for an extensive green roof (pH > 6.5), but was quite different at lower pH. The
test results for the 2 media were described by a simple linear equation with a correction factor for
the medium.
Medium pH = 7.18 - 50.36 x H+ + M
Where H+ is the acid added (meq / cc of media) and M is the media correction factor (0
for clay-based media and 0.3 for a slate-based medium). This equation can be modified and
solved to predict the number of years before a medium will reach a target pH for a given rate of
acid deposition and depth of media on the roof.
TpH − 7.18 + (M × d)
Yr =
50.36 × Hd × 0.01
Where Yr is the number of years to reach the target pH, TpH is the target pH, M is the
media correction factor, d is the depth of the media in cm, Hd is the acid deposition rate in Kg H
per hectare. With a target pH of 6.5, a medium depth of 8.6 cm, and a deposition rate of 0.495
Kg H per ha, a clay-based medium would be expected to reach the target pH in about 24 years.
Since there are other potential sources of acidification in a normal roof including fertilizers,
leaching, and plants, a manager of this roof should probably test pH after about 10-15 years.
Nine species of potential green roof plants were not adversely impacted by irrigation with
acid irrigation water. In fact, the pH of the leachate from most of these plants was higher than
from unplanted pots.
The pH of runoff from non-greened roofs without media was lower than that from green
roofs, however metal ions (Fe, Cu, Mn and Zn) were higher in the runoff from the green roofs.
Although there is ample cation exchange capacity to adsorb the metals if present in
concentrations that exceed the media solution levels in the run on to the roof (eg. contaminated
irrigation water), rain water in this study had lower concentrations of these ions than the medium
solution so there was a net leaching of these ions from the media.
xiv
CHAPTER 1.0
INTRODUCTION
Significant water quality and quantity issues result from stormwater runoff from
developed areas in North America. For the five-year period from 1997 to 2001 the rate of urban
development averaged 890,000 ha/year (2,400 ha/day) (NRCS, 2003). Development results in
water quality impairment and quantity management issues throughout the affected watershed.
For example, nutrient loading (a widespread result of agricultural runoff) may be replaced as the
critical impairment issue for a watershed by increased peak flows, flooding, and urban pollutant
loads as runoff is collected from impervious pavement and roof surfaces.
Rooftop greening has been suggested as a method to reduce these impacts by reducing
the impervious surface within a developed zone (Scholtz, 2001). The stormwater benefits offered
by green roofs include not only direct retention of a portion of the rainfall, but also delaying the
runoff peak and decreasing the peak rate of runoff from the site (PACD, 1998). Most extensive
green roofs currently being installed in North America consist of four distinct layers: an
impermeable roof cover or roofing membrane, a “drainage net,” lightweight growth medium
(about 8cm), and adapted vegetation (PACD, 1998). The drainage layer is an open, highly
permeable material that quickly channels gravitational water off the roof. The growth medium, in
addition to providing a suitable rooting zone for the selected vegetation, should be of low density
and have high water-holding capability while also providing good drainage. A light weight
medium allows for retrofit installation on older buildings, and also reduces the need for extra
structural support in new buildings. Medium depth and porosity plays an important role in
stormwater retention and plant growth. Plants provide shade to the surface below the foliage,
intercept rainfall, and slow the direct runoff from sloped roofs (Miller, 1998).
The use of green roofs in Germany is widespread and has been promoted in many cities
through financial incentives (Pederson, 2001). Economies of scale, contractor experience, and
specialized equipment have reduced the cost of installing a green roof in Germany and
throughout Europe. In contrast, installing a green roof in the U.S. can be very expensive, adding
at least $6 to more than $30-40 per square foot to the cost of the roof. Other barriers also limit
widespread use of green roofs in the U.S. Engineers, developers, and policy makers are unsure of
the actual quantifiable benefits of a green roof. Although much anecdotal information exists
detailing the benefits of green roofs, little scientifically based replicated data has been collected.
Although water retention by the medium and evaporation from it can be fairly easily modeled
and represented mathematically, the addition of plants, particularly drought tolerant crassulacean
acid metabolism (CAM) plants, with their unique ability to close stomates during the day may
greatly complicate predicting water retention.
The Center for Greenroof Research at Penn State, established in 2000, promotes
greenroof research, education, and technology transfer.
(http://hortweb.cas.psu.edu/research/greenroofcenter/index.html). It is the only facility of its type
in the U.S. with replicated small buildings for the study of extensive green roofs. The center has
1-1
collected performance data from its green roof structures over the last three years. Data collected
during 2002 and the very wet summer of 2003 indicates that the green roofs will retain
approximately 40-50% of the annual precipitation (Denardo, 2005; Jarrett et al., 2004). Retention
from individual storm events ranged from 0-100%. The green roofs also delayed runoff and
reduced peak runoff rates. Water quality data collected in 2002 and 2003 show that green roofs
can improve water quality in runoff, particularly in their ability to neutralize acid precipitation.
Our research programs focus on identifying potential benefits and limitations to adoption
of green roofs and providing the data and professional training needed to demonstrate their
effectiveness. Efforts to date suggest that the most likely cost effective driver for a developer or
zoning board to promote the use of green roofs is their ability to retain and detain stormwater,
thus reducing or eliminating the need for increased stormwater management infrastructure on the
part of the municipality or watershed management board, and reducing or eliminating the need to
set aside development land for onsite stormwater management basins. In either case the costs and
benefits are direct, easy to understand, and easy to assign to an individual entity. One of the
major limitations to promoting this use of green roofs is the lack of accepted design tools or
models to predict the effects of the green roofs, and a major limitation in current models is our
lack of a good understanding of evapotranspiration (ET) by the green roof systems. We have
begun to address this need with this research by developing accurate and dynamic estimates of
ET and green roof plant water use.
This project also addresses a major runoff water quality issue that our preliminary
research suggests green roofs could effectively remediate. Our data demonstrates the ability of a
green roof to neutralize acid rain and increase the pH of the runoff. Although the effects are clear
and consistent, the potential benefit has not been fully explored or quantified. Acid precipitation
is known to cause a number of problems in urban runoff including acid leaching of metal ions
from rooftop flashings, downspouts and other exposed metals on the roof. A green roof has the
potential to all but eliminate this pollution source. To manage a green roof for this purpose over
the long term we must understand both the exchange capacity of the medium (ability to adsorb
metal ions) and the acid buffering capacity (ability to neutralize acid rain) inherent in the green
roof system and the frequency and effects of routine maintenance procedures (liming) needed to
maintain this capacity over the life span of the green roof.
1.1 Literature Cited
Denardo, J.C., A.R. Jarrett, H.B. Manbeck, D.J. Beattie, and R.D. Berghage. 2005. Stormwater
Mitigation and Surface Temperature Reduction by Green Roofs. Transactions of the ASAE.
48(4) 1491-1496.
Jarrett, A.R., J.C. DeNardo, H.B. Manbeck, J.M. Hamlett, D.J. Beattie, and R.D. Berghage.
2004. Green Roofs: A Stormwater LID. ASAE Paper No. NABEC 04-0015.
Miller, C. 1998. Vegetated Roof Covers: A new method for controlling runoff in urbanized
areas. In Proc. Pennsylvania Stormwater Management Symposium. Villanova University.
Villanova, PA.
NRCS. 2003. National Resources Inventory: 2001 Annual NRI. USDA Natural Resource
Conservation Service. Available at: www.nrcs.usda.gov/technical/land/nri01/.
1-2
Pedersen, K. 2001. Meadows in the Sky: Contemporary Applications for Eco-Roofs in the
Vancouver Region, MS thesis, Vancouver, BC.: UBC School of Architecture. Pennsylvania
Association of Conservation Districts (PACD). 1998. Pennsylvania Handbook of Best
Management Practices for Developing Areas. Harrisburg, PA.:PACD.
Pennsylvania Association of Conservation Districts (PACD) (1998) Pennsylvania Handbook of
Best Management Practices for Developing Areas. PACD, Harrisburg, PA.
Scholz, K.B. (2001) Green roofs storm water management from the top down. Environmental
Design + Construction. January/February, 2001.
1-3
CHAPTER 2.0
GREEN ROOF PLANT WATER USE
R.D. Berghage, D. Beattie, A.R. Jarrett, F. Rezaei

Department of Horticulture
Department of Agricultural and Biological Engineering
2.1 Abstract
Extensive green roofs are being promoted as a stormwater best management practice in
North America. Research and experience from Northern Europe and Northeastern North
America has demonstrated that these roofs can retain 40-70% of the annual precipitation
depending on medium depth and plant communities. Although the plants are clearly what makes
a green roof "green", the relative contribution of plants to the stormwater function has been
described as minimal, and it has been suggested that much of the function could be achieved
with media alone. To investigate and quantify the role of common green roof plants like sedum
and delosperma, a series of weighing lysimeters were constructed in a greenhouse at Penn State
University in University Park, PA. Delosperma nubigenum, Sedum spurium, and Sedum
sexangulare in green roof modules were subjected to a dry down period during which water loss
from the roof module was recorded. The sedum and delosperma tested were found to use water
rapidly when it was available. The water loss rate from planted roof modules was about 2 times
that from unplanted modules during the first five or so days following irrigation. After five days
the rate of water loss was similar for planted and unplanted roof modules. The relative effect of
plants on the total water loss for the roof modules suggest that the plants could contribute as
much as about 40% of the stormwater retention function of the green roof. The relative affect of
plants would be greatest with relatively frequent (3-5 day) relatively small (12.7 mm, 0.5”)
storms. With longer dry periods the affect of the plants is less, i.e. the medium alone is capable
of the same or nearly the same water retention as a planted system.
2.2 Introduction
Extensive green roofs are rapidly being accepted as a stormwater BMP in North America
(PaDEP, 2006). A simple extensive green roof in North America designed as a stormwater BMP
consists of a drainage layer covered with 2-6” of a lightweight growing medium and vegetation.
Numerous studies have concluded that a green roof with about 4 inches of medium can retain 40-
60% of the annual precipitation in the Northeastern U.S., with nearly 90% of many summer
storms retained (Denardo, et al., 2005). In addition green roofs have been shown to detain runoff,
reducing peak flows. It is no accident that this 40-60% retention is very similar to the amount of
annual precipitation used by evapotranspiration in the Northeast. It appears that green roofs
function in this sense, by restoring the evapotranspirative component of the hydrologic cycle
2-1
(Kramer and Boyer, 1995)). It has been suggested that most of this retention in a green roof is a
function of the medium (VanjWoert et al., 2005). The lightweight media used are designed to
retain as much as 50% water by volume (FLL, 2002), so a 4” roof in theory could retain 2” of
precipitation if all the storage was available at the time of the rain. In practice this is seldom that
case, event frequency, environmental conditions between events and tightly held matric water in
the media reduce this figure to something closer to 0.5-1.25” retained in most summer storms by
a 4 inch thick roof in Central Pennsylvania (Denardo et al., 2005). Media components also
contribute to storage capacity in different ways. In a summary of test results from the Penn State
Agricultural Analytical Testing Laboratory the average water holding capacity for 39 multi-
course green roof media samples (standard extensive roof media test) was 46.1% with a low of
14.7% and high of 65.2% (Berghage, 2007). Although media water storage capacity obviously
affects retention in any given storm, it has surprisingly little effect on total annual retention
(Jarrett et al., 2006). Using a model based on ET and stormwater records Jarrett, et al., 2006
reported very little increase in annual retention as media storage capacity was increased from 40
to 79mm. In fact even with only 3mm of storage more than 30% of the annual precipitation was
predicted to be retained in State College.
Although the majority of the water retention capacity of a green roof is contributed by the
medium, plants also store water. The plants most commonly used on extensive green roofs are
low growing succulents like sedum, delosperma, sempervivum, etc. (Snodgrass and Snodgrass,
2006). These succulent plants can store considerable water in their tissues. A mature population
of S. spurium can weigh 1g/cm of roof surface of which 80-90% can be water. As with the soil
storage only a portion of this is available for atmospheric exchange. Many of these succulent
plants are well adapted to living in drought and have adapted a variety of strategies to reduce
water loss including lignified, waxy tissues and CAM metabolism where stomata can remain
closed during the day to reduce water loss and photosynthetic gas exchange can occur at night
(Larcher, 1995). Sedums can live for weeks or months without rain (Snodgrass and Snodgrass,
2006). This ability to minimize water loss during drought and lose an appreciable percentage of
stored water without plant death makes the concept of permanent wilting point difficult to define
with these species and makes it difficult to place a value on the portion of the plant stored water
that is exchangeable with the atmosphere.
The biggest contribution of plants to green roof water retention is most likely through the
affects of evapotranspiration on media water storage. Plants use soil moisture both for growth
and metabolism, and as a cooling system. Water is extracted from the soil by the root system,
moves through the vascular system and exits through pores in the tissues called stomates
(Kramer and Boyer, 1995). The driving force for this movement of water is the vapor pressure
differential between the water saturated plant tissue and the relatively drier external air. The rate
of water use is thus a function of the open surface area of the stomata and the vapor pressure of
the surrounding air. Plant architecture therefore plays a large role in potential evapotranspiration
and the ability of a plant community to use media water and recharge the media storage
potential. Plants with large exposed surfaces and a high density of stomata have the potential to
use far more water than plants with a high tissue volume to surface area ratio and few stomata.
Low growing species with densely packed foliage present less exposed surface and hence lose
less water. It would seem then, at first glance, that sedums and other succulents are poor choices
for recharging media water storage because they are architecturally and metabolically adapted to
reduce water loss, however a green roof is only green when the plants are alive. It would seem
that the ideal plant for a non-irrigated green roof would have the ability to use water when
2-2
available, but to conserve when water was scarce.
Although we know that sedums and other succulents can tolerate drought, and are very
good at surviving with very little water, and in fact have been outstanding plants on green roofs
in Europe and the temperate Northeastern U.S., little is known about the rate these plants actually
use water on a green roof when it is available. Are they always conservative, or do they use more
water if it is readily available? Do different species contribute more or less to the recharge of the
media storage capacity? What is the actual contribution of the plants to the total function of the
roof and does a sedum roof in fact work significantly better than a medium roof without plants?
Figure 1. Weighing lysimeters in PSU greenhouse.

S. sexangulare planted modules are shown in this
2.3 Materials and Methods photo. Note suspension cables connecting roof
A series of eight weighing lysimeters modules to load cells.
were constructed in a greenhouse at The
Pennsylvania State University in University
Park, PA (Figure 1). Each lysimeter consisted of
a load cell (LCEB-150, Omega Engineering
Company) connected to a data logger (Campbell
Scientific). Green roof modules were constructed
from wood and suspended from the load cells
with metal cable. Modules attached to each load
cell could be changed by releasing the
suspension cables and installing a different
module. Modules were 1.05 x 0.54 x 0.10 m
(LxWxH) with a 10 mm (0.5”) drainage slit at
one end. Each module was filled with a 12 mm
(0.5”) thick drainage layer (Enka drain 9715;
Cold Bond, ENKA – North Carolina) and 89 mm
(3.5”) of a commercial green roof medium
(Gerick Corp., Ohio). The media had a bulk density of 0.534 g/cc and a volumetric water content
at field capacity of 28%. The total water storage potential for the module was thus about 25 mm
of water. A total of 16 modules were constructed. Four modules were planted with Sedum
spurium, 4 modules were planted with S. sexangulare, 4 modules were planted with a mixture of
80% Delosperma nubigenum and 20% S album, and 4 modules were left unplanted with just the
drainage layer and medium. Modules were grown until plants covered 95-100% of the surface of
the module before any measurements were made. Delospoerma dominated the mixed vegetation
modules when they had reached 90-100% coverage, so although some S. album was still present
the responses reported are largely Delosperma, hence these roof modules will be referred to as
Delosperma for the rest of this report. Vegetated roof modules were installed in the weighing
lysimeters one species at a time. Modules were installed with a 1:12 slope (8%). The 4 unplanted
modules were used as controls for each of the planted series. After each planted module change,
load cell module units were recalibrated with standard brass weights between 100 and 2000g. A
light meter (LI-COR quantum sensor Q25338), and 6 copper-constantan (Omega) thermocouples
were also installed.
Modules were fully saturated followed by a dry-down period of 14-21 days. Each species
was subjected to multiple saturation and dry-down cycles at different times of the year (different
2-3
environmental conditions). Saturation to field capacity was achieved by irrigating the roof
modules to runoff, allowing gravitational water to drain for 24-48 hours followed by another
irrigation to runoff. This process was repeated at least 2-3 times before each measurement
period. After the final irrigation, modules were allowed to drain for 2-4 hours (until dripping
stopped) before measurements were collected. During the measurement period module weights
were recorded every 10 minutes. Weight changes were converted to mm of water. Analysis of
variance was used to compare planted and unplanted water loss and least squares regression was
used to fit linear and log functions to dry down curves.
2.4 Results
Evapotranspiration rates varied between species in the modules and with environmental
conditions in the greenhouse, however the general form of the responses were remarkably
consistent across species and climatic conditions. In every evaluation with non-dormant plants
the rate of water loss was rapid for the first 5-6 days, with planted modules losing significantly
more water than unplanted modules. This rapid loss phase was followed by a slower more or less
linear rate of loss where the rate of water loss was not statistically different between planted and
unplanted roof modules.
2.4.1 Species Responses
2.4.1.1 S. spurium
The rate of water loss was about two times greater for planted roof modules than
unplanted for the first 6 days (Figure 2a,b). After 6 days the planted modules had lost on average,
13 mm of water compared to 7 mm for unplanted modules (Figure 2a). After 6 days the rate of
water loss rapidly converged between planted and unplanted roof modules (Figure 2b) with both
loss rate curves approaching zero after about 20 days. The difference in water loss rates between
planted and unplanted roof modules also decreased rapidly with time from about 1.4 mm/day
immediately after irrigation to less than 0.2 mm/day by day 10 (Figure 2c). The hourly water loss
rates from planted and unplanted roof modules immediately after irrigation (day 2) were similar
during the night (0:00 hours to 07:00 and 18:00 to 0:00) (Figure 3a,b). Planted roof modules lost
more water during the morning and early afternoon (07:00 to 14:00), similar amounts during mid
afternoon (15:00), and more during the late afternoon (16:00-18:00) than unplanted modules.
The difference in hourly water loss rates peaked in the early afternoon (13:00hr) (Figure 3c).
Greater water loss from planted compared to unplanted modules resulted in about 47% more
water loss from planted modules (2.6 mm compared to 1.8 mm, respectively). Ten days after
irrigation planted green roof modules were losing only slightly more than unplanted modules
(Figure 4a) and these differences were no longer significant (Figure 4b,c).
2.4.1.2 D nubigenum
8 mm of water compared to 6 mm for unplanted modules (Figure 5a). After 6 days the rate of
water loss rapidly converged between planted and unplanted roof modules (Figure 5b) with both
loss rate curves approaching zero after about 10 days. Two equations were fitted to each line to
describe the function, a quadratic equation was used to describe water loss for the first 5 days
and a log equation was used for the remaining period. The equations for the fitted lines were
2-4
[Water loss (mm/day) = 3.41 - 1.12 x Day + 0.102 x Day2] for the unplanted roof boxes during
the first 5 days; [Water loss (mm/day) = 3.41 - 0.76 x Day + 0.049 x Day2] for planted roof
modules during the first 5 days; [Water loss (mm/day) = 0.636 x 0.94Day] for unplanted
modules after the first 5 days and [Water loss (mm/day) = 0.794 x 0.911Day] for planted roof
modules after day 5. The difference in water loss rates between planted and unplanted roof
modules also decreased rapidly with time from about 1 mm/day immediately after irrigation to
close to 0 mm/day by day 10 (Figure 5c). The hourly water loss rates from planted and unplanted
roof modules immediately after irrigation (day 2) were similar during the night (0:00 hours to
07:00 and 20:00 to 0:00) (Figure 6a,b). Planted roof modules lost more water during the morning
and early afternoon (07:00 to 14:00), similar amounts during mid afternoon (15:00-17:00), and
more during the late afternoon (18:00-20:00) than unplanted modules. The difference in hourly
water loss rates peaked around noon (11:00-13:00hr) (Figure 6c). Greater water loss from
planted compared to unplanted modules resulted in about 80% more water loss from planted
modules (2.0 mm compared to 1.1 mm, respectively). Ten days after irrigation planted and
unplanted green roof modules were losing very little water (Figure 7a) and differences were no
longer significant (Figure 7a,b,c).
2-5
25
a Figure 2.
20
Unplanted
Predicted a. Cumulative water loss
Planted
15 (mm) from roof modules
Predicted
planted with S. spurium.
10
b. Daily water loss (mm/day)

5 rates from roof modules
planted with S. spurium.
0
Fitted daily water loss
0 5 10 15 20 25 functions for unplanted
Days after irrigation modules [(Daily water loss
(mm/day)) = 1.94 x
3.5
0.852Day] and for planted
Unplanted modules [(Daily water loss
3
Predicted (mm/day)) = 3.52 x
Planted
2.5 Predicted 0.849Day];
2
c. Difference in water loss

1.5
rate (mm/day) between
1 planted (S. Spurium) and
media only roof modules.
0.5
b
0 Average temperature during
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
the measurement period
Days after irrigation
was 27C, with a minimum
temperature of 11oC and a
1.6
maximum of 40oC.
1.4
c
1.2
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
2-6
3
unplanted
Figure 3.
2.5 planted
a. Water loss on day

2
2 after irrigation for
green roof modules
1.5 planted with S.
spurium and
1 unplanted (media
only);
0.5
a
0 b. Hourly water loss
1 3 5 7 9 11 13 15 17 19 21 23 25 rates on day 2 after
Hour of the day irrigation;
0.6
b c. Difference in
0.5 hourly water loss
unplanted rates between
0.4 Planted planted and
unplanted S. spurium
0.3 roof modules on day
2.
0.2
0.1
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.1
Hour of day
0.25
c
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
-0.15
Hour of day
2-7
3
Figure 4.
Unplanted a
Planted
2.5
a. Water loss on day 10

2 after irrigation for
greenroof modules
1.5 planted with S. spurium
and unplanted (media
only).
1
0.5 b. Hourly water loss

rates on day 10 after
0 irrigation.
1 3 5 7 9 11 13 15 17 19 21 23 25
Hour of day
c. Difference in hourly
water loss rates
0.6
unplanted
between planted and
b unplanted S. spurium
Planted
0.5
roof modules on
day 10.
0.4
0.3
0.2
0.1
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.1
Hour of day
0.25
c
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
-0.15
Hour of day
2-8
14
a Figure 5.
12
10 a. Cumulative water
loss(mm) from roof modules
8 planted with D. nubigenum;
b. Daily water loss (mm/day)

4
Unplanted rates from roof modules
Predicted planted with D. nubigenum;
2 Planted
Predicted
0
0 5 10 15 20 25
c. Difference in water loss
rate (mm/day) between
planted (D. nubigenum) and
3
unplanted media only
modules.
b
2.5
Average temperature during

2
the measurement period
was 26.8C, with a minimum
1.5 temperature of 13.3oC and a
Unplanted maximum temperature of
1 Predicted 41.1oC.
Planted
Predicted
0.5
0
0 5 10 15 20 25
1.8
c
1.6
1.4
1.2
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
-0.2
-0.4
2-9
2.5
unplanted a Figure 6.
planted
2
a. Water loss on day 2 after

1.5 irrigation for green roof modules
planted with D. nubigenum and
unplanted modules;
1
0.5 b. Hourly water loss rates on day 2

after irrigation;
0
1 3 5 7 9 11 13 15 17 19 21 23 25
c. Difference in hourly water loss
-0.5 rates between planted and
Hour of day unplanted D. nubigenum roof
modules on day 2 after irrigation.
0.3
unplanted b
0.25 Planted
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
-0.15
Hour of day
0.25
c
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
Hour of day
2-10
2.5
a
Figure 7.
unplanted
planted
2
a. Water loss on day 10 after irrigation

1.5 for green roof modules planted with D.
nubigenum and unplanted (media only)
modules;
1
0.5 b. Hourly water loss rates on day 10

after irrigation.
0
1 3 5 7 9 11 13 15 17 19 21 23 25
c. Difference in hourly water loss rates
-0.5 between planted and unplanted D.
Hour of day nubigenum roof modules on day 10
after irrigation.
0.3
b
0.25
unplanted
0.2 Planted
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
-0.15
Hour of day
0.25
c
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.05
-0.1
Hour of day
2-11
12
a Figure 8.
10
a. Cumulative water loss

8
(mm) from roof modules
planted with S. sexangulare;
6
4 b. Daily water loss (mm/day)

Unplanted
rates from roof modules
predicted
2
Planted
planted with S. sexangulare;
Predicted
0
0 5 10 15 20 25 c. Difference in water loss
Days after irrigation rate (mm/day) between
planted (S. sexangulare)
1.6 and unplanted (media only)
unplanted b modules. Average
1.4 predicted Temperature during the
Planted measurement period was
1.2 predicted 27oC with a minimum of
1 21oC and a maximum of
38oC
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
0.45
c
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
1 3 5 7 9 11 13 15 17 19 21 23
2-12
1.4
Figure 9.
a
1.2
1 a. Water loss on day 1 after

irrigation for green roof modules
0.8 planted with S. sexangulare and
unplanted (media only modules);
0.6
0.4
b. Hourly water loss on day 1
unplanted
0.2 after irrigation;
Planted
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.2
c. Difference in hourly water loss
rates between planted and
Hour of day
unplanted S. sexangulare roof
modules on day 1.
0.18
unplanted b
Planted
0.13
0.08
0.03
-0.02 1 3 5 7 9 11 13 15 17 19 21 23 25
-0.07
Hour of day
0.14
c
0.12
0.1
0.08
0.06
0.04
0.02
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.02
-0.04
Hour of day
2-13
1.4
Figure 10.
unplanted a
1.2 planted
1 a. Water loss on day 10 after

irrigation for green roof modules
0.8 planted with S. sexangulare and
unplanted (media only);
0.6
0.4
b. Hourly water loss on day 10
0.2 after irrigation.
0
1 3 5 7 9 11 13 15 17 19 21 23 25 c. Difference in hourly water loss
-0.2 rates between planted and
Hour of day unplanted S. sexangulare roof
modules on day 10.
0.18
b
unplanted
0.13 planted
0.08
0.03
-0.02 1 3 5 7 9 11 13 15 17 19 21 23 25
-0.07
Hour of day
0.14
c
0.12
0.1
0.08
0.06
0.04
0.02
0
1 3 5 7 9 11 13 15 17 19 21 23 25
-0.02
-0.04
Hour of day
2-14
2.4.1.3 S. sexangulare
4.2 mm of water compared to 2.7 mm for unplanted modules (Figure 8a). After 6 days the rate of
water loss converged between planted and unplanted roof modules (Figure 8b) with both loss
rate curves approaching 0.2 mm/day after about 15 days. The difference in water loss rates
between planted and unplanted roof modules also decreased rapidly with time from about 0.4
mm/day immediately after irrigation to 0.1 mm/day by about day 10 (Figure 8c). Two equations
were fitted to each line to describe the function one to describe water loss for the first 5 days and
a second for the remaining period. The equations for the fitted lines were [Water loss (mm/day)
= 1.08 x 0.829Day] for the unplanted roof boxes during the frst 5 days; [Water loss (mm/day) =
1.87 - 0.611 x Day + 0.078 x Day2] for planted roof modules during the first 5 days; [Water loss
(mm/day) = 0.633 x 0.94Day] for unplanted modules after the first 5 days and [Water loss
(mm/day) = 1.01 x 0.932Day] for planted roof modules after day 5.The hourly water loss rates
from planted and unplanted roof modules immediately after irrigation (day 1) were similar
during the night (0:00 hours to 08:00 and 22:00 to 0:00) (Figure 9a,b). Planted roof modules lost
more water during the morning and early afternoon (08:00 to 16:00), and similar amounts during
mid afternoon and evening (17:00-22:00) than unplanted modules. The difference in hourly
water loss rates peaked around noon (12:00-13:00hr) (Figure 9c). Greater water loss from
planted compared to unplanted modules resulted in about 56% more water loss from planted
modules (1.3 mm compared to 0.9 mm, respectively).
Figure 11. Cumulative water loss (mm) from roof modules planted with D. nubigenum. Average temperature
during the measurement period was 8.6oC, with a minimum temperature of 4.7oC and a maximum
temperature of 16.7oC.
20
15
10
5 Unplanted
Planted
0
0 5 10 15 20 25
2-15
18
a Figure 12.
16
14
12
a. Cumulative water loss (mm)
from green roof modultes
10 planted with D. nubigenum;
8
6
Unplanted b. Daily water loss(mm/day)
4 Predicted rates from roof modules
2
Planted planted with D. nubigenum;
Predicted
and
0
0 5 10 15 20 25
c. Difference in water loss rate
(mm/day) between planted and
4
unplanted (media only)
b modules.
3.5
2.5 Average temperature during

the measurement period was
2
Unplanted 18.6C, with a minimum
1.5
Predicted temperature of 15.1oC and a
Planted maximum temperature of
1 Predicted
34.8oC.
0.5
0
0 5 10 15 20 25
1.6
c
1.4
1.2
0.8
0.6
0.4
0.2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
-0.2
2-16
Ten days after irrigation planted and unplanted green roof modules were loosing very little water
(figure 10a) and differences were no longer significant (Figure 10a,b).
2.4.2 Seasonal Effects

In the winter, with cool temperatures and dormant plants, there was little difference in
water loss rates between planted and unplanted roof modules until the surface of the unplanted
media became visibly dry. For example with D. nubigenum roof modules about 8 mm was lost
from both planted and unplanted modules in 10 days following irrigation. The rate of loss was
essentially linear. Over the next 11 days the planted modules lost more water than the unplanted,
media only modules presumably because the plants accessed water from deep in the media which
was less available to evaporation from the dry media surface (Figure 11). With active plants
cooler temperatures had very little effect on the pattern of water loss with rapid losses in the first
5 days following irrigation, about 2 times as much water loss from planted compared to
unplanted and most of the differences occurring in the morning and early afternoon of the first 5-
6 days (Figure 11a-c; Figure 12a,b; Figure 13a,b).
Figure 13. Sedum sexangulare. Figure 14. Delosperma nubigenum
2.5 Discussion
At the onset of this study it was not known if the drought-tolerant succulent plants used for
green roofs would conserve water, maintaining a slow water loss rate to protect against the
effects of drought regardless of water availability. With the three species studied this was clearly
not the case. When water was readily available the plants used it at a relatively rapid rate
compared to evaporation from the media of the control roof modules. As water became more
limiting, the water loss rate of the planted modules was reduced to a rate not significantly
different than the unplanted modules. Water loss during the day when water was readily
available followed the expected pattern for ET (Larcher, 1995) with water use exceeding
evaporation in the morning or early afternoon, reduced relative to evaporation or equal to
evaporation during the heat of the day, and higher again in the evening. This is an interesting and
very promising result for the use of green roofs planted with succulents for stormwater
2-17
management. These species seem by their nature to be very well suited for this use. When it
rains, water is rapidly used and released back to the atmosphere, when it doesn’t rain the plants
conserve remaining water to ensure survival until the next rain. Although sedums are reported to
be CAM plants there was no evidence to support higher water loss in the evening/night under
water limiting conditions observed in this study. This may have been due to the resolution limits
of the weighing systems used. The minimum weight change measurable by the system was
between 100 and 200g (0.17-0.35 mm) depending on wind and other vibrational noise in the
greenhouse. Under water limiting conditions average hourly water loss rates were often well
below this threshold. Although the pattern of water loss was consistent for all the species studied,
the total water loss and water loss rates were different for each of species. S. sexangulare had the
lowest water loss rate, followed by D. nubigenum and finally S. spurium. This result is not
surprising given the differences in plant architecture between these species. S. sexangulare is a
low growing species with very small cylindrical leaves tightly packed in six-sided whorls around
the stem axis (Figure 13), D. nubigenum also is low growing and has cylindrical leaves, however
they are much larger and more exposed (Figure
14), while S. spurium is the most upright of the
Figure 15. Sedum spurium
species and has relatively flat broad leaves (Figure
15). Thus the differences observed in rate of water
loss were likely a function of exposed surface for
evaporation. The results of this study suggest that
the contribution of plants to green roof stormwater
management potential is largely a function of rain
event size and interevent interval. For example if
one were to consider a typical 12.7 mm rain (0.5”)
in calculating the retention using the log equations
for water loss for S. spurium (Figure 1b). The
maximum relative effect of the plants would be
about a 40% reduction in runoff compared to an
unplanted roof (just media) which would occur
about 6 days after the last saturating rain event. At that time the planted roof would retain 100%
of the rain event while the medium without plants would only retain about 60% of the rain
(Figure 16a). Plants would only contribute a 10% increase in retention for the same rain event
occurring 1 day after the last saturating rain (23% retained by the planted roof and 13% retained
by the media) and 10% after about 23 days (100% retained by planted roof and 90% retained by
an unplanted media roof). Compare this to a 6.35 mm rain (0.25”) (Figure 16b). With this rain
the maximum affect of the plants is a 38% greater reduction in runoff from planted roofs
compared to plain media roofs occurring 2 days after the last rain (87% retained by the planted
roof and 49% retained by the medium roof). In this case there will be no affect of plants for rain
events that occur 5 or more days after the last saturating rain (both planted and media roofs will
retain 100%). It seems clear then that the contribution of plants to the green roof as a stormwater
management tool are greatest when numerous small rain events are spaced in time to maximize
the difference between planted roof evapotranspiration and potential evaporation from unplanted
media or an equivalent ballast (3-6 days). This research suggests that the maximum contribution
of sedums and other similar plants under these conditions will be about a 40% increase in rain
retention compared to an unplanted media or ballast on the roof. The other way to look at this is
that 60% or more of the function of a green roof for stormwater management can be obtained
with media and no plants. It follows then that planted green roofs will be most effective in a
2-18
climate like the northeast U.S. or Northern Europe with frequent (but not too frequent) relatively
small rain events. In a drier climate with infrequent rain events the contribution of plants to the
system will be relatively diminished and more of the stormwater function will be attributable to
the media alone. Of course without plants other benefits of the green roof, particularly the
aesthetics will be impacted. The benefits of a green roof for temperature reduction will also be
clearly reduced without plants. Since the temperature reduction is largely a function of
evaporation and the latent heat required, the temperature reductions from a planted roof will be
roughly double those of a wet unplanted medium for the first 5 or so days following a saturating
rain or irrigation event.
2.6 Summary and Conclusions

In summary, green roof plants like sedums contribute to the potential of the green roof to
function as a stormwater management tool. These plants (like other less drought tolerant species)
use water relatively rapidly when it is available. The affect of plants is thus greatest for the first 5
or so days following a rain event when the plants essentially double the rate of recharge of the
media moisture holding capacity allowing more of the next rain event to be stored and less to run
off. The pattern of water use by the plants examined was similar, however the total water loss
rate varied with species and was likely related to plant architecture. Additional plants, and plant
communities should be evaluated to examine the potential to select individual species and
communities to maximize the effectiveness of planted green roofs for stormwater management.
2-19
Figure 16a. Calculated water retention for a 12.7 mm rain (0.5”) occuring following various dry periods
for unplanted (media only) and vegetated (S. spurium) green roofs and the relative contribution of S.
spurium to total retention; 16b. Calculated retention from a 6.35 mm (0.25”) rain.
100.00%
a
80.00%
60.00%
Planted (S. spurium)
Unplanted (media only)

40.00%
Plant contribution (planted-unplanted)
20.00%
0.00%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
Days afer last saturating rain
100.00%
80.00%
60.00%
Planted (S. spurium)

40.00%
Unplanted (media only)
Plant contribution (planted-unplanted)

20.00%
0.00%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33
Days after last saturating rain
2-20
Berghage, R.D. 2007. ASTM media analysis workshop. Greening rooftops for sustainable
communities conference. Minneapolis, MN.
48(4) 1491-1496.
FLL. 2002. Guidelines for the Planning, Execution and Upkeep of Green-Roof Sites.
Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau, Bonn.
Jarrett, A.R., W.F. Hunt, and R.D. Berghage. 2006. Annual and Individual-Storm Green Roof
Stormwater Response Models. ASABE Annual International Meeting, Portland Or. Paper #
062310.
Kramer, P.J. and J.S. Boyer. 1995. Water Relations of Plants and Soils. Academic Press, San
Diego, CA 92101. p98.
Larcher. W.L. 1995. Physiological Plant Ecology. Springer, New York. P63-71.
PA. DEP. 2006. Pennsylvania Stormwater Best Management Practices Manual. Document
number 363-0300-002. The Pennsylvania Department of Environmental Protection, Harrisburg,
PA.
Snodgrass, E.C. and L.L Snodgrass. 2006. Green Roof Plants. Timber Press. Portland Or. 97204.
VanWoert, E.L. D.B Rowe, J.A. Andresen C.L. Rugh, R.T. Fernandez and L Xiao. 2005. Green
Roof Stormwater Retention: Effects of Roof Surface Slope, and Media Depth. Journal of
Environmental Quality 34(3):1036-1044.
2-21
CHAPTER 3.0
ANNUAL AND INDIVIDUAL-STORM GREEN ROOF

STORMWATER RESPONSE MODELS
A. R. Jarrett, Professor, Agricultural and Biological Engineering, Penn State University
W. F. Hunt, Assistant Professor, Biological and Agricultural Engineering, North Carolina State
University
R. D. Berghage, Associate Professor, Department of Horticulture, Penn State University
This paper was presented at the 2006 ASABE Annual International Meeting Sponsored by
ASABE Portland Convention Center Portland, Oregon, 9 - 12 July 2006
3.1 Abstract
The departments of Agricultural and Biological Engineering and Horticulture at the
Pennsylvania State University and Biological and Agricultural Engineering at North Carolina
State University have combined efforts to quantify the stormwater attenuation capabilities of
extensive green-roof systems. This green-roof system consisted of a conventional flat-roof
covering, a 12-mm thick Enka-drainage layer, 89 mm of porous medium, and Sedum spurium
planted 75 mm on center. The combined layers of this green roof had a retention storage capacity
of 40 mm and a saturated hydraulic conductivity of 11 mm/s.
The green roof system was modeled using a checkbook approach with daily rainfall depth
as an input along with daily ET rate and roof runoff as the output. The Annual Model was
applied to 28 years (1976-2003) of rainfall data in State College, PA and Raleigh, NC and
showed that 45-55% of the annual rainfall volume (depth) can be retained on the green roof.
Increasing the volume of storage (roof media depth) does not improve the roofs ability to retain
rain water. Providing only 3 mm of roof storage will still cause 25-40% of the annual rain to be
retained on the roof.
The green roof system was also modeled using the modified Puls routing model with
inputs of the rainfall hyetograph, roof size, and daily ET rate and runoff as the output. The model
was applied to 16 storms measured at the Green Roof Research Center in Rock Springs, PA and
to a variety of synthetic type II 2-, 25-, and 100-year rains under summer and dormant-season
conditions. The Storm Model simulated the experimental results well (runoff volume r2 = 0.906;
peak runoff rate r2 = 0.847). The model also showed that this green roof can attenuate the design
stormwater events to the level of the pre-development runoff rates expected from the building
foot print.
3-1
3.2 Introduction
Green roofs are a surface treatment for rooftops involving the addition of layers of
growth media and plants to create a controlled green space. Widespread use of roof vegetation
has developed since the early 1970s, with Germany leading in the use of green roofs, specifically
in cities (Peck et al., 1999).
A green roof consists of four distinct layers: an impermeable roof covering that serves as
a root barrier, a drainage net or layer, lightweight growth media, and adapted vegetation (PACD,
1998), see Figure 1. The drainage layer is an open, highly drainable material that quickly
channels gravitational water to the roof discharge point(s). The growth medium performs several
functions. In addition to providing a suitable rooting zone for the selected vegetation, the
medium should be of low density and have high water-holding capability. The lighter weight
allows for retrofit installation on older buildings, and also reduces the need for extra structural
support in new buildings. The thickness of the medium and its capillary and gravitational water
holding capacity play an important role in stormwater retention and attenuation of extreme
rainfall events. The plants intercept rainfall, slow its movement into the rooting medium, and are
a measurable portion of the green roof’s water storage capacity (Miller, 1998).
Topics addressed by European green roof researchers include air quality, stormwater
runoff attenuation, plants as building insulation, sound insulation, and building envelope
protection. Current research planned and ongoing in North America includes modeling the
impact of green roofs on the urban heat island, modeling the amount of stormwater retained
annually, and urban agriculture. The majority of these projects are ongoing in Toronto, Canada.
(Overview of Current and Planned Research, 2001). Other ongoing research has focused on the
survival of plant species in varying substrate depths in northern latitudes (Biovin et al., 2001).
Some of this research stems from environmental concerns with air quality and water quality. It is
thought that the vegetation will filter dust particles and greenhouse gasses and serve to clean the
Vegetation
Medium
Drainage Layer
Waterproof Membrane
and Root Barrier
Figure 1. Typical Green Roof Profile.
3-2
air in urban areas.
Green roofs, as stormwater management devices, must be viewed in two different ways;
1) their ability to retain stormwater from day-to-day rainfall events, 2) their ability to attenuate
the runoff expected from extreme rainfall events. From a practical, layman’s perspective
stormwater management is most often viewed as not having excess water to deal with from the
day-to-day rainfall events. These are storms with varying depths, from a trace to the rain
expected once every year that do not tax the capacity of the engineered stormwater system, but
create nuisance flooding. From an engineering and land development perspective, stormwater
BMPs are implemented because they have the ability to attenuate peak runoff rates from storms
having frequencies ranging from 2- to 100-years. In Central Pennsylvania, these design storms
have rainfall depths ranging from 66-150 mm for a 24-hour event (Aron et al., 1986). PACD
(1998) and Jarrett et al. (2004) report that the stormwater benefits offered by green roofs include
increasing the time of concentration, thus delaying the runoff peak, and decreasing the peak rate
of runoff from the site. Also, green roofs intercept and retain stormwater, thus reducing the
volume of water running off a roof, thereby contributing greatly to the NPDES II
recommendation of infiltrating the two-year return period runoff event.
Stormwater research on green roofs has included both model simulations and actual trials
with full-scale and pilot-scale installations. Miller (1998) and Scholz-Barth (2001) reported
annual runoff reductions of 38 to 54% and 38 to 45%, respectively for a 76-mm thick green roof
media. Peak flow rate reductions approximated 50%. Moran et al. (2003) reported that based on
six April to May 2003 rain events in Goldsboro, NC, a 100-mm thick green roof was able to
retain approximately 13-15 mm of rain. They also observed up to 90% reduction in peak flow
from their experimental roofs. Additionally, Michigan State University has initiated a large green
roof research program that includes measuring stormwater retention on the Ford Motor
Company’s 11 acres extensive green roof on their new assembly plant in Dearborn, MI and the
City of Portland is encouraging the placement of green roofs on all new construction within the
city. Their Design Guidelines for Green Roofs specifically states that some jurisdictions may
reduce water and sewer charges or may provide financial incentives to developers who retain
stormwater on site and that green roofs can help reduce the size of stormwater management
ponds, thus recognizing the importance of water retention on green roofs. DeNardo et al. (2005)
reported that green roofs retained 100% of rains smaller than 15 mm and 25% of larger rains in
October and 43% of larger rains in November. Jarrett et al. (2004) reported that green roofs
retained 48, 53, and 78% of larger rains in May, June and July in central PA, respectively. These
benefits, in combinations with limited open space in cities make green roofs a practical method
for easing the pressure on storm sewer systems.
The research reported herein provides the results of a stormwater modeling study
designed to determine the ability of a green roof to attenuation extreme rainfall events and the
results of a second model designed to predict the annual depth of rain that can be retained on a
green roof in central PA and Raleigh, NC.
3.3 Green Roof Hydrologic Response Models

Following the experimental green roof research conducted on six 4.4 m2 buildings at the
Russell E. Larson Research Center of the Pennsylvania State University (DeNardo et al., 2005;
Jarrett et al., 2004) we began to extend these results to include modeling the green roof and its
influence on hydrologic events. To this end two independent models were developed to assess
3-3
the influence of green roofs on the stormwater response to local rainfall events. These included
1) an Annual Green Roof Response (AGRR) Model that predicted annual roof runoff as the sum
of the daily roof responses using daily rainfall depths and daily ET as input, and 2) a Storm
Green Roof Response (SGRR) Model that routes individual storm hyetographs through the green
roof to predict the roof’s runoff rate and volume on a routing interval basis.
The green roofs modeled in this work consisted of the waterproof membrane, a drainage
layer, the growth medium, and green-roof plants. Above the roof membrane was a 0.5-in thick
layer of plastic/geotextile Enka-drain material designed to facilitate drainage of the overlying
green-roof medium, Figure 1. Above the drainage layer was 89 mm of growth medium
consisting of 12.5% sphagnum peat moss, 12.5% coir (coconut fiber), 15% perlite, and 60%
hydrolite with a saturated weight of 1.20 kg/mm-m2. The vegetation used was Sedum spurium.
3.3.1 Annual Green Roof Response (AGRR) Model

The AGRR model was based on three assumptions; 1) that a daily (24-hour) rainfall
record was available to be used as input, 2) that a reliable estimate of daily evapotranspiration
(ET) was available, and 3) that the maximum water retention available within the roof and its
vegetation is known or available. This “checkbook-type” model computed the depth of water
storage available in the green roof and its vegetation on a daily basis. This depth of available
storage, or water deficit, Dgr is defined as the pore-space available in the drainage layer and roof
media below field capacity plus the water holding capacity of the plants. Both the capillary and
hygroscopic water in the drainage layer and roof media were considered to be part of the
retention storage and could be depleted by evaporation and transpiration. In addition, the deficit,
Dgr included the water within the plants used as vegetation on the roof. These plants are ideal for
use on green roofs because, like other arid climate plants their stomata close during the hot, dry
daylight hours to limit transpiration and conserve water. One unique feature of these plants is
that they increase and decrease in size depending on the amount and availability of water. When
water is readily available (it has rained or the soil is well watered) the plants swell to maximum
size and provide excellent cover to the green roof. When water is not readily available (during
drought conditions) the plants actually take a portion of their needed water from within
themselves for plant functions and transpiration, thus from day to day they decrease in physical
size. By the later stages of an extended drought, these plants may only contain 70 to 80% of the
plant mass (and volume) they had when fully watered. When a drought period is followed by a
wetter period, these plants quickly (within a day or so) re-expand to their full size. Therefore the
plants used on our green roofs actually provided up to 10 mm of water retention roof storage.
The daily roof deficit, Dgr can be expressed as
Dgri = Dgri-1 + ETi - Ri (1)
where Dgri-1 is the roof water deficit on Day i-1, ETi is the evapotranspiration on Day i, Ri is
the rain on Day i, and Dgri is the roof water deficit on Day i. The daily deficit is not permitted to
exceed the retention capacity in the roof (Dgri may not be larger than Dmax). Rain on the roof
decreases the daily deficit, but the daily deficit may never be less than zero (0), the condition that
represents the green roof system filled to field capacity. If, on any day, the daily deficit reaches
zero (0), any remaining water is water the green roof cannot retain and thusly becomes runoff.
3-4
This logic was applied to each day during the year in question to estimate how much of
each day’s rain was expected to runoff the green roof. Rainy days following several days without
rain had more storage available, thus less runoff. Rainy days following other rainy days yielded a
large portion of the rain as runoff.
3.3.2 Storm Green Roof Response (SGRR) Model

The SGRR model is based on three assumptions; 1) that a storm hyetograph is available
with uniform times steps between 6 and 60 minutes to be used as input, 2) that a reliable estimate
of daily evapotranspiration (ET) can be provided, and (3) that the month of the storm and the
number of days since the last rain is known. The SGRR model is a Modified Puls Reservoir
Routing Model (Jarrett, 2000) adapted to a green roof.
The rainfall hyetograph input can either be rainfall intensities for a series of uniform time
steps during an actual rain event or rainfall intensities from a synthetic rainfall distribution
similar to those used to estimate pre- and post-development stormwater hydrographs for ungaged
development sites. These rainfall intensities must have a uniform time step between 6 and 60
minutes.
The stage-storage relationship for the green roof was developed from the green roof
drainage layer and roof media characteristics reported by DeNardo et al. (2005). The influence of
water stored in the green roof plants was developed from data reported by Rezaei et al. (2005).
The 12-mm thick drainage layer had a porosity of 78% and field capacity of 5.2%. The 89-mm
growth media had a porosity of 55% and field capacity of 34%. The plants growing in the media
were able to give up and then recover 10 mm of water. The daily ET rate was used to reduce the
water in the green roof starting at field capacity prior to each annual simulation. The ET was
estimated for each month based on the experimental ET results of Rezaei et al. (2005) and the
number of days since the last rain event.
3.4 Results and Discussion

3.4.1 Annual Green Roof Response (AGRR) Model
Twenty eight years (1976-2003) of daily rainfall data in State College, PA and Raleigh,
NC were evaluated using the 100
AGRR model. The State College
input rainfall series had an
average annual rainfall depth of
Green Roof Retention (%)
1024 mm of which 527 mm, or

52.8%, was retained on the green
roof. The Raleigh, NC rainfall
series had an average rainfall State College PA
depth of 1084 mm of which 483 State College PA
mm, or 45.4%, was retained. The Raleigh NC
Log Persson Type III return Raleigh NC
periods were determined for the
10
annual rainfall depths and these 1 10 100
are plotted against the percent of Return Period (yrs)
rain retained on the green roof in Figure 2. Percent retention as a function of the precipitation return
Figure 2. Percent retention, R was period.
3-5
related to return period, T as R = 71.0T-0.0947; r2 = 0.578 in State College and R = 50.4T-
0.117; r2 = 0.678 in
Raleigh.
50
These results can 45
Rain = 967 mm
Runoff = 458 mm
be viewed very positively 40
by considering that 53% of
35
the rain falling on a green
30
Depth (mm)
roofed building in State
College, PA (45% in 25
Raleigh, NC) will be 20
retained on the roof and 15
this portion of rainwater 10
does not require any 5

stormwater attention. The 0
stormwater piping 1/1 3/2 5/1 6/30 8/29 10/28 12/27
infrastructure can be Date

Figure 3. Rainfall and green roof runoff for 1999, a typical year, in State
smaller. Forty-five to College.
sixty-five percent less

water will runoff from
development sites than from
60
similar development sites
Annual Rain Retained (%)
without green roofs. 50
Another, less positive 40

way of looking at the
stormwater impacts of green 30
roofs, is to remember that rain 20
water that falls on and is State College
retained on a green roof has 10 Raleigh
no opportunity to infiltrate 0
into the soil profile and 0 20 40 60 80
becoming part of the local Roof Storage Capacity (mm)
water supply. Figure 4. Percentage of annual rain retained as a
function of the roof's storage capacity.
Figure 3 shows the
daily rain depths and the associated runoff depths for each rainfall event in 1999 in State
College. 1999 was a typical year with average rainfall. It should be noted that only larger rain
events produced runoff from the green roof. Though the runoff results vary greatly, green roof
runoff is usually limited to larger events and rains that occur immediately following rainy days,
when the roof media has not had sufficient time to recover its retention storage capacity.
Before leaving the AGRR Model, it is very useful to note that this model was setup so
that the green roof’s retention storage capacity could be varied. The green roof modeled to
produce the results shown above had a retention storage capacity of 40 mm. We varied the roof’s
retention storage capacity, which was equivalent to making the green roof (primarily the media
depth) thicker (> 89 mm) of less thick (< 89 mm). The roof’s retention capacity was varied from
a low of 3 mm to a high of 76 mm. The percent of the annual rainfall depth retained on the roof
for each retention capacity is shown in Figure 4. There are two rather striking results that come
3-6
from this evaluation. First, when the roof’s retention capacity was increased (simulating a green
roof with thicker media) there was not a great deal of decrease in the runoff expected from the
roof. In other words making the roof thicker did not greatly improve the roof’s ability to retain
rain on the roof. Secondly, when the roof’s retention capacity was decreased, in our case to as
low as 3 mm, there was still an
important reduction in annual 35
runoff caused by this small amount 30
of roof storage. The horticulture
Rain & Runoff (mm)

25
professionals make it clear that
these plants (most plants in fact) 20
need at least 70-90 mm of media to 15
provide adequate rooting and Rain
support. Thus a roof with only 3-6 10
Runoff Observed
mm of retention storage would no 5 Runoff Modeled
longer be a green roof, but this 0
analysis shows that placing one or 0 6 12 18 24
two layers of a heavy-weight Time (Hrs)
geotextile or providing only a few Figure 5. Observed and modeled results for the
June 2, 2003 rain.
millimeters of roof storage would
cause the roof to retain 25-40% of the annual rainfall.
3.4.2 Storm Green Roof Response (SGRR) Model

For engineers responsible for creating stormwater management plans for individual
development sites, the primary focus is not so much on the portion of rainwater captured and
retained annually, but on how a BMP will attenuate the peak runoff rates from large storms that
often cause flooding and considerable damage. To more fully understand how a green roof will
attenuate a specific rain event, the SGRR model was developed as described earlier. This flood
routing based model was applied to two groups of storms; 1) a series of 16 actual rain events that
occurred at the Russell Larson Research Center 10 miles southwest of State College, PA, and 2)
several synthetic 2-, 25-, and 100-year return period rain events developed for use in a local
stormwater management plan.
3.4.2.1 Actual Storm Simulations

Sixteen rain events that occurred between October 2002 and July 2003 were inputted into
the SGRR model. The actual rainfall hyetograph, each with a 1-hr time step, was used as input in
each case. The other input parameters were the month of the event and the number of days since
the last rain. The model assumed up to 1 mm of interception. The interception was decreased as
the time between events increased because the plants were assumed to reduce in size as the
availability of water decreased. The stage-storage relationship used in the routing model assumed
no runoff was possible until the rain had increased the water content in the media and drainage
layer to field capacity.
3-7
The experimentally observed and modeled depth results for the 28.7-mm June 2, 2003
storm are shown in Figure 5. The rain and observed results were collected on an hourly basis.
The modeled results were, likewise, computed on an hourly basis. The runoff collected from this
storm totaled 22.6 mm and the modeled runoff total for this storm was 20.5 mm. The model was
able to correctly show the delay in the start of runoff until the fourth-hour after the start of the
rain event. It also was able to track the runoff in time as it came from the green roof. The
comparison of the predicted versus observed runoff depths for all 16 storms evaluated are shown
in Figure 6. This figure shows that for storms ranging from 3 to 41 mm of rain over the months
of October to November 2002 and 35
May to July 2003, the SGRR model
30
predicted the observed runoff very
well (r2 = 0.906). The model 25
predicted the runoff responses best
Predicted (mm)
20
for rains smaller than 21 mm. There
was greater scatter for larger rains. 15
These results clearly establish the 10

validity of the SGRR model.
5
0
0 5 10 15 20 25 30 35
Observed (mm)
3.4.2.2 Stormwater Design Storms Figure 6. Predicted and observed green roof runoff depths.
Simulations.
The 2-, 25-, and 100-year design storm hyetographs were developed for the State
College, PA area. During most stormwater design procedures, these hyetographs would have
been used with the USDA-SCS Soil Cover Complex method or a similar algorithm to compute
pre- and post-development runoff hydrographs for the site being developed. These hydrographs
would then be used in a flood routing procedure to design a stormwater basin and its spillway
system to properly attenuate the post-development hydrograph peaks to the specified level
related to the pre-development peak rates of runoff for all return periods. Because in the case of
green roofs the rain falls directly on the green roofs, the intermediate runoff hydrograph step was
unnecessary.
2-Year Rain After 5 Dry Feb. Days 2-Year Rain After 5 Dry Feb Days
70 70
60 60
Cumulative Rain & Runoff (mm)
50 50 Rain
Flow Rate (L/s)
40 40 Runoff
30 30
20 20
Runoff
10 10
Rain
0 0
0 6 12 18 24
10 11 12 13 14 15
Time (Hrs) Time, t (hrs)
Figure 7. Rainfall and runoff rates and cumulative rainfall and runoff depths for a 2-year rain applied to our green
roof after 5 days without rain in February.
3-8
100-Year Rain After 5 Dry Days in July 100-Year Rain After 5 Dry Days in July
140
140
Cumulative Rain & Runoff (mm)

120
120
100
100 Rain
Flow Rate (L/s)
80
80 Runoff
60 60
40 40
20 20 Runoff
Rain
0 0
10 11 12 13 14 15 0 6 12 18 24
Time, t (hrs) Time (Hrs)
Figure 8. Rainfall and runoff rates and cumulative rainfall and runoff depths for a 100-year rain applied to our green
roof after a 5-day dry period in July.
In our simulations we applied each rainfall hyetograph directly to the green roof and
routed the rain through the green roof to yield the runoff event in time. Figure 7 shows the
rainfall and runoff rates as well as the cumulative rainfall and runoff depths for our green roof on
a 70- by 270-ft roof subjected to (1) a 2-year simulated rain following 5 dry days in February.
This synthetic storm had a peak rainfall rate of 58 L/s and a peak runoff rate of 27 L/s. The peak
runoff rate for an undeveloped parcel of this size using a time of concentration of 5 minutes and
a CN = 79 (HSG = C) is 17 L/s using TR-55 (USDA-SCS, 1986) and 22 L/s using TR-20
(USDA-SCS, 1983). Thus the green roof will adequately attenuate the 2-year storm even when
the ET from the green roof is limited by February weather conditions.
Figure 8 shows the rainfall and runoff rates as well as the cumulative rainfall and runoff
depths for our green roof on a 70’ x 270’ roof subjected to a 100-year simulated rain following 5
dry days in July. This synthetic storm had a peak rainfall rate of 119 L/s and a peak runoff rate of
50 L/s. The peak runoff rate for an undeveloped parcel of this size using a time of concentration
of 5 minutes and a CN = 79 (HSG = C) is 59 L/s using TR-55 (USDA-SCS, 1986) and 76 L/s
using TR-20 (USDA-SCS, 1983). Thus the green roof will attenuate the 100-year storm under,
what could be considered the best ET weather conditions.
The green roof’s ability to attenuate the peak runoff rates was evaluated for the 2-, 25-,
and 100-years return period storms in State College, PA. These storms were evaluated for four
different climatic conditions including; February ET conditions with 1 and 5 dry days before the
design event and July ET conditions with 1 and 5 dry days before the design event. In all cases
the runoff rates from the green roof were less than the pre-development peak runoff rate for the
same sized parcel of land. Figure 8. Rainfall and runoff rates and cumulative rainfall and runoff
depths for a 100-year rain applied to our green roof after a 5-day dry period in July.
Because of the way a green roof temporarily detains excess water in its gravitational pore
spaces and the rate at which this water is released from the media as runoff, green roofs have the
ability to adequately attenuate high intensity rains and considerably reduce the volume of runoff
from large volume rains.
The results of the SGRR Model have shown that a Modified Puls routing model can be
adapted to simulate the hydrologic response of a green roof. In the large-storm cases examined
3-9
the green roof was able to attenuate the peak roof outflow rates to or below the pre-development
runoff rates expected from these parcels.

The water retention and detention properties of extensive green roof have been
demonstrated to greatly improve stormwater conditions on developing sites. The AGRR Model
showed that an 89-mm thick green roof with 40 mm of retention storage capacity will retain
between 45 and 55% of average annual rainfall depth in State College, PA and Raleigh, NC. This
simple check-book model was also able to show that roofs with more retention capacity will not
greatly improve the roofs ability to retain rainwater. In addition this model also showed that
roofs with smaller retention capacities can have an important effect on retaining annual rainfall
depth; even to the point where 3 mm of retention storage capacity can retain as much as 25 to
40% of the annual rainfall depth. The SGRR Model, based on the Puls Modified routing routine
was able to account for 90% of the variability between measured and simulated results from
individual storms. This model also showed that peaks runoff rates from 2-, 25-, and 100-year
storms will be adequately attenuated to the level of the pre-development runoff peaks; thus
making green roofs a substitute for traditional stormwater basins.

Aron, G., D.J. Wall, E.L.White, C.N, Dunn, and D.M. Kotz. 1986. Field Manual of Pennsylvania
Department of Transportation. Pennsylvania Department of Transportation and Federal Highway
Administration.
Biovin, M., M. Lamy, A. Gosselin, and B. Dansereau. 2001. Effect of Artificial Substrate Depth
on Freezing Injury of Six Herbaceous Perennials Grown in a Green Roof System.
HortTechnology, 11(3):409-412.
DeNardo, J.C., A.R.Jarrett, H.B. Manbeck, D.J. Beattie, and R.D. Berghage. 2005. Stormwater
mitigation and surface temperature reduction by green roofs. Transactions of ASAE 48(4): 1491-
1496.
Jarrett, A.R. 2000. Water Management, 2nd edition. Kendall Hunt Publishing Co, Dubuque, IA.
Jarrett, A.R., J.C. DeNardo, H.B. Manbeck, J.M. Hamlett, D.J. Beattie, and R.R. Berghage.
2004. Green Roofs: A Stormwater LID. ASAE Paper No. NABEC 04-0015.
Miller, C. 1998. Vegetated Roof Covers: A new method for controlling runoff in urbanized
areas. Proceedings of the 1998 Pennsylvania Stormwater Management Symposium, Villanova
University.
Moran, A., W.F. Hunt, G.D. Jennings. 2003. A North Carolina field study to evaluate green roof
runoff quantity, runoff quality, and plant growth. Proceedings of the 2003 EWRI Conference,
Philadelphia, PA.
Overview of Current and Planned Research, Green Roofs for Healthy Cities. 2001.
(www.peck.ca/grhcc/research/overview.htm).
3-10
Peck, S., C. Callaghan, B. Bass, M. Kuhn. 1999. Greenbacks from green roofs: forging a new
industry in Canada. Canada Mortgage and Housing Corporation.
Pennsylvania Association of Conservation Districts (PACD), in association with others. 1998.

Pennsylvania Handbook of Best Management Practices for Developing Areas. Harrisburg, PA.
Rezaei, F., A.R. Jarrett, R.D. Berghage, and D.J. Beettie. 2005. Evapotranspiration Rates from
Extensive Green Roof Plant Species. ASAE Paper No. 05-2150. Tampa, FL July 17-20, 2005.
Scholz-Barth, K. 2001. Green roofs: stormwater management from the top down. Environmental
Design & Construction IV(1):63, January-February, 2001
USDA-Soil Conservation Service. 1983. TR-20 Computer Program for Project Formulation
Hydrology, Technical Release No. 20. USDA-SCS, U.S. Government Printing Office,
Washington, D.C.
USDA-SCS 1986. Urban hydrology for small watersheds. Technical Release No. 55. U.S.
Government Printing Office, Washington, D.C.
3-11
CHAPTER 4.0
GREEN ROOF CAPACITY TO NEUTRALIZE ACID RUNOFF

R.D. Berghage1, D. Beattie1, and Ayako Negassi2
1
Associate Professor, Penn State Department of Horticulture
2
Graduate Student, University of Sheffield
4.1 Abstract
Acid precipitation is common in many parts of the Northeastern and Midwestern United
States. Average pH of rain in much of this region is well below a pH of 5 and in many cases may
be pH 4.5 or less. One of the key runoff water quality benefits offered by a green roof is the
neutralization of acid rain. This benefit is of course limited by the buffering potential of the
green roof media. If this benefit is to be maintained for the 50+ year life span of a green roof it
will be necessary to replenish the media buffer through liming, much as we do with ground level
gardens and agricultural fields. We have developed and demonstrated a relatively simple test
procedure to evaluate the potential of a green roof media to neutralize acid rain. The procedure
involves the addition of small aliquots of acid to a known volume of media. The media acid
slurry is allowed to rest for a period of at least 24hr following each acid addition to let the system
come to equilibrium with the media buffer. This slow titration should be continued to a stable
end point of pH 6.0 or less. The resulting response curve, combined with acid deposition data,
provides the basis for estimating the time a roof will neutralize acid precipitation before liming is
required. A clay based green roof medium and a slate based medium were used to develop the
test. The response of the 2 media to acid additions was very similar in the range of pH desirable
on an extensive green roof. The best fit was obtained with a linear equation [media pH = 7.18 –
50.36 x meq (acid added / cc media) + (media correction factor)] R2 = 0.82. The medium
correction factor for the slate based media was 0.3. Nine green roof plant species were grown in
these media and subjected to acid irrigation for six months. There was little effect of the acid
irrigation on plant growth. The leachate from planted containers had a higher pH than unplanted
controls for most species. Runoff from green roofs had higher concentrations of several common
plant nutrient metal ions (Cu, Fe, Zn, Mn) than runoff from non-green roofs. The medium
solution concentrations of these ions were much higher than what was leached from metal
downspouts and gutters by the acid precipitation. Green roof media can however reduce the
concentration of leachate of metal ions from high concentration water filtered through the
medium. Green roof medium cation exchange is similar to that of other soils and like other soils
they can adsorb these ions removing them from solution.
4-1
4.2 Introduction
Acid precipitation is common in many parts of the Northeastern and Midwestern United
States. Average pH of rain in much of this region is well below a pH of 5 and in many cases may
be pH 4.5 or less (NADP, 2007) (Figure 1). For example in central PA the average annual pH of
precipitation was about 4.4 from 2002 – 2005 (NADP 2007). The average total acidity in
precipitation from 2000 – 2005 in central PA was 0.495Kg/Ha (NADP, 2007). Acid precipitation
can have major impacts on surface waters and streams and can damage forests, and buildings
(U.S. EPA, 2007). Episodic acidification of surface waters caused by runoff can cause short-term
problems even in areas where soil buffering capacity may protect base flows. Episodic
acidification may be even more problematic in urban areas where impervious surfaces increase
runoff even from relatively small storms.
The use of extensive green roofs has been proposed as a means to reduce runoff from flat
roofs in highly developed urban areas. An extensive green roof is characterized by a relatively
shallow, usually less than 6” deep media (manufactured soil) layer topped with drought resistant
plants like sedums and sempervivums. The medium is composed primarily of a highly porous,
lightweight aggregate, most often an expanded clay, slate or shale, and a small amount of organic
compost (Beattie et al., 2005). The medium is designed to provide an appropriate chemical and
physical environment for root growth, and yet be sufficiently lightweight that building structural
Figure 1. pH of rain in 2005. (NADP, 2007)
costs remain reasonable. Most green roof media hold about 35-45% moisture by volume at field
capacity and have a high rate of hydraulic conductivity (FLL, 2002). A 4” extensive green roof
will retain roughly 60% of the annual rainfall it intercepts (Denardo et al., 2005). During the
summer months nearly 90% may be retained while in the winter 20-30% retention is common.
4-2
Precipitation runoff from the roof passes through the medium and flows off most roofs through a
drainage course installed below the planted medium. The drainage course may be a course
aggregate similar to the
medium but without organic Figure 2. pH of runoff from green roofs and non-green asphalt roofs at the
material or fine particle Center for Green Roof Research in Rock Springs PA.
sizes, or a synthetic Runoff pH
geotextile (usually plastic).
The practical result of this is 8.5
Green
that runoff is influenced by 8 Non-Green
the medium and the medium 7.5
has the potential to affect 7
the quality of the runoff 6.5
from a green roof. Among
6
the water quality parameters
green roofs have been 5.5
reported to influence, the 5
buffering of runoff pH is 4.5
one of the most consistent 4
(Figure 2) (Berghage et al.,

3/ /05
3/ 05
/2 5
3/ /05
/2 5
/2 5
/2 5
/0 5
/1 5
/0 5
/1 5
/1 5
8/ /05
8/ /05
9/ /05
9/ /05
10 /05
/2 5
/1 5
4/ /05
6
03 2/0
04 /0
05 4/0
05 3/0
06 3/0
06 7/0
07 0/0
07 6/0
07 6/0
10 8/0
11 6/0
/0
/
14
5
28
7
16
19
26
29
6
25
2007; Van Seters et al.,
1
/
2/
2007). Although the affects Event Date
of a green roof on acid

runoff are clear and
consistent, the potential
benefit has not been fully explored or quantified. Acid precipitation is known to cause a number
of problems in urban runoff including acid leaching of metal ions from rooftop flashings,
downspouts and other exposed metals on the roof. A green roof has the potential to all but
eliminate this pollution source. To manage a green roof for this purpose over the long term we
must understand both the exchange capacity of the medium (ability to adsorb metal ions) and the
acid buffering capacity (ability to neutralize acid rain) inherent in the green roof system and the
frequency and effects of routine maintenance procedures such as liming needed to maintain this
capacity over the life span of the green roof. The objective of this study was to better quantify
the potential for green roofs to neutralize the acidity in runoff from acid rain.
4.3 Materials and Methods

4.3.1 Media Titration
A commercial clay based aggregate green roof media was titrated with sulfuric acid. A
250 cc sample of media was placed in a 1-liter container and saturated with 500 ml of deionized
water. pH was measured directly in the media slurry. Incremental acid additions (0.02M sulfuric
acid) were made to the media slurry. After each addition the pH was measured when the pH
became stable (about 4-7 minutes). The first 15 additions were 1 ml of acid, followed by 10
additions of 5 ml and 3 additions of 10 ml , 1 of 15 ml, and 1 of 20 ml, for a total of 130 ml of
acid added. The final pH was 4.41. The sample was left on the lab bench and after 24, 48, and 72
hours the media slurry was stirred and pH was measured. The test was repeated with a 40 cc
media sample in 400 ml of deionized water. Incremental additions of 0.5, 1, and 2 ml of acid
were used with the smaller sample.
4-3
4.3.2 Accelerated Media Aging Test
Two commercially available green roof aggregates were evaluated using an accelerated
aging test. The aggregates were an expanded clay and an expanded slate. Ten 200 cc samples of
each aggregate were placed in individual 1-liter containers and saturated with 500 ml of
deionized water. pH was measured directly in the media slurry after 1 hour. Acid (2 ml of
0.05M sulfuric acid) was added to each sample, the samples were stirred and allowed to
equilibrate for 1 hour, stirred again and the pH was measured in the media slurry. After 24 hours
samples were stirred again and pH was measured again in each media slurry and then 2 ml of
0.05M sulfuric acid was added to each, stirred, and allowed to equilibrate, and pH was again
measured. This process was repeated daily until the sample pH became relatively stable, 47 days
for the clay aggregate and 75 days for the slate aggregate. This procedure was repeated with acid
additions every 4 days and once per week to further evaluate the affects of recovery time. All
acid additions were converted to meq/cc of media for presentation in this report. Data were
analyzed with least squares regression and Analysis of Variance (EXCEL, Microsoft)
4.3.3 Simulated Acid Precipitation and Planted Green Roof Media

Plants of nine species suitable for use on an extensive green roof were planted in each of
the two media (expanded slate based and expanded clay based) used in the accelerated aging test.
Five plants of each species were planted in each media in 10 cm nursery pots. Plants evaluated
were Artemisia stelleriana, Agastache rugosa, Potentilla argentea, Dianthus deltoides, Sedum
album, Sedum spurium, Talinum parviflorum, Delosperma nubigenum, and Festuca idahoensis.
In addition 5 pots were filled with each media but not planted. Plants and unplanted media were
irrigated using an acidified irrigation water on an as needed basis (every 2-5 days). The acidified
irrigation water was made with deionized water adjusted to pH 4 with sulfuric acid. Irrigation
was applied to runoff (~200 ml). At various intervals after planting leachate was collected from
an irrigation and pH was determined. Data were analyzed with least squares regression and
Analysis of Variance (EXCEL, Microsoft)
4.3.4 Metals in Runoff and Media Cation Exchange

Runoff samples from small buildings, three with green roofs and two with flat asphalt
roofs were collected and analyzed by ICP (inductively coupled plasma) for metals in the runoff.
Runoff from the buildings was collected through standard metal gutters and down spouting.
Runoff from five rain events were sampled and analyzed for Cu, Zn, Mn, and Fe. Green roof
media samples were sent to the Penn State Agricultural Analytical Lab and tested for total CEC
using standard lab procedures (PSU Ag analytical laboratory, 2006).
4.4 Results and Discussion

4.4.1 Media Titration
Green roof media (expanded clay based) pH dropped from 7.5 to 4.1 as the media was
titrated with 0.013 meq of sulfuric acid. There was a little initial buffering for the first 2-3 steps
in the titration (0.0003 meq), followed by a relatively rapid drop to pH 6.3 (0.0013 meq),
followed by another relatively stable pH zone and then a more or less linear decline in pH with
additional acid added (Figure 3). After titration the pH of the media slurry increased back up to
4-4
6.5 over 72 hours (Figure 4). This suggests that a simple rapid titration is not an adequate method
to estimate the buffering potential of a green roof media. The pH recovery indicates some of the
potential buffering capacity of the green roof media depends on a slower process than can be
measured with a simple titration so a slower accelerated ageing test was developed to better
evaluate the total buffering potential.
4.4.2 Accelerated Aging Tests

When acid was added daily with a 24-hour recovery period an interesting pattern in pH
affects was observed with each media. With the clay-based media, the initial pH (after recovery
and before the next acid addition) remained relatively stable for the first 4-5 days (0.01meq acid
added per cc of media) (Figure 5). The drop in pH with the daily acid addition was about 0.4 pH
units and recovery was complete in 24 hours. After 5 days the pH in the media began to drop in a
more or less linear fashion with a pH difference between before and after acid addition of about
1pH unit. This response occurred from acid additions of 0.01 meq to about 0.05 meq. Beyond
acid additions of 0.05 meq, the change in pH was reduced and the difference in pH between
before and after acid addition was smaller (0.4 pH units). The general response of the slate based
media to acid additions was similar (Figure 6). The system was well buffered initially with little
change in pH as acid was added up to about 0.01 meq. This was followed by a more or less linear
decrease in pH with a large response to the acid addition and a large recovery. The difference in
pH from before to after acid addition was greater than in the clay based medium (about 1.5 pH
units) and the recovery in pH from acid addition continued for a longer period in the slate passed
medium (0.01 meq acid /cc media added through 0.14 meq acid added). For the first 0.028 meq
of acid added the pH response of the two media after recovery from each acid addition was about
the same, with the slate based medium having a slightly higher pH throughout. After 0.028 meq
of acid added, the pH of the clay-based media after recovery dropped faster than the pH of the
slate based media (Figure 7). This difference was a difference in the recovery rather than the
response to daily additions (Figure 8). The pH of the media was lower after 24 hr recovery than
the pH after 96 hours or 168 hours (4 days, 1 week) for both media (Figure 9, 10). Although the
difference was significant, it was fairly small (0.25 pH units or less), and the difference between
96 and 168 hours of recovery was smaller and in fact was not significant in either media.
Although the response appears to be quadratic, the best-fit least squares regression line for this
data was a simple linear equation. Polynomial and log functions did not significantly improve the
fit of the prediction line. For the clay based media the equation for the line was media pH = 7.24
– 53.95 x (meq acid added / cc media) with an R2 of 0.84 (Figure 9). For the slate based media
the equation was media pH = 7.42 – 46.78 x (meq acid added / cc media) with an R2 of 0.77
(Figure 10).
4-5
Figure 3. Titration of a clay based medium with sulfuric acid.
7.5
6.5
5.5
4.5
4
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014
meq acid/cc media
Figure 4. Green roof medium pH recovery after titration with a total of 0.013 meq sulfuric acid / cc
media.
6.5
5.5
4.5
4
0 10 20 30 40 50 60 70 80
Hours after titration
4-6
Figure 5. pH of a clay-based green roof medium before and after daily additions of sulfuric acid.
7 clay - before
clay - after
6
0
0 0.02 0.04 0.06 0.08 0.1
meq acid / cc media
Figure 6. pH of a slate-based green roof medium before and after daily additions of sulfuric acid.
7 slate - before
slate - after
6
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
meq acid / cc media
4-7
Figure 7. Comparison between clay and slate based medium pH before daily acid additions.
7
clay - before
slate - before
6
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
meq acid / cc media
Figure 8. Comparison between clay and slate based medium pH after daily acid additions.
7 clay - after
6 slate - after
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
meq acid / cc media
4-8
Figure 9. pH of a clay based medium before acid additions. Media was allowed to rest (equilibrate) for 1,4,
or 7 days between acid additions.
7.5
6.5
5.5 Everyday - before

5 Every 4 days - before
Everyweek - before
4.5 Predicted linear
4
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
meq acid / cc media
Figure 10. pH of a slate based medium before acid additions. Media was allowed to rest (equilibrate) for
1,4, or 7 days between acid additions.
8
7.5
7
6.5
6
Everyday - before
5.5
Every 4 days - before
5
Everyweek - before
4.5 Predicted linear
4
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04
meq acid / cc media
4-9
4.4.3 Simulated Acid Precipitation and Planted Green Roof Media
The pH of leachate collected decreased with time and was significantly influenced by the
species of plant. There was little difference between the pH of leachate from slate or clay based
media with any of the plant species or the unplanted blank sample (Figure 11) so data for slate
and clay were combined for statistical analysis. Plant performance was not influenced by the
simulated acid precipitation over the course of the experiment (Figure 12a,b) with the exception
of Talinum which had died by the end of the experiment. It was however not clear that the acid
irrigation water was the cause of the death of the Talinum which could just as easily have died
from other environmental considerations. By the end of the experiment many of the herbaceous
plants were showing effects of low winter light (the experiment was terminated in February).
The pH of the leachate from planted containers of green roof media followed 2 basic patterns.
For Delosperma and Talinum the leachate pH was statistically different but the differences were
so small that leachate was basically the same as leachate from unplanted containers (Figure 13,
14). With the other species, pH of the leachate from planted containers was similar to pH of
leachate from unplanted containers early in the experiment, but was significantly higher than that
from unplanted containers in later measurements (Figures 15-21). For example, pH of leachate
from containers planted with Agastache was more or less constant through out the measurement
period (Figure 16), resulting in about a 1 pH unit higher leachate pH from the planted than the
unplanted containers. Other species generally showed similar responses although the magnitude
was different. Leachate from S. spurium and Festuca planted containers, for example, was only
about 0.5 pH units higher at the end of the trial than leachate from unplanted containers. In
contrast to this general pattern, leachate from Artemisia planted containers had a lower pH than
Figure 11. pH of acid irrigation water leachate from pots of green roof media but no plants. Pots were irrigated as needed
(every 3-5 days) with water adjusted to pH 4.
7.5
Clay
7
Slate
6.5
5.5
4.5
4
0 50 100 150 200
Time (days)
4-10
leachate from unplanted containers initially but higher pH by the end of the trial.
Figure 12 a and b. Plants growing in green roof media irrigated with pH 4 adjusted
water. Photos taken after 5 months of acid irrigation.
4-11
Figure 13. Leachate from Delosperma in green roof media irrigated with pH 4 acid adjusted irrigation
water.
7
Delosperma
6.5
Blank
6
5.5
5
4.5
4
0 50 100 150 200
Time (days)
Figure 14. Leachate from Talinum in green roof media irrigated with pH 4 acid adjusted irrigation water.
7
6.5
6
5.5
5 Talium
4.5 Blank
4
0 20 40 60 80 100 120
Time (days)
Figure 15. Leachate from Artemisia in green roof media irrigated with pH 4 acid adjusted irrigation water.
7
Blank
6.5 Artemisia stelleriana
6
5.5
4.5
4
0 50 100 150 200
Time (days)
4-12
Figure 16. Leachate from Agastache in green roof media irrigated with pH 4 acid adjusted irrigation
water.
7.5
7
6.5
6
5.5
5
Agastache rugosa
4.5
Blank
4
0 50 100 150 200
Time (days)
Figure 17. Leachate from Potentilla in green roof media irrigated with pH 4 acid adjusted irrigation water.
7.5
7
6.5
6
5.5
5
Potentilla argentea
4.5
Blank
4
0 50 100 150 200
Time (days)
Figure 18. Leachate from Dianthus in green roof media irrigated with pH 4 acid adjusted irrigation water.
7.5
7
6.5
6
5.5
5
Dianthus deltoides 'Erectus'
4.5
Blank
4
0 50 100 150 200
Time (days)
4-13
Figure 19. Leachate from Sedum album in green roof media irrigated with pH 4 acid adjusted irrigation
water.
7.5
7
6.5
6
5.5
5 Sedum album
4.5 Blank
4
0 50 100 150 200
Time (days)
Figure 20. Leachate from Sedum spurium in green roof media irrigated with pH 4 acid adjusted
irrigation water.
7
6.5 S. spurium
Blank
6
5.5
5
4.5
4
0 50 100 150 200
Time (days)
Figure 21. Leachate from Sedum album in green roof media irrigated with pH 4 acid adjusted irrigation
water.
7
6.5 Festuca
Blank
6
5.5
5
4.5
4
0 50 100 150 200
Time (days)
4-14
4.4.4 Metals in Runoff and Media Cation Exchange
The concentration of Cu, Zn, Mn, and Zn, was higher in most samples from green roofs
compared with samples from a standard asphalt roof (Figure 22). Detection limits for the ICP
analysis were Cu < 0.002 mg/L, Fe < 0.005 mg/L, Mn < 0.001 mg/L, and Zn < 0.005 mg/L and
many of the 15 samples analyzed for each ion from non-green roofs and quite a few from green
roofs were below these limits. Average concentrations of Fe, Mn and Zn in runoff from green
and non-green roofs were not different from each other or zero. While there was more Cu in the
runoff from green roofs the average concentration was very small (just over 0.08 mg/L). The
total cation exchange in the green roof media was 17.56 (meq/100g) and 16.22 (meq/100g).
Figure 22. Metal ions in runoff from green and non-green roofs at the Center for Green Roof Research in
Rock Springs, PA. Average of 5 sampled storms with 3 roofs sampled per storm.
0.14
0.12 Green
Non-green
0.1
0.08
0.06
0.04
0.02
0
Cu Fe Mn Zn
-0.02
Metal ions
4.5 Discussion
The ability of a soil, or in this case a green roof media to buffer pH change is a function
of the potential buffers in the medium, the relative solubility of these buffers, and the structure of
the medium, particularly as it relates to the accessibility of the media components to the soil or
medium solution. Since one of the potential runoff water quality benefits of a green roof is to
neutralize acid rain it is important to understand the total potential of these manufactured soils to
buffer pH. This is even more important since many of the sedums used on green roofs grow best
in neutral to alkaline soils (Brickell and Zuk, 1996). While it is possible to use an acid digestion
procedure to estimate the total exchangeable buffer in a medium, this test tells you little of how
the pH in the medium will gradually change with long term small acid additions as occur in a
green roof medium exposed to acid rain. In reality the goal for successful green management is
to understand how much acid can be neutralized before the pH of the medium drops below an
appropriate level for plant growth or water quality considerations, rather than the total buffering
4-15
capacity of the medium. A better approach might be to do a simple titration where small amounts
of acid are added in sequential aliquots and allowed to come to equilibrium in the medium. With
both the green roof media evaluated in this study a simple quick titration proved inadequate.
Although the pH of the medium slurry was “stable to read” within about 5 minutes of an acid
addition (Figure 3), substantial pH recovery (medium buffering) occurred when longer
equilibration times were allowed. Additional buffering was observed as equilibration times were
increased from 24 hours (1 day) to 96 hours (4 days) and 168 hours (1 week) although the
differences were relatively small, particularly as equilibration time increased from 96 to 168
hours (Figures 9,10). These small differences suggest that an accelerated acid aging test for green
roof media can be successfully done using an equilibration time between 1 day and 7 days and
that an equilibration time of 24 hours is probably adequate. There were clear differences in the
nature of the buffering capacity of the two media tested (Figure 5-8), however it is important to
keep in mind that the desirable range of pH for a multi-course intensive green roof media is pH
6.5-8.0 (FLL, 2002). It is interesting to note that nearly all of the differences in media buffering
occurred after the medium pH had dropped below the desirable range. In fact most of the
differences between the media were observed below a pH of 6.0. This is further demonstrated by
the very similar regression functions that describe the response between medium pH and meq of
acid added / cc of media. The 95% confidence intervals for these two regression lines overlap so
it seems reasonable to use a single pH response model rather than individual models for each
media. The difference between the two media was however, statistically significant so a media
correction factor was added to the combined regression equation. The result was a the following
equation [media pH = 7.18 – 50.36 x meq (acid added / cc media) + (media correction factor)].
Figure 23. Hydrogen ion deposition in the US in 2005.
4-16
The medium correction factor for the slate based media was 0.3 which essentially means that the
pH of the slate based media was 0.3 pH units higher than the clay based medium. This combined
function had an R2 of 0.82. The 95% confidence intervals for the model parameters were 7.01-
7.36 for the intercept, -55.88 to – 45.14 for the slope and 0.2-0.4 for the media correction factor.
Using this function a simple equation can be developed relating annual acid deposition to
predicted medium pH. This function can then be easily used to determine when a green roof
would drop below the recommended pH and should be tested to determine the need for lime
additions and can suggest how much lime is needed to restore the lost buffering capacity. The
function for this predictor is: [Years = (((Target pH – (7.18 + medium correction factor)) x
(medium depth in cm)) / (50.36 x ((H deposition kg/ha/yr) x 0.01))]. For example, in Central
Pennsylvania acid deposition from 2000 to 2005 averaged 0.495 kg H / ha / yr (NADP, 2007)
(Figure 23). Using this average rate of deposition a 3.5 inch deep (8.6cm) clay based medium
would be predicted to reach a target pH of 6.5 in 24 years and a slate based medium in 35 years.
The 95% confidence interval on the regression coefficients indicates that for the clay medium
there is a 95% probability that pH 6.5 will be reached in between 13 and 38 years in the clay
medium, and between 19 and 54 years for the slate. The same evaluation for Eastern North
Carolina where the average H deposition from 2000-2005 was .28 kg H / ha / yr results in
reaching the predicted pH of 6.5 in 43 years in a clay based medium and 62 years with slate. The
lower 95% confidence interval for this calculation is 23 years for clay and 34 years for slate. A
standard testing protocol for these media on a green roof might be to test pH based on the
predicted lower confidence interval, i.e. 13 years and 19 years for clay and slate based media in
Central PA, and 23 and 34 years in Eastern North Carolina. This of course presumes that no
other sources of acidification are at work in either location. It is likely that any fertilizer added
will be acidic in reaction (many are), and it is quite possible that some of the plants used may
either directly acidify the media or may acidify the media as they decompose. In this study the
planted media was in fact better buffered, had a higher pH at the end of the acid irrigation trial,
than unplanted controls. With few exceptions, i.e. Delosperma and Talinum, the pH of leachate
from planted pots was as much as 1 pH unit higher than from the unplanted pots. This outcome is
slightly surprising, however it might be explained if you consider that the plants were using some
of the irrigation water, and hence reducing both the quantity of acid and time of exposure to the
medium. This would suggest that the largest plants, and plants with the highest transpiration
rates, should have had the highest pH. The Agastache was the largest plant at the end the
experiment and had the highest pH in the media leachate. Plant roots may have provided access
to media buffer constituents or may have directly modified the media solution pH. Some plants
like Pelargonium, are known to acidify their root zone making nutrients more available. There is
no evidence in this study to suggest that this is the case with any of the potential green roof
plants studied in this evaluation. Since many of the plants used on green roofs, particularly
sedums, are known to prefer neutral to alkaline soils, we were concerned that these plants would
suffer from exposure to acidified irrigation. None of the sedums suffered ill effects of the acid
irrigation over the period evaluated. Although the total acidity applied to the pots was not
measured, it is likely that the total volume of irrigation applied was equivalent to approximately
2-3 years of rain and that the total acidity of the water greatly exceeded a normal rain. It seems
then that there is little reason to be concerned with direct effects of acid irrigation on sedums in a
green roof until the pH drops below problematic levels.
Acid runoff from a roof frequently comes in contact with metal devices and conveyances
like flashings, gutters and downspouts. Acid in the runoff might contribute to leaching of metals
from these materials. Since the green roof neutralizes the acidity in the runoff, (at least for some
4-17
number of years) if the leaching of these materials from the medium was less than the leaching
from the metal gutters and downspouts, there would be a net improvement in runoff water
quality. Our data suggests that leaching from the media exceeds any acid induced leaching from
gutters and downspouts. The average pH of the runoff from the storms measured was 6.9 from
the green roofs and 5.3 from the non-green roofs, but despite the increased acidity in the non-
green runoff, the runoff from the green roofs had higher concentrations of iron, copper, zinc and
manganese. The medium cation exchange for the green roof media was similar to that which
might be expected in a relatively fertile soil rich with organic matter (NRAES, 1995) especially
if you consider the lightweight nature of the green roof medium. It is therefore possible to trap
metals on the exchange of the medium, however since the concentration in the precipitation is
lower than that in the medium solution (in equilibrium with the media exchange) the runoff
contains more of these material than the incoming rain. If the precipitation were to become
significantly more polluted with a metal cation, or if a contaminated irrigation water source were
used for irrigation the medium would, like any other soil remove some of the metal ions through
exchange. For example when we used a simulated acid rain with a heavy copper load, 0.138 mg
applied in 2” of water on a column filled with 4” of roof medium 79% of the copper was retained
by clay based medium and 66% was retained by a slate based medium.

One of the key runoff water quality benefits offered by a green roof is the neutralization
of acid rain. This benefit is of course limited by the buffering potential of the green roof media.
If this benefit is to be maintained for the 50+ year life span of a green roof it will be necessary to
replenish the media buffer through liming, much as we do with ground level gardens and
agricultural fields exposed to acid fertilizers and rain. We have developed and demonstrated a
relatively simple test procedure to evaluate the potential of a green roof media to neutralize acid
rain. The procedure involves the addition of small aliquots of acid to a known volume of media.
The media acid slurry is allowed to rest for a period of at least 24hr following each acid addition
to let the system come to equilibrium with the media buffer. This slow titration should be
continued to a stable end point of pH 6.0 or less. The resulting response curve, combined with
acid deposition data, provides the basis for estimating the time a roof will neutralize acid
precipitation before liming is required. The response curves for the two media evaluated in this
study were remarkably similar even though one was based on an expanded clay aggregate and
the other on an expanded slate. The model response developed for these two media presented in
this report has a correction factor for the difference in pH between the media, with the slate
having a pH of about 0.3 pH units higher than the clay based media. Although the difference was
significant and is included in the model developed in this study, the function was nearly as robust
(R2 = .75 vs R2 = .82) for a simple linear model without a media correction (data not presented).
The model suggests that a green roof in Central PA should have the medium pH evaluated after
about 13 to 19 years for clay based and slate based media respectively, and will likely need to be
limed at that time. Additional media should be evaluated using this procedure to determine if the
general model described here has broader application.
Acid irrigation of green roof plants (simulated acid rain) did not cause plant injury over
the 6-month study period. In fact nearly all the green roof plants evaluated had increased leachate
pH relative to unplanted containers. Although other species might not respond the same way,
these nine species which represent a variety of plant habits and forms, are not likely to be
adversely impacted, or to adversely impact the media if used on green roofs in areas with acid
4-18
precipitation at least in the short term. It remains to be seen, however if long-term impacts would
be greater.
The acid runoff from non-green roofs did not result in increased leaching of metals from
metal gutters and downspouts compared to runoff passing through a green roof which had a
neutral pH. In fact the runoff (leachate) from the green roof had higher concentrations of the
metals tested. In effect the metals on the media exchange in equilibrium with the medium soil
solution were much higher than anything leaching from the gutters and downspouts. This doesn’t
however mean that the media cannot act as an ion exchange filter if challenged with high
concentration rain or irrigation water. The media cation exchange is similar to that of other soils
and if a solution with, for example high copper, is filtered through the media a significant
percentage of the Cu can be retained (60-70% in this study). This suggests that the roof might be
used as a part of a wastewater treatment program to remove metals, but because the
concentration in normal rain is so low, rainwater passing through the media is bound to pick up
additional ions from the media exchange.
4-19
4.7 Appendix A
4.7.1 Green Roof Acid Rain Buffer Test
Background and General Description:

One of the key runoff water quality benefits offered by a green roof is the neutralization
of acid rain. This benefit is of course limited by the buffering potential of the green roof media.
If this benefit is to be maintained for the 50+ year life span of a green roof it will be necessary to
replenish the media buffer through liming, much as we do with ground level gardens and
agricultural fields exposed to acid fertilizers and rain. This relatively simple test procedure can
be used to evaluate the potential of a green roof media to neutralize acid rain. The procedure
involves the addition of small aliquots of acid to a known volume of media. The media acid
slurry is allowed to rest for a period of at least 24hr following each acid addition to let the system
come to equilibrium with the media buffer. This slow titration should be continued to a stable
end point of pH 6.0 or less. The resulting response curve, combined with acid deposition data,
provides the basis for estimating the time a roof will neutralize acid precipitation before liming is
required.
Materials:
Green roof media samples (200 cc) (The test should be run in triplicate to provide an estimate of
the error)
Deionized water
Sulfuric acid solution (0.05M)
Clean 1L beakers
Burette or pipette capable of 2 ml measurements
pH meter and electrode
Spatula or spoon
Procedure:
1. Measure 3 (or more) 200 cc media samples. Place each sample in a 1L beaker.
2. Add 500ml of DI water to each beaker with the media sample, stir to mix completely.
3. Measure pH on the slurry – record
4. Add 2 ml of 0.05M Sulfuric acid. Stir to mix completely. Measure pH and record.
5. Allow the media water acid slurry to rest at least 24 hours.
6. Repeat steps 3-5 until 3 or more consecutive samples are below the target pH (suggested
value of 6)
Calculations:
Convert daily acid additions to meq acid / cc media (for the procedure described above the
conversion is 0.002 meq acid added / cc of media each day).
Plot meq acid / cc vs. media pH
Determine the linear regression for the response
Use this regression to determine the number of years of acid deposition required to reach the
target pH (use NADP data for H deposition; the conversion from Kg H/ha to meq/cc is (H
Kg/ha/yr x 0.01)/(medium depth in cm)).
4-20
Beattie, D.J., R.D. Berghage, and E Snodgrass. 2005. Plants and substrates are the heart of the
green roof. In: Combined proceedings of the International Plant Propagator’s Society 55:82-86.
Berghage, R.D., D.J. Beattie, A.R. Jarrett, and T. O’Connor. 2007. Green roof runoff water
quality. Greening Rooftops for Sustainable Communities, Minneapolis, Mn.
Brickell, C., J.D. Zuk. 1996. A-Z Encyclopedia of Garden Plants. The American Horticultural
Society. DK publishing.
48(4) 1491-1496.
FLL. 2002. Guidelines for the Planning, Execution and Upkeep of Green-Roof Sites.
Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau, Bonn.
National Atmospheric Deposition Program (NRSP-3). 2007. NADP Program Office, Illinois
State Water Survey, 2204 Griffith Dr., Champaign, IL 61820.
NRAES. 1995. Recommended Soil Testing Procedures for the Northeastern United States.
NRAES. No. 493.
Van Seters, T, L. Rocha, and G. MacMillan, 2007. Evaluation of the Runoff Quantity and
Quality Performance of an Extensive Green Roof in Toronto, Ontario. Greening Rooftops for
Sustainable Communities, Minneapolis, Mn.
U.S. EPA. 2007. Acid Rain website. http://www.epa.gov/acidrain/. Accessed June 2007.
4-21
CHAPTER 5.0
BACKGROUND EDUCATIONAL AND PROMOTIONAL

MATERIALS FOR GREEN ROOFS: A SERIES OF
ARTICLES TO PROMOTE UNDERSTANDING OF THE
BENEFITS OF USING GREEN ROOFS
Jennifer Vidmar, Kathy Kelley, Robert Berghage
Department of Horticulture The Pennsylvania State University
5.1 An Introduction to Green Roofs: What They Are, Where They Originated, and
Why They are Still Relevant Today
Green roof systems are extensions of existing roofing systems that allow plants to be
grown on a rooftop. Generally, the components of a green roof include a waterproofing layer, a
root repellant layer, a layer of drainage material, and a filter membrane topped with the growing
medium and plants (Green Roofs for Healthy Cities, 2005). This basic formula can be modulated
based on the purpose and type of green roof being installed, either an extensive or intensive
green roofing system (Green Roofs for Healthy Cities, 2005). Extensive green roofing systems
are not accessible or built to withstand human traffic, as they typically consist of low-growing
plants, such as Sedum album or Sedum sexangulare as well as other species and cultivars, in a
shallow, lightweight media less than six-inches deep (R.D. Berghage, personal communication,
(Wong et. al., 2003). Intensive green roofs, on the other hand, can be planted with deeply rooted
plants such as trees and shrubs, and are usually designed to withstand human traffic (Wong et.
al., 2003). Typically, green roofs built exclusively for environmental benefits are extensive, as
these require lower initial investment than intensive green roofs. Extensive green roofs also
require less structural loading capacity, and frequently do not require additional support to be
added when converting an existing roof to a green roof (Wong et. al., 2003). An intensive green
roof provides additional benefits since it can be tailored to provide an area for the community to
socialize, garden, grow food, and enjoy water features such as fountains and ponds. Intensive
roofing systems tend to provide more social benefits than extensive roofing systems because they
are designed to withstand foot traffic and can provide areas for social interaction (Huang, 2005;
Wang et. al., 1999; Yuen et. al., 2005.)
5.1.1 History
Green roof technology is not a new concept. Throughout history, many cultures have
utilized green roofs in some form. The first historical reference to gardens built above ground
level are the ziggurats of ancient Mesopotamia, which were built beginning in the fourth
millennium up until approximately 600 B.C. (Wark et. al., 2003). They were built in the
courtyards of temples located in major cities, and consisted of stepped pyramids made from
5-1
stone, spiraling upwards towards the zenith. Archaeological evidence suggests that plantings of
trees and shrubs on landings of these stepped terraces provided resting areas for those climbing
the pyramid (Osmundson, 1999). Evidence of plants growing on rooftops can also be traced back
to the Hanging Gardens of Babylon (approximately 500 B.C. to 90 B.C.). In ancient Babylon,
terraced structures were built atop beams on which soil and plant material were placed over
waterproofed layers consisting of tar and reeds (The Garland Company, 2005). The Hanging
Gardens of Babylon were constructed by Nebuchadrezzar II to mimic the mountain scenery of
the native land of his homesick wife, Amytis. A Greek historian, Diodorus Siculus, provided
detailed descriptions of the gardens, making it clear that they were built in part on terraces over
vaults, the highest of which was approximately 70’ and carried the entire weight of the garden
(Osmundson, 1999).
In the 13th century, some Benedictine abbeys in France utilized rooftop gardens (Wark et.
al., 2003). Rooftop gardens were also built in at least three Italian cities from the 1300s to 1500s,
namely Pienza, Lucca, and Careggi. Other European countries occasionally used green roofs, and
from 1600 to 1875 greened roof space could be found in several places in Germany and Russia.
Even regions in South America had ancient rooftop gardens. The major Aztec city of
Tenochtitlan had many roof gardens, although none survived the Spanish invasion. Cortes,
himself, wrote a letter to King Charles I of Spain wherein he described the rooftop gardens on
native dwellings. In his letter, Cortes stated, “…there are many rich citizens who also possess
very fine houses. All these houses, in addition to having very fine and large dwelling rooms,
have very exquisite flower gardens both on the upper apartments as well as down below”
(Osmundson, 1999).
Centuries ago, Norwegians designed sod roofs to help insulate their buildings against
temperature extremes. Use of sod roofs as insulation also occurred in North America, as settlers
of the Great Plains utilized sod roofs during the mid to late 1800s. Settlers constructed dwellings
that were built into the side of a hill or had sod roofs. The soil covered rooftops provided
insulation from temperature extremes (Osmundson, 1999). These roofs were planted with grasses
to prevent erosion, but leakage was a problem when it rained so most buildings were abandoned
with the advent of modern heating technology.
More recently, in the early 1900s, rooftops were used for summer entertainment in major
cities of the United States. The term roof gardens was coined around 1893 and referred
specifically to the theater roof plantings of the period, in fact a number of theaters, such as
Madison Square Gardens in New York City, derived their names from their rooftop gardens
(Osmundson, 1999). The Rockefeller Center in New York City also installed rooftop gardens
during the early 1900s. A total of five rooftop gardens were installed between 1933 and 1936,
designed to provide pleasant views to high-rise tenants at premium prices (Wark et. al., 2003).
Frank Lloyd Wright and Le Corbusier, two influential architects, designed buildings that
incorporated rooftops as functional and garden space. In many of his designs, Wright used roof
areas as an extension of indoor living space, and although not true gardens, these rooftop living
areas incorporated plantings. The Midway Gardens building, designed by Frank Lloyd Wright,
and built in Chicago in 1914, utilized rooftop plantings. Unfortunately, Midway Gardens met a
fate similar to other Frank Lloyd Wright buildings, such as the Larkin Building in Buffalo, NY
and the Imperial Hotel in Tokyo, Japan, all of which were destroyed in the mid to late 1900s. Le
Corbusier, a Swiss architect, also believed in using rooftops as an additional living space, and
therefore many of his designs incorporated plants. Some of his buildings, such as the Villa
Savoye in Poussy, France, the Domino houses, the Pessac workers’ housing estate, the Unite
5-2
d’Habitation apartments in France, and his government building in Punjab, India, incorporated
plant material as an incidental part of his rooftop design (Osmundson, 1999).
5.1.2 The Current State of Green Roof Technology

Currently, green roofs are starting to come back into favor and are being installed on
roofs of new buildings. While green roof technologies are well established in Europe, the
benefits and construction are poorly understood in North America (Penn State, 2004). Green roof
technology is successful in Europe due in part to supportive legislation and financial benefits for
building owners who choose to install a green roof (Green Roofs for Healthy Cities, 2005). In
Germany, it is estimated that 10% of roofs are greened, and the Penn State Center for Green
Roof Research stated, “between 1989 and 1999, German roofing companies installed nearly 350
million square feet of green roofs” (Penn State, 2004). Green roof technology is also well
established in France, Austria, Norway, and Switzerland (Wong et. al., 2003).
The city of Chicago has been innovative in its early acceptance of green roof technology
in the U.S. (Walsh, 2004/2005). Chicago’s City Hall has a green roof, and the city’s government
has adopted policies encouraging, and in some cases requiring, green roofs for new
developments (Walsh, 2004/2005). Chicago currently has more than 80 green roofs totaling more
than 1 million square feet. In 2005, Chicago started offering $5,000 grants to residential or small
commercial property owners to green their rooftops, encouraging a wider population to use green
roof technology (City of Chicago, 2005). Currently green roofs in general, although more
specifically extensive green roofs, are experiencing a nationwide surge in popularity, and that
trend can be expected to continue. Recent press coverage has hailed the green roof as an
environmental benefit, a trendy design element, and an economic building decision.
Pennsylvania State University, University Park Campus, has plans to install a green roof, and has
committed to building another green roof on a new student medical center (R.D. Berghage,
personal communication). Other educational institutions have also embraced green roofs.
Swarthmore College’s Alice Paul Dormitory, Swarthmore, PA; Carnegie Mellon’s Hamerschlag
Hall, Pittsburgh, PA; U.S. Naval Academies’ Chauvenet Hall, Annapolis, MD; Wayne
Community College, Goldsboro, NC; Harvard University, Cambridge, MA; and MIT,
Cambridge, MA, constitute part of a growing list of educational facilities that utilize green roofs
on one or more buildings on their campus. The growth of green roof usage extends beyond the
world of academia (Green Roof Plants, 2006; The Green Scene, 2005). Data from a recent
survey of the green roof industry indicated that total green roof square footage in the U.S. grew
81% between 2004 and 2005 (from 1,186,738 installed in 2004 to 2,149,585 installed in 2005)
(Green Roofs for Healthy Cities, 2006). As this market grows, it provides opportunities for
growers, landscapers, landscape designers, wholesalers and retailers to market to a new customer
base and fill niches in a growing industry still in its infancy. In order to continue to expand this
market, it is important to fully understand the benefits of green roofs, including environmental,
economic, and human, all of which will be presented in future articles.
5.1.3 Literature Cited
City of Chicago. 2005. Building Healthy, Smart, and Green: Chicago's Green Building Agenda
2005. 28 December 2005.
<http://egov.cityofchicago.org/webportal/COCWebPortal/COC_EDITORIAL/BHSGAgenda_1.
pdf>.
5-3
Green Roofs for Healthy Cities. 2005. About green roofs. 9 October 2005.
<http://www.greenroofs.net/index.php?option=com_content&task=view&id=26&Itemid=40>.
Green Roofs for Healthy Cities. 2006. Final report: green roof industry survey 2004 & 2005.
Toronto.
Green Roof Plants. 2006. University/research green roofs. 18 August 2006.

<http://www.greenroofplants.com/university.htm>.
Huang, L.S. 2005. A study of outdoor interactional spaces in high-rise housing. Landscape and
Urban Planning. 4 October 2005. <http://www.sciencedirect.com/science>.
Osmundson, T. 1999. Roof gardens: history, design, and construction. W.W. Norton &
Company, Inc. New York, NY.
Penn State University. 2004. About green roofs. 15 September 2005.

<http://hortweb.cas.psu.edu/research/greenroofcenter/history.html>.
The Garland Company, Inc. 2005. History of green roofs. 5 November 2005.
<http://www.garlandco.com/green-roof-history.html>.
The Green Scene. 2005. The Hamerschlag Hall living roof. 17:1-2.
<http://www.cmu.edu/greenpractices/green_scene/GreenScene17.pdf>.
Walsh, S. 2004/2005. When Sod Goes High-Tech: Could an Eco-Friendly 'Green' Roof Be in
Your Future?. Bright Ideas. Winter 2004/2005:50-52.
Wang, M.S., H.T. Chien. 1999. Environmental behaviour analysis of high-rise building areas in
Taiwan. Building and Environment 34(1):85-93.
Wark, C.G., W.W. Wark. 2003. Green roof specifications and standards. The Construction
Specifier 56(8):1-12.
Wong, N.H., D.K.W. Cheong, H. Yan, J. Soh, C.L. Ong, A. Sia. 2003. The effects of rooftop
garden on energy consumption of a commercial building in Singapore. Energy and Buildings
35(4):353-364.
Yuen, B., N.H. Wong. 2005. Resident perceptions and expectations of rooftop gardens in
Singapore. Landscape and Urban Planning 73(4):263-276.
5-4
5.2 Green Roofs: How They May Be a Benefit to Businesses and Industry
In order to maintain and expand their businesses, large and small companies need to:
operate efficiently, minimize waste, maximize resources, increase employee productivity,
increase their customer base, create brand/company recognition, provide memorable and
convenient service, and lastly, be innovative and competitive with the goods and/or services they
offer. Green roofs offer several advantages to businesses that utilize them, which may include
increased employee productivity, higher employee retention, increased retail sales, higher
building occupancy, and other benefits (Gilhooley, 2002; Lohr, 1996).
5.2.1 Improved Working Conditions

There are potential benefits, through increased worker satisfaction and retention, for
businesses which install green roofs and provide employees with views of natural landscape
(Lohr, 1996). Although the effects of green roofs on worker satisfaction, worker productivity,
and the office environment have not been directly studied, other research indicates the
importance of green spaces to these subjects. Employee’s wages and benefits are the most
expensive business assets a company will acquire, therefore employee retention, satisfaction, and
productivity are paramount to business success (Gilhooley, 2002).
In one study that illustrates the importance of plants in the working environment,
participants were given a timed computer task to complete. During the task, the participant’s
emotional states were surveyed using questionnaires. The control group completed the task in a
windowless room without plants, while the test group performed the same task in a room with
plants. Pre-task surveys showed no significant difference between the two groups, but after
completion of the task, those in the presence of plants had attentiveness scores significantly
higher than participants in rooms without plants. In the same study, participants who completed
the task in the presence of plants were more productive, with a reaction time 12% faster than
participants in a bare room (Lohr, 1996).
A study conducted at Surrey University in the United Kingdom examined the stress
levels of participants completing a complex task in the presence of plant material. Participants
were allowed to rest in a room for 10 minutes, during which a baseline measure of stress was
recorded prior to beginning the task. After the task, participants rested for another 10 minutes in
the office to allow for the collection of post-task stress measurements. Participants who
completed the task in the office with plants experienced lower stress levels throughout the task
and recovered from stress faster after the task than those in the office without plants (Russell,
1999). Results from another study showed that participants were more vigilant in proofreading a
document after a break in a room with plants (Oxford Brooks University, n.d.).
At the BMW headquarters in Munich, Germany, administrators received high levels of
employee complaints related to the workplace. In an attempt to alleviate employee
dissatisfaction, BMW sponsored an internal study that compared planted areas to non-planted
areas in regards to employee well-being, health, motivation, and absenteeism. Administrators
found that employees experienced increased well-being in the greened areas, with 93% of those
working in the planted areas feeling healthier and more motivated than they did prior to the
installation of plants. BMW stated, “The human factor is the No. 1 criterion in determining a
company's relative success” (Plantscapes, n.d.).
5-5
Increased worker productivity and attentiveness saves businesses time and money, and in
addition, workplace stress plays a major role in worker retention and absenteeism. Surveys have
shown that one out of every five workers in the U.S. has left a job in the past year due to
workplace stress. Data also shows that one out of every eight workers in the U.S., or 12%, have
taken time off due to stress in the workplace (Gilhooley, 2002). It is possible that the presence of
green spaces could increase worker satisfaction and reduce workplace stress, which could
increase worker retention, decrease absenteeism, and potentially save money.
5.2.2 Increased Retail Sales

The presence of plants and landscapes can affect a person’s perception of the quality of
products a retailer offers. Consumers perceive a building with interior planting as having a more
expensive appearance, and as such, research has shown that perceptual responses in humans are
related to price acceptance and patronage behaviors (Oxford Brooks University, 1999).
Researchers investigating the perceived quality of different shopping districts found that product
quality ratings were 30% higher in districts with tree-lined sidewalks versus those with barren
sidewalks. Customer service was also perceived as being better in landscaped shopping districts.
These perceptions of higher quality translate into a customer’s willingness to shop and spend.
Respondents reported a willingness to drive farther to shop at the landscaped shopping areas, and
reported a desire to shop for a longer period of time. Similarly, respondents also reported that
they were willing to pay more for equivalent goods in business districts with trees. According to
survey results, landscaped shopping districts could charge 11.95% more than districts with bare
sidewalks for the same goods (Wolf, 2002). By including the use of green spaces in a retail
atmosphere, retailers can create an environment that promotes shopping and spending.
Use of green building technology, such as green roofs, can help a business increase sales
and improve public perception. Giant Eagle, a grocery chain based in Pittsburgh, PA, used
enough green building technology at their Brunswick, OH location to earn an LEED (Leadership
in Energy and Environmental Design) certification, yet the building cost only 2% more to erect.
Since being built, the Brunswick, OH Giant Eagle has become one of the chain’s best
performers, with sales exceeding projections by 20%. A Giant Eagle executive speculates that
the high sales can be attributed to the environmentally friendly building technology used at the
Brunswick store (Dinardo, 2005). This phenomena is not limited to businesses in the U.S. J.
Bryson, manager of an extensively planted shopping center in the United Kingdom, said, “The
annual cost of maintaining the planting of the center is just under 25,000 pounds (44,600
dollars), or 4.3% of the annual service charge. The fact that this massive sum of money has never
been queried…implies a complete acceptance of the existence of plants…as a fundamental factor
in the success of the center.” He also stated, “Through the medium of plants the center…relaxes
almost everybody, all ages and types of people relate to the atmosphere.” This extensively
planted shopping center also has the second highest net profit per square foot in Britain (Bryson,
1992).
5.2.3 Increased Occupancy Rates

It has also been observed that the presence of a green roof can increase interest from
prospective guests and tenants of a given property, and thus increase the value of a building. In
Portland, OR an affordable housing complex, Hamilton West Apartments, reported higher levels
of interest from prospective tenants due to the installation of a green roof. Sean O’Neill, building
manager for Hamilton West Apartments, said, “Everyone loves the eco-roof, not just tenants but
5-6
prospective tenants. I think it is a major selling point for the building, and it’s been very helpful
in attracting new tenants” (Walsh, 2004/2005).
Building owners and managers are often in competition to find occupants for their living
units and office spaces. It is well documented that interior plantings can help attract and retain
tenants. Reduced operating costs, improved environment, and the ‘feel-good factor’ that is
associated with green buildings makes them more attractive to potential occupants (Freeman,
2005). Numerous case studies, including one involving the Opryland Hotel in Nashville, Tenn.,
have shown that tenant occupancy and retention improved 17% with the addition of interior
plantings (Evans, 1992).
The Opryland Hotel is one of the most financially successful hotels for meetings and
conventions in the U.S. It has earned multiple awards including the Golden Key Award from
Meetings and Conventions magazine, the Mobil four-star award, the AAA four-diamond award,
and was named one of the 10 best hotels in the country by readers of Corporate Meetings and
Incentives magazine. The Opryland Hotel has over one million dollars invested in interior and
exterior landscapes in America, boasting of 25 acres of outdoor garden space and over 18,000
interior plants. Fifty-two employees are responsible for maintaining the plants with access to a
budget of about 1.2 million dollars. Results from the study at Opryland concluded that rooms
overlooking the gardens are the first to be reserved by repeat visitors to the hotel, even at a
premium cost. For example, a single night in a double occupancy room with two double beds
reserved for June 10, 2006 costs $199 for a ‘traditional’ view and $264 for a view of the ‘atrium
garden’. In addition to getting premium rates for garden suites, the occupancy rate for Opryland
Hotel has exceeded 85% every year, whereas the occupancy rate in general for hotels in the U.S.
only averages 68% (Evans, 1992). There is evidence that the effect of a green roof on occupancy
may be similar to the increase in occupancy rates at the Opryland Hotel. The Ritz-Carlton Hotel
and Towers located in Boston, Mass. adjacent to the Commons, has reported that many guests
request a room overlooking the green roof on one of their rooftops (Reidy, 2004).
5.2.4 Public Relations

Large corporations are also investigating the benefits of a green building. Wal-Mart has
built an experimental store in McKinney, Texas that incorporates multiple sustainable building
technologies, including a green roof. Wal-Mart spokeswoman Tara Stewart stated, “We wanted
to learn how [we] could improve in the area of sustainability.” Wal-Mart has experienced failing
public perception over the past 10 years, and it has been stated that the current experimentation
with eco-friendly stores is simply a good public relations effort (Dinardo, 2005). It cannot be
denied that green roofs have garnered the attention of journalists and become the subjects of
newspaper articles in New York, Chicago, Seattle, and areas of California. This constitutes free
publicity advertising the environmental initiative of the businesses installing the eco-friendly
technology.
There are other economic benefits of green roofs that could potentially benefits builders,
building owners, business owners, and homeowners. Green roofs have been effective in reducing
energy usage, increasing roof life, and reducing heating/cooling costs. A future article will
discuss these additional economic benefits of green roofs.
5-7
Bryson, J. 1992. Am I running a greenhouse or a shopping center? BALI Seminar, London, U.K.
1992. < http://www.plants-in-buildings.com/whyplants.php>.
Dinardo, A. 2005. Green giant: Will Wal-Mart's experiment help retailers save the planet?
VM&SD. Oct. 2005. 30-34.
Evans, M.R. 1992. People and plants: A case study in the hotel industry. Symp. The Role of
Hort. in Human Well-being and Social Dev. <http://www.plantsatwork.org/hotels.htm>.
Freeman, K. 2005. Green buildings: the role of interior plants. 5 Oct. 2005. <http://www.plants-
in-buildings.com/whyplants-greenbuildings.php>.
Gilhooley, M.J. 2002. Green green grass of work. Facilities Design and Mgt. 21(9): 26-30.
Lohr, V., C. Pearson-Mims, and G. Goodwin. 1996. Interior plants may improve workers
productivity and reduce stress in a windowless environment. J. of Environmental Hort. 14: 97-
100.
Oxford Brookes University. 1999. Green plants for the feel good factor. Interiorscape Magazine.
9 Oct. 2005. < http://www.interiorscape.com/rentokil/>.
Oxford Brookes University. (n.d.). Impact of workplace plants on: Perception and use of planted
space: Recovery from poor attention and fatigue and used vigilance tests, including a proof
reading. 15 Oct.r 2005. <http://www.plantsatwork.org/pdf/perception.pdf>.
Plantscapes. (n.d.). BMW in Munich goes green. Plantscapes. 5 Jan. 2006. <
http://www.plantscapes-officeplants.co.uk/htmlsite/office-plants-bmw.htm>.
Reidy, C. 2004. Roof gardens: Looking good, saving money. The Boston Globe. 6 Sept. 2004. 3:
D1.
Russell, H. 1999. The Psychological Effects of Plants on People in Offices. Interiorscape

Magazine. 9 Oct. 2005. < http://www.interiorscape.com/rentokil/>.
Walsh, S. 2004/2005. When sod goes high-tech: Could an eco-friendly 'green' roof be in your
future? Bright Ideas. Winter 2004/2005: 50-52.
Wolf, K.L. 2002. Retail and urban life: Creating a consumer habitat. Proc. 2002 People/Plant
Symposium, Amsterdam, Netherlands. 17 Nov.r 2005.
<http://www.planterra.com/SymposiumWolf.pdf>.
Wolf, K.L. 2003. Public response to the urban forest in inner-city business districts. J. of
Aboriculture. 29(3): 117-126.
5-8
5.3 Green Roofs: Longer Roof Life and Reduced Energy Consumption Yields
Valuable Economic Benefits
Green roofs, aside from possibly benefiting businesses, may also provide economic
benefits by providing a longer roof life, and allowing for savings on heating and cooling costs
(Pennsylvania State University, 2004). Installing a green roof reduces heat flux through the roof,
reducing cooling costs in summer, and lowering heating costs in the winter. A study in Ottawa,
Canada, found that a green roof reduced heat losses by 26% and heat gains were reduced by 95%
(Liu et. al., 2003). Other economic benefits that occur in some green roof installations include
reducing the amount of insulation used, reducing or eliminating the need for roof drains, the
ability to incorporate cooling or water treatment functions, receiving positive media exposure,
meeting regulatory requirements for stormwater management, and reducing community
resistance to new development (Green Roofs for Healthy Cities, 2005). The city of Chicago is
creating an expedited, consolidated permit approval process for developers utilizing green
building technologies. This expedited approval process will save time and money for building
owners and developers by bypassing stages of approval that may postpone construction (City of
Chicago, 2005). The American Society of Landscape Architects cited economics as one of the
reasons they installed a green roof on their downtown Washington, D.C. headquarters. They
stated that their choice was economically efficient since a green roof lasts longer and requires
less maintenance than a regular roof, helps reduce heating and cooling costs, and decreases the
amount of insulation required (Greener World Media, 2005).
5.3.1 Longer Roof Life

Some studies reported that green roofs last more than twice as long as regular roofs, and
despite significantly higher initial costs, studies have shown them to be similar to, or less
expensive than, traditional roofs over time (Wong et.al., 2003). There is a general lack of
understanding of both the direct and indirect benefits of green roof, making the expense of a
green roof less appealing. Another difficulty in quantifying the cost of green roofs is assigning a
dollar value to the environmental benefits of green roofing systems, such as stormwater retention
and a cooler microclimate. In some European countries entire service industries, such as green
roof maintenance and installation companies, have been formed around green roofs, reducing the
initial costs (Wong et.al., 2003).
It has been proven that green roofing systems last at least three times longer than their
traditional counterparts. This longevity is in part due to the effect of the plants and media
shielding the roof from solar exposure and temperature extremes often reached on traditional
roofs. Protection offered from vegetative cover neutralizes the thermal intensity of rooftop
exposure and minimizes temperature fluctuations. In Singapore, thermal intensities of up to
100ºC were reduced and temperature fluctuations were decreased to 5ºC due to the cover
afforded by a green roof. Increasing the lifespan of the roofing system means that future
maintenance and replacement costs are minimized. Singapore, like the U.S., does not benefit
from a widespread use of green roof technology, and since economies of scale are not present,
the initial cost of installation can be a barrier to utilization. The similarity between Singapore and
the U.S. in regards to the high initial cost of a green roof makes it an applicable comparison for
economic studies of green roof usage.
The National University of Singapore examined the initial cost implications of a green
roof compared to a flat roof (Wong et.al., 2003). Two types of life cycle cost analysis were
5-9
undertaken to compare green roofs to an exposed flat roof; the first does not account for energy
cost savings, while the second does. When energy cost savings are not accounted for, both
intensive and extensive green roofing systems have a higher life cycle cost than an exposed roof;
however, the cost difference between the extensive green roof and the bare roof was marginal
(less than 2.4%). When the life cycle cost analysis was completed and accounted for energy cost
savings, the extensive green roof was more cost effective than the traditional flat roof (Wong
et.al., 2003). Although no studies have been reported on the life cycle costs of green roofs
versus traditional roofing systems in the U.S., the U.S. EPA stated that, “future summertime
energy savings brings the price of a green roof closer to that of a traditional roof.” The U.S. EPA
has also recognized that green roofs may be more affordable over the life of a building due to a
longer life span (Walsh, 2004/2005).
5.3.2 Insulation/Heating and Cooling Benefits

Green roofs have a long history of being used as a means of insulation against
temperature extremes, such as sod roofs used by Norwegians and settlers of the Great Plains,
U.S. (Osmundson, 1999). Although roofing technology is far more advanced than it was in the
days of the early U.S. settlers, green roofs still have the ability to moderate temperature
extremes. Lower energy costs are seen as one of the advantages of green roof technology.
Another advantage stated by architects is the possibility that the lack of heat buildup on a green
roof leads to more efficient operation of air-cooling and ventilation systems (Reidy, 2004).
During a U.S. energy crisis in the 1970s, there was a massive rise in earth-shelter (houses and
buildings located underground) technology and research due to the rise in the cost of heating
fuels. Green roofs can serve a similar purpose by reducing the temperature fluctuations within a
building and reducing the need for heating and cooling (Osmundson, 1999). Being able to better
control temperature within buildings, thereby saving energy, can also lead to additional savings
for the business. In 1993 a study was completed illustrating that slight reductions in office
temperature, of about 1.5ºC, reduces worker absenteeism due to health complaints, thereby
boosting workplace productivity (as reported in Bergs, 2002).
Along with U.S. cities, Chicago and Portland, OR, the city of London in the United
Kingdom is looking into the widespread and legislated use of green roofs since they reduce
energy use thereby cutting costs, are an asset to stormwater management programs, and provide
environmental benefits. It is widely recognized that green roofs have the potential to reduce the
energy consumption of cities, especially when more than one roof in the same area are greened.
Reduction of energy consumption and the lowering of urban temperatures have become a major
topic of discussion among policy makers in large cities. The urban heat island effect can increase
temperatures by 12ºC compared to suburban temperatures, increasing the need to provide
energy-driven artificial cooling during hot weather (Murray, 2005). A study conducted by the
City of Toronto, Canada, and Environment Canada (a weather forecasting and environmental
inspection provider) states that if half of all roofs in Toronto were greened, a reduction of at least
1-2ºC would occur in the urban heat island (as reported in Green Roofs for Healthy Cities, 2005).
According to Leslie Hoffman, executive director of Earth Pledge (the organization leading the
Silvercup Studios project which will put the largest green roof in New York on the building
made famous by “The Sopranos”), the benefits of green roofs in reducing temperatures on the
rooftop and in the surrounding environment, as well as reducing energy use, is increased when
more roofs exist in one area. When large areas of greened rooftops are clustered in one area, it
creates a microclimate that reduces the temperature further than isolated areas of greenery would,
5-10
thereby increasing the energy cost savings, air pollution reduction, and other benefits derived
from the installation of eco-roofs (as reported in Chamberlain, 2005).
Possible energy savings from the use of green roofs vary depending on a multitude of
factors such as climate, building size, media depth, vegetation used, humidity, location, and
season changes. This makes developing a satisfactory model for the purposes of estimating
energy and cost savings difficult (R.D. Berghage, personal communication; Chamberlain, 2005;
Niachou et.al., 2001; Onmura et.al., 2001). Dr. Brad Bass of Environment Canada’s Adaptation
and Impacts Research Division (at the University of Toronto’s Centre for Environment) stated
that the projected annual energy saving for an eight-story building in Madrid, Spain would be
6.4%, but that estimated energy savings based a simulation conducted for a single hot day would
be closer to 10% (as reported in Murray, 2005). Using the Micro Axess Simulation model,
Environment Canada estimated a one-story building with 3.9” of media would experience a 25%
reduction of summer cooling needs. Dr. Bass believes the reduction of urban temperatures and
the reduction in temperature fluctuations resulting from green roof implementation would reduce
the demand on the power grid during a heat wave, making green roofs appealing to power
companies (as reported in Murray, 2005). Supporting evidence indicates temperatures are
measurably more stable on a green roof than on a traditional roof. Green roof surface
temperatures during peak daytime hours in July are 19-31% cooler (Chamberlain, 2005).
Researchers have continued to explore and develop models which evaluate performance
of green roofs in regard to the thermal protection of buildings. A green roof system is difficult to
model due to the many variables that affect its performance. For instance, evaporative cooling
and evapotranspiration of vegetative material greatly improves the effectiveness of a green roof.
In times of drought or minimal rainfall, the amount of cooling from plant material and
evaporation will be lower. An improved model found, in one simulation, an average of 7.2ºC
reduction in indoor air temperature due to the addition of an eco-roof with the maximum
reduction observed during the hottest period of the day, 12:00 to 15:00 hours. This model also
presumed that the heating flux entering a green roof was four times less than that entering a bare
roof. Green roofs also have very little fluctuation in the amount of heat flux when compared to
traditional roofing systems. One aspect this model carefully considered is the affect of L.A.I.
(leaf area index, or the amount of leaf tissue per a given amount of space) on the amount of heat
flux and the efficiency of a green roof. A green roof exhibits peak performance when the L.A.I.
is high, as the vegetative material acts as insulation and a cooling system (Kumar et.al., 2005).
One researcher investigated the performance of green roof thermal properties and the
variations in energy savings using a model. It was demonstrated that some of the variation in
energy savings experienced are due to vegetation, building construction (greater savings are
possible in older building with less existing insulation), and other factors, although it was
focused on how the amount of insulation a building affects the added benefits of installing a
green roof. The model was used to study energy savings pertaining to three different building
types: heavily insulated (as might be expected in modern buildings built to be energy efficient),
moderately insulated, and minimally insulated (as is often the case in older buildings) (Niachou
et.al., 2001). Buildings with minimal insulation would most likely experience a 45% energy
savings when a green roof was installed, conversely, a building that was heavily insulated
received only a 2% benefit. Retrofitting older buildings, and buildings with poor insulation,
could provide a true benefit to business owners in regards to energy savings (Niachou et.al.,
2001).
5-11
Green roofs may provide cost savings, but in the U.S. initial costs remain high. While the
price may be a source of hesitation for some building owners, there are laws and subsidies in
several areas of the U.S. that necessitate or provide monetary incentives for businesses or private
property owners to install a green roof. Since major cities are experiencing overburdened
wastewater management systems, green roofs are one way to provide relief and minimize the
impact of new development on existing water management systems (R.D. Berghage, personal
communication). Currently, the concept of green roofs is receiving more attention from city
authorities in the U.S. in part because they help mitigate stormwater. In addition, green roofs are
used in some areas to promote biodiversity and other environmental benefits. These functions of
green roofs will be featured in the next article of the series.
Barrio, E.P. 1998. Analysis of the green roofs cooling potential in buildings. Energy and
Buildings. 27(2): 179-193.
Bergs, J. 2002. The effect of healthy workplaces on the well-being and productivity of office
workers. 2002 Plants for People Intl. Symp. 9 Oct. 2002.
<http://scholar.google.com/scholar?hl=en&lr=&safe=off&q=cache:8xr9AyMG_4oJ:plantsatwor
k.org/pdf/SymposiumBergs.pdf+author:%22Bergs%22+intitle:%22The+Effect+of+Healthy+Wo
rkplaces+on+the+Well-being+and+...%22>.
Chamberlain, L. 2005. A roof garden? It's much more than that. The New York Times. Section
C, Business: 8.
Chamberlain, L. 2005. 'Green' roofs sprouting across U.S. skylines. The New York Times.
Finance: 15.
City of Chicago. 2005. Building healthy, smart, and green: Chicago's green building agenda
2005. 28 Dec. 2005.
pdf>.
Green Roofs for Healthy Cities. 2005. About Green Roofs. 9 Oct. 2005.
<http://www.greenroofs.net/index.php?option=com_content&task=view&id=26&Itemid=40>.
Greener World Media. 2005. News: Architectural group to install green roof at D.C.
headquarters. 9 Nov. 2005.
<http://www.greenerbuildings.com/news_detail.cfm?NewsID=2791&print=true>.
Liu, K. and B. Baskaran. Thermal performance of green roofs through field evaluation. North
Amer. Green Roof Infastructure Conf., Chicago, 29-30 May 2003. <http://irc.nrc-
cnrc.gc.ca/ircpubs>.
5-12
Kumar, R. and S.C. Kaushik. 2005. Performance evaluation of green roof and shading for
thermal protection of buildings. Building and Environ. 40(11):1505-1511.
Murray, S. 2005. Rooftop gardens that create urban cool. The Financial Times, Business Life
Science and Technology. London 1: 14.
Niachou, A., K. Papakonstantinou, M. Santamouris, A. Tsangrassoulis, and G. Mihalakakou.

2001. Analysis of the green roof thermal properties and investigation of its energy performance.
Energy and Buildings. 33(7): 719-729.
Onmura, S., M. Matsumoto, and S. Hokoi. 2001. Study on evaporative cooling effect of roof
lawn gardens. Energy and Buildings. 33(7): 653-666.
Pennsylvania State University. 2004. About green roofs. 15 Sept. 2005.

<http://hortweb.cas.psu.edu/research/greenroofcenter/history.html>.
Reidy, C. 2004. Roof gardens: Looking good, saving money. The Boston Globe. 6 Sept. 2004. 3:
D1.
Wong, N. H., D.K.W. Cheong, H. Yan, J. Soh, C.L. Ong, and A. Sia. 2003. The effects of
rooftop garden on energy consumption of a commercial building in Singapore. Energy and
Buildings. 35(4): 353-364.
5-13
5.4 Green Roofs: Reducing the Impact of Urbanization and What is Being Done to
Encourage Usage
Cities throughout the world are showing high levels of interest in increasing the use of
green roofs. Green roofs manage the effect of rain water falling on the increasing amount of non-
porous surface area in cities and thereby reduce the impact of runoff on wastewater management
systems. Managing the effect of additional buildings on wildlife populations (e.g., the Black
Redstart, a bird common to Britain, U.K.) might also be achieved by the use of green roofs.
Birds, insects, and reptiles, etc., could use the green roofs as habitat or breeding grounds (Evans,
2005). For these and other reasons, some U.S. cities are seriously considering how they can
promote the use of green roofs on buildings by offering grants and subsidies.
5.4.1 Stormwater Mitigation

The majority of attention regarding green roofing systems has focused on its major
contribution to the urban environment, primarily stormwater mitigation. Since many major cities
are experiencing overburdened wastewater management systems, green roofs are one way to
provide relief and minimize the impact of new development on existing systems. Currently,
green roofs are getting more attention from city authorities in the U.S. because of potential
energy savings and stormwater mitigation. For example, in New York City there are drainage
basins where as little as one-twentieth of an inch of rain could cause overflow (Murray, 2005).
Some cities, including Seattle, WA., Portland, OR., and Chicago, IL., are already including green
roofs as part of their stormwater management programs (Walsh, 2004/2005). Although some of
the other benefits of green roof technology may be difficult to assign dollar figures, the savings
in sewage treatment costs are more tangible as they have the potential to be measured in dollars
per gallon.
Green roofs absorb water during a rainfall event, delaying and reducing the run-off.
When this run-off is unabsorbed and a stormwater or sewage system overflows, the overflow
becomes an expensive management problem for the city as it can pollute waterways. This
pollution has caused the U.S. EPA to pressure cities to initiate better stormwater management
practices (Murray, 2005). Recently Balmori & Associates, a landscape design firm in New York
City, conducted a comprehensive assessment of New York City’s flat roofed buildings. They
concluded that Long Island City alone has 667 acres of flat roof surfaces that are suitable for
greening. If this amount of roof were greened in one area, substantial reductions in stormwater
runoff, air pollution, the urban heat island effect, and energy grid loading could be expected
(Chamberlain, 2005).
5.4.2 Biodiversity and Other Environmental Benefits

Green roofs can be used as alternative habitats for wildlife and threatened species (Evans,
2005; Frith et. al., 2005). Some buildings in Asia and Europe have unique amenities that attempt
to attract, or facilitate migration of, wildlife to green roofs. For example, building owners have
created networks of ‘ladders’ to enable lizards and other creatures to scale the building and reach
the green roof, whereas other green roofs may consist of piles of rocks and woody debris that
mimic the breeding habitat of birds (Berghage, personal communication). In London, U.K.,
much attention has been given to the possibility that green roofs may be able to increase the
biodiversity in urban areas by providing wildlife habitat above the ground (Donald, 2005).
5-14
Green roofs also have the ability to neutralize acid rainfall and filter out some impurities
and toxins (R.D. Berghage, personal communication; Oster, 2005), as the roof media can
effectively neutralize acid rain and maintain a media pH that facilitates plant growth and
development. After several years of exposure to such precipitation, the ability of the media to
neutralize acid rain will decline. Based on results from soil tests, media can be recharged without
replacement or major maintenance if a combination of quick acting and slow release lime is used
to raise the pH to an acceptable level (R.D. Berghage, personal communication).
Green roofs may eventually aid in the reduction of the urban heat island effect in large
cities. Long Island City, N.Y. has 667 acres (270 hectares), equivalent to more than 75% of
Central Park, of empty flat-roof surfaces that could easily be converted to a green roof
(Chamberlain, 2005b). If all of the suitable roofs in Long Island City, N.Y. were greened, a
measurable reduction of the urban heat island effect would occur. An article by ‘Look Japan’
estimated the temperature in Tokyo, Japan, could be lowered by 0.11 to 0.84ºC if 50% of all
available rooftop space were greened, which would result in the savings of approximately
$953,380 each day for the city, just in electricity (as reported in Yuen et. al., 2005). By amassing
large numbers of green roofs in one geographic area, the creation of a microclimate could reduce
some of the environmental impacts of green space loss due to rampant urbanization.
5.4.3 Law and Subsidies

Laws, regulations, and subsidies that support or mandate the installation of green roofs
are in place in several European countries, including Great Britain, Switzerland, and Germany.
These regulations have various purposes, including habitat preservation, stormwater
management, reducing the urban heat island effect, providing social areas, and energy
conservation. In the U.K., the endangered Black Redstart, Phoenicurus ochruros (a bird), nests
on the ground in abandoned industrial sites and are being displaced by reconstruction of former
industrial areas. To encourage preservation of the bird’s population, which is estimated to be less
than 100 pairs still nesting in Britain, the bird and its habitat are protected under Schedule 1 of
the Wildlife and Countryside Act of 1981 (Frith, 2005). Builders who want to obtain permits to
reconstruct old industrial sites must first create new habitats for the birds if the building in
question currently serves as a nesting area (Evans, 2005). Due to the high value of land, the most
economical fashion of complying with this law is to install a green roof on the new building with
the objective of creating habitat for this endangered species (R.D. Berghage, personal
communication). In some areas of Germany and Switzerland, a green roof is required in order to
get a permit to build, and in other areas a flat roof over certain dimensions must, by law, be
greened (R.D. Berghage, personal communication; Donald, 2005).
Even in the U.S., regulations supporting the installation of green roofs are being created,
as cities such as Portland, Chicago, Seattle, and Boston have issued green roof guidelines (Frith
et. al., 2005). Municipal codes in the city of Chicago are currently under review, with changes
already made to energy conservation requirements, and all new flat-roofed city buildings are
being outfitted with green roofs. A 2005 pamphlet from the City of Chicago on green building
states, “The city has adopted a policy that encourages and, in some cases, requires green roofs
and adherence to green building standards in developments undergoing Department of Planning
and Development review.” The pamphlet outlines some specifics, including the stipulation that
discount mass merchandisers must cover at least 50% of their roof surface in greenery if they
receive LEED certification, and 75% of the roof must be greened without LEED certification
(City of Chicago, 2005a). The city of Toronto, Canada is also beginning to adopt policies
5-15
requiring the installation of greenery on all new city buildings with a suitable roof. It is also
going to require that green roofs be installed as a provision of certain types of low-interest loans
that support eco-friendly buildings (City of Toronto, 2005).
Various grants have also been able to help defray the costs of green roof installation. A
$500,000 grant was recently awarded to aid in the design and construction of a 35,000-square
foot green roof on Silvercup Studios in Queens, New York (Chamberlain, 2005a). Chicago’s
Department of Environment began offering $5,000 grants for residential and small commercial
building owners to aid in the planning and installation of green roofs (Merritt, 2005). Chicago is
also exploring other financial and policy incentives that can be utilized to further green building
initiatives. Current and developing incentives for green buildings are not targeted at the
developing community alone, but also benefit business owners, homeowners, financial
institutions, and insurance providers (City of Chicago, 2005). Toronto is also creating grant
programs to support the implementation of green roof systems, where such pilot programs will
be focusing on retrofitting existing buildings with an eco-roof. Other incentives are being
developed to encourage the use of green roofs in Toronto, they include reduced water rates for
buildings with green roofs, certain building permit processes will be hastened or simplified, and
free training will be provided (City of Toronto, 2005). Portland has been developing a
stormwater discount program since 2000, and expects to implement it in 2006. This program will
provide discounts to home and business owners that reduce water runoff from their rooftops, or
use alternative management methods (Portland Bureau of Environmental Services, 2004). Grants
from the City of Portland, the Portland Development Commission, and the Energy Trust of
Oregon are also being made available as financial incentives for various green building
initiatives (Portland’s Office of Sustainable Development, 2005).
In addition to making cities better through decreased loads on the wastewater
management system and additional habitat for flora and fauna, green roofs may provide
significant human benefits. While there exists a rich body of evidence that plants and views of
nature serve to calm the human psyche and enrich human well-being, there is no direct research
on the effect of green roofs on human health and well-being (Fjeld, 2005). In the last article of
the green roof series the potential benefits of greened rooftops on human health and well-being
will be discussed along with the potential for future research.
Chamberlain, L. 2005a. A roof garden? It's much more than that. The New York Times. Section
C, Business: 8.
Chamberlain, L. 2005b. 'Green' roofs sprouting across U.S. skylines. The New York Times.
Finance: 15.
City of Chicago. 2005a. Building Healthy, Smart, and Green: Chicago's Green Building Agenda
2005. 28 December 2005.
pdf>.
City of Chicago. 2005b. Green roofs grant program 2005. 4 Nov. 2005.
<http://egov.cityofchicago.org/city/webportal/portalContentItemAction.do?BV_SessionID=@@
5-16
@@0534030788.1131735627@@@@&BV_EngineID=ccceaddgedhjdhecefecelldffhdfgm.0&c
ontentOID=536930896&contenTypeName=COC_EDITORIAL&topChannelName=Dept&block
Name=Environment%2FI+Want+To&context=dept&channelId=0&programId=0&entityName=
Environment&deptMainCategoryOID=>.
City of Toronto. 2005. Milestones of policy making. 3 Jan. 2006.

<http://www.toronto.ca/greenroofs/policy.htm>.
Donald, C. 2005. How green does your roof grow? Sunday Times, Features: Home: 52.
Evans, P. 2005. Bird's-eye view. The Guardian, London, U.K. 2 Oct. 2005.
<http://society.guardian.co.uk/environment/story/0,14124,1526913,00.html>.
Fjeld. T. 2005. Do plants in offices promote health? 15 Nov. 2005. < http://www.plants-in-
buildings.com/documents/FjeldDoPlantsinOfficespromoteHealth.pdf?PHPSESSID=e653e7b957
ce5bc2b6>
Frith, M., P. Sinnadurai, and D. Gedge. 2005. Advice for [Redstart] conservation in London. 28
Dec. 2005. <http://www.wildlondon.org.uk/resourcefiles/20040625132153Black+Redstart.doc>.
Greener World Media. 2005. News: Architectural group to install green roof at D.C.
headquarters. 9 Nov. 2005.
<http://www.greenerbuildings.com/news_detail.cfm?NewsID=2791&print=true>.
Merritt, L. 2005. City launches green roof grant program. 15 Nov. 2005.
<http://egov.cityofchicago.org/city/webportal/portalContentItemAction.do?topChannelName=H
omePage&contentOID=536932287&Failed_Reason=Invalid+timestamp,+engine+has+been+rest
arted&contenTypeName=COC_EDITORIAL&com.broadvision.session.new=Yes&Failed_Page
=%2fwebportal%2fportalContentItemAction.do>.
Murray, S. 2005. Rooftop gardens that create urban cool. The Financial Times, Business Life
Science and Technology. London 1: 14.
Oster, D. 2005. Green roofs; Save energy, cut pollution and look good. Pittsburgh Post-Gazette.
Lifestyle: C-1. 5 Oct. 2005. <http://www.post-gazette.com/pg/05218/549461.stm>.
Portland Bureau of Environmental Services. 2004. Stormwater Management Manual. 4 Nov.

2005. < http://www.portlandonline.com/shared/cfm/image.cfm?id=55741>.
Portland’s Office of Sustainable Development. 2005. Financial incentives for commercial

development. 3 Jan. 2006. < http://www.green-rated.org/incentives.asp?md=commercial>.
Yuen, B., and N.H. Wong. 2005. Resident perceptions and expectations of rooftop gardens in
Singapore. Landscape and Urban Planning. 73(4): 263-276.
5-17
5.5 Green Roofs: Potential Benefits to Human Health and Well-being
Questions arise as to whether green roofs, like other studies conducted with
interiorscapes, gardens, and natural settings, could improve human quality of life, social
interactions, health event outcomes, stress levels, and moods. There exists a rich body of
evidence that plants and views of nature serve to calm the human psyche and enrich human well-
being (Fjeld, 2005). Increased urbanization and the high value of land in some parts of the world
make setting aside large tracts of land for parks, stormwater mitigation, and natural habitat
impossible. By utilizing unused rooftops, the amount of available green space can be extended
(R.D. Berghage, personal communication). The use of intensive green roofing systems makes the
existence of rooftop parks and public areas possible, thereby enriching the urban social
environment (Yuen et. al., 2005).
Substantial evidence exists illustrating that the environment has a substantial affect on
stress levels, recovery, and immunization (Bergs, 2002; Freeman, 2005; Gilhooley, 2002;
Lothian et. al., 2005; Oxford Brookes University, 1999; Parsons, 1998; Plants-in-buildings,
2005; Plantscapes, (n.d.); Ulrich, 1984). One theory that explains the relationship between the
natural environment and human psychological health hints at a long history of human/nature
interaction. Plants and nature represent part of the ecosystem in which mankind evolved as a
species. It is believed that human evolution (pre- Homo sapiens) began 4.5 million years ago,
and it is also estimated that plant life has existed for 150 million years. This sequence of events
means that the evolutionary history of humans has been closely linked to the natural
environment. Modern human species, Homo sapiens, has existed for about 100,000 years, and
the characteristics of human kind have changed little in the past 10,000 years. Mankind today is
almost biologically identical to those who lived thousands of years ago, long before the creation
of an industrial society which occurred about 250 years ago. While Homo sapiens have not
changed considerably in the recent past, the living environment has. Within the last century
humans interactions with nature have been reduced or severed in industrialized nations, meaning
humans spends a large portion of their time in an artificial environment (Fjeld, 2005). Artificial
environments can pose a risk to human psychological welfare; studies in the field of
environmental psychology have shown that our surroundings have a significant effect on
emotional stability, stress limits, and the sense of well-being (Bergs, 2002; Craig, 2003; Parsons,
1998; Whitehouse et. al., 2001). An explanation for the link between human well-being and
environment is psychological identity. The concept of psychological identity is that human
minds, as well as human bodies, adapted to ensure survival in the wild. In a natural setting, the
human psyche may switch into an automatic mental state that recognizes nature and natural
components as something familiar. Conversely, placed in an unfamiliar environment more
mental energy is expended to maintain a higher state of awareness, thereby increasing stress
levels (Fjeld, 2005). Expanding the amount of exposure mankind has to nature and natural
elements could have significant impacts on stress levels and psychological health.
While there is no direct body of evidence which supports the reduction of stress from the
view of a green roof, there is evidence that views of natural environments can differentially
affect stress levels and stress recovery in varying contexts. Studies have shown that
psychological and physiological stress recovery is hastened by views of nature. In one study,
participants were exposed to mild stress while viewing environments dominated by natural or
urban scenery. Those who viewed the videotapes of nature-dominated habitats recovered more
quickly and completely than those who viewed videotapes dominated by urban artifacts (Ulrich
et. al., 1991). Another study compared the affect of roadside environment on stress reduction and
5-18
immunization. Participants were evaluated for stress response by measuring: three channels of
facial electromyographic activity; two channels of electrooculographic activity; and the results of
electrocardiogram, blood pressure, and skin conductance tests. This study videotaped typical
landscaped roadside environments (this constituted the nature dominated drive scenery), and
roadside environments dominated by buildings, road signs, and construction (which constituted
the simulated drive dominated by urban artifacts). After viewing a video of either the urban or
landscaped scenery, a passive or active stressor was administered to analyze stress immunization.
Results indicated that observing environments dominated by natural elements, as they would be
viewed by the driver of a vehicle, reduced stress and helped immunize against future stressors
when compared to drivers viewing roadside environments dominated by urban elements
(Parsons, 1998). It could be inferred that the effect of roadside scenery on stress levels might be
similar to the effect of a green roof on stress levels.
There is substantial evidence that some portion of a motorist’s attention is devoted to
non-task oriented environmental factors (Parsons, 1998). Studies on human interaction with
workplace, medical, retail, and social environments provide evidence that a portion of their
attention is also devoted to non-task oriented aspects of the environment (Bergs, 2002; Bryson,
1992; Craig, 2003; Evans, 1992; Freeman, 2005; Gilhooley, 2002; Lothian et. al., 2005; Milligan
et. al., 2004; Ousset et. al., 1998; PLANET, 2005; Plants-in-buildings, 2005a, b; Russell, 1999;
Sherman et. al., 2005; Ulrich, 2000; Ulrich, 1984; Westphal, 1999; Whitehouse et. al., 2001;
Wolf, 2002; Yuen et. al., 2005). This indicates that the incidental or indirect view of green roof
space could reduce stress.
Reductions in stress may be responsible for better health outcomes that have been
observed among patients with views of nature. Hospital environments are stressful in part due to
the fact that they are complex, technical, and unfamiliar (Kiecolt-Glaser et. al., 1998). It is
thought that sustained exposure to a hospital environment can result in mental fatigue and
cognitive dysfunctions. Evidence also suggests that significant distress or anxiety before and
after surgery resulted in a more complicated and prolonged postoperative recovery (Kiecolt-
Glaser et. al., 1998). Studies have shown that patients with a view of nature have lower rates of
infection, require fewer analgesics, and may have shorter hospital or intensive care stays (Ulrich,
2000; Ulrich, 1984). One study observed surgical patients in a suburban Pennsylvania hospital
matched based on age, pre-surgical health, gender, and other qualities after surgery. Hospital
confinements for surgical patients often limit their access to the outdoors entirely to views from a
window. Records of patients assigned to rooms on the second and third story were divided into
two groups: those patients who had a room on the side of the hospital wing that overlooked a
stand of deciduous trees, and those who had a room on the side of the wing with a window view
of a brick wall. All patients on a given floor received care from the same nurses, and their rooms
were nearly identical with a window placed in such a way that patients lying in a hospital bed
had an unobstructed view of the outdoors. For the purposes of this study only records of patients
that underwent a cholecystectomy, a surgery for gall bladder ailments, during the months that
trees had foliage (May 1 through October 20), were used. Patient’s records were then placed into
matching pairs, based on sex, age, smoking status, weight, nature of previous hospitalizations,
the year of surgery, and floor level. These records were then compared and analyzed for the
number of days of hospitalization, number and strength of analgesics a day, number and strength
of doses of anxiety medication each day, minor complications, and the nurses’ notes on the
patient’s condition and recovery. Negative comments written by nurses about a patient’s
recovery were more common in those with a view of the brick wall than in those with a view of
trees (3.96 per patient viewing the wall compared to 1.13 per patient with a view of trees).
5-19
Moderate and strong doses of painkillers, on days two through five of the hospital stay, were
used significantly less by those with a view of nature than those with a view of the brick wall.
Conversely, the group who viewed trees used more doses of weak painkillers, such as aspirin and
acetaminophen while the group who viewed the wall was given more doses of potent narcotics.
Shorter hospital stays were observed in the tree group: those with a view of the deciduous trees
spend an average of 7.96 days in the hospital, and those with a view of the brick wall spent an
average of 8.70 days in the hospital. Patients who had rooms looking out on a natural area
required fewer painkillers, slightly fewer minor complications, and left the hospital sooner than
participants with a view of other hospital areas. Reducing the amount of drugs used and the
length of hospital stays has the potential to boost human well-being and lower the cost of
healthcare. By influencing the patient’s emotional state and stress level, a patient’s recovery can
be hastened (Ulrich, 1984). Hospital outcomes, staff morale, patient satisfaction, and the
impressions of hospital visitors are all positively impacted by the addition of interior and exterior
landscapes (Plants-in-buildings, 2005a).
Investigators who studied the use of healing gardens in a pediatric cancer center observed
the number of people who visited the hospital’s gardens and the number of patients with views of
these areas who either chose to keep their window blinds open or closed. It was found that there
was an inverse relationship between the number of people in the garden and the number of open
blinds. It appears that the desire for privacy was more important than the ability to receive
natural light and have access to natural views. By using green roofs, patients could have access
to the view of a garden-like natural setting without sacrificing privacy (Sherman et. al., 2005).
Another study researched the utilization and consumer satisfaction of visitors to a children’s
hospital garden. Reasons for visiting the garden were related to relaxation, coping mechanisms,
and stress reduction (a combined 64% of responses); furthermore 90% of the participants
reported positive mood changes, resulting from visiting the garden (Whitehouse et. al., 2001).
The increased well-being may be related to the positive effects of a natural setting which may
promote rest and relaxation.
In a study designed to discover the landscape and design preferences of assisted living
residents, it was found that residents consistently chose interior and exterior designs that
provided views. Using paired photographs: 1) a view digitally altered to include more plant
material; 2) another exterior view (e.g., into another yard, onto a porch, or beyond a fence), or
3) a view that included less plant material, but included additional features (e.g., more walkways,
benches, or a swing), assisted living residents were asked to choose which scene they preferred.
In all cases where participants were shown an unaltered view or the digitally altered photograph
with the added greenery or features, the scene with the additions was chosen (Rodiek et. al.,
2004). Ability to view additional areas and see a natural or landscaped setting is important to
assisted-living residents who spend a majority of their time in a confined setting.
A study by Amanda Read, a student of The Royal Agricultural College, Cirencester,
England illustrated that the presence of plants in a classroom can encourage student attendance
and in-class attentiveness. A group of 34 students was observed over the course of an academic
year during weekly lectures. Class was alternately held in a room with plants and in a room
without plants. In order to make accurate observations, the students (audience) actions were
recorded and the tapes were later analyzed for signs of inattentiveness such as talking, yawning,
or fidgeting. Results showed that in the planted room student inattentiveness was reduced by
70%. Class attendance was also higher for lectures in the planted room, with an attendance rate
of 97.8% compared to 86.4%for the lectures in the unplanted classroom (as reported in Plants-in-
buildings, 2005b). Results of this study further the conclusions of a study done by Virginia Lohr
5-20
in which interior plantings were found to reduce stress and increase productivity in office
workers (Lohr et. al., 1996).
In an urban environment, greening rooftops may be one of a few options left to increase
green space. Research has shown that greenery helps to make an urban environment more livable
for residents, and reduces the negative effects of living in urban centers. As a population’s way
of living becomes more urban, the desire and need for contact with nature increases. It has been
postulated that people will visit urban green spaces on a regular basis when it is within a three to
five-minute walk of home or work (as reported in Yuen et. al., 2005). In Singapore, some city
officials are pushing architects and developers to include rooftop gardens to extend the amount
of park space available to city residents. A study there used focus group discussions and surveys
to explore city residents perceptions and expectations of rooftop gardens. A household survey
was conducted of 333 residents living near (less than a five-minute walk) and further away
(greater than five but less than or equal to a 20-minute walk) from a rooftop garden. Survey
results showed a high awareness level of the rooftop garden, but low rates of utilization.
Residents voiced some concerns over the utilization of rooftop park space, namely the high heat,
the need to climb stairs to gain access to the garden, and safety concerns (Yuen et. al., 2005).
Data also suggested that rooftop garden usage was more prevalent among those 35-54 years old,
and that men were more likely to visit the roof gardens than women. Reasons residents visited
the garden were also investigated, and the primary reasons included taking the children out to
play, getting exercise, and finding peace and quiet. A possible reason for the low usage rates of
rooftop gardens in Singapore may be related to the lack of amenities. Survey respondents and
focus group members suggested the addition of more landscaping, areas for fitness
routines/classes, barbeque pits, snack areas, garden statues, water features, and outdoor exercise
equipment. When asked about benefits of green roofs and rooftop gardens in Singapore,
respondents included better air quality, land use optimization, beautifying the environment, and
the addition of greenery and nature views (Yuen et. al., 2005). T. Osmundson has concisely
summed one benefit of rooftop gardens in urban settings in his book entitled Roof gardens:
history, design, and construction, stated, “A feeling of isolation from the…general confusion of
the typical downtown city street can be sensed in most roof gardens above ground level. It is one
of their major attributes and one which a downtown park at street level can rarely achieve.”
Part of the reason for the need to increase green space in an urban environment is related
to the social breakdown that occurs in highly populated areas and high-density housing. Research
has illustrated that negative effects of residential crowding are due in part to the collapse of
social support systems. In fact, Lin reported that residents of high-rise housing in Taipei do not
desire close relationships with their neighbors, and many believe the opportunity for social
contact is unnecessary (as reported in Huang, 2005). The breakdown of a social construct in
urban residents concerns social psychologists, and could create significant social problems over
time. A study on behavior in high-rise buildings showed that 51.67% of the residents of high-rise
complexes are not satisfied with their living environments. Of the nine reasons stated for
resident’s dissatisfaction, the lack of open spaces ranks number one (Wang et. al., 1999).
Research conducted on the number of social interactions taking place on green roofs in Taiwan
shows that the second highest percentage of social interaction takes place in planted areas
(25.63%). To increase the quantity of social interaction by the use of intensive green roof space,
the study concluded that areas of visual focus, plants, play areas, and open areas encourage social
behavior (Huang, 2005). Provision of greenery and open spaces increases the opportunity for
social interaction and thereby enhances the social construct. Creation of open spaces, park-like
settings, and scenic views may significantly increase the quality of life for urban dwellers. Both
5-21
intensive and extensive green roofs can help fulfill the need of urban dwellers for green spaces.
Studies are needed to discover what types and features of green roofs best fit urban needs.
Despite the fact that the green roof industry is still in its infancy in the U.S. and many
other countries around the world, there has been considerable research into many aspects of
green roofs including possible environmental benefits, energy benefits, stormwater management
benefits, the water quality of green roof runoff, media formulas, and plant selections. Several
questions remain that have not been answered by research to date: Economic impacts of green
roof installations need to be quantified for the U.S. market, human benefits of green roofs (both
extensive and intensive) have yet to be investigated, and in-depth studies of wildlife utilization of
green roofs needs to be researched. With very strong evidence of the effect of views of nature on
human health, well-being, and happiness, it can be postulated that green roofs may have a
significant beneficial effect for those occupying space overlooking the eco-roof. Research
specifically on the human benefits of green roofs is needed to quantify the possible benefits to
human well-being. Once the benefits to human well-being, actual or perceived, are investigated,
further studies can be done to research the possibility of increased productivity, concentration,
and attentiveness that are critical in the workplace environment.
Bergs, J. 2002. The effect of healthy workplaces on the well-being and productivity of office
workers. 2002 Plants for People Intl. Symp. 9 Oct. 2002.
<http://scholar.google.com/scholar?hl=en&lr=&safe=off&q=cache:8xr9AyMG_4oJ:plantsatwor
k.org/pdf/SymposiumBergs.pdf+author:%22Bergs%22+intitle:%22The+Effect+of+Healthy+Wo
rkplaces+on+the+Well-being+and+...%22>.
Bryson, J. 1992. Am I running a greenhouse or a shopping center? BALI Seminar, London, U.K.
1992. < http://www.plants-in-buildings.com/whyplants.php>.
Craig. 2003. Green therapy. The American gardener. 82(5): 46-50.
Evans, M.R. 1992. People and plants: A case study in the hotel industry. Symp. The Role of
Hort. in Human Well-being and Social Dev. <http://www.plantsatwork.org/hotels.htm>.
Fjeld. T. 2005. Do plants in offices promote health? 15 Nov. 2005. < http://www.plants-in-
buildings.com/documents/FjeldDoPlantsinOfficespromoteHealth.pdf?PHPSESSID=e653e7b957
ce5bc2b6>
Freeman, K. 2005. Green buildings: the role of interior plants. 5 Oct. 2005. <http://www.plants-
in-buildings.com/whyplants-greenbuildings.php>.
Gilhooley, M.J. 2002. Green green grass of work. Facilities Design and Mgt. 21(9): 26-30.
Huang, S.L. 2005. A study of outdoor interactional spaces in high-rise housing. Landscape and
Urban Planning. 18 Jan. 2006. <http://www.sciencedirect.com/science>
5-22
Kiecolt-Glaser, J.K., G.G. Page, P.T. Marucha, R.C. MacCallum, and R. Glaser. 1998.
Psychological influences on surgical recovery: Perspectives from psychoneuroimmunology.
American Psychologist. 53(11): 1209-1218.
Lohr, V., C. Pearson-Mims, and G. Goodwin. 1996. Interior plants may improve workers
productivity and reduce stress in a windowless environment. J. of Environmental Hort. 14: 97-
100.
Lothian, M., and K. Freeman. 2005. Plants reduce stress and increase productivity in the
workplace. 7 Nov. 2005. < http://www.plants-in-buildings.com/whyplantsstressreduction.php>.
Milligan, C., A. Gatrell, and A. Bingley. 2004. 'Cultivating Health': Theraputic landsapes and
older people in northern England. Social Science and Medicine. 58: 1781-1793.
Ousset P.J., F. Nourhashemi, J. Albarede, and P. Vellas. 1998. Theraputic Gardens. Geriatrics. 6:
369-372.
Oxford Brookes University. 1999. Green plants for the feel good factor. Interiorscape Magazine.
9 Oct. 2005. < http://www.interiorscape.com/rentokil/>.
Parsons, R. 1998. The View from the Road: Implications for Stress Recovery and Immunization.
J. of Environmental Psychology. 18: 113-139.
PLANET. 2005. Interior Plants in Offices Are Vital to Human Comfort and Health. 5 Oct. 2005.
<www.alca.org>.
Plants-in-buildings. 2005. Why plants in hospitals are good for you. 22 Nov. 2005.
<http://www.plants-in-buildings.com/news-hospitals-010205.php>.
Plants-in-buildings. 2005. Research news: plants may improve student performance. 12 Oct.
2005. <http://www.plants-in-buildings.com/news-230805-racstudy.php>.
Plantscapes. (n.d.). BMW in Munich goes green. Plantscapes. 5 Jan. 2006. <
http://www.plantscapes-officeplants.co.uk/htmlsite/office-plants-bmw.htm>.
Rodiek S., and J. Fried. 2004. Access to the outdoors: Using photographic comparison to assess
preferences of assisted living residents. Landscape and Urban Planning. 73: 184-199.
Russell, H. 1999. The Psychological Effects of Plants on People in Offices. Interiorscape

Magazine. 9 Oct. 2005. < http://www.interiorscape.com/rentokil/>.
Sherman, A., J. Varni, R.S. Ulrich, and V. Malcarne. 2005. Post-occupancy evaluation of healing
gardens in a pediatric cancer center. Landscape and Urban Planning. 73: 167-183.
5-23
Ulrich, R.S. 1984. View through a window may influence recovery from surgery. Science.
224(4647): 420-421.
Ulrich, R.S., R.F. Simons, B.D. Losito, E. Fiorito, M.A. Miles, and M. Zelson. 1991. Stress
recovery during exposure to natural and urban environments. J. of Environmental Psychology.
11(3): 201-230.
Ulrich, R.S. 2000. Evidence based environmental design for improving medical outcomes. Conf.
at McGill University Health Centre, Montreal, Can.
Wang, M.S. and H.T. Chien. 1999. Environmental behaviour analysis of high-rise building areas
in Taiwan. Building and Environ. 34(1): 85-93.
Westphal, L.M. 1999. Growing Power: Social benefits of urban greening projects. Univ. of Ill.,
Chicago, PhD Diss.
Whitehouse, S., J. Varni, M. Seid, C. Cooper-Marcus, M. Ensberg, J. Jacobs, and R.

Mehlenbeck. 2001. Evaluating a children's hospital garden environment: utilization and
consumer satisfaction. J. of Environmental Psychology. 21: 301-314.
Wolf, K.L. 2002. Retail and urban life: Creating a consumer habitat. Proc. 2002 People/Plant
Symposium, Amsterdam, Netherlands. 17 Nov.r 2005.
<http://www.planterra.com/SymposiumWolf.pdf>.
Yuen, B. and N.H. Wong. 2005. Resident perceptions and expectations of rooftop gardens in
Singapore. Landscape and Urban Planning. 73(4): 263-276.
5-24

04 Dec 10sg

Uploaded by

Copyright:

Available Formats

04 Dec 10sg

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

04 Dec 10sg

Uploaded by

Copyright:

Available Formats

National Decentralized Water Resources Capacity Development

Project (NDWRCP) Research Project

QUANTIFYING EVAPORATION AND

Dr. William Hunt

Figure 13 Sedum sexangulare ......................................................................................... 2-17

Figure 14 Delosperma nubigenum .................................................................................. 2-17

Figure 15 Sedum spurium ............................................................................................... 2-18

Figure 2 Percent retention as a function of the precipitation return period..................... 3-5

Figure 3 Titration of a clay based medium with sulfuric acid......................................... 4-6

their tissues to survive where other 1.5

The relative contribution of the 0.5

Medium pH = 7.18 - 50.36 x H+ + M

1.1 Literature Cited

GREEN ROOF PLANT WATER USE

R.D. Berghage, D. Beattie, A.R. Jarrett, F. Rezaei

The Pennsylvania State University

Figure 1. Weighing lysimeters in PSU greenhouse.

2.4.1 Species Responses

b. Daily water loss (mm/day)

c. Difference in water loss

a. Water loss on day

a. Water loss on day 10

0.5 b. Hourly water loss

b. Daily water loss (mm/day)

Average temperature during

a. Water loss on day 2 after

0.5 b. Hourly water loss rates on day 2

a. Water loss on day 10 after irrigation

0.5 b. Hourly water loss rates on day 10

a. Cumulative water loss

4 b. Daily water loss (mm/day)

1 a. Water loss on day 1 after

1 a. Water loss on day 10 after

2.5 Average temperature during

2.4.2 Seasonal Effects

Figure 13. Sedum sexangulare. Figure 14. Delosperma nubigenum

2.6 Summary and Conclusions

Planted (S. spurium)

Unplanted (media only)

Planted (S. spurium)

Plant contribution (planted-unplanted)

ANNUAL AND INDIVIDUAL-STORM GREEN ROOF

Figure 1. Typical Green Roof Profile.

3.3 Green Roof Hydrologic Response Models

3.3.1 Annual Green Roof Response (AGRR) Model

Dgri = Dgri-1 + ETi - Ri (1)

3.3.2 Storm Green Roof Response (SGRR) Model

3.4 Results and Discussion

1024 mm of which 527 mm, or

Raleigh, NC) will be 20

retained on the roof and 15

this portion of rainwater 10

does not require any 5

infrastructure can be Date

sixty-five percent less

without green roofs. 50

Another, less positive 40

Rain & Runoff (mm)

3.4.2 Storm Green Roof Response (SGRR) Model