UNIVERSITY OF CINCINNATI

Transcription

1 UNIVERSITY OF CINCINNATI Date: May 17, 2006 I,, Christine Robertson hereby submit this work as part of the requirements for the degree of: Master of Community Planning in: College of Design, Art, Architecture and Planning It is entitled: A green roof build-out analysis for the University of Cincinnati: Quantifying the reduction of stormwater runoff This work and its defense approved by: Chair: Carla Chifos, PhD Virginai Russell Marilyn Wall Hale Thurston, PhD

2 A green roof build-out analysis for the University of Cincinnati: Quantifying the reduction of stormwater runoff A thesis submitted to the Division of Research and Advanced Studies at the University of Cincinnati in partial fulfillment of the requirements for the degree of MASTER OF COMMUNITY PLANNING in the School of Planning College of Design, Architecture, Art and Planning 2007 by Christine Robertson B.A. English and French, University of California Los Angeles, 2003 Thesis Committee: Chair: Carla Chifos, PhD Virginia Russell, MLA Marilyn Wall Hale Thurston, PhD

3 ABSTRACT Urbanization has dramatically altered the natural landscape by replacing pervious land cover with impervious materials. This has lead to an increase in stormwater runoff which raises a number of environmental and economic concerns. Traditionally, stormwater has been treated with costly engineering solutions including expanding infrastructure and increasing treatment capacity. However, within the last decade green infrastructure, has begun to surface throughout cities across the United States offering the possibility of minimizing the economic and environmental concerns associated with development, while providing a myriad of environmental benefits that traditional solutions exclude. This study looks at the potential contribution that green roofs can have on stormwater runoff reduction through a case study. It examines the University of Cincinnati, comparing existing runoff conditions with a projected green roof coverage scenario using TR-55. It concludes by providing policymakers with a general framework to support a green roof policy for the City of Cincinnati. iii

4 iv

5 ACKNOWLEDGEMENTS I would like to thank the chair of my thesis committee Dr. Carla Chifos for her continual guidance and support throughout this study, as well as fellow committee members Jennie Russell and Marilyn Wall who both helped to plant the green roof seed in my head. Finally, I would like to express my appreciation to Dr. Hale Thurston and Dr. William Shuster of the U.S. Environmental Protection Agency, Cincinnati, Ohio for sharing their guidance and expertise through the methodological part of this study. Lastly, I would like to thank my parents for all their years of unconditional love and encouragement and for instilling within me an appreciation of the environment at a very young age. Thank you. v

6 Everything under the heavens that is horizontal belongs to nature. One must be persistent in the quest to green, or forest, all rooftops so that from a bird s-eye-view, one would only recognize a natural, green landscape. When one creates green roofs, one doesn t need to fear the so-called paving of the landscape: the houses themselves become part of the landscape. People must use the roofs to return to nature what we unlawfully took from her by constructing our homes and buildings the layer of earth for grasses and trees. Friedensreich Hundertwasser vi

7 TABLE OF CONTENTS THE RISE OF IMPERVIOUS SURFACES...1 DEVELOPMENT AND THE LOSS OF THE NATURAL LANDSCAPE...1 IMPERVIOUS SURFACES AND THEIR IMPACT ON WATER QUALITY...2 POSSIBLE SOLUTIONS...8 GREEN ROOF STUDY INSPIRATIONS...9 Toronto, Canada...10 Washington, D.C...11 PROBLEM STATEMENT...14 RESEARCH QUESTION...17 OBJECTIVES...18 STUDY OUTLINE...18 METHODOLOGY...19 Task 1: Literature Review...19 Task 2: Identification of Case Study Site and Determination of Boundary...20 Task 3: Quantification of Stormwater Runoff at University of Cincinnati...21 Task 4: Identification of Green Roof Criteria...23 Task 5: Green Roof Building Inventory...24 Task 6: Determination of a Curve Number for Extensive Green Roofs...25 Task 7: Quantifying Stormwater Runoff after Projected Green Roof Coverage...25 Task 8: Analysis and Findings...26 Task 9: Recommendations and Conclusions...27 GARDENS IN THE SKY...28 DEFINITION...28 TYPES...28 COMPONENTS...30 BENEFITS...33 Amenity Value...35 Food Production...36 Aesthetic Quality...37 Increased Roof Life...37 Insulation and Energy Efficiency...39 Strengthened Public Relations...40 Job Creation...41 Habitat and biodiversity...41 Stormwater Runoff...43 Air Quality...44 Urban Heat Island Effect...45 Fire Prevention...48 CONSIDERATIONS...49 Function...49 Load-Bearing...49 Slope...50 Wind...50 vii

8 Plant Selection...51 Irrigation...53 Schedule...53 Maintenance...54 Cost...54 Building Regulations...56 GREEN ROOF WORLD HISTORY...57 Germany and Europe...59 North America...61 Japan...62 CLOSING REMARKS...63 THE SCIENCE OF HYDROLOGY...65 DEFINITION...65 BACKGROUND...66 HYDROLOGIC MODELS...66 SURFACE AND SUBSURFACE MODELS...67 STORMWATER MODELS...69 TECHNICAL RELEASE CLOSING REMARKS...71 APPLICATION OF THE TR-55 TO THE CASE STUDY...73 EXISTING CONDITIONS...73 Identification of case study site and determination of boundary...73 Quantification of stormwater runoff at University of Cincinnati...76 PROJECTED CONDITIONS...80 Identification of Green Roof Criteria...80 Green Roof Building Inventory...80 Determination of a CN for Extensive Green Roofs...81 Quantifying Stormwater Runoff after Projected Green Roof Coverage...83 ANALYSIS AND FINDINGS...84 LIMITATIONS...94 Data...94 Assumptions Made to Execute the Methodology...95 CONCLUSIONS AND RECOMMENDATIONS...97 CONCLUSION...97 RECOMMENDATIONS Research Policy GLOSSARY BIBLIOGRAPHY viii

9 1 chapter THE RISE OF IMPERVIOUS SURFACES If planners are to approach their field in a sustainable way, they must first understand the relationship between the built environment and the natural landscape. By examining the spatial distribution of land use, planners can determine the ecological, social and economic impact on the location in question (Leotão et al. 2006). DEVELOPMENT AND THE LOSS OF THE NATURAL LANDSCAPE Development has dramatically changed the natural landscape. The natural environment is comprised of forests, grasslands, wetlands and other porous land cover. When it rains in these natural areas, over 90 percent of rainwater is absorbed and filtered through the ground, and over time, is eventually infiltrated into the nearest aquifer (EPA 2003; Benfield et al. 1999). The amount of rainfall that is converted to runoff is less than 10 percent of the rainfall volume (EPA 2003). This cycle of absorption and filtration helps to decontaminate the water, slowly transporting it back into nature and filtering out pollutants in the process. As natural areas become urbanized, however, porous land cover is replaced by large paved surfaces which are impervious to stormwater. In urban areas, impervious surfaces often account for more than 45 percent of land cover (Kloss and Calarusse 2006). The primary components of impervious cover are vehicle related infrastructure such as roads, parking 1

10 lots, driveways and sidewalks which typically account for 60 to 70 percent of the total impervious surface in a watershed (Benfield et al. 1999). Another significant source of impervious cover in the urban environment comes from rooftops. The National Oceanic and Atmospheric Administration (NOAA) and Pennsylvania State University estimate there are 25 million acres of impervious land cover in the continental United States (Kloss and Calarusse 2006). This quantity represents nearly one-quarter of the more than 107 million acres almost 8 percent of non-federal land in the contiguous United States that had been developed by 2002 (USDA 2002). As this percentage of land cover increases it carries with it a number of environmental concerns in addition to the issues concerning water quality. Examples of these problems include the urban heat island effect, erosion, diminished air quality, increased energy consumption and loss of green space. IMPERVIOUS SURFACES AND THEIR IMPACT ON WATER QUALITY The increase in impervious surfaces tremendously alters the natural process of the hydrologic cycle (Benfield et al. 1999). Instead of filtering through the ground, stormwater is trapped above the paved surface, accumulates and eventually runs off at alarming rates. On a developed acre of impermeable surface in a 1-inch storm, 25,000 gallons of water run off the site. By contrast, only 2,700 gallons of stormwater runs off an undeveloped acre of land in the same 1-inch storm (Cathcart 2006). According to the Environmental Protection Agency (EPA), over 10 trillion gallons of urban stormwater runoff flow into the nation s rivers, lakes and oceans each year (EPA 2004). This runoff carries with it a number of pollutants including vehicle-related fluids such as oil, grease, gasoline and hydrocarbons, along with other contaminants such as copper, zinc, lead, 2

11 cadmium, chloride, and nitrate (see Table 1). Bacteria and other pathogens, pesticides, fertilizers, nutrients, sediment and debris are often washed off as well. Sediment clouds the water, killing plants and destroying habitat, while nutrients contribute to algae blooms that can deplete oxygen leading to eutrophication and dead zones where life is no longer able to thrive in that ecosystem. 67 toxic pollutants have been identified in urban runoff (Benfield et al. 1999). These pollutants can become detrimental to aquatic ecosystems and pose a major threat to human health. The EPA reports that in 2002, 21 percent of all swimming beach advisories and closings were attributed to stormwater runoff (EPA 2003). TABLE 1: MAJOR URBAN STORMWATER POLLUTANTS Pollutant Source Bacteria Pet waste, wastewater collection systems Metals Automobiles, roof shingles Nutrients Laws, gardens, atmospheric deposition Oil and grease Automobiles Oxygen-depleting substances Organic matter, trash Pesticides Lawns, gardens Sediment Construction sites, roadways Toxic chemicals Automobiles, industrial facilities Trash and debris Multiple sources Source: Kloss and Calarusse 2006 Research has shown that as the amount of impervious surface cover increases in a watershed, the velocity and volume of surface runoff increases; flooding erosion and pollutant loads in receiving waters increase; groundwater recharge and water table decline; stream beds and flows are altered; and aquatic habitat is impaired (Benfield et al. 1994). This illustrates a strong correlation between the level of impervious cover in a drainage basin and the health of its receiving stream. In fact, scientists have documented that water quality begins to degrade as impervious cover in a watershed exceeds a mere 10 percent (Benfield et al. 1999). Above 10 percent fish 3

12 species begin to decline. When the watershed reaches 25 percent imperviousness additional species will decline, and when levels exceed 30 percent the watershed may be considered degraded. As a result of these findings stormwater runoff has evolved into one of the most pressing problems which impact water quality that cities around the country must address (see Table 2). This problem is gaining attention nationwide due to the EPA s recently promulgated National Pollutant Discharge Elimination System stormwater measures. Phase II of this system requires communities with a population less than 100,000 residents to meet new criteria for stormwater runoff (Thurston 2006). TABLE 2: URBAN STORMWATER S IMPACT ON WATER QUALITY Water Body Type Stormwater s Rank as Pollution Source % of Impaired Waters Affected Ocean shoreline 1 st 55% (miles) Estuaries 2 nd 32% (sq. miles) Great Lakes shoreline 2 nd 4% (miles) Lakes 3 rd 18% (miles) Rivers 4 th 13% (miles) Source: Kloss and Calarusse 2006 The reason behind the new criteria for stormwater runoff is twofold. One motive is to address issues as they currently exist. Another is to prevent or abate these potential impacts in the future. Recent studies indicate that stormwater pollution may begin to increase at a higher rate (Kloss and Calarusse 2006). Over the past two decades, the rate of development has been two times greater than the rate of population growth (USDA 1997). Between 2000 and 2025 U.S. population is predicted to increase by 22 percent which would add an additional 68 million acres of development if the current rates of population growth and land consumption continue (Kloss and Calarusse 2006). By 2030, half of the total square footage of buildings 200 billion square feet will have been built after the year 2000 (Nelson 4

13 2004). As development continues at this rate, impervious surfaces will most likely expand in parallel. HISTORY OF AN ANTIQUATED SYSTEM An additional impact of stormwater runoff is its contribution to combined sewer overflows (CSOs) as well as sanitary sewer overflows (SSOs). Estimates indicate that CSO discharges are generally composed of percent sewage and percent stormwater (Kloss and Calarusse 2006). This combination of sewage and stormwater can present particular threats to human health and the integrity of aquatic ecosystems (see Table 3). TABLE 3: POLLUTANTS IN CSO DISCHARGES Pollutant Pathogenic bacteria, viruses, parasites Fecal coliform (indicator bacteria) Oxygen depleting substances Median CSO Concentration Treated Wastewater Concentration 215,000 colonies/100ml < 200 colonies/100ml 43 mg/l 30 mg/l (BOD 5 ) Suspended solids 127 mg/l 30 mg/l Toxics Cadmium 2 µg/l 0.04 µg/l Copper 40 µg/l 5.2 µg/l Lead 48 µg/l 0.6 µg/l Zinc 156 µg/l 51.9 µg/l Nutrients Total Phosphorus 0.7 mg/l 1.7 mg/l Total Kjeldahi Nitrogen 3.6 mg/l 4 mg/l Trash and debris Varies None Source: Kloss and Calarusse 2006 Combined sewer systems (CSSs) are the remnants of an aged infrastructure designed to transport stormwater and sewage through the same pipe eventually emptying into the water. These systems started to emerge in the mid-1800's, particularly in the Northeast and Great 5

14 Lakes regions 1. At the time, CSSs were believed to be a significant improvement to the health and sanitary conditions of the community compared to the preexisting methods of disposal which included dumping human fecal waste into the streets, alleys and waterways. As population and development continued to increase and people began to realize the health costs associated with polluting drinking water sources, municipalities began to separate sewer and stormwater systems, sending sewage to wastewater treatment plants and directing stormwater into waterways. While the cost of the separate system was less than the combined, the cost to separate existing infrastructure proved too financially impossible to many cities. Hence, many cities maintained existing infrastructure while modernizing the new. The American Public Works Association (APWA) conducted a survey of CSSs in 1967 and found that there were more than 1,300 overflows nationwide, resulting in 1.2 trillion gallons of untreated sewer overflows annually. According to the US EPA, CSOs are a major threat to water quality and human health for 40 million people throughout the approximate 772 communities in the United States that operate combined sewer systems (see Figure 1). CSOs are responsible for gastrointestinal diseases, many beach closings, shellfishing restrictions and limitations of other recreational activities. With the transition from CSSs to Sanitary Sewer Systems (SSS), a new form of pollution emerged. As cities continued to grow the need for sewer capacity increased 2. However, pipe 1 United States Environmental Protection Agency. Combined sewer overflows. [Online] (accessed December 3, 2006). 6

15 Figure 1: Location of CSO communities throughout the United States Source: USEPA, capacity did not increase. Similar to CSO points, SSO points were engineered to relieve surcharge throughout the system. This surcharge can be a result of insufficient capacity throughout the system or rainfall dependent infiltration and inflow (RDII) caused by leaky, cracked pipes which allows water to seep through. US EPA estimates that between 23,000 and 75,000 SSOs occur each year, resulting in between 3 to 10 billion gallons of untreated overflows. The untreated sewage from these overflows can contaminate the water and backup into people s basements, causing property damage and threatening public health (EPA 2004). The Clean Water Act (CWA) requires communities to prevent the occurrence of SSO and CSO events. Failure to do so can result in fines and other penalties imposed by the government for violation of federal law. Cities that fail to comply with the law may enter into a consent decree requiring them to formulate a Long Term Control Plan (LTCP) 2 United States Environmental Protection Agency. Sanitary sewer overflows. [Online] (accessed December 5, 2006). 7

16 requiring: (a) Characterization, monitoring, and modeling of the combined sewer system, (b) Public participation, (c) Consideration of sensitive areas, (d) Evaluation of alternatives to meet CWA requirements using either the "presumption approach" or the "demonstration approach"; (e) Cost/performance considerations; (f) Operational plan, (g) Maximizing treatment at the existing Publicly Owned Treatment Works (POTW) treatment plant, (h) Implementation schedule, and (i) Post-construction compliance monitoring program 3. POSSIBLE SOLUTIONS Growth management tools such as urban growth boundaries (UGBs), urban service boundaries (USBs) and land preservation are several examples of how cities can limit development, in the hope of spurring higher density development and conserving open space. However for many cities around the country these solutions are not in place and they must instead try to deal with the consequences of decades of urban sprawl. Cities in this situation must address how to deal with high levels of impervious surfaces under current conditions. Traditionally, the approach to deal with stormwater has been to treat it as a waste product. This has led to expensive engineering projects which involve massive amounts of concrete, highly structured piping improvements and other mechanisms to channel runoff and its pollutants from impervious land into the water. However, newer, flexible, and more effective alternatives to manage stormwater and CSOs, known as green infrastructure, are currently being adopted in North America (Kloss and Calarusse 2006). 3 United States Environmental Protection Agency. Combined sewer overflows. [Online] (accessed December 3, 2006). 8

17 Green infrastructure such as rain gardens, green roofs, wetlands, porous pavement and increased vegetative cover throughout the urban environment, is an environmental strategy which addresses the source of stormwater and sewer overflow pollution while bringing with it other environmental benefits such as decreased energy consumption, reduction in the urban heat island effect, improved air quality, increased urban biodiversity and improved aesthetics. Cities across the United States such as Portland, Seattle, Chicago, Minneapolis and Boston, are beginning to introduce green infrastructure as a component of comprehensive stormwater management plans aimed at reducing runoff and sewer overflow events (Kloss and Calarusse 2006). In particular, greening the tops of buildings by replacing traditional asphalt cover with a green roof has become a growing trend around the world because it is an effective way to decrease levels of impervious surface in an urban environment without sacrificing developable land. In most cases, rooftops go unused and are simply wasted space. By expanding the current roof structure to include a high quality water proofing and root repellant system, a drainage system, filter cloth, a lightweight growing medium and plants, an impervious area is quite easily transformed into a pervious one, contributing many environmental benefits to surrounding areas. GREEN ROOF STUDY INSPIRATIONS Recently, there have been two major studies completed which quantify the benefits of green roofs. The first is a study performed in Toronto, Canada which quantifies the benefits of widespread implementation of green roofs and then ascribes a monetary benefit to the city. 9

18 The second is a study performed in Washington D.C. funded by the EPA which looks at the contribution that green roofs can make to reduce the rate and volume of stormwater runoff and consequent implications for decreasing the volume and frequency of sewer overflows. Toronto, Canada In 2005, the City of Toronto approached a team from Ryerson University to measure the cost and benefits of green roofs and develop a further understanding of the potential monetary savings to the municipality through their use (Banting et al. 2005). After performing a literature review to identify the benefits which would address immediate needs for the City of Toronto, the team then performed an extensive inventory of the types of buildings located throughout the city as well as their geographic distribution using Geographic Information Systems (GIS), then developed a model to compute the monetary value of the city-wide benefits. TABLE 4: CATEGORY OF BENEFIT AND PROJECTED SAVINGS Category of benefit Initial Annual Air quality $0 $2,500,000 Building energy $69,000,000 $21,500,000 Combined sewer overflows (CSOs) $46,000,000 $700,000 Stormwater $117,000,000 $0 Urban heat island effect $80,000,000 $12,300,000 Total $312,000,000 $37,000,000 Source: Banting et al The benefits were calculated based on the assumption that 100 percent of available green roof area be used (Banting et al. 2005). The available area included flat sloped roofs greater than 350 square meters or roughly 3,767 square feet. Assuming 75 percent of that area would be greened, the total area was determined to be 50 million square meters or roughly 10

19 538,195,521 square feet. Under this scenario, the study predicted that the city would save an initial $312,000,000 and a recurrent annual savings of $37,000,000 (see Table 4). Finding that the monetary, environmental and aesthetic benefits were significant, the researchers assigned a monetary value to stormwater management, combined sewer overflows, air quality, building energy and the urban heat island effect. Researchers noted that although benefits such as aesthetic improvements of the urban landscape, increase in property values, benefits of increased amenity spaces, use for food production and increased biodiversity were not quantified, they ought to be taken into consideration (Banting et al. 2005). Washington, D.C. The District of Columbia is a 61.4 square mile area, home to 572,000 residents. Currently, the district does not meet water quality standards for the Anacostia, Potomac and Rock Creek Rivers and must invest in a $1.9 billion Long Term Control Plan to manage its combined sewer overflows (Deutsch et al 2006). This sparked interest in the potential contribution that green infrastructure such as green roofs and increased tree canopy could make towards water quality improvements for the area. Using a grant administered by the US EPA, the city set out to quantify the contribution green roofs could make toward improving air and water quality at different green roof coverage scenarios, assess the benefits under the different coverage scenarios, create a basis for a Green Roof Vision for DC, and highlight areas for further study. The existing area available for green roofs was determined by considering existing and proposed development greater than or equal to 10,000 square feet, (generally large 11

20 commercial, industrial or government buildings.) This criteria was selected due to the relative ease of implementation associated with larger construction. The projected green roof coverage was assumed to cover 80 percent of the building s footprint. This percentage was designed to allot for standard features found on rooftops such as HVAC systems and roof access. According to the criteria, 75 million square feet, 29 percent of the total building area, were determined to be eligible for green roof installation. The city then came up with six different scenarios to model, ranging from 100 percent coverage on available structures to 20 percent. For each scenario, it was implicit that 80 percent of green roofs would be extensive (soil depth of approximately 3 inches) and 20 percent would be intensive (soil depth of approximately 6 inches or more.) Each of the six scenarios then quantified the benefit on water quality using a model built specifically for the District of Columbia s sewer system based on reduction in stormwater runoff and combined sewer overflow events, assuming 2 inches of rainwater storage for extensive roofs and 4 inches for intensive. The decrease in CSOs was determined by comparing the increased storage capacity allotted by the green roofs to the volume of overflows at each overflow point. On a per roof basis, the stormwater model predicted that an extensive green roof could potentially reduce roof runoff volumes by roughly 65 percent and an intensive roof by 85 percent. Using the 80/20 ratio previously identified, the district s runoff was predicted to decrease by up to 69 percent. Cumulatively, at the minimum coverage scenario (20 percent of available buildings) 23 million gallons were estimated to be added to the city s storage capacity, equaling 297 million gallons of precipitation that would be captured by the roof 12

21 instead of entering the system. At 100 percent coverage the city s storage capacity increased to 113 million gallons, totaling 1,485 million gallons of precipitation prevented from entering the system. Next, the city modeled each of the six scenarios to determine how this increased stormwater retention could potentially decrease sewer overflow events in various watersheds. Depending on the watershed in question, the number of rainfall events that were shown to trigger sewer overflows decreased by 2-10 events under the 20 percent coverage scenario and by 7-16 events under the 100 percent coverage scenario, thus significantly decreasing levels of bacteria, depletion of dissolved oxygen and rate of stream bank erosion. The number of CSO discharges triggered by annual rain events was shown to decrease by 13 percent with a volume decrease of 75 million gallons under the 20 percent coverage scenario, while the 100 percent coverage scenario decreased the number of events by 28 percent with a volume decrease of 334 million gallons. The DC Water and Sewer Authority (WASA) acknowledged that the stormwater benefits are considerable for both the CSO and Municipal Separate Storm Sewer Systems (MS4) areas with regard to meeting water quality, and that they could potentially provide significant savings in capital investments in the LTCP. The Department of Health agreed with the sewer district s sentiments and has stated that the findings of this analysis will serve as the basis for creating additional incentives and regulations to support wide-scale implementation of green roofs. 13

22 The final part of the study was to determine a target green roof coverage based on the benefits associated with the different scenarios, green roof coverage in other cities and ease of local implementation. Findings showed that the 20 percent coverage scenario, providing the same benefit on air quality as 19,500 trees, was the preferred scenario. Assuming the 20 percent scenario, with 20 percent of existing development retrofitted and 80 percent of proposed development greened, 21.7 million square feet of rooftops would have a green roof. Under this scenario, 30 million gallons would be added to the city s stormwater storage capacity and 430 million gallons of stormwater could be stored over the course of the year, decreasing citywide runoff by 1.7 percent and CSO discharges by 15 percent. Based on these findings, the study recommended an objective of 20 percent coverage (20 million square feet) in 20 years. As a result of the study, the recommendation will be followed by an implementation strategy, coordinating efforts with city and federal agencies and stakeholders. PROBLEM STATEMENT Ohio has about 87 overflow communities ranging from small rural communities to large metropolitan areas such as Cincinnati (MSD 2006b). Cincinnati s first sewers were built in the 1800s to carry rainwater away from homes, businesses and streets. At that time the city did not have sewage treatment or indoor plumbing. When indoor plumbing began to surface, home and business owners hooked their sewage lines to the existing storm sewers, combining stormwater and raw sewage into the same pipe. These pipes emptied directly into the nearest river, stream or ditch, which eventually transported the untreated sewage to 14

23 the Ohio River. Although the city s first wastewater treatment plants began operating during the post-war boom of the 1950s, the combined sewer systems remained in place, and overflow pipes were installed to relieve the system when it reached capacity. Today, the Metropolitan Sewer District of Greater Cincinnati (MSD) is responsible for wastewater collection, treatment and disposal of over 200 million gallons of wastewater each day for 800,000 customers throughout 44 municipalities and townships in Hamilton County, Ohio 4. Out of the 3,000 miles of sewers, approximately one-third of MSD s sewer s are combined sewers. The remaining two-thirds are sanitary sewers. Currently, the system contains approximately 80 sanitary overflow points and 212 combined overflow points. This translates to over 105 annual sewer overflow occurrences, totaling over 14.4 billion gallons of sewer overflows (MSD 2006b). On December 3, 2003, the Justice Department, the United States Environmental Protection Agency, the State of Ohio and the Ohio River Valley Water Sanitation Commission (ORSANCO) announced a final settlement with the Board of Commissioners of Hamilton County and the City of Cincinnati that will end violations of the Clean Water Act and address raw sewage backing up into residents basements. Under this settlement and a previous partial settlement from February 2002, MSD is expected to spend more than a billion dollars to bring its aging sewer system into compliance with the Clean Water Act and other applicable laws and regulations. 4 Metropolitan Sewer District of Greater Cincinnati. (accessed December 3, 2006). 15

24 In June 2006, MSD submitted a draft Long Term Control Plan projected to spend $1.99 billion over the next 23 year, which focuses on traditional engineering solutions including system separation and increased pipe capacity (MSD 2006a). However, MSD s plan was engineered with the anticipation that ORSANCO an interstate agency responsible for the facilitation of water quality standards for the Ohio River would lower water quality standards by the end of June. Due to the significant amount of adverse public comment, the Commission tabled the June vote for October During the October meeting, the Wet Weather Standards Committee recommended that the Commission hold off on the majority of the wet weather components previously proposed, until recommended changes were made. The pollution control standards are not scheduled to come under review for another three years 5. MSD claims that if they continue to increase sewer utility fees to the point that it would take to meet current water quality standards, it will create an economic crisis, driving people out of the county. As the plan is currently formulated, rates are projected to increase to 2 percent of the city s median household income. The city currently has a poverty level of 21.9 percent (U.S. Census 2000) implying that projected rate increases may create an immense economic burden on many families, making it difficult for them to acquire basic needs and services. However, EPA guidelines state that rates should be determined based on the entire service area, which far exceeds the city boundary. To date, the plan has not yet been approved by the EPA and the discrepancy concerning rate structure has yet to be seen. 5 Ohio River Valley Water Sanitation Commission. Pollution control standards. (accessed December 3, 2006). 16

25 As discussed earlier, solutions to deal with runoff and sewer overflows have always been limited to traditional engineering solutions. These solutions, including the expansion of stormwater or sewage pipes or the construction of treatment facilities, are extremely expensive (Snodgrass and Snodgrass 2006). However, recent studies like the ones performed in Toronto and D.C. have shown that there are more sustainable alternatives to manage stormwater such as detention ponds, rain gardens, vegetated swales and green roofs. Due to the fact that roofs in particular comprise about percent of impervious cover in urban areas, they can play an important role in minimizing the amount of stormwater which leaves the site (Dunnett and Kingsbury 2004). As highlighted earlier, green roofs in particular can capture up to 90 percent of the stormwater that falls on its surface (Snodgrass and Snodgrass 2006). The percent that does runoff does so at a much slower rate, thereby increasing ground absorption and decreasing the impact of the first flush. In general, a 4-inch extensive green roof stores approximately 60 percent of annual rainfall while even a 2-inch system can retain as much as 50 percent of a ½-inch rainfall (Snodgrass and Snodgrass 2006). By incorporating green roof infrastructure in Cincinnati, costs to manage stormwater and minimize sewer overflows could perhaps be significantly decreased while providing a myriad of environmental, aesthetic and economic benefits that traditional improvements do not entail. RESEARCH QUESTION There are many large roofs located at the University of Cincinnati. These roofs account for a large portion of impermeable surfaces at the university and contribute a large amount of stormwater runoff entering the Mill Creek sewershed. 17

26 What is the potential contribution of green roofs to reduce stormwater runoff at the University of Cincinnati if all eligible institutional buildings are greened on both East and West campus? Do these findings support the proposition that a local policy for green roofs would be worthwhile for the city? OBJECTIVES The first objective of this study is to determine the existing runoff for the university. The second objective is to take an inventory of the buildings located at the university and of those that meet the criteria for green roofs, define the total area and subsequent reduction in stormwater runoff. The third objective is to compare the two scenarios and quantify reduction of stormwater runoff. The fourth objective is to make the study available to policy makers such as the EPA, OEPA, the city and the county. STUDY OUTLINE Chapter one introduces the issues surrounding impermeability and stormwater and defines the problem statement, research questions and objectives. Chapter two presents the inspiration for the study and outlines the methodology used to carry out the study. Chapter three reviews the literature relevant to green roof infrastructure. Chapter four provides a background on hydrology, hydrologic models and supplies a critique of TR-55. Chapter five walks the reader through the methodology and presents an analysis of the findings. Chapter six concludes the study by underlining the study s findings and contribution to the literature on green roofs and provides recommendations for future study, funding mechanisms and policy. 18

27 2 chapter METHODOLOGY This chapter covers the methodology used to conduct this study. Time constraints and the scope of this study did not permit an analysis to be performed at the same scale as the studies carried out in Washington D.C. or Toronto. In an effort to replicate these efforts on a significantly smaller scale, it was decided that a small portion of Cincinnati would be looked at in an effort to imply how green roofs could reduce stormwater runoff under local conditions. This study was thus synthesized and completed in the following eight tasks. Task 1: Literature Review An extensive literature review was performed prior to data collection. The first part of the literature review found in chapter three, examines green roofs. It begins by defining green roofs and elaborating on the components of a green roof and how they function collectively. It continues to review published literature pertaining to the economic, environmental, and aesthetic benefits of green roof technology. Next, the literature review highlights various economic and architectural considerations that should be addressed. The chapter concludes by providing a brief but informative history on green roofs including a review of the international status of green roofs. 19

28 Chapter four, the second portion of the literature review discusses the history and evolution of the science of hydrology and examines current hydrologic models with particular attention to those used to measure stormwater runoff and how planners need to incorporate stormwater models into land use planning. Due to the time constraints of this project (which is designed to be carried out and completed in less than nine months time) shadowed by the researcher s limited familiarity with hydrologic models, Technical Release 55 (TR-55) is used to analyze the current and proposed situation, because it is a fairly simple process as well as the most widely used and reliable model to predict the runoff-rainfall relationship for small urbanized watersheds (USDA 1986). Subsequently, a detailed explanation about hydrologic model TR-55 is provided, including a review of its strengths and limitations. Although the model provides the steps to calculate peak discharge, this step was excluded from the study for the reasons outlined above. Task 2: Identification of Case Study Site and Determination of Boundary For the purpose of this study, it was decided that a small fraction of MSD s service area, which includes the City of Cincinnati as well as Hamilton County, would be chosen, and structures which met the proposed green roof criteria (outlined in Task 3) would be hypothetically greened. A number of criteria were established to find a workable case study site for this project. The criteria for selection of a case study site was as follows: To be located within Hamilton County and within the political boundaries of the City of Cincinnati To be located within a single sewershed 20

29 To be located within an area which has a large amount of industrial, institutional or commercial development To be located within an area where such development is substantial in area, generally indicating a sizeable amount of large, flat rooftops After looking at a number of different options, the University of Cincinnati was chosen as the case study site for this project. The University of Cincinnati is located within the boundary of the City of Cincinnati in the upper region of the Mill Creek sewershed. Due to the high propensity of institutional buildings associated with the university, the vast majority of development met the necessary size requirements and information was easy to obtain. The boundary was determined from university published maps. Task 3: Quantification of Stormwater Runoff at University of Cincinnati Tasks three through eight are covered in chapter five, the case study section of the research. The third task was performed using data available from the Cincinnati Area Geographic Information System (CAGIS). The analysis was performed by converting aerial photography into polygon shapefiles and then quantifying stormwater runoff using TR-55. The first step in executing the methodology required for TR-55 was to determine the existing land cover on campus. Land cover was calculated using the following steps: Open the Mr. Sid aerial photo in CAGIS Create new shapefile in ArcCatalog to delineate the case study boundary and trace the boundary over the aerial photo using campus maps to determine the boundaries of each campus 21

30 Create two additional shapefiles in ArcCatalog, one for open space and one for buildings Modify both shapefiles using the editor toolbar, tracing each land cover as a new polygon Using the dissolve function in ArcCatalog, combine the two shapefiles into one Create a new shapefile in ArcCatalog and clip it using the boundary shapefile, then minus the combined shapefile to give the third shapefile for pavement, roads and parking lots Calculate the areas for all three shapefiles, using the function in ArcMap Export three tables to an excel file Calculate total area for each of the three land use categories The second step in executing the methodology required for TR-55 was to plug soil type and land cover into a worksheet to determine the composite CN. The composite CN was calculated using the following steps: Assign a CN to each land use (open space or impervious cover, i.e. buildings, roads, parking lots) according to soil type found in Table 2-2a in TR-55 manual Multiply the total area of each land use by the according CN Sum the two areas Sum the two products Divide the total product by the total area to get the composite CN The third step was to determine the existing runoff for a 2-year storm event. According to the Rainfall frequency atlas of the Midwest a 2-year rainfall in Cincinnati is 2.86 (Huff and Angel 22

31 1992). Existing runoff was calculated by using rainfall (2.86 inches) and CN (92) with Table 2-1 in the TR-55 manual. The fourth step was to convert runoff depth into total runoff volume for the entire case study. To convert the depth in inches to feet runoff depth was divided by twelve. This was then multiplied by the total area of the case study in square feet to obtain the volume in cubic feet. Task 4: Identification of Green Roof Criteria To qualify for a green roof a number of criteria based on the available literature were established to ensure that the study would project workable results. The criteria below has been set to select buildings with roofs that are considered by the green roof industry as (a) easily greenable and b) potentially cost-effective. Because the green roof industry is still expanding in the United States, it is generally agreed not to be cost-effective to green smaller structures. Additionally, although the lifespan of the green roof offsets the overall higher cost in comparison with that of a traditional roof, the upfront cost can be more than double, making it difficult for residents to afford. It is important to note that for the purpose of this study the projected green roof coverage would be comprised solely by extensive green roofs with a 3-inch growing medium because these systems do not generally require additional structural support or irrigation systems. For these reasons the criteria for green roofable structures was set according to the following: 23

32 Must be non-residential development (owners of non-residential development may more easily be able to offset the upfront cost of a green roof) The roof must be greater than or equal to 500 square feet with a slope less than 30 for ease of implementation Assume that 80 percent of the footprint will be greened (allowing for typical rooftop structures such as heating, ventilating and air conditioning (HVAC) systems, railings and access points for maintenance, etc.) Task 5: Green Roof Building Inventory The fifth task of the study was to perform an inventory of existing structures in the study area. Once the university buildings were identified, those that qualify for green roofs were chosen based on the predefined criteria needed to support green roof infrastructure. The green roof building inventory was completed using the following steps: Open the Mr. Sid aerial photo in ArcMap Create a copy of the buildings shapefile in ArcCatalog and rename it green roofs Import the green roof shapefile into ArcMap Use the select by attribute function in ArcMap to identify structures under 500 square feet Open the editor toolbar and select modify feature to remove the structure(s) Set the building layer at 70 percent transparency and remove the polygons determined to have pitched roofs Save the edits Open the attribute table and export the table in excel format 24

33 Task 6: Determination of a Curve Number for Extensive Green Roofs In TR-55, mass rainfall is converted to mass runoff by the use of a curve number (CN) an empirical parameter used to predict runoff and infiltration (USDA 1986). Determination of a CN is based on soils and cover conditions which the model represents as hydrologic soil, cover type, treatment and hydrologic condition (USDA 1986, 1-1). However, a green roof is not included in the land use classification system in TR-55 and a CN was not provided. Because green roofs vary in depth and retention capabilities, it is unrealistic to assume that they have the same CN as open space for example. Therefore, it was necessary to determine what the curve number for an extensive green roof was using a 2-year storm event. The determination of a curve number for a 3-inch extensive green roof was determined according to a paper written by Charlie Miller founder and president of Roofscapes, Inc. which directly relates the intensity of the storm event to the green roof s retention capacity. For this study, a CN of 66 was applied assuming a growing depth of 3 inches. Task 7: Quantifying Stormwater Runoff after Projected Green Roof Coverage Once the total area of green roofable buildings was identified, a green roof build-out analysis was performed. To determine the change in runoff after suitable roofs had been greened, a second analysis was performed. An extensive green roof was hypothetically placed on the structures that met the predefined green roof criteria and runoff volume was recalculated. 25

34 To quantify stormwater runoff with the projected green roof coverage the same methodology was used as outlined above in Task three to quantify existing runoff. The composite CN was calculated using the following steps: Create a new category for green roof coverage Assign a CN to each land use according to soil type found in Table 2-2a in TR- 55 manual (except for the green roof coverage which does not take HSG into account, only the predetermined CN of 66) Multiply the total area of each land use by the according CN Sum the three areas Sum the three products Divide the total product by the total area to get the composite CN The remainder of the worksheet was carried out according to the steps outlined in Task three. Task 8: Analysis and Findings This portion of the study looks at the current level of impervious area and stormwater runoff and compares it to the projected green roof scenario. Existing conditions are presented, followed by projected conditions. Findings were then examined in greater depth to imply what the projected scenario could mean for the city and county in terms of reducing stormwater runoff and potentially impacting the rate and frequency of sewer overflows. A total cost estimate to green the case study is approximated based on available literature according to cost per square footage. The chapter concludes by noting the additional benefits reflected, but not quantified, in the study. 26

35 Task 9: Recommendations and Conclusions Chapter six discusses the implications this study has for long term land use planning and environmental mitigation techniques in Cincinnati and discusses the importance for planners to employ these techniques. The chapter concludes by highlighting the study s conclusions and consequent recommendations. Suggestions for further study are provided followed by recommendations for implementing a green roof policy. 27

36 3 chapter GARDENS IN THE SKY This chapter examines green roofs. It begins by defining green roofs and elaborating on the components of a green roof and how they function collectively. It continues to review published literature pertaining to the economic, environmental, and aesthetic benefits of green roof technology. Next, the literature review highlights various economic and architectural considerations that should be addressed. The chapter concludes by providing a brief but informative history on green roofs including a review of their international status. DEFINITION The term green roof refers to a system which is comprised of several layered components which function collectively, typically including a deck, waterproofing, insulation, protection and storage, drainage, root permeable layer, growing medium, and vegetation (Snodgrass and Snodgrass 2006). Green roofs may also be referred to as eco-roofs, vegetated roofs or roof gardens. TYPES There are two types of green roofs, intensive and extensive (see Table 5). Intensive green roofs presume that the area to be greened will be used for aesthetic or recreational purposes similar to a traditional garden. Soil depth is generally at least 6 inches deep and often composed of a lightweight growing media known as substrate. Because of the soil depth, 28

37 intensive green roofs can support diverse vegetative communities and must be treated individually. Intensive green roofs have the ability to house grass and flowers and can even include trees and park benches. Consequently, the roof requires a high level of maintenance. Often times, intensive roofs require additional structural support as well as intricate irrigation systems. TABLE 5: COMPARISON OF EXTENSIVE AND INTENSIVE GREEN ROOF SYSTEMS Extensive Intensive Thin soil Deep soil Little or no irrigation Irrigation required Description Stressful conditions for plants More favorable conditions for plants Advantages Disadvantages Lightweight Suitable for large areas Suitable for roofs with 0-30 slope Low maintenance Often no need for irrigation and drainage systems Relatively little technical expertise needed Often suitable for retrofit projects Can leave vegetation to develop spontaneously Relatively inexpensive Looks more natural Easier for planning authority to demand green roofs be a condition of planning approvals More limited choice of plants Usually no access for recreation and other uses Unattractive to some, especially in winter Greater diversity of plants and habitats Good insulation properties Can simulate a wildlife garden on the ground Can be made very attractive Often visually accessible Diverse utilization of roof (i.e., for recreation, growing food, as open space) Greater weight on roof Need for irrigation and drainage systems, hence greater need for energy, water, materials, etc. Higher cost More complex systems and expertise required Source: Peck et al Extensive green roofs on the other hand, are not presumed to have the same aesthetic or recreational purposes as intensive green roofs and are often not intended to be seen. They usually have a soil depth between 3-4 inches which dramatically narrows the variety of 29

38 species the system can support. Because of the soil depth the additional weight is significantly less for extensive green roofs. Semi-extensive roofs employ the same lightweight technology and growing media as extensive roofs, but have a slightly increased soil depth which enables a slightly wider variety of species selection. COMPONENTS The initial and most important layer on a green roof is its decking, which can be comprised of concrete, wood, metal, plastic, gypsum, or composite. Plywood, the most commonly used material for residential construction in the United States, typically does not have enough structural support to support a green roof on its own. Consequently, additional engineering is often required to support a green roof on structures with plywood decks (Snodgrass and Snodgrass 2006). The subsequent layer is a waterproofing layer. Just as with any roof construction, the waterproofing membrane must be completely waterproof since waterproofing is the main function of any roof (Velasquez 2005). However, should a leak be detected after construction of the green roof is complete, repair will most likely entail removal of the entire system including vegetation which can be extremely costly (Snodgrass and Snodgrass 2006). To avoid this issue, it is imperative that adequate flood testing be done to ensure the material is impervious and will be able to provide water tightness for a minimum of 24 hours (Velasquez 2005). Alternatives for waterproofing materials include modified bitumen, rubberized asphalt, polyvinyl chloride (PVC), ethylene propylene diene monomer (EPDM, a water resistant rubber), and thermoplastic polyolefin (TPO). 30

39 If organic material such as wood, bitumen or asphalt is used it is crucial that an additional root protection barrier is installed to separate the plant layer and membrane from possible root intrusion and potential puncturing of the membrane (Dunnett and Kingsbury 2004). Certain systems have been set upon metal or plastic base plates to completely separate the vegetative layer from the membrane. However, more commonly, root barriers are comprised of PVC which range in thickness from 0.8 mm (0.03 in) to 1.0 mm (0.04 in). This has raised certain environmental concerns due to the ecological implications of using PVC. However, the argument in favor of PVC points to its recyclability, elimination of further costs, and minimization of leaks. The drainage layer follows and is particularly important for a green roof for a variety of reasons (Dunnett and Kingsbury 2004) (see Table 6 for main materials used for drainage). Drainage layers in green roofs perform several functions (Miller 2003). First, they sustain the overlying growth media in a drained condition, which prevents overwatering of the vegetation. Secondly, they provide the primary means for releasing rain related runoff which includes the elimination of surface flow and minimization of flow that seeps into the growth media. Lastly, they retain water following irrigation or rainfall events. Some green roofs may also require a root-permeable filter layer (Dunnett and Kingsbury 2004). This is generally a cloth or type of fabric made out of semi-permeable polypropylene fabric. It allows separation between the growing medium and the drainage layer to avoid blockage. It can also decrease the growing medium s mobility thereby reducing the chance of washing down the drain. If it is believed to be necessary, the root-permeable filter layer can adhere to the drainage layer or act as a separate layer. 31

40 TABLE 6: THREE MAIN TYPES OF DRAINAGE MATERIALS Materials Comments Granular materials (gravel, stone chips, broken clay tiles, clinker, scoria (lava rock), pumice, expanded shale, expanded clay granules) and hospitable environment. Porous mats Lightweight plastic or polystylere drainage modules Contain large amounts of air/pore space to allow for water to move in from the vegetation and substrate and promote drainage. The most low tech and simple. Good means to increase the amount of root space. Provides a well aerated Mats which act like sponges, absorbing water into their structure. Can be constructed from many materials including clothing and car seats. Some mats may be too absorbent, sucking moisture from the growing medium and impacting plant growth. Vary tremendously in design and appearance. Some can retain water and others cannot. Some can be filled with granular material. They are rigid enough to support the growing medium and vegetation, isolating these layers from the surface. Offer permanent free-flowing lightweight drainage. Some may permit subsurface irrigation. Source: Dunnett and Kingsbury 2004 Following either the drainage layer or the root permeable layer is the growing medium or substrate layer which must be lightweight and able to store water efficiently while maintaining free-draining properties (Dunnett and Kingsbury 2004) (see Table 7 for materials used as substrate). This layer is vital to the lasting success of the green roof (Velasquez 2005). Contrary to ground-level gardens which tend to thrive off topsoil, green roofs use a higher rate of inorganic materials to organic because of the properties mentioned above, and may exclude organic material entirely. The final and most vital and distinct layer of the green roof is of course the vegetative layer (Velasquez 2005). While tried and true construction materials exist for the previous layers, the vegetation chosen for a green roof must be selected according to the site s climate (Dunnett and Kingsbury 2004). To select plants and vegetation that will survive in a harsh rooftop environment, an extensive knowledge of local conditions is required and expertise from a landscape architect should be sought. Further considerations regarding species selection is covered later in the chapter. 32

41 TABLE 7: SOME MATERIALS USED AS A BASIS FOR GREEN ROOF SUBSTRATES Materials Comments Natural Mineral Sand Fine texture can result in lack of pore space and problems of saturation of the substrate if drainage is poor. Conversely, coarse sands can be so free-draining as to require constant irrigation. Lava (scoria) and pumice Lightweight and valuable if locally available. Gravel Relatively heavy. Artificial minerals Perlite Particles tend to collapse over time. Vermiculite Very lightweight, but has no water- or nutrient-holding capacity Light expanded clay granules (LECA) Expanded Shale Rockwool Recycled or waste materials Crushed clay brick or tiles, brick rubble Crushed concrete Subsoil and again may disintegrate over time. Lightweight, produce large amounts of pore space because of their size, and absorb water because of their porous nature. Very lightweight but energy-intensive production and no nutrientholding capacity. Stable and uniform, some nutrient and moisture retention. Brick rubble may contain mortar and cement, which will raise the ph of the substrate. Limited moisture retention and nutrient availability, alkaline. However, cheap and available in quantity as a demotion material. Heavy, low fertility, readily available as by-product of construction. Source: Dunnett and Kingsbury 2004 BENEFITS There are many claims regarding the benefits of green roofs. However, while there is no doubt that green roofs have many benefits over traditional rooftops the peer reviewed literature which quantifies the degree of those benefits is limited (Dunnett and Kingsbury 2004). Until recently, there has been little evidence to support the benefits of green roofs in North America although evidence from Germany has been available since the 1970s. However, since the mid-1990s research undertaken in North America has continued to increase and many of the previously inaccessible German publications have been translated into English. 33

42 This research, performed in North America and overseas, documents that the benefits of green roofs function at a variety of different scales. Certain benefits are more pronounced if large numbers of green roofs are implemented (Dunnett and Kingsbury 2004). These benefits can be seen at the neighborhood level, city scale or both. Other benefits may be seen at an individual scale. Peck and Kuhn (2000) further divide the advantages into private and public sector benefits which can be used to gage the interest of both. Private benefits include savings in energy costs, extension of the life of the roof and aesthetic enhancement to property. These benefits are likely to promote the use of green roofs by giving economic or personal advantages to owners and developers who choose to implement them. Public benefits include stormwater management, urban climate change mitigation and increase in habitat and biodiversity. These benefits are more likely to foster the propagation of planning regulations by local and city governments to encourage the implementation of green roofs. Dunnett and Kingsbury (2004) divide the benefits of green roofs into three categories: Aesthetic, environmental, and economic. The aesthetic benefits of green roofs cited by the authors include the amenity value of green roofs, food production and enhanced aesthetic quality. Economic benefits consist of an increased roof life, improved insulation and energy efficiency, and strengthened public relations. Although, Dunnett and Kingsbury do not place job creation under this umbrella, research suggests that job creation figures into the economic benefits as well (Peck at al. 1999). Finally, the environmental benefits are increased habitat and biodiversity, stormwater management, enhancement of air quality, decrease in 34

43 the urban heat island effect, mitigation of noise pollution and fire prevention. As the authors point out, there is often crossover between the three categories. Amenity Value Rooftops are a highly underutilized resource in urban and suburban settings (Dunnett and Kingsbury 2004). Generally, with the exception of HVAC systems and minimal maintenance access, rooftops are left barren and unused. However, intensive green roofs can play an important role in the urban environment. Recreational activities which can transpire atop intensive green roofs include barbecuing, eating, drinking, dog exercising, golf, bowling, sun bathing, reading and relaxing while offering views of the cityscape (Hutchinson et al. 2003). One green roof in Minneapolis even acts as a theatre, projecting movies against a neighboring wall. Even when a green roof is not designed for intensive recreational use, it will still suffice for limited, individual recreation (Dunnett and Kingsbury 2006). Urbanites in high-density, compact towns may find it the only way to personalize the exterior of their homes. With an enormous pressure on green space at ground level, green roofs are a major untapped resource to provide people with the prized amenity of recreational green space amidst a grey environment. In the same vein, the amenity aspect of a green roof offers a number of economic benefits as well. Green roofs enhance the value of the structure they inhabit, allow builders and hotel owners to charge higher rates for rental space because of the attraction and help business 35

44 owners to attract customers (Osmundson 1999). For these reasons, green roofs can be an important addition to urban renewal areas. Food Production Many social, economic and environmental concerns have risen concerning the quality of food, how it is produced and the distance that it travels (Dunnett and Kingsbury 2004). Transporting food over long distances has many implications in terms of energy, pollution, and negated nutritional value. Roof surfaces provide the opportunity to grow local, healthy, organic food within the urban environment. While community gardens have proved to yield inner-city residents with outstanding amounts of locally grown produce, they still require open green space which can be hard to preserve in light of development pressures felt by many cities. Green roofs can offer the same benefit of community gardens without feeling similar pressures from developers and localities. In countries such as Haiti, Colombia, Thailand and Russia, rooftops and balconies have been used to produce a variety of marketable products from fruits, vegetables and plants (Dunnett and Kingsbury 2004). In Vancouver, Canada, the Fairmont Hotel has a roof garden which spans 2,098 square feet, providing the hotel with all the herbs used in the hotel which saves the establishment approximately $25,000-$35,000 (Canadian) annually. Still, it is important to take into account load bearing restrictions when considering a roof garden (Velasquez 2005). Intensive green roofs have the ability to yield a greater crop than extensive roofs simply due to the additional weight that their carrying capacity allots for (Dunnett and Kingsbury 2004). However, although extensive roofs cannot bear the same 36

45 use intensity as intensive roofs, extensive roofs can still be used for food production by replacing ornamental plants with food-producing plants. Dunnett and Kingsbury (2004) note that green roof space can even be leased for food production or other amenity benefits, which has the potential to open up untapped commercial possibilities. Aesthetic Quality While many urbanites spend their time in the city looking up at the cityscape, others may spend the majority of their time looking across urban roofscapes from office downtown settings in high-rise buildings especially in larger, denser cities. These large, flat, dark rooftops are extremely unattractive. Green roofs transform these otherwise black surfaces into lush green settings, bringing nature into the roofscape. Even when these roofs are inaccessible, yet visible, they provide a beneficial therapeutic effect including stress reduction, lowering of blood pressure, relief of muscle tension, and increase in positive feelings (Dunnett and Kingsbury 2004). In fact, studies show that worker productivity in green buildings is significantly higher than in traditional type buildings. Because of the therapeutic effects, properties which boast views of a green roof may also experience financial gains. Increased Roof Life One of the greatest selling points of green roofs is their increased life span over that of a conventional roof. An extensive green roof should last years, while a conventional roof lasts only years before widespread repairs are needed (Dunnett and Kingsbury 2004). The major reason for the increase in roof life is due to the protective qualities of the green roof and thereby absence of exposure to natural elements including temperature 37

46 fluctuations and precipitation. Exposure to heat can accelerate the aging in bituminous material which reduces its durability while ultraviolet radiation can change its mechanical properties accelerating deterioration (Dunnett and Kingsbury 2004). Because the roof membrane of a green roof is generally unexposed it receives a significantly lower level of heat and radiation than traditional exposed roofs (Liu and Baskaran 2003). Table 8 below shows the results of a study conducted in Toronto, Canada where the membrane of a greened roof and an ungreened roof were measured over the course of 660 days. Notably, the ungreened membrane was light grey in color. Had it been black, as many are, the temperatures would have likely been even higher. TABLE 8: STATISTICS OF TEMPERATURES ON AN UNGREENED AND GREENED ROOF IN TORONTO, CANADA OVER A PERIOD OF 660 DAYS Ungreened roof Greened roof Ambient Temp. > # of days % of days # of days % of days # of days % of days 30 C/86 F C/104 F C/122 F C/140 F C/158 F Source: Liu and Baskaran 2003 Temperature fluctuation which is characterized by regular contracting and swelling is wearing on roof materials and eventually results in disintegration, cracking, delamination and splitting (Dunnett and Kingsbury 2004). Green roofs reduce the wear on the membrane by moderating temperature fluctuation (Liu and Baskaran 2003). German researchers found that diurnal temperature variations were reduced by up to 94 percent averaging a decrease of 54 F in temperature fluctuations between day and night (Dunnett and Kingsbury 2004). 38

47 However, the degree of temperature diminution is largely dependent on species selection (Dunnett and Kingsbury 2004). Research shows that a covering that features a wide variety of grasses and forbs (weeds) is able to achieve a significant reduction in temperature compared to a conventional roof, while more homogeneous communities achieve a significantly lower reduction. Another reason for the increased roof life of a green roof is its preventative nature. One of the downfalls of a conventional roof is their tendency to pool stormwater, rather than allow it to runoff, which increases the likeliness of exploiting any weakness in the surface (Dunnett and Kingsbury 2004). Green roofs decrease stormwater s ability to pool by storing most of the water in the roof s substrate and vegetation. Finally, green roofs act as a protective layer shielding the membrane from human traffic. Oftentimes maintenance crews need to work atop the roof to access HVAC systems. Substrate and vegetation provide a protective layer between traffic and the roof. Additionally, the build-up of organic matter, one of the main culprits of drainage system failure for traditional systems, can be incorporated into the green roof s substrate as humus. Insulation and Energy Efficiency Several properties of green roofs contribute to their thermal characteristics including direct shading of the roof, evaporative cooling from plants and growing medium, additional insulation from both the plants and growing medium, and the thermal mass effects of the growing medium (Liu and Baskaran 2003). Green roofs can decrease heat gain through shading, insulation, evapotranspiration and thermal mass. Yet green roofs only reduce heat 39

48 loss through insulation and diminished radiation heat losses. Therefore, green roofs are more effective in keeping structures cool during spring and summer than preventing heat loss in the fall and winter. There are several key factors that play into the degree of a green roof s thermal performance. If the vegetation is not primarily evergreen, then the insulating and evaporative function may be eliminated while climactic fluctuations such as prolonged freezing conditions or snow cover can potentially eliminate energy benefits on extensive roofs with little vegetation cover (Dunnett and Kingsbury 2004). The insulating effect of green roofs in reducing energy cost for buildings on an individual scale represents one of the strongest economic arguments for wide spread implementation (Dunnett and Kingsbury 2004). In fact, research has demonstrated that improving roof albedo (level of reflectivity) provides the greatest energy savings than other efforts to lower ambient temperatures (Wong 2005). However, when used as an economic incentive, experts stress the importance of collecting data on energy performance under a variety of climactic situations to clearly demonstrate the direct economic benefit. Strengthened Public Relations The United States Green Building Council (USGBC) is an organization dedicated to promote sustainable development through the environmental performance of buildings. In recent years, Leadership in Energy and Environmental Design (LEED), a national 69-point system set by the USGBC to rate the environmental performance of a building, has begun to get widespread national coverage (Earth Pledge 2005). The implementation of a green roof 40

49 which covers at least 50 percent of the building results in the attainment of one point for heat island reduction and an additional point for stormwater management (Dunnett and Kingsbury 2004). LEED certification is a good way to promote positive public relations, although implementation of a green roof in particular is a tangible, visible way to convey a sound environmental attitude to the public. Job Creation There has been some literature which points to the positive impact that the creation of a green roof market can have on the workplace. The creation of a green roof industry has been noted to affect the market for suppliers of green roof materials as well as other professionals who may be involved in the overall process such as landscape architects, architects, roofers, buildings, gardeners and engineers (Peck et al. 1999). Green roofs have also been linked to a potential growth in jobs because of their ability to produce food. Habitat and biodiversity The habitat of a rooftop is very similar to that of seasonally dry environments with shallow soil. These environments should respectively be used as models for green roofs (Dunnett and Kingsbury 2004). Interestingly, these types of environments happen to be at risk for a number of reasons in both urban and rural areas. Researchers suggest that as these habitats continue to dwindle, green roofs can be valuable in conserving or restoring endangered habitats and vegetation types, while offering supplemental habitat for rare or endangered species that depend on them (Burke 2003; Shriner 2003). Furthermore, extensive green roofs, which are generally absent from human use can provide an untouched home for birds and insects. 41

50 One of the most intensive studies to document the impact that green roofs can have on increasing urban biodiversity took place in Basel, Switzerland (Brenneisen 2003). Within the first three years of the study researchers documented 78 spiders and 254 beetles, of which fourteen of the spiders (18 percent) and 27 of the beetles (11 percent) were classified as rare or endangered. Due to the increase in insects, bird activity increased dramatically, introducing species such as black redstarts, wagtails, rock doves and house sparrows species which generally occur in higher mountain areas, on river banks or on steppes with grasslands and bare stony ground and patchy vegetated areas. The value of this impact was highly noted in London where green roofs have been planned to provide the black redstart, a rare and protected bird, with breeding habitat (Gedge 2003). This program, known as the Black Redstart Action Plan, was launched in London to protect biodiversity from the negative repercussions of urban development. Since the plan s initiation more than 15,000 square meters, or 161,400 square feet, of green roofs have been planned. Respectively, the plan has been essential in moving green roofs into the mainstream. The improvement of urban biodiversity through the use of green roofs is largely dependent upon construction, species selection and vegetation type (Dunnett and Kingsbury 2004). Introducing a variety of slopes and heights, open stony unvegetated areas, a variety of vegetation and freely and poorly drained areas maximizes the ecological function of green roofs. In addition to attracting invertebrates and birds, green roofs can also play an important role in linking fragmented habitat which is especially important for migratory birds, and help to connect green space in urban areas, blending the urban environment with 42

51 the natural, and making the city a more hospitable environment to humans and nonhumans alike (Sharp 2003). Stormwater Runoff Green roofs are highly valuable in the mitigation and even elimination of stormwater runoff problems associated with development and the rise of impervious surfaces (Snodgrass and Snodgrass 2006; Peck et al. 1999). Throughout the past several years research in this field has begun to multiply in the United States as researchers, nonprofit organizations and local governments rapidly seek an affordable solution to the problem of stormwater runoff pollution and sewer overflows. By limiting the volume of runoff and decreasing the rate of runoff green roofs have been shown to have tremendous potential to reduce flooding and protect water quality. This is accomplished in a variety of ways. Green roofs are effective in managing the problem of stormwater in two ways. First, green roofs retain stormwater and to a certain degree prevent it from entering the system. The water that is retained by the green roof is stored in the soil and eventually transevaporated by the vegetation on the roof. Even green roofs with 4 inches of substrate can retain as much as 60 percent of all rainwater, which has significant potential for addressing stormwater runoff and water quality (Moran et al. 2005). As important is the fact that green roofs decrease the peak rate at which the stormwater that is not retained leaves the premises, slowing the amount of runoff entering the system and decreasing the impact of the first flush the highly contaminated initial runoff (Graham and Kim 2003). Green roofs have revealed a delay in runoff anywhere from 30 minutes up to 4 43

52 ½ hours decreasing the rate of runoff by percent (Moran et al. 2003). Researchers note that engineers can plan for this reduction and delay of runoff when planning for stormwater and sewer infrastructure (Mather 2006). Green roofs have also been noted to contribute other benefits to water quality. Due to the fact that lead, zinc and copper are often used in roof construction, green roofs prevent these metals from washing off roof surfaces and contaminating the water (Peck et al. 1999). Additionally, because traditional rooftops can reach temperatures up to 180 F they have been shown to raise the temperature of stormwater runoff, a growing concern because of its degradation to water quality and aquatic life. A rooftop at a temperature of 100 F can heat stormwater from 73 F to 95 F (Wong 2005). Green roofs alleviate this problem. Much of the research pertaining to stormwater management benefits has been concerned with the depth requirements needed to achieve the best results. While intensive roofs can retain a greater amount of rainfall and slow the rate of runoff to a greater extent, these systems are often not generally regarded as an affordable option. While, intensive systems will most generally achieve greater results and prove more effective from prolonged winter storms, researchers seem to have been more compelled to demonstrate the effects that extensive systems can have on stormwater runoff and retention due to ease of implementation. Air Quality Particulate matter, primarily derived from vehicles, carries with it a host of health problems including respiratory problems and difficulties breathing, while heavy metals found in vehicle 44

53 emissions are toxic even in relatively low concentration (Dunnett and Kingsbury 2004). Particulates and ozone, the aggravating component of smog that is produced on hot sunny days, are both associated with increased death rates from respiratory illness and disease, especially during hot weather. The vegetation provided on green roofs helps to filter out particulate matter. As the airborne matter passes over the vegetation it settles onto the leaf and stem surfaces, which is then washed off by rainfall where it filters into the soil. Plant foliage is also able to absorb gaseous pollutants by storing the material in their tissue. However, there is little data to support the claims that green roofs can play a major role in this process. The majority of these claims are primarily inferred from studies performed in different locations on other types of vegetation. However, green roofs have been shown to play a major role in the trapping of heavy metals, a growing pollution problem in urban areas due to road travel. Research shows that green roofs can capture up to 95 percent of cadmium, copper and lead and 16 percent of zinc (Peck et al. 1999). These elements are thus prevented from washing off the roof during a storm event and prevented from entering the sewer network. Urban Heat Island Effect Heat is the number one killer related to weather in the United States 6. This problem is exacerbated by the phenomenon known as the urban heat island effect. Urban heat islands 6 United States Environmental Protection Agency. Heat Island Effect. [Online] (accessed February 8, 2007). 45

54 occur in urban and suburban areas as vegetation is replaced by dark impervious surfaces that absorb the heat instead of reflecting it. On average, temperatures are 2 F-10 F higher in urban environments than in surrounding rural settings which results in an increase in cooling devices and energy consumption. In certain areas, urban temperatures can even be up to 22 F hotter than surrounding rural areas (Wong 2005). This elevated temperature leads to elevated night-time temperatures, higher levels of humidity, polluted air and an increase in particulate matter (Dunnett and Dunnett 2004). Urban heat islands also increase ozone, a harmful pollutant found in smog. A solution to the heat island effect is to increase the percentage of vegetated land cover in urban areas. As water moves through the roots of plants and other vegetation, it makes its way up to the leaves where water is released as a vapor (Dunnett and Dunnett 2004). As plants use this process, known as evapotranspiration, they cool themselves as well as the surrounding atmosphere (Earth Pledge 2005). Rooftops in particular are generally the hottest areas in the urban environment (Wong 2005). As highlighted earlier, traditional asphalt roofs can easily reach 160 F during the summer while vegetated surfaces, including green roofs, rarely exceed 80 F (Lui and Baskaran 2003). A study performed in Canada, suggested that if 6 percent of total available roof space in Toronto had green roofs, summer temperatures could be reduced by 1.8 F to 3.6 F (Earth Pledge 2005). Furthermore, the results from the study indicate that a temperature reduction of 1.8 F would subsequently lead to a reduction in cooling demands by 5 percent. Scientists at the U.S. Department of Energy s Lawrence Berkeley National Laboratory (LBNL) 46

55 estimate that if ambient air temperatures in cities were reduced by a mere 5.4 F, the nation would save up to $6 billion per year in energy costs (Scholz-Barth 2004). If scientific predictions prove to be true, global warming will intensify the urban heat island effect by 5 C (Peck et al. 1999). To address this growing concern, cities around the world are trying to find ways to increase vegetative cover in urban environments (Deutsch et al. 2006). However, developable land is highly coveted. Green roofs provide the increase in vegetation, while preserving developable land. While other methods of roofing such as white or cool roofs do provide a similar effect, they do not offer the wide array of other benefits associated with green roofs (Snodgrass and Snodgrass 2006). The U.S. Environmental Protection Agency is trying to address the issue of heat islands through their Heat Island Reduction Initiative (HIRI) Program established in 1997 (Wong 2005). This program seeks to educate, promote strategies and support research geared to abate the heat island effect. Although the EPA has historically prioritized the use of cool roofs over green roofs due mainly to the barriers associated with implementing green roofs in the United States, they have supported the technology and have sponsored the GRHC conference annually. The EPA has noted that green roofs have begun to play an important role in the HIRI strategy and will continue to become more important as research and technology of the infrastructure advances. As the EPA continues to support green roof technology their endorsement will most likely increase the technology s legitimacy with the public. 47

56 Noise Pollution Noise pollution is another consequence of urban development which can threaten human health and welfare, interfering with the ability to function physically and psychologically and can lead to hearing loss 7. While urban areas often reflect sound, green roofs, particularly deeper intensive roofs, can absorb sound through both the substrate as well as the vegetation (Dunnet and Kingsbury 2004). Although research in this area is limited, a German study has shown that a 4-inch green roof reduced sound transmission into buildings by at least 5 decibels while Canadian claims suggest a 40 decibel reduction with a 4.8-inch substrate and a decibel reduction with an 8-inch substrate. Fire Prevention Although dry vegetation is often seen as a fire hazard, evidence from Europe shows that green roofs can help to prevent the spread of fire, chiefly when the roof is saturated (Peck and Kuhn 2000). Fire breaks such as gravel and concrete pavers, sprinkler systems and fire resistant plants can help to alleviate concern. Literature pertaining to the green roof s ability to retard the spread of fire is minimal. However, a recent study found that the threat of fire is times higher on bare roofs with fully adhered bituminous waterproofing membranes than on extensive green roofs with grasses, perennials and sedums (Breuning 2007). In fact Germany, where there is no evidence of fire ever associated with a green roof, often offers a percent discounted rate on fire insurance when an extensive green roof system is installed. As data in this area is insubstantial researchers hypothesize that this may be an area of future study. 7 United States Environmental Protection Agency. EPA Identifies noise levels affecting health and welfare. [Online] (accessed February 8, 2007). 48

57 CONSIDERATIONS Despite the many benefits associated with green roofs, the design and construction of a green roof is an intricate process and many issues to be considered prior to implementation. The design process of a green roof is extremely site specific and requires open and ongoing communication between the owner and design team (Velasquez 2005). This is particularly important to avoid otherwise preventable problems and increased cost. Issues and considerations which must be addressed prior to implementation include intended function, load bearing, slope, wind, plant selection, irrigation, schedule, maintenance, cost and building regulations. Function The first thing that the building owner needs to consider when designing a green roof is the function (Velasquez 2005). The intended function of the green roof will directly impact the system choice. It is vital that the intended function is decided prior to design and construction and that the necessary parties are continuously involved throughout the entire process. The function of the roof will in large part determine support requirements, plant selection, irrigation needs, schedule and costs. Furthermore, whether or not the roof will be accessible for people or not will help to determine concerns about access, safety, loads, circulation, and climate (Osmundson 1999). Load-Bearing As noted earlier, one of the most important things to consider when building a green roof is the structural composition of the building to be greened. Load-bearing is the most significant consideration for any green roof (Snodgrass and Snodgrass 2006). It relates 49

58 directly to whether an extensive or intensive green roof is suitable for that particular building or if structural reinforcement will be necessary (Dunnett and Kingsbury 2004). For new construction, load bearing is not such a limiting factor because the building can be designed to support the system of choice. For existing roofs however, it is essential that the green roof conform to the roof s existing load carrying capacity or undergo a structural upgrade, which can be quite costly. In both cases, it is vital to have a structural engineer or licensed practitioner determine the roof s maximum load-bearing capabilities, in addition to local snow loading requirements (Velasquez 2005). Slope Sloped roofs present the potential for slippage (Dunnett and Kingsbury 2004). Although extensive systems are suitable for roof slopes up to 30, certain precautions must be taken to prevent problems with sliding on the substrate and vegetative layers (Peck et al. 1999). On a roof with a slope greater than 20 it is imperative that the roofer ensure that the vegetative layer stays in place, particularly when the substrate is saturated, through the use of horizontal strapping, laths, battens or grids (Dunnett and Kingsbury 2004). Roofs with a slope greater than 30 can be built on, but will require special media mixes and specialized devices. Wind Wind is an important consideration for green roofs. Due to the fact that they protrude from the surface, green roofs have a high degree of wind exposure (Osmundson 1999). High wind speeds can dry out the substrate and vegetative layers as well as injure plants (Dunnett and Kingsbury 2004). Large organic parts and lightweight aggregate material can be blown 50

59 away fairly easily and any sustained wind damage should be attended to immediately following the event (Breuning 2007). However, wind uplift can be minimized by bonding the waterproofing layer to the roof beneath and surrounding the edges of the roof with strips of gravel, stones or pavers to shield the plants. Additionally, wind should be taken into consideration as it is a potential source of discomfort should the roof be intended for recreational purposes (Osmundson 1999). Plant Selection Green roofs are still so new to North America that no specific plant lists exist (Snodgrass and Snodgrass 2006). While species selection, growing medium and depth have been studied comprehensively in Germany, many of the publications are not readily available to the English-speaking world. However, plants suitable for green roofs typically share a number of characteristics (Dunnett and Kingsbury 2004) (see Table 9). For a green roof to achieve its function, Dunnett and Kingsbury (2004) note four fundamental characteristics for rooftop vegetation. To begin with, they must cover the substrate surface within a fairly short period of time after planting occurs. Secondly, they must be able to form a self-repairing mat so that damaged areas will be covered by new growth. Next, vegetation must be able to attract and transpire the amount of water that the structure is designed to maintain. And lastly, the vegetation must be durable enough to survive even the harshest weather conditions. Drought-tolerant plants tend to encompass the desired characteristics because they are able to survive while undergoing transformation to their form and physiology. Sedums happen 51

60 TABLE 9: FEATURES TO LOOK FOR IN PLANT SELECTION OF A GREEN ROOF Type Comments Low mat-forming or Stems that root into the substrate as they grow are arguably more cushion growth useful than those that root only from a central point, as this allows the plant to maintain good coverage of the surface after damage. Low, carpeting or mounded plants will be less susceptible to wind damage and uprooting than taller plants. Many ground-hugging and mat forming plants are also well adapted to drought conditions. Succulent leaves and or other water-storage capacity Compact twiggy growth and small evergreen leaves held close to the stem on ground-hugging plants Grey or silver foliage Geophytes Shallow rooting Evergreen foliage Short life cycles and effective reproduction Attractive characteristics These characteristics are typical of a wide range of subshrubs from habitats that experience either heat or wind-induced water shortages. This coloring is due to either minute hairs on the leaf surface or a waxy coating, both adaptations that reduce water loss. They are also visually attractive. Species which lie down to bulbs or tubers during the winter or during a hot and dry season, are often visually striking and can play an important secondary role in roof-greening vegetation. Except in the deeper soils, tap-rooted species will be less successful than species with spreading root systems. Such plants are know as shelf-plants because of their ability to grow in very thin layers of substrate. For green roofs to work year-round, plants must also be functioning year-round. This is particularly true in terms of water management evapotranspiration only being effective when plants are actively photosynthesizing. This will result in gaps in the vegetation being effectively filled, promoting the maintenance of long-term vegetation cover. Where roofs are visible and actively used, the species, as well as promoting the functioning of the roof, should also be visually attractive. Source: Dunnett and Kingsbury 2004 to be particularly effective due the fact that they store water in their leaves and are shallow rooted, while this is typically not the case for other drought-tolerant plants. It is noteworthy that although a particular species may not meet all requirements, a combination of several different species will be able to perform the necessary function as a collective whole. Of course, as is often the case in green roofs, the desired function of the green roof will play an important part in the plant selection process. Load bearing considerations, which typically reflect potential soil depth, may also narrow the choice of viable plants, generally 52

61 leaving aesthetics as the final consideration. For an in-depth discussion of plant and vegetation selection for green roofs, see Green roof plants by Snodgrass and Snodgrass. Irrigation Extensive green roof systems require watering during the first couple of years for the plants to establish themselves properly (Velasquez 2005). However, if extensive green roofs are planned and constructed correctly there is no need for additional irrigation systems with the exception of significantly dry climates (Dunnett and Kingsbury 2004). Techniques can include drip irrigation or a hose, which can funnel collected rainwater from a cistern to the roof. On the other hand, intensive systems, with greater soil depth and more elaborate vegetation will require additional irrigation systems. At times, these systems may be quite intricate and can imply a significant additional cost. Schedule Due to the fact that green roofs are multifaceted the construction of a green roof requires additional planning and a well thought out schedule. However, this can present a problem. Unlike Europe, the United States lacks a systematic approach that accounts for environment, vegetation, manufacturing and distribution methods. In North America, the green roof industry has not matured and every project requires a custom job primarily due to the variation in climate across the U.S. (Russell 2007). In the past, there has been no systematic approach for the planning, design and construction of green roofs. In an effort to counteract this issue, GRHC released a guideline in 2005, for the successful planning, execution and maintenance of green roofs known as the critical 53

62 path. This has provided the industry with a uniform approach that can be plugged into most regional conditions in North America, and set minimum standards with which to comply. GRHC is currently working to establish a certification and accreditation process in place to police green roof design and construction. However, this is not yet in place. In the meantime, the critical path helps to ensure that public health, safety and welfare guidelines are considered, as well as cost effective methods, best practices, and other project management considerations (Russell 2007). Maintenance Another factor that must be considered is maintenance. Intensive green roofs require a significant amount of maintenance similar to ground-level gardens, including separate irrigation systems and regular gardening. Although extensive roof systems are virtually selfsustaining after the initial two years, they are not maintenance free. Nevertheless, extensive systems can be considered low-maintenance, requiring four annual tasks: Feeding, plant protection from pest and disease, inspection of proper drainage and weeding requirements (Dunnett and Kingsbury 2004). Cost Cost restraints may also play into the intended scope of the green roof. The cost of a green roof is influenced by a number of different factors (see Table 10). The additional upfront cost associated with green roofs is arguably the biggest obstacle the technology faces. Extensive green roof systems generally do not require additional structural support and do not support a wide variety of vegetation. Accordingly, costs for extensive systems in North America range from $10-$25 per square foot. Intensive green roofs however, because of the 54

63 TABLE 10: GREEN ROOF COST RANGES AND FACTORS 8 Element Price Range Factors Growing medium Extensive $2-$12 / ft. 2 Intensive $2-$20 / ft. 2 Volume / type of growing medium, shipping distances and method of conveyance to roof (crane, blower truck, manual, etc.) Vegetation Extensive $0-$5 / ft. 2 Intensive $1.25-$10 / ft. 2 May not be required. Type and size of plants of plants, time of year, seeds, cuttings, plugs, mats, pots, shrubs, trees may require containers and / or anchorage. Extensive $2.40-$6.40 / ft. 2 Installation Intensive $6.40-$14.40 / ft. 2 ( % of material cost) Modular green roof system Extensive $10 + / ft. 2 (including vegetation, planting, growing medium Intensive $13 + / ft. 2 and root repellant layer) Structural reinforcement of existing roof Cost is highly dependent on existing structure. Size of project, sophistication of design, type of planting approach, nature of access to roof. Sophistication of design, shipping, installation, plant species and density. May not be necessary. Consult a structural engineer to determine the load carrying capacity of any roof. Erosion protection layer $0-$0.30 / ft. 2 medium is not left exposed or May not be necessary if growing vegetation is well established. Curbs/Borders $0-$20 per linear foot May not be necessary. Type (pre-cast concrete, aluminum edging, gravel, timber borders, modular systems, recycled products, etc.) and length. Walkways $0-$10.20 / ft. 2 concrete unit pavers, natural stone, wood decking, recycled products etc.) May not be necessary. Type (pre-cast and length. Railings Maintenance Irrigation system $0-$65.45 per linear foot Extensive $0.25-$4.10 / ft. 2 for the first two years Intensive $1-$4.10 / ft. 2 $0-$5 per linear foot * Costs may vary significantly due to regional differences May not be necessary. Materials (aluminum, brass, wrought iron, welded steel, etc.) Thickness of railing. Number of rails. Roof deck penetration. Size of roofs, types of plants, nature of access. May not be necessary. Type of irrigation system used and size of project. support requirements, can cost anywhere between $25-$100 per square foot or more (Velasquez 2005). 8 Green Roofs for Healthy Cities. Green Roof Design 101: Introductory Course, 2 nd edition Participant s Manual. 55

64 Increased installation costs and slow investment return have been a challenge for green roofs in the past. Although at first glace green roofs seem to be quite a bit more costly than traditional rooftops, the increase in the life of the roof offsets additional upfront cost significantly. Additionally, the benefits of green roofs should be factored into the cost. GRHC has recently announced the preliminary formulation of a Life Cycle Costing (LCC) model for green roofs to be built by the Athena Institute (Velasquez 2007). This initiative aims to create an economic means to evaluate project investment alternatives over a designated study period. The end product is planned to be web accessible to allow the user to factor in the study period, discount rate, general inflation rate, energy price inflation, investment cost data, financing data, residual or resale value, recurring operating and maintenance costs, and replacement costs as well as certain soft costs such as social or amenity value as well. Researchers expect the tool to be groundbreaking in the green roof marketplace. Building Regulations Green roofs are subject to the same safety restrictions as any development, generally requiring a separate egress (Osmundson 1999). Issues concerning building codes, height ordinances, and safety restrictions are city-specific and must be well researched prior to the design or construction of a green roof. However, railings, parapets or blockades are generally required. Furthermore, in some cities, green roofs are considered an additional floor and may be subject to height restrictions. 56

65 GREEN ROOF WORLD HISTORY Green roofs date back thousands of years to early civilizations of the Tigris and Euphrates valleys, which included the legendary hanging gardens of Babylon constructed in the seventh and eighth centuries (Dunnett and Kingsbury 2004). The home of other ancient roof gardens included the ziggurats of ancient Mesopotamia, built circa 600 B.C., and the Villa of the Mysteries in Pompeii (Osmundson 1999). Greeks, Romans, Persians and other cultures used roof gardens to cool their environment during periods of extreme heat (Snodgrass and Snodgrass 2006). During the Middle Ages and throughout the Renaissance, green roofs could be found in places of divine worship such as the Mont-St-Michel in France, to meet the demands of Christian ecclesiastical architecture of the time (Osmundson 1999, 115). Green roofs also spread next door throughout Italy. One of the first and best preserved green roofs can be found in the Palazzo Picholomini which was commissioned by Pope Pius II. Others were present in the Tower of the Guinigis, a family of silk merchants, as well as in the garden of Cosimo de Medici, the first of the Medici dynasty (Osmundson 1999). During the 17 th and 18 th centuries, roof gardens were seen as an exquisite amenity by the nobility. In czarist Russia a 10-acre garden was built on the upper level of Kremlin palace which featured a 1,000 square foot pond complete with fountains fed by water from the Moscow River. The lower garden was nearly half the size, yet sported the same amenities. Amazingly, these gardens required over 10 tons of lead for waterproofing. Green roofs were also commissioned by other nobility at the time, such as Catherine II of Russia and King Ludwig II of Germany. 57

66 By the mid-nineteenth century, concrete began to surface as a roofing material and flat rooftops began to emerge throughout Northern Europe and the United States (Dunnett and Kingsbury 2004). The World Exhibition in Paris, held in 1868 featured a concrete planted nature roof, which marked the first of several similar experiments throughout Western Europe. A block of Parisian apartments were built in 1903 featuring planted terraces and a roof garden. After returning from a visit to Paris, Rudolph Aronson, a New York conductor and musician came to the conclusion that the sole means to overcome the dense downtown land use was to design rooftops to accommodate a summer outdoor garden theater. In 1882, he designed the Casino Theater (Osmundson 1999). From then on the popularity of these rooftops for summertime entertainment continued to flourish as designed by architectural greats such as Frank Lloyd Wright and Le Corbusier. During this time, green roofs experienced widespread popularity and the term roof garden was coined. However, the Depression halted the majority of large-scale construction efforts. Thus the modern day reemergence of green roofs did not truly begin to flower until the mid-twentieth century following World War II. Due to the post-war building campaign which transpired in Switzerland, Germany and Austria, green roofs began to gain popularity as an alternative to traditional methods of roofing, gaining attention for their practical, aesthetic and environmental purposes (Dunnett and Kingsbury 2004). Here, in the German-speaking European countries was the birth of the modern day green roof movement. 58

67 Germany and Europe Germany has been and remains the leader in the green roof movement worldwide (Lawlor et al. 2006). Today, with green roofs included on more than 7 percent of all new roof construction (Velasquez 2005), Germany continues to lead the way in the green roof movement, not only in developing green roof technologies and systems, but by enacting state and federal policy to mandate and incentivize their installation (Herman 2003). The green roof market in Germany has grown significantly since 1982, averaging percent annual growth (Peck et al. 1999). By 1989, 1 million square meters of roofs were greened and by 1996, the number had exploded to 10 million square meters. By 2001, 14 percent of all German flat roofs had been greened and 13.5 million square meters of green roofs were installed in that year alone (Herman 2003; English Nature 2003). Today, there are at least 2 billion square feet of extensive green roofs built in Germany (Breuning 2007). This significant growth is attributable to a number of factors. Approximately 70 percent of German municipalities offers various types of subsidies to help offset the higher upfront cost associated with green roofs while others offer grants for green roof construction and will even pay up to 50 percent of capital costs (Velasquez 2005). Moreover, roughly 30 cities throughout Germany have zoning districts which require green roofs. Moreover, the cost of a green roof in Germany is significantly lower than the costs in North America (Velasquez 2005). On average, an extensive green roof costs between $4-$13 per square foot in Germany, while, as noted earlier, costs in North America range from $10-$25 per square foot with intensive green roofs costing anywhere between $25-$100 per square 59

68 foot or more. This cost difference is attributable to the well established green roof market in Germany. The Germans have also developed the most comprehensive guidelines for green roof planting and installation through the Forschungsgesellschaft Landschaftsentwicklung Landschaftsbau (FLL) or the Landscape Research, Development and Construction Society (Snodgrass and Snodgrass 2006). This non-profit organization has worked for over 25 years to develop standards for green roof technology (Philippi 2006). In 2004, the FLL guidelines were translated into English and are currently the only existing guidelines accessible to the English speaking world. However, while Germany proves to be the leader in green roofs, European neighbors are not far behind. In Switzerland, 12 percent of flat roofs are greened and authorities require that 25 percent of new commercial development be greened (Snodgrass and Snodgrass 2006). In Vienna, world renowned Austrian architect Friedensreich Hundertwasser built what would become the most influential green roof in Vienna housing 992 tons of soil and 250 trees and shrubs (Dunnett and Kingsbury 2004, 13). In Austria green roofs are also common where cities offer many of the incentives and policies as Germany (Dunnett and Kingsbury 2004). Norway is another common place to see green roofs where the idea of the green roof is imbued with romantic notions of nature (Dunnett and Kingsbury 2004, 17). In fact, 60

69 Scandinavian countries have a long-tradition of using sod-covered buildings to keep them dry and warm. However, with the exception of Norway, most of the remaining green roofs are regarded as historical structures and modern day green roof applications rarely take place. When it comes to the advancement of green roof infrastructure, Britain and the Netherlands have not been able to keep up with other European countries (Dunnett and Kingsbury 2004). Arguably similar to the United States, green roofs remain enigmatic to many and unknown to more. In southern Europe, in countries such as Greece, Italy, Spain and Portugal, green roof development has lagged as well, and Russian neighbors have very little in the way of green roofs. In Europe, green roof research and development began in the 1950s as part of a wider movement that recognized the ecological and environmental value of urban habitats, and in particular the benefits to plants and wildlife (Dunnett and Kingsbury 2004, 15). This research began to take a turn a decade later to look into the techniques and practicalities of growing plants in thin substrate layers on roofs (Dunnett and Kingsbury 2004, 15). What resurfaced in Europe as a counter-culture movement has made its way into the mainstream as an effective way to balance development with the natural world. North America The green roof industry in North America has lagged significantly behind Germany. However in recent years, the industry has experienced a significant growth. This is in part due to the success transpiring in Europe. Researchers, politicians and advocates have 61

70 traveled to Europe with the desire to spread the green roof gospel to North America. While at the same time, many European markets have begun to expand their market overseas. While some cities in the United States and Canada have embraced the green roof as part of a comprehensive plan to sustainable development, many others remain ignorant of the green roof s potential. Although high profile developments including the Ford Motor Company complex in Dearborn Michigan and Chicago s City Hall have helped to raise national attention to green roof infrastructure, many still regard it as foreign and exotic. Nevertheless, cities such as Chicago, Portland, Seattle, Minneapolis and others have enacted policy to support widespread green roof installation. Many organizations in North America share a vested interest in green roofs. Most notably is GRHC, a member organization based in Toronto which is committed to expanding the green roof market and the Center for Green Roof Research at Penn State University. Of particular contribution to the research on green roofs in North America is GRHC s annual Greening Rooftops for Sustainable Communities conference where stakeholders come from around the country and around the world to exchange the latest research. Non profit organizations have become vital to the advancement of green roof infrastructure in North America as well as educators and advocates. Japan With a population of over 12 million people, Tokyo is the most populated city in the world (Lawlor et al. 2006). Over the last 100 years, the temperature in Tokyo has increased by 3 C. In 2001, the Tokyo Metropolitan Government (TMG) responded to this growing problem 62

71 by instituting a green roof policy to abate the heat island effect. Under this policy, green roofs are required to cover 20 percent of new development larger than 10,764 square feet for private developments and 2,691 square feet for public developments. Failure to comply with the regulation results in a fine of approximately $2,000 (US). The compelling thing about Tokyo s green roof policy is the process involved by which it was promulgated. Unlike other cities, Tokyo bypassed the formal research and analysis phase and did not seek public support. After a telling demonstration project, steps were taken to educate the development sector and immediately incorporate green roofs into policy. By the end of 2001, one year after the policy was enacted, the total net area of green roofs had increased from 564,000 square feet to over 1.1 million square feet. As an effort to increase green space, Tokyo hopes to increase green roof coverage to 130 million square feet. CLOSING REMARKS Due to the vast benefits associated with green roofs, many cities around the United States have begun to research, advocate and incorporate green roofs into government policy, create incentives for residents, spearhead educational campaigns and in some cases mandate green roofs on new development. Oftentimes green roof policy is part of a larger attempt to manage stormwater, decrease energy consumption or promote sustainable design. 63

72 Cities like Toronto, Chicago and Portland tend to be at the forefront of the green roof movement in North America, boasting several different opportunities for the public and private sectors to incorporate green roof infrastructure. Examples include reduced fees, tax credits, grants, stormwater credits and low interest loans. Other examples include regulatory or process benefits, such as, density bonus allowances, green space waivers, or variances, and expedited permit processing. For these cities, green roofs offer a practical means of achieving economic savings in energy and water treatment costs, while ameliorating environmental conditions for their residents (Snodgrass and Snodgrass 2006). Nevertheless, while Japan, Germany and other European countries have embraced the green roof as a solution to a number of environmental problems, the green roof industry still faces many challenges in the United States. Many policy makers have not yet embraced the green roof as a tool to manage both environmental and economic issues. Some claim the amount of data does not support the initiation of a policy designed to advance green roof infrastructure. Others argue that performance is difficult to quantify making it difficult to assign a monetary value. Still, the industry s recent growth in North America has generated a compelling argument to advance the body of research to show how these benefits apply under local conditions. 64

73 4 chapter THE SCIENCE OF HYDROLOGY This chapter defines hydrology and provides a background on the science. It continues to discuss models used to monitor surface and subsurface flow. Special attention is given to stormwater and the need for planners to incorporate stormwater modeling into land use planning. The chapter concludes by reviewing literature specific to TR-55, including strengths and limitations. DEFINITION Hydrology is the science which studies the movement of water through the hydrologic cycle. The hydrologic cycle is composed of seven steps: evaporation, precipitation, interception, runoff, infiltration, recharge and discharge. Water evaporates from the oceans and the land surface, is carried over the Earth in atmospheric circulation as water vapor, precipitates again as rain or snow, is intercepted by trees and vegetation, provides runoff on the land surface, infiltrates into soils, recharges groundwater, discharges into streams, and ultimately, flows out into the oceans from which it will eventually evaporate again (Maidment 1993, 1.3). Like all science, hydrology attempts to predict what will occur within the hydrologic cycle and then monitor actual results. Hydrologist s principal concern is to measure and observe water use, water control and pollution control. Hydrologists measure these three areas through the use of hydrologic modeling. 65

74 BACKGROUND Hydrology has been a subject of interest for many years. In 1851, the concept of time of concentration was introduced to the science of hydrology as well as an initiative to measure the intensity of rainfall (Maidment 1993). These two basic components make up the fundamental core of the methods currently used by hydrologists to compute stormwater runoff. However, the hydrologist s ability to predict the hydrologic cycle has evolved tremendously over the last 150 years. This transition has gone from the reliance on flawed experimental observations to the ability to accurately represent different scenarios through sophisticated modeling (Dixit 2003). Specifically since the 1950 s, the science has continued to advance parallel to the explosion in the computer-science related field. Despite the evolution of the science of hydrology, there is still a lot that remains a mystery concerning the hydrologic process. In other cases, the knowledge of various processes may exist, yet those processes have not yet been fused into a workable model (Maidment 1993). HYDROLOGIC MODELS A hydrologic model is a mathematical depiction of the movement of water over land or below its surface (Maidment 1993). In other words, it is a tool used to predict the behavior of the hydrologic cycle. Hydrologic principles and working equations can be derived from the fundamental laws of physics, or they can result from a synthesis of field observations and phenomenon. Once an adequate process description has been constructed, it is often incorporated into computer models whose parameters are determined using local data for each application (Maidment 1993, 1.9). 66

75 When using any quantitative expression of a process or phenomenon one is observing, analyzing or predicting, certain assumptions are made (Overton and Meadows 1973, 4). Since no process can be completely observed, any mathematical expression of a process will involve some element of stochasticism, i.e., uncertainty (Overton and Meadows 1973, 4). This is because of the complexity of hydrologic processes. Respectively, alternative mathematical formulations are proposed to statistically reproduce historical records (Maidment 1993). SURFACE AND SUBSURFACE MODELS There are many models available to measure the flow of various processes within the hydrologic cycle. In addition to being able to monitor processes such as precipitation, evaporation, infiltration, groundwater flow, snow and floating ice flow, stream flow, flood runoff, and flow routing, models can also be employed to measure things such as climate which is strongly determined by the earth s hydrologic cycle (Maidment 1993). As technology has advanced, so has the capability of the science to predict these cycles. Computer models have become essential for analyzing complex problems because they provide a quantitative framework for synthesizing a large set of parameters that describe spatial variability within the environment, as well as spatial and temporal trends in hydrologic parameters and stresses (Maidment 1993, 22.1). New computer programs and improvements to existing programs are being created every year. Tables 11 an 12 below highlight several computer program packages in hydrology that are available to hydrologists to measure both surface and subsurface flow. The scope of the models vary accordingly. 67

76 TABLE 11: COMPUTER MODELS FOR SURFACE WATER No. Model Source 1 Single event rainfall runoff models 1.1 HEC-1 U.S. Army Corps of Engineers, Davis, CA 1.2 TR-20/TR-55 Soil Conservation Service, USDA, Washington D.C. 1.3 Illudas Illinois State Water Survey, Champaign, IL 1.4 DR3M U.S. Geological Survey, Reston, VA 2 Continuous stream flow simulation 2.1 SWRRB Agricultural Research Service, USDA, Temple, TX 2.2 PRMS U.S. Geological Survey, Reston, VA 2.3 SHE Institute of Hydrology, Wallingford, England 3 Flood hydraulics 3.1 Steady flow HEC-2 U.S. Army Corps of Engineers, Davis, CA WSPRO U.S. Department of Transportation, Washington D.C. 3.2 Unsteady flow DMBRK U.S. National Weather Services, Silver Spring, MD DWOPER U.S. National Weather Services, Silver Spring, MD 4 Water quality 4.1 SWMM University of Florida Water Resources Center, Gainesville, FL 4.2 HSPF U.S. EPA Environmental Research Laboratory, Athens, GA 4.3 QUAL2 U.S. EPA Environmental Research Laboratory, Athens, GA 4.4 WASP U.S. EPA Environmental Research Laboratory, Athens, GA Source: Dixit 2003, 33-34; Maidment 1993 TABLE 12: COMPUTER MODELS FOR SUBSURFACE WATER No. Model Source 1 Groundwater flow 1.1 PLASM International Groundwater Modeling Center (IGMC), Colorado School of Mines, Golden, CO 1.2 MODFLOW U.S. Geological Survey, Reston, VA 1.3 AQUIFEM-1 2 Groundwater contamination transport 2.1 AT123D International Groundwater Modeling Center (IGMC), Colorado School of Mines, Golden, CO 2.2 BIO1D Geotrans, Inc., Herndon, VA 2.3 RNDWALK International Groundwater Modeling Center (IGMC), Colorado School of Mines, Golden, CO 2.4 USGS MOC U.S. Geological Survey, Reston, VA 2.5 MT3D S.S. Papadopulis and Associates, Inc., National Water Well Association 2.6 MODPATH U.S. Geological Survey, Reston, VA 3 Variably saturated flow and transport 3.1 VS2D U.S. Geological Survey, Reston, VA 3.2 SUTRA U.S. Geological Survey, Reston, VA; National Water Well Association Source: Dixit 2003, 33-34; Maidment

77 STORMWATER MODELS Stormwater models are used to measure the direct response to rainfall (Overton and Meadows 1973). It is vital to include stormwater models in land use planning to measure the degree by which water resources will be affected by development and to employ effective strategies by which to alleviate the intensity of environmental degradation. Because the hydrologic cycle is tied together with the characteristics of the land it is imperative that planners use these models to simulate alternative methods of development. There are two theoretical approaches that have been used to develop stormwater models. Urban planning often employs deterministic modeling. This is a type of model which attempts to simulate scenarios based upon physical laws and measures of initial and boundary conditions and input (Overton and Meadows 1973, 7). A deterministic model is usually derived from first principals (like Newton's Law), and are therefore approximations of what we observe in nature, yet taken as exact solutions (Shuster 2007). The second is parametric modeling. Parametric models are somewhat less rigorously developed and generally simpler in approach (Overton and Meadows 1973, 8). The differences between the two are generally minor. TECHNICAL RELEASE 55 TR-55 is a lumped-parameter empirical model based on a statistical assessment of rainfallrunoff patterns observed for different types of land use created by the United States Department of Agriculture (USDA) (Shuster 2007). A simplified version of TR-20 its predecessor, TR-55 presents uncomplicated methods to approximate stormwater runoff and peak discharges in small watersheds (USDA 1986). The model is a single-event rainfall 69

78 runoff model that uses a design storm as rainfall input to model a single watershed (Maidment 1993). The fairly simple process allows those who do not have an extensive background or knowledge of hydrology to perform a hydrologic analysis with respect to land cover (USDA 1986). The ease of use has allowed for less informed users to apply the model (Fennessey et al. 2001). In TR-55, mass rainfall is converted to mass runoff by the use of a curve number (CN) an empirical parameter used to predict runoff and infiltration (USDA 1986). Determination of a CN is based on soils and cover conditions which the model represents as hydrologic soil, cover type, treatment and hydrologic condition (USDA 1986, 1-1). Soils are classified into four different categories according to surface intake rates and subsurface permeability (see Table 13). Once soil type is determined from the Natural Resource Conservation Service (NRCS) soil survey, available for download in GIS format, the Hydrologic Soil Group (HSG) can be established by looking up the name of the soil in the glossary found at the end of the model. The next factor needed to determine CN is cover type. There are several methods to obtain information concerning cover type. The most common methods are through the use of field reconnaissance, aerial photographs and land use maps (USDA 1986, 2-2). Once cover type is established the modeler simply assigns a predetermined CN based on a matrix of cover type and HSG. This CN is multiplied by the area of the cover type for each cover category. This total is then summed and divided by the total area to give a composite CN. 70

79 TABLE 13: HYDROLOGIC SOIL GROUPS Soil group Textures Description Sand, loamy Have low runoff potential and high infiltration rates even when A sand, or sandy thoroughly wetted. They consist chiefly of deep, well to loam excessively drained sand or gravel and have a high rate of B C D Silt loam or sandy clay loam Sandy clay loam Clay loam, silty clay loam, sandy clay, silty clay, or clay water transmission (greater than.030in/hr). Have moderate infiltration rates when thoroughly wetted and consist chiefly of moderately deep to deep, moderately well to well drained soils with moderately fine to moderately coarse textures. They have a moderate rate of water transmission ( in/hr). Have low infiltration rates when thoroughly wetted and consist chiefly of soils with a layer that impedes downward movement of water and soils with moderately fine to fine texture. Have a low rate of water transmission ( in/hr). Have high runoff potential. They have very low infiltration rates when thoroughly wetted and consist chiefly of clay soils with a high swelling potential, soils with a permanent high water table, soils with a claypan or clay layer at or near the surface, and shallow, soils over nearly impervious material. These soils have a very low rate of water transmission ( in/hr). Note: Some soils are in group D because of high water table that creates a drainage problem. Once these soils are effectively drained, they are placed in a different group. Source: Dixit 2003, United States Department of Agriculture 1986 Once the composite CN is determined, the modeler obtains 24-hour rainfall amounts for the storm event in question and looks up the runoff depth in the table provided in the manual. When the runoff depths for CNs or rainfall amount is not shown, the modeler interpolates the values that are shown (USDA 1986). CLOSING REMARKS Curve numbers are useful mechanisms for design purposes related to land use planning and stormwater management (USDA 1986). However, there are certain limitations. The equation does not include an expression for time and, thus it does not take rainfall duration or intensity into account. Furthermore, the CN cannot be used to estimate runoff from snowmelt or rain on frozen ground and is less accurate when runoff is less than 0.5 inches. When the weighted CN is less than 40, another procedure should be used. 71

80 Another one of the limitations of the TR-55 is the amount of data required to execute the methodology (Brabec et al. 2002). This can limit the users of the model, particularly in smaller jurisdictions where access to data is not readily available. Although the use of GIS has become standard in many cities across the United States many smaller jurisdictions may not have access to the data or the may be restricted financially. All things considered, TR- 55 is an asset to the stormwater modeling field. Its ease of use has gained widespread acceptance within the engineer and design community (Fennessey et al. 2001). In fact, studies suggest that the use of TR-55 accounts for over half of conducted hydrologic studies. Some concern has been raised concerning the model s accuracy and precision, however, when put to the test, TR-55 proves as accurate as other more complicated models (Fennessey et al. 2001). 72

81 5 chapter APPLICATION OF THE TR-55 TO THE CASE STUDY This chapter quantifies existing levels of impermeability at the University of Cincinnati and uses the TR-55 to quantify existing stormwater runoff. The second half of this chapter identifies the buildings that are eligible for a green roof and applies a CN of 66 to that area to determine the reduction in stormwater runoff for a 2-year storm. The chapter concludes with a discussion of the study s findings and limitations. EXISTING CONDITIONS Identification of case study site and determination of boundary Once the literature review was complete, the case study site was identified and the boundary was determined using campus maps available on the university s web site. Based on the criteria set forth in chapter two the University of Cincinnati was chosen as the case study site for this project due to the following: It is located within Hamilton County and within the political boundaries of the City of Cincinnati It is located within a single watershed It is located within an area which has a large amount of institutional development 73

82 MAP 1: LOCATION OF CASE STUDY IN RELATION TO CINCINNATI WATERSHEDS 74 Source: CAGIS 2007

83 MAP 2: AERIAL VIEW OF THE UNIVERSITY OF CINCINNATI Source: CAGIS

84 The University of Cincinnati is located in the Uptown Cincinnati, Ohio in the upper Mill Creek sewershed (see Map 1). The university is divided into two campuses, East and West, located caddie corner to one another on opposite sides of Martin Luther King Jr. Boulevard. West campus is framed by Clifton Avenue running North-South along the western periphery, Calhoun Street running East-West along the southern border and Jefferson Avenue to the East (see Map 2). East campus is bordered by Vine Street on the western edge and Burnett Avenue on the eastern, with several collector streets defining the northern edge. The total case study area for the University of Cincinnati, including both East and West campus was determined to total acres. Quantification of stormwater runoff at University of Cincinnati The first step in executing the methodology required for TR-55 was to determine the existing land cover on campus. Land cover was calculated using the following steps: Open the Mr. Sid aerial photo in CAGIS Create new shapefile in ArcCatalog to delineate the case study boundary and trace the boundary over the aerial photo using campus maps to determine the boundaries of each campus Create two additional shapefiles in ArcCatalog, one for open space and one for buildings Modify both shapefiles using the editor toolbar, tracing each land cover as a new polygon Using the dissolve function in ArcCatalog, combine the two shapefiles into one 76

85 Create a new shapefile in ArcCatalog and clip it using the boundary shapefile, then minus the combined shapefile to give the third shapefile for impervious cover including roads, sidewalks and parking lots Calculate the areas for all three shapefiles, using the function in ArcMap Export three tables to an excel file Calculate total area for each of the three land use categories The second step in executing the methodology required for TR-55 was to plug soil type and land cover into a worksheet to determine the composite CN. The soils layer in CAGIS indicated that the entire campus is built on Rossmoyne soil type, which is in HSG C, one of the model inputs (see Map 3). Roughly half of the campus was located on a 3-8 percent slope located on the outer West and South region of West campus and in the eastern half of East campus. With the exception of a small area located in the northeast corner of West campus, the remaining soil has a slope ranging from 8-15 percent, although slope was not one of the inputs needed to quantify runoff volume. The composite CN was calculated using the following steps (see Table 14): Assign a CN to each land use (open space or impervious cover, i.e. buildings, roads, sidewalks and parking lots) according to soil type found in Table 2-2a in TR-55 manual Multiply the total area of each land use by the according CN Sum the two areas Sum the two products Divide the total product by the total area to get the composite CN 77

86 TABLE 14: EXISTING CURVE NUMBER Soil and HSG Cover description CN Area (%) Product Rossmoyne, C Open space, good condition Rossmoyne, C Impervious cover Total CN (weighted) = total product/total area CN 92 Based on the coverage in the table above, the CN for existing conditions was determined to be 92. The third step was to determine the existing runoff for a 2-year storm event. According to the Rainfall frequency atlas of the Midwest a 2-year rainfall in Cincinnati is 2.86 (Huff and Angel 1992). Existing runoff was calculated by using rainfall (2.86 inches) and CN (92) with Table 2-1 in the TR-55 manual. Based on these inputs the depth of runoff was determined to be 1.97 inches (see Table 15). TABLE 15: EXISTING RUNOFF Factors Storm Event Frequency - yr. 2 Rainfall, P (24-hour) - in Runoff - in The fourth step was to convert runoff depth into total runoff volume for the entire case study. To convert the depth in inches to feet, 1.97 was divided by twelve giving a quotient of feet. This was then multiplied by the total area of the case study in square feet (1,540,639.47) to obtain the volume in cubic feet. The total volume was determined to equal 252, cubic feet which was multiplied by to produce a total of 1,890,064 gallons of stormwater runoff. 78

87 MAP 3: SOIL TYPE FOR THE UNIVERSITY OF CINCINNATI Source: CAGIS

88 PROJECTED CONDITIONS Identification of Green Roof Criteria As outlined in chapter two, a number of criteria based on available literature were established to ensure that the study would project workable results. The buildings 9 deemed suitable for green roofs were: Non-residential development Greater than 500 square feet With a slope less than 30 In this study a green roof was defined as an extensive system with a growing medium depth of 3 inches. Green Roof Building Inventory The next step was to perform the building inventory to determine structures suitable for green roofs. The first step was to do a data query for the building layer in GIS using the select by attribute feature to remove all structures which did not meet the minimum size requirement. Only one small building located near the southern border on East campus was less than 500 square feet. Once this building was removed, the layer was superimposed over the aerial image at a transparency level of 70 percent. The buildings were then surveyed and those with pitched roofs were removed from the shapefile. The green roof building inventory was completed using the following steps: Open the Mr. Sid aerial photo in ArcMap Create a copy of the buildings shapefile in ArcCatalog and rename it green roofs 9 This study defined parking structures as buildings, and were therefore suitable for green roofs assuming all other criteria were met. 80

89 Import the green roof shapefile into ArcMap Use the select by attribute function in ArcMap to identify structures under 500 square feet Open the editor toolbar and select modify feature to remove this polygon Set the building layer at 70 percent transparency and remove the polygons determined to have pitched roofs Save the edits Open the attribute table and export the table in excel format The total area for the new building layer was calculated (3,507, square feet) and deemed as the building area suitable for green roofs (see Table 16). Next, 20 percent of the building area suitable for green roofs (701, square feet) was subtracted to account for area needed for HVAC systems, railings, access or other non-greenable functions. This provided a remainder of 2,806, square feet of green roofable area. TABLE 16: BUILDING INVENTORY Description Square Feet Acres % (TBA) Total Building Area (TBA) 3,645, Area not suitable for green roof 137, % Building area suitable for green roof 3,507, % (Minus 20% for HVAC and other) 701, Total Green Roofable Area 2,806, % Total Non Green Roofable Area 839, % Determination of a CN for Extensive Green Roofs As mentioned in chapter two, TR-55 does not provide the CN for a green roof so it was necessary to determine a useful CN for this study. The determination of a curve number for 81

90 a 3-inch extensive green roof was determined according to a paper written by Charlie Miller founder and president of Roofscapes, Inc the revered green roof design, engineering and installation firm based out of Philadelphia. Although this paper was not in a peer reviewed journal Miller is a member of the Technical Advisory Committee for the Center for Green Roof Research at Pennsylvania State University, and the GRHC Coalition and was deemed to be a credible source by members of this committee. Furthermore, after looking over various CNs, the members of this thesis committee decided that 66 was a reasonable estimation. According to Miller, the CN for a green roof is directly related to the intensity of the storm event (Miller 2006). Using the input and output from the 2- year storm event, the associated CN coefficient can be back-calculated because the rainfall distribution is known in detail, i.e., it was specified as the NRCS Type II distribution. The CN is computed for the condition immediately after the peak when 65.5 percent of the rainfall has occurred (Miller 2006, 2). The operating equation used in Miller s paper to determine the CN for a 3.25-inch extensive green roof is: Q = (P IA*S)2 / (P + (1-IA)*S): where Q = cumulative runoff in inches S = 1,000/CN 10 or CN = 1000/(S + 10) CN = runoff curve number P = cumulative rainfall in inches IA = initial capture of rainfall, assigned a value of 0.1 in this analysis 82

91 Quantifying Stormwater Runoff after Projected Green Roof Coverage To quantify stormwater runoff with the projected green roof coverage the same methodology was used as outlined to quantify existing runoff. The composite CN was calculated using the following steps (see Table 17): Create a new category for green roof coverage Assign a CN to each land use according to soil type found in Table 2-2a in TR- 55 manual (except for the green roof coverage which does not take HSG into account, only the predetermined CN of 66) Multiply the total area of each land use by the according CN Sum the three areas Sum the three products Divide the total product by the total area to get the composite CN TABLE 17: PROJECTED CURVE NUMBER WITH GREEN ROOF COVERAGE Soil and HSG Cover description CN Area (%) Product Rossmoyne, C Open space, good condition Rossmoyne, C Impervious Cover NA Green Roofs (3") Total CN (weighted) = total product/total area CN 84 Based on the projected coverage in the table above, the CN for projected conditions, assuming 100 percent green roof coverage on 80 percent of the green roofable buildings was determined to be 84, compared to the existing CN of 92. The next step was to determine the existing runoff for a 2-year storm event as discussed. Projected runoff was calculated by using rainfall (2.86 inches) and CN (84) with Table 2-1 in 83

92 the TR-55 manual (see Table 18). Based on these inputs the depth of runoff was determined to be 1.39 inches. TABLE 18: PROJECTED RUNOFF WITH GREEN ROOF COVERAGE Factors Storm #1 Frequency - yr. 2 Rainfall, P (24-hour) - yr Runoff - in The next step was to convert runoff depth into total runoff volume for the entire case study. As mentioned earlier in the chapter, to convert the runoff depth in inches to feet, 1.39 was divided by twelve giving a quotient of feet. This was then multiplied by the total area of the case study in square feet (1,540,639.47) to obtain the volume. The total volume was determined to equal 178, cubic feet or 1,334, gallons a decrease of 555, gallons or 29 percent compared to existing conditions. ANALYSIS AND FINDINGS The results of the case study show a volume decrease of 29 percent compared to existing conditions. This decrease in stormwater runoff by nearly 30 percent is extremely significant. The outcome of this study shows that by replacing eligible traditional rooftops with green roofs the university can decrease the amount of stormwater runoff for a 2-year storm by nearly one-third. These results suggest that a decrease in volume of this degree may potentially remove a high percentage of pollutants that would otherwise be carried into the stormwater system. Furthermore, a 30 percent decrease in runoff may very well likely reduce the amount of combined sewer overflows entering the Mill Creek sewershed. Although this study does not 84

93 attempt to quantify how the decrease in stormwater runoff will affect the number of CSO occurrences in the Mill Creek sewershed, the study shows that there is a strong case to investigate how this decrease in stormwater runoff could decrease CSOs. The combination of the decrease in stormwater runoff and potential reduction in CSOs implies that there may be a significant improvement to water quality should green roofs be implemented on a large scale. Again, this study did not aim to quantify the improvement to water quality. However, the literature shows that the retention capacity of green roofs has the ability to remove a significant portion of water pollutants by eliminating their transport. Additionally, a decrease in CSOs will result in reduced levels of bacteria, namely e. Coli, found in human feces. Levels of phosphorus and nitrogen may also be reduced with the decrease in CSOs. As mentioned previously, even if the number of CSO occurrences was not reduced, water quality could show an improvement simply due to the absence of nearly one-third of stormwater that would otherwise enter the system. By preventing roughly 30 percent of stormwater from entering the system, pollutants are simultaneously prevented from contaminating nearby waterbodies. However, reduction in stormwater runoff and potential reduction in CSOs are not the only benefits that this study highlights. Of perhaps equal importance is the degree by which green roofs transform the university s level of open space and habitat. The university s existing conditions show that nearly three-fourths of the campus is impermeable. The campus currently has acres in impervious cover, translating into 73 percent of total 85

94 land cover (see Table 19, Figure 2 and Map 6). Pervious cover makes up the remaining 27 percent with a total of acres. Buildings account for roughly 44 percent of the total amount of impervious surfaces while sidewalks, roads and parking lots make up the remaining 56 percent. TABLE 19: EXISTING IMPERMEABLE COVER Type Square feet Acres Percentage Pervious 3,166, % Impervious 8,374, % Total 11,540, % FIGURE 2: PERCENTAGE OF EXISTING IMPERVIOUS AND PERVIOUS COVER Buildings currently account for acres or 32 percent of land cover at the university while the predominant land cover is comprised of sidewalks, roads and parking lots which account for acres or 41 percent of the total area. Open space, which includes grass, and other vegetated areas, is the smallest category and accounts for acres or 27 percent of total land cover at the campus (see Table 20, Figure 3 and Map 4). The limited open space that is present on campus is primarily located along the campus peripheries, most likely for the aesthetic street views. 86

95 TABLE 20: EXISTING LAND COVER Type Square feet Acres Percentage Open Space (grass vegetation) 3,166, % Pavement (sidewalks, roads, parking lots) 4,728, % Buildings 3,645, % Total 11,540, % FIGURE 3: EXISTING LAND COVER The limited open space, as well as its placement on campus, is extremely fragmented (see Map 6). However, with the addition of the green roofs, the percentage of open space increases by nearly 25 percent making the proportion of impermeable surfaces to permeable surfaces nearly equal (see Map 5). The increase in vegetative cover transforms the campus from a grey fragmented environment to one with a high level of connectivity and significant vegetative cover (see Map 6 and 7). This increase in vegetative cover is important for all of the environmental, aesthetic and economic benefits outlined in chapter three. For this reason it is imperative that the university look at the cost and benefits holistically should they consider green roof implementation for the campus. 87

96 MAP 4: EXISTING TYPE OF LAND COVER AT THE UNIVERSITY OF CINCINNATI 88 Source: CAGIS 2007

97 MAP 5: PROJECTED LAND COVER AT THE UNIVERSITY OF CINCINNATI 89 Source: CAGIS 2007

98 MAP 6: EXISTING IMPERMEABLE COVER AT THE UNIVERSITY OF CINCINNATI 90 Source: CAGIS 2007

99 MAP 7: PROJECTED IMPERMEABLE COVER AT THE UNIVERSITY OF CINCINNATI Source: CAGIS

100 TABLE 20: COMPARISON OF EXISTING AND PROJECTED COVER CONDITIONS Existing Conditions Projected Conditions Type Square feet Acres Percentage Square feet Acres Percentage Pervious 3,166, % 5,972, % Impervious 8,374, % 5,567, % Total 11,540, % 11,540, % FIGURE 4: PROJECTED LAND COVER WITH GREEN ROOFS As noted earlier, the green roof coverage projected in this study has the ability to transform the land cover at UC by nearly doubling the amount of pervious land at the university (see Table 20 and Figure 4). This increase in permeability translates to a significant decrease in stormwater runoff. Under the green roof scenario, the total volume of stormwater runoff was decreased by 555, gallons or 29 percent compared to existing conditions. With green roof implementation the university is projected to contribute 1,334, gallons in stormwater runoff compared to existing runoff conditions which contribute about 1,890,064 gallons of stormwater runoff for a 2-year storm event. 92