Assessing Utah's 90th percentile storm to understand bioretention design and cost to meet Utah's modified stormwater management rule

This research dissects and improves understanding of the implications of Utah's modified stormwater permit rule. By analyzing the 90th percentile (90th-P) storm depth calculation and modeling an urban watershed on the foothills of Salt Lake City, Utah; this project provides initial guidance to local governments and developers to plan and design bioretention facilities. The 90th-P storm depth calculation was analyzed through the manipulation of four variables: location, inclusion of snow, sample time period, and depression storage value. Analysis shows that the 90th-P storm depth was between 0.47 to 0.80 inches for eleven different stations across the state of Utah, between 0.54 to 0.69 inches for several 5-year sample time periods, and between 0.53 and 0.60 inches for varying depression storage values. These values are significantly different from the State of Utah Division of Environmental Quality, Division of Water Quality's (Division) suggested 90th-P storm depth of 0.6 to 0.7 inches. It is recommended that permit requirements are adjusted or expanded to accommodate these findings. The maximum storm depth recommended by the Division was modeled in a Storm Water Management Model (SWMM) to predict the size and cost of bioretention facilities needed to control the 90th-P storm depth. Bioretention facilities occupied anywhere from 0.16% to 1.09% of the subcatchment. It was found that the sizing is highly dependent on the runoff volume before bioretention implementation. This runoff volume is affected by many variables including impervious area, upstream characteristics, slope, and existing pervious areas. Additional research is need to fully understand this runoff volume. This research can guide developers to abide by these new permit restrictions.

ASSESSING UTAH’S 90TH PERCENTILE STORM TO UNDERSTAND BIORETENTION DESIGN AND COST TO MEET UTAH’S MODIFIED STORMWATER MANAGEMENT RULE by Sierra A. Jensen A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Civil and Environmental Engineering Approved: ______________________________ Dr. Steven J. Burian, PhD, PE Thesis Faculty Supervisor _____________________________ Dr. Michael E. Barber, PhD, PE Chair, Department of Civil and Environmental Engineering _______________________________ Dr. Michael E. Barber, PhD, PE Honors Faculty Advisor _____________________________ Sylvia D. Torti, PhD Dean, Honors College May 2018 Copyright © 2018 All Rights Reserved ABSTRACT This research dissects and improves understanding of the implications of Utah’s modified stormwater permit rule. By analyzing the 90th percentile (90th-P) storm depth calculation and modeling an urban watershed on the foothills of Salt Lake City, Utah; this project provides initial guidance to local governments and developers to plan and design bioretention facilities. The 90th-P storm depth calculation was analyzed through the manipulation of four variables: location, inclusion of snow, sample time period, and depression storage value. Analysis shows that the 90th-P storm depth was between 0.47 to 0.80 inches for eleven different stations across the state of Utah, between 0.54 to 0.69 inches for several 5-year sample time periods, and between 0.53 and 0.60 inches for varying depression storage values. These values are significantly different from the State of Utah Division of Environmental Quality, Division of Water Quality’s (Division) suggested 90th-P storm depth of 0.6 to 0.7 inches. It is recommended that permit requirements are adjusted or expanded to accommodate these findings. The maximum storm depth recommended by the Division was modeled in a Storm Water Management Model (SWMM) to predict the size and cost of bioretention facilities needed to control the 90th-P storm depth. Bioretention facilities occupied anywhere from 0.16% to 1.09% of the subcatchment. It was found that the sizing is highly dependent on the runoff volume before bioretention implementation. This runoff volume is affected by many variables including impervious area, upstream characteristics, slope, and existing pervious areas. Additional research is need to fully understand this runoff volume. This research can guide developers to abide by these new permit restrictions. ii TABLE OF CONTENTS ABSTRACT ii AKNOWLEDGEMENTS iv INTRODUCTION 1 METHODS 4 RESULTS 15 RECOMMENDATIONS AND FUTURE WORK 24 REFERENCES 26 iii ACKNOWLEDGEMENTS The completion of this report would not be possible without the guidance from my research mentor Dr. Steven Burian, department chair Dr. Michael Barber, peer Hessam Tavakol-Davani, and all the faculty in the Civil and Environmental Engineering Department at the University of Utah. Thank you for your support. iv 1 INTRODUCTION City expansion and urbanization is accompanied by an increase in impervious surfaces which can cause historically larger volumes of stormwater runoff with increased pollutant levels. Water that once infiltrated into the soil, now flows across roads and parking lots picking up anything in its path. Pollutants typically found in urban runoff include debris, oil and grease, sediments, heavy metals, and nutrients including nitrogen and phosphorus [1]. This water eventually makes its way to rivers, lakes, and groundwater which can lead to an increase in eutrophication and cause more frequent algal bloom outbreaks downstream. Polluted waters affect local farmers who rely on water sources for irrigation in addition to the general population who may use the water bodies for recreation or drinking. Algal blooms affect 65% of United States estuaries costing the nation nearly $2.2 billion dollars per year [2] [3]. These problems are most common in large cities which are challenged to control large stormwater volumes with high pollution concentrations. This research focused on the state of Utah and some of the water problems it has recently faced. With Utah as one of the fastest growing states in the nation, it has had problems dealing with stormwater volumes and polluted downstream waters. Utah’s population growth rate has increased steadily over the last six years. From 2011 to 2012 the population increased at a rate of 1.4%. More recently from 2016 to 2017 it reached a maximum increase of 1.9% [4]. In 2018 the population of Utah reached just over 3.1 million people. Although Utah is the 13th largest state in the U.S. by land mass, 80% of its population lives on a 100-mile stretch of Interstate-15 [5]. This creates a highly 2 concentrated urban area that continues to attract more people. This concentrated population growth has caused recent water problems. Salt Lake County and Utah County share the same drainage: the Jordan River, which, like many locations across the nation, has had outbreaks of toxic algal blooms in the last few years [6]. Across the nation, many cities are looking to solve these water problems by implementing systems such as Green Infrastructure (GI) or Low Impact Development (LID). There have been several cities that have been able to implement these practices. Chicago has seen success through Green Roofs, Los Angeles through rainwater harvesting, Philadelphia through permeable pavers, Phoenix through constructed wetlands, and Denver through rain gardens [7] [8]. These successes, especially in the semi-arid climates of Phoenix and Denver, are a testament to the success that Utah will be able to achieve through LID. There are a few places across Utah which have already implemented GI. On the University of Utah campus, rain gardens have been implemented near the civil engineering building to capture and filter stormwtaer runoff. And in South Salt Lake, the newly built light rail features rain gardens to control and treat stormwater. Additionally the LDS Church Conference Center in downtown Salt Lake features a green roof, and the Natural History Museum has a permeable parking lot. These projects will be important models for other developments as GI becomes important and necessary across the state. Utah has taken the first steps to introduce more LID projects which will hopefully mitigate stormwater quantity and quality problems across the state. The State of Utah Department of Environmental Quality, Division of Water Quality (Division) first 3 published the Small Municipal Separate Storm Sewer Systems (MS4) General Utah Pollutant Discharge Elimination System (UPDES) Permit Number UTR090000 in August of 2010. The permit outlined certain responsibilities of MS4s which included organizing a Storm Water Management Program and maintaining records and reporting. In September 2016, a modification to the permit was issued requiring on-site retention of the 90th percentile (90th-P) storm event for new and redevelopment projects that disturb greater than one acre [9]. This permit modification requires developers to evaluate the application of an LID strategy. LID attempts to mimic nature to decrease runoff and provide natural stormwater treatment through methods including bioretention facilities, rain gardens, vegetated rooftops, rain barrels, and permeable pavements. In general, this permit requires stormwater to be controlled on site through infiltration, evapotranspiration, and/or harvest and reuse. These modifications will be in full effect across the state on March 1, 2019. The implications of this permit have not been quantitatively assessed. The 90th-P storm is not a depth that is easily found in hydrologic tables. Additionally, the methods to reach this calculation are not standardized by the Division, making the application and results dependent on user choices and thus uncertain. In April of 2016, the Division released some guidelines to understand the calculation of the 90th-P storm depth [10]. While this document is helpful to perform calculations, it is not a legal document. The permit modification simply states that MS4s must control the 90th-P storm, but does not describe the steps to calculate this value. Developers will want to build the smallest GI facility while still controlling the 90th-P storm to reduce overall project cost. Without these methods written out in the permit itself, developers have an avenue to use a different 4 method to calculate this storm volume and as a result use a smaller depth to represent the 90th-P storm. One objective of this research is to examine the 90th-P storm across different parameters to understand the variability and sensitivity of this value. Four different variables will be analyzed: location, inclusion of snow, sample time period, and depression storage value. This research will then break down the relationship between bioretention size and other characteristics of the subcatchment. A small urban watershed on the University of Utah campus has been modeled in a Storm Water Management Model (SWMM) and is used to estimate bioretention cell design. Lastly, the cost of the bioretention facilities will be reported. These collected results will lead to the final goal which is to increase understanding of the permit implications across the state of Utah. METHODS DATA COLLECTION All precipitation data was collected from the National Oceanic and Atmospheric Administration (NOAA) formerly known as the National Climatic Data Center (NCDC) [11]. Daily precipitation data were exported from eleven stations across the state. Most of these stations have a record history dating back to 1948 with the exception of Provo where the record began in 1980. These stations track many variables including precipitation and snowfall depth. For each snowfall depth, NOAA provided a converted precipitation depth. 5 90th PERCENTILE STORM DEPTH Division of Water Quality Method. The Division hasoutlined methods to calculate the 90th-P storm depth [10]. These guidelines were supported by technical guidance from the Environmental Protection Agency (EPA) in addition to standard practices from the Center for Watershed Protection [12] [13]. Each source outlines methods to calculate various percentile storm events. These methods were incorporated into the Division’s methods. The Division’s method to calculate the 90th-P storm event is as follows. First, “obtain long-term daily rainfall data” which the EPA recommends to be at least 30 years [12]. Next, sort the data low to high and remove all snowfall and small storm events (less than 0.1 inches). Lastly use the PERCENTILE function in excel to calculate the 90th-P storm depth. The division estimates that 90th-P storm depth to be between 0.6-0.7 inches for all regions across the state [10]. This method will be used as the control and one variable will be manipulated at a time to understand the sensitivity and variability of this method. Location. The first variable explored was the storm depth across the state. Figure 1 shows eleven stations and their distribution across the state of Utah. Seven of the eleven sites are along I-15, and the remaining four are spread across other smaller cities. The station info is described in Table 1. The Division’s method was performed for all available data. The following four parameters defined the methods to evaluate location:     NOAA Station(s): All Utah Stations 1-11 Snow Storms: Excluded Date Range: Any data available from 1948-2018 Assumed depression storage value: 0.1 inch 6 Figure 1. NOAA Weather Stations Across Utah Table 1. NOAA Weather Stations Summary -111.8033 Elevation (ft) 1460 Start Date 1/1/1948 41.2439 -111.9466 1325.9 1/1/1948 2/13/2011 40.7781 -111.9694 1287.8 1/1/1948 4/12/2018 40.26231 -112.08973 1487.4 10/1/1950 2/28/2018 PROVO BYU 40.2458 -111.6508 1392.9 10/1/1980 4/8/2018 USC00422578 EPHRAIM 39.3583 -111.5991 1675.8 9/1/1949 4/9/2018 7 USC00425654 MILFORD 38.40168 -113.01624 1526.7 1/1/1948 4/10/2018 8 USW00023170 38.36921 -110.71447 13109.4 1/1/1948 4/12/2018 9 USW00093129 37.7086 -113.0944 7/1/1948 4/12/2018 10 USC00420738 HANKSVILLE CEDAR CITY MUNICIPAL AIRPORT BLANDING 37.613 -109.4847 1/1/1948 4/11/2018 11 USC00427516 ST. GEORGE, 37.119 -113.6068 1/1/1948 4/13/2018 # NOAA ID Name Latitude Longitude 1 USC00425186 41.7456 2 USC00426404 3 USW00024127 4 USC00422696 LOGAN UTAH ST U OGDEN PIONEER PH, UT US SALT LAKE CITY INTERNATIONAL AIRPORT FAIRFIELD 5 USC00427064 6 End Date 4/12/2018 7 Season. The Division’s method excludes snowstorms, which do not immediately produce runoff, from the 90th-P storm calculation. Many storms in Utah will produce snow rather than rain. Snow is a large contribution to the total precipitation in Utah. Eventually snow will melt and contribute to stormwater runoff. This section seeks to understand if including snowmelt in the 90th-P storm depth will change the value. The following four parameters defined the methods to evaluate the inclusion of snowfall:     NOAA Station(s): All Utah Stations 1-11 Snow Storms: Included Date Range: Any data available from 1948-2018 Assumed depression storage value: 0.1 inch rain or equivalent snow after melting Time. The next variable is the sample time period. In this urbanized world, there is evidence of climate and weather patterns differing from years past. According to a newsletter published by the EPA in August of 2016, climate change will cause different precipitation patters in Utah. For example, it certainly has changed annual snowpack and has caused snowpack to melt earlier [14]. The 90th-P storm calculation from 80 years ago will differ from a calculation over the last decade. This section of research explores how this precipitation differs depending on the time period evaluated. Two different analyses were made. First, 5-year blocks of time were analyzed and the 90th-P storm was calculated according to the Division’s method. Second, the 90th-P storm depth was calculated from first a 10-year and then up to a 70-year sample. This data was compared to understand the difference between older data and data from the last decades. The EPA states that only data over the last 30-years should be included in percentile calculations. The following four parameters defined the methods to evaluate time period: 8     NOAA Station(s): SLCIA, Station #3 Snow Storms: Excluded Date Range: Any data available from 1948-2018, evaluated over varying time periods. Assumed depression storage value: 0.1 inch Depression Storage. The Division has stated in their methods that small storm events (typically less 0.1 in) should be removed from the 90th-P storm calculation. These events do not immediately produce runoff and are captured in depression storage. According to the American Society of Civil Engineers (ASCE) in their guide to Urban Stormwater Management Systems, a typical depression storage value ranges from 0.05 to 0.10 inches for impervious surfaces. For lawns, which this study area has a small amount of, the value can range from 0.10 to 0.20 inches [15]. This variable will be evaluated by removing different sized storm events to understand how the depression storage value affects the 90th-P storm depth. The following four parameters defined the methods used in this section:     NOAA Station(s): SLCIA station #3 Snow Storms: Excluded Date Range: Any data available from 1948-2018 Assumed depression storage value: 0.02, 0.04, 0.06, 0.09, 0.10, 0.11, and 0.12 SWMM of URBAN WATERSHED Model Setup. EPA SWMM 5.1.012 was selected as the modeling platform for this study. SWMM has the capability to model LID such as bioretention cells. SWMM allows the user to design a bioretention cell directly in the software by adjusting surface, soil, storage, and drainage components. A previously developed model constructed by Youcan Feng and later adjusted by Hessam Tavakol-Davani was used in this research. Both were students of Civil and Environmental Engineering at the University of Utah. The model 9 was first used in research studying the potential of GI in a semi-arid urban environment [16]. The study area is shown in Figure 2. Figure 2. Study Area with Pervious and Impervious Areas Shown Youcan Feng defined the following parameters to build this model. The model is of an urban watershed located on the University of Utah campus. Salt Lake City has a semi- 10 arid climate. It received an average 16.1 inches of precipitation annually from 1981 to 2010, and had an average air temperature of 52.7 °F [16]. The USDA’s Web Soil Survey identified the primary soil group to be Bingham gravelly loam. This soil has a hydraulic conductivity of 0.354 in/hour [17]. From a nearby USGS groundwater station, the water table was measured to be 125.5 ft below the land surface. Land cover was determined and verified by Youcan Feng using imagery from Utah Automated Geographic Reference Center (AGRC). Average building and tree height were estimated to convert wind speeds. A storm drainage system directs runoff into Red Butte Creek which runs south of the project area. The system was divided into catchments based on terrain, locations of storm drain inlets, curb and gutters, and other features. These site characteristics were verified with several site visits. There were a few modifications made to fit this model to the scenario in this specific research. The model was adjusted so that the 90th-P storm would most accurately represent the Division’s methods. The depression storage values were adjusted to 0.1 inches for pervious surfaces in all subcatchments. All model inputs were inspected and verified using best engineering judgment. All areas, drainage paths, percent impervious, and slopes remained as they existed in the previous model. Design Storm Distribution. The previously calculated 90th -P storm depth was next modeled in a typical rainfall distribution for a 24-hour storm. The Farmer-Fletcher (1971) design storm distribution is a short, high-intensity rainfall event. Typically, 1-hour, 3hour, and 6-hour storms are modeled using this distribution. Farmer and Fletcher’s work later expanded using precipitation data gathered in the Great Basin Experimental Area 11 (GBEA) allowing further development of a distribution that represented longer 12-hour and 24-hour storm events. This modified Farmer-Fletcher distribution, later called the GBEA, is based on local data collection and is believed to be the best available storm distribution for Utah storms lasting 24 hours [18]. This design storm can be summarized with a dimensionless distribution over a dimensionless time step. To calculate the time step, the total storm duration (24 hours) is multiplied by 60 minutes/hour and then divided by the total number of time steps (48 units). Table 2 shows the dimensionless table for the GBEA design storm distribution. Table 2. GBEA Storm, Dimensionless Time Incremental Cumulative Step Precipitation Precipitation 0 0.0000 0.0000 1 0.0010 0.0010 2 0.0025 0.0035 3 0.0040 0.0075 4 0.0044 0.0119 5 0.0045 0.0164 6 0.0046 0.0210 7 0.0050 0.0260 8 0.0058 0.0318 9 0.0062 0.0380 10 0.0063 0.0443 11 0.0065 0.0508 12 0.0070 0.0578 13 0.0075 0.0653 14 0.0080 0.0733 15 0.0090 0.0823 16 0.0100 0.0923 17 0.0110 0.1033 18 0.0115 0.1148 19 0.0130 0.1278 20 0.0140 0.1418 21 0.0160 0.1578 22 0.0190 0.1768 23 0.0250 0.2018 24 0.0300 0.2318 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 0.0500 0.0600 0.0650 0.0675 0.0700 0.0690 0.0650 0.0500 0.0350 0.0280 0.0230 0.0210 0.0190 0.0180 0.0170 0.0155 0.0150 0.0145 0.0140 0.0130 0.0110 0.0100 0.0090 0.0087 0.2818 0.3418 0.4068 0.4743 0.5443 0.6133 0.6783 0.7283 0.7633 0.7913 0.8143 0.8353 0.8543 0.8723 0.8893 0.9048 0.9198 0.9343 0.9483 0.9613 0.9723 0.9823 0.9913 1.0000 12 The incremental precipitation is equal to the proportion of rainfall during a certain time step. The cumulative precipitation is the sum of all previous incremental values. To find a specific value for a storm even with a known depth, take the product of the incremental precipitation and the total storm depth. The 90th-P storm was distributed as a GBEA 24-hour storm. Although the exact value for the 90th-P storm depth is unknown for this study area, the maximum value of 0.7 inches given by the Division was assumed as a conservative estimate. Figure 3 shows the 24hour distribution of this design storm. This distribution was formatted in a *.dat file which could be easily read by the SWMM. Incremental Precipitation (in) 0.06 0.05 0.04 0.03 0.02 0.01 0.00 0 5 10 15 20 Time (hour) Figure 3. GBEA, 24-hour, 90th Percentile Design Storm Distribution Bioretention Design. The bioretention design is controlled in two places in the SWMM. It is controlled in the LID Control Editor in addition to the LID Usage Editor which is specific to each subcatchment. First, the LID Controls were designed. The bioretention 13 cell has four components: surface, soil, storage, and drain. There was no drain in these bioretention cells; precipitation would instead infiltrate to the groundwater table. A summary of the bioretention design can be found in Figure 4. A berm height of 6 inches is standard for smaller bioretention cells. The vegetation volume is an assumed value based on guidelines provided by the EPA SWMM handbook [15]. The surface roughness and surface slope are constants based on the site properties and were provided by previous research [16]. The seepage rate of 0.6 inches per hour was found in the USGS Soil Survey database [17]. This is a conservative estimate for all the soils in the area. All other values including porosity, wilting point, conductivity, suction head, etc. were determined in previous research [16]. Figure 4. Bioretention Cell Constants Next the LID Usage Editor was used to iterate different bioretention sizes for each subcatchment. The GBEA 90th-P storm distribution was added, and the bioretention area was iterated until zero runoff was achieved in all the subcatchments. In this scenario of zero runoff, all precipitation would be controlled on-site and the permit requirements would be met. It was assumed that 90% of the impervious area would be routed to the 14 bioretention cells. Due to this assumption, there were a few subcatchments where the runoff could not be reduced to zero, but rather 0.1 inches. COST ESTIMATION In 2009, the Water Environment Research Foundation released a set of cost estimation spreadsheets for LID. Rain garden (or bioretention cell) costs were estimated using this spreadsheet. Method A was used which relies on the drainage area and performs a simple cost estimate. Figure 5 shows all assumed costs used in the estimate. The only input that was modified was the garden area and drainage area. Figure 5. EPA Rain Garden Cost Estimator, Simple Cost Next the annual maintenance schedule was defined. It was decided that maintenance and installation would be performed professionally for all sites. This means that every year the bioretention cell would have a one-person crew perform two hours of maintenance. The visit would cost about $31/hour for labor plus and an additional $10 for material costs. These costs were accurate in 2019, and according to the Bureau of Labor Statistics, there is a total inflation increase of 16.1% from 2009 to 2018 [19]. Accounting for this inflation, the cost of each visit today is equal to $84/visit. Additionally, infrequent visits 15 are also added to the estimate. These visits will occur every three years at a cost of $541 per visit and every five years at a cost of $260 per visit. These visits would take care of larger maintenance activities such as tilling the soil and replacing mulch. These costs were estimated over a 50 year life-cycle cost and then converted to the present value of the cost. RESULTS 90th PERCENTILE STORM DEPTH Location. The Divison has suggested that the 90th-P storm depth is equal to 0.6-0.7 inches for all locations across the state. As expected it varied more than the recommended value set by the Division. Figure 6 shows the spread of 90th-P storm depths at the locations across the state. The shading is the range that the Division has stated to be the average 90th -P storm depth across the state. Only four of the eleven stations are in this range. Figure 6. 90th Percentile Storm Depths Across Utah 16 Season. Snow does not affect the 90th-P storm depth a significant amount. Some locations had the same 90th-P storm depth when snow was included where others had an increase or decrease. This is likely because the snowstorms were distributed similar to rain events. By eliminating all snow events (small and large), the 90th-P storm depth did not change much. Figure 7 shows the comparison across the state. As mentioned previously, the blue shading shows the 90th-P storm depth suggested by the Division. Figure 7. 90th-P storm Depth Calculation Including and Excluding Snowmelt Time. The time period from which precipitation data is sampled is very important. It can affect the result significantly. Figure 8 displays how much this depth can vary from one year to another. In the early 1960s, the 90th-P storm depth was less than 0.55 inches; whereas, just 10 years later it spiked up to just under 0.65 inches. A different analysis was performed which evaluates the validity of older data. Figure 9 shows how the calculation varies depending on how much data is sampled. Each line represents a time 17 period that the 90th-P storm was evaluated over. The 90th-P storm depth was relatively constant for several years, but when only the last thirty years were sampled, a there is an obvious shift. This is significant because it shows that time period matters and will affect the calculated 90th-P storm depth. 90th Percentile Storm Depth (in) 0.70 0.65 0.60 0.55 0.50 0.45 0.40 1948 1958 1968 1978 1988 Year Figure 8. 90th Percentile Storm in 5-year Time Blocks 1998 2008 2018 18 90th Percentile Storm Depth (in) 0.65 0.63 0.60 1948-2017 1988-2017 1998-2017 2008-2017 0.58 0.55 1948 1958 1968 1978 1988 1998 2008 2018 Year Figure 9. 90th-Percentile Storm Depth by Time Period Depression Storage. As expected, the smaller the depression storage value, the more smaller data that is kept in the sampel set, and the smaller the calculated 90th-P storm depth. Figure 10 displays the calculated 90th-P storm depth based on different depression storage values. According ASCE’s guidelines, the depression storage value can range from 0.05 to 0.10 for impervious surfaces [15]. The darker shading displays this range, and the lighter shading displays the Division’s recommended range. The figure displays that if terrain were to have a depression storage value anything less than 0.1 inches, the 90th-P storm depth would fall below 0.6 inches and be less than the Division’s recommendation. 19 90th Percentile Storm Depth (in) 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0 0.02 0.04 0.06 0.08 0.1 0.12 Depression Storage (in) Figure 10. 90th Percentile Storm Based on Varying Depression Storage Values SWMM OF URBAN WATERSHED A clear trend was not apparent in the sizing of these bioretention cells. Table 3 provides a summary of the subcatchments, the runoff before LID implementation, and the bioretention size that eventually controlled the 90th-P storm depth. Not all of the subcatchments in the area produced runoff. Most of the subcatchment runoff was routed to pervious areas. Table 3 shows all of the subcatchments that produced runoff. These pervious areas infiltrated some of the water without the help from added LID. Each bioretention cell was sized to the closest 5 square feet. 20 Table 3. Summary of SWMM Bioretention Sizing Subcatchment Number Total Area (ac) Total Area (ft2) % Impervious Impervious Area (ft2) No LID, Runoff (in) Bioretention Area (ft2) 30 31 33 34 35 36 37 38 40 41 42 43 44 45 46 47 48 50 51 52 53 3.86 0.18 0.6 0.26 0.13 0.25 0.12 0.13 0.74 0.2 0.19 0.14 0.54 0.4 0.27 0.54 0.25 0.16 0.37 0.98 0.13 168,142 7,841 26,136 11,326 5,663 10,890 5,227 5,663 32,234 8,712 8,276 6,098 23,522 17,424 11,761 23,522 10,890 6,970 16,117 42,689 5,663 85 90 90 90 90 90 90 90 70 90 90 90 40 60 50 80 90 10 90 70 90 142,920 7,057 23,522 10,193 5,097 9,801 4,704 5,097 22,564 7,841 7,449 5,489 9,409 10,454 5,881 18,818 9,801 697 14,505 29,882 5,097 0.08 0.03 0.15 0.04 0.02 0.05 0.05 0.04 0.08 0.04 0.10 0.04 0.03 0.02 0.10 0.04 0.04 0.02 0.03 0.02 0.09 1015 30 445 50 10 65 35 30 270 35 90 30 70 40 70 120 50 20 60 70 55 It was challenging to find a trend that described the bioretention sizing. Figure 11 shows the relationship between the bioretention area and the total subcatchment area. There is no linear, exponential, or other obvious trend. Several different relationships were tested including bioretention size versus impervious area and runoff depth versus percent impervious. The main conclusion that can be made from this graph is that most subcatchments up to 30,000 square feet can be controlled using a bioretention cell less than 100 square feet. There are many variables that will affect this estimation and additional models will be necessary to fully understand this trend. 21 1200 Bioretention Area (ft2) 1000 800 600 400 200 0 0 5,000 10,000 15,000 20,000 25,000 Total Subcatchment Area (ft2) 30,000 35,000 Figure 11. Bioretention Area versus Total Subcatchment Area Figure 12 displays the relationship between the bioretention area and the runoff volume before LID implementation. There is a positive linear relationship as expected. As the runoff volume increases, a larger bioretention basin is needed to control this volume. 1200 y = 3350x + 21.4, R² = 0.9777 Bioretention rea (ft^2) 1000 800 600 400 200 0 0 0.05 0.1 0.15 0.2 Runoff Volume (acre-in) Figure 12. Bioretention Area versus Initial Runoff Depth 0.25 0.3 0.35 22 The runoff volume is the variable that is difficult to find. Additional research is necessary to understand this value and predict depths in different scenarios. COST ESTIMATE A simple cost estimate was performed. Initially the goal was to create a graph that developers could use to understand the cost of installation when attempting to control the 90th-P storm on a certain land size. Runoff volume is less dependent on the area and more dependent on a combination of other variables. Figure 13 shows a graph that compares total subcatchment area with total present value of bioretention cost over 50 years. Unfortunately, subcatchment size is not the main component to affect bioretention size, and so no clear trend can be found. This graph does show however that a majority of subcatchments can be controlled by a $5,000 bioretention cell. Figure 14 shows the relationship between bioretention size and both capital cost and maintenance cost. About 80% of the cost for these bioretention cells comes from the 50-year maintenance and only 20% comes from the initial installation. Present Value of Total Cost over 50 Years 23 $25,000 $20,000 $15,000 $10,000 $5,000 $0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Total Subcatchment Area (ft^2) Figure 13. Total Subcaatchment Cost versus Total Subctchment Area $20,000 Initial Capital Costs Cumulative Present Value Cost (USD) $15,000 $10,000 $5,000 $0 0 100 200 300 400 Bioretention Cell Size (ft^2) Figure 14. Capital Costs and 50-year Life-Cycle Cost versus Bioretention Cell Size 500 24 RECOMMENDATIONS AND FUTURE WORK The implications of Utah’s stormwater permit modification were explored through a sensitivity analysis of the 90th-P storm depth calculation and through a SWMM of a small urban watershed on the University of Utah campus. The 90th-P storm depth calculation was explored through the manipulation of four variables: location, inclusion of snow, sample time period, and depression storage value. The 90th-P storm depth varied significantly from the value recommended by the Division. The 90th-P storm depth was between 0.47 to 0.80 inches for the eleven stations evaluated, between 0.54 to 0.69 inches for different 5-year sample time periods for the Salt Lake City International Airport (SLCIA) station, and between 0.53 and 0.60 inches for varying depression storage values at the SLCIA. The inclusion or exclusion of precipitation as snow did not affect the 90thP storm depth. This evaluation of the 90th-P storm depth could be improved through many different methods. First, it is important to understand how missing data may affect the calculation. If data points are flagged in the system, how should users deal with this data? Next, the methods to include snowstorms could be improved. By comparing snowstorms to temperature data, the date of snowmelt could be found. The methods in this research treated all snowstorms like rainstorms and assumed immediate runoff. And finally, the constant addition of more data throughout the state of Utah and specific to certain project areas would improve this calculation overall. The maximum storm depth recommended by the Division (0.70 inches) was modeled in a Storm Water Management Model (SWMM) to predict the size and cost of bioretention 25 facilities needed to control the 90th-P storm depth. Bioretention facilities occupied anywhere from 0.16% to 1.09% of the subcatchments that produced runoff. It was found that the bioretention sizing is highly dependent on the runoff volume before implementation. This runoff volume is affected by many variables including impervious area, upstream subcatchment characteristics, slope, and existing pervious areas. Additional research is needed to fully understand this runoff volume and its relationship to these characteristics. In conclusion, this research has provided additional information for developers to understand the implications of Utah’s new stormwater management permit. It is recommended that clear guidelines for the 90th-P storm depth calculation are written and published in parallel to the permit. This will make the calculation dependent on clear guidelines rather than user decisions. Also recommended is further research to understand the relationship between subcatchment characteristics and LID cost. These recommendations will be valuable to developers across the entire state of Utah. 26 REFERENCES [1] City of Los Angeles, Development Best Management Practices Handbook: Low Impact Development Manual. Los Angeles, 2011. [2] W. K. Dodds et al., "Eutrophication of U.S. Freshwaters: Analysis of Potential Economic Damages," vol. 43, no. 1, pp. 12-19, 20009. [3] S. Bricker et al., "Effects of Nutrient Enrichment in the Nation's Estuaries: A Decade of Change," 2007. [4] Courtney Tanner. (2017, December) The Salt Lake Tribune. [Online]. https://www.sltrib.com/news/politics/2017/12/20/utahs-population-is-growingfaster-than-every-state-except-for-idaho-and-nevada-say-new-census-estimates/ [5] (2018) World Population Review. [Online]. http://worldpopulationreview.com/states/utah-population/ [6] Amy Joi O'Donoghue. (2016, July) Deseret News. [Online]. https://www.deseretnews.com/article/865658346/Effects-of-algal-blooms-continueto-spread-throughout-Wasatch-Front.html [7] Christopher Economides, Tess Russo, and Upmanu Lall. (2014) Green Infrastructure: Sustainable Solutions in 11 Cities across the United States. [Online]. http://water.columbia.edu/files/2014/04/Green_Infrastructure_FINAL.pdf 27 [8] United States Environmental Protection Agency. [Online]. https://www.epa.gov/green-infrastructure/green-infrastructure-semi-arid-west [9] UPDES General Permit for Discharges from Small MS4s: Permit Modification, 2016, Permit No. UTR090000. [10] Jeanne Riley. (2016, April) Utah Department of Environmental Quality. [Online]. https://deq.utah.gov/legacy/topics/fact-sheet/docs/handouts/2016/05may/calculation90-percentile-storm-event.pdf [11] National Centers for Environmental Informaion. [Online]. https://www.ncdc.noaa.gov/cdo-web/datasets [12] United States Environmental Protection Agency, "Technical Guidance on Implementing the Stormwater Runoff Requirements for Federal Projects under Section 438 of the Energy Independence and Security Act," Washington, D.C., 2009. [13] Tom Schueler , David Hirschman, Michael Novotney, and Jennifer Zielinski. (2016, March) Center for Watershed Protection. [Online]. https://owl.cwp.org/mdocsposts/urban-subwatershed-restoration-manual-series-manual-3/ [14] United States Environmental Protection Agency, "What Climate Change Means for Utah," EPA 430-F-16-046, 2016. 28 [15] Lewis A. Rossman, "Storm Water Management Model User's Manual Version 5.1," Office of Research and Development, Environmental Protection Agency, Cincinnati, EPA-600/R-14/413b, 2015. [Online]. www2.epa.gov/water-research [16] Youcan Feng, Steven Burian, and Christine Pomeroy, "Potential of Green Infrastructure to Restore Predevelopment Water Budget of a Semi-Arid Urban Catchment," Journal of Hydrology, vol. 542, pp. 744-755, 2016. [17] United States Department of Agriculture Natural Resource Conservation Service. Web Soil Survey. [Online]. https://websoilsurvey.sc.egov.usda.gov/App/WebSoilSurvey.aspx [18] "City of Draper Drainage Design Criteria," Draper City, 2012. [19] (2018) US Department of Labor Bureau of Labor Statistics. [Online]. https://www.bls.gov/cpi/