A Watershed Approach to Mitigating Stormwater Impacts (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

64 Determining Off-Site Mitigation Options in the WBSMT The WBSMT utilizes several key datasets, user inputs, and computational approaches to link project impacts and mitigation targets in order to provide potential off-site mitigation alternatives. This chapter outlines the methods for linking project impacts to off-site mitigation options, along with a clearly defined procedure for ranking and prioritizing mitigation opportunities. This chapter provides a brief discussion on how DOTs might use the WBSMT or similar approach for assessing on-site and off-site BMPs along with alternative mitigation options. By defining the project area, stormwater management goals, and watershed characteristics, baseline conditions and mitigation targets can be established and used to evaluate both in-kind and out-of-kind stormwater management alternatives. This allows the user to establish a ranking and prioritization system for selecting the preferred watershed- based mitigation options for a specific transportation project. Project Characterization The first step of project characterization is to specify the total project acreage, the annual aver- age daily traffic (AADT) count, and the area-weighted average hydrologic soil group (HSG) found within the project. The project area must be defined according to the stormwater management requirements and in many cases may include the area of disturbance and the existing roadway. For example, in a roadway expansion as shown in Figure 6, the project area may consist of both the striped areas, which represent the expansion, and the solid gray region, which represents the existing roadway. There are three ranges of AADT that can be specified for a project: (1) 0–30K, (2) 30–90K, and (3) 90K+. Default runoff concentrations for each of these AADT classifications are based on median stormwater concentrations from highways summarized from the HRDB and NSQD as reported in Geosyntec et al. (2015). The area-weighted average or dominant HSG for the project refers to the pervious ROW areas and therefore is a relatively insensitive parameter if the project area is mostly impervious. If the average project HSG is unknown, a procedure for determining this parameter is discussed under Soils Characteristics. The next step is specifying the acreages of various land types (highway, urban, natural land, cropland, and pasture) which were replaced by the project and each land type’s percent imper- viousness. In this step, the acreage of the existing pre-project impervious area is included in the highway land type. The degree of imperviousness affects the project runoff coefficient and the Linking Project Impacts to Mitigation Alternatives in the WBSMT C H A P T E R 6 To find the WBSMT, go to the TRB website and search for NCHRP Research Report 840.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 65 estimated footprint of in-kind on-site BMPs and, with the acreage, is used to calculate on-site loadings pre- and post-BMP. Figure 7 shows the project characterization table in the WBSMT. Stormwater Management Goals Specification of stormwater management goals is the final project characterization step and consists of five questions used to develop stormwater impact mitigation priorities. These relate to: 1. Reduced downstream flooding 2. Reduced hydromodification impacts 3. Reduced downstream TSS and particulate bound pollutant loads and concentrations 4. Reduced downstream nutrient impacts 5. Improved/minimized downstream temperature impact The stormwater management goals are given a score of one or zero depending on if the pro- vided answer is a yes or no, and are used to rank out-of-kind mitigation measures discussed in the section Watershed Assessment – Stormwater Management Goals in Appendix A. Watershed Characterization Geographic Location The geographic location of the project is specified by four descriptors: state, county, HUC-12 ID, and rain gauge. Dropdown lists for each of these descriptors are populated in the WBSMT Watershed Characteristics sheet by selecting the project’s region in the map of the contiguous United States. Care should be taken to select the rain gauge that is geographically close to the Figure 6. Example of a typical roadway expansion project area. Figure 7. Project characterization in the WBSMT.

66 A Watershed Approach to Mitigating Stormwater Impacts project and whose average annual precipitation most closely matches that of the project loca- tion. If the HUC-12 ID is not known, a lookup link is provided to help identify and select the HUC-12 ID for the project location (see Figure 8). Once the HUC-12 ID has been specified, most of the watershed characteristics are automatically populated from associated EnviroAtlas national datasets. Two watershed characteristics must be approximated by the user: 1) the watershed area-weighted average or dominant HSG and 2) the percent of watershed natural land cover not associated with the project that is disturbed or actively eroding (if any). These are discussed further in this chapter. EnviroAtlas Datasets Eight of the 169 available national datasets in EnviroAtlas were leveraged for use in the WBSMT and are summarized in Table 20. The EnviroAtlas datasets are indexed using the HUC-12 subwatershed ID furnished by the user. In addition to the eight national datasets, the subwatershed total acreage is used to calculate the acreage of existing land cover. Soils Characteristics Successful application of the WBSMT requires the input of the area-weighted average or dom- inant HSG in the project area. If unknown, the average HSG can be determined using the USDA Note: The HUC-12 boundaries are outlined with a thick black line. Source: USGS Figure 8. Lookup of HUC-12 ID using the USGS’s Hydrography National Map Viewer.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 67 NRCS Web Soil Survey. The soil data explorer application allows the user to define the area of interest and generate a soil report of the soil’s estimated hydrologic properties. This provides the HSG of each soil unit (as shown in Figure 9). Disturbed Areas Disturbed areas are unstable areas within the watershed that are not associated with active construction. These areas could be a landslide or logged area that is an ongoing source of ero- sion and sediment to the receiving water. Disturbed areas are not provided in the EnviroAtlas national datasets and must be estimated by the user. The estimated annual soil loss (tons/acre) of disturbed areas within the watershed can be estimated by entering appropriate values for the five parameters of RUSLE2 (Revised Universal Soil Loss Equation, Version 2) into the Advanced sheet of the WBSMT. An example of the RUSLE2 parameters and estimated soil loss based on the parameters is presented in Figure 10. Hyperlinks are provided in the sheet to the RUSLE2 website, which includes guidance for an explanation on the use of the equation and the definitions of each parameter. Within the WBSMT, if there are disturbed areas identified by the user, the opportu- nity score for upland stabilization out-of-kind mitigation measure becomes non-zero. See the subsequent section titled Off-Site Assessment (Out-of-Kind Mitigation) for details. In addition to the RUSLE2 parameters, the default mass fractions of model constituents [TSS, total phosphorus (TP), and total nitrogen (TN)] can be altered in the Advanced sheet of the WBSMT to estimate pollutant loading from the identified disturbed areas. The TSS concentration Dataset Description Average annual precipitation (in./yr) An estimate of the total precipitation (snow and rain) that falls within the 12-digit HUC each year. (Note that this value may be adjusted by the user if it does not agree with local data or if there is interest in evaluating climate change scenarios.) Percent cropland Based on the Cropland Data Layer created by the USDA, the layer uses all crop types to depict the percentage of cropland in the 12-digit HUC. Percent natural land cover Percent of total land that has natural land cover (forests, shrubs, grasslands, barren land, and wetlands) within each 12-digit HUC. Excludes agriculture and developed land. Percent natural land cover in buffer Percent of naturally covered land within 30 meters of streams, rivers, or other hydrologically connected water bodies within each 12-digit HUC. Percent pasture Percent of land managed as pasture (areas planted for livestock grazing or seed/hay crop production) in each 12-digit HUC. Percent potentially restorable wetlands Estimation of the percent of land that may be suitable for wetland restoration within each 12-digit HUC. Percent urban area Estimation of the percent of developed land within each 12-digit HUC including open spaces, parks, golf courses, single family homes, multifamily housing units, retail, commercial, industrial, and associated infrastructure. Urban areas are not confined to city limits. Percent impervious area Percent of total impervious land within each 12-digit HUC; includes buildings, roads, and sidewalks. Source: USEPA 2015 Table 20. EnviroAtlas datasets used for watershed characterization in the WBSMT.

68 A Watershed Approach to Mitigating Stormwater Impacts is the percent of the erosional sediment mass that is predicted to be delivered to the receiving water as suspended sediment. This is also known as the sediment delivery ratio and depends on numerous factors, such as basin size, slope, soil type, particle size distribution, vegetative cover, gully and channel density, sediment storage areas, and rainfall depth and intensity. Walling (1983) explored the challenges associated with estimating sediment delivery and summarized various approaches that researchers often use. He postulated that ratios could range from a few percent to over 100% (due to bank and channel erosion), recommending that site-specific Note: Includes breakdown of the project’s soil types and their respective HSG ratings. Source: NRCS 2015 Figure 9. Screenshot from NRCS’s Web Soil Survey webpage. Figure 10. RUSLE2 parameters for estimating annual soil loss in disturbed areas.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 69 assessments be completed to the extent possible to arrive at locally-derived delivery ratios for a watershed. The default ratio in the WBSMT is 10 percent. Users are encouraged to modify this default with more regionally appropriate datasets, if available. The default TP concentration in soils is based on the median phosphorus content in A-horizon soils across the United States as estimated by Smith et al. (2014). The default TN concentration in soils is based on the median nitrogen content of the mean nitrogen mass concen- tration in 0-100 cm soil profiles as reported by Batjes (1996) for the most abundant soil classes of North America as reported by the Food and Agriculture Organization of the United Nations (FAO/UNESCO 1992). These default concentrations are provided in Figure 11. As with the TSS mass fraction, the TP and TN mass fractions should be updated with local data if available. Defining Load Reduction Targets Four desired project target reduction percentages are entered by the user for stormwater runoff volume, TSS, TP, and TN to estimate expected regulatory BMP performance goals. The project loads are considered to originate solely from the highway land use and are calculated as follows: ( ) = × ×Project Area Runoff Volume ft HA AAP RC (7)3 ( ) = × ×Project Area Pollutant Load lbs Project Area Runoff Volume PCC CF (8) where HA is highway area (sq.ft), AAP is average annual precipitation (ft), RC is long-term aver- age runoff coefficient, PCC is pollutant characteristic concentration (mg/L), and CF is conversion factor from mg/L to lbs/ft3, 6.2428E-05. Long-term runoff coefficients have been calculated for each rain gauge provided in the WBSMT using each HSG and impervious values ranging from 0 percent to 100 percent. The coefficients were calculated using a 30-year SWMM simulation over one-acre drainage areas for 340 rain gauges across the United States. Using the average HSG and percent imperviousness provided under proj- ect characterization, the closest matching runoff coefficient is determined and applied in Equation 7. The pollutant characteristic concentrations are presented in Figure 12 and found in sheet 9-Advanced in the WBSMT where the default concentrations can be updated using site-specific or regional data, if available. The WBSMT uses AADT to determine the concentrations to use as the value for characteristic pollutant concentration. Conducting On-Site BMP Assessment Supported BMP Types and Performance Assumptions The In-kind, On-site Options sheet allows the user to set the design basis and preferred storm- water BMPs for on-site stormwater mitigation efforts under typical circumstances when on-site BMPs are feasible and desirable for the project. For the purpose of sizing, the entire project area Figure 11. Sediment mass loading fractions for constituents.

70 A Watershed Approach to Mitigating Stormwater Impacts is assumed to be routed to an on-site BMP (defined in sheet 1-Project Details). If none or only a portion of the project would actually be routed to an on-site BMP, this portion can be defined in Step 7, Analysis. The stormwater design basis is defined as the percentage of average annual runoff volume that would be captured and/or reduced by a BMP designed according to the sizing criteria applicable to the project. If sizing criteria are based on a design storm depth or intensity, these percentages can be determined by continuous simulation modeling or by using one of the NCHRP Report 792 BMP evaluation tools (Taylor et al. 2014). By default, the percent capture and percent volume reduction are set to 80 percent and 20 percent, respectively. The other on-site BMP information required is the preferred BMP type and the ratio of the BMP footprint to the effective impervious drainage area. The ratio is used to size the footprint of the BMP using Equation 9 below: ( ) = × ×BMPFootprint ac BMP FR DA I (9) where: BMP FR is the BMP footprint ratio, DA is drainage area (ac), and I is the average percent imperviousness of drainage area. There are five supported BMP types available for on-site in-kind mitigation: • Swale • Wetland • Dry pond • Bioretention • Media filter The default effluent quality and characteristics for each BMP type are summarized in Table 21. These defaults can be modified to site conditions in the sheet 9-Advanced. Effluent concentrations are based on a summary of data from the International Stormwater BMP Database (Geosyntec et al. 2015). Computations Pollutant load reductions occur by two means: (1) volume reduction and (2) stormwater treatment. The reductions are determined by first breaking up project runoff volumes into three categories: Figure 12. Land use characteristic constituent concentrations (mg/L) used in the calculation of pre-BMP constituent loadings.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 71 • Bypassed volume: runoff which is not captured by the BMP • Reduced volume: runoff which is captured by the BMP and locally infiltrated • Treated volume: runoff which is captured by the BMP, treated, and released The bypassed, reduced, and treated volumes are calculated using Equations 10, 11, and 12, respectively. These calculations also use the percent capture and percent volume reduction design bases defined by the user. Post-BMP loads (treated plus bypassed) are then calculated using Equations 8–10. All concentrations associated with the bypassed stormwater volume remain the same as the project runoff concentrations (i.e., bypassed volumes receiving no treatment). Loads determined to be a part of volume reduction are considered to be completely removed by the BMP. Loads associated with the treated volume are calculated using the appropriate BMP effluent quality concentration listed in Table 21. The post-BMP loads are then compared to the project’s target load reductions. If the targets are not met, off-site mitigation measures must be implemented to achieve the difference as described in the following section. ( ) ( )= × −Bypassed Volume ft Project Runoff Volume 100% %Runoff Capture (10)3 ( ) = ×Reduced Volume ft Project Runoff Volume %Volume Reduction (11)3 ( ) = − −Treated Volume ft Project Runoff Volume Bypassed Volume Reduced Volume (12)3 ∑( ) ( )= × ×PostBMP Load lbs Bypassed Volume PCC CF (13)bypassed All land uses where PCC is pollutant characteristic concentration (mg/L) and CF is conversion factor from mg/L to lbs/ft3, 6.2428E-05. ( ) =PostBMP Load lbs 0 (14)reduced ( ) = × ×PostBMP Load lbs Treated Volume BMP Effluent Quality CF (15)treated ( ) = + +PostBMP Load lbs PostBMP Load PostBMP Load PostBMP Load (16)total reduced bypassed treated Off-Site Assessment (In-Kind Mitigation) Off-Site Drainage Characterization Off-site mitigation occurs outside of the project area so the off-site drainage area and BMP characteristics need to be properly defined. The same procedure used for the characterization of the original project area when calculating on-site BMPs also applies to calculating the off-site drainage area. This procedure includes defining the total tributary acreage, the AADT for any Table 21. Default BMP effluent concentrations and footprint ratios. Reference Swales Wetlands Dry Ponds Bioretention Media Filter Effluent Quality (mg/L) Total Suspended Solids (TSS) 21.6 10.9 23.3 9.9 8.4 Total Phosphorus (TP) 0.171 0.091 0.197 0.24 0.089 Total Nitrogen (TN) 0.87 1.2 1.6 0.92 1.04 Characteristic BMP Footprint Ratio 0.05 0.1 0.07 0.08 0.065

72 A Watershed Approach to Mitigating Stormwater Impacts highway land use, dominant HSG, and the breakdown of the total drainage area into individual land use acreages and average imperviousness values as shown in Figure 13. Supported BMP Types and Performance Assumptions The BMP types, performance characteristics, and design basis for in-kind off-site mitigation options are identical to those for in-kind on-site mitigation, but the influent to off-site BMPs dif- fers. Default effluent quality concentrations and footprint ratios are listed in Table 21. The BMP type, footprint ratio, and stormwater design basis for the off-site BMP can be modified by the user similar to the on-site BMP options. An off-site load reduction adjustment factor is located in Step 3 of sheet 5-In-Kind Off-Site Options. The purpose of the factor is to reduce the magnitude of the effective load removed by the specified BMP to account for uncertainty in achieving an equivalent level of water quality improvement in the watershed by going off-site. The calculated load reductions, based on BMP performance characteristics, are divided by the load reduction adjustment factor to determine the adjusted load reduction. The default off-site load reduction adjustment factor is 1, indicating no adjustment, which is reasonable considering the off-site BMP may already need to be sized to treat a larger volume of water to achieve the same load reduction as an on-site BMP. Computations In-kind off-site mitigation efforts are calculated using the same overall process as in-kind on-site efforts. Unlike the project’s pre-BMP runoff volumes, which consider the entire Figure 13. In-Kind Off-Site Options from WBSMT.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 73 project area as highway, the in-kind off-site mitigation approach calculates pre-BMP runoff volumes for each specified land use (i.e., highway, urban area, natural land cover, cropland, and pasture) based on Equation 7 and the provided land use acreage and percent impervi- ousness. Pre-BMP loads are then calculated using Equation 8 and the appropriate pollutant characteristic concentrations. Load reductions occur by the two previously defined means: (1) volume reduction, and (2) stormwater treatment, using Equations 10, 11, and 12. Post-BMP loads are then calculated using Equations 13–16. The total post-BMP loading is the sum of the categories’ individual load- ings (Equation 16). The post-BMP loads are then compared to the project’s target load reductions. If the targets are not met, out-of-kind mitigation measures must be implemented to make up the difference. Off-Site Assessment (Out-of-Kind Mitigation) There are four out-of-kind mitigation measures supported in the WBSMT: • Stream improvement techniques – restoration of stream bed and banks, improved floodplain connectivity, riparian enhancements, and dechannelization/restoring channel morphology. • Upland stabilization – rehabilitation of disturbed areas that contribute to sedimentation of downstream watercourses. • Reducing impervious surface connectivity – reducing total impervious area and routing run- off from impervious areas to pervious areas to allow infiltration and filtration of stormwater. • Wetland restoration/creation – increasing wetland functionality in treating pollutants and/or construction of a wetland to treat pollutants and improve habitat. Ranking of Mitigation Measures The out-of-kind measures are ranked based on HUC-12 watershed metrics and user inputs in sheet 6-Out of Kind Options. The sheet consists of two primary sections: (1) the weighting of watershed metrics (Figure 14), and (2) the ranking of out-of-kind mitigation measures based on priority and opportunity scores (Figure 15). The user must first assign a score relating the applica- bility and health of the nine watershed beneficial uses shown in Step 2 in Figure 14. The user can assign an applicability score of 0, 1, or 2 based on the definitions in Table 22. Because beneficial uses are defined by states and not the USEPA, they have not been standardized. The user must select the beneficial uses that most closely match those defined for the receiving water. The values assigned to the beneficial uses directly affect the default priority scoring of ecosystem services in the watershed, shown in Step 3 of Figure 14. The default scores reflect the estimated dependence that ecosystem services have on the distribution of land uses within the watershed and the ben- eficial uses of the receiving water. The user can choose to implement the default priority scores or override them by entering a User Priority Score based on a more detailed assessment of ecosystem services. The scores are normalized and must sum up to 1.0. After ecosystem services have been ranked, mitigation measure priorities must be assessed (Figure 15). The default priorities are estimated from the relationship between mitigation measures and watershed processes that have been weighted by ecosystem services priorities, out- standing loads to be mitigated, and stormwater management goals. The default opportunity scores are based on an assessment of the land use areas that are most amenable to the various mitigation measures. These land use areas are used as surrogate measures of potential out- of-kind mitigation opportunities. A detailed description of the priority and opportunity score computations is provided in Appendix A: Out-of-Kind Mitigation Assessment Approach. The final score for each mitigation measure provided in Step 5 in Figure 15 is the product of the user

Figure 14. Ranking of watershed beneficial uses and ecosystem services. Figure 15. Ranking of mitigation measures and final out-of-kind mitigation scoring. Applicability Score Description 0 Beneficial use is not applicable or important 1 Beneficial use is applicable or important 2 Beneficial use is applicable and impaired Table 22. Beneficial use applicability score definitions.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 75 priority score and user opportunity score (Equation 17) divided by the sum of all of the mitiga- tion measure’s final scores (Equation 18). = ×Final Score Mitigation Options Priority Opportunity Score (17)MM MM MM where MM is mitigation measure. ∑ ( )=Final Weighted Score Final Score Final Score (18)MM MM MMAll MM Analysis of Mitigation Efforts Sheet 7-Analysis provides the user with an overview of the project loadings and target load reductions and the opportunity to review mitigation measure results and alter the levels of implementation between on-site, off-site, and out-of-kind mitigation as shown in Figures 16 and 17. The Watershed Estimated Load/imperv area and Watershed Estimated Load/total area are two metrics provided in the overview to help the user assess the degree of imperviousness in the HUC-12 watershed and its effect on pre-BMP loading. Load Reduction Calculations All of the BMP footprint areas and load reductions presented in Figure 16 are calculated by scaling the original BMP footprint area (Equation 9) and post-BMP loads (Equation 16) by the mitigation measure’s percent implementation. The load removal ranking is the sum of the BMP’s ability to meet the total desired load reduction for each load type and is calculated using Equation 19. If a BMP met all load reduction targets, it would receive a load removal ranking of 4.0. ∑ ( )( )=Load Removal Rank BMPLoad Reduction lbs or cu. ft Target Load Reduction lbs or cu. ft (19) All Load Types If there are still outstanding loads after the implementation of on-site and off-site in- kind mitigation measures, an outstanding equivalent impervious area is calculated for out-of-kind mitigation (Figure 17). This area is calculated for each load type using Equa- tion 20. The acreage reported in the analysis sheet is the maximum determined for the four load types. ( ) ( )=     Outstanding Equivalent Imperv Area ac Outstanding Load lbs or cu. ft Watershed Estimated Load Impervious Area lbs ac or cu. ft ac (20) Recommended Out-of-Kind Mitigation Using the outstanding equivalent impervious acreage, the estimated footprint of each out- of-kind mitigation measure is calculated by multiplying the outstanding equivalent acreage by the mitigation measure ratio (Equation 21). The mitigation measure ratio is equal to the user footprint ratio input in sheet 6-Out of Kind Options and is shown in Figure 15. ( ) = ×Estimated Footprint ac OEIA Mitigation Measure Ratio (21) where OEIA is outstanding equivalent impervious area (ac).

76 A Watershed Approach to Mitigating Stormwater Impacts Figure 16. Project loading overview and in-kind load reduction reporting from sheet 7-Analysis.

Linking Project Impacts to Mitigation Alternatives in the WBSMT 77 Figure 17. Out-of-kind mitigation overview from sheet 7-Analysis. The in lieu multiplier applies only to areas with an established in lieu program (e.g., a munici- pal stormwater fee in lieu program) and the multiplier should be defined by the program. The default is set to 1.1. The in lieu estimated footprint is the estimated footprint calculated using Equation 21 multiplied by the in lieu multiplier. The final score is determined using Equations 17 and 18 and is discussed in the Off-Site BMP Assessment section. The rank of each out-of-kind mitigation measure is dependent on the measure’s final score, with a higher score equating to a higher rank as defined in Table 23. The actual footprint of the out-of-kind mitigation measure is the estimated footprint multiplied by the percent implementation. Final Score Rank Given Rank 1 H 2 H 3 M 4 L Note: L = low, M = medium, H = high Table 23. Ranking of out-of-kind mitigation measures.

78 A Watershed Approach to Mitigating Stormwater Impacts Reporting The report sheet in the WBSMT summarizes all of the project conditions and load reduction targets, and quantifies the total load reduction (TLR) percentage attributed to each mitigation type (on-site in-kind, off-site in-kind, and out-of-kind in the form of outstanding reductions), as shown in Figure 18. This information is broken down into three sections. The first section, Project Location and Watershed Characteristics, contains all of the relevant project details and watershed characteristics necessary to determine the project loadings. The second section, Load Reduction Targets, reiterates the design target reductions (percentages) and resultant load reduction values (lbs or cu.ft). The final section in the sheet, Mitigation Measures and Banking Status Summary, presents a breakdown of each mitigation measure’s load reductions and how much of the TLR these reductions constitute. Results are presented visually in the form of bar graphs and pie charts so that the user can easily and quickly ascertain which type of mitigation measure provided the highest reduction per load type. A table at the bottom of the sheet outlines the calculated footprints for each specified mitigation measure for the project. Figure 18. Summary of project mitigation found in sheet 8-Report.