Guidelines for Evaluating and Selecting Modifications to Existing Roadway Drainage Infrastructure to Improve Water Quality in Ultra-Urban Areas (2012)

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.

69 Section 4 describes the available BMP options from which retrofit planners must identify, adapt, and ultimately select BMPs that can meet the retrofit goals and constraints. The initial screening of BMP options should be based primarily on the treatment capabilities of the BMPs using the following criteria: 1. Does the BMP include relevant treatment processes that address the target pollutants and conditions of concern; and 2. Does the BMP provide an acceptable level of treatment performance? This section describes approaches for evaluating the treatment effectiveness of BMPs and presents general BMP effectiveness information and ratings that can be used for initial screening of the BMP options and identification of candidate BMPs. 5.1 Evaluating the Relevant BMP Treatment Processes 5.1.1 Fundamental Unit Operations to Guide BMP Selection A common criterion for BMP selection is the ability to meet stipulated BMP performance criteria for pollutants of concern or surrogate pollutants, for example 80% TSS removal. However, a more fundamental approach is to base BMP selection on the unit operations (UOPs) or processes that will be effective for the highway POCs (Strecker et al., 2005). The basic steps are: 1. Characterize the form of the pollutants and conditions of concern (e.g., particulate, dissolved, immiscible, chem- ical species, particle size, specific gravity, volume, and discharge); 2. Identify the fundamental unit operations that are effective for the pollutant forms and conditions of concern; and 3. Select BMPs that include the appropriate unit operations. Characterizing retrofit BMPs by their unit operations allows stormwater managers to first identify which treatment processes are needed to achieve retrofit treatment objectives and then to select the BMPs that can feasibly and/or effec- tively provide those processes. Table 5.1 summarizes the unit operations that are inherent in the retrofit categories. The four fundamental unit operations categories—hydrologi- cal, physical, biological, and chemical—are described in the following paragraphs (for additional details, see the WERF Guidance Manual on Critical Assessment of Stormwater Treatment and Control Selection Issues). Hydrological Operations: Hydrologic operations are common and significant unit operations for stormwater treatment, where the goals include reducing peak flows, reducing runoff volumes, and mitigating hydromodifica- tion. Hydrologic operations alter the discharge hydrographs, which affects water quality through direct changes in pollut- ant loadings and/or concentrations. • Flow attenuation: Flow attenuation is the reduction of peak discharge events (peak shaving) by detention, con- veyance, and interception. Stormwater detention and gradual release of runoff is the most common approach. Flow attenuation occurs during conveyance by extend- ing the travel time and by infiltration where it occurs. Interception is a form of detention storage that occurs on vegetation, and to a lesser extent on pavement, especially porous and open-graded pavements. Detention retrofits, infiltration retrofits, and capture and use retrofits are the most effective approaches for flow attenuation. • Flow reduction: Flow reduction is a decrease in total run- off volume to storm sewers and receiving waters through infiltration, evaporation, retention, and use. Detention S e c t i o n 5 Evaluating BMP Effectiveness

Table 5.1. Ranking of BMP retrofit categories according to unit operation effective level. Fundamental Process Category Unit Operations or Processes Targets C a t c h B a s i n R e t r o f i t G S R D R e t r o f i t H y d r o d y n a m i c R e t r o f i t O i l - W a t e r S e p a r a t o r V e g e t a t e d F i l t e r R e t r o f i t M e d i a F i l t e r R e t r o f i t D e t e n t i o n R e t r o f i t s I n f i l t r a t i o n R e t r o f i t s P a v e m e n t R e t r o f i t s C a p t u r e & U s e R e t r o f i t s Typical Location in Treatment Train P P P P P/S S S S P S Hydrology / Hydraulics Flow attenuation (hydrograph matching) Peak shaving 0 0 0 1 1 0 4 to 5 3 to 5 1 3 to 5 Reduce total volume of runoff Volume reduction 0 0 0 0 3 to 5 0 1 to 2 (A) 0 (U) 5 1 to 2 3 to 5 Flow-duration control and design Volume attenuation 0 0 0 1 1 0 3 to 5 4 to 5 0 3 to 4 Physical Operations Screening Trash, debris, coarse sediment 1 to 5 4 to 5 3 to 5 0 1 to 3 0 to 3 0 0 0 0 Flotation and skimming Trash, debris, oil & grease 0 0 3 to 5 3 0 0 0 0 0 0 Sedimentation/Settling Sediment, debris, sorbed metals and nutrients 0 to 2 0 2 to 3 2 to 3 3 to 5 0 3 to 5 2 0 2 Filtration Sediment, sorbed metals and nutrients 0 to 3 0 0 0 3 to 5 5 0 4 to 5 3 to 4 0 Sorption processes (absorption) Dissolved metals, nutrients, organics 0 0 0 0 3 to 5 3 to 5 0 3 to 5 0 0 Volatilization/Aeration Oxygen demand, VOCs, PAHs 0 0 1 0 1 to 3 0 3 (A) 0 (U) 0 0 0 Physical disinfection (heat and UV radiation) Pathogens 0 0 0 0 1 0 2 (A) 0 (U) 1 (A) 0 (U) 2 1 Biological Processes Microbially mediated transformations Dissolved nutrients, organics, metals 0 0 0 0 0 to 2 2 2 to 4 (A) 0 (U) 2 0 0 Uptake and storage Nutrients, organics, metals 0 0 0 0 0 to 2 1 2 (A) 0 (U) 1 0 0 Chemical Processes Sorption processes Nutrients, organics, metals 0 0 0 0 3 4 to 5 1 to 2 (A) 0 (U) 4 0 0 Flocculation / Precipitation Fine sediment, nutrients 0 0 0 0 0 0 0 to 2 (A) 0 (U) 0 0 0 Chemical disinfection (ozone, chlorine) Pathogens 0 0 0 0 0 0 0 0 0 0 P – Primary treatment, S – Secondary treatment; (A) Aboveground application; (U) Underground application 0 - BMP does not include unit process OR is not recommended for that process due to operations and maintenance issues (e.g., a filter should not be used to screen) 1 - BMP includes unit process, but likely provides poor effectiveness. 2 - BMP includes process, but likely provides marginal effectiveness. 3 - BMP designed to include unit process, but other BMPs may be more effective. 4 - BMP is specifically designed to include unit process, but the design is not optimal. 5 - BMP is specifically designed to include unit process and is among the best alternatives available. Adapted from Strecker et al. (2005)

71 facilities, infiltration BMPs, and vegetated BMPs provide significant infiltration and evaporation losses. Retention basins and cisterns permanently capture stormwater, where it is reduced through infiltration and evaporation, or used for irrigation or other activities. Infiltration retrofits, veg- etated filtration retrofits, and capture and use retrofits are the most effective flow reduction approaches. • Hydromodification control: The goal of hydromodification control is to mitigate alterations to the runoff hydrograph by reducing flows and volume to better match predevelop- ment discharge conditions, usually through attempts to rep- licate the predevelopment flow-duration relationship. One approach is through volume attenuation, which requires both flow attenuation and flow reduction. For example, LID practices that reduce runoff volume may be combined with detention facilities to control discharges. Flow-duration con- trol is a more accurate sizing and flow control approach that seeks to match the full range of pre- and post- development flows over the long-term, rather than controlling individual design events. Detention retrofits and infiltration retrofits are the most effective approaches for mitigating modifica- tions to the predevelopment hydrograph. Physical Operations: Physical operations are treatment processes based on physical mechanisms such as screening and sedimentation. Physical unit operations are the domi- nant forms of treatment in most stormwater BMPs. • Screening: Screens are intended to physically exclude sol- ids (including trash, particles, debris, and organisms) that have a dimension larger than a selected screen opening (Strecker et al., 2005). Coarse screening is used to remove larger trash and debris in a way that will not clog storm- water control facilities and will be easy to maintain. Typi- cal applications are upstream trash racks for pretreatment (e.g., upstream of detention or media filtration retrofits). Other retrofits specifically target trash and debris using a variety of screening devices such as bags, socks, nets, or screens located at stormwater inlets, within stormwater conveyances, or at the outfalls. • Floatation and skimming: Floatation processes utilize the net buoyancy of some pollutants or pollutants attached to floatable materials by trapping particles, debris, or fluids with a specific gravity of less than 1. The floating pollut- ants are subsequently removed by skimming (Strecker et al., 2005). Targeted highway POCs are typically trash and debris and separate-phase oils and grease. BMPs that include floatation processes are oil-water separators, baf- fled tanks, hydrodynamic systems, and inverted elbows used in catch basins (water quality inlets). • Sedimentation: Sedimentation is the settling of particles due to density differences. Under quiescent conditions, particle removal is a function of particle density, particle size, and fluid viscosity, the latter being affected by temper- ature. Under turbulent conditions, removal is dependent upon surface hydraulic loading, particle settling velocity, and shear stress (Urbonas and Stahre, 1993). Sedimenta- tion is most effective for higher pollutant concentrations (> 400 mg/L) and larger particle sizes (> 50 µm) (Urbonas and Stahre, 1993). However, the efficiency of any settling system is generally a function of residence time, which depends on the size and design of the system (Huber et al., 2005). Therefore, small-footprint BMPs such as oil-grit separators, small baffled tanks, and hydrodynamic sepa- rators that have small volumes and short retention times will often have reduced sedimentation performance. More effective sediment removal occurs with detention retrofits that have longer detention times. • Filtration: Filtration is the physical straining of particles through porous media such as sand, gravel, and permeable asphalt. Filtration processes target removal of particulates and highway POCs that adhere to particulates including metals, organics, and nutrients. Particulates are lodged and trapped in the pore space of the media, which over time can cause clogging and crusting of the porous media, reducing performance and necessitating maintenance. However, with proper pretreatment and maintenance, filtration pro- cesses are very effective for particulates, particularly fine particulates that can be difficult to remove by sedimenta- tion processes. Filtration processes are incorporated into proprietary and non-proprietary media filters, infiltration facilities, bioretention BMPs, and porous asphalts. • Volatilization/Aeration: Volatilization and aeration refer to mass-transfer across air-fluid interface. Volatilization is the process whereby pollutants that are dissolved in water or present in an immiscible phase (oils) vaporize and escape to the atmosphere. Pollutants of concern that are susceptible to volatilization are VOCs and SVOCs. Vola- tilization processes are most prominent in BMPs such as surface detention facilities that have large air-water inter- faces and good airflow and wind action, and are exposed to elevated temperature or direct sunlight. Aeration is the process of entraining air in the water column, primarily to increase dissolved oxygen and to decrease the BOD. Aeration would be applicable for discharges and receiving waters with an excessive oxygen demand, for example to promote nitrification of ammo- nia. Stormwater BMPs typically do not include active aer- ation processes (in some cases, wet ponds have included active aeration systems). Passive aeration is most effective in BMPs with large air-water interfaces and good airflow and wind action, such as surface detention facilities. • Physical agent disinfection: Physical agent disinfection is the destruction of stormwater-borne pathogens (and

72 indicators) through non-chemical agents including sun- light, ultraviolet (UV) light, and heat. Physical disinfection is suitable for BMP applications where the influent has low turbidity; otherwise pretreatment is required, e.g., media filtration followed by UV disinfection. Heat and UV from sunlight also promote disinfection. Shallow detention facilities that allow penetration of sunlight and promote heating will be a less friendly environment for pathogens. Biological Operations: Biological operations use living organisms (plants, algae, and microbes) to transform or remove organic and inorganic pollutants. • Microbially mediated transformations: Microbially mediated transformations are chemical transformations by bacteria, algae, and fungi, for example, the degradation of organic compounds and the denitrification of nitrate. Microbial processes are most significant for converting dissolved nutrients (nitrogen), metals, and some organic compounds. The transformations occur slowly, on the order of days, and are greatly affected by temperature. Thus, the most effective BMPs are wet detention facilities with long detention times (wet basins and constructed wetlands), located in warm weather climates. Bioreten- tion with an intentional saturated condition has also been examined for nitrate reduction. • Uptake and storage: Uptake and storage refers to the assim- ilation of organic and inorganic pollutants by plants and microbes. Uptake and storage can be used to remove dis- solved metals, nutrients (phosphorus and nitrogen), and organic compounds. The processes occur where soil prop- erties and water quality are adequate to support vegetative and microbial growth, and residence times are adequate. BMPs that include uptake and storage processes include constructed wetlands, wet basins, and bioretention. Chemical Operations: Chemical operations are treatment processes based on chemical interactions. Chemical opera- tions used in stormwater treatment include sorption, coagu- lation and flocculation, and chemical agent disinfection. • Sorption: Sorption refers to both absorption and adsorp- tion. Absorption is a physical process wherein a substance of one state is incorporated into the structure of another substance of a different state. Adsorption is the physio- chemical adherence or bonding of ions and molecules (ion exchange) onto the surface of another molecule. Sorption is a key process for the removal of dissolved highway pol- lutants of concern including dissolved metals, soluble nutrients, and soluble organics. Sorption processes are integrated with media filtration and infiltration BMPs, and media are often selected on the basis of their sorp- tion capacity. For example, sorption media used in BMPs include perlite, compost, zeolite, and activated carbon. Engineered media may also use iron, manganese oxide, ion exchange media, and media coatings. The effective- ness of sorption processes in BMPs varies greatly and depends on the chemical properties of the media, the sur- face area of the media, and the system design. BMPs that are designed to enhance sorptive processes are proprietary and non-proprietary media filters; bioretention BMPs; and compost-amended filters, swales and filter strips. Sorption processes can be combined with sedimentation to seques- ter pollutants. • Precipitation, coagulation, and flocculation: These indi- vidual processes occur rapidly and are accelerated through chemical additions. Precipitation is the transformation of a pollutant from a dissolved state to a solid state. Coagula- tion is the destabilization of colloidal particles so that par- ticle growth can occur, and flocculation is the process by which fine particles collide to form larger particles that can be removed through filtration and/or settling. Precipita- tion, coagulation, and flocculation processes are used in advanced stormwater BMPs for treatment of fine and col- loidal particulates, dissolved metals, and phosphorus. The effectiveness of these processes depends on the chemicals being used, pH, temperature, and hardness. These pro- cesses can potentially generate significant quantities of sludge, which must be properly handled and disposed. • Chemical agent disinfection: Chemical disinfection refers to the destruction of stormwater-borne pathogens through the use of chemical agents such as chlorine and ozone. The effectiveness of chemical disinfection depends on the dose, mixing, contact time, chemical characteristics of the influ- ent, particles in the influent, and the characteristics of the target organisms. 5.1.2 Treatment Train Sequence to Maximize Effectiveness BMP rankings in Table 5.1 can assist with the identification and selection of retrofit BMPs that provide the fundamental unit operations for meeting water quality and/or quantity objectives. However, Table 5.1 also shows that BMPs may sup- ply multiple unit operations, and some unit operations may occur more effectively in some BMPs than others. As such, the placement or order of one or more BMPs within a treat- ment system should be considered to maximize the effective- ness of the retrofit design. The recommended approach is to base BMP selection and configuration on the concept of the treatment train. The treatment train targets the most easily treated pollutant first and progressively targets smaller par- ticles and pollutants that are more difficult to treat (Strecker et al., 2005):

73 1. Reduce flows from the drainage area (hydrological con- trol). Runoff volume reduction is a principal storm water control strategy because it directly reduces pollutant loads and treatment structure requirements. Along rural high- ways, volume reduction can be achieved by dispersion to the natural landscape. However, runoff dispersion is almost never practical in dense urban highway settings. Porous pavements and porous overlays can contribute to volume reduction to varying extent within the highway catchment, but these are not yet widely accepted prac- tices for dense highways. Similarly, capture and use sys- tems using cisterns are also a potential volume reduction strategy, but these too have limited feasibility due to the lack of sufficient demand for harvested water, particularly within the right-of-way/DOT control. Volume reduction for ultra-urban highways is most likely to be achieved in conjunction with downstream treatment BMPs such as infiltration practices, vegetated BMPs, and surface deten- tion facilities, in situations where these BMPs are feasible. 2. Remove bulk or gross solids (pretreatment: > 5 mm). Pre- treatment to remove bulk pollutants such as litter, debris, and other large solids (often termed “gross solids”) is imple- mented early in the treatment process to reduce the potential for clogging and/or obstruction of downstream treatment system components. Effective pretreatment can be achieved with screens and racks, GSRD systems, some types of catch basin retrofits, and hydrodynamic systems where under- ground retrofits are required or where the grade allows. 3. Remove settleable solids and liquid floatables (primary/ secondary treatment: >50 µm): Easily settleable solids and floatables are addressed by most common and core treat- ment BMPs that provide primary and/or secondary levels of stormwater treatment. Primary treatment removes the coarser settleable solids and floatables and includes catch basin retrofits, hydrodynamic retrofits, small vaults and settling areas, oil-grit separators, and pavement retrofits. Detention retrofits and vegetated treatment retrofits pro- vide secondary levels of treatment that are more effective for finer settleable solids. In addition, they can provide vol- ume reduction through infiltration where feasible, which can reduce all surface discharge loads of all pollutants. 4. Remove fine particulates (secondary conventional treat- ment: <50 µm): Secondary treatment BMPs that address fine sediments involve filtration processes. BMPs in this cat- egory include infiltration practices, sand filters, bioreten- tion systems, and proprietary and non-proprietary media filtration systems. Infiltration and bioretention BMPs that include volume reduction are preferred where feasible. 5. Remove dissolved, colloidal, and pathogenic constitu- ents (enhanced treatment): Enhanced treatment prac- tices address pollutants that are not effectively removed in conventional BMPs. Dissolved metals and dissolved nutrients (nitrate, soluble phosphorus) can be addressed by media filtration BMPs where the media is engineered and tailored to enhance sorption processes. BMPs that include biological processes may also be considered, such as wetlands. Coagulation/flocculation processes address the removal of colloidal particles, and disinfection pro- cesses address pathogenic pollutants. BMP selection and configuration for retrofit applications are obviously affected by site constraints, DOT policies, and maintenance requirements. But within the context of site- specific conditions, more effective retrofit treatment systems are obtained when BMPs are selected on the basis of unit operations for specific targeted pollutants, and the configu- ration of BMPs is based on a treatment train approach. 5.2 Evaluating BMP Performance When comparing treatment performance of candidate BMPs, the design engineer should consider the various met- rics that establish BMP performance as well as regulatory and certification requirements that may establish acceptable BMP types. BMP performance metrics include hydraulic perfor- mance metrics and treatment performance metrics described in the following subsections. 5.2.1 Capture Efficiency Capture efficiency (or “percent capture”) is a BMP hydrau- lic performance metric. For online BMPs, the capture effi- ciency is the fraction of total runoff that is processed and treated/managed by the BMP. Runoff volumes or flows that exceed the design capacity of the BMPs are considered untreated and are referred to as bypass or overflow. For exam- ple, detention basins typically have a water quality outlet at the bottom for treated outflows and an overflow outlet at the top of the basin for bypass flows. Some of the flows may also leave the basin through infiltration and evaporation, which are also considered treated/managed outflows. The percent capture for the detention basin is calculated by: Percent Capture = −( )[ ]100 1 V Vby t where Vt is the total influent volume and Vby is the total bypass volume leaving the basin through the overflow outlet. This is typically evaluated by conducting a long-term simulation of the performance of the system. Flow-based BMPs such as swales and media filtration sys- tems are sized on the basis of maximum discharge rate that is managed for water quality. The percent capture of flow-based BMPs is calculated with the same equation above, where flows in excess of the design discharge are considered untreated bypass

74 discharges. To calculate the percent capture, the inflow hydro- graph is integrated to determine the total inflow volume and bypass/overflow volume for flows above the design discharge. A properly designed BMP should generally result in cap- ture efficiencies of about 70% to 90% of the long-term flows from the watershed. However, in severely space-constrained or otherwise constrained situations, lower capture efficien- cies may still be appropriate in some locations as part of watershed-wide solutions. 5.2.2 Volume Reduction Volume reduction is simply the volume loss occurring in the BMP due to infiltration and to a minor extent, by evaporation. Volume reduction is both a hydraulic and treatment metric. Volume loss directly reduces the discharge volume to receiving waters, which is a key process for meeting hydromodification goals. Volume loss also directly reduces the pollutant mass dis- charged to receiving waters, providing water quality benefits. Pollutant mass reduction is particularly relevant to meeting TMDL objectives where there are mass load allocations but also for reducing the frequency of discharges in some cases. The volume reduction capacity of BMPs is controlled by the long-term infiltration rate of the underlying soils or by demand drawdown of captured stormwater. Infiltration facilities, which are designed to achieve high levels of volume reduction, must therefore be sited in areas with good infil- tration capacity. As noted earlier, finding a demand for har- vested stormwater in the highway environment is difficult, but in some cases may be possible. Other BMPs that are not sited on the basis of infiltra- tion capacity can provide significant volume reduction (Table 5.2). Table 5.2 is based on all reported events; gener- ally higher percent volume reductions will occur for smaller storms. In particular, unlined surface detention facilities (dry extended-detention basins) and biofiltration BMPs have potential volume reductions of 30% or more. BMPs with per- manent pools (e.g., wet basins and wetland basins) are gen- erally observed to have lower levels of volume reduction in the range of about 10%. Fine sediments that collect in these facilities may partially seal and impede infiltration losses. BMPs that are lined or contained in vault structures, such as hydrodynamic devices and proprietary media filters, provide no volume reduction (Strecker et al. 2004a,b). Volume reduction can provide significant hydrologic and water quality benefits to receiving waters. Designers should consider these benefits during BMP selection, when surface detention and biofiltration BMPs are feasible options. 5.2.3 Pollutant Levels Retrofit planners can use various approaches to quantify pollutant levels in BMP influent and effluent discharges. These approaches include actual monitoring and use of available/appropriate runoff concentrations for highways and other contributing land uses (e.g., Driscoll et al., 1990; Granato et al., 2003a,b). Concentration: For most highway pollutants of concern, the average concentration of pollutants in BMP influent and effluent discharges is most accurately determined through automated composite samplers. Composite samplers allow for the calculation of an event mean concentration (EMC), which is the flow-proportional average concentration of a given parameter during a storm event. An EMC is the total constituent mass divided by the total runoff volume. The EMC is frequently the most useful means of quantifying the pollutant levels in runoff events and BMP discharges. Some highway POCs are not amenable to automated sam- pling and are more commonly measured by individual grab samples. Oil and grease and coliform bacteria are parameters that are not routinely measured with automated samplers. Oil and grease tends to adsorb to tubing in samplers, and bac- teria analyses require a short holding time, typically less than the storm duration. Trash and debris are also not amenable to common measurement methods and are usually quantified on a mass load basis. Sediment: Because of the regulatory emphasis on sediment, designers should be aware of the challenges and approaches of characterizing particulate mass. Sediment is commonly mea- sured as TSS concentration, and most regulations are based BMP Category Number of Monitoring Studies 25th Percentile Median 75th Percentile Biofilter—Grass Strips 16 18% 34% 54% Biofilter—Grass Swales 13 35% 42% 65% Bioretention (with underdrains) 7 45% 57% 74% Detention Basins—Surface, Grass Lined 11 26% 33% 43% Retention (Wet) Ponds—Surface 20 2% 11% 18% Wetland Basins/Channels 11 3% 4% 5% Relative volume reduction = (Study Total Inflow Volume & Study Total Outflow Volume)/(Study Total Inflow Volume) Source: Geosyntec & Wright Water (2010) Table 5.2. Average BMP volume losses of media filtration retrofits.

75 on TSS standards. However, as discussed in Section 2.2.2, TSS measurement may under-represent particulate mass, and SSC may be a more appropriate measure. Also, a more complete picture of BMP effectiveness is gained when turbidity and PSD measurements are included as part of the BMP assessment. Pollutographs: Concentration measurements at individ- ual points in time can be useful for BMP efficiency evalu- ations. Concentrations resulting from samples collected at specific times during an event allow for the generation of a pollutograph, a plot of the pollutant concentration versus time. Pollutographs are useful for analyzing intra-event tem- poral variations, for example to determine the presence or absence of first flush by pollutant type and form. Pollutant Loads: Pollutant load is the mass of pollutant discharged over a specified duration (e.g., lbs per year). Pol- lutant loads are typically calculated by using an average con- centration multiplied by the total volume of flow over the averaging period. Thus, the load calculation depends on the method used to calculate total flow and average concentra- tion. Pollutant loads are useful for assessing the impact to receiving waters where long-term loadings can cause water quality problems outside of discrete storm events. Determin- ing pollutant load reductions in BMPs is relevant to assess- ing compliance with load-based TMDLs. For example, the Stochastic Empirical Loading and Dilution Model (SELDM), developed by the USGS and FHWA as a highway runoff qual- ity model, can be used to estimate storm flows, concentra- tions, and pollutant loads from highway facilities, as well as the resulting risk of exceeding water quality criteria in the receiving waters with and without user-defined BMPs (see http://ma.water.usgs.gov/FHWA/SELDM.htm). 5.2.4 Percent Removal Percent removal (or reduction) is a widely used BMP per- formance metric. Percent removal (or efficiency ratio) is sim- ply the percentage difference in influent and effluent pollutant levels, expressed on a concentration (e.g., influent and effluent EMC on a storm-by-storm basis) or mass load basis. Regu- latory agencies frequently use percent removal standards for BMP performance objectives and certification of proprietary devices. For example, 80% TSS reduction or 60% total phos- phorus removal are common BMP performance standards. The use of percent removal methods should (but often do not) include an appropriate non-parametric (or parametric, if applicable) statistical test to establish if differences in influent and effluent pollutant levels are statistically significant. Note that it is better to show the actual level of significance found, rather than just noting if the result was significant (assuming a 0.05 level). Parametric tests usually require transformation of the data so that tests are carried out on data with normal distributions. The most commonly observed data distribution is log-normal, so computing the mean and standard devia- tion of log transformations of the sample EMC data and then converting them to arithmetic estimates often results in a bet- ter estimate of the mean of the population due to these more typical distributional characteristics (Geosyntec and Wright Water, 2009). However, the geometric mean (the retrans- formed mean of the logarithms of data) may cause a negative bias in load estimates because the geometric mean gives more weight to low outliers and less weight to high outliers, thereby under-representing the storms that may carry the bulk of the annual load. Because of such difficulties associated with data transformations, non-parametric tests are commonly used to analyze hydrologic and water quality data because these tests do not depend on the assumption that data are from a specific probability distribution (Helsel and Hirsch, 2002). Percent removal, by itself, is not a fully adequate BMP treat- ment performance metric because it does not take into account the magnitude of the influent and effluent concentrations (Strecker et al., 2004b; Jones et al., 2008). For example, when influent pollutant levels are large, BMPs can achieve a high per- cent removal but often with poor effluent quality that would not meet overall water quality goals. Conversely, when influent pol- lutant levels are small, BMPs may achieve a low percent removal but often with good effluent quality. For example, a BMP with an influent and effluent TSS concentration of 1000 mg/L and 100 mg/L, respectively, achieves a 90 percent removal, but an effluent TSS concentration of 100 mg/L is poor. Conversely, a BMP with an influent and effluent TSS concentration of 50 mg/L and 25 mg/L achieves a 50% reduction, but with good effluent quality in terms of TSS. To counter these limitations, some regulatory agencies and testing organizations consider the magnitude of the influent concentrations by bracketing the allowable TSS influent concentrations in the calculation of per- cent removal effectiveness (e.g., NJDEP, 2009a; Ecology, 2008). Percent removal is an important criterion for comparison with regulatory criteria, but percent removal should not be used as a primary treatment performance metric, or should be used in conjunction with appropriate consideration of influent and effluent quality. 5.2.5 Effluent Quality and Effluent Probability Method Strecker et al. (2004a) recommend BMP effluent quality as the most appropriate treatment performance metric. Effluent quality as a performance metric does not have the problems associated with percent removal metrics, and effluent quality is directly comparable to other BMPs, receiving water condi- tions, and water quality objectives (Strecker et al., 2001). Geosyntec and Wright Water (2009) advocate the effluent probability method as the most useful approach for quantifying BMP efficiency. The effluent probability method is a technique

76 that provides a statistical view of influent and effluent qual- ity (Oregon State University et al., 2006). The approach is to first determine if the BMP is providing treatment by applying appropriate non-parametric (or parametric, if applicable) sta- tistical tests to indicate if any perceived differences in the influ- ent and effluent mean of the EMCs are statistically significant. Next, the cumulative distribution function of influent and effluent quality or a standard parallel probability plot is pre- pared and examined. For example, Figure 5.1 shows standard parallel probability plots for BMP influent and effluent data. A normal probability plot should be generated showing the log-transformed data of both inflow and outflow EMCs for all storms for the BMP. Equivalently, a normal probability plot can be constructed with a logarithmic concentration axis. In either case, lognormality can be evaluated by seeing if the data points can be approximated using a straight line. If the log- transformed data deviate significantly from normality, other transformations can be explored to determine if a better dis- tributional fit exists (Geosyntec and Wright Water, 2009). In Figure 5.1, the plot for suspended solids (SS) shows the BMP is effective at reducing SS across the entire range of the influent distribution. The plot for COD shows the BMP pro- vides poor removal at low concentrations (less than about 30 mg/L) but increasing levels of removal at higher concen- trations. The effluent probability method provides a clearer picture of the ultimate measure of BMP effectiveness over the full range of the influent quality. However, there is not a one-to-one correlation between the percentiles in the influ- ent data and the percentiles in the effluent data. For example, the median influent concentration and the median effluent concentrations may not occur in the EMC samples collected during the same storm. Although the influent and effluent concentrations in a probability plot are not paired values, the relative position and slope of the two populations are a good indication of the effectiveness of the BMP. 5.2.6 Scour Tests Sediment resuspension and washout caused by high dis- charge diminishes the effectiveness of BMPs. Washout is of particular concern for small-footprint BMPs with limited sediment storage capacity. Some laboratory testing protocols of manufactured devices require scour tests (NJDEP, 2009b). For example, the NJDEP protocols require scour tests to deter- mine the maximum treatment flow rate that can be conveyed through the device without causing excessive scouring (i.e., less than 10% washout at 125% of the maximum treatment flow rate). When evaluating small-footprint BMPs with limited sed- iment storage capacity, designers should strongly consider the potential for sediment flushing and should actively investigate the availability of scour test data or evaluation procedures. 5.2.7 Scaling and Performance Functions Studies have found the effectiveness of small-footprint proprietary BMPs (hydrodynamic separators, small vault sedimentation systems) for sediment removal is largely a function of the size of the device (Yu and Stopinski, 2001; Note: Data are ranked for both the inlet and outlet series and are not necessarily paired. Source: Geosyntec and Wright Water (2009) Figure 5.1. Probability plots for TSS and COD.

77 Charbeneau et al., 2004). Accordingly, researchers have applied scaling theory as a means to estimate the perfor- mance of proprietary BMPs, particularly for hydrodynamic devices, as a function of the size device. Wilson et al. (2007) used controlled testing of proprietary BMP devices to develop performance curves (see Figure 5.2). These curves relate the sediment removal efficiency to a dimen- sionless Peclet number that captures the relevant processes and dimensions of the device. The Peclet number is defined as: Pe V hd Qs= where Vs is the particle settling velocity, h is the settling length scale of the device, d is the horizontal diameter of the device, and Q is the discharge. The performance curves can, in principal, be applied in sev- eral ways (Gulliver et al., 2009). First, the curves can be used to scale the performance of devices that are measured at one size to another size. Thus, laboratory tests that are performed on small-scale devices can be scaled to full-scale devices. Secondly, the performance curves can potentially be used for BMP sizing, assuming the particle size distributions used to develop the per- formance curves are representative of actual field conditions. For example, a BMP planner with design discharge and an aver- age sediment particulate diameter (e.g., 100 µm) can use the performance curves to calculate the required device dimensions to achieve a target sediment removal efficiency (e.g., 80%). The application of scaling theory to proprietary BMPs is still a developing field. Future studies are needed to verify the predictive capabilities of performance curves in actual field installations. If the approach is found to be practical, the ability to scale the performance of manufactured devices can potentially help DOTs design and size manufactured devices. 5.3 BMP Testing Protocols and DOT Certification One of the greatest limitations of small-footprint proprietary BMPs is the lack of independent verifiable performance infor- mation (NRC, 2008). Available testing data are often inconsis- tent in monitoring protocols and reporting. In response to this limitation, a number of states and DOTs regulate the allowable use of proprietary BMPs through formal testing and certifica- tion procedures. However, testing protocols and testing pro- grams to evaluate and/or certify proprietary BMP performance are not uniform. Various testing and certification programs are established, including the following: • The Technology and Reciprocity Partnership: The Tech- nology Acceptance Reciprocity Partnership (TARP) is an Figure 5.2. Performance curves for the BaySaver Model 1k. Source: Wilson et al. (2007)

78 interstate agreement on the reciprocal evaluation, accep- tance, and approval of innovative energy and environ- mental technologies. The TARP agreement resulted in the development of minimum sampling and testing proto- cols (TARP Tier II Protocol) for evaluating BMP perfor- mance and verifying manufacturer claims (TARP, 2003). The TARP II Protocols are a common benchmark for field evaluation of proprietary BMPs. • Environmental Technology Verification Program: The USEPA Environmental Technology Verification Program (ETV) develops testing protocols and verifies the perfor- mance of innovative environmental technologies. ETV is a voluntary program that makes objective performance information available to support decision making. ETV does not endorse, certify, or approve technologies. ETV has developed testing protocols for stormwater BMPs (ETV, 2002), and verification reports and statements for a number of propriety BMPs are published on the ETV website. • New Jersey Department of Environmental Protection: The New Jersey Department of Environmental Protection (NJDEP) has established both laboratory and field test protocols for proprietary BMPs. NJDEP actively certifies BMPs based on a TSS removal standard. Laboratory test- ing of BMPs is conducted by the New Jersey Corporation for Advanced Technology (NJCAT), a non-profit public/ private partnership created to provide third-party credible and independent verification of vendors’ technology per- formance claims. Notable testing criteria include limits on influent TSS concentrations, requirements for PSD mea- surements, limits on maximum particle sizes, and sedi- ment scour tests. • Washington State Department of Ecology’s Technology Assessment Protocol: Washington’s Technology Assess- ment Protocol–Ecology (TAPE) is a field test protocol for certifying proprietary BMPs (Ecology, 2008). TAPE cer- tifies and approves proprietary BMPs in accordance with treatment performance goals as shown in Table 5.3. For TSS performance goals, PSD measurements are required to demonstrate representativeness to typical runoff condi- tions. Evaluation of field monitoring data must include a statistical analysis to show statistical significance of perfor- mance results. The status of various proprietary devices is given at http://www.ecy.wa.gov/programs/wq/stormwater/ newtech/technologies.html. • BMP monitoring protocols from the International BMP Database: The International Stormwater Best Manage- ment Practices Database Project (BMP Database) is a cooperative effort between the American Society of Civil Engineers (ASCE) and the USEPA to develop a repository of scientifically sound BMP performance monitoring data. A detailed BMP performance monitoring guidance doc- ument (Geosyntec and Wright Water, 2009) was devel- oped to provide a basis for consistent monitoring and reporting protocols for studies that are adopted into the database. The monitoring guidance document provides detailed protocols for BMP monitoring, data evalua- tion, and interpretation of BMP performance. No BMP “acceptance” or “approvals” are provided. The purpose of the database was solely to develop a database of rigor- ous and consistent monitoring and reporting protocols to improve information available for analysis and deci- sion making. The BMP evaluation protocols and testing programs above demonstrate a continuing and increasing effort to evaluate and certify the performance of proprietary BMPs. There is also an ongoing effort within the ASCE to unify evaluation and certification protocols (Guo et al., 2008). Similarly, the NRC (2008) study Urban Stormwater Management in the United States also advocated for a national testing program that would include common protocols to verify the perfor- mance of proprietary BMPs. Despite these efforts, however, current testing and certification protocols are disjointed Performance Goal Influent Requirements Effluent Requirements Pretreatment TSS = 100 to 200 mg/L 50% TSS reduction TSS < 100 mg/L TSS = 50 mg/L Basic Treatment TSS = 100 to 200 mg/L 80% TSS reduction TSS < 100 mg/L TSS = 20 mg/L Enhanced Treatment Dissolved copper: 0.003 to 0.02 mg/L Dissolved zinc: 0.02 to 0.3 mg/L Meet basic treatment condition for TSS Exceed basic treatment for dissolved copper and zinc. A numeric target is not specified. It is suggested that available data from vendors and the BMP Database be used to help determine if the device demonstrates significantly higher removal rates (Ecology, 2008). Phosphorus Treatment TP = 0.1 to 0.5 mg/L 50% TP Reduction Oil Treatment No recurring visible sheen Avg. daily TPH < 10 mg/L Max. discreet TPH < 15 mg/L Table 5.3. TAPE performance goals for proprietary BMP approval.

79 across the country. Consequently, DOTs face a range of regu- latory policies regarding the use of proprietary BMPs. The flexibility to use or test non-approved BMPs in retrofit appli- cations is a key consideration in BMP selection. 5.4 BMP Performance Data The most informative BMP performance information is data collected through DOT monitoring studies and pilot testing programs or through local/regional BMP perfor- mance assessments. In the absence of local data, the most robust BMP performance monitoring data are compiled in the International BMP Database (http://www.bmpdatabase. org/) because all monitoring studies included in the BMP Database are evaluated for consistency with monitoring and reporting protocols. The BMP Database includes more than 400 BMP studies (and continues to grow), performance analysis results, tools for use in BMP performance studies, monitoring guidance, and other study-related publications. The authors recommend engineers and designers conduct a search of the BMP Database to obtain performance data for specific regions and BMPs of interest. Table 5.4 shows median influent and effluent concentra- tions for various BMPs and common highway POCs from the BMP Database as of June 2008. The degree of pollutant removal depends on the pollutant species/form and the level of treatment provided by the BMP. Design features such as pond surface area, length-to-width ratio, vegetation and soil types, and the use of a forebay or other enhancements may affect the type and level of treatment provided. In some instances, effluent medians may exceed influent medians, such as for some nutrients and dissolved constituents. This may suggest that the BMP provides no effective treatment, or possibly the BMP is exporting pollutants from internal sources within the BMP (e.g., nutrients released from bio- filters) or due to washout of previously captured materials. BMP performance data for specific types of proprietary BMP devices and innovative BMP technologies are available from the following sources: • Manufacturers: Many manufacturers will supply indepen- dent performance test reports that have been conducted for certification purposes. • Certification agencies: Regulatory agencies that certify proprietary BMPs provide test reports through their web- sites, for example, the NJDEP: http://www.nj.gov/dep/ stormwater/treatment.html. • Massachusetts Stormwater Technology Evaluation Proj- ect: The Massachusetts Stormwater Technology Evaluation Project (MASTEP) is a web-based stormwater technolo- gies clearinghouse (http://www.mastep.net/). The website includes a searchable database of proprietary technologies including performance data summaries and reports. The database also provides scoring (0 to 4) of BMP technologies indicating whether there are sufficient TARP- compliant or similar reliable field or laboratory data to be able to evalu- ate pollutant removal efficiency claims. • Independent research studies: Independent studies have compiled and evaluated available performance data of proprietary BMPs (Brueske, 2000; Yu and Stopinski, 2001; Charbeneau et al., 2004). • DOT pilot studies and research programs: The most applicable BMP assessments are studies conducted in high- way settings. DOTs across the country regularly conduct and sponsor a variety of BMP technology assessments, including retrofit pilot studies and pilot studies of inno- vative BMPs and manufactured devices. Table 4.13 lists potential DOT resources. 5.5 BMP Evaluation Guidance The recommended criteria for evaluating the effectiveness of BMPs include the following: • The amount of runoff that receives treatment or is bypassed. The capture efficiency of the BMP affects the quantity of pollutant load reduction, as bypass discharges are untreated. BMPs are ideally sized to maximize capture efficiency to an optimal point at which further increases in size provide diminishing returns in terms of treatment and cost. In space-limited settings, the optimized capture efficiency may be highly constrained, and capture efficien- cies below the design storm standard may be warranted. • The BMP’s ability to reduce runoff volumes via infiltration and evapotranspiration. Reducing runoff volume helps to attenuate flows, mitigate hydromodification impacts, and directly decreases pollutant loads to receiving waters, which can benefit compliance with TMDL allocations. • The effluent quality of treated runoff. The ability of the BMP to achieve acceptable effluent quality for target POCs is the most direct measure of BMP effectiveness. Effective BMPs are those with effluent quality that meet regulatory water quality objectives and effluent quality that is pro- tective of receiving water quality. Estimates of expected effluent quality are ideally based on local pilot tests or BMP effectiveness studies that are representative of antici- pated conditions. In the absence of such data, general BMP effectiveness ratings shown in Table 5.5 can be used for comparative guidance. The ratings are based on published BMP performance data and the unit operating processes utilized in the BMP that are expected to reduce runoff volumes and/or contaminant levels.

Constituent Point of Discharge Dry Detention Basin (n = 25)1 Wet Pond (n = 46)1 Wetland Basin (n = 19)1 Biofilter (n = 57)1,2 Media Filter (n = 38)1,3 Hydrodynamic Devices (n = 32)1,4 Porous Pavement (n = 6)1 Total Suspended Solids (mg/L) Influent 72.65(42–104) 34.13 (19–49) 37.76 (18–53) 52.15 (41–63) 43.27 (27–60) 39.61 (22–76) xx Effluent 31.0(16–46) 13.37 (7.3–19) 17.77 (9.3–26) 23.92 (15–33) 15.86 (9.7–22) 37.67 (21–54) 16.96 (5.9–49) Total Copper (µg/L) Influent 20.14(8.4–32) 8.91 (5.3–12.5) 5.65 (2.7–39) 31.93 (25–39) 14.57 (11–18) 15.42 (9.2–22) xx Effluent 12.10(5.4–18.8) 6.36 (4.7–8.0) 4.23 (0.62–7.8) 10.66 (7.7–14) 10.25 (8.2–12) 14.17 (8.3–20) 2.78 (0.88–8.8) Dissolved Copper (µg/L) Influent 6.66 (0.73–13) 7.33 (5.4–9.3) xx 14.15 (10–18) 7.75 (4.5–11) 13.59 (9.8–17) xx Effluent 7.37 (3.3–11.4) 4.37 (3.7–5.7) xx 8.40 (5.6–11.4) 9.00 (7.3–10.7) 13.92 (4.4–23) xx Total Lead (µg/L) Influent 25.01 (12–38) 14.36 (8.3–20) 4.62 (1.4–12) 19.53 (10–29) 11.32 (6.1–16.5) 18.12 (5.7–30) xx Effluent 15.77 (4.7–27) 5.32 (1.3–9.0) 3.26 (2.3–4.2) 6.70 (2.8–10.6) 3.76 (1.1–6.4) 10.56 (4.3–17) 7.88 (1.6–38) Dissolved Lead (µg/L) Influent 1.25 (0.33–2.2) 3.40 (1.12–5.7) 0.50 (0.33–0.67) 2.25 (0.77–3.7) 1.44 (1.05–1.8) 1.89 (0.83–3.0) xx Effluent 2.06 (0.93–3.2) 2.48 (0.98–5.4) 0.87 (0.85–0.89) 1.96 (1.3–2.7) 1.18 (0.77–1.6) 3.34 (2.2–4.5) xx Total Zinc (µg/L) Influent 111.56 (52–172) 60.75 (45–76) 47.07 (25–91) 176.71 (128–225) 92.34 (52–132) 119.08 (74–165) xx Effluent 60.20 (21–100) 29.35 (21–38) 30.71 (13–67) 39.83 (28–52) 37.63 (17–58) 80.17 (53–108) 16.60 (5.9–47) Dissolved Zinc (µg/L) Influent 26.11 (5.2–75) 47.46(38–57) xx 58.31 (32–79) 69.27 (40–100) 35.93 (5.0–67) xx Effluent 25.84 (11–41) 32.86(18–48) xx 25.40 (19–32) 51.25 (29–73) 42.46 (10.4–75) xx Total Phosphorus (mg/L) Influent 0.19(0.17–0.22) 0.21 (0.13–0.29) 0.27 (0.11–0.43) 0.25 (0.22–0.28) 0.20 (0.15–0.26) 0.24 (0.01–0.46) xx Effluent 0.19(0.12–0.27) 0.12 (0.09–0.16) 0.14 (0.04–0.24) 0.34 (0.26–0.41) 0.14 (0.11–0.16) 0.26 (0.12–0.48) 0.09 (0.05–0.15) Dissolved Phosphorus (mg/L) Influent 0.09 (0.06–0.13) 0.09(0.06–0.13) 0.10 (0.04–0.22) 0.09 (0.07–0.11) 0.09 (0.03–0.14) 0.06 (0.01–0.11) xx Effluent 0.12 (0.07–0.18) 0.08 (0.04–0.11) 0.17 (0.03–0.31) 0.44 (0.21–0.67) 0.09 (0.07–0.11) 0.09 (0.04–0.13) xx Total Nitrogen (mg/L) Influent 1.25 (0.83–1.7) 1.64 (1.4–1.9) 2.12 (1.6–2.7) 0.94 (0.94–1.7) 1.31 (1.2–1.4) 1.25 (0.33–2.2) xx Effluent 2.72 (1.8–3.6) 1.43 (1.2–1.7) 1.15 (0.82–1.62) 0.78 (0.53–1.0) 0.76 (0.62–0.89) 2.01 (1.4–2.6) xx TKN (mg/L) Influent 1.45 (0.97–1.9) 1.26(1.0–1.5) 1.15 (0.81–1.5) 1.80 (1.6–2.0) 1.52 (1.1–2.0) 1.09 (0.52–1.7) xx Effluent 1.89 (1.6–2.2) 1.09 (0.87–1.3) 1.05 (0.82–1.3) 1.51 (1.2–1.8) 1.55 (1.2–1.8) 1.48 (0.87–2.5) 1.23 (0.44–3.4) Nitrate-Nitrogen (mg/L) Influent 0.70 (0.35–1.05) 0.36 (0.21–0.51) 0.22 (0.01–0.47) 0.59 (0.44–0.73) 0.41 (0.30–0.51) 0.40 (0.06–0.73) xx Effluent 0.58 (0.25–0.91) 0.23 (0.13–0.37) 0.13 (0.07–0.26) 0.60 (0.41–0.79) 0.82 (0.60–1.05) 0.51 (0.08–1.3) xx 1 Actual number of BMPs reporting a particular constituent may be greater or less than the number reported in this table, which was based on number of studies reported in the database based on BMP category. 2 The biofilter BMP category includes vegetated swales, filter strips, and vegetated buffers. 3 The media filter BMP category includes sand filters, organic filters, gravel filters, bioretention filters, and proprietary filters. 4 The hydrodynamic device BMP category includes a wide range of proprietary and non-proprietary hydrodynamic device types, catch basins, and oil-water separators. Notes: xx–Lack of sufficient data to report median and confidence interval. Values in parenthesis are the 95% confidence intervals about the median. More information about the confidence interval can be found in Helsel and Hirsch (2002). Differences between median influent and effluent concentrations does not necessarily indicate there is a statistically significant difference between influent and effluent. Table 5.4. Median of average BMP influent and effluent concentration with 95% confidence interval about the median.

Table 5.5. Qualitative BMP effectiveness ratings (low, medium, and high). Retrofit Category BMP Type Report Section with additional performance information References S S T s t n e i r t u N e t a l u c i t r a P d e v l o s s i D s t n e i r t u N h s a r T r e t t i L d n a s l a t e M l a t o T D s l a t e M d e v l o s s i a i r e t c a B c i n a g r O s t n a n i m a t n o C s t l a S d e v l o s s i D n o i t a u n e t t A w o l F n o i t c u d e R e m u l o V Catch Basin Inserts4 4.2.2 Caltrans (2004); Various reports on MASTEP; EC&T (2005); CSU (2005) L-M L 0 M-H L 0-L 0-L L-M 0 0 0 Sumped catch basins 4.2.3 Smith (2002); Caltrans (2003c) L-M L L L L-M L NA L-M L L L Proprietary treatment systems4 4.2.4 EC&T (2005); Pitt and Khambhammettu (2006); Yu and Stanford (2007) L-H L-M L M-H L-M L L L-M L L L GSRD Proprietary trash capture systems, non-proprietary GSRDs 4.3 Caltrans (2005) NA NA NA H NA NA NA NA NA NA NA Hydrodynamic Various proprietary devices4 4.4 BMP Database; Caltrans (2004); Various reports on MASTEP M L-M L H L-M L NA NA L L L Oil-Water Separator Baffled tanks, coalescing plate separators4 4.5 Smith (2002); Caltrans (2004); ETV (2005b) L-M L-M L H L-M NA NA L L L L Detention Surface detention basins (unlined) 4.6.1 BMP Database M-H L-M L H M-H L M M L H L-H3 Wet basins 4.6.1 BMP Database H M-H L-M H H M M H L L-H1 L Surface wetland 4.6.1 BMP Database H M-H M-H H H NA M H L M-H1 L Underground detention pipes/tanks for flow attenuation 4.6.2 NA NA NA NA NA NA NA NA NA H L Underground detention tanks/wet vaults for water quality4 4.6.2 Wright Water (2001, 2002); Li et al. (2008a) M-H L-M NA NA M-H NA NA M-H NA M L Vegetative Filtration Filter strips 4.8.1 BMP Database H L-M 0-L H M M M M L L-H3 L-H3 Swales (unlined) 4.8.1 BMP Database H M 0-L H H M M H L L-H3 L-H3 Bioretention with 4.8.1 BMP Database H H 0-L H H M H H L L-M L-M (continued on next page)

Retrofit Category BMP Type Report Section with additional performance information References S S T s t n e i r t u N e t a l u c i t r a P d e v l o s s i D s t n e i r t u N h s a r T r e t t i L d n a s l a t e M l a t o T D s l a t e M d e v l o s s i a i r e t c a B c i n a g r O s t n a n i m a t n o C s t l a S d e v l o s s i D n o i t a u n e t t A w o l F n o i t c u d e R e m u l o V underdrains Media Filtration Sand filters (surface or underground) 4.7.2 BMP Database; Caltrans (2004); Shoemaker et al. (2002) H M L-M H H L-M M H L L L Sand and organics, and engineered media filter drains 4.7.3 Herrera Environmental Consultants (2006); Shoemaker et al. (2002); Claytor and Schueler (1996) H H L H H M-H M H L L L-M Underground proprietary media filters4 4.7.4 Caltrans (2004); ETV (2005a); Various reports on MASTEP M-H L-M L M-H M-H L M M L L L Infiltration Pavement Infiltration basins, trenches, dry wells, infiltration vaults, and bioretention without underdrains 4.9 BMP Database H H H2 H H H H H H2 H H Porous pavement (assuming runoff is infiltrated) 4.10 BMP Database H H H2 L H H H H L-H2 H H Permeable overlays 4.10 Barrett (2006); Stanard et al. (2008) H L-M 0-L NA M-H L NA NA NA 0 0 NA = not applicable or not available; 0, L, M, H = zero, low, medium, and high levels of BMP effectiveness. This is a qualitative scale based on the types of unit processes included in the BMP, available literature information, and available performance monitoring data. Zero indicates the BMP does not include applicable unit processes. A low rating indicates the BMP generally provides little to no treatment and in some cases may export constituents. A medium rating indicates the BMP generally provides some removals but other BMP options can provide more effective treatment. A high rating indicates the BMP provides effective treatment when properly designed and maintained. 1 Can be high when combined with extended-detention basin. 2 Removal of dissolved salts and nitrate may be limited if groundwater below the infiltration basin discharges to the receiving water body. 3 Depends on soil conditions. 4 Based on a limited survey of devices. Performance may vary for manufactured systems. Device-specific information should be pursued. Table 5.5. (Continued).