Volume Reduction of Highway Runoff in Urban Areas: Final Report and NCHRP Report 802 Appendices C through F (2015)

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.

Appendix D Potential Impacts of Highway Stormwater Infiltration on Water Balance and Groundwater Quality in Roadway Environments (White Paper #2) 1 Introduction .................................................................................................................................... D-1 2 Potential Impacts of Stormwater Infiltration on Water Balance .............................................. D-2 2.1 Potential Changes in the Water Balance ...................................................................................... D-4 2.2 Potential Benefits and Impacts Associated with Increases in Deeper Infiltration ....................... D-5 3 Potential Impacts of Stormwater Infiltration on Groundwater Quality ................................... D-7 3.1 Sources, Fates, and Transport of Groundwater Contaminants of Concern in Roadway Runoff . D-7 3.2 Potential for Groundwater Impacts Associated with Existing Soil and Groundwater Contamination ...................................................................................................................................... D-14 3.3 Potential Risks to Groundwater Quality from Roadway Contaminant Spills ............................ D-15 4 Assessing and Mitigating Potential Stormwater Infiltration Impacts on Water Balance and Groundwater Quality ........................................................................................................................... D-16 4.1 Localized Groundwater Mounding ............................................................................................ D-16 4.2 Water Balance Impacts on Stream Flows .................................................................................. D-18 4.3 Pollutants in Stormwater Runoff ............................................................................................... D-19 4.4 Existing Soil and Groundwater Contamination ......................................................................... D-21 4.5 Wellhead and Spring Protection ................................................................................................ D-23 4.6 Contaminant Spills ..................................................................................................................... D-24 4.7 Special Considerations for Karst Topography ........................................................................... D-25 5 Consultation with Local Agencies ............................................................................................... D-25 5.1 Local Groundwater Suppliers .................................................................................................... D-25 5.2 Water Resources Protection Agencies ....................................................................................... D-26 5.3 Sanitation Districts and Other Underground Utilities ................................................................ D-26 6 Summary ....................................................................................................................................... D-27 7 References ..................................................................................................................................... D-27 D-i

NCHRP Project 25-41 1 Introduction Infiltration of stormwater from urban highways has many potential benefits, but also has the potential to result in environmental and infrastructure impacts associated with the volume of water infiltrated (i.e., water balance impacts) and the introduction and/or mobilization of pollutants into groundwater (i.e., groundwater quality impacts). Therefore, a thorough assessment of potential negative impacts of infiltration is recommended as part of evaluating the feasibility and desirability of employing infiltration techniques for a given project site and within watersheds in general. The purpose of this white paper is to provide guidance for identifying potential impacts related to these factors, particularly in the urban roadway environment, and provide recommendations for project planners and designers with respect to assessing and avoiding and/or mitigating these potential impacts. Two key factors should guide the user’s interpretation and application of this material: First, it is critical to balance the benefits of stormwater infiltration with its risks. This white paper is intended as resource for understanding a wide range of potential issues. However, the purpose of this white paper is not to discourage or introduce unnecessary barriers to stormwater infiltration where it makes sense. Not all of the issues identified would necessarily apply to a given site. In most cases, the groundwater-related risks associated with stormwater infiltration can be mitigated or avoided once they are identified and given careful consideration. Additionally, it is important to consider the watershed-scale context of potential issues. Some of the potential issues identified in this white paper would be of limited concern for stormwater infiltration at isolated sites, but cumulative effects could lead to a significant issue with widespread application. Analogously, the most effective way to mitigate these issues and balance benefits of stormwater infiltration may be the watershed and regional planning scale. Indeed, we recommend that many of these factors should be considered, and appropriate studies done, as part of developing regulations that would mandate infiltration. The remainder of this white paper is organized as follows: • Section 2 introduces and identifies potential impacts related to changes in the natural water balance. • Section 3 introduces and identifies potential impacts related to groundwater quality. • Section 4 discusses the factors that should be considered in in evaluating whether infiltration of roadway runoff is infeasible or undesirable from the perspective of site water balance or groundwater quality. This section also provides recommendations for classifying the relative risk that a given project poses and how impacts can be potentially avoided and/or mitigated as part of the project development process. • Section 5 provides guidance for consulting with local agencies, such as water suppliers and resource agencies, with respect to potential water balance or groundwater quality impacts associated with stormwater infiltration. • Section 6 provides a brief summary of this white paper. This white paper is based on a synthesis of published literature, experiences of the research team, and selected stormwater guidance documents. This white paper is not intended to provide a comprehensive set D-1

NCHRP Project 25-41 of criteria for evaluating infiltration feasibility that are applicable in all cases. Project planning and design professionals should exercise appropriate judgment in considering potential water balance and groundwater quality impacts associated with infiltration of stormwater from highways. 2 Potential Impacts of Stormwater Infiltration on Water Balance The “water balance” refers to the fate of precipitation that falls on a given area of land over a given period of time. The major components of the water balance include (1) direct runoff to surface waters, (2) evapotranspiration (ET), and (3) deeper infiltration, including water that recharges a groundwater aquifer and/or discharges as baseflow to stream channels (Gobel et al. 2004). A water balance can be computed at a wide range of scales, from the scale of a site or small subwatershed up to the scale of a major river basin or even continent. In the context of highway project development, the water balance at a site-scale or small watershed-scale is typically the most meaningful, as a project may have the greatest potential to cause negative impacts at this scale. A water balance can also be computed for a range of timescales, from very short (i.e., minutes, hours) to long term (i.e., decades, or longer). Analysis at the scale of days or weeks may be most appropriate for assessing acute impacts such as localized groundwater mounding, while analysis of long term averages (i.e., years, decades) may be most appropriate timescale for evaluating long term changes in watershed hydrologic regime, such as changes in baseflow or long term subsurface soil wetting, that may result in chronic or acute issues. The “natural” or “undeveloped” water balance varies greatly by region, watershed conditions, and the scale of the system that is being considered. To illustrate this variability, annual fluxes in water balance components were compared between several case studies of mostly undeveloped watersheds throughout the United States (Table 1). While studies reported fluxes in different combinations, the variability in conditions is evident. In several studies, ET represents the largest component of the water balance, ranging from 30 to more than 90 percent. This trend is especially prevalent in warmer locations that receive less rainfall (e.g., California, Texas). While it was infrequently separated from baseflow in the studies reviewed, direct surface runoff is typically a minor element in natural water balances, particularly in the arid west. At a site level, water balance may differ substantially from regional averages as a function of soil properties, local surface geology and hydrogeology, vegetation properties, presence of impervious surfaces, relative magnitudes and patterns of potential ET and rainfall, and other factors. Table 1. Evaluation of Variability in Undeveloped Water Balance by Region based on Selected Studies Source: Location: Annual Fluxes (as percent of precipitation): Church et al. 1995 North Eastern US Runoff and Baseflow = 55% ET and Recharge = 45% Jefferson et. al. 2008 Northwest (Cascade Mountains); The two study watersheds adjoin each other in the upper McKenzie River watershed on the west side of the Oregon Cascades Runoff and Baseflow = 70%; ET = 30%; Water Storage Change = 0% Milly 1994 East of the Rocky Mountains Runoff and Baseflow = 27%; ET = 73%; Water Storage Change = 0% Mohseni & Stefan 2001 The Baptism River watershed in northern Minnesota. The watershed is heavily timbered with both deciduous and coniferous trees. Runoff and Baseflow = 55% ET and Recharge = 45% Mohseni & Stefan 2001 The Little Washita River watershed in Oklahoma. One third of the watershed is cultivated and the rest Runoff and Baseflow = 7% D-2

NCHRP Project 25-41 Source: Location: Annual Fluxes (as percent of precipitation): is either pasture or wooded pasture. ET and Recharge = 93% Najjar 1999 Susquehanna River Basin Runoff and Baseflow = 49% ET = 51% Water Storage Change = 0% Ng & Miller 1980 Southern California Chaparral (average of 2 years of monitoring) South facing: Runoff and Baseflow = 3%; ET = 97%, Storage change negligible. North facing: Runoff and Baseflow = 9%, ET = 83% and storage change = 8%. Rose 2009 South Eastern US: five-state study area (Georgia, South Carolina, North Carolina, Virginia, and Maryland) Runoff and Baseflow = 37% ET and Recharge = 63% Ward 1993 Texas Runoff = 12.5 %; ET = 86 %; Recharge = 1.5% Sanford and Selnick (2013) estimated the estimated long term fraction of precipitation lost to evapotranspiration at the county scale for the conterminous United States (Figure 1). This graphic illustrates the variability in the relative proportions of different fluxes in of the water balance. It also demonstrates that ET is at or above 40 percent of the long-term fraction of precipitation for the majority of the United States. Given that the post-development ET fraction is approximately proportional to the amount of vegetation remaining, the ET fraction could be reduced to less than 10 to 20 percent for moderate to dense development. With this change in ET, the potential to impact water balance is significant for both runoff and infiltration approaches. Figure 1. Estimated Long Term Fraction of Precipitation Lost to Evapotranspiration (Sanford and Selnick, 2013) D-3

NCHRP Project 25-41 2.1 Potential Changes in the Water Balance Project activities have the potential to alter the water balance of a site (and smaller watersheds) as a result of changes in land cover (i.e., addition of impervious surface, compaction of pervious areas) and/or as a result of the addition of stormwater controls (i.e., infiltration systems). Project changes together with other watershed developments may impact overall watershed characteristics and should be considered in concert with cumulative impact analyses. In what ways can stormwater infiltration result in changes in the water balance? How are roadways unique in their impacts on the water balance? Many transportation improvements include the addition of impervious surfaces, either as a result of construction of a new roadway or the addition of lanes. In addition adjacent roadside unpaved right-of- way is compacted. An increase in impervious surface (and compacted soils) typically results in an increase in the surface runoff component of the water balance and tends to decrease the amount of water that enters the ground, resulting in reductions in deeper infiltration (groundwater recharge and/or baseflow). Where vegetation is removed and replaced by impervious surface, ET is also generally reduced due to a reduction in the amount of rainfall intercepted in plant leaves and in the upper soil layer that contains plant roots (i.e., the root zone) and mulches – these storage elements hold water during a rain storm and make it available for subsequent ET. While evaporation occurs from residual ponded water on impervious surfaces as well, the relative storage provided from ponding on impervious surfaces tends to be significantly less than the storage provided in the root zone below vegetated areas (Gobel et al. 2004) or even in unvegetated soils. The result is that project development, without stormwater controls, tends to result in an increase in surface runoff with corresponding decreases in both ET and deeper infiltration. Stormwater regulations are increasingly emphasizing management approaches based on “mimicking pre-development hydrology”. By “hydrology” the regulations typically narrow hydrology to what is really surface runoff hydrology (i.e. groundwater and ET components of hydrology are not addressed). This goal is generally accomplished by using volume reduction practices that rely on infiltration, ET, or consumptive uses to reduce the amount of runoff discharged directly to surface waters (Dietz 2007; U.S. EPA 2010). In most cases, compliance with these regulations is demonstrated based on the volume of direct surface runoff, without reference to or direct consideration of the other elements of the water balance or overall hydrology (SARWQCB 2009; WADOE 2012; U.S. EPA 2012c). Infiltration BMPs can be effective in mitigating increases in direct surface runoff volume and the corresponding reductions in the amount of water infiltrated. However infiltration BMPs may result in proportions of ET and deeper infiltration that are different than natural conditions. When surface runoff volume is held fixed between natural and proposed conditions and ET is reduced (as discussed above), an increase in deeper infiltration elements of the water balance (recharge and/or baseflow) must occur. A study conducted in Recklinghausen, Germany demonstrated that an “average density” development without infiltration caused the area-averaged groundwater recharge to decrease from 221 mm per year (28 percent of rainfall) in the natural condition to 163 mm year (20 percent of rainfall) in the developed condition. When runoff from the developed impervious area was infiltrated, the area-averaged groundwater recharge nearly doubled from the proposed condition without controls (163 mm per year) to the proposed condition with controls (245 mm per year, 31 percent of rainfall), which also exceeded the natural condition recharge of 221 mm per year (Gobel et al. 2004). In semi-arid climates, natural recharge may be significantly less than this study, which could result in a more substantial change in groundwater recharge as a result of the use of infiltration BMPs. For example, Ng and Miller (1980) found that in D-4

NCHRP Project 25-41 Southern California chaparral, ET made up 83 to 97 percent of the water balance, and deeper infiltration was less than 10 percent. This is consistent with the estimates made by Sanford and Selnick (2013) shown in Figure 1. In evaluations that we have conducted in Southern California, deeper infiltration has been estimated to increase by as much as three times over pre-project conditions when stormwater in infiltrated to meet water quality design requirements. Also, sites that develop with a greater impervious cover (such as roadways), tend to result in a greater reduction in ET and therefore a more substantial shift toward increased infiltration when project goals include matching pre-development volume to the post-project volume of direct surface discharge. The proportional split between deeper infiltration and ET that occurs in a BMP is a function of the underlying infiltration rate of site soils, soil moisture retention properties, plant root depths, rainfall intensity, and facility design characteristics, specifically the footprint and the depth of the BMP (Clark et al. 2006). When shallow BMPs with larger surface areas are used, the level of ET tends to increase due to the additional retained moisture content in the top layer of soils in closer contact with the atmosphere (Strecker and Poresky, 2009). In contrast, when deeper BMPs with smaller footprints are used, or when BMPs do not contain amended soil and vegetation elements a greater portion of the water balance is associated with deeper infiltration; ET plays a more minor role. Roadway environments, particularly in urban areas, represent a unique category of developed areas because of the typically high degree of “connectedness” of impervious area with hardened drainage systems and relatively small footprints that are typically available for infiltration BMPs. Directly connected impervious areas (DCIAs) have been shown to generate a significant amount of the total runoff from developed areas (Lee and Heaney 2003). The level of connectedness is important for water balance, as runoff from DCIAs, by definition, is not routed across pervious surfaces before being routed to the stormwater drainage system, which limits opportunities for ET and infiltration. Limited space in typical roadway environments also limits the amount of ET that may occur from BMPs. While relatively small footprints may be sufficient to achieve infiltration where soil conditions allow (i.e., infiltration rates are high, other constraints do not exist), ET is primarily driven by surface area, therefore the ET element of the water balance generally decreases as the BMP footprint decreases. As a result, where infiltration is used in the roadway environment to mimic pre-development surface runoff volumes, a substantial increase in deeper infiltration compared to natural conditions would be expected; this change may be particularly acute in semi-arid and arid areas where deeper infiltration tends to be a smaller element of the water balance in the natural condition and ET a larger element in the natural condition. 2.2 Potential Benefits and Impacts Associated with Increases in Deeper Infiltration When considering potential impacts of stormwater infiltration on water balance, the most significant consideration is the increase in deeper infiltration that may occur as part of projects that are designed to mimic natural surface water discharge (as introduced in Section 2.1). An increase in deeper infiltration may have both positive and negative impacts. An increase in deeper infiltration (when done safely) may have the important positive benefit of augmenting groundwater supplies, improving the quantity and temperature of baseflow in streams, especially in areas where baseflows have already been depleted due to previous development and/or groundwater extraction, and/or improving the groundwater quality. However, there are also potential negative impact related to elevated groundwater tables and changes to dry weather hydrologic regimes in stream channels. How do the changes in the water balance express themselves in terms of positive and negative impacts? D-5

NCHRP Project 25-41 Potential benefits. In regions where aquifers are actively managed for water supply, an increase in infiltration may be desirable and even encouraged. For example, the Los Angeles (CA) region is underlain by a number of productive aquifers that are actively managed for water supply. Based on model estimates conducted by the Los Angeles and San Gabriel River Watershed Council (2010), if the first 0.75 inches of each rainstorm within the Los Angeles Region were captured and infiltrated into the regional aquifer system, the natural percolated volume of 194,000 acre-feet could be increased to 578,000 acre- feet. This addition could potentially supply enough additional water for 1.5 million people, which amounts to approximately $311 million in new water supply for a region that continually faces water supply challenges. The study identified a range of other benefits associated with infiltration in these conditions. In some areas including the Central Valley in California and the Ogallala aquifer in South Dakota, Nebraska, Wyoming, Colorado, Kansas, Oklahoma, New Mexico, and Texas, where groundwater elevations have been severely reduced, such augmentation over natural infiltration recharge rates may be extremely beneficial. In regions where streams are perennial, urbanization has often been found to result in a reduction in baseflow and associated water quality issues during dry weather (for example, Spinello and Simmons 1992). Note that in some locations with wastewater discharges or significant irrigation return flows this is not the case. Implementing stormwater measures that increase deeper infiltration compared to natural conditions in selected locations may be an important part of regional strategies to augment baseflow and address dry weather water quality considerations. Where localized increases in infiltration volume do not cause negative impacts (such as those discussed below), this localized increase in infiltration may help contribute to regional improvements in watershed health by offsetting regional decreases in infiltration volume caused by historic urbanization. Another potential positive impact of increased infiltration volume is the potential to improve groundwater quality through dilution of groundwater contaminants, provided that the stormwater itself does not result in water quality impacts (as discussed in Section 3). In some cases, dilution of contaminants is not desirable, as groundwater contamination plumes can be more efficiently addressed if they remain localized and concentrated. However, in some cases, dilution of groundwater contaminants is an acceptable management strategy that can be supported through practices that result in an increased infiltration component of the water balance (Fischer et al. 2003). Nightingale (1987) found that infiltration of stormwater in the Fresno, California, area was actually diluting nitrate levels in groundwater that were elevated due in large part to surrounding agricultural activities. Potential negative consequences. Potential negative consequences of increases in deeper infiltration are generally most acute in areas where local subsurface conditions have limited ability to accept additional infiltrated volumes or where minor changes in hydrogeologic conditions would result in potential impacts. On a subwatershed scale, an increase in deeper infiltration volume has the potential to change the local hydrogeologic regime, which can have significant adverse impacts on streams. For example, in the arid southwest, many channels are naturally ephemeral, meaning that they flow during and after storm events or for some period during the wet season, but are normally dry for much of the year. The riparian ecosystems in these areas are specifically adapted for these conditions. An increase in deeper infiltration in these areas has the potential to extend the duration of surface baseflows which can result in a “type change” of the stream from ephemeral to intermittent or perennial channels. This can result in colonization of the channel with different vegetative and terrestrial species, changes in hydrologic regime, and other changes that threaten the functions and values of a riparian area as well as the species present D-6

NCHRP Project 25-41 therein. As a specific example, environmental clearance documents for the Rancho Mission Viejo planned community in Orange County, California (California FEIR No 589) identified potential adverse impacts to the endangered Arroyo Toad (which prefers dry wash habitats) if the baseflow regime of ephemeral intermittent creeks within the project were significantly altered as a result of development. This project included carefully planned stormwater management features to manage the overall water balance of the site to avoid significant reductions or increases in deeper infiltration volumes so that natural stream baseflows are maintained. On a more localized basis, the volume of infiltrated water can result in groundwater “mounding” – localized increases in the elevation of the groundwater table below infiltration BMPs. The potential for groundwater mounding is increased where there is shallow groundwater, a shallow restricting layer (bedrock, clay lens), poor soils for infiltration, a relatively shallow groundwater gradient (i.e., slope of the groundwater surface), and/or the BMP footprint is relative large (Carleton 2010). Elevated groundwater levels can lead to a number of severe problems, including flooding and damage to structures and utilities through buoyancy and moisture intrusion, increase in inflow and infiltration into municipal sanitary sewer systems, and flow of water through existing utility trenches, including sewers, potentially leading to formation of sinkholes (Gobel et al. 2004). Groundwater mounding also has the potential to impact groundwater quality as a result of a reduction in the separation between BMPs and the groundwater table and/or mobilization of contaminants as a result of submergence of contaminated soils, as discussed in Section 3. Infiltration may also increase the risk of geotechnical hazards such as subsidence, liquefaction, slope instabilities, foundation and subbase issues, and infrastructure damage associated with expansive clays (OCPW 2011; Oregon State University et al. 2006). White Paper No. 3 provides more information regarding potential geotechnical impacts and hazards related to stormwater infiltration. Section 4 provides guidance for assessing the potential for water balance impacts recommended measures to help mitigate these potential impacts. 3 Potential Impacts of Stormwater Infiltration on Groundwater Quality While infiltration of stormwater has the potential for improvements in surface water quality, it also has the potential for unintended consequences for groundwater quality (Clark et al. 2006). Infiltration of stormwater has the potential to impact groundwater quality as a result of an influx of pollutants contained in stormwater and/or mobilization of pollutants that are present in soils or groundwater. This section identifies the groundwater contaminants of concern that are typically encountered in runoff from urban roadways and discusses their fate and transport relative to potential impacts on groundwater quality (Section 3.1). It also discusses the potential for stormwater infiltration to mobilize and spread soil and groundwater contaminants (Section 3.2) and the potential for contaminant spills to impair groundwater quality (Section 3.3). 3.1 Sources, Fates, and Transport of Groundwater Contaminants of Concern in Roadway Runoff Potential for stormwater infiltration to contaminate groundwater has been studied extensively (Weiss et al. 2008). Pitt et al. (1999) and Pitt and Clark (2007) have characterized the risk of groundwater contamination from infiltration of stormwater as a function of the following factors: D-7

NCHRP Project 25-41 • Pollutant mobility – pollutants that are more mobile in the vadose zone (the unsaturated zone between the groundwater table and the ground surface) have a higher potential to contaminate groundwater than those which do not move through the vadose zone as readily. • Pollutant abundance – pollutants that are highly abundant in terms of concentration and detection frequency in stormwater have higher potential to impact groundwater quality. • Pollutant partitioning – pollutants that are present primarily in soluble fractions tend to have a higher potential to contaminate groundwater. This section identifies the main categories of contaminants which may potentially pose a risk to groundwater quality as a result of stormwater infiltration. What are the most common contaminants of concern for groundwater in roadway environments? How are roadways unique in their impacts on groundwater quality? In what ways can stormwater infiltration impact groundwater quality? The fate and transport of individual groundwater contaminants varies significantly depending on the pollutant, characteristics of water and soil, treatment facility type, and other factors. Although the composition and concentration of pollutants found in stormwater runoff are highly site-specific, the following categories of pollutants have been frequently detected in urban stormwater runoff and may pose concerns for groundwater quality: • Nutrients • Pesticides • Organic compounds • Pathogenic microorganisms • Heavy metals • Salts Note that particulates and particulate-bound pollutants are not included on this list because infiltration BMPs are generally highly effective at removing particulates from stormwater prior to stormwater reaching groundwater. With the exception of conditions with direct connections between surface water and groundwater (such as Karst topography, discussed in Section 4.7 or injection wells with little or no pretreatment), there is limited potential for particulates to migrate to groundwater and impair groundwater resources. The sections below provide a discussion of the potential for groundwater contamination for the six primary pollutants categories listed above based on a synthesis of literature sources and evaluation of unit processes provided in stormwater BMPs. Table 2, at the end of this section, provides a summary of the pollutants of concern and the relative risk to groundwater quality that each is believed to possess. 3.1.1 Nutrients The two nutrients with the most significant potential to contaminate groundwater are nitrogen and phosphorus. Nitrogen compounds, specifically nitrate, are the most commonly encountered nutrient contaminants in groundwater due to their application as fertilizers on developed landscaped areas and agricultural land uses and the deposition of vehicular exhaust on roadways and surrounding soils (Pitt et al. 1994, 1996); they also occur from the breakdown of organic debris. Naturally occurring sources of nitrogen, the atmosphere and soils, can also lead to groundwater contamination if environmental D-8

NCHRP Project 25-41 conditions are conducive. Compost amendments in shoulder treatments and/or stormwater controls can also be sources of nutrients at levels of concern if not appropriately specified and sourced. Nitrogen compounds in stormwater may be removed through a number of processes, such as uptake by plants and microbes, microbially-mediated nitrification/denitrification, and volatilization, which are highly dependent on soil composition and hydrologic properties (Weiss et al. 2008). Nitrate is the nitrogen species of greatest concern for groundwater quality as it is highly mobile in the vadose zone and groundwater, and it is not readily converted to other nitrogen species, except where biological processes are active and cyclical changes in redox conditions occur (Pitt et al. 1994, 1996). The leaching of nitrogen into groundwater and soils is most common during cool, wet seasons because lower temperatures reduce the rates of denitrification, ammonia volatilization, microbial immobilization, and plant uptake. An accumulation of nitrate in groundwater may result in health risks associated with groundwater consumption and may contribute to detrimental nutrient loadings in surface waters that receive inputs from contaminated groundwater. Nitrite and ammonia are also mobile in soils, but tend to be present in relative low levels because in most cases, they undergo oxidation to nitrate relatively rapidly in the aerobic vadose zone. Fertilizer sources of nitrate can be controlled by limited and careful application of fertilizers. Phosphorus is also of concern for groundwater contamination in some cases. Sources of phosphorus in the urban highway environment include detergents in gasoline, motor oil, fertilizer, bird droppings, animal remains, and compost amendments. The most common form of dissolved, mobile phosphorus present in stormwater, orthophosphate, can be removed from infiltrating water through precipitation or chemical adsorption onto soils (Weiss et al. 2008). As this is a sorption process, the relative phosphorus saturation of native soils can be used as an indicator of the level of potential removal of dissolved phosphorus that will result as water percolates through the soil. Soils that have been historically used for agricultural purposes may contain relatively high levels of phosphorus and may provide relatively little capacity to sorb additional phosphorus or may contribute to phosphorus loadings. Like nitrate, elevated dissolved phosphorus levels in groundwater may contribute to detrimental nutrient loadings in surface waters that receive groundwater discharges. Overall, nutrients in highway runoff pose relatively limited potential to impact groundwater quality in most cases due to relatively low concentrations in urban stormwater in comparison to groundwater concentrations and groundwater quality objectives (Pitt et al. 1999). Risks are elevated when stormwater includes runoff from industrial and/or agricultural land uses that may periodically contain very high levels of nutrients or where aquifers or receiving waters have limited capacity for additional nutrients. 3.1.2 Pesticides Pesticides contamination of groundwater, such as 2,4-D, lindane and chlordane, usually originate from weed, pest or fungus control in landscaped areas. Sources in the urban highway environment may include roadside maintenance of landscaping, medians, and other landscaped or managed areas. Concentrations of pesticides found in groundwater vary significantly based on pesticide usage (quantity and type), underlying soil texture, total organic carbon of the soil, contaminant persistence, and the depth to groundwater (Pitt et al. 1994, 1996, 1999). Pesticides with high solubility and low affinity for organic matter (low partitioning coefficients) tend to pose the greatest risk. Pesticides tend to have the highest mobility in coarse-grained or sandy soils without a hardpan layer, with low clay and organic matter content and high permeability. D-9

NCHRP Project 25-41 Decomposition is possible in both soil media and water, but the time frame can range between days and years depending on the conditions and the specific compounds (Pitt et al. 1994, 1996). Studies conducted on the half-life of most pesticides are generally applicable to surface and near-surface conditions and do not generally account for the reduced microbial activity deeper in the vadose zone. Where there is concern regarding pesticide contamination, risks can generally be mitigated by careful selection and application of pesticides. Pesticides with low solubility and high affinity for organic matter tend to pose limited risk of groundwater contamination. Where soils have limited organic matter, the use of amended media in an infiltration BMP has the potential to limit transport of pesticides. 3.1.3 Organic Compounds Some organic compounds in groundwater are naturally occurring, such as those originating from decomposing animal wastes, leaf litter, vegetation, and organisms in the soil, but many are man-made and originate from sources like landfills, sewage systems, agricultural runoff, and urban stormwater runoff, including highway runoff (Pitt et al. 1999). The most common organics detected in groundwater are phthalate esters, such as bis(2-ethylhexyl)phthalate, and phenolic compounds, such as phenol and 2,4- dimethyl phenol. Volatile organic compounds, such as benzene, chloroform, methylene chloride, trichloroethylene (TCE), tetrachloroethylene, toluene, and xylene, are also detected in groundwater. Finally, polycyclic aromatic hydrocarbons (PAHs, a class of semi-volatile organic compounds), such as fluoranthene and pyrene, are commonly detected in urban stormwater, particularly from roadway and parking lot runoff due to their production via combustion processes, and have been detected in groundwater (Pitt et al. 1999). Uncombusted petroleum products (oils, lubricants, etc.) are also sources of some types of PAHs. For example, testing in New Jersey demonstrated that groundwater receiving runoff from roadways and parking lots contained elevated concentrations of petroleum hydrocarbons, such as benzene and toluene, when compared to background groundwater levels (Fischer et al. 2003). The majority of organic compounds can be removed via volatilization, sorption and degradation to levels that would not affect groundwater quality. Treatment processes found in organically-active soils (sorption, microbial degradation) generally trap hydrocarbons within the first few centimeters of soil and do not allow them to percolate far enough to contaminate groundwater (Weiss et al. 2008). PAHs, in particular, have been successfully removed from infiltrating stormwater through degradation by naturally developed microbial communities where there is a prolonged aerobic, sulfate reducing, and denitrifying environment (Miklas and Grabowiecki 2007). Volatilization is another pathway that has been observed to remove organics, but the rate of volatilization for many hydrocarbons decreases with lower temperatures by nature, resulting in higher detectable rates in colder months (Fischer et al. 2003). The presence of sandy soils and a high water table have been found to be correlated with greater prevalence of groundwater contamination by organic compounds (Pitt et al. 1994, 1996). While roadways are a known source of organic compounds, particularly petroleum hydrocarbons and combustion byproducts, the risk posed by infiltration of organic compounds in stormwater is relatively low due to relatively low stormwater concentrations and relatively low mobility of most organic compounds in the vadose zone. There is an elevated level of contamination risk where specific sources of organic compounds are present, where there is a shallow depth to groundwater, where soils are sandy, and where subsurface injection/infiltration is used – each of these factors reduces the effectiveness of removal processes in the vadose zone that would protect groundwater quality (Pitt et al. 1994, 1996). D-10

NCHRP Project 25-41 3.1.4 Pathogenic Microorganisms Pathogens, including bacteria and viruses, are ubiquitous in urban stormwater runoff. They originate from anthropogenic sources, such as human waste, dog waste, and failing septic systems, as well as natural sources, such as bird and animal droppings. Pathogenic contamination of groundwater has been linked to stormwater infiltration: for example, pathogens, most commonly enteroviruses, have been found to occur more frequently in groundwater where there is a high groundwater table and near MS4 outfalls (Pitt et al. 1999). Contamination depends on a number of site-specific environmental factors and pathogen characteristics, but it is believed that infiltration through BMPs may create a pathway for groundwater contamination (Weiss et al. 2008). The survival and persistence of pathogens in the vadose done and in groundwater is sensitive to a number of factors, including pH, interactions with soil microflora, moisture content, temperature, dissolved oxygen content, and concentrations of organic matter. Viral pathogens, specifically enteroviruses have a high likelihood for groundwater contamination if they are present in the stormwater runoff as they are highly mobile and are less sensitive to environmental factors than bacterial pathogens (Pitt et al. 1999). Bacterial pathogens are often removed through sedimentation or sorption to soils during percolation through the upper layer of soils due to their size. The pathogens that are successfully filtered and sorbed are inactivated and killed as soil dries and necessary survival factors are eliminated (Pitt et al. 1999). For pathogens reaching the groundwater table, the distribution and movement of pathogens in groundwater is controlled by convection, sorption, and dispersion in the liquid phase. Pathogen contamination from stormwater infiltration, particularly by enteroviruses, may be of significant concern for groundwater quality in many urban roadway areas due to the potential for elevated loadings from human waste, pet waste, and garbage. Like other contaminants, the risk of pathogen contamination is elevated where soils are coarser and depth to groundwater is less. However, where groundwater is not used for human consumption or where groundwater is already disinfected prior to use, the material consequences of bacterial contamination may be limited. Additionally, for controlled access facilities, such as limited access highways and freeways, the source of human pathogens would be very limited, with associated reduced concern for groundwater contamination. When highways include runoff from adjacent areas, there is more potential for human pathogen sources. 3.1.5 Heavy Metals Heavy metals, such as copper, lead, and zinc, are commonly found in stormwater runoff. Sources of heavy metals in urban runoff include automobile parts, building materials, exposed metal products, fertilizers, fungicides, atmospheric deposition from industrial and vehicle exhausts and other sources. In the roadway environment, specific sources included brake pads and treated woods (copper), tires (zinc), and galvanized guardrails (zinc). Leaded gasoline was historically a major source of lead and arsenic loading, however allowable levels of lead have been greatly reduced in gasoline in the United States and, in the last 20 years, lead is not commonly detected in runoff at levels of concern. Lead may still be a concern as a legacy pollutant in roadside soils in some case and/or if there are sensitive receptors that could be impacted by lead at low levels. While roadways are recognized to be an elevated source of metals, concentrations in urban runoff tend to be much less than drinking water standards, therefore risks of human health issues from stormwater infiltration are minor. During the infiltration process, when infiltrated through soils, metals are typically removed via filtration of particulate-bound metals, adsorption of dissolved metals to soil particles, chemical precipitation, diffusion into solid particles, and biological uptake (Weiss et al. 2008), decreasing their D-11

NCHRP Project 25-41 likelihood of groundwater contamination. However, metals are of increased concern when infiltration facilities are located in rapidly-infiltrating inert materials, such as sand or gravels, or when infiltration occurs in close proximity to sites with elevated stormwater concentrations, such as industrial sites and maintenance yards. Among the heavy metals of potential concern to groundwater quality, zinc has the highest solubility in stormwater and tends to be the most mobile in the vadose zone (Pitt et al. 1999). Soils that have a high cation exchange capacity (CEC) and organic content tend to provide greater potential for removing heavy metals through sorption and other processes, however metals have been found to be well retained in systems without organic content, such as permeable pavements (Dierkes et al. undated). The use of organic materials in soil media for infiltration systems can improve the performance of removal of heavy metals. 3.1.6 Salts and Dissolved Minerals Inorganic dissolved minerals, including chloride, sulfate, and sodium, have been detected in groundwater at concentrations of concern (Pitt et al. 1999). Sources of salts in groundwater include natural salts in soils, addition of fertilizers to agricultural fields, evaporation of irrigation water (leaving salts behind), addition of salts to waste streams through consumptive uses specifically water softening, and application of salts to roads and other surfaces during cold weather. Among stormwater-related sources, road salting is the most significant source of salt loading and has been found to significantly affect chloride content and salinity of groundwater (Pitt et al. 1994, 1996). The potential for long-term accumulation of salts in groundwater is a function of the nature of the aquifer and the loading of saline water versus fresh water – aquifers that exist in closed or relatively closed basins are more susceptible to long term increases in salts. Reclaimed water (treated wastewater) is also a source of salts and may be a consideration in areas where reclaimed water is used to irrigate roadside vegetation or other landscaping tributary to roadway systems. Roadway runoff in cold climates, where salt is used, has a high potential for contaminating groundwater because salts are water soluble, non-filterable, not readily sorbed to solids, and can leach into groundwater as infiltration occurs (Weiss et al. 2008; Pitt et al. 1994, 1996). Because conventional treatment methods are not effective at removing salts, the potential for salt contamination of groundwater may be an overriding factor in determining the feasibility of stormwater infiltration. In areas where salts are not applied to roadways, the infiltration of stormwater may improve groundwater quality through dilution, as stormwater typically has relative low dissolved mineral concentrations (Pitt et al. 1994, 1996). 3.1.7 Summary of Stormwater Contaminants of Concern for Groundwater Quality Impacts Table 2 provides a summary of the potential risks to groundwater quality posed by the stormwater contaminants. The level of risk posed by each category of contaminant is discussed, as well as the degree to which risks can potentially be addressed through mitigation measures such source controls, pretreatment, and separation to groundwater. Specific mitigation measures are identified in Section 4. D-12

NCHRP Project 25-41 Table 2. Summary of Potential Stormwater Contaminants of Concern for Groundwater Quality Constituent Roadway-related Sources Relative Abundance in Roadway Stormwater Runoff in Soluble Phase Mobility through Vadose Zone Relative Stormwater-Related Contamination Risk Potential for Remaining Risk after Mitigation Measures (See Section 4.4) Nitrogen • Fertilizers • Vehicle exhaust • Petroleum products • Plant materials • Animal droppings and remains • Compost amendments Low to Moderate Moderate to High, as nitrate Greatest during cooler weather Low to moderate; while nitrate is has a high potential for leaching to groundwater, relatively low concentrations are typically observed in stormwater runoff which tend to be well below groundwater quality objectives. Agricultural or industrial land uses in the tributary watershed are an indicator of elevated risk. Limited risk, except where groundwater is very sensitive to nitrogen inputs, or where natural sources of nitrogen exist in soils that may be mobilized by stormwater infiltration. Phosphorus • Fertilizers • Detergents in gasoline • Motor oil • Plant materials • Animal droppings and remains • Compost amendments Low to Moderate Low to Moderate Greatest where soils have limited potential to sorb dissolved phosphorus. Low; phosphorus concentrations tend to be relatively low in highway runoff and mobility in subsurface environments tends to be limited. Where soils have been used for historic agricultural uses, risks may be elevated. Limited risk with appropriate mitigation measures, except where soils have received historic phosphorus loadings and may be a source of phosphorus. Pesticides • Pesticide use in landscaped areas Low to Moderate Low to High Greatest in sandy soils, high water table, and pesticides with high solubility and low affinity for organic matter. Low to moderate; concentrations in highway stormwater runoff tend to be low, and pesticide concentrations can be managed through application, however some combinations of pesticide types and soil conditions are conducive for pesticides to migrate to and persist in groundwater. Limited risk with appropriate mitigation measures. Organic Compounds • Oils, gasoline • Asphalt, coal tar • Agricultural runoff • Other Moderate to High Low to Moderate Greatest in sandy soils and where there is a high water table. Low to moderate; most organics are removed in soils prior to reaching groundwater. Limited risk with appropriate mitigation measures. Pathogens • Human waste • Animal droppings and remains • Septic systems Highly variable; may be highest in urban areas due to human waste and pet waste Low for bacterial pathogens Moderate to High for enteroviruses Moderate to high; pathogens are commonly present in stormwater runoff. While roadways do not have elevated risks compared to other land uses, pathogens may be present at elevated levels. Enteroviruses are highly mobile and have a high risk of groundwater contamination if infiltrated. Limited risk for bacterial pathogens with appropriate mitigation measures Enteroviruses may pose an unavoidable risk where they are present and their introduction into groundwater is not tolerable. Metals (copper, zinc, lead • Vehicles • Roadway debris • Roadway paint • Roadway materials Moderate to High relative to surface water toxicity criteria Generally Low relative to drinking water criteria Low, removed through sorption and filtration Low to moderate; metals are generally removed through sedimentation, filtration, sorption and precipitation prior to reaching groundwater. Indicators of elevated risk include industrial land uses, inorganic soils, and shallow depth to groundwater. Zinc appears to exhibit the highest risk. Limited risk with appropriate mitigation measures. Salts • Roadway salting • Fertilizers • Natural mineral leaching • Reclaimed water irrigation Seasonally High High Seasonally high; where colder temperatures and roadway freezing requires salting, added loading of salts increases the risk of groundwater contamination. Risks of salt loadings may be difficult to mitigate, except where it is feasible to: (1) divert runoff from infiltration BMPs during winter months, (2) avoid using salt (risks of alternative approaches should be evaluated), or (3) coordinate stormwater management approach with a local salt management plan. D-13

NCHRP Project 25-41 3.2 Potential for Groundwater Impacts Associated with Existing Soil and Groundwater Contamination Legacy contamination of soil and groundwater is common in areas that have previously been used for urban development or agricultural uses. The most common constituents of concern are volatile organic compounds; especially chloroform, solvents like PCE and TCE, and the gasoline oxygenate MBTE. These compounds have been observed to persist in groundwater and have been detected in aquifers throughout the United States (Moran, et al. 2006). Subsurface contamination may also take the form of septic systems, cemeteries, and informal municipal solid waste disposal sites, with similar considerations. Where soil or groundwater contamination exists below a project site or in the project vicinity, stormwater infiltration may have a number of negative consequences that should be carefully evaluated (Gobel et al. 2004): 1. Infiltration of stormwater can mobilize pollutants in contaminated soils and provide a mechanism to transport these pollutants to groundwater. This can threaten groundwater quality and complicate soil cleanup efforts. 2. Infiltration of stormwater can influence the behavior of existing groundwater pollutant plumes through the addition and/or concentration of infiltrating water over current conditions. This can result in a plume shifting its direction of movement, accelerating movement and/or spreading and becoming more diffuse. Where cleanup efforts are underway, this can complicate these efforts and make isolation of the plume from groundwater sources more challenging. Where cleanup activities have not started, the spreading of a plume generally makes cleanup efforts more challenging and costly and may threaten drinking water supplies. 3. Where stormwater infiltration results in a localized increase in groundwater table, there may be potential for the water table to intersect with contaminated soil and/or utilities that would otherwise be “high and dry”. Finally, infiltration may also disrupt the natural degradation process of contaminants in soil or groundwater by introducing additional waters, nutrients, or limiting reagents to the environment. Soil and groundwater contaminants are not solely anthropogenic. In some cases, historic geologic deposits may contain elevated levels of natural contaminants such as selenium and arsenic which can be mobilized through stormwater infiltration. For example, the central portion of Orange County (California) is underlain by a selenium plume that originated from selenium contained in natural sediments deposited over geologic time, which has been released as a result of changes in water table with agricultural development and urbanization. Infiltration of stormwater in this vicinity has the potential to further mobilize selenium from soils as well as increase the volume of contaminated groundwater discharged to creeks. It is possible that in some cases, stormwater infiltration could be used as part of a solution to address groundwater or soil contamination, either through dilution or by strategically influencing the movement of a plume. Such an approach should generally be implemented in close coordination with groundwater management and/or cleanup authorities. Soil and groundwater contamination are highly site-specific considerations that should be evaluated carefully for each project site where historical data suggests that contamination is present or may be present. D-14

NCHRP Project 25-41 3.3 Potential Risks to Groundwater Quality from Roadway Contaminant Spills The U.S.DOT estimates that 7 percent of all trucks travelling the nation’s roadways are carrying hazardous material (Federal Highway Administration 2009). As such, contaminant spills are a constant risk in the roadway environment, with the potential to deposit high concentrations and volumes of pollutants onto the roadway and into the roadway drainage system within a short period of time. While most state DOTs have procedures defined for reporting and handling spills, contaminants may begin to infiltrate before responders are able to contain and remove the source. Infiltration of spilled contaminants may occur whether there are infiltration BMPs present or not, however, in theory, the use of infiltration BMPs may accelerate the rate at which the contaminants are able to percolate into groundwater and also may concentrate these contaminants into small areas such that the natural ability of soils to attenuate and remove contaminants are more limited. Conversely, the presence of infiltration BMPs may help prevent spilled pollutants from draining to surface waters where they may have impacts on the downstream environment and receiving waters. Much of the spill may be contained within the infiltration system if runoff is not present at the time of the spill. Based on a review of 10 years of roadway spill records in Alabama, Becker et al. (2001) found that hydrocarbons (including diesel oil, road tar, gasoline, fuel oil, asphalt, liquefied petroleum gas, jet fuel, hydraulic oil, and creosote) were the most commonly spilled constituents in highway incidents by a large margin and were also released in the greatest volumes. Ammonia and ammonium nitrate were also released, but at much less frequency and lower quantities. Other types of spills are possible; however data suggests that spills other than petroleum hydrocarbons tend to be quite rare. Per the discussion in Section 3.1.3, petroleum hydrocarbons (organic compounds) are generally retained within a relatively thin layer of surficial soils and would not be expected to pose a significant risk to groundwater quality in most cases where infiltration occurs through soils and/or media. BMPs that are underlain by a thick layer of soil or that have amended media filtration processes would theoretically provide a high level of control for petroleum hydrocarbon spills and limit the potential for groundwater contamination. Higher risk of contamination may occur when groundwater table is high and soils are sandy or gravelly with limited organic content. Where these conditions prevail, it may be desirable to include BMP design components isolate groundwater contamination pathways such as including a containment vault with isolation valves upstream of underground infiltration systems to contain inflow of contaminants. However, the general measures that are used to address potential impacts from chronic stormwater loadings in sandy soils and high groundwater conditions (e.g., pretreatment and amended soils) would also tend to be effective for controlling and containing most spills. Miklas and Grabowiecki (2007) evaluated the potential for permeable pavements to capture and degrade petroleum hydrocarbons and concluded that relatively large hydrocarbon spills can be contained in permeable pavement systems. They found that the gravel, sand, and soil that make up permeable pavements function as hydrocarbon traps and in-situ bioreactors under some conditions. Similar findings would likely apply to bioretention areas and similar systems that have amended soils and plant roots. In the event of any significant spills, the systems would need to be remediated to remove contaminated soil and media. Literature was not identified that specifically considered the potential risks posed by spills of solvents or other mobile contaminants; however these types of spills are believed to be very uncommon and highway designs may not be able to provide risk mitigation measures for very infrequent and unpredictable occurrences. D-15

NCHRP Project 25-41 4 Assessing and Mitigating Potential Stormwater Infiltration Impacts on Water Balance and Groundwater Quality This section is intended to provide guidance for assessing the potential for impacts of stormwater infiltration related to water balance and groundwater quality. This section also provides recommendations for mitigating these potential impacts to potentially improve the level of infiltration that can safely be achieved in the highway environment. What and how should factors be considered in evaluating whether infiltration of roadway runoff is infeasible or undesirable related to site water balance and groundwater quality? How can potential impacts be mitigated to reduce risks and improve feasibility and desirability of infiltration? This section is divided into seven key factors that should be considered in assessing the feasibility and desirability of infiltration related to water balance and groundwater quality: • Localized groundwater mounding, • Water balance impacts on streamflow, • Contamination from stormwater runoff pollutants, • Soil and groundwater contamination, • Wellhead and spring protection, • Special considerations for Karst aquifers, and • Local groundwater management objectives and criteria. Key to summary tables in this section: The summary tables included within each subsection below contain guidance for evaluating the potential level of risk associated with each feasibility factor and provides recommendations for additional analysis and mitigation measures where site-specific factors suggest that risks may be elevated. The intent of these tables is to help users develop planning level classifications of potential risk. Each of these factors may warrant more extensive analysis than described in this planning level document, but the tables and guidance here will assist in project planning by focusing attention where the potential for impacts is greatest. Where elevated risk indicators are not present, a simplified assessment may be adequate. The greater number of elevated risk indicators present for a given site indicates a higher level of risk, which may require a more extensive assessment to demonstrate that infiltration can be safely done or may represent a technical basis for infiltration to be considered infeasible. These tables are intended for planning level screening only -- the final determination of whether to implement infiltration in an urban roadway environment should be based on site specific information and analysis, coordination with other applicable agencies, and best professional judgment. 4.1 Localized Groundwater Mounding The use of infiltration BMPs has the potential to result in elevated local groundwater tables that exceed the natural seasonal high groundwater table or extend the duration of ponding at a site. This can have a number of negative consequences, as introduced in Section 2.2, including: D-16

NCHRP Project 25-41 • Infrastructure damage, • Geotechnical hazards, • Inflow and infiltration into the sanitary sewer, • Conveyance of water in utility trenches, • Mobilization of groundwater contaminants, • Reduced separation to groundwater, resulting in greater risk of contamination (See Section 4.3), and • Undesirable surface ponding. The mounding potential beneath a particular BMP is a function of the facility design, the infiltration rate, the precipitation rate, and the hydrogeologic conditions at the site. Table 3 provides indicators for assessing whether groundwater mounding may be of specific concern for a project. Table 3. Risk Indicators for Localized Groundwater Mounding Lower Risk Indicators ↔ Elevated Risk Indicators Risk Factors • Depth to seasonally high groundwater relatively large (greater than 10 feet) (Pitt et al. 2004) • Relatively steep groundwater gradient; groundwater mounding quickly dissipates • Relatively simple geology; roughly uniform infiltration expected • Smaller footprint and linear BMPs • Storm events are well- distributed throughout the year • Depth to seasonally high groundwater is low to moderate (less than 10 feet) (Pitt et al. 2004) • Relatively little groundwater gradient • Complex geology with impermeable lenses, potential faults or other barriers to vertical or lateral dissipation • Larger footprint BMPs • Storm events tend to arrive in clusters, creating critical periods for localized groundwater mounding Potential Additional Studies and Mitigation Measures for Sites with Elevated Risk • Conduct more detailed investigation of groundwater depths and gradients to refine qualitative classification of the potential for mounding. • Conduct computational analysis of groundwater mounding or conduct large scale pilot infiltration testing (See White Paper 1) to assess capacity of subsurface geology to safely accept water. • Use BMPs that are more distributed (smaller footprint area per unit) or are more linear in nature to help reduce mounding potential. Smaller facilities are those located at intervals within the normal right-of-way width, whereas linear facilities are those that are continuous over much of the length of the right-of- D-17

NCHRP Project 25-41 way. These BMPs include bioretention areas (a smaller footprint area), or permeable shoulders (a linear facility). • Develop a stormwater management strategy involving less infiltration; infiltration of the full water quality or hydromodification control volume may not be feasible or desirable in conditions with unavoidable potential for groundwater mounding. Where predominantly lower risk indicators exist, the use of a simplified method is generally appropriate. For example, USGS has developed a spreadsheet tool for rapid assessment of mounding potential (Carleton 2010). The tool (available at http://pubs.usgs.gov/sir/2010/5102/) estimates the maximum expected mounding below an infiltration BMP by solving simplified groundwater flow equations that incorporate a number of simplifying assumptions about the hydrogeology of the site. For more complex or critical conditions, it may be appropriate to utilize a more robust modeling framework, such as the USGS MODFLOW model, which is a finite-difference groundwater flow model (http://water.usgs.gov/nrp/gwsoftware/modflow.html). It is appropriate for the civil and geotechnical engineers to use best professional judgment regarding the selection of the methods used to assess this factor. 4.2 Water Balance Impacts on Stream Flows As discussed in Section 2.2, the use of infiltration systems to reduce surface water discharge volumes may result in additional volume of deeper infiltration compared to natural conditions, which may result in impacts to receiving channels associated with change in dry weather flow regimes. This is a function of the local climate, but also the local hydrogeologic conditions. To assist in project planning, a number of indicators may be used to help evaluate the potential for impacts, as summarized in Table 4. A relatively simple survey of hydrogeologic data (piezometer measurements, boring logs, regional groundwater maps) and downstream receiving water characteristics is generally adequate to determine whether there is potential for impacts and whether a more rigorous assessment is needed. Since linear highway projects typically make up a relatively minor portion of any watershed, a single project is unlikely to significantly change the overall watershed water balance. More typically, the practical question is whether infiltration of highway runoff would exacerbate an existing water balance condition rather than necessarily create a new condition of concern. A review of local watershed conditions may be adequate to determine if this factor warrants further considerations. If water balance concerns do exist in the watershed, it may be most appropriate to follow the lead of local jurisdictions in terms of watershed-scale priorities and approaches. Where water balance conditions appear to be sensitive to development impacts and there is an elevated risk of impacts, a computational analysis may be warranted to evaluate the feasibility/desirability of infiltration. Such an analysis should account for precipitation, runoff, irrigation inputs, soil moisture retention, ET, baseflow, and change in groundwater recharge on a long-term basis. Because water balance calculations are sensitive to the timing of precipitation versus ET, it is most appropriate to utilize a continuous model simulation rather than basing calculations on average annual or monthly normal conditions. The USGS Soil Water Balance Model (http://pubs.usgs.gov/tm/tm6-a31/), the EPA Stormwater Management Model (SWMM, http://www.epa.gov/nrmrl/wswrd/wq/models/swmm/), or other models may be appropriate for these calculations. Where cumulative watershed impacts are of D-18

NCHRP Project 25-41 concern, other studies prepared by local jurisdictions may be informative to understand the feasibility and desirability of infiltrating highway runoff. Table 4. Risk Indicators for Water Balance Impacts on Stream Flows Lower Risk Indicators ↔ Elevated Risk Indicators Risk Factors • Actively managed aquifer with adequate capacity for additional infiltrated volume • Perennial streams in vicinity that naturally receive base flow from groundwater discharge year round • Relatively simple geology; predictable vertical migration of water • Higher order (larger) streams • Existing proportion of precipitation lost to ET is lower (See Figure 1) • Proposed BMPs provide substantial ET losses; less potential to increase deeper infiltration compared to natural conditions • Current conditions to do not show signs of water balance issues • Perched groundwater or shallow, local aquifer • Streams in the vicinity are ephemeral or intermittent and have sensitive species/habitat issues • More complex and problematic geology; e.g., sloping strata with potential for lateral migration to channel banks • Lower order (smaller) streams • Existing proportion of precipitation lost to ET is higher (See Figure 1) • Future development is projected over a significant proportion of watershed (potential for cumulative effects) • Current conditions show signs of water balance issues Potential Additional Studies and Risk Mitigation Measures for Sites with Elevated Risk • Conduct more detailed investigation of hydrogeology and water fluxes. • Conduct computational analysis of water balance on a long-term basis. • Select BMPs with greater quantity of amended soil media, greater vegetation, and shallower depth to achieve a more natural balance between infiltration and ET. • Develop a stormwater management strategy involving less infiltration; infiltration of the full water quality or hydromodification control volume may not be feasible or desirable in cases where water balance impacts are significant and cannot be avoided. • If cumulative development impacts are of concern, potentially collaborate with local jurisdictions to evaluate potential water balance impacts, or adhere to findings of studies prepared by local jurisdictions. 4.3 Pollutants in Stormwater Runoff As discussed in Section 3.1, the concentration of stormwater pollutants is highly dependent on the land uses and activities present in the area tributary to an infiltration BMP. Likewise, the potential for groundwater contamination is a function of pollutant abundance, speciation of pollutants in soluble forms, and the mobility of the pollutant in the subsurface soils. Therefore, an evaluation of a large number of D-19

NCHRP Project 25-41 potential conditions that may be encountered with a range of associated risks and a site-specific assessment is recommended. To assist in project planning, Table 5 provides risk indicators to help evaluate the potential for impacts and identify pollutants that may warrant greater analysis. This table also provides potential mitigation measures to help reduce risks – these include general mitigation measures, as well as pollutant-specific mitigation measures. Table 5. Potential for Contamination from Stormwater Runoff Pollutants Lower Risk Indicators ↔ Elevated Risk Indicators Risk Factors • Lower traffic volumes, lower heavy truck traffic and contaminant sources1 • Soils have substantial pollutant attenuation capacity1 • Depth to mounded seasonal high groundwater exceeds 10 feet • Climate does not necessitate significant roadway salting • Anthropogenic sources of pathogens are limited in tributary area • Higher traffic volumes and contaminant sources1 • Soils have limited pollutant attenuation capacity1 • Depth to mounded seasonal high groundwater is less than 10 feet • Climate necessitates significant roadway salting • Sources of anthropogenic pathogens are present in tributary area • Karst topography (see Section 4.7) Potential Additional Studies and Risk Mitigation Measures for Sites with Elevated Risk • Characterize depth to groundwater – locate and design BMPs to maintain separation of 10 to 20 feet to seasonal high groundwater – greater separation appears to result in lower risk for all constituents of concern. • Review data and/or conduct monitoring to characterize expected runoff quality for the project; evaluate expected stormwater quality vs. local groundwater quality objectives. • Utilize BMPs that provide treatment to runoff prior to deeper infiltration to address pollutants of concern, such as amended soil media to augment removal capacity of native soil – effective for pesticides, metals, many organics and bacterial pathogens; limit subsurface infiltration/injection. • Utilize practices to reduce pollutant loadings, such as permeable friction course overlays, which result in less splashing and lesser quantity of pollutants. • Utilize pollutant-specific source controls, as applicable: o Nutrients - limit fertilizer applications, use of slow-release fertilizers. o Pesticides - limit application of pesticides; utilize pesticides with lower mobility and shorter half-life. o Pathogens – provide sanitary facilities; provide animal waste disposal; implement program to clean up animal remains. o Salts – limit application, consider alternative methods, and potentially divert runoff from salted roadways in cold weather. • Identify expected high source areas, such as maintenance yards and gas stations, and design drainage system to hydrologically isolate these areas from infiltration BMPs. • Develop a stormwater management strategy involving less infiltration; infiltration of the full water quality or hydromodification control volume may not be feasible or desirable in cases where potential contamination cannot be mitigated. 1 – Indicators of loading and soil pollutant attenuation capacity are included in the following bullet lists. D-20

NCHRP Project 25-41 The following list includes indicators of elevated sources that may be applicable in the urban highway environment (SARWQCB 2009; WADOE 2012; U.S. EPA 2009): • Average daily traffic volume - A traffic volume threshold of 25,000 or 30,000 average daily traffic (ADT) has been applied in a number of permits and guidance documents as an indicator of conditions that could potentially threaten groundwater quality. Note that some researchers attributed potential water quality concerns in these areas as general urban sources vs. the specific traffic volumes (i.e. with ADTs of this level, monitoring sites were located in more dense urban land use areas.) • Commercial or industrial sites subject to an expected average daily traffic count (ADT) ≥100 vehicles/1,000 ft² gross building area (trip generation). Particularly those with heavy truck traffic. • Other areas with potential high threat to water quality, such as industrial or light industrial activities; fleet maintenance, storage, and/or wash yards; and nurseries. • Material storage areas, such as asphalt or salt storage areas. • Other locally-applicable guidance at the discretion of the project engineer. The following list includes indicators of the pollutant attenuation capacity of soils (WADOE 2012): • Cation exchange capacity (CEC) of the treatment soil should be at least 5 milliequivalents CEC/100 g dry soil (U.S. EPA Method 9081). CEC values of greater than 5 meq/100g are expected in loamy sands, according to Rawls, et al. Lower CEC content may be considered if it is based on a soil loading capacity determination for the target pollutants that is accepted by the local jurisdiction. • Organic content of the treatment soil (ASTM D 2974): Organic matter can increase the sorptive capacity of the soil for some pollutants. A minimum of 1.0 percent organic content is recommended. • Depth of organically-active soils. An assessment of CEC and organic content should encompass all distinct layers below the base of the facility to a depth of at least 2.5 times the maximum design water depth, but not less than 6 feet. • Other locally-applicable guidance at the discretion of the project engineer. 4.4 Existing Soil and Groundwater Contamination Section 3.2 describes the potential impacts that may result when stormwater is infiltrated in the vicinity of existing groundwater plumes, contaminated soils, septic systems, cemeteries, municipal solid waste disposal sites, and other potential sources of groundwater pollution. When conducting a site assessment to assess the feasibility of stormwater infiltration, the following sources may contain information that is useful in determining the potential risks associated with existing contamination: • Records of previous uses of the site. • Locations of current or historic septic systems, cemeteries, historic (informal) refuse dumps. D-21

NCHRP Project 25-41 • Mapping of contaminated groundwater plumes and soils, which may be available from local groundwater management agencies or state environmental quality agencies. For example, spatial databases and layers containing contaminated sites were identified in California, New Jersey, Connecticut, and Alaska (see References) as part of this effort; similar resources likely exist or are under development in other states. • U.S. EPA Envirofacts database, which contains locations and information regarding RCRA, CERCLA, brownfields, and cleanup sites (U.S. EPA 2012a). Individual states may also have more detailed information available. Table 6 summarizes risk factors that may indicate an elevated degree of risk and provides recommendations for potential mitigation measures related to infiltrating near groundwater plumes or contaminated soils. Table 6. Groundwater Contaminant Plumes and Soil Contamination Lower Risk Indicators Moderate Risk Indicators ↔ Elevated Risk Indicators Risk Factors • Contamination not suspected at site • Local data confirms no contamination • Historic land uses do not include industrial, agricultural or other uses suggesting potential for contamination • Septic systems, cemeteries, municipal solid waste sites not present in project vicinity • Historic site uses included industrial, agricultural, or other uses suggesting potential contamination, but contamination is unknown • Historic contamination is documented but has been fully remediated • Known contamination at the site • Groundwater plumes (natural or unnatural) exist in site vicinity • Septic systems, cemeteries, municipal solid waste sites present in site vicinity and hydrogeologic connections with these sources are possible Potential Additional Assessments and Risk Mitigation Measures for Sites with Elevated Risk • Consult with regulatory agencies responsible for site cleanup and groundwater protection regarding potential benefits or consequences associated with stormwater infiltration at the site (see Section 5). • Conduct more detailed investigation of potential contamination to reduce uncertainties and better quantify potential risk. • Site infiltration BMPs to avoid infiltration of stormwater where there is significant unavoidable risk of mobilizing of pollutants or spreading of groundwater plumes. • Develop a stormwater management strategy involving less or no infiltration; infiltration of the full water quality or hydromodification control volume may not be feasible or desirable in cases where potential for mobilization or spreading of existing contamination cannot be avoided. D-22

NCHRP Project 25-41 4.5 Wellhead and Spring Protection Wellheads and springs, natural and man-made, are water resources that may potentially be adversely impacted by stormwater infiltration through the introduction of contaminants or alteration in water supply and levels. It is recommended that the locations of wells and springs be identified early in the design process and site design be developed to avoid infiltration in the vicinity of these resources. Setbacks of 100 to 200 feet from springs and wellheads are typical in existing guidance (OCPW 2011, VCWPD 2011; WADOE 2012). In Washington State, a wellhead protection zones are also defined by the 1-year, 5-year, and 10-year travel times. It may be appropriate to develop site-specific setbacks. Table 7 summarizes the risk indicators associated with infiltration near wellheads and springs and recommends potential mitigation measures. Table 7. Risk Indicators for Wellhead and Stream Protection Lower Risk Indicators ↔ Elevated Risk Indicators Risk Factors • No springs or wellheads in close proximity to infiltration BMPs (greater than 100 to 200 feet; and greater than 1- year travel time, as applicable) • Low levels of contaminants expected in stormwater runoff after effective treatment • Infiltration rates are low and soils have substantial pollutant attenuation capacity • Groundwater is relatively deep (greater than 10 to 20 feet) • Springs and/or wellheads in close proximity to infiltration BMPs (less than 100 to 200 feet; or within 1-year travel time, as applicable) • High levels of contamination in stormwater runoff even after effective treatment • Infiltration rates are high, and soils have more limited pollutant attenuation capacity • Groundwater is relatively shallow (less than 10 to 20 feet) • Karst topography (see Section 4.7) Potential Risk Mitigation Measures for Sites with Elevated Risk • Locate infiltration BMPs to maintain recommended setbacks. • Coordinate with local water provider or health department responsible for wellhead protection. • Follow mitigation measures for stormwater pollutants identified in Section 4.4. • Follow mitigation measures for existing soil and groundwater contamination in Section 4.3. • Develop a stormwater management strategy involving less infiltration; infiltration of the full water quality volume may not be feasible or desirable if recommended setbacks cannot be preserved and/or risks to drinking water supplied or springs cannot be avoided. D-23

NCHRP Project 25-41 4.6 Contaminant Spills As discussed in in Section 3.3, contaminant spills are a constant risk on roadways, with the potential to deposit high concentrations and volumes of pollutants before responders can control the source of the spill. However, the most common spills are petroleum hydrocarbons, and the risk of these spills contaminating groundwater can generally be mitigated with appropriate measures. Table 8 provides indicators of elevated risk and recommendations for potential mitigation measures to reduce these risks. Table 8. Risk Indicators for Contaminant Spills Lower Risk Indicators ↔ Elevated Risk Indicators Risk Factors • Lower traffic roadways with lower heavy truck traffic • Groundwater is relatively deep (greater than 10 to 20 feet) • Tributary area does not contain fueling stations, warehouses, storage tanks, or similar potential sources of spills • Higher traffic roadways, particularly for truck traffic • History of spills in nearby area may be indicative of higher risk • Groundwater is relatively shallow (less than approximately 10 feet) • Tributary area contains fueling stations, warehouses, storage tanks, or similar potential sources of spills • Karst topography (see Section 4.7) Potential Risk Mitigation Measures for Sites with Elevated Risk • Include amended media treatment layer below the infiltration media, as needed when inert sandy or gravely soils are present, to improve sorption capacity to retain spills and improve bioremediation potential. • Consider treatment train options: oil-water separator, oil contamination boom, media filters, biofiltration systems, vegetated filter strip, upstream of infiltration system. • Develop spill response plan for project if a general spill response plan is not already administered by the project owner; include provisions for protection and remediation of stormwater BMPs in the event of a spill. • Include isolation/diversion mechanism that can be activated by emergency responders in the event of a spill to prevent contamination from either entering the BMP if adequate storage is available and/or capturing the spill within the BMP prior to discharge to either groundwater or a surface receiving water; note it may not be practicable to isolate or divert flow from some BMP types, such as permeable pavement shoulders or filter strips which receive runoff as sheet flow. D-24

NCHRP Project 25-41 4.7 Special Considerations for Karst Topography Karst topography refers to a specific geologic formation that has been shaped by the dissolution of soluble bedrock elements. Karst topography is most frequently associated with limestone or dolomite rock, but may be present in other types of rock as well. The Karst landscape is characterized by containing sinkholes, underlying caves, and springs. Karst topography is present in various locales throughout the United States and is not characteristic of any one region. Karst topography has a number of unique considerations from the perspective of BMP implementation, including elevated potential for groundwater contamination from stormwater infiltration. This elevated risk is a result of thin or non-existent surface soil layers beneath the surface prior to the Karst formation, which results in little to no natural filtration of runoff. Karst aquifers also provide the potential for direct hydraulic connections from the surface to groundwater via sinkholes, springs, and caves (Donaldson 2004; Weiss et al. 2008). The EPA NPDES Menu of BMPs states that infiltration BMPs may not be used in regions with Karst topography due to the potential to create sinkholes or cause groundwater contamination (http://cfpub.epa.gov/npdes/stormwater/menuofbmps/index.cfm?action=factsheet_results&view=specific &bmp=69). Unless site-specific analyses determine that infiltration can be safely achieved without impacts through the use of robust treatment and spill containment methods, infiltration of roadway runoff should be avoided in Karst areas. 5 Consultation with Local Agencies As introduced above, infiltration may play an important role in improving groundwater resources, and projects may be able to demonstrate multiple benefits as a result of careful application of infiltration approaches. However infiltration may pose significant risks to groundwater resources and environments, and infrastructure that can be impacted by changes in groundwater levels. In general, it is a best practice for DOTs to coordinate with agencies responsible for local groundwater management and underground utilities and resource agencies whenever infiltration is considered for a project. These agencies have a vested interest in protecting groundwater supplies and underground infrastructure and usually have extensive knowledge about these resources. 5.1 Local Groundwater Suppliers Consulting with applicable groundwater supply agencies early in the project development process can help simplify the process of evaluating feasibility and desirability of infiltration. These agencies may be able to provide information to assist in evaluating the feasibility of infiltrating stormwater and may have locally-applicable criteria, maps, or other resources that have already been developed. Groundwater supply agencies with groundwater management authority may include local governments, special water supply districts, or others. A potential model for inter-agency coordination was developed in Orange County (CA) as part of development of the Orange County Technical Guidance Document (Orange County Public Works 2011). The Technical Guidance Document was developed by Orange County Public Works (OCPW), and includes guidance of evaluating the feasibility of infiltration stormwater, with consideration of groundwater quality, among other factors. The Orange County Water District (OCWD) is a public agency responsible for providing water to more than 20 cities and more than 2 million residents of Orange County. OCWD is responsible for management of the water quality and basin yield of the groundwater D-25

NCHRP Project 25-41 basin that underlies the Santa Ana River in northern Orange County. OCWD was actively involved in the development of the Technical Guidance Document, as they have an interest in enhancing groundwater supplied through infiltration and have significant concerns about the potential for groundwater contamination due to stormwater infiltration. Through a collaborative approach, OCPW and OCWD developed criteria acceptable to both agencies, including the use of groundwater levels and plume locations provided from OCWD records to help develop screening tools. The agencies also established a process by which projects identified as having an elevated risk of groundwater impacts would be submitted for review by OCWD. 5.2 Water Resources Protection Agencies Groundwater protection requirements are typically administered by state environmental quality agencies or regions of U.S. EPA. These agencies are commonly the same agencies responsible for administering stormwater and surface water regulations; however different departments may be responsible for surface water quality versus groundwater quality. Consulting with local resource agencies responsible for groundwater quality protection can also help streamline the process of evaluating feasibility and desirability of stormwater infiltration, for many of the same reasons introduced in Section 5.1. Water resources protection agencies are typically responsible for establishing groundwater quality objectives and developing plans to protect or improve groundwater quality, where needed. These agencies are also typically responsible for administering the cleanup of contaminated sites. Because of their multiple-resource purview, these agencies may be able to help provide input and strike a balance between the surface water quality benefits of stormwater infiltration versus potential impacts to water balance or groundwater quality. Consultation can also help identify if stormwater infiltration would have the potential to exacerbate existing problems that resource protection agencies are working to address. Additionally, depending on the location and design, infiltrating facilities may be considered “Class V Injection Wells” under the federal Underground Injection Control (UIC) program. Class V wells are defined as systems that are used to inject non-hazardous fluids underground, which could include dry wells or deep infiltration basins. The minimum requirements for compliance are that fluid injected underground may not endanger underground drinking water sources and that the owners/operators must submit the required inventory information to their permitting authority. Some areas may have additional requirements and more information is available on the EPA website (U.S. EPA 2012b). UIC programs are generally administered by state environmental quality agencies in most states. 5.3 Sanitation Districts and Other Underground Utilities Infiltration of stormwater from roadway projects has the potential to elevate local groundwater tables, which can increase the amount of inflow and infiltration (I&I) to the sanitary sewers and/or cause impacts to other utilities. This risk is particularly high where groundwater levels are already relatively close to the elevation of sewer and/or utility elevations. Infiltrated stormwater can migrate into sewer lines, or flow within utility trenches. The local operators of sewer collection and treatment systems should be consulted early in the planning process to evaluate whether stormwater infiltration may pose a concern related to increases in inflow and infiltration. An increase in inflow and infiltration may place an additional burden on these agencies, with respect to hydraulic conveyance capacity and waste water treatment plant treatment capacity, potentially resulting in increased incidence of sanitary sewer overflows. As a result, these agencies may have locally- D-26

NCHRP Project 25-41 applicable criteria, maps, or other resources that have already been developed to assist in identifying areas where increased stormwater infiltration may be undesirable. Similarly, agencies or companies that operate other underground utilities may have specific concerns regarding infiltration of stormwater and/or resources to assist in evaluating feasibility and desirability. 6 Summary The potential risks of stormwater infiltration on water balance and groundwater quality are influenced by many site-specific factors, as introduced in this white paper. While careful attention should be paid to any site-specific actors that indicate elevated risks, in general the risks are relatively low and can be effectively mitigated in a number of ways, including via pretreatment, soil amendments, selection of appropriate infiltration sites, and observing minimum separation criteria between infiltration BMPs and the groundwater table. Additional, watershed-scale plans prepared by local jurisdictions may provide guidance for how cumulative risks can be mitigated to an acceptable level. In cases where elevated risks cannot be mitigated and/or specific local prohibitions are in place, stormwater infiltration may not be feasible and/or desirable. 7 References Alaska Department of Environmental Conservation (2012). Spill Prevention and Response, http://dec.alaska.gov/spar/guidance.htm. Becker, S., R. Pitt, and S. Clark (July 2001). Environmental Health, Public Safety, and Social Impacts Associated with Transportation Accidents Involving Hazardous Substances. University Transportation Center of Alabama and US Dept. of Transportation. UTCA Project No. 00214. Tuscaloosa, AL. California State Water Resources Control Board (2012a). EnviroStor, http://www.envirostor.dtsc.ca.gov/public/. California State Water Resources Control Board (2012b). GeoTracker, http://geotracker.waterboards.ca.gov/. Carleton, G.B. (2010). Simulation of Groundwater Mounding Beneath Hypothetical Stormwater Infiltration Basins, United States Geological Survey, Scientific Investigations Report 2010-5012. Church, M.R., G.D. Bishop and D.L. Cassell (1995). Maps of regional evapotranspiration and runoff/ precipitation ratios in the northeast United States. Journal of Hydrology, 168, 283–298. Clark, S.E., and R. Pitt (2007). Influencing factors and a proposed evaluation methodology for predicting groundwater contamination potential from stormwater infiltration practices. Water Environment Research, 79, 29–36. Clark, S.E.; K.H Baker, J.B. Mikula, and C.S. Burkhardt (2006). Infiltration vs. Surface Water Discharge: Guidance for Stormwater Managers, Water Environment Research Fund Report 04-SW-3. Connecticut Department of Energy and Environmental Protection, List of Contaminated or Potentially Contaminated Sites in Connecticut, http://www.ct.gov/dep/cwp/view.asp?A=2715&Q= 325018, 2012. Dierkes, C.; A. Holte, and W.F. Geiger (undated). Heavy Metal Retention within a Porous Pavement Structure, Department of Civil Engineering, Urban Water Management, University of Essen. Dietz, M.E. (2007). Low impact development practices: A review of current research and recommendations for future directions, Water Air Soil Pollut, 186, 351–363. D-27

NCHRP Project 25-41 Donaldson, B.M. (2004). Highway Runoff in Areas of Karst Topography, Virginia Transportation Research Council, VTRC 04-R13. Federal Highway Administration (FHWA) (2009). Traffic Incident Management in Hazardous Materials Spills in Incident Clearance FHWA-HOP-08-058, January 2009 http://ops.fhwa.dot.gov/publications/fhwahop08058/default.htm. Fischer, D., E.G. Charles, and A.L. Baehr (2003). Effects of Stormwater Infiltration on Quality of Groundwater Beneath Retention and Detention Basins, Journal of Environmental Engineering. 129(5), 464–471. Göbel, P., H. Stubbe, M. Weinert, J. Zimmermann, S. Fach, C. Dierkes, H. Kories, J. Meßer, V. Mertsch, W.F. Geiger, and W.G. Coldewey. (2004). Near-natural stormwater management and its effects on the water budget and groundwater surface in urban areas taking account of the hydrogeological conditions. - Journal of Hydrology, 299, 267–283. Jefferson, A., A. Nolin, S. Lewis and C. Tague (2008). Hydrogeologic controls on streamflow sensitivity to climate variation. Hydrol. Process., 22, 4371–4385. Lee, J.G.; and J.P. Heaney (2003). Estimation of Urban Imperviousness and Its Impacts on Storm Water Systems, Journal of Water Resources Planning and Management, 129(5), 419–426. Los Angeles and San Gabriel Rivers Watershed Council (LASGRWC 2010), Water Augmentation Study-Research, Strategy and Implementation Report, Los Angeles and San Gabriel Rivers Watershed Council. http://www.usbr.gov/lc/socal/reports/LASGwtraugmentation/report.pdf. Miklas, S. and P. Grabowiecki (2007). Review of permeable pavement systems, Building and Environment, 42, 3830–3836. Milly, P.C.D. (1994). Climate, soil water storage, and the average annual water balance. Water Resources Research, 30(7), 2143–2156. Mohseni, O. and H.G. Stefan (2001). Water budgets of two watersheds in different climate zones under projected climate warming. Climatic Change, 49, 77–104. Moran, M.J., P.A. Hamilton, and J.S. Zogorski (2006). Volatile Organic Compounds in the Nation’s Ground Water and Drinking-Water Supply—A Summary, United States Geological Survey. USGS Fact Sheet 2006-3048. http://pubs.usgs.gov/fs/2006/3048/. Najjar, R.G. (1999). The water balance of the Susquehanna River Basin and its response to climate change. Journal of Hydrology 219, 7–19. Oregon State Univeristy, GeoSyntec Consultants, University of Florida, and The Low Impact Development Center, Inc. (2006), NCHRP Report 565: Evaluation of Best Management Practices for Highway Runoff Control, Transportation Research Board of the National Academies, Washington, D.C. New Jersey Department of Environmental Protection (NJDEP) (2012), Site Remediation Program, http://www.nj.gov/dep/srp/kcsnj/. Ng, E. and P.C. Miller (1980). Soil Moisture Relations in the Southern California Chaparral. Ecology, 61(1), 98–107. Nightingale, H.I. (1987). Water Quality Beneath Urban Runoff Management Basins. Water Resources Bulletin, 23(2), 197–205. Orange County (CA) Public Works (OCPW) (2011). Technical Guidance Document for Preparing WQMPs, http://www.ocwatersheds.com/DocmgmtInternet/Download.aspx?id=638. D-28

NCHRP Project 25-41 Pitt, R., S. Clark, and K. Parmer (1994). Protection of Groundwater from Intentional and Nonintentional Stormwater Infiltration, U.S. Environmental Protection Agency, EPA/600/SR-94/051. PB94-165354AS, Storm and Combined Sewer Program, Cincinnati, OH. 187 pp. Pitt, R., S. Clark, K. Parmer, and R. Field (eds.) (1996). Groundwater Contamination from Stormwater Infiltration, Ann Arbor Press, Ann Arbor, MI, 219 pp. Pitt, R.; S. Clark, and R. Field (1999). Groundwater contamination potential from stormwater infiltration practices, Urban Water, 1, 217–326. Rose, S. (2009). Rainfall–runoff trends in the south-eastern USA: 1938–2005. Hydrol. Process., 23, 1105–1118. Sanford, W.E. and D.L. Selnick (2012). Estimation of Evapotranspiration Across the Conterminous United States Using a Regression with Climate and Land-Cover Data. Journal of the American Water Resources Association (JAWRA) 49(1), 217–230. DOI: 10.1111 ⁄ jawr.12010 Santa Ana Regional Water Quality Control Board (SARWQCB) (2009). Orange County Municipal NPDES Storm Water Permit (Order No. R8-2009-0030, Amended by Order No. R8-2010-0062). http://www.waterboards.ca.gov/rwqcb8/water_issues/programs/stormwater/oc_permit.shtml. Spinello, A.G. and D.L. Simmons (1992). Base flow of 10 South-Shore streams, Long Island New York, 1976-85, and the effects of urbanization on base flow and flow duration. USGS Water-Resources Investigations Report 90-4205. Stephenson, J.B., and B.F. Beck (1995). Management of the discharge quality of highway runoff in karst areas to control impacts to groundwater–A review of relevant literature. Multidisciplinary Conference on Sinkholes and the Engineering and Environmental Impacts of Karst, 297–321. Strecker E.W. and A.L. Poresky (2009). Stormwater Retention On-Site, The Water Report, 2009. U.S. EPA (2010). MS4 Permit Improvement Guide U.S. Environmental Protection Agency, Office of Water, Office of Wastewater Management, Water Permits Division, April 2010, EPA 833-R-10-001. http://www.epa.gov/npdes/pubs/ms4permit_improvement_guide.pdf. U.S. EPA (2012a), Envirofacts Database, http://www.epa.gov/enviro/. U.S. EPA (2012b). Class V Wells, http://water.epa.gov/type/groundwater/uic/class5/index.cfm. U.S. EPA (2012c). U.S. EPA Permit No. DC0000221. Authorization to Discharge Stormwater Under the National Pollutant Discharge Elimination System, Municipal Separate Storm Sewer System (MS4) Permit. U.S. EPA Region 3, November 9, 2012. Ventura County (CA) Watershed Protection Division (2011). Technical Guidance Manual for the Preparation of Stormwater Quality Management Plans. http://vcstormwater.org/index.php?option=com_content&view=article&id=32&Itemid=45. Washington Department of Ecology (WADOE) (August 2012). Stormwater Management Manual for Western Washington. Publication Number 12-10-030. Ward, G.H. (1993). A water budget for the state of Texas with climatological forcing. The Texas Journal of Science, 45(3), 249–264. Washington State Department of Transportation (WSDOT) (2011). Highway Runoff Manual. M 31- 16.03. http://www.wsdot.wa.gov/publications/manuals/fulltext/M31-16/HighwayRunoff.pdf. Weiss, P.T., G. LeFevre, and J.S. Gulliver (2008). Contamination of Soil and Groundwater Due to Stormwater Infiltration Practices, University of Minnesota, St. Anthony Falls Laboratory, Project Report No.515. D-29