Stormwater Best Management Practices in an Ultra-Urban Setting: Selection and Monitoring

2.    Stormwater Management in the Ultra-Urban Environment

2.1    Introduction

The objective of this chapter is to provide a description of the characteristics of the ultra-urban environment and the types of best management practices (BMPs) typically found in ultra-urban areas. To characterize ultra-urban areas, water quality constituents typically found in stormwater runoff are defined. An overview of typical urban stormwater BMPs is presented, along with a description of the physical characteristics and BMP design considerations specific to the ultra-urban environment. Additional design criteria for each technology are presented in Chapter 3.

Federal, state, and local agencies responsible for watershed management and pollution control programs are increasingly becoming aware of the significant effects that urbanization has on the natural balance between stormwater runoff and the ecosystem of wetland and stream systems. Land use changes from agricultural to urban (urbanization) result in the conversion of pervious spaces, such as vegetated and open forested areas, to increased areas of impervious surface, resulting in increased runoff volumes and pollutant loadings.

As urbanization occurs, the quantity of stormwater runoff from the surrounding watershed increases due to the reduction in the amount of pervious spaces available to infiltrate rainwater and snowmelt. The greatly increased runoff volumes and the subsequent erosion and sediment loadings to surface waters that accompany these changes are of concern. Hydrologic and hydraulic changes result from site clearing, grading, and the addition of impervious surfaces and maintained landscapes. Hydrological changes to the watershed are directly related to an increased amount of impervious surface. Roads, parking lots, sidewalks, rooftops, and other impervious surfaces decrease the infiltrative capacity of the ground and result in changes to peak runoff frequency, time to peak, runoff volume, and runoff velocity, disturbing the receiving stream channel and wetlands. Stream channels respond by either increasing their cross-sectional area to accommodate the higher flows or down-cutting the channel. This channel instability begins a cycle of streambank erosion and habitat degradation, and may increase the frequency and severity of flooding.

In response to these detrimental ecological stresses that urbanization places on a watershed, BMPs have been developed to reduce water quantity impacts and water quality constituents normally associated with stormwater runoff from urbanization. The "ultra-urban" environment (a term coined by the city of Alexandria, Virginia; see box below) has been used to describe metropolitan areas of the country where space for stormwater BMP implementation is limited. These heavily urbanized areas present special challenges to those responsible for stormwater management. Stormwater management in these ultra-urban areas may necessitate retrofits to existing stormwater control and conveyance systems.

2.2    The Ultra-Urban Environment

Ultra-urban areas are characterized by high densities of paved surfaces or buildings that result in a high degree of imperviousness. Buildings, parking facilities, urban streets, highways, or walkways cover a majority of the land area, with imperviousness typically greater than 50 percent in ultra-urban areas, and up to 100 percent in some cases. These impervious surfaces can provide an effective environment to collect and accumulate constituents from atmospheric deposition, vehicular traffic, or other sources. Figure 1 illustrates these changes in runoff resulting from increased impervious area. High runoff conditions efficiently transport many water quality constituents. Several factors have been identified as major influences on the types of constituents and their concentrations in urban runoff. Among these are site-specific characteristics, such as land use practices. Ultra-urban areas typically contain higher population densities. These areas exhibit high levels of trash and debris, which tend to clog stormwater control structures and pollute receiving streams. In addition, the pets of the people living in ultra-urban areas are a potential concern since they deposit fecal matter in the urban environment. This fecal matter is washed off during storm events and contributes pathogenic bacteria to stormwater runoff.

Figure 1. Changes in runoff flow resulting from increased impervious area (adapted from North Carolina Department of Natural Resources and Community Development, as cited in Livingston and McCarron, 1992)

Traffic characteristics are another major influence on constituent loadings in stormwater runoff. Though mass transit methods such as subways are frequently implemented in ultra-urban areas, automobile usage is typically very high. Traffic densities are highest in urban areas, due to commuter traffic and people traveling to commercial/business areas for personal business. Increased automobile usage contributes to the constituent loadings deposited in the urban environment.

Identifying these constituent sources aids in characterizing the runoff from ultra-urban areas. This information helps to determine the most effective technologies for removing constituents from stormwater runoff, one key element in determining the type of BMP necessary to achieve water quality benefits. Fish and aquatic life concerns may also be relevant to BMP selection in some areas. Lack of oxygen and high temperatures are inter-related and very important for aquatic life. For fish and other aquatic life, temperature can be one of the most significant pollutants and presents difficult challenges in ultra-urban areas.

2.2.1    Target Water Quality Parameters

The characteristics of highway runoff have been the focus of several studies (Barrett et al., 1995). Stormwater runoff from roadways and impervious surfaces in heavily developed areas has been shown to contain significant levels of constituents such as street litter, animal and bird waste, atmospheric deposition, and inputs from urban road runoff (Shaver, 1994). Among the constituents found in highway runoff are particulates, chromium, copper, cadmium, lead, nickel, nitrogen and phosphorus, zinc, manganese, petroleum hydrocarbons, and rubber. A list of these constituents and their primary sources is included in Table 1.

Major sources of constituents on highways are vehicles and atmospheric deposition. Vehicles generate water quality constituents on highways both directly and indirectly. Vehicles contribute constituents directly from normal operation and wear of frictional parts. Cars and other vehicles were suggested as the source of over 50 percent of the total load to the Lower San Francisco Bay of three metals—copper, cadmium, and zinc (Woodward Clyde, 1992). Reportedly, tire wear could account for at least half of the total cadmium and zinc loads deposited in the bay each year, with the copper load being linked to brake pad wear. Metals that are strongly linked to cars, such as cadmium and copper, are found to have higher concentrations in runoff from streets and parking lots and minimal concentrations in roof and lawn runoff. Fish are very sensitive to metals such as copper. Vehicles can also pick up and carry solids from parking lots, urban roadways, construction sites, farms, and dirt roads and deposit them onto urban streets (Barrett et al., 1995). Through this indirect mechanism, vehicles can contribute solids and associated water quality constituents to highway surfaces. The results of several studies characterizing highway runoff constituent concentrations are presented in Table 2.

Several factors affect overall constituent loadings in street runoff. Street runoff is strongly influenced by emissions and leaks from vehicular traffic. Streets are usually directly connected to stormwater drainage systems by curb and gutter. Curb and gutter systems are not very effective at trapping and retaining fine particles that are deposited in them. Often these particles are washed into storm drains.

Disconnecting impervious surfaces and directing runoff from impervious surfaces to pervious surfaces can provide the opportunity for infiltration of stormwater runoff, reducing both stormwater quantity and constituent loading to the receiving stream. Unfortunately, there are limited opportunities for directing stormwater runoff to pervious surfaces in the ultra-urban environment since most of the land area is already covered with impervious surfaces.

In one study, Bannerman et al. (1993), streets were identified as a significant source of urban constituents in residential, commercial, and industrial areas. The study collected over 300 runoff samples from 46 micro-sites in two watersheds, sampling runoff from lawns, driveways, rooftops (residential and flat industrial), commercial and industrial parking lots, and a series of street surfaces (feeder, collector, and arterial). Streets produced some of the highest concentrations of phosphorus, suspended solids, bacteria, and several metals. In addition, streets generated a disproportionate amount of the total runoff volume from the watershed. Parking lot areas had moderately high concentrations of all constituents.

Other studies have found the concentrations of some of the metals and nutrients significantly correlated with that of total suspended solids (TSS). These results suggest that controlling TSS may result in reducing other constituents with the same particle sizes. The City of Austin (1990) found the event mean concentration (EMC) values of total phosphorus (TP), total Kjeldahl nitrogen (TKN), total organic carbon (TOC), lead (Pb), and zinc (Zn) are related to the values of TSS EMC. This correlation indicates that these constituents may be removed along with the particulates by filtration technologies such as sand filters.

The impact on the environment from the use of road salt or other deicing agents is another issue to address when characterizing runoff from streets and highways. Although sodium chloride, typically used as the primary chemical deicer in northern states, is an inexpensive and effective choice, concerns have been raised about the potential negative impacts (from chloride) on the environment, human health, roadway infrastructure, and vehicles. The Michigan Department of Transportation identified some of these impacts in a recent study (Public Sector Consultants, 1993), and those potential impacts of road salt are listed in the adjacent box.

Stormwater "hotspots" are another issue affecting ultra-urban runoff characterization. Land uses or activities that generate higher-than-normal concentrations of hydrocarbons, trace metals, or toxicants have been defined as "hotspot" areas. Increased constituent loadings from these areas may generate concerns about sediment toxicity, groundwater contamination, or toxicity in receiving surface waters. More effective stormwater treatment may be required in these areas. A preliminary list of potential stormwater hotspots is included below (Claytor and Schueler, 1996):

• Airport deicing facilities.
• Auto recycler facilities.
• Commercial nurseries.
• Commercial parking lots.
• Fueling stations.
• Fleet storage areas (bus, truck).
• Industrial rooftops (depending on the roof surface).
• Marinas.
• Outdoor container storage of liquids.
• Outdoor loading/unloading facilities.
• Public works storage areas.
• Vehicle service and maintenance areas.
• Vehicle and equipment washing/steam cleaning facilities.

Identifying "hotspot" areas will aid in determining the most effective BMP, in terms of constituent removal capability, in addition to determining the most appropriate location for the BMP. While the physical characteristics of ultra-urban areas help determine the water quality constituents contained in stormwater runoff, these physical characteristics may also limit the feasibility of various BMPs.

2.2.2    Urban Stormwater Management

The hydrologic effects of development can cause a multitude of problems, including significant flooding potentially endangering life and property. In the ultra-urban environment, stormwater runoff must be routed efficiently and effectively to minimize flooding. Therefore, when considering BMP alternatives for a specific site, both water quantity and water quality issues are taken into consideration.

Efficient collection and routing of stormwater runoff in ultra-urban areas are essential to minimize localized flooding and provide efficient drainage to properties. Increased impervious surfaces within ultra-urban areas can lead to increases in storm runoff volumes and higher runoff velocity due to increased imperviousness and reduced areas for infiltration of runoff. For example, approximately 55 percent of the rain that falls each year in forested basins in King County, Washington, eventually appears as streamflow; for an impervious basin, approximately 85 to 90 percent of annual rainfall eventually appears as streamflow (King County, Washington, 1996).

In some cases BMPs can be used to mitigate the downstream effects of increased peak flows in receiving waters. For example, detention facilities can help maintain the rate and/or duration of flows at predevelopment levels. The basic concept of a detention facility is to collect water from developed areas and release it at a slower rate than the rate at which it enters the system. The difference between the inflow and outflow is then temporarily stored in a pond or vault. Due to space limitations in the ultra-urban environment, BMPs are frequently designed to provide multiple benefits. Whenever possible, BMPs provide both water quantity and water quality benefits.

Several studies have documented the "first flush" phenomenon, indicating pollutant concentrations tend to be much higher at the beginning of a storm compared to the middle or the end (Barrett et al., 1995). This has led to requirements in some states to capture and treat the "first flush" or water quality volume (WQV) of a storm, typically the first 12.7 mm (0.5 in) of runoff from the impervious area in a drainage basin. Based on this definition of WQV, the WQV for each impervious hectare is just under 126 m³ per impervious hectare (1,800 ft³ per impervious acre). In other states, the WQV of a storm is defined as the first 25.4 mm (1 in) of runoff from the impervious area in a drainage basin. In ultra-urban areas stormwater quality requirements are frequently limited to treating only the WQV of a storm event.

Requirements for design of water quality BMPs vary around the country. For areas of existing development, requirements are not specifically identified because of the constraints of the ultra-urban environment. Local conditions, availability of funding, and problem pollutants vary widely in developed communities. Suitable areas for structural treatment systems are often unavailable in heavily urbanized areas. Retrofitting existing conveyance systems with new BMPs to provide water quality benefits may provide the only opportunity to improve the water quality of receiving streams.

Retrofitting is a process that involves the modification of existing control structures or conveyance systems, initially designed to safely convey or temporarily store stormwater runoff to minimize flooding. Retrofitting existing conveyance systems and installing a new BMP designed for water quantity control and/or water quality treatment is an option used in the ultra-urban environment. These BMPs must fit into the existing storm drain system, and match the existing hydraulic gradient. Ultra-urban BMPs are frequently configured off-line and designed to treat a certain portion (usually the "first flush") of a storm. The remainder of the runoff bypasses the water quality BMP. Where existing development or financial constraints limit the feasibility of locating different BMP options, it might be necessary to evaluate and prioritize various factors to determine the most appropriate retrofit for a particular site.

2.2.3    Ultra-Urban BMP Technologies

BMP technologies fall into two distinct categories, as illustrated in the box below. The first group of ultra-urban BMPs are control measures that are mainly associated with structural practices. These BMPs can be installed on-line, retaining and treating the entire storm event, or they can be configured off-line, treating only a portion of the storm event, with the rest of the runoff bypassing the BMP.

Ultra-urban stormwater BMPs focus on the collection, pretreatment, storage, and eventual treatment to remove constituents of a specific quantity, typically the WQV or first 12.7 mm (0.5 in) of runoff from impervious areas. Isolating the WQV of a storm requires the construction of an isolation/diversion weir, set to allow overflow when the BMP is completely full (Bell, 1996). Many BMPs designed to treat only the WQV portion of a storm can be more effective if designed off-line for this reason. These structural BMPs are generally implemented under conditions where land space requirements are considered to be a constraint. Such is the case in an ultra-urban setting, where retrofitting is a common practice.

Structural BMP technologies typically use one or more of the following treatment mechanisms to achieve water quality benefits:

• Detention (particle settling).
• Adsorption (chemical and physical processes).
• Biological removal mechanisms.
• Filtration (physical process).

Infiltration technologies include infiltration basins, trenches, and bioretention. Infiltration technologies use the interaction of the chemical, physical, and biological processes between soils and water to filter out sediments and other soluble constituents from urban runoff. As the stormwater percolates into the ground, fine material suspended in stormwater is captured within the soil. The resulting treated runoff percolates through to the groundwater. Infiltration trenches are well-suited to the ultra-urban environment since they can be located completely underground. However, they are limited to areas that have specific soil types and groundwater table characteristics and may have higher maintenance costs because they are completely underground. Bioretention is a relatively new type of infiltration technology potentially suited to ultra-urban areas. Bioretention areas manage stormwater runoff by using a conditioned soil layer that contains a mixture of detritus, humus, and mineral and biological complexes in a shallow depressed area. The soil layer and the microbes living in the soil enhance filtration, and the vegetation aids constituent removal. Since small bioretention areas can be located in medians, parking lot islands, or grassy areas along streets, they are ideal for the constricted ultra-urban environment.

Detention and retention practices temporarily store stormwater to control runoff, and settle and retain suspended solids and associated constituents. Stormwater ponds, including retention or wet ponds, dry detention ponds, and pond/wetland combinations, have been the traditional detention and retention practices used to provide both water quantity and water quality control. Generally, dry ponds are designed to provide stormwater hydrologic control through detention and retention or wet ponds are designed with a permanent pool to also provide for treatment of stormwater pollution. Wetlands and shallow marsh systems use the nutrient uptake of vegetation to enhance constituent removal. Ponds are effective technologies for reducing constituent loadings from stormwater runoff. They can be implemented in ultra-urban areas by siting a number of smaller ponds rather than one large pond system to provide some measure of water quality control. Opportunities exist, however, where sufficient land area is available to implement a pond BMP as part of an urban park setting, which provides both aesthetic and educational benefits. Ponds or pond/wetland combinations should be considered as a possibility where these opportunities exist.

Stormwater filtering systems have been developed and used successfully in ultra-urban areas due to their relatively small footprint and moderate physical requirements (modest head requirements and no soil restrictions). A number of filtering systems have been developed for use in heavily urbanized areas. These include the Delaware sand filter, Austin sand filter, packed bed filter, leaf compost filter, and vertical sand filter. Each of these filters provides the same basic components: (1) a sedimentation area to retain the largest particles, which may clog the filter medium, and (2) a filter chamber containing the filter medium, that filters and removes soluble constituents. Most stormwater filters are designed to treat only a portion of a storm event, usually the WQV, and therefore are configured off-line. The Delaware sand filter is an example of a stormwater BMP that has been modified to fit into the ultra-urban environment. Its shallow configuration, with the sedimentation and filtration chambers below ground but relatively accessible for periodic maintenance, fits in well with the limited space environment of ultra-urban areas.

Vegetated practices include technologies such as grassy swales and filter strips. Vegetated swales and filter strips are designed to capture and filter runoff, with a portion of the runoff infiltrating into the soil. Vegetation is used to enhance biological uptake of stormwater constituents. These BMPs can be easily used along roadway corridors and require minimal maintenance (mowing to maintain vegetation at a certain height). Buffer strips of only a few meters can remove a significant amount of suspended constituents from highway runoff (Yu et al., 1995).

Manufactured/pretreatment technologies include water quality inlet BMPs such as oil/grit separators, water quality access holes, and catch basin inserts. Water quality inlets typically use detention to enhance removal of both coarse and fine sediments, trap debris and trash, and separate oil and grease from the runoff. Though most of these technologies are well suited to the ultra-urban environment due to their minimal space requirements and physical restrictions, only limited independent evaluation of these technologies has been performed to date. To maintain their effectiveness, provisions must be made for frequent cleaning and inspection. FHWA has previously not recommended the use of water quality inlet BMPs such as oil/grit separators for highway applications, although they may perform adequately in maintenance yards with proper maintenance after installation. Finally, many states recommend they be considered only for pretreatment applications or as a last alternative.

Porous pavements are included in the category of infiltration technologies but are unique in their design. While conventional pavement results in increased runoff, porous pavements allow stormwater to percolate through the pavement and infiltrate into the soil below. Porous asphalt, concrete, and interlocking paving stones allow streets, parking lots, sidewalks and other impervious surfaces to retain their natural infiltrative capacity, while also providing the functional features necessary for automobile and pedestrian traffic. Porous pavements must be correctly sited, designed, and installed, as well as periodically maintained, for them to function properly over their life span. They continue to be studied and evaluated to determine whether any reduction in infiltrative capacity occurs over time due to an accumulation of sediments; the longevity of porous materials is part of this evaluation. Porous asphalt, concrete, and pavers have been used in urban and ultra-urban areas. In the highway setting, porous pavements could be used on shoulders and rest areas or parking areas for cars.

The second group of BMPs are viewed as preventive measures and are to a large extent associated with nonstructural practices. Nonstructural measures, such as streetsweeping, have been implemented in urban areas to reduce constituent loadings in stormwater runoff, thereby reducing the need for more expensive structural measures. In a study of stormwater characteristics for various land uses in the city of Austin (City of Austin, 1990), constituent median EMCs were reduced in areas where streetsweeping occurred at least once per week, versus those areas that did not receive maintenance. Streetsweeping technologies, adapted from those used to remove spilled coal and coal dust from along railroad tracks, have recently been used in stormwater management applications and may reduce the need for more expensive structural controls.

The feasibility of various BMP technologies in the ultra-urban environment is limited by particular design considerations specific to each site. Conventional structural BMPs, such as extended detention dry ponds, wet ponds, and infiltration basins are often impractical to implement in ultra-urban environments. With older cities frequently located in river valleys, high water tables and the prevalence of marine clays may preclude the use of infiltration technologies (Bell, 1996). These limitations have generated modifications to existing structural BMPs, and in some cases have led to the design of new BMPs that can properly treat or dispose of urban stormwater constituents.

2.2.4    Ultra-Urban Design Considerations

Design considerations determined by the physical characteristics of ultra-urban areas fall into several categories. The considerations for a given ultra-urban site include:

• Space limitations.
• Economic considerations.
• Conflicts with existing utilities.
• Safety issues.
• Maintenance requirements for BMPs.

Space Limitations: The limited space available for BMPs can be a result of physical limitations, particularly in retrofit situations where no prior planning for BMP requirements has been performed. Build-out conditions usually exist in metropolitan areas, particularly in older business districts of the cities. Lot line-to-lot line structures are the norm, leaving limited space available for BMP implementation. BMPs are frequently located below ground, often the only appropriate and cost-effective location in ultra-urban areas.

Retrofitting an existing stormwater conveyance system with a water quality BMP involves designing the BMP to fit in with the existing storm drain system. Existing hydraulic gradients between source areas and final discharge to the receiving stream may limit the type of BMP feasible for a specific site. Some BMPs, such as filtering systems, have modest head requirements that can fit into the existing hydraulic gradient of the storm drain system. Head requirements for other BMPs might preclude their application in certain areas.

Economic considerations associated with BMP implementation often determine whether to locate installations below ground. Since there is less land available, property values are at a premium. The cost of real estate in areas of high land values and the lost opportunity costs of additional development that must be given up for conventional BMPs located at the surface frequently outweigh the cost of more expensive BMP options such as sand filters. Property values typically are over $215 per square meter ($20 per square foot) in ultra-urban areas (Bell, 1996). Right-of-way width for highways implemented in ultra-urban areas is usually minimized to reduce land requirements. This practice results in limited space available for stormwater BMP implementation. An example illustrating economic considerations in determining the feasibility of various BMP options is shown in the adjacent box.

Locating BMPs below ground may require additional structural measures to ensure stormwater management structures can withstand vehicular traffic in areas where they are implemented. Delaware sand filters are typically designed to be located along the periphery of parking areas. Alaska Marine Lines located Delaware sand filters along the perimeter of a paved area used to ship, handle, and store cargo containers. Concrete lids (AASHTO H-20) are segmented for ease in removal and cover the complete trench area, with vertical scuppers for stormwater inflow. The lid material is designed to withstand loadings from pedestrians, bicycles, and occasional light vehicles (Spearman and Beard, 1995). An example illustrating the economic considerations resulting from additional structural measures is provided in the adjacent box.

Conflicts with existing utilities may also limit opportunities for BMP installation. Utilities are frequently located below ground, which may also be the only feasible location for stormwater BMPs. In ultra-urban environments, water and sewer piping, natural gas lines, and telephone and electrical conduits are frequently located in rights-of-way, also often the only available space for a BMP. The BMP might need to be modified to fit into the space available without disrupting existing utilities, incurring additional design costs. Or the utilities might need to be relocated in order to install the BMP, adding to the construction cost of the installation.

Two main factors concerning safety issues should be considered in evaluating the feasibility of various BMP options. First, ultra-urban areas are heavily populated, adding to safety concerns when considering potential BMPs such as ponds, wetlands, and surface sand filters. These open surface systems may require additional measures such as fencing to ensure the safety of the public. Second, locating BMPs underground, often the only feasible location in ultra-urban environments, presents additional maintenance requirements that trigger worker safety regulatory requirements. Depending on the type of BMP, these installations may be considered confined spaces. Confined spaces have specific requirements to ensure safe access to the unit, which must be followed each time the BMP is inspected or maintenance is performed.

Maintenance requirements must be carefully planned and implemented when BMPs are located completely below the surface and access is limited to access hole openings or the removal of concrete panels. As previously mentioned, underground BMPs may be considered confined spaces and require additional measures to ensure safe access for inspection or maintenance. Due to these potential restrictions or additional measures, BMP technologies that require periodic maintenance on an annual or semiannual basis are often preferred to those requiring more frequent maintenance efforts. Difficulty in performing the maintenance (increased level of effort) increases the cost of the required maintenance.

Stormwater management in the ultra-urban environment is determined by a number of different factors, including runoff characteristics, site design considerations, and the feasibility of implementing various BMP options based on these considerations. These factors are interdependent and may restrict the types of BMPs that can be implemented in a given location. The site design considerations and management issues related to each particular site must be analyzed and prioritized to provide sufficient information to evaluate the feasibility of various BMP options.

References

Bannerman, R.T., D.W. Owens, R.B. Dodds, N.J. Hornewer. 1993. Sources of Pollutants in Wisconsin Stormwater. Wat. Sci. Tech., Vol. 28, No. 3-5, pp. 241-259.

Barrett, M.E., R.D. Zuber, E.R. Collins III, J.F. Malina, Jr., R.J. Charbeneau, and G.H. Ward. 1995. A Review and Evaluation of Literature Pertaining to the Quantity and Control of Pollution from Highway Runoff and Construction. 2nd ed. Technical Report CRWR 239. Center for Research in Water Resources, The University of Texas at Austin.

Bell, W., L. Stokes, L.J. Gavan, and T.N. Nguyen. 1995. Assessment of the Pollutant Removal Efficiencies of Delaware Sand Filter BMPs. City of Alexandria, Department of Transportation and Environmental Services, Alexandria, VA.

Bell, W. 1996. BMP Technologies for Ultra-urban Settings. In Proceedings of Effective Land Management for Reduced Environmental Impact, Tidewater's Land Management Conference on Water Quality, August 22, 1996.

Bell, W., D.M. Grove, and T.N. Nguyen. 1998. Appropriate BMP Technologies for Ultra-Urban Applications. Prepared for A Regional Conference on Minimizing Erosion, Sediment, and Stormwater Impacts: Protection and Enhancement of Aquatic Resources in the 21st Century, sponsored by the State of Delaware Department of Natural Resources and Environmental Control, Division of Soil and Water Conservation, September 15-18, 1998.

City of Austin. 1990. Stormwater Pollutant Loading Characteristics for Various Land Uses in the Austin Area. Environmental Resources Management Division, Environmental and Conservation Services Department, City of Austin, Austin, TX.

Claytor, R.A., and T.R. Schueler. 1996. Design of Stormwater Filtering Systems. The Center for Watershed Protection, Silver Spring, MD.

King County. 1996. Surface Water Design Manual. Draft. King County Surface Water Management Division, King County, WA.

Livingston, E.H., and E. McCarron. 1992. Stormwater Management: A Guide for Floridians. Florida Department of Environmental Regulation, Tallahassee, FL.

Pitt, R.E. 1996. The Control of Toxicants at Critical Source Areas. In Proceedings of Effects of Watershed Development and Management on Aquatic Ecosystems Symposium, presented by the Engineering Foundation, Snowbird, UT, Aug. 4-9, 1996.

Public Sector Consultants. 1993. The Use of Selected Deicing Materials on Michigan Roads: Environmental and Economic Impacts. Michigan Department of Transportation, Lansing, MI.

Shaver, E. 1994. Sand Filter Design for Water Quality Treatment. Proceedings of the Urban Runoff and Receiving Systems Conference, ed. E. Herricks. ASCE, New York, NY.

Spearman, J.W., and S.R. Beard. 1995. Design, Construction, and Evaluation of a Sand Filter Stormwater Treatment System. Part I, Design and Construction. Report to Alaska Marine Lines, Seattle, WA.

USEPA. 1993. Guidance Specifying Management Measures For Sources Of Nonpoint Pollution In Coastal Waters. EPA-840-B-92-002. U.S. Environmental Protection Agency (USEPA), Office of Water, Washington, DC.

Woodward Clyde Consultants. 1992. Source Identification and Control Report. Santa Clara Valley Nonpoint Source Control Program.

Yu, S.L., R.J. Kaighn, S.L. Liao, and C.E. O'Flaherty. 1995. The Control of Pollution in Highway Runoff Through Biofiltration. Volume I: Executive Summary. Virginia Transportation Research Council, Charlottesville, VA.