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.
6 Innovative Stormwater Management and Regulatory Permitting There are numerous innovative regulatory strategies that could be used to improve EPAâs stormwater program. This chapter first outlines a substantial departure from the status quo, namely, basing all stormwater and other wastewa- ter discharge permits on watershed boundaries instead of political boundaries. Watershed-based permitting is not a new concept, but it has been attempted in only a few communities. Development of the new permitting paradigm is fol- lowed by more modest and easily implemented recommendations for improving the stormwater program, from a new plan for monitoring industrial sites to en- couraging greater use of quantitative measures of the maximum extent practica- ble requirement. The recommendations in the latter half of the chapter do not preclude adoption of watershed-based permitting at some future date, and indeed they lay the groundwork in the near term for an eventual shift to watershed- based permitting. WATERSHED PERMITTING FRAMEWORK FOR MANAGING STORMWATER At its initial meeting in January 2007, the committee heard opinions that collectively pointed in a new direction for managing and regulating stormwater that would differ from the end-of-pipe approach traditionally applied by regula- tory agencies under the National Pollutant Discharge Elimination System (NPDES) permits and be based instead on a watershed framework. Indeed, the U.S. Environmental Protection Agency (EPA) has already given substantial thought to watershed permitting and issued a Watershed-Based NPDES Permit- ting Policy Statement (EPA, 2003a) that defined watershed-based permitting as an approach that produces NPDES permits that are issued to point sources on a geographic or watershed basis. It went on to declare that, âThe utility of this tool relies heavily on a detailed, integrated, and inclusive watershed planning process. Watershed planning includes monitoring and assessment activities that generate the data necessary for clear watershed goals to be established and per- mits to be designed to specifically address the goals.â In the statement, EPA listed a number of important benefits of watershed permitting: More environmentally effective results; Ability to emphasize measuring the effectiveness of targeted actions on improvements in water quality; 475
476 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES Greater opportunities for trading and other market-based approaches; Reduced cost of improving the quality of the nationâs waters; More effective implementation of watershed plans, including total maximum daily loads (TMDLs); and Other ancillary benefits beyond those that have been achieved under the Clean Water Act (e.g., integrating CWA and Safe Drinking Water Act [SDWA] programs). Subsequent to the policy statement, EPA published two guidance docu- ments that lay out a general process for a designated state that wishes to set up any type of permit or permits under CWA auspices on a watershed basis (EPA, 2003b, 2007a). It also outlined a number of case studies illustrating various kinds of permits that contain some watershed-based elements. Box 6-1 de- scribes in greater detail the more recent report (EPA, 2007a) and its 11 âoptionsâ for watershed-based permitting. Unfortunately, the EPA guidance is lacking in its description of what constitutes watershed-based permitting, who would be covered under such a permit, and how it would replace the current program for municipalities and industries discharging stormwater under an individual or general NPDES permit. Few examples are given, some of which are not even watershed-based, with most of the examples involving grouping municipal wastewater treatment works under a single permit with no reference to stormwa- ter. Most of the 11 options are removed from the fundamental concept of water- shed-based permitting. Finally, the guidance fails to elaborate on the policy statement goal to make water quality standards watershed-based. The commit- tee concluded that, although the EPA documents lay some groundwork for wa- tershed-based permittingâespecially the ideas of integrated municipal permits, water quality trading, and monitoring consortiaâthe sum total of EPAâs analy- sis does not define a framework for moving toward true watershed-based per- mitting. The guidance attends to few of the details associated with such a pro- gram and it has made no attempt to envision how such a system could be ex- tended to the states and the municipal and industrial stormwater permittees. This chapter attempts to overcome these shortcomings by presenting a more comprehensive description of watershed-based permitting for stormwater dis- chargers. The approach proposed in this chapter fits within the general framework outlined by EPA but goes much further. First, it is intended to replace the pre- sent structure, instead of being an adjunct to it, and to be uniformly applied na- tionwide. The proposal adopts the goal orientation of the policy statement and then extends it to root watershed management and permitting in comprehensive objectives representing the ability of waters to actually support designated bene- ficial uses. The proposal builds primarily around the integrated municipal per- mit concept in the policy statement and technical guidance. Like EPAâs outline, the committee emphasizes measuring the effectiveness of actions in bringing improvements, but goes on from there to recommend a set of monitoring activi-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 477 BOX 6-1 EPAâs Current Guidance on Watershed-Based Permitting Rather than explicitly define watershed based permitting, the EPAâs recent guidance (EPA, 2007a) groups a large number of activities as having elements of watershed-based permitting, and defines how each might be utilized by a community. They are NPDES permitting development on a watershed basis, Water quality trading, Wet weather integration, Indicator development for watershed-based stormwater management, TMDL development and implementation, Monitoring consortium, Permit synchronization, Statewide rotating basin planning, State-approved watershed management plan development, Section 319 planning, and Source water protection planning. Taking these topics in order, the first option is generally similar to that in EPA (2003a,b), but with some more detail on possible permitting forms. âCoordinated individual permitsâ implies that individual permits would be made similar and set with respect to one another and to a holistic watershed goal. The nature of such permits is not fully described, and there are no examples given. An âintegrated municipal permit,â also presented in the earlier policy statement, would place the disparate individual NPDES permits in a munici- pality (e.g., wastewater plants, combined sewer overflows, municipal separate storm sewer systems [MS4s]) under one permit. However, such a permit is not necessarily watershed- based. Finally, the âmulti-source permitâ could go in numerous directions, none of which are described in detail. In one concept, all current individual permittees who discharge a common pollutant into a watershed would come under one new individual permit that regu- lates that pollutant, while keeping the existing individual permits intact for other purposes. The Neuse River Consortium is given as an example. Alternatively, a multi-source permit could cover all dischargers of a particular type now falling under one individual permit that regulates all of their pollutants (no examples are given). In yet another application, this permit could be a general permit, and it would be identical to the existing general permits, except that it would be organized along watershed boundaries. As above, it could be re- fined on the basis of pollutant or discharger type. The other ten options are more distant from the fundamental concept of watershed- based permitting. The water quality trading description is minimal, though it does mention a new EPA document that gives guidance to permittees for trading. Wet weather integra- tion, the third topic, can mean any number of things, from creating a single permit to cover all discharges of pollutants during wet weather in a municipality, as described above for âcoordinated individual permits,â to just having all the managers of the systems get together and strategize. Although a stated goal is to reduce the amount of water in the sewer sys- tem after a storm, this integration is not particularly well defined in the document, nor is it well differentiated from other activities that would normally occur under an MS4 permit. continues next page
478 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-1 Continued Indicator development for watershed-based stormwater management refers to identi- fying indicators that are better than one or a few pollutants at characterizing the degree of impairment wrought by stormwater. Stormwater runoff volume is one indicator being de- veloped by Vermont, and percent impervious surface is another. As discussed in Chapter 2, some states have long used biological indicators that integrate the effects of many pol- lutants as well as physical stresses such as elevated flow velocities. Indicators can be used as TMDL targets or as goals in NPDES permits. Identifying and adopting indicators is, essentially, a prerequisite to implementing some of the other options listed above. Regarding the next topic on the list, the option of TMDL development is obvious, since the TMDL program is by definition watershed based. If it can be made the highest priority, and if stormwater is a contributor, then the implementation plan can be an excellent way to combat stormwater pollution on a watershed basis. Reducing the contribution of the pollut- ant from a stormwater source can involve water quality trading, better enforcement of exist- ing permits, or creating new watershed-based permits. Hence, again, there is considerable overlap with the previously discussed options. Developing a monitoring consortium is an option that works when sufficient data are not available to do much else. The concept mainly refers to monitoring of ambient waters. The activity is shared among partners (e.g., all wastewater plants in a region), with the goal of collecting and analyzing enough data to improve management decisions on a watershed basis, instead of for a single plant. The following topic, permit synchronization, refers to having all permits within a water- shed expire and be renewed simultaneously. This approach could be helpful for streamlin- ing administrative, monitoring, and management tasks associated with maintaining the permits. Some states have operated in this way, whereas others have decided not to. It is one way to coordinate permits in cases where other types of watershed-based permitting would not work. Similarly, the statewide rotating basin approach, used by many states, relies on a five-year cycle. The state is divided into major watersheds, and each watershed is in a different stage of the cycle every year. It is a way to distribute the workload such that there is never a year when, for example, every watershed would require monitoring. Since it is a statewide program, how it relates to a watershed-based permitting situation is not at all clear. ties designed to support active adaptive management to achieve objectives, aswell as to assess compliance. Credit trading, indicator development, the rotat- ing basin approach, and monitoring should be part of management and permit- ting programs within watersheds, and ideas are advanced to develop these and other elements. In addition to building on the work of EPA, the proposed approach tackles many of the impediments to effective watershed management identified in the National Research Council (NRC) treatise on watershed management (NRC, 1999). That report noted that watershed approaches are easiest to implement at the local level; thus, the approach developed in this chapter is a bottom-up proc- ess in which programmatic responsibility lies mainly with municipalities. Be- cause the natural boundaries of watersheds rarely coincide with political juris- dictions, watersheds as geographic areas are less useful for political, institu- tional, and funding purposes, such that initiatives and organizations directed at watershed management should be flexible. The proposed approach recognizes this reality and makes numerous suggestions for pilot testing, funding, and insti- tutional arrangements that will facilitate success. Finally, NRC (1999) notes the
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 479 With regard to the next topic, there has been a great deal of watershed planning around the nation and tremendous variety in form and comprehensiveness. Plans gener- ally contain some information on the state of the watershed, goals for the watershed, and activities to meet those goals. Development of such plans in areas that do not have them could facilitate watershed-based permitting by providing much needed information about conditions, sources of pollutants, and methods to reduce pollution. According to EPA, a watershed plan may or may not indicate the need for watershed-based permitting. The Section 319 Program refers to voluntary efforts to reduce pollution from nonpoint sources. The program in and of itself is not relevant to NPDES permits, since it deals strictly with activities that are not regulated. However, these activities could be traded with more traditional stormwater practices as part of a watershed-based effort to reduce overall pollution reaching waterbodies. Many watershed plans must consider guidance for the 319 program in order to get funding for their management activities. If the watershed in question contains a drinking water source (either surface water or groundwater), then a good source water protection plan can have a significant impact on NPDES permitting in a watershed. Information collected during the assessment phase of source water protection could be used to help inform watershed-based permitting. Also, NPDES permits could be rewritten taking into account the proximity of discharges to source water intakes. Following its coverage of the 11 options, EPA (2007a) gives a hypothetical example of picking six of the options to develop permitting for a watershed. It discusses how the op- tions might be prioritized, but in a very qualitative manner, according to considerations such as availability of funding and personnel, stakeholder desires, environmental impacts, and sequencing of events. Chapter 1 of the report ends with a list of performance goals that might apply to the 11 options. Chapter 2 further explains the multi-source watershed-based permit, discussing, for example, who would be covered by it, who would administer it, and how credit trading fits in. The chapter has a lot of practical, although quite intuitive, information about how to write such a permit. Much of the decision making is left to the permit writer. There are discussions of effluent limitations, monitoring requirements, reporting and record keeping, special conditions, and public notice. Chapter 3 follows by presenting case studies, al- though fewer than appeared in 2003 and not all truly watershed based. need to âdevelop practical procedures for considering risk and uncertainty in real world decision-making in order to advance watershed management.â The proposed revised monitoring system presented later in this chapter is designed to provide information in the face of ongoing uncertainty, i.e., adaptive manage- ment in a permitting context. Watershed Management and Permitting Issues There are many implications of redirecting the stormwater management and regulatory system from a site-by-site, SCM-by-SCM approach to an emphasis on attainment of beneficial uses throughout a watershed. Most fundamentally, the programâs focus would shift to a primary concentration on broad goals in terms of, for example, achieving a targeted condition in a biological indicator associated with aquatic ecosystem beneficial uses or no net increase in elevated
480 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES flow duration. Application of site-specific stormwater control measures (SCMs) would no longer constitute presumptive evidence of permit compliance, as is often the case in permits now, although it would still be an essential means to meeting goals. Achieving those goals, however, would form the compliance criteria. In recognition of the demonstrated negative effects of watershed hydrologic modification on the attainment of beneficial uses, the proposal steps beyond the generally prevailing practice by embracing water quantity as a concern along with water quality. The inclusion of hydrology is consistent with the CWA on several grounds. First, elevated runoff peak flow rates and volumes increase erosive shear stress on stream beds and banks and directly contribute particulate pollutants to the flow (such as suspended and settleable solids, as well as nutri- ents and other contaminants bound to the soil material). Conversely, reduced dry-weather flows often occur in urban streams as a result of lost groundwater recharge and tend to concentrate pollutants and, hence, worsen their biological effects. Moreover, pollutant mass loading is the product of concentration and flow volume, and thus increased wet-weather surface runoff directly augments the cumulative burden on receiving waters. Finally, regulatory precedent for incorporating hydrology exists, as demonstrated by Vermontâs stormwater pro- gram (LaFlamme, 2007). At this time, stormwater management and regulation are divorced from the management and regulation of municipal and industrial wastewater. A true wa- tershed-based approach would incorporate the full range of municipal and indus- trial sources, including (1) public streets and highways; (2) municipal stormwa- ter drainage systems; (3) municipal separate and combined wastewater collec- tion, conveyance, and treatment systems; (4) industrial stormwater and process wastewater discharges; (5) private residential and commercial property; and (6) construction sites. These many sources represent an array of uncoordinated permits under the current system and a strong challenge to developing a water- shed-based approach. As pointed out in Chapter 2, multi-source considerations are an implicit facet of TMDL assessments, wherein states must consider both point and nonpoint sources. EPA (2003b) identified, among other possible per- mit types, an Integrated Municipal NPDES Permit, which would bundle all re- quirements for a municipality (e.g., stormwater, combined sewer overflows, biosolids, pretreatment) into a single permit. The Tualatin River watershed in Oregon has faced this challenge, at least in part, through an innovative water- shed permit that combines both wastewater treatment and stormwater, brings in management of agricultural contributions to thermal pollution, and allows for pollutant trading among sources (see Box 6-2). It appears that the various par- ticipating parties did not use their energies in trying to allocate blame but instead determined the most effective and efficient ways of improving conditions. For example, the municipal permittees willingly offered incentives to agricultural landowners to plant riparian shade trees as an alternative to more expensive means of reducing stream temperatures under their direct control. Indeed, with agriculture not being regulated by the Clean Water Act, watershed permitting
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 481 BOX 6-2 Watershed-Based Permitting in Oregon Clean Water Services is a wastewater and stormwater utility that covers a special ser- vice district of 12 cities and unincorporated areas in urban Washington County, Oregon. It was originally chartered in the 1970s as the Unified Sewerage Agency to consolidate the management of 26 âpackageâ wastewater treatment facilities. Its responsibilities expanded to stormwater management in the early 1990s and it now serves nearly 500,000 customers. There are four wastewater treatment plants (WWTPs) in the district, with a dry weather capacity of 71 million gallons per day (MGD). During low-flow months, the discharge from these plants can account for 50 percent of the water in the Tualatin River. The district also own rights to one-quarter of the stored water in Hagg Lake. The land use in the watershed is about one-third urban, one-third agriculture, and one-third forest. In 2001, the region was faced with TMDLs on the Tualatin River or its tributaries for to- tal phosphorus, ammonia, temperature, bacteria, and dissolved oxygen. By 2002, the area was also dealing with four expired NPDES permits and one expired MS4 permit (all of which had been administratively extended), approval of a second TMDL, and an Endan- gered Species Act (ESA) listing. The region decided that it wanted to try to integrate all of these programs using a watershed-based regulatory framework. This would include a TMDL implementation mechanism, an ESA response plan, and integrated water resources management (meaning that water quantity, water quality, and habitat considerations would be made at the same time). Prior to integration, water quality was covered by the TMDL and NPDES programs, but these programs did not cover water quantity and habitat issues. The ESA listing addressed the habitat issues, but it was done totally independently of the TMDLs and NPDES permits. Thus, the region applied for an integrated municipal NPDES permit that bundles all NPDES permit requirements for a municipality into a single permit, including publicly owned treatment works (POTWs), pretreatment, stormwater, sanitary sewer overflows, and biosol- ids. Initially, it encompassed the four WWTP permits, the one MS4 permit, and the indus- trial and construction stormwater permits. The hope was that this would streamline multiple permits and capture administrative and programmatic efficiencies; provide a mechanism for implementing more cost-effective technologies and management practices including water quality credit trading; integrate watershed management across federal statutes such as the CWA, SDWA, and ESA; and encourage early and meaningful collaboration and coopera- tion among key stakeholders. This case study was successful because a single entityâClean Water Servicesâwas already in charge of what would have otherwise been a group of individual permittees. Furthermore, all the NPDES permits had expired and the TMDL had just been issued, pro- viding a window of opportunity. The state regulatory agency was very willing, and EPA provided a $75,000 grant. Finally, there was a robust water quality database and modeling performed for the area because of the previous TMDL work. The watershed-based permit, the first in the nation, was issued February 26, 2004. Among its unique elements are an intergovernmental agreement companion document signed by the Oregon Department of Environmental Quality (DEQ), water quality credit trading, and consolidation of reporting requirements. The water quality trading is one of the most interesting elements, and sev- eral variations have been attempted. Biological oxygen demand (BOD) and NH3 have been traded both intra-facility and inter-facility. The temperature TMDL on the Tualatin River is a particularly interesting example of trading because it helped to bring agriculture into the process, where it would otherwise not have been involved. Along the length of the river, there are portions that exceed the tem- perature standard. A TMDL allocation was calculated that would lower temperatures by the continued next page
482 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-2 Continued same amount everywhere, such that there would be no point along the river that would be in exceedance. Options for reducing temperature include reducing the influent wastewater temperature (which is hard to do), reducing the total WWTP discharge to the Tualatin River (which is not practical), mechanically cooling or refrigerating WWTP discharge (which would require more energy), or trading the heat load via flow augmentation and increased shading (which is what was attempted). Clean Water Services choose to utilize a market-based, watershed approach to meet the Tualatin temperature TMDL. It was market-based because it had financial incentives for certain groups to participate, it was cost-effective, and it provided ancillary ecosystem services. It was a watershed-based approach because it capitalized on the total assimila- tive capacity of the basin. What was done was to (1) provide cooling and in-stream flow augmentation by releasing water from Hagg Lake Reservoir, and (2) trade riparian stream surface shading improvement credits. They also reused WWTP effluent in lieu of irrigation withdrawals. For the riparian shading, they developed an âenhancedâ CREP program to increase the financial incentives to rural landowners (with Clean Water Services paying the difference over existing federal and state programs). Clean Water Services also made incentive payments to the Soil and Water Conservation District to hire people to act as agents of Clean Water Services. Oregon DEQâs Shadalator model was used to quantify thermal credits for riparian planting projects, which required that information be collected at 100-foot increments along the stream on elevation, aspect, wetted width, Nordfjord-Sogn Detachment Zone, channel incision, and plant type and planting corridor width. To summa- rize, over the five-year term of the permit, Clean Water Services will release 30 cfs/d of stored water from Hagg Lake each July and August and shade roughly 35 miles of tributary riparian area (they have already planted 34 miles of riparian buffer). This plan involved an element of risk taking, since the actions of unregulated parties (such as farmers) have sud- denly become the responsibility of Clean Water Services. and initiatives of this type represent the best, and perhaps only, mechanism for ameliorating negative effects of agricultural runoff that, left unattended, would undo gains in managing urban runoff. The Neuse River case study, discussed later in this chapter, is another example of bringing agricultural contributions to aquatic degradation under control, along with urban sources, through a water- shed-based approach. Significant disadvantages of the current system of separate permits for mu- nicipal, construction, and industrial activities are (1) the permits attack the prob- lem on a piecemeal basis, (2) they are hard to coordinate because they expire at different times, (3) they are not designed to allow for long-term operation of SCMs, and (4) they do not cover all discharges. A solution to these problems would be to integrate all discharge permitting under municipal authority, as is proposed here. The lead permittee and co-permittees would bear ultimate re- sponsibility for meeting watershed goals and would regulate all public and pri- vate discharges within their jurisdictions to attain them. Municipalities are the natural focus for this role because they are the center of land-use decisions throughout the nation. Municipalities must be provided with substantially greater resources than they have now to take on this increased responsibility. Beyond funding, regula-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 483 tory responsibilities must be realigned to some degree. The norm now is for states to administer industrial permits directly and generally attend to all aspects of permit management. However, states, more often than not, are unable be- cause of resource limitations to give permittees much attention in the form of inspection and feedback to ensure compliance. At the same time, some states, explicitly or implicitly, expect municipal permittees to set up programs to meet water quality standards in the waters to which all land uses under their jurisdic- tions discharge.1 It only makes sense in this situation to have designated states (or EPA for the others) specify criteria for industrial and construction permits but revise regulations to empower and support municipal co-permittees in com- pliance-related activities. This paradigm is not unprecedented in environmental permitting, as under the Clean Air Act, states develop state implementation plans for implementation by local entities. For this new arrangement to work, states would have to be comfortable that municipalities could handle the respon- sibility and be able to exercise the added authority granted. The committeeâs opinion is that municipalities generally do have the capability, working together as co-permittees with a large-jurisdiction lead permittee and with guidance and support from states. It bears noting at the outset that the proposed new program would not re- duce the present systemâs reliance on general permits. Whereas a general permit now can be issued to a group of municipalities having differing circumstances, under the new system a permit could just as well be formulated in the same way for a group of varying watersheds. General industrial and construction permits would be just as prevalent too. Toward Watershed-Based Permitting Watershed-based permitting is taken in this report to mean regulated allow- ance of discharges of water and wastes borne by those discharges to waters of the United States, with due consideration of (1) the implications of those dis- charges for preservation or improvement of prevailing ecological conditions in the watershedâs aquatic systems, (2) cooperation among political jurisdictions sharing a watershed, and (3) coordinated regulation and management of all dis- charges having the potential to modify the hydrology and water quality of the watershedâs receiving waters. 1 For example, the second Draft Ventura County [California] Municipal Separate Storm Sewer System Permit states (under Findings D. Permit Coverage), âProvisions of this Or- der apply to the urbanized areas of the municipalities, areas undergoing urbanization and areas which the Regional Water Board Executive Officer determines are discharging storm water that causes or contributes to a violation of a water quality standard â¦ .â The permit further states (under Part 2âReceiving Water Limitations), â1. Discharges from the MS4 that cause or contribute to a violation of water quality standards are prohibited. â¦ 3. â¦ This Order shall be implemented to achieve compliance with receiving water limitations. If exceedence(s) of water quality objectives or water quality standards persist â¦ the Permit- tee shall assure compliance with discharge prohibitions and receiving water limitations â¦ .â
484 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES Determining Watershed Scale for Permitting A fundamental question that must be answered at the outset of any move to watershed permitting is, What is a watershed? Hydrologically, a watershed is the rain catchment area draining to a point of interest. Hence, the question comes down to, Where should the point of interest be located to define water- sheds for permitting purposes? If placed close to the initial sources of surface runoff (e.g., on each first-order stream just above its confluence with another first-order stream), attention would be very specifically directed. However, there would be little flexibility to devise solutions for the greatest good. For example, trading of the commodities runoff quantity and quality would be very restricted. If on the other hand the point of interest is placed far downstream, thus defining a very large watershed, a welter of issues, and probably also of involved jurisdictions, would overly confuse the management and regulatory task. The U.S. Geological Survey (USGS) delineates watersheds in the United States using a nationwide system based on surface hydrologic features. This system divides the country into 21 regions, 222 subregions, 352 accounting units, and 2,262 cataloging units. These hydrologic units are arranged within each other, from the smallest (cataloging units) to the largest (regions). USGS identifies each hydrologic unit by a unique hydrologic unit code (HUC) consist- ing of 2 to 16 digits based on the four levels of classification in the hydrologic unit system. Watersheds thus delineated are typically of the order a few square kilometers in area. This system is now being linked to the National Hydrogra- phy Dataset (NHD) and the National Land Cover Dataset to produce NHDPlus, an integrated suite of application-ready geospatial datasets. The USGS system provides a starting point. Ultimately, though, what con- stitutes a watershed will best be answered with reference to specific biogeo- physical conditions and problems and by personnel at relatively close hand (i.e., state or regional oversight agency staff). A general guideline might be the catchment area of a waterbody influenced by a set of similar subwatersheds. Similar subbasins would presumably be amenable to similar solutions and trad- ing off reduced efforts in some places for compensating additional efforts else- where, as well as to analysis and monitoring on a representative basis, instead of exhaustively throughout. Often, a watershed defined in this way would flow into another watershed and influence it. Thus, there would have to be coordina- tion among managers and regulators of interacting watersheds. It would be common for several watersheds ranging from relatively small to large in scale to be nested. Each would have its management team, and a committee drawn from those teams should be formed to coordinate goals and actions. A prerequisite to moving toward watershed permitting, then, is for states or regions within states to delineate watersheds. California took this step early in the NPDES stormwater permitting process and offers a model in this respect, as well as in encompassing all jurisdictions coordinated by a lead permittee. First, the state organized its California EPA regional water boards on a watershed ba-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 485 sis. Furthermore, since 1992 it has been common in California to establish one jurisdiction as the lead permittee (e.g., Los Angeles County in the Los Angeles region, Orange County in the Santa Ana Region, and San Diego County in the San Diego Region) and all of the politically separate cities as co-permittees. The lead permittee has typically been the jurisdiction most widely distributed geographically in the region and large enough to develop compliance mecha- nisms and coordinate their implementation among all participants. Box 6-3 de- scribes the approach taken to delineating management units within the Chesa- peake Bay watershed, which comprises parts of Pennsylvania, Maryland, Vir- ginia, and the District of Columbia. The case study illustrates well the approach advocated here of focusing on the outcome in the receiving water and consider- ing all aspects of land and water resources management that determine that out- come. Steps Toward Watershed-Based Permitting Once a watershed is defined, a further question arises regarding how much and what part of its territory to cover formally under permit conditions. Under the present system substantial development occurring outside Phase I or Phase II municipal jurisdictions is escaping coverage. Failing to control relatively high levels of development both outside a permitted jurisdiction and upstream of more lightly developed areas within a permitted area is particularly contrary to the watershed approach. Areas having a more urban than rural character are already essentially treated as urban in water supply and sewer planning, and the same should occur in the area of stormwater management. Accordingly, the permit should extend to any area in the watershed, even if outside Phase I or II jurisdictions, zoned or otherwise projected for development at an urban scale (e.g., more than one dwelling per acre). States do have authority under the CWA to designate any area for Phase II coverage based on projected growth or the presence of impact sources. They should be required to do so for nationwide uniformity and best protection of water resources. It is essential to clarify that watershed-based permitting as formulated in this chapter differs sharply from what has been termed watershed (or basin) planning. According to EPA, watershed planning âidentifies broad goals and objectives, describes environmental problems, outlines specific alternatives for restoration and protection, and documents where, how, and by whom these ac- tion alternatives will be evaluated, selected, and implementedâ (http://www.epa. gov/watertrain/planning/planning7.htm). Drawing up such a plan is a time- consuming process, which has often become an end in itself, instead of a means to an end. Completing a full watershed plan, as usually construed, should not be a prerequisite to watershed-based permitting. Rather, the anticipated process would spring much more from comprehensive, advanced scientific and technical analysis of the water resources to be managed and their contributing catchment areas than from a planning framework.
486 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-3 Watershed Delineation for the Chesapeake Bay The âTributary Strategy Teamâ approach of the Chesapeake Bay Watershed provides a specific example of a watershed-scale approach to implementation of water quality con- trol measures. Some background on this longstanding program is first provided, before turning to how watersheds were delineated. In 1983, the states of Virginia, Maryland, and Pennsylvania; the District of Columbia; and EPA signed an agreement to form the Chesa- peake Bay Program with a goal to restore and protect the bay, which was suffering from nutrient overenrichment, severely reduced submerged aquatic vegetation, and contamina- tion by toxics. In 1987 the program established a target of a 40 percent reduction in the amount of nutrients entering the Bay by 2000. In 1992 the bay program partners agreed to continue the 40 percent reduction goal beyond 2000 by allocating nutrient reduction targets to the bayâs tributaries. In Chesapeake 2000, the most recent version of the Chesapeake Bay agreement, the nutrient reduction goals were reaffirmed, and an additional goal of sediment reduction was established. New York, Delaware, and West Virginia, locations of the bayâs headwaters, also became involved in nutrient and sediment reduction. Cap load allocations for nutrients (nitrogen and phosphorus) and sediment to be reached by 2010 were agreed upon by the states. The states began developing 36 voluntary watershed- based tributary strategies to meet the state cap load allocations covering the entire 64,000- square-mile Chesapeake Bay watershed. Watershed-based tributary strategies are developed in cooperation with local water- shed stakeholders. For rural areas, where stakeholders include farmers, nutrient strategies include promotion of management practices such as maintaining cover crops on recently harvested cropland to reduce soil erosion, reduction in nitrogen applications, conservation tillage, and establishment of riparian buffers. For urban-area stakeholders such as home- owners and municipalities, tributary strategies include practices such as enhanced nutrient removal at WWTPs, low-impact development (LID) practices, erosion and sediment control practices, and septic system upgrades. The first cut at delineating the watershed, which was based on hydrography and to- pography, defined the eight major areas draining to the Chesapeake Bay: six major basins (Susquehanna, Potomac, York, James, Rappahannock, and Patuxent) plus smaller areas not draining to a major river on the Eastern and Western Shores of the bay in Maryland. These subdivisions are disparate with respect to size (the Susquehanna can engulf almost the entire other seven), but direct drainage to the bay was the criterion at this level. The next cut was made at state borders. For example, the Susquehanna traverses three states and was subdivided at the New YorkâPennsylvania and Pennsylvaniaâ Maryland political boundaries. Further cuts were subsequently made within some states. The criteria for these cuts varied from state to state, but generally involved a combination of smaller political jurisdictions (e.g., county, township), subwatershed basin borders, and other local considerations, such as local interest and investment (e.g., watershed associa- tions). The resulting delineations are highly variable in size but apparently satisfactory to the local parties who decided on the areas. They represent individual âtributary strategy areasâ but are also nested within the larger eight designations and involve interjurisdictional and interstate coordination where a subbasin is divided by a political boundary. Although the example of the Chesapeake Bay is at a very large scale, the principles of watershed de- lineation it illuminates apply at all scales.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 487 Effective watershed-based permitting as outlined in this report is composed of: Centralizing responsibility and authority for implementation with a municipal lead permittee working in partnership with other municipalities in the watershed as co-permittees; Adopting a minimum goal in every watershed to avoid any further loss or degradation of designated beneficial uses within the watershedâs component waterbodies; Assessing waterbodies that are not providing designated beneficial uses in order to set goals aimed at recovering these uses; Defining careful, complete, and clear specific objectives to be achieved through management and permitting; Comprehensive impact source analysis as a foundation for targeting so- lutions; Determining the most effective ways to isolate, to the extent possible, receiving waterbodies from exposure to those impact sources; Developing and appropriately allocating funding sources to enable the lead permittee and partners to implement effectively; Developing a monitoring program composed of direct measures to as- sess compliance and progress toward achieving objectives and diagnosing rea- sons for the ability or failure to meet objectives, in support of active adaptive management; and Developing a market system of trading credits as a tool available to municipal co-permittees to achieve watershed objectives, even if solutions can- not be uniformly applied. The system proposed herein is a significant departure from the road traveled in the 20 years since CWA amendments began to bring stormwater under direct regulation. This reorganization is necessary because of the failure of the present system to achieve widespread and relatively uniform compliance (see Chapter 2) and, ultimately, to protect the nationâs water resources from degradation by mu- nicipal, industrial, and construction runoff. The workload associated with adopting this approach will be considerable and will take some time to com- plete. The structure of the new program should be fully in place within five years, which is considered to be a reasonable period to complete the work. It could be fully implemented throughout the nation within ten years. However, interim measures toward its fulfillment should occur sooner, within one to two years. Such measures should be applied to each land-use and impact-source category (i.e., existing residential and commercial development, existing indus- try, new development, redevelopment, construction sites). For example, meas- ures such as an effective impervious area limit or a requirement to maintain pre- development recharge to the subsurface zone could make early progress in man-
488 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES aging new development, and lead toward the ultimate, objective-based manage- ment and permitting strategy for that category. Advanced source control per- formance standards would be appropriate interim measures for existing devel- opment. One innovative approach to watershed-based management that can ease the burden of the proposed new system is the rotating basin approach. As described by EPA (2007a), this option entails delineating state watershed boundaries and grouping the watersheds into basin management units, usually by the state water pollution control agency. Next, states implement a watershed management process on a rotating schedule, which is usually composed of five activities: (1) data collection and monitoring, (2) assessment, (3) strategy development, (4) basin plan review, and (5) implementation. Over time, different waterbodies are intensively studied as part of the rotation. Data collected can be used to support a number of different reporting and planning requirements, including a finding of attainment of water quality standards, a determination of impairment, or pos- sible delisting if the waterbody is found not to be impaired. Florida offers a good example of the rotating basin approach. The Florida Department of Envi- ronmental Protection has defined five levels of intensity, or phases, each taking about one year to complete, and it has divided the state into 30 areas based on HUCs. At any one time six areas are in each phase before rotating to a subse- quent phase. This division of effort would help alleviate the burden of moving to a new system of watershed-based permitting by programming the work over a period of years. It could certainly be organized on a priority basis, in which the watersheds of greatest interest for whatever reason (e.g., having the highest re- source values, being most subject to new impacts) would get attention first. An Objective-Based Framework The proposed framework for watershed-based management and regulation of stormwater relies on broad goals to retain and recover aquatic resource bene- ficial uses, backed by specific objectives (e.g., water quality criteria) that must be achieved if the goals are to be fulfilled. Meeting the objectives and overarch- ing goals is intended to become the basis for determining permit compliance, instead of the current reliance on implementation of SCMs as presumptive evi- dence of compliance. The broad goals of retaining and recovering beneficial uses are entirely con- sistent with the antidegradation clause of the CWA. Antidegradation means that the current level of water quality shall be maintained and protected, unless wa- ters exceed levels necessary for maintaining their beneficial uses and the state finds that allowing lower water quality is necessary to accommodate important economic or social development. In accordance with the antidegradation clause, a major pillar of the proposed concept is the goal of preventing degradation from the existing state of biological health, whatever it may be, to a lower state. Thus, fully and nearly pristine watersheds are to remain so and, at a minimum,
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 489 partially or highly impaired ones are to suffer no further impairment. Beyond this minimum, impaired waters should be assessed to determine if feasible ac- tions can be taken to recover lost designated beneficial uses or at least improve degraded uses. As discussed in Chapter 2, beneficial uses relate to the social and ecological services offered, or intended to be offered, by waterbodies. For example, Cali- fornia has 20 categories of beneficial uses embracing water supply for various domestic, agricultural, and industrial purposes; provision of public recreation; and support of aquatic life and terrestrial wildlife (CalEPA, Central Coast Re- gional Water Board Basin Plan). That beneficial uses are usually assigned at the state level by waterbody classes or specific waterbodies would not change under the proposed permitting program revision. Most waters have several beneficial uses encompassing some water supply and ecological functions and, perhaps, some form of recreation. Unlike most current stormwater programs where at- tainment of beneficial uses is only implicit, these goals would become explicit in the altered system and officially promulgated by the authority operating the permit program (a designated state, in most cases, or EPA). The permitting au- thority would then partner with municipal permittees to determine the conditions that must be brought to bear to attain beneficial uses, set objectives or criteria to establish those conditions, and follow through with the tasks to accomplish ob- jectives. The proposed frameworkâs reliance on achieving objectives that reflect the cumulative aquatic resource effects of contributing watershed conditions sug- gests the following related concepts: In whatever manner watershed boundaries are set, the full extent of the watershed from headwaters onward should be considered in defining objectives. This is important even where watershed scale and boundaries are based on local and/or regional hydrogeomorphic circumstances and their associated manage- ment and regulatory needs. Watersheds can and often will be defined and nested at different scales (e.g., streams tributary to a lake, a river flowing into an estu- ary or marine bay). The scale of objectives must be consistent with the scale and recog- nized beneficial uses of the watershed(s) in question; for example, sustaining salmonid fish spawning could be the basis for a stream objective, while retaining an oligotrophic state could be the essential objective for a lake to which the stream is tributary. Whenever beneficial uses pertain to living organisms (aquatic life or humans), representing the vast majority of all cases, objectives should be largely in biological terms. That is not to say that supplementary objectives cannot be stated otherwise (e.g., in terms of flow characteristics, chemical water quality constituents, or habitat attributes), but the ultimate direct thrust of the program
490 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES should be toward the biota. Objectives must be carefully chosen to represent attributes of impor- tance from a resource standpoint, limited in number for feasibility of tracking achievement, and defined in a way that achievement can be measured. For ex- ample, nitrogen is generally the nutrient limiting algal growth in saline systems and in excess it stimulates growth that can reduce dissolve oxygen, killing fish and other aerobic organisms. In this case the most productive objectives would probably target reduction of nitrogen concentration and mass flux and mainte- nance of dissolved oxygen. For waterbodies designated for contact recreation, fecal coliform indicators (although not directly pathogenic when waterborne) have proven to be an effective means of assessing condition and should continue to form the basis for objectives to protect contact recreation until research pro- duces superior measures. If drinking water supply is a designated beneficial use of a lake, it will better serve that function in a lower than a higher state of eutro- phication, which can be managed, according to a long limnological research record, by restricting water column chlorophyll a as an objective. Where the beneficial use is fish protection and propagation, biological criteria might in- clude (1) maintenance of a specific population size of a resident fish species when that speciesâ population can be assayed conveniently; (2) maintenance of a numerical index (e.g., benthic index of biotic integrity) when a fish species of ultimate interest cannot be assessed so conveniently but is known or reasonably hypothesized to be associated with the index; or (3) a related parameter, such as eelgrass beds, which are important fish nursery areas in estuarine waters, such that areal coverage by these beds would be an appropriate objective to track over time. An intermittent waterbody could have biological criteria related to, for example, fish migration or amphibian reproduction. The achievement of objectives, or lack thereof, is the basis for follow- up and prescription of remedies in an active adaptive management mode; that is, falling short of objectives would trigger a search for reasons throughout the wa- tershed, followed by identification of actions necessary and sufficient to remedy the shortfall, assessment of their ability to reach objectives, and the cost of doing so. In the course of this assessment it may be concluded that the objective itself is faulty and should be restated, replaced, or discarded. Basing the watershed framework principally on biological objectives grows out of the CWAâs fundamental charge to protect the biological (as well as physical and chemical) integrity of the nationâs waters. The tie between specific physical and chemical conditions and the sustenance of aquatic biological com- munities is not well established through an extensive, well-verified body of re- search. Moreover, living organisms consuming or living in water are subject to a vast multitude of simultaneous physical and chemical agents having the poten- tial to harm them individually and interactively. There are no realistic prospects
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 491 for research to determine the levels of these numerous agents that must be main- tained to support beneficial uses. Therefore, their integrative effects must be determined using measures of biological populations or communities of interest. By and large, state water quality standards as now promulgated would not serve the proposed objective-based system well. They are usually not phrased in biological terms or with respect to hydrologic variables now known to have in- strumental negative effects on aquatic organisms, but instead mostly as concen- trations of selected chemical elements or compounds. However, there is no pro- hibition of biological or hydrologic standards in the law. The recommended emphasis is consistent with and informed by the tiered aquatic life uses system applied by some states and illustrated for Ohio in Box 2-1. The use of such sys- tems must expand greatly to support the recommended framework. An opportu- nity to do so exists through the triennial review already required for each stateâs water quality standards. Certain special considerations affect the development and use of objectives as the device to carry forward watershed-based stormwater management and regulation. First, other elements of the CWA beyond the stormwater program and other laws may very well be involved in a watershed (see Chapter 2). Mu- nicipal and industrial wastewater discharges will often be contributors along with stormwater. Aquatic organisms may be listed as threatened or endangered under the federal ESA or state authority. Both objectives and the management and regulatory program designed to achieve objectives should reflect any such circumstances. Instituting the proposed permitting program will require converting the TMDL program to one more suitable for its purposes and structure. The TMDL program is watershed based and hence offers some precedent and experience applicable to the new system. However, for the most part, it has operated only on waters declared to be impaired for specific pollutants, and it relies on man- agement of specific physical and chemical water quality variables. Furthermore, in its current mode it takes no account of potential future impact sources. The TMDL program should be replaced with one adapted to the objective-based framework proposed here. This new program should apply to all waters as- signed objectives, âimpairedâ or not, and formulate limits in whatever terms are best to achieve objectives. Hence, although the program would expand in cov- erage area, the efficient tailoring of objectives directly to beneficial uses could compensate for the expansion by targeting fewer variables. Finally, the new program should look to the future as well as the present by encompassing the anticipated impacts of prospective landscape changes. The nature of a program to replace TMDLs can be glimpsed from a few at- tempts to move in the anticipated direction even under the existing structure. For example, Connecticut collected data directly linking impervious cover to poor stream health in Eagleville Brook (Connecticut Department of Environ- mental Protection, 2007). The streamâs TMDL was developed using watershed impervious cover as a surrogate parameter for a mix of pollutants conveyed by stormwater. The intention is to reduce effective imperviousness by disconnect-
492 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES ing impervious areas, installing unspecified SCMs, minimizing additional dis- turbance, and enhancing in-stream and riparian habitat. Flow was used as a sur- rogate for stormwater pollution in the Potash Brook, Vermont TMDL (Vermont DEC, 2006). In this waterbody, the impairment was based on biological indices that were then related to a hydrologic condition believed to be necessary to achieve the Vermont criteria for aquatic life. The TMDL will be implemented via the use of runoff-volume-reduction SCMs throughout the watershed. Impact Sources The CWA provides for regulating, as specific land-use types, only desig- nated industrial categories, with construction sites disturbing one acre or more considered to be one of those categories. Otherwise, it gives authority to regu- late municipal jurisdictions operating separate storm sewer systems. Generally speaking, these jurisdictions encompass, in addition to the industrial categories, the full range of urban land-use types, such as single- and multiple-family resi- dential, various kinds and scales of commercial activity, institutional, and parks and other open space. All of these land uses and the activities conducted on them are, to one degree or another, sources of the agents that physically and chemically modify aquatic systems to the detriment of their biological health. Hence, most of the impact sources to which these aquatic systems are subject are not directly regulated under CWA authority as are industrial sources, but instead are indirectly regulated through the municipal program. Also, as already dis- cussed, the situation is further complicated by the presence of municipal and industrial wastewater sources along with landscape sources contributing flow and pollutants to receiving waters via stormwater discharges. The watershed-based framework envisioned here relies on municipalities led by a principal permittee. Thus, a fundamental task that municipal permittees charged with operating under a watershed-based permit must do is to find indus- tries and construction sites in the watershed that have not filed for permit cover- age and bring them under regulation. Furthermore, municipal co-permittees, with leadership by a watershed lead permittee, must classify industries and con- struction sites within their borders according to risk and accordingly prioritize them for inspection and monitoring (methods for doing this are discussed later in the chapter). Municipal permittees must have better tools than they have had in the past to assess the various impact sources and formulate strategies to manage them that have a reasonably high probability of fulfilling objectives. The pre- sent state of practice and research findings offers some directions for choosing or more completely developing these tools. However, by no means are all the necessary elements available, and substantial new basic and applied research must be performed. From the literature come several possibilities to improve source analysis in the complex urban environment. Some examples of apparent promise, drawn from Clark et al. (2006) include the following:
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 493 Nirel and Revaclier (1999) used the ratio of dissolved rubidium (Rb) to strontium (Sr) to identify and quantify the impact of sewage effluents on river quality in Switzerland. Rubidium was present in larger quantities than strontium in feces and urine, making the ratio of these two elements an effective tracer that does not vary with river flow for a given water quality condition. Using the ratio alone produced the same conclusions regarding impact as measuring a host of physicochemical water quality variables. The researchers estimated that the Rb:Sr ratio must be lower than 0.007 if biological diversity is to be maintained, which could be the basis of an objective to manage river water quality. Al- though this case pertains to municipal wastewater and the technique works best in waters with a naturally low Rb:Sr ratio (e.g., calcareous regions), it success points out a potential avenue of research to simplify stormwater management on the basis of quantitative objectives related to biological integrity. Cosgrove (2002) described the approach used in New Jersey to charac- terize the relative contribution of point and nonpoint sources of pollutants in the Raritan River Basin. Twenty-one surface water sampling locations within the watershed were monitored four to five times per year from 1991 to 1997. These data were evaluated by comparing the median concentration at each sampling location with land-use statistics. Cumulative probability curves were also de- veloped for each pollutant to demonstrate the probability that the concentration at a given location would be below a certain level (e.g., a stream standard). These probability curves were useful in determining the risk that a given loca- tion would violate a particular standard. The concentration data, coupled with continuous flow monitoring records, were utilized to determine the total load for each constituent. Regression analysis was used to develop a relationship be- tween the total in-stream loads and flow. Such an analysis provided an indica- tion of municipal or industrial discharge versus diffuse-source-dominated loca- tions. Pollutant loads could then be converted to yield (load per unit area) to normalize the results for comparison from one station to another. The âscreen- ing levelâ methodology uses only existing data and, not requiring advanced modeling techniques, can be used to understand where to focus more rigorous modeling techniques. Maimone (2002) presented the overall approach that was used to screen and evaluate potential pollutant sources within the Schuylkill River watershed as part of the Schuylkill River Source Water Assessment Partnership. The partner- ship performed source water assessments of 42 public water supply intakes for the Pennsylvania Department of Environmental Protection. The watershed en- compasses over 1,900 square miles with more than 3,000 potential point sources of contamination. In addition, runoff from diverse land uses such as urban and agriculture had to be characterized using the Stormwater Management Model. For all 42 surface water intakes, potential point sources were identified using existing databases. The list was first passed through a series of Geographic In-
494 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES formation System-based âscreeningâ sieves to limit the sources to only those considered to be high priority (including proximity and travel time from source to intake). Ten categories were identified that cover the range of the most im- portant contaminants that might be found within the watershed, and a represen- tative or surrogate chemical was identified whose properties were used to stand in for the category. Beyond the geographic screening, a more sophisticated screening was needed to limit the number of sites, using a decision support computer software program called EVAMIX. The greatest benefit of EVAMIX, compared to other software, is that it allows mixed criteria evaluation, qualita- tive and quantitative, to be considered concurrently. EVAMIX produced source rankings representing an organized and consistent use of both the objective data and the subjective priorities of decision makers. Hetling et al. (2003) investigated the effect of water quality manage- ment efforts on wastewater discharges to the Hudson River (from Troy, New York to the New York City Harbor) from 1900 to 2000. The paper demon- strated a methodology for estimating historic loadings where data are not avail- able. Under these circumstances, estimated historic sewered and treated popula- tions and per capita values were used to calculate wastewater flow and loadings for 5-day biochemical oxygen demand (BOD5), total suspended solids (TSS), total nitrogen, and total phosphorus. The analysis showed that dispersed land- scape sources have become the most significant contributors of the first two contaminants to the river, while municipal wastewater plants remain the largest sources of nutrients. The methodology presented in this paper could be used by co-permittees to estimate present-day sources of various types and contribute to moving toward a comprehensive permit incorporating multiple sources. Zeng and Rasmussen (2005) used multivariate statistics to characterize water quality in a lake and its tributaries. Tributary water was composed of three components. Factor analysis demonstrated that stormwater runoff was the predominant cause of elevation of a group of water quality variables in a factor including TSS, the measurement of which is a convenient surrogate for all vari- ables in the factor. Similarly, municipal and industrial discharges could be char- acterized by total dissolved solids, and groundwater by alkalinity plus soluble reactive phosphorus. These sources can thus be distinguished through meas- urement of just four common water quality variables. Reducing the number of analytes reduces laboratory costs and allows resources to be freed up for other purposes. Cluster analyses performed on the data indicated that further savings could be realized by sampling just one among several stations in a cluster and sampling at just one point in time over a period of relatively stable water quality (e.g., a relatively dry period). A key research need associated with applying the proposed framework is assessment of these and other mechanisms for sorting out the contributions of
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 495 the variety of impact sources in the urban environment. Leading this effort would be a natural role for EPA. Impact Reduction Strategies The philosophical basis for impact reduction under a modified permitting system centered on a lead municipal permittee and associated co-permittees is to avoid, as far as possible, exposing receiving waters to impact sources or to oth- erwise minimize that exposure. The concept embraces both water quantity and quality impact sources and specifically raises the former category to the same level of scrutiny as traditionally applied to water quality sources. Furthermore, the endpoints upon which success and compliance would be judged are directly related to achievement of beneficial uses. This approach to impact reduction, where the direct focus is on reducing the loss of aquatic ecosystem functioning supportive of beneficial uses, fundamentally contrasts with the currently prevail- ing system. What are primary concerns in the existing system (e.g., discharge concentrations of certain chemical and physical substances, technological strate- gies from a menu of practices) are still prospectively important, but only as a means toward realizing functional objectives, not as endpoints themselves. To be sure, attaining beneficial uses will require wise choices among tools to de- crease discharges and contaminant emissions. However, the ultimate proof will always be in biological outcomes. As made clear in Chapters 3 and 4, linkages among myriad stressing agents, impact receptors, and specific mitigating abilities of technological fixes are poorly understood and not easily understandable. The proposed new paradigm acknowledges that the linkages are not established among the voluminous ele- ments in an exceptionally complex system ranging from impact sources, through environmental transport and fate mechanisms, to ecosystem health. However, it is intuitively and theoretically clear that minimizing the generation of impacts in the first place and slowing their progression into aquatic environments can break the chain of landscape alteration that leads to increased runoff and pollutant pro- duction, modifies aquatic habitat, and ultimately causes deterioration of the bio- logical community. Landscapes can be managed in a preventive, integrated fashion that deals with the many undifferentiated agents of impact and avoids, or at least reduces, the damage. Although the application of these theories may not automatically and quickly stem biological losses, the powerful mechanism of adaptive management, if correctly applied, can be used to make course correc- tions toward meeting the defined objectives. An earlier National Research Council (NRC) committee examined the sci- entific basis of EPAâs TMDL program and recommended âadaptive implemen- tationâ (AI) to water quality standards (NRC, 2001a). That committee drew AI directly from the concept of adaptive management for decision making under uncertainty, introduced by Holling and Chambers (1973) and Holling (1978) and described it as an iterative process in which TMDL objectives and the imple-
496 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES mentation plans to meet those objectives are regularly reassessed during the on- going implementation of controls. Shabman et al. (2007) and Freedman et al. (2008) subsequently extended and refined the applicability of AI for promoting water quality improvement both within and outside of the TMDL program. In that broader context, AI fits well with the framework put forward here. Indeed, the proposed revised monitoring system presented later in this chapter is de- signed to provide information to support adaptive management in a permitting context. The Stages of Urbanization and Their Effects on Strategy In waterbodies that are not in attainment of designated uses, it is likely that the physical stresses and pollutants responsible for the loss of beneficial uses will have to be decreased, especially as human occupancy of watersheds in- creases. Reducing stresses, in turn, entails mitigative management actions at every life stage of urban development: (1) during construction when disturbing soils and introducing other contaminants associated with building; (2) after new developments on Greenfields are established and through all the years of their existence; (3) when any already developed property is redeveloped; and (4) through retrofitting static existing development. Most management heretofore has concentrated on the first two of those life stages. The proposed approach recognizes three broad stages of urban development requiring different strategies: new development, redevelopment, and existing development. New development means building on land either never before covered with human structures or in prior agricultural or silvicultural use rela- tively lightly developed with structures and pavements (i.e., Greenfields devel- opment). Redevelopment refers to fully or partially rebuilding on a site already in urban land use; there are significant opportunities for bringing protective measures to these areas where none previously existed. The term existing de- velopment means built urban land not changing through redevelopment; retrofit- ting these areas will require that permittees operate creatively. What is meant by redevelopment requires some elaboration. Regulations already in force typically provide some threshold above which stormwater man- agement requirements are specified for the redeveloped site. For example, the third Draft Ventura County Municipal Separate Storm Sewer System Permit defines âsignificant redevelopmentâ as land-disturbing activity that results in the creation or addition or replacement of 5,000 square feet or more of impervious surface area on an already developed site. The permit goes on to state that where redevelopment results in an alteration to more than 50 percent of the im- pervious surfaces of a previously existing development, and the existing devel- opment was not subject to postdevelopment stormwater quality control require- ments, the entire site becomes subject to application of the same controls re- quired for new development. Where the alteration affects 50 percent or less of the impervious surfaces, only the modified portion is subject to these controls.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 497 All urban areas are redeveloped at some rate, generally slowly (e.g., roughly one or at most a few percent per annum) but still providing an opportunity to amelio- rate aquatic resource problems over time. Extending stormwater requirements to redeveloping property also gradually âlevels the playing fieldâ with new de- velopments subject to the requirements. As pointed out in Chapter 2, some ju- risdictions offer exemptions from stormwater management requirements to stimulate desired economic activities or realize social benefits. Such exemp- tions should be considered very carefully with respect to firm criteria designed to weigh the relative socioeconomic and environmental benefits, to prevent abuses, to gauge just how instrumental the exemption is to gaining the socioeco- nomic benefits, and to compensate through a trading mechanism as necessary to achieve set aquatic resource objectives. It is important to mention that not only residential and commercial proper- ties are redeveloped, but also streets and highways are periodically rebuilt. Highways have been documented to have stormwater runoff higher than other urban land uses in the concentrations and mass loadings of solids, metals, and some forms of nutrients (Burton and Pitt, 2002; Pitt et al., 2004; Shaver et al., 2007). Redevelopment of transportation corridors must be taken as an opportu- nity to install SCMs effective in reducing these pollutants. Opportunities to apply SCMs are obviously greatest at the new development stage, somewhat less but still present in redevelopment, but most limited when land use is not changing (i.e., existing development). Still, it is extremely im- portant to utilize all readily available opportunities and develop others in static urban areas, because compromised beneficial uses are a function of the devel- opment in place, not what has yet to occur. Often, possibly even most of the time, to meet watershed objectives it will be necessary to retrofit a substantial amount of the existing development with SCMs. To further progress in this overlooked but crucial area, the Center for Watershed Protection issued a practi- cal Urban Stormwater Retrofit Practices manual (Schueler et al., 2007). Practices for Impact Reduction As described in Chapter 5, in the past 15 to 20 years stormwater manage- ment has passed through several stages. First, it was thought that the key to suc- cess was to match postdevelopment with predevelopment peak flow rates, while also reducing a few common pollutants (usually TSS) by a set percentage. Find- ing this to require large ponds but still not forestalling impacts, stormwater man- agers next deduced that runoff volumes and high discharge durations would also have to decrease. Almost simultaneously, although not necessarily in concert, the idea of LID arose to offer a way to achieve actual avoidance or at least minimization of discharge quantity and pollutant increases reaching far above predevelopment levels. For purposes of this discussion, the SCMs associated with LID along with others are named Aquatic Resources Conservation Design (ARCD). First, this term signifies that the principles and many of the methods
498 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES apply not only to building on previously undeveloped sites, but also to redevel- oping and retrofitting existing development. Second, incorporating aquatic re- sources conservation in the title is a direct reminder of the central reason for improving stormwater regulation and management. ARCD goes beyond LID to encompass many of the SCMs discussed in Chapter 5, in particular those that decrease surface runoff peak flow rates, volumes, and elevated flow durations caused by urbanization, and those that avoid or at least minimize the introduc- tion of pollutants to any surface runoff produced. This concentration reduction, together with runoff volume decrease, cuts the cumulative mass loadings (mass per unit time) of pollutants entering receiving waters over time. The SCM cate- gories from Table 5-1 that qualify as ARCD include: Product Substitution, Watershed and Land-Use Planning, Conservation of Natural Areas, Impervious Cover Minimization, Earthwork Minimization, Reforestation and Soil Conservation, Runoff Volume ReductionâRainwater Harvesting, Vegetated, and Subsurface, Aquatic Buffers and Managed Floodplains, and Illicit Discharge Detection and Elimination. The menu of ARCD practices begins with conserving, as much as possible, existing trees, other vegetation, and soils, as well as natural drainage features (e.g., depressions, dispersed sheet flows, swales). Clustering development to affect less land is a fundamental practice advancing this goal. Conserving natu- ral features would further entail performing construction in such a way that vegetation and soils are not needlessly disturbed and soils are not compacted by heavy equipment. Using less of polluting materials, isolating contaminating materials and activities from contacting rainfall or runoff, and reducing the in- troduction of irrigation and other non-stormwater flows into storm drain systems are essential. Many ARCD practices fall into the category of minimizing im- pervious areas through decreasing building footprints and restricting the widths of streets and other pavements to the minimums necessary. Water can be har- vested from impervious surfaces, especially roofs, and put to use for irrigation and gray water system supply. Harvesting is feasible at the small scale using rain barrels and at larger scales using larger collection cisterns and piping sys- tems. Relatively low traffic areas can be constructed with permeable surfaces such as porous asphalt, open-graded Portland cement concrete, coarse granular materials, concrete or plastic unit pavers, or plastic grid systems. Another im- portant category of ARCD practices involves draining runoff from roofs and pavements onto pervious areas, where all or much can infiltrate or evaporate in many situations.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 499 If these practices are used, but excess runoff still discharges from a site, ARCD offers an array of techniques to reduce the quantity through infiltration and evapotranspiration and improve the quality of any remaining runoff. These practices include (1) bioretention cells, which provide short-term ponded and soil storage until all or much of the water goes into the deeper soil or the atmos- phere; (2) swales, in which water flows at some depth and velocity; (3) filter strips, broad surfaces receiving sheet flows; (4) infiltration trenches, where tem- porary storage is in below-ground gravel or rock media; and (5) vegetated (âgreenâ) roofs, which offer energy as well stormwater management benefits. Natural soils sometimes do not provide sufficient short-term storage and hydrau- lic conductivity for effective surface runoff reduction because of their composi- tion but, unless they are very coarse sands or fine clays, can usually be amended with organic compost to serve well. ARCD practices should be selected and applied as close to sources as pos- sible to stem runoff and pollutant production near the point of potential genera- tion. However, these practices must also work well together and, in many cases, must be supplemented with strategies operating farther downstream. For exam- ple, the City of Seattle, in its ânatural drainage systemâ retrofit initiative, built serial bioretention cells flanking relatively flat streets that subsequently drain to âcascadesâ of vegetated stepped pools created by weirs, along more sloping streets. The upstream components are highly effective in attenuating most or even all runoff. Flowing at higher velocities, the cascades do not perform at such a high level, although under favorable conditions they can still infiltrate or evapotranspire the majority of the incoming runoff (Horner et al., 2001, 2002, 2004; Chapman, 2006; Horner and Chapman, 2007). Their role is to reduce runoff from sources not served by bioretention systems as well as capture pol- lutants through mechanisms mediated by the vegetation and soils. The success of Seattleâs natural drainage systems demonstrates that well-designed SCMs can mimic natural landscapes hydrologically, and thereby avoid raising discharge quantities above predevelopment levels. In some situations ARCD practices will not be feasible, at least not entirely, and the SCMs conventionally used now and in the recent past (e.g., reten- tion/detention basins, biofiltration without soil enhancement, and sand filters) should be integrated into the overall system to realize the highest management potential. The proposed watershed-based program emphasizing ARCD practices would convey significant benefits beyond greatly improved stormwater man- agement. ARCD techniques overall would advance water conservation, and infiltrative practices would increase recharge of the groundwater resource. ARCD practices can be made attractive and thereby improve neighborhood aes- thetics and property values. Retention of more natural vegetation would both save wildlife habitat and provide recreational opportunities. Municipalities could use the program in their general urban improvement initiatives, giving incentives to property owners to contribute to goals in that area while also com- plying with their stormwater permit.
500 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES Municipal Permittee Roles in Implementing Strategies Municipal permittees sharing a watershed will have key roles in promoting ARCD under the proposed new system. First, the lead permittee and its partners would be called upon to perform detailed scientifically and technically based watershed analysis as the programâs foundation. The City of San Diego (2007) offers a model by which permittees could operate with its Strategic Plan for Wa- tershed Activity Implementation. The plan consists of: Activity location prioritizationâlocations prioritized for action based on pollutant loading potential; Implementation strategy and activity prioritizationâtiered approach identifying activities directed at meeting watershed goals over a five-year pe- riod; Potential watershed activitiesâgeneral list of activities required and potentially required to meet goals as guidance for planning and budgeting; Watershed activity mapsâspecified locations for activities; and Framework for assessment monitoringâa plan for development of the monitoring and reporting program. Municipal permittees would be required under general state regulations to make ARCD techniques top priorities for implementation in approving new de- velopments and redevelopments, to be used unless they are formally and con- vincingly demonstrated to be infeasible. In that situation permit approval would still require full water quantity and quality management using conventional practices. Beyond regulation, municipalities would be called upon to give pri- vate property owners attractive incentives to select ARCD methods and support to implement them. Furthermore, they should supplement on-site ARCD instal- lations with municipally created, more centralized facilities in subwatersheds. Other municipal roles in the proposed program revolve around the promi- nence of soil infiltration as a mechanism in ARCD. Successful use of infiltra- tion requires achieving soil hydraulic conductivity sufficient to drain the runoff collector quickly enough to provide capacity for subsequent storms and avoid nuisance conditions, while not so rapid that contaminants would reach ground- water. One important task for municipal co-permittees will be defining water- shed soils and hydrogeological conditions to permit proper siting and design of infiltrative facilities. A great deal of soils information already exists in any community but must be assembled and interpreted to assist stormwater manag- ers. U.S. Department of Agriculture soil surveys, while a start, are often insuffi- ciently site-specific to characterize the subsurface accurately at a point on the landscape. More localized data available to municipalities come from years of recorded well logs, soil borings, and percolation test results. Municipalities should tap these records to define, to their best ability, soil types, hydraulic con- ductivities, and seasonal groundwater positions. Although abundant and valu-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 501 able, these data are unlikely to be sufficient to define subsurface attributes across a watershed. Thus, municipalities should collect additional data (soil borings, soils analyses, and percolation tests) to obtain a good level of assurance of the prospects for infiltrative ARCD. Part of the task for municipalities will be overcoming opposition to infiltra- tion if it is unjustified. Some opponents discourage infiltration based on coarse soil survey data that may not apply at all at a locality, or they fail to take into account that the well-established ARCD practice of soil amendment, generally with organic compost, can improve the characteristics of somewhat marginal soils sufficiently to function well during infiltration. While such amendment cannot increase hydraulic conductivity sufficiently in restrictive clay soils, the technique has proven to effectuate substantial infiltration and attendant reduc- tion in runoff volumes and peak flow rates in Seattleâs natural drainage systems, discussed above. These systems lie on variable soils, including formations cate- gorized by the Natural Resources Conservation Service (2007) as being in hy- drologic group C. This group generally has somewhat restricted saturated hy- draulic conductivity in the least transmissive layer between the surface and 50 centimeters (20 inches) of between 1.0 micrometers per second (0.14 inches per hour) and 10.0 micrometers per second (1.42 inches per hour). Furthermore, additional runoff reduction often occurs through evapotranspiration, which is enhanced by the vegetation in ARCD systems. Another objection sometimes raised to infiltrating stormwater is its per- ceived potential to compromise groundwater quality. Whether or not that poten- tial is very great depends upon a number of variables: rate of infiltration, ability of the soil type to extract and retain contaminants, distance of travel to ground- water, and any contaminated layers through which the water passes. It is unlikely that urban stormwater, with its prevailing pollutant concentrations, will threaten groundwater if it travels at a moderate rate, through soils of medium or fine textures without contaminant deposits, to groundwater at least several me- ters below the surface. To ensure that groundwater is not compromised when surface water is routed through infiltrative practices, municipalities must estab- lish where appropriate conditions do and do not exist and spot infiltration oppor- tunities accordingly. Records of past waste disposal, leaks, and spills must be consulted to clean up or stay away from contaminated zones. There are alterna- tives even if documented soils or groundwater limitations rule out infiltrative practices. Much can be accomplished to reduce the quantities of contaminated urban runoff discharged to receiving waters through impervious surface reduc- tion, water harvesting, and green roofs. One additional problem to infiltrating stormwater runoff exists in some rela- tively dry areas and must be countered by municipalities. Overirrigation of lawns and landscape plantings has already increased infiltration well over the predevelopment amount and raised groundwater tables, sometimes to problem- atic levels. This unnecessary use of irrigation not only wastes potable water, often scarce in such areas, but reduces capacity to infiltrate stormwater without further water table rise. Municipalities should set up effective programs to con-
502 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES serve water and simultaneously increase stormwater infiltration capacity. A final element of an integrated management and permitting program under municipal control is use of capacity in the sanitary sewer and municipal waste- water treatment systems to treat some stormwater. This initiative must be pur- sued very carefully. One reason for care is that municipal treatment works have historically been overburdened with stormwater flows in combined sewers and have not yet broken free of that burden through sewer separation programs. A second reason is that municipal sewage treatment plants are generally designed to remove particulates and decompose organic wastes and not to capture the array of pollutants in stormwater, many dissolved or associated with the finest and most difficult to capture particles. Toxic contaminants can damage mi- crobes and upset biological treatment plants. Nonetheless, capacity exists in many WWTPs to treat stormwater. The delivery of pollutants the plant was not designed to handle can be managed by pretreatment requirements, applied to industrial stormwater dischargers particularly. Dry weather flows, consisting mostly of excess irrigation runoff, can be diverted to treatment plants to prevent at least some of the nutrient and pesticide contamination that otherwise would flow to receiving waters. Additional capacity to treat stormwater can be gained by repairing defective municipal wastewater pipes that allow groundwater entry. Special Considerations for Construction and Industrial Land Uses All of the principles discussed above apply to industrial and construction sites as well: minimize the quantity of surface runoff and pollutants generated in the first place, or act to minimize what is exported off the site. Unfortunately, construction site stormwater now is managed all too often using sediment barri- ers (e.g., silt fences and gravel bags) and sedimentation ponds, none of which are very effective in preventing sediment transport. Much better procedures would involve improved construction site planning and management, backed up by effective erosion controls, preventing soil loss in the first place, which might be thought of as ARCD for the construction phase of development. Just as ARCD for the finished site would seek to avoid discharge volume and pollutant mass loading increase above predevelopment levels, the goal of improved con- struction would be to avoid or severely limit the release of eroded sediments and other pollutants from the construction site. Chapter 5 discusses construction- phase stormwater management in more detail. Other industrial sites are faced with some additional challenges. First, in- dustrial sites usually have less landscaping potentially available for land-based treatments. Their discharges are often more contaminated and carry greater risk to groundwater. On the other hand, industrial operations are amenable to a vari- ety of source control options that can completely break the contact between pol- lutants and rainfall and runoff. Moving operations indoors or roofing outdoor material handling and processing areas can transform a high-risk situation to a
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 503 no-risk one. It is recommended that industrial permits strongly emphasize source control (e.g., pollution prevention) as the first priority and the remaining ARCD measures as secondary options (as outlined in Table 5-9). Together these measures would attempt to avoid, or minimize to the extent possible, any discharge of stormwater that has contacted industrial sources. It is likely that the remaining discharges that emanate from an industrial site will often require treatment and, if relatively highly contaminated, very efficient treatment to meet watershed objectives. Some industrial stormwater runoff car- ries pollutant concentrations that are orders of magnitude higher than now pre- vailing water quality standards. In these cases meeting watershed objectives may require providing active treatment, which refers to applying specifically engineered physicochemical mechanisms to reduce pollutant concentrations to reliably low levels (as opposed to the passive forms of treatment usually given stormwater, such as ponds, biofiltration, and sand filters). Examples now in the early stages of application to stormwater include chemical coagulation and pre- cipitation, ion exchange, electrocoagulation, and filtration enhanced in various ways. These practices are undeniably more expensive than source controls and other ARCD options and traditional passive treatments. If they must be used at all, it is to the advantage of all parties that costs be lowered by decreasing con- taminated waste stream throughput rates to the absolute minimum. Administrative and Funding Arrangements A number of practical, logistical considerations pertain to converting to the permitting and regulatory system discussed above. These considerations in- clude: What design and performance standards should be placed on the man- agement systems? What administrative vehicles offer the best prospects for success? What funding arrangements are necessary to support the revised per- mitting and management system? Design and Performance Standards It has already been asserted under the discussion of objectives above that ul- timate performance standards should be based on results in the aquatic systems under protection. The report further advocates promulgating these standards primarily in terms of biological health (for protection of human health, aquatic life, or both), supplemented by measures of conditions well known to influence biological health quite directly, such as hydrologic variables. It was further pro- posed that active adaptive management be applied in relation to the degree of
504 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES achievement of water resource objectives. However, it would not be wise to standardize entirely on this level and leave all questions of the means to the end to individual permittees. Certain design-level standards would also be appropri- ate. An example is provided by the recently issued draft municipal permit for Ventura County, California. In that permit, application of low-impact methods to new development and redevelopment is specified to hold the effective imper- vious area to 5 percent of the total contributing catchment. While technical ex- perts may disagree on the precise number, the point is that adopting such a stan- dard gives a straightforward design requirement on an evidentiary basis. Results in the receiving waters would still be tracked and used in active adaptive man- agement if necessary, but effective application of the design standard would provide some level of initial assurance that the aquatic health standards can be met. Forging Institutional Partnerships At the heart of the proposal for a new system of regulating discharges to the nationâs waters is issuing permits to groups of municipalities in a watershed operating as co-permittees under a lead permittee. Furthermore, the proposal envisions these municipal permittees assuming responsibility for and imple- menting the permits for all public and private dischargers in their jurisdictions. These admittedly sweeping changes in the way waters have been managed al- most everywhere in the nation raise serious issues of acquiescence to the new arrangements, compatibility, and devising a sufficient and stable funding base. This section draws from the small number of examples where arrangements like those proposed here have been attempted. The Los Angeles County Municipal Storm Water Permit offers a case study in how to aggregate municipalities in a co-permittee system while still allowing prospective members latitude should they perceive their own interests to deviate, even considering the advantages of group action. The permit, first issued in 1990, presently covers five watersheds and 86 municipal permittees. During the process of reissuing the 1996 permit, the City of Long Beach challenged the provisions of the Los Angeles County MS4 permit. The city was given the op- tion of applying for its own individual permit, which it did. Long Beach was issued its own individual MS4 permit in 1999 with provisions similar to the Los Angeles County MS4 permit. As another example, a small coastal municipality (Hermosa Beach) covered by the Los Angeles County Municipal Storm Water Permit investigated the possibility of withdrawing from the county permit in 2000 to be reclassified as a Phase II municipality. Just as with Long Beach, Hermosa Beach was given the option of applying for an individual permit as a Phase I MS4, but in the end Hermosa Beach elected to remain within the are- awide permit. Although this report strongly encourages cooperative participa- tion of municipalities as co-permittees, it does not mandate it. Rather, the flexi- bility illustrated above should be retained in the proposed new permitting pro-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 505 gram. What matters for compliance with the CWA is that a municipality man- age discharges in a manner at least equivalent to other permittees in the water- shed. Stephenson and Shabman (2005) gave thought to the dilemma of entities who may not naturally work well together being asked to cooperatively solve a problem that all have had a share in creating. They argued that new organiza- tional forms that consolidate multiple regulated entities under a single organiza- tional umbrella could be used to coordinate and manage jointly the collective obligations of a group of regulated parties at lower costs to members. Private and public regulated entities alike could benefit from participation in these new organizations. Such cooperative organizations could offer participating parties financial incentives and decision-making flexibility through credit trading pro- grams. Two larger-scale compliance associations exist in the Neuse and Tar- Pamlico river basins in North Carolina (Stephenson and Shabman, 2005). In both programs the state was concerned about nutrient enrichment of estuary wa- ters and imposed an aggregate cap on industrial and municipal wastewater dis- chargers equivalent to a 30 percent reduction in nitrogen loads. In both pro- grams, the state granted individual point source dischargers a choice: (1) accept new requirements to control nitrogen through individual NPDES permits or (2) form and join a discharger association. The rigidities associated with individual NPDES permits provided enough incentive for most point source dischargers to opt for the second choice. Compliance associations were then created and is- sued permits. The Neuse River rules cover nonpoint agricultural sources as well as point discharges. Counties are responsible for reducing nutrient loads, and farmers must either join county associations that apply different strategies or individu- ally contribute to meeting objectives by setting aside 50- to 100-foot buffers along all streams. North Carolina requires compliance associations to meet a single mass load cap. In the Tar-Pamlico case, the legal requirement to meet the cap was estab- lished by an enforceable contractual agreement signed by the association and the state. In the Neuse program, a single âgroup compliance permitâ was issued to the association. Both legal mechanisms established financial penalties for the two associations if aggregate discharges of the group exceed the association cap. A key advantage of the association is similar to that of a formal effluent trading programâgranting dischargers flexibility to decide how best to meet the aggre- gate load cap. To date, the associations have managed to keep nitrogen loads considerably below their respective caps. Compliance costs have also fallen below original projections. Further, there is some evidence that the association concept is producing incentives for strong cooperative behavior that did not ex- ist prior to implementation. The case studies presented here illustrate ways in which both public and private entities subject to regulation can exercise options for operating autono- mously should they not wish to incorporate with a group, while still contributing
506 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES to the achievement of watershed objectives. The case studies suggest that most dischargers conclude in the end that group membership offers considerable ad- vantages. Funding Considerations The existing stormwater permit program is characterized, in most of the na- tion, by municipal Phase I and now Phase II permittees operating mostly alone. In contrast the new system envisions coalitions of permittees that share a water- shed operating in concert, under the coordination and leadership of a principal permittee. The present structure tends to bring about duplication in effort and staff, whereas cooperation should stimulate efficiencies that could defray at least part or even much of the extra local costs associated with new responsibilities for municipal permittees. As explored in the preceding section, municipalities may not necessarily wish to join in co-permittee arrangements; and mechanisms are proposed to al- low them to operate individually, as long as watershed objectives are met. However, the state could encourage participation through financial inducements, for example, by estimating the resources needed to meet the requirements of each watershed permit and pointing out to permittees how shared resources can save each contributor money. The state should also set preferences and better terms for grants in the favor of municipalities who join together. To the questions of administrative vehicles and funding arrangements, stormwater utilities are the preferred mechanism, and regulations should support creating stormwater utilities. It should be added that, with watershed-based permitting as proposed here, utilities should also be regionalized on a watershed basis. A utility draws funds from the entities served in direct relation to the cost of providing the services, here management of the quantity and quality of stormwater discharged to natural waterbodies. These funds must be dedicated to that purpose and that purpose only, and cannot be redirected to general agency coffers or for any unrelated use. Not only are more funds from more reliable sources needed, but monies should be redirected in ways differing from their allocation under the current system. It was proposed earlier that a lead municipal permittee, working with other municipal co-permittees, be given responsibility for coordinating permit- ting and management of municipal, industrial, and construction stormwater permits, and even permits involving other sources, such as industrial process and municipal wastewaters. Those entities would hence be doing work now devolv- ing to individual private developers and industrial plants and other public au- thorities. They would need to attract the revenue from those other bodies in proportion to the added work taken on. A utility structure would provide a well- tested means of carrying out this reallocation. Stormwater utility fees are generally assessed according to a simple for- mula, such as a flat rate for all single-unit dwellings and in proportion to imper-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 507 vious area for commercial property. Some municipalities have investigated charging more directly according to the estimated quantity and quality of stormwater discharged into the public drainage system. Municipal permittees may choose to formulate such a system, but the development process itself is not a trivial task and, being based on general (and usually quite simple) hydrologic and water quality models, can generate considerable arguments from rate payers. Going through this process is probably not necessary or even advisable for most municipal permittees, who will have many new functions should the proposed system be adopted. Instead, they should concentrate on implementing a fee structure based on a simple formula like the one above and then capture addi- tional revenues for special functions that they will take over from industrial and construction permittees. As discussed previously, in the proposed program municipal co-permittees, with leadership by a watershed lead permittee, will be asked to classify indus- tries and construction sites within their borders according to risk and accord- ingly prioritize them for inspection and monitoring. It is proposed in the section on Measures of Achievement, below, that inspection include reviewing and ap- proving industrial and construction site stormwater pollution prevention plans (SWPPPs). While many municipalities now inspect construction sites for stormwater compliance and some inspect industries, this work will increase sig- nificantly in the new system, and SWPPP review and approval will be a com- pletely new element. Moreover, municipalities would perform some industrial monitoring now conducted by the industries themselves and may monitor high- risk construction sites. These special functions would require different institu- tional arrangements and substantial new revenue that could not be fairly charged to all rate payers. There are several possible sources for these funds. One way would be to increase industrial and construction permit fees and direct large proportions to municipalities to support inspection and monitoring. The permit- ting authority (designated state or EPA) would still hold ultimate authority, and municipalities could refer industrial and construction permittees found during inspection to be out of compliance to the permitting authority for enforcement. Another means would be to form consortia of industries of similar type and as- sess fees directly applicable to inspection and monitoring. For example, scrap- yards under the jurisdiction of the California EPA Los Angeles Regional Water Board formed a monitoring consortium under which sample collection by a qualified contractor rotates among the members, with funding by all. While the members operate this system, it could be adapted to operation by municipal co- permittees. A second-level funding concern is, once revenues are generated, how should they be put to use? It is very important that funds largely be devoted directly to the tasks at hand regarding the achievement of objectives instead of into excessive administrative and bureaucratic structure. These tasks are scien- tific and technical and are highly oriented toward what is actually going on in the drainage systems and their receiving waters. Thus, the majority of funds should be directed to making scientific and technical judgments based on obser-
508 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES vations and monitoring results obtained in the field (see the discussion below). Measures of Achievement Critique of the Current Monitoring System No area exemplifies the differences between the present and proposed new stormwater permitting and monitoring systems more than the measures used to gauge achievement. The current monitoring system is characterized by scattered and uncoordinated measurements of discharges from Phase I MS4s and some industries, and some visual observations of construction sites. The system pro- posed to take its place would emphasize monitoring of receiving water biologi- cal conditions as a data source for prescribing management adaptations to meet specified biological objectives. The discussion here first critiques the prevailing system to construct part of the rationale for changing it. It then proceeds to out- line a recommended monitoring structure to replace it. To expand very briefly on the point that the present system is scattered and uncoordinated, monitoring under all three stormwater permits is according to minimum requirements not founded in any particular objective or question. It therefore produces data that cannot be applied to any question that may be of importance to guide management programs, and it is entirely unrelated to the effects being produced in the receiving waters. Phase I municipal permit hold- ers are generally required to monitor some storms at some discharges for no stated purposes but to report periodically to the permitting agency (Phase II mu- nicipalities have no monitoring requirements, although they may represent the major or even only impact sources in a given watershed). The usual model for industries across the nation is to collect a few discharge grab samples a year and send the results to the permitting authority, plus occasionally to make observa- tions for obvious signs of pollution (e.g., oil sheen, odor). Construction site monitoring is less standardized and often involves no water quality monitoring at all. Again, no permittee under any of the three programs is obligated accord- ing to national standards to check the effects of its discharges on receiving wa- ters. Since the individual effects of any discharger are often not distinguishable from any other, the scattershot system would usually not be able to discern re- sponsibility for negative effects in the receiving water ecosystem. Input to the committee conveyed the strong sense that monitoring as it is be- ing done is nearly useless, burdensome, and producing data that are not being utilized. For example, the City of Philadelphia conducts substantial amounts of wet weather monitoring, which is very expensive, but it can barely monitor for TSS in many of its heavily impacted streams (Crockett, 2007). The resources to monitor for the more exotic pollutants do not exist. Smaller municipal permit- tees without the resources and sophistication of a big-city program have diffi- culty performing even the most basic monitoring. City water managers believe
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 509 that the traditional stormwater program places too much emphasis on monitoring of individual chemicals rather than looking at ecological results (Crockett, 2007). Industry representatives have also described several problems they see in industrial stormwater monitoring as it is performed now (Bromberg, 2007; Longsworth, 2007; Smith, 2007). One concerns the high degree of variability, from the methods used to what is actually measured (Stenstrom and Lee, 2005; Lee et al., 2007). Opponents have been quite critical of the benchmarks to which industrial monitoring data are compared, believing that the benchmarks have no basis in direct measurements associating stormwater with impacts. Some have suggested replacing monitoring with an annual stormwater docu- mentation report to the permitting authority. It seems that industry personnel disrespect the current monitoring framework for some good reasons and feel it conveys a burden for little purpose. There was some implication that industry would be receptive to measures offering more meaningful information in place of poorly conceived monitoring requirements (Bromberg, 2007; Longsworth, 2007; Smith, 2007). Proposed Revised Monitoring System A structure in several tiers is proposed as a monitoring system to serve the watershed-based permitting and management framework. Progress Evaluation Tier. This tier would represent the ultimate basis for judgment on whether the objectives adopted for the watershed are being met. Because these objectives would mainly be expressed in terms related to direct support of beneficial uses, so too would monitoring in the Progress Evaluation Tier principally emphasize direct measurements of ecological health. The pre- ferred model for this evaluation would be the paired watershed approach, which is based on the classic method of scientific experimentation and was developed for water resource management investigations by EPA (Clausen and Spooner, 1993). Ideally, conditions in the waterbody under evaluation would be com- pared to conditions in the same waterbody before imposition of a permit and management scheme (before versus after comparison), as well as to conditions in a similar waterbody not subject to human-induced changes (affected system versus reference system comparison). At least one of these comparisons must be made if both cannot. If the objectives involve improving conditions, and not just avoiding more degradation, the reference should represent that state to which the objective points. This function has traditionally been the province of the permitting authority (i.e., the designated state or EPA). In the new program, the function is assigned to municipal permittees, guided by the lead permittee, to conduct or contract, but with a substantial contribution by the permitting authority in the form of mate- rial support and guidance. The primary vehicle envisioned to perform the pro-
510 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES gress assessment is a well-qualified monitoring consortium serving the water- shed, and perhaps other watersheds in the vicinity. Case studies below present examples of successful joint ventures in monitoring that can serve as models. The proposal is based on the belief that monitoring should be more manageable and effective at the watershed compared to the state level and, furthermore, that utilizing a consortium approach should make it feasible for a coalition of mu- nicipal co-permittee partners to commission monitoring. Findings of objective shortfall would trigger development of active adaptive management strategies. Generally, an assessment should be conducted to de- termine what additional measures should be put in place in regulating new de- velopment and redevelopment, as well as increasing coverage of existing devel- opments with retrofits. Diagnostic Tier. The second tier would be designed to provide the munici- pal permittees with the necessary information to formulate active adaptive man- agement strategies, and they would be responsible for this second tier as well as the first. The Diagnostic Tier would be composed of assessment of information from the Compliance Reporting Tier, plus some specific field monitoring to determine the main reasons for ability or failure to meet objectives. Some highly directed monitoring of receiving water conditions could determine the need to improve management of water quantity, water quality, or both. A tool like the Vermont flow-duration curves is an example of a potentially useful de- vice for diagnostic purposes. To allow the use of such a tool, it is important that continuous flow recorders be installed on key streams in the watershed. The techniques described in the Impact Sources section above, once they are further developed, would also be useful in Diagnostic Tier monitoring. An important dimension of this tier would be prioritized inspection and monitoring of potentially high-risk industrial and construction sites. In addition, data submitted by the industrial and construction permittees according to the Compliance Reporting Tier would assist in targeting dischargers to bring about the necessary improvements in water quantity and/or quality management. Compliance Reporting Tier. It is proposed that the first step in compli- ance reporting be submission of SWPPPs by all construction and industrial per- mittees (plus municipal corporation yards as an industrial-like activity) to the jurisdictional municipal permittee for review and approval. It is further pro- posed that the industrial permittees and municipal corporation yards be relieved of sample collection, if they develop SWPPPs making maximum possible use of ARCD practices, supplemented by active treatment as necessary, and the mu- nicipal permittee approves the SWPPP. Construction sites would be given a similar sampling dispensation if they develop an approved SWPPP along the lines of Box 5-3. Otherwise, the permittees would be required to perform scientifically valid sampling and analysis and report results to the watershed co-permittees. This more comprehensive and meaningful monitoring would increase the burden al-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 511 ready felt by permittees and create a strong incentive to apply excellent SCMs. This burden could be relieved to a degree through participation with other simi- lar dischargers in the watershed in a monitoring coalition. As an example, in North Carolina coalitions of wastewater dischargers are working with the state Division of Water Quality (DWQ) to create and manage coalition-led watershed monitoring programs that operate in conjunction with DWQâs ambient chemis- try and biological programs (Atkins et al., 2007). Lee et al. (2007), after an as- sessment of industrial stormwater and other monitoring data, concluded that selecting a subset of permittees from each monitored category would yield better results at lower overall cost compared to monitoring at every location. This strategy would permit the use of more advanced sampling techniques, such as flow-weighted composite samplers instead of grab sampling, to estimate repre- sentative loads from each category with improved accuracy and reduced vari- ability. All permittees would still make observations of the SCMs and discharges and keep records. The final proposed step in compliance reporting is an annual report covering observations, SCM operation and maintenance, SWPPP modifi- cations, and monitoring results (if any), to be sworn as to correctness, notarized, and submitted to the lead municipal permittee. The Massachusetts Environ- mental Results Program (April and Greiner, 2000) offers a possible model for compliance reporting and verification. This program uses annual self- certification to shift the compliance assurance burden onto facilities. Senior- level company officials certify annually that they are, and will continue to be, in compliance with all applicable air, water, and hazardous waste management performance standards. The state regulatory agency reviews the certifications, conducts both random and targeted inspections, and performs enforcement when necessary. Research Tier. The final tier would be outside the permit system and exist to develop broad mechanistic understanding of stormwater impacts and SCM functioning important to assist permittees in reaching their objectives. EPA and state agencies designated to operate the permit system would have charge of this tier. These agencies would develop projects and contract with universities and other qualified research organizations on a competitive basis to carry out the research. Instructive Case Studies for the Proposed Revised Monitoring System Many municipalities, even large ones, would be challenged and burdened by taking on comprehensive watershed monitoring. The Southern California Coastal Water Research Project Authority (SCCWRP, http://www.sccwrp.org) offers an excellent model of how co-permittees in a watershed or an even broader area could organize to diffuse these challenges and burdens. SCCWRP
512 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES is a joint-powers agency, one that is formed when several government bodies have a common mission that can be better addressed by pooling resources and knowledge. In SCCWRPâs case, the common mission is to gather the necessary scientific information so that member agencies can effectively and cost- efficiently protect the Southern California marine environment. Key goals adopted by SCCWRP are defining the mechanisms by which aquatic biota are potentially affected by anthropogenic inputs and fostering communication among scientists and managers. Comprised of a multidisciplinary staff, SCCWRP encompasses units specializing in analytical chemistry, benthic ecol- ogy, fish biology, watershed conditions, toxicology, and emerging research. SCCWRPâs current mission stems from the results of a 1990 NRC review of marine environmental monitoring programs in the Southern California Bight (NRC, 1990). It was determined that although $17 million was being spent an- nually on marine monitoring, it was not possible to provide an integrated as- sessment of the status of the Southern California coastal marine environment. Most monitoring was associated with NPDES permit requirements and directed toward addressing questions about site-specific discharge sources. As a result, most monitoring in the bight was restricted to an area covering less than 5 per- cent of the bightâs overall watershed, making it difficult to draw conclusions about the system as a whole. The limited spatial extent of monitoring was also found to limit the quality of local-scale assessments, since the boundaries of most monitoring programs did not match the spatial and temporal boundaries of the important physical and biological processes in the bight. NRC (1990) further found that there was a lack of coordination among ex- isting programs, with substantial differences in the parameters measured among programs, preventing integration of data. Even when the same parameters were examined, they were often measured with different methodologies or with dif- ferent (or unknown) levels of quality assurance. Moreover, the NRC found that even when the same parameters were measured in the same way, substantial differences in data storage systems among monitoring programs limited access to the data for more comprehensive assessment. To avoid repetition of these shortcomings, the SCCWRP example should be given very thorough considera- tion as a template for the Progress Evaluation, Diagnostic, and Research Tiers in the proposed revised monitoring program. The San Gabriel River Regional Monitoring Program (SGRRMP, http://www.lasgrwc.org/SGRRMP.html) is a watershed-scale counterpart to the larger-scale regional monitoring efforts in Southern California. The SGRRMP incorporates local and site-specific issues within a broader watershed-scale per- spective. The program exists to improve overall monitoring cost effectiveness, reduce redundancies within and between existing monitoring programs, target monitoring efforts to contaminants of concern, and adjust monitoring locations and sampling frequencies to better respond to management priorities in the San Gabriel River watershed. Five core questions provide the structure for the re- gional program: What is the environmental health of streams in the overall watershed?
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 513 Are the conditions at areas of unique importance getting better or worse? Are receiving waters near discharges meeting water quality objectives? Are local fish safe to eat? Is body-contact recreation safe? The workgroup convened to establish the program recommended monitoring designs to answer the core questions effectively and efficiently. The resulting program is a multilevel monitoring framework that combines probabilistic and targeted sampling for water quality, toxicity, and bioassessment and habitat con- dition. The City of Austin, Texas, has more than 20 years of stormwater monitor- ing experience and offers additional guidance on designing and implementing watershed monitoring programs (City of Austin, 2006). Austin performs de- tailed periodic synoptic sampling in the watersheds it manages to track trends in stormwater quantity and quality. The city uses the results to evaluate the im- pacts of land development on stormwater quantity and pollution, establishing statistical relationships between measures of these conditions and the amount of impervious cover. Trend assessment over time leads to recommended changes to the City of Austin Environmental Criteria Manual as needed. Creating Flexibility and Incentives Within a Watershed Approach A watershed-based permitting approach to stormwater management focuses attention on watershed objectives and endpoints. To be able to achieve these goals, observable performance measures beyond the success of an individual SCM need to be identified that are consistent and necessary to meet designated uses. These might include watershed-level numeric limits on the amount of a particular pollutant allowed to enter a waterbody (e.g., pounds of phosphorus) or various measures of allowable volume of discharge. A watershed focus shifts attention away from specific SCM performance and site-specific technological requirements to achieving a larger watershed goal. As a consequence, there is considerable management flexibility in deciding how these goals will be achieved. Indeed, this flexibility was cited by the NRC (1999) as a prerequisite to successful watershed management. One way of exercising this flexibility is to create an âincentive-basedâ or âmarket-basedâ approach to choose how watershed goals are met. It is recog- nized throughout the environmental management field that entities subject to regulation do not necessarily have equal opportunities and qualifications to comply sufficiently to sustain resources. To compensate for this, the market- based approach allows individual discretion to select how effluent (or runoff volume) will be controlled (choice of technology, processes, or practices) and where they will be controlled (on site or off site). That is, any discharger legiti-
514 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES mately unable to meet discharge quantity and quality allocations would be able to finance offsets elsewhere to achieve the watershed goals. An important ele- ment and challenge is to couple this decision-making flexibility with personal (typically financial) incentives so that people willingly make choices supportive of the watershed objectives. Broadly stated, the idea is to create financial rea- sons and decision-making opportunities to lower compliance costs and create or implement new effluent/volume control options (Shabman and Stephenson, 2007). Because incentive-based policies require a shift in emphasis from technolo- gies and practices to outcomes (e.g., volume or quantity of effluents), the mu- nicipal manager would not be responsible for deciding what SCM will be im- plemented in specific areas or hand picking specific practices to promote. Rather the stormwater program managerâs responsibilities shift to establishing watershed goals, developing metrics to measure outcomes and performance, and performing necessary inspection and enforcement activities. Effluent trading, sometimes called âwater-quality trading,â is one type of incentive-based policy. In an ideal form, effluent trading requires government to establish a binding aggregate limit or cap on an outcome (e.g., mass load of effluent, volume of runoff) for an identified group of dischargers. The cap or aggregate allowable discharge is set to support and achieve a socially deter- mined environmental goal. Because it is fixed, the cap provides the public as- surances that environmental objectives will be achieved in the face of a growing and changing economy. The total allowable discharge is then divided into dis- crete and transferable units, called allowances, and either distributed or auc- tioned to existing dischargers. All dischargers must own sufficient allowances to cover their discharges. For instance, any new or expanding source must first purchase allowances (and hence effluent or volume reductions) from another source before legally discharging. The requirement to hold allowances on the condition to discharge and the positive allowance price creates financial incen- tives for pollution prevention. Dischargers holding allowances rather than re- ducing discharge face forgone revenues that could have been achieved from the sale of allowances. Conversely, expanding dischargers have incentives to invest in pollution prevention in order to avoid the cost of purchasing additional allow- ances. In the context of the revised permit system advocated here, achievement of objectives (generally of a biological nature) will require some combination of strategies such as no net increases in hydrologic parameters (e.g., peak flow rates, durations, volumes), water pollutants, forest cover loss, and effective im- pervious area. If one entity is unable to contribute adequately to meeting its share of compliance, then it must obtain the necessary credit by buying it from another similar entity that is able to contribute more than its designated share. Ideally, all sources of a waterbodyâs problems, not only stormwater, would come under the trading system. Implementing the market system requires development of a resource-based currency, a nontrivial exercise but one for which models are available in other
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 515 fields, especially air emissions. For example, emission trading has been a criti- cal element of the nationâs strategy to limit sulfur dioxide and nitrogen oxide emissions (Ellerman et al., 2000). Carbon trading is a cornerstone policy in the European Union effort to limit greenhouse gas emissions. The EPA promotes the use of trading to help achieve the goals of the CWA and has issued several policy statements and recently published guidance on how trading programs can be grafted within existing NPDES permitting programs (EPA, 2003a, 2007b). However, compared to the air program, experience and success with trading in the water program have been limited (Shabman et al., 2002). Furthermore, programs labeled trading have been implemented in a multitude of ways in the nationâs water quality program (Woodward et al., 2002; Stephenson et al., 2005; Shabman and Stephenson, 2007). In many instances, trading programs are case- specific and isolated âtradesâ that do not fundamentally change the choice and incentives facing dischargers in a conventional permitting system. The extent to which trading policies can be effectively employed on a watershed scale is lim- ited not only by the physical differences between air and water mediums, but also by the unique legal structure of the CWA (Stephenson et al., 1999). For example, the CWA is oriented around imposing technology-based performance requirements on specific subset of discharge sources. Individual NPDES per- mits require sources to achieve these agency-identified levels of performance and may specify how performance is achieved. The statute also places limits and disincentives on the degree to which permit agencies can deviate from these limits (e.g., âantibackslidingâ). Thus, the focus of the NPDES permitting system has been on individual source control and technologies, unlike the air program, which has a stronger statutory orientation around achieving broader air quality goals (ambient air quality standards). The orientation of the NPDES program limits the flexibility and incentives for regulated parties that might make market-oriented trading possible. It turns out that some of the more successful applications of trading in the water program have occurred because of permitting innovations that effec- tively avoid some of these rigidities (see discussion of North Carolina point source control program on the Neuse River, above). Trading programs of various types have been proposed or suggested for stormwater (Thurston et al., 2003; Parikh et al., 2006). Although conceptual models of a comprehensive trading program based on the total volume of allow- able water to be discharged have been proposed, no working examples have yet to be implemented. More limited versions of trading programs, however, have been developed. These programs provide compliance flexibility for new sources of stormwater runoff. In some locations, new developments face a requirement to provide a specific level of volume or effluent control from the parcel to be developed. The regulated entity is typically obligated to meet this requirement with the applications of on-site SCMs. Trading programs create opportunities for regulated entities to meet their regulatory requirement off site (off the parcel to be developed), called here an offset. In some trading programs, the off-site controls can be accomplished by the creation of an in lieu fee program. Such
516 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES programs typically occur for dischargers that are not required to hold or obtain individual NPDES permits. In lieu fee programs offer some opportunity for regulated parties to make a financial payment (fee) to a local government entity in lieu of implementing on- site controls. The fees are collected and used to implement stormwater controls in other areas of the watershed. Controlling runoff at a regional level rather than through the construction of many small on-site controls may be more cost- effective given the economies of scale associated with some SCMs (see Chapter 5 pages 362â363). The option for off-site controls also allows the stormwater program to direct investments in stormwater control to specifically targeted ar- eas of the watershed. Examples of in lieu fee programs include Santa Monica, California, the Neuse River Basin in North Carolina, and Williamsburg, Virginia. Santa Monicaâs program requires new and redevelopment projects to treat a specific volume of runoff. The program first requires the regulated entity to take all fea- sible steps to meet the requirement through the implementation of on-site infil- tration practices. If the regulated party can demonstrate why it is economically and physically infeasible to install any type of infiltration or treatment SCM, the regulated party can pay a fee based on the volume of water that needs to be con- trolled (the total mitigation volume is the volume that would have been attenu- ated via an SCM). The fee set by Santa Monica is $18/gallon of total required mitigation volume. The $18 reflects the cost of constructing an SCM and main- taining it over 40 years (DeWoody, 2007). Presumably these fees are used to construct infiltration measures elsewhere. The Neuse River Program requires all new land development to meet a ni- trogen export standard of 3.6 pounds per acre per year (North Carolina Division of Water Quality, 1999). The water quality goal for the Neuse basin is to reduce mass nitrogen loads by 30 percent in order to improve water quality in the estu- ary. The export standard was set to achieve a 30 percent reduction from the av- erage nitrogen load from lands prior to development. Developers have the op- tion to meet this export standard either through the application of on-site SCMs or by paying a fee into a state-administered Riparian Buffer Restoration Fund (see 15A North Carolina Administrative Code 02B .0240), which would be used to reduce nitrogen loads elsewhere in the basin. Developer discretion, however, is not unlimited. Under no circumstances may developers discharge more than an estimated 6.0 pounds per acre per year from a residential site. The Williamsburg program has an in lieu fee program for total phosphorus loads created by new development (Frie et al., 1996; Stephenson et al., 1998). For every new development, the increase in total phosphorus load from storm- water runoff from impervious surfaces is estimated. Developers have the choice to meet the phosphorus load reduction requirement through the application of on-site controls or by paying a fee to the city. The fee is set at $5,000/lb of phosphorus, with the fees earmarked to the construction of regional stormwater facilities or for the preservation of open space within the city. The presence of a fee option could also provide incentives for developers to implement source
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 517 reduction practices. The above programs differ in some important ways. For example, the Santa Monica program requires regulated entities to undergo a âsequencingâ process that places regulatory preference on on-site controls before being able to use the fee option. The Williamsburg program allows regulated entities the option to select between constructing on-site controls and paying the fee without a regula- tory preference for on-site controls. Sequencing rules tend to limit control op- tions and thus the cost-effectiveness of these types of programs. In lieu fee programs are distinguished from other offset programs in that it is the responsibility of the local government (or more generally, any designated fee service provider such as a nongovernmental organization) to provide the off- site SCMs. In lieu fee programs, common in the U.S. wetlands program, face a number of implementation and design challenges (Shabman and Scodari, 2004). For example, enforcement sometimes becomes a concern because the local stormwater management agency responsible for constructing and maintaining the SCMs is also responsible for monitoring and enforcement. These dual re- sponsibilities create potential conflicts of interest; if an off-site mitigation pro- ject fails, there maybe no apparent overseeing agency to enforce corrective ac- tions. The lack of transparency in accounting to determine whether the offset projects provide enough compensation is also sometimes a challenge. Finally, the ability to fully offset the volume of effluent discharge from a new develop- ment is contingent on collecting enough revenue from the fee to pay for the con- struction and maintenance of offsite SCMs. The delay between impacts and compensation and lack of full public cost accounting complicate the challenges of setting an appropriate fee. Ensuring that in lieu fee programs provide the necessary mitigation could be accomplished in a number of ways. For example, an oversight agency may be designated to establish tracking and reporting requirements and monitor in lieu fee program performance. Or, the potential conflicts of interest inherent in the lieu fee program design could be avoided by separating the provision of the off- site mitigation service from the monitoring and enforcement. It is possible to imagine that the private sector, rather than an in lieu fee administrator, could provide off-site stormwater reduction services to those subject to the stormwater control requirements. In this case, the private sector would provide stormwater detention/retention services above and beyond what is required by law. These private service providers would receive stormwater runoff credits for these in- vestments (âabove baselineâ) that could be sold to developers who might wish to meet their control obligations in ways other than on-site controls. In essence, the role of searching, designing, and constructing offsite SCMs would be trans- ferred to the private-sector stormwater credit providers. The local stormwater managers, however, would retain full authority to monitor, verify, and enforce to ensure that these offsets are successfully implemented. The flexibility provided by in lieu fee and trading programs requires that pollutant loads or runoff volume created at one site be reduced at another site. Thus, a design issue confronting these types of programs is the consideration of
518 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES the spatial extent in which offsetting activities can occur. The extent of the spa- tial range of offsetting activities in turn will depend partly on the nature and type of service being offset. For example, in the Neuse example nitrogen is a re- gional, basinwide concern with minimal localized effects. In such cases, the offsetting activities might be allowed basinwide (after adjusting for nitrogen attenuation through the basin). In other situations where localized concerns maybe a greater concern (say from localized flooding), the flexibility offered by such programs may be more limited. However, such spatial flexibility might also be a way to implement and achieve watershed planning objectives. For example, development may be encouraged in high-impact areas, and offsetting fees could be used to protect and enhance water quality objectives in other areas. This last point deserves further explanation. Although this chapter advo- cates that biological conditions in waterbodies should be maintained or im- proved, there are many urban areas where local waterbodies cannot achieve the same designated uses as less developed areas. If a goal-setting entity chose to do so, beneficial uses for waters in these areas could be set at levels that ac- knowledge this highly altered condition, such that these streams would not be expected to achieve the same biological condition as streams outside the urban core (see Chapter 5 pages 364-366). This might be done to encourage develop- ment in high impact areas; San Jose, CA, provides an example (see Chapter 2). In that cityâs stormwater program, in urban areas where on-site control is either technically impossible (due to soil or space constraints) or prohibitively costly, the developers can meet the post-construction treatment standard by providing volume control either through participation in a regional stormwater project or by providing equivalent projects off site (e.g., stream restoration). It is also possible to design a stormwater offset program that allows the dif- ferent functions of stormwater management to be separated to achieve watershed objectives. For example, management of peak flow serves mostly to prevent localized flooding while more stringent volume control maybe required to pro- tect stream channels and aquatic life. Control of peak flow might be required on site or within a narrow geographic region. In areas targeted for development, however, the volume control needed for channel protection might be transferred off site and into areas where watershed planning has identified the need for higher levels of stream channel protection or enhancement (more stringent water quality standards). A similar watershed approach based on functional assess- ment was recommended for wetland compensation (NRC, 2001b). Regulatory and Legal Implications of Proposed Watershed-Based Permitting Framework for Managing Stormwater EPA, the states, and municipal permittees would all have tasks to perform to transform the framework set forth in this report to a fully developed and func- tioning program. These efforts would be rewarded with a program that is rooted
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 519 in science, transparent in its aims, fairer for all than the current program, and better for the aquatic environment. This section of the report outlines the tasks necessary to carry the proposal forward to full development. EPA should seek significant congressional funding to support the states and municipalities in undertaking this new program, in the nature of the support dis- tributed to upgrade municipal WWTPs after the 1972 passage of the Federal Water Pollution Control Act. Beyond financial support, EPAâs tasks emphasize broad policy formulation, regulatory modifications and adaptations necessary to initiate the new program, and guidance to the states and permittees. The princi- pal adaptation needed in the regulatory arena involves converting the current TMDL program to a form suitable for the new system. Guidance would be needed in a number of crucial areas, and it is EPAâs natural role to develop it. States (or EPA for states without delegated authority) would have broad re- sponsibilities to translate policies and federal regulations into their own regula- tory and management systems. A key task in this regard would be to recast wa- ter quality standards into objectives most directly supporting sustenance and improvement of beneficial uses. States already have considerable background for performing this task through their present definitions of beneficial uses, the Section 303(d) process for assessing waterbody compliance with water quality standards, and the triennial review of those standards. However, the added prominence of biological aspects of beneficial uses and associated objectives will require additional analysis. Other prominent state tasks will involve defin- ing the watersheds subject to permits, forming bodies of co-permittees associ- ated with the watersheds, and appointing the lead permittee. Many other state tasks entail cooperative work with the permittees to support and assist them in funding and conducting their activities. Many aspects of the municipal permitteesâ roles in implementing strategies were explored above in a section titled accordingly. That section especially fo- cused on activities to advance the use of ARCD methods. More broadly, the permittees will be coordinators of all permits pertaining to the watershedâs aquatic resources, collectively pointed toward meeting objectives that the per- mittees adopt under state oversight. Other categories of tasks assigned to the municipalities under the proposed system include monitoring, in the contexts of both inspections and sampling performed through a consortium, and enforce- ment actions and program adaptations to promote progress toward achieving objectives. Box 6-4 provides a listing of anticipated tasks for the municipal permittees as well as the states and EPA. A Pilot Program as a Stepping Stone The shift of responsibility for stormwater regulation to municipalities under the watershed-based approach may lead to some surprises in implementation and enforcement. Primarily because of this, EPA is well advised to institute a pilot
520 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-4 Government Agencies Roles during the Operation of a Watershed-Based Permitting System EPA 1. Petition Congress for significant funding support for states and municipal permit- tees, and develop a program of fairly distributing funds based on environmental and financial needs at the watershed level. 2. Initiate regulatory modifications and clarifications necessary to establish the system. 3. Set policies for watershed permitting based on this reportâs recommendations. 4. Adapt TMDL program for use in the new program. 5. Produce guidance to assist the states and municipal permittees in the areas of: a. Developing a rotating basin approach; b. Developing an integrated municipal NPDES permit incorporating the full range of sources; c. Developing stormwater utilities and other funding mechanisms; d. Using impact source analysis (e.g., using reasonable potential analysis and new research results, industrial and construction site risk assessment); e. Using ARCD techniques for new development, redevelopment, and retrofitting; f. Developing monitoring consortia; g. Developing a credit trading system; h. Developing an active adaptive management program Designated States (or EPA otherwise) 1. Define watersheds for which permits will be issued and set up a rotating basin ap- proach to govern watershed analysis in support of subsequent steps. 2. Formulate and formally adopt goals relative to avoiding any further loss or degrada- tion of designated beneficial uses in each watershedâs component waterbodies and recover- ing lost beneficial uses. 3. Use the results of the existing Section 303(d) process and supplementary work to assess the extent of designated beneficial use achievement in each watershed and set goals for protection and recovery. 4. Match municipal permittees to watersheds and designate a lead permittee for each watershed. 5. Estimate resource needs to fulfill permit requirements in each watershed. 6. Develop a grant program, drawing on EPA and state funds, to support municipal permittees, with incentives for joining co-permittee associations. 7. Identify areas outside the jurisdictions of permitted municipalities that should be brought into the program because of projected development or the existence of problem sources that would compromise the protection and recovery of beneficial uses. 8. Use the triennial review process to modify water quality standards to the objective basis, emphasizing biological outcomes recommended in this report. 9. Revise the TMDL program in accord with the needs of the new program. 10. Set requirements for credit trading systems. 11. Set up an integrated municipal NPDES permit incorporating the full range of sources. 12. Work with municipal permittees to establish specific objectives as the basis for pro- gress assessment. 13. Work with municipalities to develop adaptive management programs responding to progress assessment results.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 521 14. Write municipal permits incorporating the above elements. 15. Write industrial and construction general or individual permits incorporating the recommendations in this report. 16. Allocate a substantial portion of industrial and construction permit fees to munici- pal permittees to oversee those sectors. 17. Set requirements for municipalities and private properties to opt out of the de- fined program without compromising the achievement of objectives. 18. Provide consultation, support, and guidance (adapted from EPA materials or origi- nally produced) to municipal permittees in the areas of: a. Developing stormwater utilities and other funding mechanisms; b. Using impact source analysis (e.g., industrial and construction site risk as- sessment); c. Using ARCD techniques for new development, redevelopment, and retrofit- ting; d. Developing monitoring consortia; e. Developing a credit trading system 19. Perform enforcement actions on non-complying dischargers referred by munici- pal permittees. 20. Assess performance of municipal permittees and specify corrections, rewards, and penalties accordingly. Municipal Co-permittees (led by Lead Permittee) 1. Adopt specific objectives as the basis for program progress assessment. 2. Convert ordinances and regulations as needed to implement the modified pro- gram. 3. Supplement and reorganize staffing to emphasize progress and compliance as- sessment as the principal functions of the program. 4. Perform or contract detailed scientifically and technically based watershed analy- sis as a foundation for permit compliance. 5. Assemble existing data on soils and hydrogeologic properties and supplement with additional data collection as necessary to assess infiltration prospects across the mu- nicipality. 6. Create incentives for private property owners to maximize the use of ARCD methods in new development and redevelopment. 7. Build subwatershed-scale, publicly owned ARCD works to supplement on-site management measures and as retrofits. 8. Develop capacity for stormwater management in municipal WWTPs by reducing groundwater inflows to sanitary sewer lines. 9. In areas experiencing excessive infiltration and groundwater table rise resulting from non-stormwater flows, develop capacity for stormwater management through infiltra- tion by formulating water conservation programs. 10. Identify industries and construction sites that are required to apply for permits but have not done so and compel their filing. 11. Establish or enhance existing programs to inspect and oversee industries and construction sites; report non-complying dischargers to the state for enforcement actions. 12. Set up or join a monitoring consortium structured to implement the progress evaluation and diagnostic tiers of the proposed monitoring program. 13. Annually report monitoring results to the permitting authority; submit a compre- hensive progress assessment triennially.
522 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES program that provides some experience in municipality-based stormwater regu- lation before instituting a nationwide program. This pilot program will also al- low EPA to work through more predictable impediments to this watershed-based approach. The most obvious impediment arises from the inevitable limits of an urban municipalityâs responsibility within a larger watershed: substantial growth and accompanying stormwater loading may occur on the outside periphery of a municipalityâs designated boundaries. If an urban authority lacks legal authority over this future growth, and if this growth contributes significantly to water quality degradation, then a considerable share of the urban stormwater problem could remain poorly addressed. A pilot program should help identify the extent of this jurisdictional slippage and help identify ways to overcome it. Second, it is possible that some municipalities will balk at the added responsibility in- volved with the watershed-based approach, even with adequate funding. Unless the objective performance standards are rigid, the monitoring requirements sub- stantial, and the rewards for compliance compelling for municipalities that meet the standards, it is quite possible that noncompliance or bare minimal compli- ance will be the norm. A pilot program provides a less politically charged at- mosphere to experiment with the benefits of watershed-based regulation at the local level and to generate local government support for the approach. Finally, because the watershed-based approach necessitates legislative amendments to the CWA, instituting a pilot program in the interimâboth to improve the design of a watershed-based program as well as to generate enthusiasm for itâseems a sensible course. The pilot program should target those local governments that are most eager to redress water quality degradation in their watersheds, but feel stymied by what they perceive as inadequate legal authority and flexibility to make the nec- essary improvements. Willing municipalities or regional governments would thus opt-in to the program. The pilot program entices these more progressive municipalities to participate by allowing them to serve as the lead authority and providing them with much greater flexibility to determine how to meet their performance-based water quality goals with fewer legal constraints. Under the pilot program, a municipal government or similar legal authority would apply to EPA or a delegated state to be designated as the lead agency for that portion of the watershed within its legal jurisdiction. In the application it- self the municipality would establishâusing modeling and ambient dataâhow it plans at a general level to maintain or exceed its water quality goals (objective performance standards). These goals must be at or above the state water quality goals, or if they are different (i.e., use biological criteria when the state adopts chemical criteria), the municipality must demonstrate how its performance stan- dards will attain the equivalent of the state water quality goals at the down- stream edge of the municipalityâs border. The municipality would also be re- quired to provide assurance of sufficient infrastructure and funding to allow it to develop a water quality plan, implement that plan, issue permits, and enforce the requirements within its boundaries. Finally, municipal plans, once finalized, would need to meet minimum federal procedural requirements. For example,
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 523 the plans must be transparent and provide opportunities for public comment; they must be enforceable; and they must establish monitoring programs that will track whether they in fact meet the objective performance standards. If a mu- nicipality fails to meet any of its performance standards by the requisite dead- line, the state and EPA would have the option of revoking the municipalityâs program, and reinstituting federal requirements. Ideally, federal guidance would also be available to municipalities to provide direction on how they might insti- tute a watershed-based plan within their boundaries, while still reserving consid- erable flexibility to allow them to develop creative and progressive stormwater solutions. For example, municipalities would be encouraged to form stormwater utilities that are financed from point and even nonpoint sources that assist them in establishing rigorous permitting and enforcement of their water quality plan. Municipalities that voluntarily take on this role as lead authority will be re- warded with few legal constraints on how they meet their performance-based objectives. NPDES permits for major sources will still be required and must meet federal minima (technology-based controls) to avoid possible hot spots surrounding large dischargers, and states would remain listed as the lead permit- tee for these permits, but the lead municipality or other regional government would be able to propose new, more stringent limits that are presumptively fa- vored in revised NPDES permits. Stormwater permits would also be mandatory, but their substantive requirements would be left wholly within the discretion of the lead municipality. Finally, states and municipalities would not be required to comply with all of the federal regulations governing TMDLs (they would make a basic load calculation for pollutants contributing to degraded conditions, 33 U.S.C. Â§ 1313(d), but would not be required to do more). Instead, the water- shed-based program would be considered the functional equivalent of TMDLs for at least the municipalityâs portion of the watershed since the program ensures that water quality objectives are met. Municipalities could even be allowed to set interim goals over a period of a decade or more so that TMDLs need not be achieved in a single permit cycle. Other than federal minimum standards for major NPDES sources, munici- palities would have primary if not exclusive authority to decide what types of sources (including nonpoint) require permits, whether certain land uses might be taxed for stormwater management fees, and whether and how to create trading programs among the contributors to water quality impairments within their wa- tershed. Municipalities would also have legal authority to petition EPA to re- strict upstream sources that contribute significantly to water quality degradation in ways that make it difficult for them to reach their goals. Upstream govern- ments or sources could be subject to more rigorous federal or state TMDLs and could be vulnerable to tort and related claims from downstream municipalities. This added flexibility and authority for municipalities to control water qual- ity problems within their legal jurisdictionâcoupled with objective performance standardsâshould lead to more creative approaches to stormwater management that create significant benefits to the municipality (i.e., more green-space buffers along waterways for recreation) and stronger planning and taxation of new de-
524 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES velopments that otherwise might be uncontrolled. Municipal green space, parks, and a variety of other public goods that both reduce stormwater and enhance the public enjoyment of the surface waters could result from allowing a municipal- ity the freedom to determine how best to regulate sources within its local boundaries. For example, rather than automatically allowing federally approved SCMs that have little aesthetic or recreational qualities, alternative approaches to SCMs that retain their effectiveness but provide other qualities (particularly qualities that draw the public outdoors for recreation or relaxation) are more likely to be encouraged or even required by a municipality that serves as lead over implementation of its water quality program. Although a national watershed-based approach to stormwater regulation is likely to require legislative amendments, the pilot program may not necessitate additional legislative authorization. It is possible that through regulation, EPA may be able to develop âin lieu ofâ or âfunctional equivalentâ requirements that allow a rigorous watershed plan to substitute for the bare federal requirements governing stormwater regulation, general permits, and TMDL planning laid out in the CWA. This type of intricate legal analysis, however, is beyond the scope of this document. Final Thoughts The watershed-based stormwater permitting program outlined above is ul- timately essential if the nation is to be successful in arresting aquatic resource depletion stemming from sources dispersed across the landscape. EPA is called upon to adopt the framework now and set in motion a process to move it toward implementation over the next five to, at most, ten years. This chapter deals with some but not the entire realm of political, legal, regulatory, and logistical issues raised by converting to a fundamentally different system of management and permitting. Ideas are contributed regarding piloting and transitioning toward the new program, altering institutional arrangements to accommodate it, and incen- tives for effective participation. For watershed-based permitting to take hold, specific actions will have to be undertaken by EPA, state permitting authorities, and municipal permittees during the adoption and transition process. The proposed program could be implemented by EPA in a number of ways, ranging from making it mandatory without any exception in all states and juris- dictions to leaving it entirely voluntary. The committee recommends neither extreme and believes the best course would be: (1) pilot test and refine the pro- gram as described in the report section titled âA Pilot Program as a Stepping Stone;â (2) make the refined program the default to be followed by all desig- nated states (and EPA in others) and all municipal, industrial, and construction permittees, unless a state permitting authority convincingly demonstrates to EPAâs satisfaction than an alternative approach will accomplish the programâs overall goal of retaining and recovering aquatic resource beneficial uses; (3) develop very significant incentives for states and permittees to participate; and
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 525 (4) require objective demonstration by any state opting for an alternative that it is broadly achieving the goal to at least the same extent as states within the pro- gram, with appropriate sanctions for noncompliance. ENHANCEMENT OF EXISTING PERMITTING BASIS The current federal stormwater regulatory framework has been in place since 1990, and the point source NPDES program under which it is being im- plemented has existed since 1972. The U.S. Congress deliberately acted in 1987 to amend the federal CWA with the goal of addressing stormwater pollution because it had been identified as a leading cause of surface water impairments, and regulations were inadequate to address it effectively. The total rethinking of the current framework of regulating stormwater pollution described above may require changes in statute and take a long time to implement. Thus, in addition to the longer-term approach that integrates a watershed-wide planning and per- mitting strategy into the program, several near-term solutions are also offered, with the objective of improving the current regulatory implementation and which at most might require changes in regulation. Problems Complying with Both Municipal and General Industrial Permits The NPDES permitting authority issues (1) separate individual permits or general permits to impose discharge requirements on small, medium, and large MS4s; (2) general permits that require construction activity operators who dis- charge stormwater to waters of the United States, including those who discharge via MS4s, to implement SCMs; and (3) general permits for operators of storm- water discharges associated with industrial activity who discharge to waters of the United States, including those who discharge via MS4s, to implement SCMs. The MS4 operators in turn are also required under the terms of their MS4 per- mits to require industries and construction site operators who discharge storm- water via the MS4 to implement controls to reduce pollutants in stormwater dis- charges to the maximum extent practicable, including those covered under the permitting authorityâs NPDES general permits. This dual-coverage scheme ap- pears intended to recognize the separation of governmental authorities. Unfor- tunately, in practice it is duplicative, inefficient, and ineffective in controlling stormwater pollution that enters the MS4 from diffuse and dispersed sources. Particularly in the area of monitoring of water quality, the dual approach seems to have resulted in a lack of prioritization of high-risk industrial sources and the purposeless collection of industrial stormwater monitoring data or the poor use of it to strategically reduce the discharge of stormwater pollutants to the MS4. The preference of EPA to use general NPDES permits to alleviate the ad- ministrative burden associated with permitting more than a 100,000 point
526 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES sources discharging stormwater is understandable. It would have been prudent to have some form of prioritization to select some subset of the whole as high- risk or have a strategy for identifying a subset for individual NPDES permits to better achieve the objective of ensuring compliance with water quality standards on the basis of potential risk. As discussed in Chapter 2, there are no federal guidelines for prioritization (determining what industries are high-risk for stormwater discharges), and the state permitting authorities have largely not prioritized because of the overwhelming burden of administering a very expan- sive stormwater permitting program. In the existing permitting scheme, the MS4 operator cannot be faulted for having a reasonable expectation that the permitting authorityâs general NPDES permits that regulate industrial activities and construction that discharge to the MS4 would require, at a minimum, a sufficient level of identification and im- plementation of SCMs to facilitate the MS4 operatorâs compliance with the MS4 permit. However, such controls are not identified by the NPDES permitting authority and rather are left to the choice of the industrial facility and construc- tion site operators. Furthermore, the NPDES permitting authority imposes weak to no discharge sampling requirements on industrial facility and construction activity operators, which greatly impairs the MS4âs ability to determine and control the worst regulated stormwater discharges to the MS4. Similarly, the NPDES permitting authorityâs general permit for construction activity encour- ages construction facility operators to consider post-construction stormwater controls, but it does not require them, even though the MS4 permitâs program- matic measures mandate new development planning and post-construction con- trols as essential elements of the MS4 program. The lack of integration among stormwater permits and the absence of objective measures of compliance that are quantifiable is a glaring shortcoming in current stormwater permits and ren- ders them difficult to enforce for water quality protection. The California EPA State Water Board asked an expert panel to evaluate the extent of implementation success of the stormwater program in California and the feasibility of numeric effluent limits in stormwater permits. In its report (CA SWB, 2006), the panel concluded that the flexible approach of allowing a per- mittee to self-select SCMs for the purpose of controlling stormwater pollution was largely ineffective. The reasons stated were: (1) the SCMs were selected without proper consideration of design, performance, hydraulics, and function; (2) the MS4 permittees were not accountable for the performance of the SCMs; (3) the industrial and construction permittees were not responsible for the per- formance of the SCMs; and (4) the SCMs were seldom maintained properly except for aesthetic purposes. In other words, the flexibility provided by self- determination, self-evaluation, and self-reporting did not assure that SCMs were being implemented to effectively reduce stormwater pollutants to the MEP. Rather, the flexibility resulted in a lack of coordination of purpose and account- ability between the MS4 permittees who owned or operate the MS4 and the in- dustry and construction permittees who discharge to the MS4. Although typi- cally enforcement by the permitting authority would have restored the integrity
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 527 of the stormwater program, that remedy is likely to be ineffective here because the choice of SCMs is left too much to discretion and there are no quantifiable performance or design criteria for water quality purposes. Integration and Dissemination of Authority This section offers a near-term alternative solution to the problem cited above that utilizes the existing framework of the NPDES stormwater program. The strategy builds on the authority of MS4s over industry and construction sites to implement an integrated permitting scheme to reduce stormwater pollution into the waters of the United States. Unlike the first section of this chapter, it does not take a watershed approach to protecting water quality, even though the municipal stormwater programs may be more cost-effective if implemented on a watershed scale. It also addresses a significant shortcoming of the current scheme, that is, failure to recognize the enormous staff resources that it would take at the federal and state level for successful implementation in the absence of the leadership of local governments. Further, federal and state NPDES per- mitting authorities do not presently have, and can never reasonably expect to have, sufficient personnel under the principles of democratic governance, such as in the United States, to inspect and enforce stormwater regulations on more than 100,000 discrete point source facilities discharging stormwater. A better structure would be one where the NPDES permitting authority empowers the MS4 permittees, who are local governments working for the public good, to act as the first tier of entities exercising control on stormwater discharges to the MS4 to protect water qualityâan approach here called âintegration.â The central concept of integration is to give the MS4s controlling jurisdic- tion and responsibility over discharges from construction and industry to the MS4 in addition to their responsibility to implement the programmatic minimum measures identified in regulation. This approach would be similar to the current NPDES permitting scheme for publicly owned WWTPs, where a WWTP opera- tor controls the quality of wastewater inputs (industrial waste streams) to make sure that the total output will not exceed water quality standards (see Box 6-5 on the National Pretreatment Program). The WWTP operators establish additional criteria such as local limits, require discharge monitoring of industrial wastes, and conduct inspections to make sure industrial discharges implement adequate wastewater treatment technologies, so that treated effluent from the wastewater treatment can comply with water quality standards to protect receiving waters. The same could be done for stormwater, except here the WWTP is replaced by the MS4, and the other inputs in this case are all industrial and construction dis- charges of stormwater into the MS4. The criteria by which the outputs of the industries are judged could be either water quality- or technology-based criteria. This arrangement puts the burden on the MS4 to identify high-risk industries because the MS4 is now responsible for the overall output (which could be, for example, the concentration of pollutants in stormwater monitored during
528 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-5 National Pretreatment Program EPAâs NPDES Permitting Program requires that all point source discharges to waters of the United States (i.e., âdirect dischargesâ) must be permitted. To address âindirect dis- chargesâ from industries to Publicly Owned Treatment Works (POTWs), EPA, through CWA authorities, established the National Pretreatment Program as a component of the NPDES Permitting Program. The National Pretreatment Program requires industrial and commercial dischargers to treat or control pollutants in their wastewater prior to discharge to POTWs. In 1986, more than one-third of all toxic pollutants entered the nationâs waters from POTWs through industrial discharges to public sewers. Certain industrial discharges, such as slug loads, can interfere with the operation of POTWs, leading to the discharge of un- treated or inadequately treated wastewater into rivers, lakes, etc. Some pollutants are not compatible with biological wastewater treatment at POTWs and may pass through the treatment plant untreated. This âpass throughâ of pollutants impacts the surrounding envi- ronment, occasionally causing fish kills or other detrimental alterations of the receiving waters. Even when POTWs have the capability to remove toxic pollutants from wastewa- ter, these toxics can end up in the POTWâs sewage sludge, which in many places is land- applied to food crops, parks, or golf courses as fertilizer or soil conditioner. The National Pretreatment Program is unique in that the general pretreatment regula- tions require all large POTWs (i.e., those designed to treat flows of more than 5 MGD) and smaller POTWs with significant industrial discharges to establish local pretreatment pro- grams. These local programs must enforce all national pretreatment standards (effluent limitations) and requirements, in addition to any more stringent local requirements neces- sary to protect site-specific conditions at the POTW. More than 1,500 POTWs have devel- oped and are implementing local pretreatment programs designed to control discharges from approximately 30,000 significant industrial users. EPA has supported the pretreatment program through development of more than 30 manuals that provide guidance to EPA, states, POTWs, and industry on various pretreat- ment program requirements and policy determinations. Through this guidance, the pre- treatment program has maintained national consistency in interpretation of the regulations. The general pretreatment regulations establish responsibilities of federal, state, and local government, industry, and the public to implement pretreatment standards to control pollutants that pass through or interfere with POTW treatment processes or that may con- taminate sewage sludge. The general pretreatment regulations apply to all non-domestic sources that introduce pollutants into a POTW. These sources of âindirect dischargeâ are more commonly referred to as industrial users (IUs). Since IUs can be as simple as an unmanned coin-operated car wash to as complex as an automobile manufacturing plant or a synthetic organic chemical producer, EPA developed four criteria that define a significant industrial user (SIU). Many of the general pretreatment regulations apply to SIUs as op- posed to IUs, based on the fact that control of SIUs should provide adequate protection of the POTW. Unlike other environmental programs that rely on federal or state governments to im- plement and enforce specific requirements, the Pretreatment Program places the majority of the responsibility on local municipalities. Specifically, Section 403.8(a) of the general pretreatment regulations states that any POTW (or combination of treatment plants oper- ated by the same authority) with a total design flow greater than 5 million MGD and smaller POTWs with SIUs must establish a local pretreatment program. As of early 1998, 1,578 POTWs were required to have local programs. Although this represents only about 15 percent of the total treatment plants nationwide, these POTWs account for more than 80 percent (i.e., approximately 30 billion gallons a day) of the national wastewater flow. Consistent with Section 403.8(f), POTW pretreatment programs must contain the six minimum elements described below (EPA, 1999):
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 529 1. Legal Authority The POTW must operate pursuant to legal authority enforceable in federal, state, or local courts, which authorizes or enables the POTW to apply and enforce any pretreatment regulations developed pursuant to the CWA. At a minimum, the legal authority must enable the POTW to: i. deny or condition discharges to the POTW, ii. require compliance with pretreatment standards and requirements, iii. control IU discharges through permits, orders, or similar means, iv. require IU compliance schedules when necessary to meet applicable pretreatment standards and/or requirements and the submission of reports to demonstrate compliance, v. inspect and monitor IUs, vi. obtain remedies for IU noncompliance, and vii. comply with confidentiality requirements. 2. Procedures The POTW must develop and implement procedures to ensure compliance with pre- treatment requirements, including: i. identify and locate IUs subject to the pretreatment program, ii. identify the character and volume of pollutants contributed by such users, iii. notify users of applicable pretreatment standards and requirements, iv. receive and analyze reports from IUs, v. sample and analyze IU discharges and evaluate the need for IU slug control plans, vi. investigate instances of noncompliance, and vii. comply with public participation requirements. 3. Funding The POTW must have sufficient resources and qualified personnel to carry out the au- thorities and procedures specified in its approved pretreatment programs. 4. Local Limits The POTW must develop local limits or document why those limits are not necessary. 5. Enforcement Response Plan (ERP) The POTW must develop and implement an ERP that contains detailed procedures indicating how the POTW will investigate and respond to instances of IU noncompliance. 6. List of SIUs The POTW must prepare, update, and submit to the approval authority a list of all sig- nificant industrial users (SIUs). In addition to the six specific elements, pretreatment program submissions must in- clude: A statement from the city solicitor (or the like) declaring the POTW has adequate authority to carry out program requirements; Copies of statutes, ordinances, regulations, agreements, or other authorities the POTW relies upon to administer the pretreatment program, including a statement reflecting the endorsement or approval of the bodies responsible for supervising and/or funding the program; continues next page
530 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-5 Continued A brief description and organizational chart of the organization administering the program; and A description of funding levels and manpower available to implement the program. The objectives of the National Pretreatment Program are achieved by applying and enforc- ing three types of discharge standards: (1) prohibited discharge standards, (2) categorical standards, and (3) local limits. Prohibited Discharge Standards All IUs, whether or not subject to any other national, state, or local pretreatment re- quirements, are subject to the general and specific prohibitions identified in 40 C.F.R. Â§Â§403.5(a) and (b), respectively. General prohibitions forbid the discharge of any pollut- ant(s) to a POTW that cause pass-through or interference. These prohibited discharge standards are intended to provide general protection for POTWs. Examples of these in- clude prohibitions on discharges of pollutants that can create fire or explosion hazards, cause corrosive structural damage, obstruct flow within the POTW, and interfere with the POTWâs biological treatment activity. However, their lack of specific pollutant limitations creates the need for additional controls, namely categorical pretreatment standards and local limits. Categorical Standards Categorical pretreatment standards (i.e., categorical standards) are national, uniform, technology-based standards that apply to discharges to POTWs from specific industrial categories (i.e., indirect dischargers) and limit the discharge of specific pollutants. Cate- gorical pretreatment standards for both existing and new sources are promulgated by EPA pursuant to Section 307(b) and (c) of the CWA. Limitations developed for indirect dis- charges are designed to prevent the discharge of pollutants that could pass through, inter- fere with, or otherwise be incompatible with POTW operations. The categorical pretreat- ment standards can be concentration based or mass based. For example, the pretreat- ment standard for the electrical and electronic component manufacturing industry (40 C.F.R. Part 469, Subparts A-D) are concentration-based daily maximum and monthly aver- age limits that vary by subpart and pollutant parameter. Local Limits Prohibited discharge standards are designed to protect against pass-through and in- terference generally. Categorical pretreatment standards, on the other hand, are designed to ensure that IUs implement technology-based controls to limit the discharge of pollutants. Local limits, however, address the specific needs and concerns of a POTW and its receiv- ing waters. Federal regulations at 40 CFR Â§Â§403.8(f)(4) and 122.21(j)(4) require control authorities to evaluate the need for local limits and, if necessary, implement and enforce specific limits as part of pretreatment program activities. Local limits are developed for pollutants (e.g., metals, cyanide, BOD5, TSS, oil and grease, organics) that may cause interference, pass-through, sludge contamination, and/or worker health and safety prob- lems if discharged in excess of the receiving POTW treatment plantâs capabilities and/or receiving water quality standards.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 531 events). If put in this position, municipalities will make intelligent choices and adopt effective strategies to identify which industries and sources to focus upon. Each of these issues is discussed in greater detail below. Determination of High-Risk Dischargers At present, the federal stormwater regulations do not specifically identify which sources would be considered high risk given the common pollutants in MS4 stormwater discharges. With the exception of the category of municipal landfills and hazardous waste treatment, storage, and disposal facilities, it does not even state that the other nine categories of industry singled out in the regula- tions for permitting under the multi-sector industrial stormwater general permit (MSGP) are really high risk. The devolution of this responsibility to the mu- nicipality is sensible because the municipality, as the land-use authority, already conducts development review and issues industrial conditional-use permits. The permitting authority would still be responsible for inspecting high-risk state, federal, and other facilities over which the MS4 permittee has no jurisdiction. In addition, the permitting authority would inspect municipal facilities such as air- ports, ports, landfills, and waste storage facilities to avoid the situation of self- inspection. Methods for ranking industries according to risk are discussed in a subsequent section. It is likely that some of the designated high-risk facilities would be better regulated by individual stormwater NPDES permits. In particular, good candi- dates for individual NPDES permits include international ports, airports, and multiphase construction land developments, which are similar (in the potential risk they pose to water quality) to traditional major wastewater facilities such as petroleum refineries and large POTWs. SCM Design Parameters, Numerical SCM Performance Criteria, and Monitoring For the integration approach to work, the permitting authority and the MS4 permittee must better delineate SCM design parameters, numerical performance criteria, and default SCMs based on best available technology or water quality standards for the discharge of industrial and construction stormwater. Both the ASCE International Storm Water Database (which is now called the WERF In- ternational Storm Water Database because it is maintained by the Water Envi- ronment Research Foundation) and the National Stormwater Quality Database (NSQD), which were developed with EPA funding, are comprehensive datasets that can be used to develop numeric technology-based effluent criteria or limits for industrial and construction stormwater discharges. The MS4 can then de- termine the compliance of industry and construction activity with its require- ments by using either some numeric criteria or a suite of SCMs that have been
532 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES presumptively determined as capable of achieving the performance criteria. The EPA MSGP includes a general list of sector-specific SCMs, but these presently have no performance criteria associated with them. It is important that the EPA continue to support both the WERF and the NSQD databases as the repositories of SCM performance and MS4 monitoring data, so that MS4s can use them to establish local limits and update the performance criteria periodically to fully effectuate the iterative approach to ensuring that MS4 discharges eventually will meet water quality standards. The proposed integration scheme will also facilitate the MS4 permitteeâs implementation of a purpose-oriented stormwater monitoring program directed toward identifying problematic industrial or construction stormwater discharges or high-risk industrial facility sectors. The current benchmark monitoring con- ducted by MSGP facilities would be eliminated. Instead, MSGP facilities would have the option of performing scientifically valid stormwater discharge sam- pling to demonstrate their compliance with performance criteria or to participate in an MS4-led monitoring program by paying in lieu fees to support the cost of the purpose-oriented MS4 monitoring program. The net effect of this alternative is to pool the resources to come up with an optimal sampling strategy to replace what is now a stormwater monitoring strategy that is haphazard and not useful. MS4 Responsibilities Under integration, the MS4 permittee would be primarily responsible for the quality of stormwater discharges that exit the MS4 to the waters of the United States. The MS4 permittee would not be responsible for stormwater dis- charges from federal and state facilities or for facilities that have been issued an individual NPDES permit for stormwater discharges. The MS4 permittee would be responsible for implementing the six minimum program measures, assisting in the oversight and inspection of facilities covered under the MSGP and the construction general permit (CGP), and implementing a strategic water quality monitoring program to identify and control pollutant discharges from high-risk sites. The permitting authority would share any fees collected under the MSGP and CGP with the MS4, and facilities covered by them would have the option to opt-out of self-monitoring and contribute equivalent funds to an MS4-led moni- toring program. Similarly, the permitting authority would be expected to sup- port research and special studies that address issues of regional or national sig- nificance through partnerships with the MS4 permittees. Some MS4s may balk at taking on more responsibility for the control of stormwater pollution, as required for integration to succeed. However, there are already several case examples that exist. The State of Oregon requires facilities that discharge industrial stormwater to file a Notice of Intent (NOI) for coverage under the MSGP with both the state and the local MS4 (Campbell, 2007). The state has an agreement with the local MS4s for the inspection of the facilities covered under the MSGP and the sharing of NOI fees. The State of Tennessee
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 533 has a statewide pilot program to partner with local MS4s for the inspection of construction sites that are covered under the CGP. Analogy to the WWTP Pretreatment Program It is certainly true that the MS4s are a more challenging point source to regulate for the discharge of pollutants than WWTPs. WWTPs have fewer out- falls discharging to waters of the United States than MS4s, and inputs into them are through discrete rather than diffuse sources as in the case of MS4s. It is thus expected to be more difficult to identify problem stormwater sources and to hold them accountable for discharges in excess of standards. This problem is not insurmountable, however. Watershed and land-use hydrologic models can be developed and refined by strategic sampling of pollutant sources for use by MS4 permittees and regulatory agencies. If EPA and state permitting authorities es- tablish measurable outcomes as expected endpoints of progress, MS4 permittees will make intelligent choices about which measures to implement in order to meet these endpoints. In large part, the lack of progress nationally towards con- trolling pollutants in stormwater discharges from the MS4s has been due to the absence of national SCM design standards, MS4 discharge performance criteria, and stormwater effluent guidelines. Presently, the MS4 permittees as owners and operators of the MS4 affirmatively approve connections to the conveyance system for rainfall runoff. Historically the issuance of the MS4 connection per- mit has been based on the sizing of the pipes for the conveyance of flood waters. There are few barriers to including water quality considerations in reauthorizing these connections and adding new ones. Note that EPA did initially consider using the WWTP pretreatment ap- proach for stormwater discharges by requiring MS4 permittees to be primarily responsible for discharges of stormwater associated with industrial activity through the MS4 (53 Fed. Reg. 49428; December 7, 1988). However, EPA de- viated from this approach in issuing its Final Storm Water Rule (55 Fed. Reg. 48006; November 16, 1990). In the absence of regulations that specifically con- fer authority on MS4 permittees to establish local limits for stormwater dis- charges to the MS4 from industry and businesses, the EPA should promulgate specific SCMs and performance guidelines with rigorous requirements for self- monitoring and compliance in order to support the integrated framework for controlling stormwater pollution from MS4s. Potential Legal Barriers A revised stormwater program that requires MS4s to play a more significant role in enforcement and oversight and that provides greater specificity in permit requirements is not only contemplated, but arguably demanded by Congress in the CWA. Specifically, Congress directs that MS4 permits be conditioned on
534 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES the requirement that the MS4s âshall require controls to reduce the discharge of pollutants to the maximum extent practicableâ 42 U.S.C. Â§ 1342(p)(3)(B)(iii). EPA has already conditioned Phase I MS4 permits on the requirement that the municipality establish that it has the legal authority to inspect discharges into the system and take regulatory and enforcement action against excessive or violat- ing sources [40 C.F.R. Â§ 122.26(d)(2)(i)]. Nevertheless, to ensure that MS4s play an even more active role, EPA should include several additional require- ments in its implementing regulations. In addition to promulgating more de- tailed and specific SCM requirements as discussed above, EPA should also re- quire that the Phase I MS4s establish that they possess sufficient funding and staff to effectuate their responsibilities [see, e.g., 40 C.F.R. Â§ 403.8(f)(2) and (3) requiring this showing for the POTW program]. Like the POTW program, states should also be authorized as MS4 permittees when the local governments are unable or unwilling to carry out their mandatory stormwater permit respon- sibilities [see, e.g., 40 C.F.R. Â§ 403.10(e) providing this authority for the POTW program]. Industrial Program The industrial stormwater permit program presently incorporates a menu of SCMs that are to be selected by the facility operator, a rudimentary monitoring program that includes visual observations, some water quality sampling for se- lected parameters for certain types of industries subject to numerical effluent limitations (see Table 2-6) or a set of pollutant-level benchmarks that are to be used as a measure to appropriately revise the SWPPP (see Table 2-5), and an- nual reporting. Neither SCM performance criteria nor the characteristics of a design storm for water quality purposes have been established. Given the broad discretion that facility operators enjoy as a result, it has been difficult to gauge compliance with the MSGP and initiate enforcement for non-compliance even though industrial stormwater discharges are required to meet effluent limitations (technology- or water quality-based) that reflect water quality standards (Duke and Beswick, 1997; Duke and Augustenborg, 2006; Wagner, 2006). Several ideas to address some of the shortcomings in the implementation of the permit- ting program for industrial stormwater discharges are offered as additions to the concept of MS4 regulatory integration discussed previously. They would sub- stantively improve the current industrial stormwater permitting program even if the integration recommendations were not acted upon. Criteria for a Water Quality Design Storm and Subsequent SCM Selection To improve the quality of stormwater discharges from industry, provide for better accountability, and advance the objectives of the CWA, it is important
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 535 first to identify the criteria for a water quality design storm as opposed to one for flood control design, where the objective is to protect human life and real prop- erty. It is important that the permitting authority designate the basis for the de- termination of the water quality design storm, and explicitly state that it would form the criteria for evaluation of compliance with technology-based standards or water quality-based standards. This is essential because the engineering de- sign decisions that determine how much stormwater is to be treated to remove toxic pollutants that pose a risk to human health or aquatic life is more a policy matter than a scientific one (Schiff et al., 2007). While modeling exercises us- ing continuous simulation methods in theory could be performed for every pro- ject or subwatershed or region to support planning decisions on how much stormwater needs to be treated for optimum water quality benefits, such a de- tailed analysis will be too cumbersome and cost-prohibitive for routine planning and implementation purposes. Thus it is recommended that the EPA establish guidelines for the selection of water quality design storms for controlling pollu- tion from MS4 and industrial stormwater discharges. This would not be a new practice for EPA because the agency has previously established design storms for certain industrial sectors when promulgating effluent guidelines (Table 2-6). Conceivably, unlike the technology limiting design storms that are set on rainfall recurrence intervals, the design storm to protect surface water quality and bene- ficial uses could be different for different eco-regions of the United States. The water quality design storm, which may be expressed as total rainfall depth, runoff volume, or rainfall intensity, incorporates the concept that extreme rainfall events are rare, and that a few times each year the runoff volume or flow rate from a storm will exceed the design volume or rate capacity of an SCM. Therefore, for the purpose of best available technology and cost-effectiveness, industrial facility operators should not be held accountable for pollutant removal from storms beyond the size for which an SCM is designed. For MS4 operators, the concept of designing MS4s for both flood control conveyance (capital flood design) and for water quality protection (water quality design) involves a fundamental shift. Whereas flood control engineers design conveyance systems with return frequencies of two years (streets), ten years (detention basins), 50 years, and 100 years (channels), the water quality design storm event is for a return frequency of six months to a year. The water quality design implicitly focuses on treating the first flush of runoff, which contains the highest load and concentration of pollutants and which occurs in the first half to one inch of runoff. In contrast, flood control designs are built to convey tens of inches of runoff. In addition to issuing the guidelines to support the setting of stormwater cri- teria for water quality design, it is important that the EPA establish SCM per- formance criteria based on best technologies and identify the âpresumptive tech- nologiesâ that have been demonstrated to achieve the performance criteria. The water quality design storm and the best available technologies with their associ- ated criteria can then form a basis for technology-based effluent limitations to be included in industrial stormwater permits. If the facility operator elects the iden-
536 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES tified presumptive technology, then compliance monitoring requirements can be scaled down to a minimum to ensure that the treatment systems are being prop- erly maintained. On the other hand, if the operator elects to go with a suite of alternative SCMs, then the monitoring requirements sufficient to demonstrate that the suite of alternative SCMs are in fact achieving the effluent quality of the selected technology can be prescribed. In such a scheme, visual monitoring will serve to ensure that the treatment systems are being properly maintained, and compliance can be reported using the same procedures as required presently for the industrial wastewater permits. How to Identify a High-Risk Industry Both the watershed-based permitting approach described previously in this chapter and the integration approach call for municipal permittees, as part of their responsibilities, to identify high-risk industrial stormwater dischargers. This involves identifying the potential sources of concern, evaluating the extent of their potential impacts, and then prioritizing them for attentionâa classic risk assessment. Municipalities would generally not be able to give equal and full attention to all sources, nor should they. Unfortunately, what constitutes high risk or any level of risk for industries covered by NPDES stormwater permits has not been defined by EPA, although the states have developed various inter- pretations (see Appendix C). Two methodologies for identifying industrial and commercial facilities that are considered high-risk for discharging pollutants in stormwater are presented below. Box 6-6 describes the âintensity of industrial activityâ method devised for the City of Jacksonville (Duke, 2007). This method uses telephone queries and a point scale system to visually score each facility based on the intensity of the industrial activities exposed to stormwater, and groups the results into cate- gories A, B, C, or D in increasing order of intensity (Cross and Duke, 2008). The categories are designed to distinguish high-risk facilities from low-risk fa- cilities, and not to make fine distinctions among facilities with similar character- istics. This typology is sufficient to distinguish facilities with little or no poten- tial for discharging pollutants associated with stormwater from facilities that might discharge those pollutants. More than half of the facilities that were sub- ject to Floridaâs MSGP were determined to be low-risk (Cross and Duke, 2008). Box 6-7 outlines an empirical methodology used by the County of Los An- geles to rank the risk of industrial facilities for stormwater pollution on the basis of pollution potential P. The pollution potential P was computed as a product of the number of on-site sources, percent imperviousness, pollutant toxicity, degree of exposure, and the number of facilities (Los Angeles County, 2001). Based on this ranking scheme, five top high-risk industries were selected: (1) automobile dismantlers, (2) automobile repair, (3) metal fabrication, (4) motor freight, and (5) automobile dealers. Stormwater discharges from six facilities in each cate- gory were characterized over a two-year period, and the effectiveness of SCMs
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 537 BOX 6-6 Risk Assessment for Industrial Dischargers of Stormwater The City of Jacksonville has had very good success in determining what industries pose the highest stormwater risks by starting with businesses having the Standard Indus- trial Classification (SIC) codes designated for permit coverage but using multiple lists of potential sources and cross checking them to target inspections and other interventions where they will have the best effect. Other clues to sources of interest include other envi- ronmental permits (e.g., wastewater NPDES permits, permits for discharge to sanitary sewer), tax records, records of fire code inspections, building permit filings, planning agency proceedings, contacts with business associations, marketing information put out by companies, Resource Conservation and Recovery Act hazardous waste reports, and tele- phone and field surveys. Duke (2007) proposed a 0- to 8-point scoring scheme (shown below) to rate the inten- sity of industrial activities exposed to stormwater. The system is based on the relative amount of exposure to precipitation and runoff by industrial materials, processes, wastes, and vehicles. Once municipalities gather the data and then classify their industries accord- ingly, they would have a very useful tool to program inspections and monitoring emphasiz- ing the industries most risking their success in achieving established objectives. A similar system could and should be developed for construction sites. 0 points 2 Small bulk waste, e.g., covered dumpster: area <100 m Hazardous waste: containers not exposed to precipitation 1 point Outdoor vehicle use: 1-2 vehicles, outdoors occasionally/never, not used in precipitation Vehicle washing outdoors, 1-2 vehicles, rarely or occasionally done 2 points Outdoor vehicles, e.g., forklifts: 1-2, outdoors occasionally/never, used in precipitation Outdoor vehicles, e.g., forklifts: 1-2, outdoors every day, not used in precipitation Outdoor vehicles, e.g., forklifts: 3-4, outdoors occasionally/never, not used in precipita- tion Vehicle maintenance or re-fueling, 1-2 vehicles, rarely or occasionally done, outside Vehicle washing outdoors, 1-2 vehicles, regularly done Vehicles washing outdoors, 3 vehicles, rarely or occasionally done 4 points 2 Storage of materials or products: area < 100m and/or < five 55-gallon drums Fixed outdoor equipment: 1-2 small or large item(s) Outdoor vehicles, e.g., forklifts: 1-2, outdoors every day, used in precipitation Outdoor vehicles, e.g., forklifts: 3-4, outdoors occasionally/never, used in precipitation Outdoor vehicles, e.g., forklifts: 3-4, outdoors every day, not used in precipitation Uncovered shipping/receiving area: 1-2 docks Vehicle maintenance or re-fueling outdoors, 1-2 vehicles, regularly done Vehicle maintenance or re-fueling outdoors, vehicles, rarely or occasionally done 2 Plant yard, rail lines, access roads: 1,000 ft Small process equipment, e.g., compressors, generators: exposed to precipitation continues next page
538 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES BOX 6-6 Continued 6 points Outdoor vehicles, e.g., forklifts: 3-4, outdoors every day, used in precipitation Outdoor vehicles, e.g., forklifts: > 5 or heavy, outdoors occasionally, used in precipitation Outdoor vehicles, e.g., forklifts: > 5 or heavy, outdoors every day, not used in precipita- tion Vehicle maintenance or re-fueling outdoors, 3 vehicles, regularly done 2 Plant yard, rail lines, access roads: 1,000 ft 8 points 2 Storage of materials or products: area 100 and/or five 55-gallon drums Boneyard of scrap, disused equipment, similar Hazardous waste: containers exposed to precipitation Fixed outdoor equipment: small or 2 large items Outdoor vehicles, e.g., forklifts: > 5 or heavy, outdoors every day, used in precipitation Uncovered shipping/receiving area: 3 docks 2 Plant yard, rail lines, access roads: 5,000 ft Manufacturing activities, e.g., cutting, painting, coating materials: exposed to precipita- tion SOURCE: Duke (2007). was assessed at a subset of them. However, the monitoring was minimal, and so much of the prioritization was based on best professional judgment about pollut- ant discharges. Industrial Stormwater Discharge Monitoring Monitoring data from Phase I MS4s have been compiled in the NSQD for several years, making possible a number of important findings about the quality of municipal stormwater (see Chapter 3). Although industry that occurs within MS4s is technically included in the NSQD, the data are lumped together and not sector specific. There is no comparable, reliable source of data specifically on industrial discharges, even though EPA requires benchmark monitoring for MSGP industrial permittees. The intent was that industrial facility operators would use benchmark exceedances as action levels to improve SCMs, but this self-directed approach has been largely a failure. Many industrial facilities re- ported repeated exceedances of benchmark values without action, and others have failed to report any monitoring data at all. In addition, the representative- ness of single grab samples taken to characterize the discharge and less-than- rigorous sample collection and quality assurance procedures have resulted in monitoring data that are not very useful. One of the only analyses of benchmark monitoring data ever done evaluated Californiaâs program between 1992 and 2001 (see Box 4-2; Stenstrom and Lee, 2005; Lee et al., 2007). The study showed no relationship between facility type and stormwater discharge quality. The cited reasons for the poor relationship included variability in sampling pa- rameters, sampling time, and sampling strategyâthat is, poor data.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 539 BOX 6-7 Los Angeles County Critical Facilities Monitoring Data One of the few sources of data on industrial stormwater discharges comes from the County of Los Angeles. A stepwise process was used to identify the highest-risk indus- trial/commercial facilities, which were then monitored to measure the quality of their storm- water discharges and to evaluate the effectiveness of SCMs. The initial list of candidate facilities was identified from their relative numbers and the extent of their outdoor activities. This list was then refined using an empirical equation for pollutant potential P: P=QxRxTxExN where Loading (Q) is the number of sources at a site and the likelihood of release; Imperviousness (R) of a site is the percent of paved area; Pollutant toxicity (T) denotes the number of toxic pollutants and the inherent toxicity of the mix; An exposure factor (E) signifies if activities are exposed to rainfall; and The Number (N) represents the total number of sites in the county. Each variable was assigned a qualitative number from 1 to 10, with 10 representing the worst condition. Based on this equation, five top âcritical sourceâ industries were determined: (1) auto- mobile dismantlers; (2) automobile repair; (3) metal fabrication; (4) motor freight; and (5) automobile dealers. Six facilities from each of these categories were monitored during five storms a year for two years. The stormwater discharge samples were analyzed for general conventional pollutants, heavy metals, bacteria, and semi-volatile organic compounds. Half of the facilities were then fitted with SCMs, which were monitored to evaluate their effec- tiveness. The highest median values were observed for total zinc (approx. 450 Âµg/L), dissolved zinc (approx. 360 Âµg/L), total copper (approx. 240 Âµg/L), and dissolved copper (approx. 110 Âµg/L) in stormwater discharges from fabricated metal sites. However, levels for total and dissolved zinc did not appear to be significantly different among the industry types. SCMs in the form of good housekeeping and spill containment measures were installed at half of the sites. For total and dissolved zinc, the median concentration lowered or stayed nearly the same with the implementation of SCMs at the auto dismantling, auto repair, and fabri- cated metals industries (i.e., in none of the circumstances was the difference significant). For total and dissolved copper, however, where the fabricated metal industry had displayed the highest median concentrations, levels were significantly reduced with the implementa- tion of SCMs. The auto dismantling and auto repair businesses showed no significant dif- ferences in copper after the implementation of SCMs. SOURCE: Los Angeles County (2001).
540 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES In the past, it has been proposed to EPA that it fund a project that would systematically collect the benchmark monitoring data across the nation, as has been done for MS4s, but these suggestions have been rejected. To get better data from specific industrial sectors, it is recommended that a small subset of industrial users and sectors be selected for composite sampling in a program directed by the MS4. Alternatively, making a trained team responsible for monitoring of small-business industrial dischargers would reduce, if not elimi- nate, current problems with quality assurance. Monitoring of industrial stormwater discharges could be streamlined by considering the adoption of a Reasonable Potential Analysis (RPA), which is already part of the existing practice in developing limits for NPDES wastewater permits (EPA, 1991). The RPA is a procedure that uses statistical distribution assumptions in association with a limited number of wastewater discharge qual- ity measurements to determine the likelihood that a receiving water quality stan- dard would be violated, which assists the permitting authority in determining what permit limitations should be set to protect receiving water quality. The effluent data from any treatment system may be described using standard de- scriptive statistics such as the mean concentration and the coefficient of varia- tion. Using a statistical distribution such as the lognormal, an entire distribution of values can be projected from limited data; limits on pollutant concentrations in discharge can then be set at a specified probability of occurrence so that the receiving water is protected. An RPA for stormwater pollutants may be particu- larly relevant in developing performance criteria for SCMs for facilities dis- charging stormwater within the integrated framework of MS4 permitting. Also, MS4 permittees could use the method to reduce the number of pollutants that high-risk industries would be required to monitor in order to demonstrate to the municipality that they are not the source of pollutants in MS4 discharges that are impairing surface waters. Construction Program The recommendations for stormwater discharges associated with construc- tion activity are very similar to those offered for stormwater discharges associ- ated with industrial activity. The integration with the MS4 program is less of a challenge because municipalities have always had primacy on land development planning and construction activity. Most municipalities have had requirements for soil erosion and sediment control plans on construction sites that precede the federal stormwater regulations. EPA regulations already allow permitting au- thorities to approve Phase I and Phase II MS4 permittee oversight of CGP con- struction sites under the qualifying local program provision (40 C.F.R. 122.44(s)) (Grumbles, 2006). The weakness in the implementation of this pro- vision currently is the absence of rigorous SCM performance criteria guidelines for MS4s permittees to meet in order to be deemed as qualifying. The construction stormwater general permit program requires the develop-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 541 ment and implementation of an SWPPP. The SWPPP, which must be prepared before construction begins, focuses on two major requirements: (1) describing the site adequately and identifying the sources of pollution to stormwater dis- charges associated with construction activity on site and (2) identifying and im- plementing appropriate measures to reduce pollutants in stormwater discharges to ensure compliance with the terms and conditions of this permit. The SWPPP must describe the sequence of major stormwater control activities and the kinds of SCMs that will be in place, and it must identify interim and permanent stabi- lization practices, including a schedule of their implementation. There is an expectation that the construction site operator will use good site planning, pre- serve mature vegetation, and properly stage major earth-disturbing activities to avoid sediment loss and prevent erosion. Post-construction stormwater controls need to be considered, but are not required. Construction site operators are re- quired to visually inspect the construction site weekly and perform a walk through before predicted storm events. No annual reports are required, but re- cords must be kept for a period of three years after permit coverage has been terminated. There are no SCM performance criteria, other than a suggestion that most SCMs should be able to achieve 80 percent TSS removal. As with indus- try, it is difficult to gauge compliance with the CGP except when inadequate SCMs result in a massive discharge of sediment from a construction site. The pollutant parameters that are of concern in stormwater discharges from construction activity are TSS, settleable solids, turbidity, and nutrients from ero- sion; pH from concrete and stucco; and a wide range of metallic and organic pollutants from construction materials, processes, wastes, and vehicles and other motorized equipment. The permitting authority, in addition to guidelines for the water quality design storm, must establish SCM performance criteria for storm- water discharges associated with construction activity. The construction site operator should be given the option of implementing SCMs that are the pre- sumptive technology, or equivalent SCMs that can achieve the performance cri- teria. For example, the recommended SCMs in Box 5-3 could serve as the pre- sumptive construction SCMs on a typical construction site that is less than 50 acres in size. If the operator elects to go with a suite of alternative SCMs, then adequate monitoring must be performed to demonstrate that the alternative SCMs are in fact achieving the performance criteria. In addition, the CGP pres- ently does not mandate or require that post-construction SCMs be integrated with the MS4 permittee requirements under its New Develop- ment/Redevelopment Program requirements. The proper planning for and im- plementation of SCMs that will help mitigate stormwater pollution from planned future use of the site will be critical to protecting water quality. Thus the post- construction requirements of the CGP should be strengthened and better inte- grated with the new development/redevelopment requirements of the MS4 per- mits.
542 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES Municipal Program Several key enhancements to the MS4 permitting program are needed to en- sure that resources are targeted to achieve the greatest on-the-ground implemen- tation of SCMs to make incremental progress in meeting water quality stan- dards. Six specific issues are discussed below; their implementation will require greater collaboration and flexibility among regulators and permitted parties. These recommendations are suggested for communities that are not ready for the integrated watershed approach proposed in the prior section, and represent a bridge toward building internal capacity to implement them. Numeric Expression of âMaximum Extent Practicableâ The ambiguity of the term âmaximum extent practicableâ (MEP) has been a major impediment to achieving meaningful water quality results in the MS4 program. The EPA should develop numerical expressions of MEP in the next round of permit renewals that can be measured and tracked. A national numeric benchmark should be avoided; states should focus on regional benchmarks that are tied to their water quality problems. Four examples of methods to define MEP in a numeric manner are provided below: the first three are applied at a regional or state level, whereas the last (impervious cover-based TMDLs) offers more flexibility to be applied at individual sites. Establish Municipal Action Levels. This approach relies on the use of a national database of stormwater runoff quality to establish reasonable expecta- tions for outfall monitoring in highly developed watersheds. The NSQD (Pitt et al., 2004) allows users to statistically establish action levels based on regional or national event mean concentrations developed for pollutants of concern. The action level would be set to define unacceptable levels of stormwater quality (e.g., two standard deviations from the median statistic, for simplicity). Munici- palities would then routinely monitor runoff quality from major outfalls. Where an MS4 outfall to surface waters consistently exceeds the action level, munici- palities would need to demonstrate that they have been implementing the stormwater program measures to reduce the discharge of pollutants to the maximum extent practicable. The MS4 permittees can demonstrate the rigor of their efforts by documenting the level of implementation through measures of program effectiveness, failure of which will lead to an inference of noncompli- ance and potential enforcement by the permitting authority. Site-Based Runoff and/or Pollutant Load Limits. This approach is pri- marily used for watersheds that are experiencing rapid development; it estab- lishes numeric targets or performance standards for pollutant or runoff reduction that must be met on individual development sites. The numeric targets may involve specific pollutant load limits or runoff reduction volumes. For example,
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 543 Virginia DCR (2007) and Hirschman et al. (2008) established a statewide com- putational method to ensure that SCMs are sized, designed, and sequenced to comply with specific nutrient-based load and runoff reduction limits. The nutri- ent load limits of 0.28 lb/acre/yr for total phosphorus and 2.68 lb/acre/yr for total nitrogen were computed using the Chesapeake Bay Model for Virginia tributaries to the bay. The design process also requires the computation of run- off reduction volumes achieved to promote the use of nonstructural SCMs. The basic concept is that new development on non-urban land must not exceed the average annual nutrient load and runoff volume for non-urban land using effec- tive SCMs in the watershed. This blended site-based runoff and load limit ap- proach has been advocated by the Office of Inspector General (2007) and Schueler (2008a) and is under active consideration by several other Chesapeake Bay states. Wenger et al. (2008) reports on a no-net-hydrologic-increase strategy to protect endangered fish species in the northern Georgia Piedmont that sets spe- cific on-site runoff reduction requirements for a range of land uses and design storm events. A similar approach has been incorporated into the recently en- acted Energy Independence and Security Act of 2007 that contains provisions that require that the âsponsor of any development or redevelopment project in- volving a Federal facility with a footprint that exceeds 5,000 square feet shall use site planning, design, construction, and maintenance strategies for the prop- erty to maintain or restore, to the maximum extent technically feasible, the pre- development hydrology of the property with regard to the temperature, rate, vol- ume, and duration of flow.â The challenge of defining MEP as a runoff reduction or pollutant load limit is that considerable scientific and engineering analysis is needed to establish the performance standards, evaluate SCM capability to meet them, and devise a workable computational approach that links them together at both the site and watershed levels. In addition, care must be taken to define an appropriate base- line to represent predevelopment conditions that does not unduly penalize rede- velopment projects or make it impossible to comply with limits at new devel- opment sites after maximum effort to apply multiple SCMs is made. Turbidity Limits for Construction Sites. Numeric enforcement criteria can be used to define what constitutes an egregious water quality violation at construction sites and provide a technical criterion to measure the effectiveness of erosion and sediment control practices. Currently, most states and localities do not specify either numeric enforcement criteria or a monitoring requirement within their CGP (see the survey data contained in Appendix C). A maximum turbidity limit would establish definitive criteria as to what constitutes a direct sediment control violation and trigger an assessment for remediation and prevention actions. For example, local erosion and sediment control ordinances could establish a numeric turbidity limit of 75 Nephelometric Turbidity Units (NTU) as an instantaneous maximum for rainfall events less than an inch (or a 25 NTU monthly average) and would prohibit visible sedi-
544 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES ment in water discharged from upland construction sites. While the exact tur- bidity limit would need to be derived on a regional basis to reflect geology, soils, and receiving water sensitivity, research conducted in the Puget Sound of Washington indicates that turbidity limits in the 25 to 75 NTU can be consis- tently achieved at most highway construction sites using current erosion and sediment control technology that is properly maintained (Horner et al., 1990). If turbidity limits are exceeded, a detailed assessment of site conditions and fol- low-up remediation actions would be required. If turbidity limits continue to be exceeded, penalties and enforcement actions would be imposed. Enforcement of turbidity limits could be performed either by state, local, or third party erosion and sediment control inspectors, orâunder appropriate protocols, training, and documentationâby citizens or watershed groups. Impervious Cover Limits and IC-based TMDLs. MS4s that discharge into TMDL watersheds also require more quantitative expression of how MEP will be defined to reduce pollutant loads to meet water quality standards. Maine, Vermont, and Connecticut have recently issued TMDLs that are based on impervious cover rather than individual pollutants of concern (Bellucci, 2007). In such a TMDL, impervious cover is used as a surrogate for increased runoff and pollutant loads as a way to simplify the urban TMDL implementation process. Impervious cover-based TMDLs have been issued for small subwater- sheds that have biological stream impairments associated with stormwater run- off but no specific pollutant listed as causing the impairment (in most cases, these subwatersheds are classified as impacted according to the Impervious Cover Model [ICM]âsee Box 3-10). A specific subwatershed threshold is set for effective impervious cover, which means impervious cover reductions are required through removal of impervious cover, greater stormwater treatment for new development, offsets through stormwater retrofits, or other means. Traditional pollutant-based TMDLs would continue to be appropriate for ânon-supportingâ and âurban drainageâ subwatersheds, although they could be modified to focus compliance monitoring on priority urban source areas or sub- watersheds that produce the greatest pollutant loads. Although EPA (2002) in- dicates that this analysis does not extend to demonstrating that changes will oc- cur in receiving waters, it does outline a rigorous process for evaluating pollut- ant discharges and SCM performance. More recent EPA guidance (2007c) rec- ommends that MS4s conduct a four-step analysis, which is distilled to its es- sence below: Step 1: Estimate loads for pollutant of concern for the watershed. Step 2: Provide a specific list of SCMs that will be applied in the listed wa- tershed. Step 3: Estimate the pollutant removal capability of the individual SCMs applied. Step 4: Compute aggregate watershed pollutant reduction achieved by the MS4.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 545 Although this is not a particularly new interpretation of addressing stormwater loads in watersheds listed as impaired and/or having written TMDLs, it is excep- tionally uncommon for individual MS4s to document the link between their stormwater discharges and water quality standard exceedances, as modified by the system of SCMs that they used to reduce these pollutants. As of 2007, EPA could only document 17 TMDLs that addressed stormwater discharges using this sequential analysis. EPA and states need to provide more specific guidance for MS4s to comply with TMDLs in their permit applications and annual re- ports. Focus MS4 Permit Implementation at the Subwatershed Level Chapter 5 noted the importance of the watershed context for making better local stormwater decisions. This context can be formally incorporated into local MS4 permits by focusing implementation on a subwatershed basis, using the ICM, as described in Box 3-10 and outlined in Table 6-1. When urban streams are classified by the ICM, this basic subwatershed planning process can be used to establish realistic water quality and biodiversity goals for individual classes of subwatersheds, as shown in Table 6-2. As can be seen, goals for water and habi- tat quality become less stringent as impervious cover increases within the sub- watershed. This subwatershed approach provides stormwater managers with more specific, measurable, and attainable implementation strategies than the one-size-fits-all approach that is still enshrined in current wet-weather manage- ment regulations. Some examples of how to customize stormwater strategies for different subwatersheds are described in Table 6-3. This approach enables MS4s to util- ize the full range of watershed planning, engineering, economic, and regulatory tools that can manage the intensity, location, and impact of impervious cover on receiving waters. In addition, the application of multiple tools in a given sub- watershed class helps provide the maximum level of protection or restoration for an individual subwatershed when impervious cover is forecast to increase due to future growth and development. The conceptual management approach shown in Table 6-3 is meant to show how urban stream classification can be used to guide stormwater decisions on a subwatershed basis. The first column of the table lists some key stormwater management issues that lend themselves to a subwatershed approach and are explained in greater detail below. Linkage with Local Land-Use Planning and Zoning. Given the critical relation between land use and the generation of stormwater, communities should ensure that their planning tools (e.g., comprehensive plans, zoning, and water- shed planning) are appropriately aligned with the intended management classifi- cation for each subwatershed. For example, it is reasonable to encourage rede- velopment, infill, and other forms of development intensification within non-
546 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES TABLE 6-1 Components of Subwatershed-Based Stormwater Management 1. Define interim water quality and stormwater goals (i.e., pollutants of concern, biodiversity targets) and the primary stormwater source areas and hotspots that cause them. 2. Delineate subwatersheds within community boundaries. 3. Measure current and future impervious cover within individual subwatersheds. 4. Establish the initial subwatershed management classification using the ICM. 5. Undertake field monitoring to confirm or modify individual subwatershed classifi- cations. 6. Develop specific stormwater strategies within each subwatershed classification that will guide or shape how individual practices and SCMs are generally assembled at each individual site. 7. Undertakes restoration investigations to verify restoration potential in priority subwatersheds. 8. Agree on the specific implementation measures that will be completed within the permit cycle. Evaluate the extent to which each of the six minimum management practices can be applied in each subwatershed to meet municipal objectives. 9. Agree on the maintenance model that will be used to operate or maintain the stormwater infrastructure, assign legal and financial responsibilities to the owners of each element of the system, and develop a tracking and enforcement system to ensure compli- ance. 10. Define the trading or offset system that will be used to achieve objectives else- where in the local watershed objectives in the event that full compliance cannot be achieved due to physical constraints (e.g., indexed fee-in-lieu to finance municipal retrofits). 11. Establish sentinel monitoring stations in subwatersheds to measure progress to- wards goals. 12. Revise subwatershed management plans in the subsequent NPDES permitting cycle based on monitoring data. supporting or urban drainage subwatersheds, whereas down-zoning, site-based IC caps, and other density-limiting planning measures are best applied to sensi- tive subwatersheds. Stormwater Treatment and Runoff Reduction MEP. Subwatershed classification allows managers to define achievable numerical benchmarks to define treatment in terms of the maximum extent practicable. Thus, a greater level of treatment is required for less-developed subwatersheds and a reduced level of treatment is applied for more intensely developed subwatersheds. This is most frequently expressed in terms of a rainfall depth associated with a given design storm. Designers are required to treat and/or reduce runoff for all storm events up to the designated storm event. This flexibility recognizes the greater difficulty and cost involved in providing the same level of treatment in an in-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 547 TABLE 6-2 Expectations for Different Urban Subwatershed Classes Lightly Consistently attain scores for specific indicators for hydrology, Impacted biodiversity, and geomorphology that are comparable to streams Subwater- whose entire subwatersheds are fully protected in a natural state sheds (e.g., national parks). Should provide for healthy reproduction of (1 to 5% IC) trout, salmon, or other keystone fish species. Consistently attain scores for specific stream indicators that are Moderately comparable to the highest 10 percent of streams in a population of Impacted rural watersheds in order to maintain or restore ecological structure, Subwater- function, and diversity of the streams. The âgood to excellentâ indi- sheds cator scores for this category of subwatersheds will be the bench- (6 to 10% IC) mark against which the relative quality of more developed subwater- sheds will be measured. Consistently attain good stream quality indicator scores to en- Heavily sure enough stream function to adequately protect downstream re- Impacted Sub- ceiving waters from degradation. watersheds Function is defined in terms of flood storage, in-stream nutrient (11 to 25% IC) processing, biological corridors, stable stream channels, and other factors. Consistently attain âfair to goodâ stream quality indicator Non- scores. Supporting Subwater- Meet bacteria standards during dry weather and trash limits sheds during wet weather. (26 to 60% IC) Maintain existing stream corridor to allow for safe passage of fish and floodwaters. Maintain âgoodâ water quality conditions in downstream receiv- Urban Drain- ing waters. age Subwater- sheds Consistently attain âfairâ water quality scores during wet weather and âgoodâ water scores during dry weather. (61 to 100% IC) Provide clean âplumbingâ in upland land uses such that dis- charges of sewage and toxics do not occur. Note: the objectives presume some portion of the subwatershed has already been developed, thereby limiting attainment of objectives. If a subwatershed is not yet developed, managers should shift expectations up one category (e.g., urban drainage should behave like non-supporting). Also, the specific ranges of IC that define each management category should always be derived from local or regional monitoring data. Note that the ranges in IC shown to define a subwatershed man- agement category are illustrative and will vary regionally. tensely developed subwatershed, as well as the fact that less treatment is needed to maintain stream condition in a highly urban subwatershed. The other key element of defining MEP is to specify how much of the treatment volume must be achieved through runoff reduction. The runoff reduc- tion volume has emerged as the primary performance benchmark to maintain predevelopment runoff conditions at a site after it is developed. In its simplest terms, this means achieving the same predevelopment runoff coefficient for each storm up to a defined storm event through a combination of canopy interception, soil infiltration, evaporation, rainfall harvesting, engineered infiltration, ex- tended filtration, or evapotranspiration (Schueler, 2008b). Once again, the physical feasibility and need to provide treatment through runoff reduction be- comes progressively harder as subwatershed impervious cover increases.
548 TABLE 6-3 Examples of Customizing Stormwater Strategies on a Subwatershed Basis URBAN STORMWATER MANAGEMENT IN THE UNITED STATES
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 549 Site-Based IC Fees. Several economic strategies can be used to promote equity and efficiency when it comes to managing stormwater in different kinds of subwatersheds. In lower-density subwatersheds, an excess impervious cover fee can be charged to individual sites that exceed a maximum threshold for im- pervious cover for their zoning category. Similarly, an impervious cover mitiga- tion fee can be levied at individual development sites in more intensely devel- oped subwatersheds when on-site compliance is not possible or it is more cost- effective to provide an equivalent amount of treatment elsewhere in the water- shed. The type of fee and the frequency that is used is expected to be closely related to the subwatershed classification. Subwatershed Trading. The degree of impervious cover in a subwater- shed also has a strong influence on the feasibility, cost, and appropriateness of restoration projects. Consequently, any revenues collected from various site IC fees can be traded among subwatersheds to arrive at the least-cost, effective so- lutions. In general, the most intensely developed subwatersheds are sending areas and the more lightly developed subwatersheds are used as receiving areas for such projects. Stormwater Monitoring Approach. Subwatershed classification can also be used to define the type and objectives for stormwater monitoring to track compliance over time. For example, in sensitive subwatersheds, it may be ad- visable to routinely measure in-stream metrics of biological integrity to ensure stream quality is being maintained or enhanced. As impervious cover increases, stormwater managers may want to shift toward tracking of subwatershed imper- vious cover and actual performance monitoring of select SCMs to establish their effectiveness (e.g., impacted subwatersheds). At even higher levels of impervi- ous cover, streams are transformed into urban drainage, and monitoring becomes more focused on identifying individual stormwater outfalls with the worst qual- ity during storm conditions. TMDL Approach. Subwatershed classification may also serve as a useful tool to decide how to apply TMDLs to impaired waters, or how to ensure that healthy waters are not degraded by future land development. For example, most lightly developed subwatersheds will seldom be subject to a TMDL, or if so, urban stormwater is often only a minor component in the final waste load alloca- tion. Antidegradation provisions of the CWA are often the best means to protect the quality of these healthy waters before they are degraded by future land de- velopment. By contrast, impaired watersheds appear to be the best candidates to apply impervious cover-based TMDLs, as described earlier in this section. As subwatershed impervious cover increases, more traditional pollutant-based TMDLs are warranted, with a focus on problem subwatersheds for non- supporting streams and priority source areas for urban drainage. Dry Weather Water Quality. The type, severity, and sources of illicit dis-
550 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES charges often differ among different subwatershed classifications, which can have a strong influence on the kind of dry weather detective work needed to isolate them. For example, in lightly developed subwatersheds, failing septic systems are often the most illicit discharges, which prompts assessments at the lot or ditch level. The storm-drain network and potential discharge source areas becomes progressively more complex as subwatershed impervious cover in- creases. Consequently, illicit-discharge assessments shift toward outfall screen- ing, catchment analysis, and individual source analysis. Addressing Existing Development. The need for, type of, and feasibility for restoration efforts shift as subwatershed impervious cover increases. In gen- eral, lightly developed watersheds have the greatest land area available for retro- fits and restoration projects in the stream corridor. Consequently, unique resto- ration strategies are developed for different subwatershed classifications (Schueler, 2004). Require More Quantitative Evaluation of MS4 Programs The next round of permit renewals should contain explicit conditions to de- fine and measure outcomes from the six minimum management measures that constitute a Phase II MS4 program. Measurable program evaluation is critical to develop, implement, and adapt effective local stormwater programs, and has been consistently requested in permits and application guidance. To date, how- ever, only a small fraction of MS4 communities have provided measurable out- comes with regard to aggregate pollutant reduction achieved by their municipal stormwater programs. CASQA (2007) defines a six-level pyramid to assess program effectiveness, beginning with documenting activities, raising awareness, changing behaviors, reducing loads from sources, improving runoff quality, and ultimately leading to protection of receiving water quality (see Figure 6-1). At the current time, most MS4s are struggling simply to organize or docu- ment their program activities (i.e., the first level), and few have moved up the pyramid to provide a quantitative link between program activities and water quality improvements. The framework and methods to evaluate program effec- tiveness for each of the six minimum management measures has been outlined by CASQA (2007). Regulators are encouraged to work with permitted munici- palities to define increasingly more specific quantitative measures of program performance in each succeeding permit cycle.
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 551 Assessment Outcome Levels Level 6 â Increasing Protecting Receiv- Difficulty ing Water Quality Level 5 â Improving Runoff Quality Level 4 â Reducing Loads from Sources Level 3 â Changing Behavior Level 2 â Raising Awareness Level 1 â Documenting Stormwater Program Activities FIGURE 6-1 Pyramid of Assessment Outcome Levels for an MS4. SOURCE: CASQA (2007). Shift Monitoring Requirements to Measure the Performance of Stormwater Control Measures The lack of monitoring requirements in the Phase II stormwater program makes it virtually impossible to measure or track actual pollutant load or runoff volume reductions achieved. While the existing Phase I outfall monitoring re- quirements have improved our understanding of urban stormwater runoff qual- ity, they are also insufficient to link program effort to receiving water quality. It is recommended that both Phase I and II MS4s shift to a more collaborative monitoring effort to link management efforts to receiving water quality, as de- scribed below: If a review of past Phase 1 MS4s stormwater outfall monitoring indi- cates no violations of the Municipal Action Limits, then their current outfall monitoring efforts can be replaced by pooled annual financial contributions to a regional stormwater monitoring collaborative or authority to conduct basic re- search on the performance and longevity of range of SCMs employed in the community. If some subwatersheds exceed Municipal Action Levels, outfall moni- toring should be continued at these locations, as well as additional source area sampling in the problem subwatershed to define the sources of the stormwater
552 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES pollutant of concern. Phase II MS4s should be encouraged to make incremental financial contributions to a state or regional stormwater monitoring research collaborative to conduct basic research on SCM performance and longevity. Although the committee knows of no examples where this has been accomplished, this pool- ing of financial resources by multiple MS4s should produce more useful scien- tific data to support municipal programs than could be produced by individual MS4s alone. Phase II communities that do not participate in the research col- laborative would be required to perform their own outfall and/or SCM perform- ance monitoring, at the discretion of the state or federal permitting authority. All MS4s should be required to indicate in their annual reports and permit renewal applications how they incorporated research findings into their existing stormwater programs, ordinances, and design manuals. CONCLUSIONS AND RECOMMENDATIONS The watershed-based permitting program outlined in the first part of this chapter is ultimately essential if the nation is to be successful in arresting aquatic resource depletion stemming from sources dispersed across the landscape. Smaller-scale changes to the EPA stormwater program are also possible. These include integration of industrial and construction permittees into municipal per- mits (âintegrationâ), as well as a number of individual changes to the current industrial, construction, and municipal programs. Improvements to the stormwater permitting program can be made in a tiered manner. Thus, individual recommendations specific to advancing one part of the municipal, industrial, or construction stormwater programs could be imple- mented immediately and with limited additional funds. âIntegrationâ will need additional funding to provide incentives and to establish partnerships between municipal permittees and their associated industries. Finally, the watershed- based permitting approach will likely take up to ten years to implement. The following conclusions and recommendations about these options are made: The greatest improvement to the EPAâs Stormwater Program would be to convert the current piecemeal system into a watershed-based permitting system. The proposed system would encompass coordinated regulation and management of all discharges (wastewater, stormwater, and other diffuse sources), existing and anticipated from future growth, having the potential to modify the hydrology and water quality of the watershedâs receiving waters. The committee proposes centralizing responsibility and authority for im- plementation of watershed-based permits with a municipal lead permittee work- ing in partnership with other municipalities in the watershed as co-permittees,
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 553 with enhanced authority and funding commensurate with increased responsibil- ity. Permitting authorities would adopt a minimum goal in every watershed to avoid any further loss or degradation of designated beneficial uses in the water- shedâs component waterbodies and additional goals in some cases aimed at re- covering lost beneficial uses. The framework envisions the permitting authori- ties and municipal co-permittees working cooperatively to define careful, com- plete, and clear specific objectives aimed at meeting goals. Permittees, with support from the permitting authority, would then move to comprehensive scientific and technically based watershed analysis as a founda- tion for targeting solutions. The most effective solutions are expected to lie in isolating, to the extent possible, receiving waterbodies from exposure to those impact sources. In particular, low-impact design methods, termed Aquatic Re- sources Conservation Design in this report, should be employed to the full ex- tent feasible and backed by conventional SCMs when necessary. This report also outlines a monitoring program structured to assess progress toward meeting objectives and the overlying goals, diagnosing reasons for any lack of progress, and determining compliance by dischargers. The new concept further includes market-based trading of credits among dischargers to achieve overall compli- ance in the most efficient manner and adaptive management to program addi- tional actions if monitoring demonstrates failure to achieve objectives. Integration of the three permitting types, such that construction and industrial sites come under the jurisdiction of their associated municipali- ties, would greatly improve many deficient aspects of the stormwater pro- gram. Federal and state NPDES permitting authorities do not presently have, and can never reasonably expect to have, sufficient personnel to inspect and enforce stormwater regulations on more than 100,000 discrete point source fa- cilities discharging stormwater. A better structure would be one where the NPDES permitting authority empowers the MS4 permittees to act as the first tier of entities exercising control on stormwater discharges to the MS4 to protect water quality. The National Pretreatment Program, EPAâs successful treatment program for municipal and industrial wastewater sources, could serve as a model for integration. Short of adopting watershed-based permitting or integration, a variety of other smaller-scale changes to the EPA stormwater program could be made now, as outlined below. EPA should issue guidance for MS4, MSGP, and CGP permittees on what constitutes a design storm for water quality purposes. Precipitation events occur across a spectrum from small, more frequent storms to larger and more extreme storms, with the latter being a more typical focus of guidance manuals to date. Permittees need guidance from regional EPA offices on what water quality considerations to design SCMs for beyond issues such as safety of human life and property. In creating the guidance there should be a good faith
554 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES effort to integrate water quality requirements with existing stormwater quantity requirements. EPA should issue guidance for MS4 permittees on methods to identify high-risk industrial facilities for program prioritization such as inspections. Two visual methods for establishing rankings that have been field tested are provided in the chapter. Some of these high-risk industrial facilities and con- struction sites may be better covered by individual NPDES stormwater permits rather than the MSGP or the CGP, and if so would fall directly under the permit- ting authority and not be part of MS4 integration. EPA should support the compilation and collection of quality industrial stormwater effluent data and SCM effluent quality data in a national data- base. This database can then serve as a source for the agency to develop tech- nology-based effluent guidelines for stormwater discharges from industrial sec- tors and high-risk facilities. EPA should develop numerical expressions to represent the MS4 stan- dard of Maximum Extent Practicable. This could involve establishing mu- nicipal action levels based on expected outfall pollutant concentrations from the National Stormwater Quality Database, developing site-based runoff and pollut- ant load limits, and setting turbidity limits for construction sites. Such numeri- cal expressions would create improved accountability, bring about consistency, and result in implementation actions that will lead to measurable reductions in stormwater pollutants in MS4 discharges. Communities should use an urban stream classification system, such as a regionally adapted version of the Impervious Cover Model, to establish realistic water quality and biodiversity goals for individual classes of sub- watersheds. The goals for water and habitat quality should become less strin- gent as impervious cover increases within the subwatershed. This should not become an excuse to work less diligently to improve the most degraded water- waysâonly to recognize that equivalent, or even greater, efforts to improve water quality conditions will achieve progressively less ambitious results in more highly urbanized watersheds. This approach would provide stormwater managers with more specific, measurable, and attainable implementation strate- gies than the one-size-fits-all approach that is promoted in current wet weather management regulations. Better monitoring of MS4s to determine outcomes is needed. Only a small fraction of MS4 communities have provided measurable outcomes with regard to aggregate flow and pollutant reduction achieved by their municipal stormwater programs. A framework and methods to evaluate program effec- tiveness for each of the six minimum management measures have been outlined by CASQA (2007) and should be adopted. In addition, the lack of monitoring
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 555 requirements in the Phase II stormwater program makes it virtually impossible to measure or track actual pollutant load or runoff volume reductions achieved. It is recommended that both Phase I and II MS4s shift to a more collaborative monitoring paradigm to link management efforts to receiving water quality. *** Watershed-based permitting will require additional resources and regulatory program support. Such an approach shifts more attention to ambi- ent outcomes as well as expanded permitting coverage. Additional resources for program implementation could come from shifting existing programmatic re- sources. For example, some state permitting resources may be shifted away from existing point source programs toward stormwater permitting. Strategic planning and prioritization could shift the distribution of federal and state grant and loan programs to encourage and support more watershed-based stormwater permitting programs. However, securing new levels of public funds will likely be required. All levels of government must recognize that additional resources may be required from citizens and businesses (in the form of taxes, fees, etc.) in order to operate a more comprehensive and effective stormwater permitting pro- gram. REFERENCES April, S., and T. Greiner. 2000. Evaluation of the Massachusetts Environmental Results Program. Washington, DC: National Academy of Public Admini- stration.. Atkins, J. R., C. Hollenkamp, and J. Sauber. 2007. Testing the watershed: North Carolinaâs NPDES Discharge Coalition Program enables basinwide monitoring and analysis. Water Environment & Technology 19(6). Bellucci, C. 2007. Stormwater and Aquatic Life: Making the Connection Be- tween Impervious Cover and Aquatic Life Impairments for TMDL Devel- opment in Connecticut Streams. Pp. 1003-1018 In: TMDL 2007. Alexan- dria, VA: Water Environment Federation. Bromberg, K. 2007. Comments to the NRC Committee on Stormwater Dis- charge Contributions to Water Pollution, January 22, 2007, Washington, DC. Burton, G. A., and R. E. Pitt. 2002. Stormwater Effects Handbook. Boca Raton, FL: Lewis/CRC Press. California EPA, State Water Board. 2006. Storm Water Panel Recommenda- tionsâThe Feasibility of Numeric Effluent Limits Applicable to Discharges of Storm Water Associated with Municipal, Industrial, and Construction Activities. Available at http://www.cacoastkeeper.org/assets/pdf/Storm- WaterPanelReport_06.pdf. Campbell, R. M. 2007. Achieving a Successful Storm Water Permit Program in Oregon. Natural Resources & Environment 21(4):39-44.
556 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES CASQA (California Stormwater Quality Association). 2007. Municipal Stormwater Program Effectiveness Assessment Guidance. Los Angeles. Available at info.casqa@org. Chapman, C. 2006. Performance Monitoring of an Urban Stormwater Treat- ment System. Master's Thesis, University of Washington, Seattle. City of Austin. 2006. Stormwater Runoff Quality and Quantity from Small Watersheds in Austin, TX. Austin, TX: Watershed Protection Department, Environmental Resources Management Division. City of San Diego. 2007. Strategic Plan for Watershed Activity Implementa- tion. San Diego, CA: Stormwater Pollution Prevention Division. Clark, S., R. Pitt, S. Burian, R. Field, E. Fan, J. Heaney, and L. Wright. 2006. The Annotated Bibliography of Urban Wet Weather Flow Literature from 1996 through 2005. Note: Publisher not shown. Clausen, J. C., and J. Spooner. 1993. Paired Watershed Study Design, 841-F- 93-009. Washington, DC: EPA Office of Water. Connecticut Department of Environmental Protection. 2007. A Total Maxi- mum Daily Load Analysis for Eagleville Brook, Mansfield, CT. Hartford: State of Connecticut Department of Environmental Protection. Available at http://www.ct.gov/dep/lib/dep/water/tmdl/tmdl_final/eaglevillefinal.pdf. Cosgrove, J. F. 2002. TMDLs: A simplified approach to pollutant load deter- mination. WEFTEC 2002 Conference Proceedings September 2002. Alex- andria, VA: Water Environment Federation. Crockett, C. 2007. The regulated perspective of stormwater management. Presentation to the NRC Committee on Stormwater Discharge Contribu- tions to Water Pollution. January 22, 2007. Washington, DC. Cross, L. M., and L. D. Duke. 2008. Regulating industrial stormwater: state permits, municipal implementation, and a protocol for prioritization. Jour- nal of the American Water Resources Association 44(1):86-106. Cutter, W. B., K. A. Baerenklau, A. DeWoody, R. Sharma, and J. G. Lee. 2008. Costs and benefits of capturing urban runoff with competitive bidding for decentralized best management practices. Water Resources Research, doi:10.1029/2007WR006343. DeWoody, A. E. 2007. Determining Net Social Benefits from Optimal Parcel- Level Infiltration of Urban Runoff: A Los Angles Analysis. M.S. Thesis. University of California, Riverside. Doll, A., and G. Lindsey. 1999. Credits bring economic incentives for onsite stormwater management. Watershed and Wet Weather Technical Bulletin 4(1):12-15. Duke, L. D. 2007. Industrial stormwater runoff pollution prevention regula- tions and implementation. Presentation to the National Research Council Committee on Reducing Stormwater Discharge Contributions to Water Pol- lution, Seattle, WA, August 22, 2007. Duke, L. D., and C. A. Augustenborg. 2006. Effectiveness of self identified and self-reported environmental regulations for industry: the case of storm water runoff in the U.S. Journal of Environmental Planning and Manage-
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 557 ment 49:385-411. Duke, L. D., and P. Beswick. 1997. Industry compliance with storm water pol- lution prevention regulations: the case of transportation industry facilities in California and the Los Angeles region. Journal of the American Water Re- sources Association 33:825-838. Ellerman, A. D., P. L. Joskow, R. Schmalensee, J. P. Montero, and E. M. Bai- ley. 2000. Markets for Clean Air: the U.S. Acid Rain Program. New York: Cambridge University Press. EPA (U. S. Environmental Protection Agency). 1991. Technical Support Document for Water Quality-Based Toxics Control. EPA-505/2-90-001. Washington, DC: EPA Office of Water Enforcement and Permits. EPA. 1999. Introduction to the National Pretreatment Program. EPA-833-B- 98-002. Washington, DC: EPA Office of Wastewater Management. EPA. 2002. Establishing total maximum daily load (TMDL) Wasteload alloca- tions (WLAs) for storm water sources and NPDES permit requirements based on those WLAs. Memorandum from Robert Wayland, Director, Of- fice of Wetlands, Oceans, and Watersheds to Jim Hanlon, Director, Office of Water, November 22, 2002. Available at www.epa.gov/npdes/pubs/ final-wwtmdl.pdf. EPA. 2003a. Watershed-Based NPDES Permitting Policy Statement. In Water- shed-Based National Pollutant Discharge Elimination System (NPDES) Permitting Implementation Guidance. EPA, Washington, DC. EPA. 2003b. Watershed-Based National Pollutant Discharge Elimination Sys- tem (NPDES) Permitting Implementation Guidance. EPA, Washington, DC. EPA. 2007a. Watershed-Based NPDES Permitting Technical Guidance (draft). EPA, Washington, DC. EPA. 2007b. Water Quality Trading Toolkit for Permit Writers. EPA 833-R- 07-004. Washington, DC: EPA Office of Wastewater Management, Water Permits Division. EPA. 2007c. Understanding Impaired Waters and Total Maximum Daily Load (TMDL Requirements for Municipal Stormwater Programs. EPA 883-F- 07-009. Philadelphia, PA: EPA Region 3. Freedman, P., L. Shabman, and K. Reckhow. 2008. Donât Debate; Adaptive implementation can help water quality professionals achieve TMDL goals. WE&T Magazine August 2008:66-71. Frie, S., L. Curtis, and S. Martin. 1996. Financing regional stormwater facili- ties. In: Managing Virginiaâs Watersheds in the 21st Century: Workable So- lutions. Proceedings from the 9th Annual Virginia Water Conference, Staunton. Grumbles, B. 2006. Qualifying Local Programs for Construction Site Storm Water Runoff. Memorandum from EPA Assistant Administrator Ben Grumbles to James Mac Indoe, Alabama Dept. of Environmental Manage- ment. May 8. Hetling, L .J., A Stoddard, and T. N. Brosnan. 2003. Effect of water quality
558 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES management efforts on wastewater loadings during the past century. Water Environment Research 75(1):30. Hirschman, D., K. Collins, and T. Schueler. 2008. Draft Virginia Stormwater Management Nutrient Design System. Prepared for Technical Advisory Committee and Virginia Department of Conservation and Recreation. Elli- cott City, MD: Center for Watershed Protection. Holling, C. S., ed. 1978. Adaptive Environmental Assessment and Manage- ment. Caldwell, NJ: Blackburn Press. Holling, C. S., and A. D. Chambers. 1973. Resource science: The nurture of an infant. BioScience 23:13-20. Horner, R. R., and C. Chapman. 2007. NW 110th Street Natural Drainage Sys- tem Performance Monitoring, with Summary of Viewlands and 2nd Avenue NW SEA Streets Monitoring. Report to Seattle Public Utilities by Depart- ment of Civil and Environmental Engineering, University of Washington, Seattle. Horner, R. R., H. Lim, and S. J. Burges. 2002. Hydrologic Monitoring of the Seattle Ultra-Urban Stormwater Management Projects, Water Resources Series Technical Report No. 170. Department of Civil and Environmental Engineering, University of Washington, Seattle. Horner, R. R., H. Lim, and S. J. Burges. 2004. Hydrologic Monitoring of the Seattle Ultra-Urban Stormwater Management Projects: Summary of the 2000-2003 Water Years. Water Resources Series Technical Report 181. Department of Civil and Environmental Engineering, University of Wash- ington, Seattle. Horner, R., C. May, E. Livingston, D. Blaha, M. Scoggins, J. Tims, and J. Max- ted. 2001. Structural and non-structural BMPs for protecting streams. Pp. 60-77 In: Linking Stormwater BMP Designs and Performance to Receiving Water Impact Mitigation. Proceedings Engineering Research Foundation Conference. American Society of Civil Engineers. Horner, R., J. Guedry, and M. Kortenhof. 1990. Improving the Cost- Effectiveness of Highway Construction Site Erosion and Sediment Control. Washington State Department of Transportation. Seattle, WA: Department of Civil Engineering, University of Washington. Keller, B. 2003. Buddy can you spare a dime? What is stormwater funding. Stormwater 4:7. LaFlamme, P. 2007. Presentation to the Committee on Stormwater Discharge Contributions to Water Pollution, January 22, 2007, Washington, DC. Lee, H., X. Swamikannu, D. Radulescu, S. Kim, and M. K. Stenstrom. 2007. Design of stormwater monitoring programs. Journal of Water Research, doi:10.1016/j.watres.2007.05.016. Longsworth, J. 2007. Comments to the NRC Committee on Stormwater Dis- charge Contributions to Water Pollution. January 22, 2007, Washington, DC. Los Angeles County Department of Public Works. 2001. Los Angeles County 1994-2000 Integrated Receiving Water Impacts Report. Available at
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 559 http://ladpw.org/wmd/NPDES/IntTC.cfm. Maimone, M. 2002. Prioritization of contaminant sources for the Schuylkill River source water assessment. Presentation at the Watershed 2002 Spe- cialty Conference, Ft. Lauderdale, FL, February 23-27. NRC (National Research Council). 1990. Monitoring Southern Californiaâs Coastal Waters. Washington, DC: National Academy Press. NRC. 1999. New Strategies for Americaâs Watersheds. Washington, DC: Na- tional Academy Press. NRC. 2001a. Assessing the TMDL Approach to Water Quality Management. Washington, DC: National Academy Press. NRC. 2001b. Compensating for Wetland Losses Under the Clean Water Act. Washington, DC: National Academy Press. Natural Resources Conservation Service. 2007. Part 630 Hydrology, National Engineering Handbook, Chapter 7, Hydrologic Soil Groups. Washington, DC: U.S. Department of Agriculture. Nirel, P. M., and R. Revaclier. 1999. Assessment of sewage treatment plant effluents impact on river water quality using dissolved Rb:Sr ratio. Envi- ronmental Science and Technology 33(12):1996. North Carolina Division of Water Quality. 1999. Neuse River Basin Model Stormwater Program for Nitrogen Control. Available at http://h2o.enr.state.nc.us/su/Neuse_SWProgram_Documents.htm. Last ac- cessed November 2007. Office of Inspector General. 2007. Development Growth Outpacing Progress in Watershed Efforts to Restore the Chesapeake Bay. Report 2007-P-000031. Washington, DC: U.S. Environmental Protection Agency. Parikh, P., M. A. Taylor, T. Hoagland, H. Thurston, and W. Shuster. 2005. Application of market mechanism and incentives to reduce stormwater run- off: an integrated hydrologic, economic, and legal approach. Environ- mental Science and Policy 8:133-144. Pitt, R., A, Maestre, and R. Morquecho. 2004. National Stormwater Quality Database. Version 1.1. Available at http://rpitt.eng.ua.edu/Research/ ms4/Paper/Mainms4paper.html. Last accessed January 28, 2008. Schiff, K., D. Ackerman, E. Strecker, and M. Leisenring. 2007. Concept devel- opment: design storm for water quality in the Los Angeles region. Southern California Coastal Water Research Project. Costa Mesa. Schueler, T. 2008b. Technical Support for the Baywide Runoff Reduction Method. Baltimore, MD: Chesapeake Stormwater Network. Available at www.chesapeakestormwater.net. Schueler, T. 2004. An Integrated Framework to Restore Small Urban Water- sheds. Manual 1. Urban Subwatershed Restoration Manual Series. Ellicott City, MD: Center for Watershed Protection. Schueler, T. 2008a. Final Bay-wide Stormwater Action Strategy: Recommen- dations for Moving Forward in the Chesapeake Bay. Baltimore, MD: Chesapeake Stormwater Network. Schueler, T., D. Hirschman, M. Novotney, and J. Zielinski. 2007. Urban
560 URBAN STORMWATER MANAGEMENT IN THE UNITED STATES Stormwater Retrofit Practices. Ellicott City, MD: Center for Watershed Protection. Shabman, L., and K. Stephenson. 2007. Achieving nutrient water quality goals: bringing market-like principles to water quality management. Journal of American Water Resources Association 43(4):1076-1089. Shabman, L., and P. Scodari. 2004. Past, Present, and Future of Wetland Credit Sales. Discussion Paper 04-48. Washington, DC: Resources for the Future. Shabman, L., K. Reckhow, M. B. Beck, J. Benaman, S. Chapra, P. Freedman, M. Nellor, J. Rudek, D. Schwer, T. Stiles, and C. Stow. 2007. Adaptive Implementation of water quality improvement plans: opportunities and challenges. Durham, NC: Duke University. Shabman, L., K. Stephenson, and W. Shobe. 2002. Trading programs for envi- ronmental management: reflections on the air and water experiences. Envi- ronmental Practice 4:153-162. Shaver, E., R. Horner, J. Skupien, C. May, and G. Ridley. 2007. Fundamentals of Urban Runoff Management: Technical and Institutional Issue, 2nd Ed. Madison, WI: North American Lake Management Society. Smith, B. 2007. Comments to the NRC Committee on Stormwater Discharge Contributions to Water Pollution. January 22, 2007, Washington, DC. Stenstrom, M. K., and H. Lee. 2005. Final Report. Industrial Stormwater Monitoring Program. Existing Statewide Permit Utility and Proposed Modifications. Stephenson, K., and L. Shabman. 2005. The use and opportunity of cooperative organizational forms as an innovative regulatory tool under the Clean Water Act. Paper presented at the Southern Agricultural Economics Association Annual Meetings Little Rock, AK, February 5-9, 2005. Stephenson, K., L. Shabman, and J. Boyd. 2005. Taxonomy of trading pro- grams: Concepts and applications to TMDLs. Pp. 253-285 In: Total Maxi- mum Daily Loads: Approaches and Challenges. Tamim Younos (ed.). Tulsa, OK: Pennwell Press. Stephenson, K., L. Shabman, and L. Geyer. 1999. Watershed-based effluent allowance trading: Identifying the statutory and regulatory barriers to im- plementation. Environmental Lawyer 5(3):775-815. Stephenson, K., P. Norris, and L. Shabman. 1998. Watershed-based effluent trading: the nonpoint source challenge. Contemporary Economic Policy 16:412-421. Thurston, H. W., H. C. Goddard, D. Szlag, and B. Lemberg. 2003. Controlling storm-water runoff with tradable allowances for impervious surfaces. Jour- nal of Water Resources Planning and Management 129(5):409-418. Vermont Department of Environmental Conservation. 2006. Total Maximum Daily Load to Address Biological Impairment Potash Brook (VT05-11). Chittenden County, Vermont. Virginia DCR (Virginia Department of Conservation and Recreation). 2007. Virginia Stormwater Management Program Permit Regulations, Chapter 60. Wagner, W.E. 2006. Stormy regulations: The problems that result when storm
INNOVATIVE STORMWATER MANAGEMENT AND REGULATORY PERMITTING 561 water (and other) regulatory programs neglect to account for limitations in scientific and technical programs. Chapman Law Review 9(2):191-232. Wenger, S., T. Carter, R. Vick, and L. Fowler. 2008. Runoff limits: an ecologi- cally based stormwater management program. Stormwater April/May. Available at http://www.stormh2o.com/march-april-2008/ecologically- stormwater-management.aspx. Woodward, R. T., R. A. Kaiser, and A. B. Wicks. 2002. The structure and practice of water quality trading markets. Journal of the American Water Resources Association 38:967-979. Zeng, X., and T. C. Rasmussen. 2005. Multivariate statistical characterization of water quality in Lake Lanier, Georgia, USA. Journal of Environmental Quality 34(6):1980-1991.