Stormwater Credit Trading Program Architecture

Stormwater credit trading programs are gaining traction as a new tool in the effort to reduce the cost of stormwater-borne pollution control in impaired watersheds at a time when new tools are sorely needed. Since the amendments of the Clean Water Act in 1987 to extend coverage to nonpoint pollution under National Pollutant Discharge Elimination System (NPDES) permits, municipal stormwater programs have been incrementally strengthened each new permit term following the iterative process in an effort to meet applicable water quality standards. However, restoration of impaired water bodies impacted by stormwater runoff is rare. One reason is that in most regions, stormwater pollution comes predominately from existing private ­development, which had been built at a time when stormwater retention or effective stormwater controls were not required. Aside from some industrial sites, stormwater programs do not directly regulate these contributing areas, and storm­water infrastructure improvements come with the pace of redevelopment.

Funding essential municipal program elements like ­public education, water quality monitoring, inspections, reporting, and reviewing development permits is a chronic challenge. Funding for the kind of capital improvements needed to mitigate the impact of stormwater pollution coming from developed land is even rarer. Stormwater credit trading is an opportunity to accelerate infrastructure improvements on those sites that are not being regulated as part of land development or industrial stormwater management programs. Credit trading also provides the opportunity to address high-priority water quality concerns in a more cost-efficient manner by incentivizing projects that have a high impact-to-cost ratio.

Ideally, these programs provide the flexibility to manage the most urgent water quality threats in a watershed with solutions that provide the greatest pollution reduction benefit per dollar spent, as well as provide flexibility to those struggling to meet water quality goals on constrained sites. Credit programs reduce the total cost of watershed restoration and encourage redevelopment of highly urbanized areas. However, special care must be exercised to avoid creating local receiving water problems and other unintended consequences while addressing larger watershed issues. The structure of existing credit trading programs varies significantly with regard to how credits are generated and traded, and when they are allowed to be used in lieu of onsite stormwater management. There are also differences in the geographic scope of credit trading programs and in approaches to ensuring that credit-generating BMPs are constructed and maintained properly.

Alternative compliance programs are also emerging as a component of low impact development (LID) or green infrastructure-based MS4 permits, and are often structured very similarly to credit trading programs. For example, in San Diego, CA, land development projects must manage as much of the design storm as possible onsite using LID approaches. Any portion of the design storm that is infeasible to manage onsite must be managed offsite with BMPs that will have at least equal effectiveness as compared to approved LID BMPs. Inherent in this approach is the need to quantify the pollutant load or volume reduction capabilities of onsite LID BMPs and to assign credit to offsite BMPs that can offset mitigation requirements that are not met onsite.

Currency
Most credit trading programs use a currency of pollutant load or runoff volume. One advantage of trading pollutant load is that calculations can compensate for different runoff sources with different concentrations, yielding more credits for treating relatively polluted water. For example, treating the same amount of runoff from a typical commercial area would reduce pollutant loads more than treating runoff from a typical low-density residential area. Trading pollutant load also facilitates trade between disparate sources. For example, industrial or agricultural discharges can be traded for urban stormwater discharges.

Trading runoff volume generally assumes equal impact of retaining a certain volume of water regardless of its source, which can simplify equivalency calculations. Because runoff retention also effectively retains most conventional pollutants, trading with a currency of runoff volume is more likely to result in management of a wide range of pollutants. Volume reduction is also an obvious currency to use in areas where peak flow reduction is desired—for example, where combined sewer overflow events are impacting water quality or where downstream flooding or erosion is an issue.

Both runoff volume and pollutant load reduction are useful credit trading currencies because they are direct measures of the impact of BMPs. Impervious area reduction is less effective as a credit currency because it is not a direct cause of impairment. In general, credits should be issued for pollutants, not site conditions that are expected to impact pollutant discharge but do not directly cause downstream impairments.

Pollutant Issues
In watersheds where one pollutant is dominant, the credit trading currency may be the load of that specific pollutant. For example, phosphorus credits are traded throughout Virginia both within the Chesapeake Bay watershed and the Southern Rivers watersheds. In watersheds where multiple TMDLs overlap, stormwater retention volume makes more sense as a traded currency as is the case in the Anacostia River watershed in the District of Columbia. Regardless of the credit trading currency, it is important to consider that urban runoff transports a wide variety of pollutants, including nutrients, trash, heavy metals, sediment, bacteria, oil, and grease. Trading with a focus on one or a small number of pollutants can result in a watershed-wide net increase in non-traded pollutants in downstream waters. To prevent harmful accumulation of non-traded pollutants, while recognizing the potential for spills or other site-level issues, a baseline level of treatment should be provided on all sites. For example, in San Diego, offsite stormwater management is an acceptable alternative to implementing retention or other LID technologies onsite; but if pursuing offsite options, treatment of the design storm with conventional treatment controls is still required onsite (San Diego Regional Water Quality Control Board 2015).

Trading pollutant load reductions across sectors is a popular approach, especially where nutrients are the credit currency, because non-urban BMPs like agricultural land conversion and streambank restoration can provide nutrient load reductions at a fraction of the cost of urban stormwater controls. In Virginia, for example, conversion of farmland to agroforestry or forest can generate phosphorus credits with a value of $10,800 to $24,000 per pound, as compared to stormwater controls, which were found to range from $20,100 to $74,900 per pound of phosphorus removed in Virginia Department of Transportation projects (Nobles et al. 2017). However, where allowed, it may leave new urban development largely untreated. This effectively sacrifices the health of smaller urban streams in favor of protecting larger receiving water body health. It also potentially neglects the impact of other non-traded pollutants because non-urban BMPs targeting traded pollutants are not likely to also reduce loads of common urban stormwater-borne pollutants. For example, agricultural land conversion may have no impact on hydrocarbons, trash, and heavy metals, all of which are likely to be generated on urban sites. To ensure that urban development does not create additional impairments, credits should not be tradable across sectors as defined in the TMDL load allocations and waste-load allocations.

Hydromodification
Land development and the associated conversion of pervious landscapes to impervious surfaces increase the frequency, duration, magnitude, and temperature of stormwater runoff (Stein et al. 2012). Downstream impacts are commonly experienced as changes in downstream channel structure, flow patterns, and sediment yield, as well as reduction in biodiversity. These hydromodification effects are best managed close to the source and consistent with the green infrastructure approach promoted by EPA (USEPA 2017) and others. In keeping with this guidance, stormwater credit trading should be utilized as an alternative compliance pathway only where onsite stormwater management is deemed infeasible. For example, the Phase I NPDES permit for Ventura County, CA, requires that LID strategies for managing the water quality design storm be exhausted prior to considering offsite compliance (Los Angeles Regional Water Quality Control Board 2010). If trading is allowed without first demonstrating infeasibility of onsite management, such trading should be limited to sites that are smaller than one acre or those that do not have a significant hydromodification impact or pollutant load-generating probability.

Even when runoff volume is the traded currency, flow control protection should still be provided onsite so that there will be no adverse hydrologic impact from land development. When nutrient and sediment load reduction credits are being generated through streambank restoration or other downstream controls, it is critical that upstream hydromodification controls be provided to help to avoid a situation where increased flows from unmitigated development undermine downstream improvements.

Factors of Safety and Timing
Some alternative compliance programs allow payment into a mitigation bank in lieu of managing stormwater onsite. The accumulated funds in the mitigation bank are then used to construct retrofit projects that have an equal or greater benefit compared to the controls that would have met onsite requirements. This approach is problematic for several reasons. Most notably, construction of the credit-generating facilities depends on the rate of participation in the program. If the value of the mitigation bank does not grow to the point that it can support new projects, land development projects could be completed without the necessary offsetting projects in place. Additionally, the timing of credit-generating projects will necessarily be delayed as compared to the construction of the land development project relying on them. Adding to the intrinsic funding-related delays, permitting, design, and construction of watershed retrofit projects can take several years, even for projects without unforeseen complications. Whether paying into mitigation banks or purchasing credits, the credit-producing projects should be completed before those credits are claimed by development projects.

Design and ­Construction Verification
Whether the traded currency is pollutant load or runoff volume, the credit quantity generated relies on assumptions about the size and effectiveness of the credit-generating projects. While these assumptions will almost certainly be grounded in sound engineering principles, it is also common for as-built details to vary somewhat from design details. These differences should be minor, but sometimes in the construction phase, site constraints are discovered and affect the project design. It is also common for contractors to substitute materials, misinterpret plans, or otherwise make decisions that affect site performance.

One solution is to require verification that as-built plans match design plans with necessary adjustments to the credits given or received to reconcile differences. This is already an element of most municipal stormwater programs, with certification required proving that proper stormwater BMP construction was completed prior to issuance of a Notice of Termination for the construction phase of the project. However, the certification methods vary. In some cases, the contractor can certify that the plans were followed. In other cases, it is the engineer of record. Preferably, an independent audit of the design is performed by the permittee or the permitee’s agent to verify the proper construction of the installed systems.

Operation and Maintenance Verification
Credit-generating facilities must be maintained so that they are in good operating condition for the duration of the time that the credits are being received. Unless there are subsequent regulatory changes or BMPs employed on the credit-receiving project, this will be equal to the life of the credit-receiving project. There are a couple of options for capturing the cost of operation and maintenance in the credit valuation.

The simplest option is to include the anticipated operation and maintenance cost in the price of a one-time credit purchase. This approach burdens the owner of the credit-generating facility with all the risk related to operational cost uncertainty. The owner must factor in assumptions about the cost and timing of routine maintenance and major or corrective maintenance, as well as inflation of those costs over time, into the credit price. The greater the operation and maintenance cost uncertainty, the greater the contingency reserve needs to be, and the greater the credit price.

A preferred option is to make credit purchase renewable at a fixed time period with proof of adequate operation of the credit-generating facility required upon renewal, and adjustments to the price as needed to reflect the operator’s costs and the current market value of credits. This has several potential advantages over a one-time purchase approach. Credit renewal should be contingent on inspection of the credit-generating facility and verification that it is operating as intended. This check would ideally be performed by an independent authority that can objectively assess the facility condition and confirm that the assumed volume reduction or treatment effectiveness is being provided.

Allowing the price of the credits to be changed at the time of renewal gives the owner an opportunity to make adjustments based on a known operation and maintenance history, which should reduce the credit price. It also provides an opportunity to correct for the market value of credits. Over time, the market value of credits should increase—assuming that the cheapest-to-generate credits are created first and the cheapest credits available are purchased first. This has driven some speculative investing in stormwater credits. For example, Prudential recently invested $1.7 million in the Washington DC credit trading program through District Stormwater to install green infrastructure controls that will generate credits that can potentially be traded for a profit (Spector 2016).

Repricing gives an opportunity to rebalance the price with the current market demand. Renewal at a five-year interval provides a good balance of flexibility and budgeting certainty and is consistent with the municipal NDPES stormwater permit renewal interval schedule. Changes to the credit trading contract should be limited outside of the point of renewal.

If either party wants to make changes to the way stormwater is managed on the site, those changes should be made coinciding with the end of a credit exchange contract. For example, retired farmland should be put back into production only when there are no sites relying on the credits generated by retiring it. Similarly, if development density is increased upstream of a regional treatment or retention facility, thereby reducing the credit-generating capacity of that facility, it should be undertaken when those credits are not already claimed. On the other hand, a site receiving stormwater credits may undergo redevelopment or redesign to add stormwater management capacity, and may no longer need stormwater credits. This retrofit capacity should be planned and constructed while a credit trade contract is in place.

Factor of Safety
Given uncertainties in hydraulic and pollutant loading calculations, maintenance diligence, BMP effectiveness, differences in the impact of practices at different locations in the watershed, and other factors, there are likely to be discrepancies between the calculated benefit of the credit-generating facility and the credit value required. To account for any differences, a factor of safety should be applied to credit trades. For example, the pollutant load or runoff volume at the credit-receiving facility can be required to be offset by a ratio of at least 1.1:1, or a discount in trading value of at least 10% could be applied to the calculated value of generated credits.

Planned retirement of credits is another option for adding a safety margin to the entire program. For example, 10% of available credits could be retired per year by the regulatory authority through direct purchase, which would help to support the credit generation market and improve watershed health without having a significant price impact on the market. Alternatively, the regulatory authority could purchase and retire credits generated at or below a certain threshold price and could count those retired credit values toward the required load or volume reductions. This would keep the credit value from bottoming out in a market and could be relatively cost-effective means of complying with load-reduction targets for public agencies. Nonprofit organizations, concerned individuals, and others should also be encouraged to purchase and retire credits because it will accelerate watershed rehabilitation.

Crediting Extra BMP Capacity
It may be tempting to oversize a required BMP to create tradable stormwater credits, but this approach can be problematic depending on how the extra capacity is credited. The simple approach of assuming the extra storage or treatment capacity is available to offset another development that would produce a design storm equal to that extra capacity is flawed. It neglects the fact that the extra capacity would be used only when the design storm is exceeded. From the perspective of pollutant load or volume reduction, one large system is less effective than multiple smaller systems with the same combined capacity because more of the small system capacity would be used more often (Dougherty et al. 2016).

A better approach would be to calculate the actual annual pollutant load reduction or volume reduction onsite that results from the oversized system and to trade that capacity.

Conclusion
The structure of existing credit trading programs varies significantly with regard to how credits are generated and traded and when they are allowed to be used in lieu of onsite stormwater management. There are also important differences in the geographic scope of credit trading programs and in construction and maintenance management of credit-generating BMPs.

To minimize the unintended negative effects of credit trading programs, a few program elements are important. The traded credit currency should be pollutant load reduction or volume reduction. Trading should be limited geographically so that smaller tributary streams are not sacrificed in order to restore larger receiving waters. Credits should be renewed at a fixed interval with inspection of credit-generating facilities when credits are first generated, and each time the credit contracts are renewed. To protect local receiving waters and to avoid accumulation of non-traded pollutants, a baseline level of treatment should be applied to all credit-receiving sites. To compensate for uncertainty in the credit exchange process, a factor of safety should be applied to credit values.

With these safeguards, stormwater credit trading can provide much-needed flexibility to manage the most urgent water quality threats in a watershed with solutions that provide the greatest pollution reduction benefit per dollar spent.

References
Dougherty, S., R. Hammer, and A. Valderrama. 2016. How to: Stormwater Credit Trading Programs. Natural Resources Defense Council Issue Brief, IB: 16-01-a.

Los Angeles Regional Water Quality Control Board. 2010. Waste Discharge Requirements for Storm Water (Wet Weather) and Non-Storm Water (Dry Weather) Discharges from the Municipal Separate Storm Sewer Systems within the Ventura County Watershed Protection District, County of Ventura and the Incorporated Cities Therein. Order No. R4-2010-0108.

Maryland Department of the Environment, Maryland Department of Agriculture. April 2017. Final Draft—Maryland Trading and Offset Policy and Guidance Manual—Chesapeake Bay Watershed.

Alicia L. Nobles, Jonathan L. Goodall, and G. Michael Fitch. 2017. “Comparing Costs of Onsite Best Management Practices to Nutrient Credits for Stormwater Management: A Case Study in Virginia.” Journal of the American Water Resources Association 53(1): 131–143. DOI: 10.1111/1752-1688.12487.

San Diego Regional Water Quality Control Board. 2015. National Pollutant Discharge Elimination System (NPDES) Permit and Waste Discharge Requirements for Discharges from the Municipal Separate Storm Sewer Systems (MS4s) Draining the Watersheds within the San Diego Region. Order No. R9-2013-0001 (as amended by Order Nos. R9-2015-0001 and R9-2015-0100).

Spector, J. 2016. “Turning Stormwater Runoff into Everyone’s Business.” Citylab, March 18, 2016. Retrieved April 2017 at http://www.citylab.com/cityfixer/2016/03/stormwater-runoff-credits-nature-conservancy-washington-dc/473700/

Stein, E.D., F. Federico, D.B. Booth, B.P. Bledsoe, C. Bowles, Z. Rubin, and A. Sengupta. 2012. Hydromodification Assessment and Management in California. Southern California Coastal Water Research Project Technical Report 667, 138.

USEPA. 2017. Green Infrastructure. Accessed May 2017 at https://www.epa.gov/green-infrastructure/