The Pollutant Removal Efficiency Conundrum

Oct. 1, 2009

Over the last several years, there has been considerable debate regarding the science of measuring the performance of a best management practice (BMP)–as a percent removal efficiency (%RE) measured by influent and effluent event mean concentration (EMC), as an effluent concentration, or as a mass balanced removal efficiency. Reasons for not using a %RE have been discussed at length by the International Stormwater BMP Database team, principally based upon the limitations of testing techniques for total suspended solids (TSS) and the attempts to use EMC measurements. Limitations of these techniques include the variability of concentrations, particle sizes, flow rates, first-flush effects, and minimum reducible concentrations, all pointing to the need to develop alternate BMP measurement methods. At least as far as TSS is concerned, researchers have demonstrated that using a %RE based upon influent and effluent EMCs has serious limitations.

Background
Let’s step back and take a look at the big picture for a moment. Why are we measuring BMP performance? The bottom line is that stormwater pollution is most often being quantified either in response to enforcement of Clean Water Act provisions, or in response to environmental lawsuits.

The rub is that, historically, as engineers, we were trained from the wastewater treatment plant perspective to think in terms of meeting pollutant standards by reducing the effluent concentrations of pollutants to a numeric limitation. That type of requirement was well suited for point sources where there were fairly constant inflow volumes and pollutant loads, as well as one outfall pipe to sample.

However, nonpoint sources are horses of a different color. Across the country, total maximum daily loads (TMDLs) are being expressed in numerous ways, depending on factors such as type of pollutant to be reduced and lawsuit settlement conditions. Because of the large number of disparate outflow points from stormwater systems, EPA and states sometimes take the position that effluent limitations are neither practical nor enforceable for stormwater discharges and have adopted strategies such as reducing stormwater pollutants based upon mass loadings and BMP mass loading %RE for enforcement expediency’s sake. A partial listing of TMDL expressions is shown below.

  1. Load reduction (tons/mi3/yr)
  2. EMC reduction (mg/l)
  3. EMC limitation (ppm or gm/day)
  4. Annual mass reduction (kg/yr)
  5. Temperature imitation (degrees)
  6. Total dissolved gas (mm hg)
  7. Bacteria (CFU/100 mg)

As can be seen, TMDLs, pollutant load calculations, and the need to calculate BMP pollutant removals are being driven in many different directions. It is this divergence between regulatory requirements and the science of BMP measurement that causes many of us to scratch our heads.

For example, TMDLs in Florida are established by assessing the health of an impaired water body, determining which pollutants cause the impairment, calculating how many pounds per year of each pollutant the water body could receive without causing impairment, and then setting requirements upon communities to reduce their pollutant discharges (either nonpoint sources or point sources or both) to those levels, either as a mass annual percentage reduction or with fixed-load reductions expressed in pounds per year. Most of the TMDLs in Florida are for nutrient reductions. Mass annual reductions are generally used in order to normalize variations in watershed uses, rainfall intensities, soil conditions, and seasonal effects, all of which affect influent concentrations and BMP performance.

The Florida Department of Environmental Protection (DEP) enforces these load allocations through basin management action plans (BMAPs). As part of the BMAP process, each community must identify specific retrofit projects it will commit to undertake to meet its load reductions. Often the BMP selection is driven by a water-quality master plan long before survey and hydraulic data are available for the traditional design process. Each project must detail pollutant loadings entering the BMP and the number of pounds of pollutants removed by the BMP. The reductions must be calculated by applying a mass annual %RE (not an EMC %RE) for each proposed BMP against the mass annual loading from the contributing sub-basin to demonstrate the pounds of pollutant removed for each BMP on an annual basis. A city’s list of proposed BMPs and its associated load reductions are approved by Florida DEP and inserted into the city’s National Pollutant Discharge Elimination System MS4 permit.

Another example of using a mass annual %RE is in EPA’s 319 grant program, where calculation of kilograms per year of individual pollutants removed for each BMP funded is required for the grant application. These planning grants again happen long before there is a designed BMP. In addition, for areas of Florida discharging to listed waters, new development permits require matching pre- versus post-development pollutant loads for total nitrogen or total phosphorus using %RE calculations. Yet another example is the requirement to match pre- versus post-pollutant loads for US Army Corps of Engineers 404 permits for properties in Florida discharging to listed waters.

As can be seen, a stormwater professional must now be able to calculate pollutant loads and reductions for proposed retrofit- or development-related BMPs with a degree of accuracy that satisfies permitting and TMDL regulators.

That is the big picture, which creates quite a conundrum. Clearly, there is a need for practitioners to use some form of %RE in pollutant loading calculations, and clearly the scientific community has demonstrated that the accuracy of EMC %RE values leaves a lot to be desired.

Recommendation
To address these conflicting needs, I propose an approach using a tiered system of pollutant load and reduction calculations based upon the accuracy level desired by the user as follows.

  1. Level 1. Low level of accuracy–Master planning
  2. Level 2. Medium level of accuracy–Design and permitting
  3. Level 3. High level of precision–Calibrated BMP or TMDL study

This approach is akin to a level-of-service approach commonly used in engineering design criteria.

Level 1 pollutant load calculations are appropriate for planning studies where low levels of BMP performance accuracy are tolerable. These types of water-quality (not hydraulic and hydrologic) studies are often used by communities, to demonstrate TMDL compliance and for screening of BMPs before surveys and site conditions are fully known. A single number for BMP performance can be used for a generic BMP class, with general assumptions made for pollutant loading rates that have far more variability than the BMP performance. A Level 1 calculation would also be appropriate for pollutant load calculations for stormwater grant applications. With these types of analyses, the use of a mass annual %RE can be more accurate than a %RE based on EMCs, and is in fact recommended by EPA for 319 Stormwater Retrofit grants.

Level 2 pollutant load calculations would be the next step up in precision and accuracy, used by practitioner engineers for design and permitting. At this level, survey and watershed data would be available for actual hydraulic and hydrologic data inputs to a model. Precise BMP performance and dimensions are used to meet specified design criteria for effluent control or pollutant removal efficiency. Another example of a Level 2 analysis would be the monitoring required for EPA 319 grants to demonstrate post-construction BMP effectiveness.

With Level 3 calculations, the highest levels of precision, accuracy, and modeling would be used to certify BMP performance, or for studies used by regulatory agencies to establish TMDL loading criteria. Examples of Level 3 analysis would be the studies used to include a BMP in the International Stormwater BMP Database, or for New Jersey Corporation for Advanced Technology (NJCAT), or Technology Acceptance and Reciprocity Partnership BMP certification. A Level 3 study would typically be performed by researchers or academians.

Additional Research
As part of the evolving science of stormwater, the research community has focused on developing more accurate methods to measure BMP performance. There has been progress in TSS measurement through the NJCAT program, the Environmental & Water Resources Institute (EWRI) task committees on Certification Guidelines for Manufactured Stormwater BMPs, and Guidelines for Monitoring Stormwater Gross Solids, as well as others.

When attention was diverted from the wastewater world to the stormwater world 20 years ago, EPA’s method 160.2 for TSS measurement was the best available science. Researchers have since determined that the TSS measurement does not adequately account for large sediment particles found in stormwater and that measuring suspended solids concentration (SSC) more accurately accounts for larger particles (Rosner et al. 2007, Guo 2007). This method works well for controlled laboratory testing, but during field sampling, autosamplers are not capable of accurately and consistently sampling sediment particles above 75 microns. We should move forward to develop more accurate methods of sampling and reporting larger sediments, such as recommended in EWRI’s Gross Solids Committee report. Note that the report still recommends the use of autosamplers for sampling sediment particles of less than 75 microns, as well as other dissolved pollutants (Rushton et al. 2009).

As long as the acronym TSS is used for both wastewater and stormwater purposes, there will be confusion as to definitions and applications. It may be better to break away from the term TSS for stormwater purposes and use another term such as stormwater suspended solids for sediment particles of less than 75 microns, with particles greater than or equal to 75 microns being called coarse solids.

Sampling of lakes and wet detention ponds sometimes shows that effluent TSS measurements exceed influent concentrations. This anomaly can usually be explained by the organic (not sediment) particle contribution to TSS from the biological (algae and phytoplankton), colloidal, and flocculation activities in the lakes. Again, the point is emphasized that using TSS is not always a suitable parameter for measuring stormwater pollutants. Further compounding the issue is that both the organic and inorganic components of TSS can lead to impairment of waterways.

Methods to measure the organic content of a water sample include the ASTM Standard Test Method D2579-93e1 for Total Organic Carbon in Water and Standard Method 5310 for Total Organic Carbon. Methods such as these should be used to refine data analysis to differentiate between organic and sediment-based solids. By accounting for these different solids components and by using a mass balanced %RE methodology, a more accurate determination of a BMP’s field performance can be calculated.

A daunting problem for BMP manufacturers, consultants, and municipal engineers is that many state and local regulations were written in the early years of stormwater management when the best available science was to require BMPs that provided 80% removal of pollutants (generally TSS). Without more detailed monitoring requirements, BMP manufacturers have used a wide variety of conflicting testing and reporting techniques for their products, often leaving regulators unable to determine if a proposed BMP meets the requirements for pollutant reductions.

Rather than causing vendors to endure great cost to test for pollutant reductions in different ways in many different states, an appropriate standard for sampling, measurement, effectiveness determination, and regulation for all BMPs should be developed as a model for communities to follow. This is part of the goal of the joint effort being undertaken by EWRI’s Urban Water Resources Research Council, EWRI’s Stormwater Infrastructure Committee, and ASTM to develop American Society of Civil Engineers (ASCE) Guidelines for Certification of Stormwater BMPs.

The first step of this work has been to develop standards to determine sediment-based TSS removal efficiency in manufactured BMPs. Significant research efforts have led to more accurate methods of testing and reporting sediments, with committee recommendations forthcoming soon. The next steps for future efforts will be to develop standards that address the organic components of TSS, nutrients, and other pollutants.

As these monitoring standards evolve, EPA, states, and communities should revise their regulations accordingly to use consistent rules for BMP measurement, allowing a more effective use of time, money, and energy by all parties to accomplish the goals of cleaning our waterways.

About the Author

Gordon England

Gordon England, P.E., D.WRE, is president of Stormwater Solutions in Cocoa Beach, FL.