Does Your Stormwater BMP Make the Grade?Options for testing BMPs and understanding the results

The pollutants in stormwater are a complex chemical and physical soup of materials that are mixed together in numerous ways. Chemicals are adsorbed onto and into solids, and particles stick to larger objects and each other. Removing the bulk of particles and larger solids actually will greatly reduce all constituents in the water that tend to bind to them in some fashion.

Stormwater Fig5 1211ww

By John Moll

The past 20 years have shown us that quantifying water quality effectiveness is probably not a bad practice. In guidance provided by the USEPA for improving water quality, the base standard recommended is the removal of 80% of the total suspended solids (TSS) load, which is allowed as a surrogate for actually removing individual pollutants. There are a few exceptions where individual constituents such as phosphorus are the basis for BMP approval but by and large, "80% TSS removal" is the mantra of water quality professionals. Other standards exist because "guidance" is not law, but most regulators have long forgotten that a TSS standard is not to be found in the actual regulations.

The pollutants in stormwater are a complex chemical and physical soup of materials that are mixed together in numerous ways. Chemicals are adsorbed onto and into solids, and particles stick to larger objects and each other. Removing the bulk of particles and larger solids actually will greatly reduce all constituents in the water that tend to bind to them in some fashion.

If we accept removing solids as a surrogate for removing all constituents, we must find a way to measure the effectiveness of a BMP at removing those solids. To determine that actual percentage of solids removal, we have to know exactly how much material the BMP captured, and exactly how much material the BMP was unable to capture and which entered the watershed downstream. This may seem simple but it isn't.

Stormwater Fig5 1211ww
Neglected BMP catching oil and trash.

For example, how would you measure how much material a wet pond had caught during a test period? For a manufactured vault, you could remove the material that it caught and weigh it, but for a pond this is not possible. Even with a vault, there are questions: Do you count a running shoe and a tail light lens that washed into the device as "TSS"? Assuming you can get a true measure of solids for both the pond and the vault, how would you measure what each BMP missed? Over the past 20 years, we have made significant progress in how we try to measure the effectiveness of BMPs but some basic challenges remain.

Stormwater Fig6 1211ww
BMP not cleaned for three years.

A typical field testing protocol will measure 15 to 20 storm events defined as meeting a certain rainfall depth and/or time duration. Storms have to be large enough to allow for sampling in the pipe or conveyance; thus, small events are typically not sampled. Very large storm events will normally produce flows that must bypass the BMP; there, again, no sampling is possible. The smallest storms may carry the bulk of dissolved materials, and the largest storms will certainly carry the largest materials, but neither type will be a part of the analysis.

Stormwater Fig4 1211ww
Neglected pond with invasive growth.

If we accept the storm definition, we then need to consider how the water is sampled. Are they grab samples or are they taken by an automated sampler? Where in the water column are the samples taken? We do not know in advance how the rate of rainfall will change during the storm, so how would we determine when to take the influent samples to make them representative of the whole storm? Similarly, the nature of the influent and the effectiveness of the BMP will change over time, so determining how and where to take an effluent sample cannot be known in advance.

Given that we believe that we have taken the samples correctly and at the right times, what is the correct amount to sample? Can we take a few liters of samples and declare them as representative of the thousands of gallons of water that flowed through the BMP during the storm events?

Stormwater Fig1 1211ww
BMP test setup at Alden Research Laboratory
(Shrewsbury, MA)

Finally, if we make all these assumptions correctly, then can we take this test in one watershed that measures a few storms and declare it valid for every site? Every climate? Every watershed? Every time?

This seems like an implausible way to assess a BMP and, in fact, it is. So, what are we to do?

Lab Testing

Over the past 20 years, we have made great strides in improving laboratory testing. It seems to offer a ready solution but for a long time, it lacked a reliable standard. In the past, various techniques were being used to conduct laboratory testing and the test data could not be compared. Results could be manipulated by varying the size of the material tested, the concentration, the sampling methods, water temperature, and many other factors. Recently, however, an ASTM standard has been published that takes the variables out of laboratory testing.

Stormwater Fig 2 1211ww
Sampling points at Alden Research Laboratory
(Shrewsbury, MA)

The idea behind a strictly controlled laboratory testing protocol is to be able to test the same BMP at two different laboratories and get the same results by using the exact same methods. If two different BMPs are tested by a standardized protocol, then their effectiveness can be accurately compared. Clearly, a standardized ASTM testing protocol is a huge step forward.

Testing under controlled laboratory conditions would seem to be the ideal solution for BMP evaluation but some valid concerns still remain. The fact that most land-based BMPs, such as a wet pond, cannot be tested in a laboratory is the most obvious drawback. Another restriction is the limited ability of most laboratories to deliver high flows of clean water, such as might be needed to test a large land-based BMP.

Stormwater Fig3 1211ww
Pond with trash and nutrients built up.

While these are clear physical limitations, there are more subtle challenges that exist for lab-tested BMPs. In a laboratory, the size and concentration of the particles fed into the BMP is strictly controlled and consistent. In nature, however, the material entering the BMP is variable from moment to moment, both in particle size and in concentration. In the laboratory, the particles fed into the BMP are typically silicon dioxide-based material with a relative density of 2.65. In the field, we know that the average density of the material captured by BMPs has proven to be about 1.70, and it consists of materials that vary from heavy metals — such as wheel weights — to natural sands, silts and clays, to plastics, trash and organics, to very light floating materials such as Styrofoam beads. While the laboratory measures performance (and sediment retention) at steady flow rates, natural flow rates vary greatly from moment to moment and from storm to storm. Laboratory testing is a good thing but it should not be mistaken for reality, nor should a "percentage of removal" be assigned based strictly on laboratory results.

The State of the Art

Today, we define the removal of solids as a surrogate for actually controlling pollution. We choose to trust testing to "verify" the performance of BMPs. Testing certainly does give some indication of what a BMP may or may not do but the real answer to their effectiveness is right at hand. All we need to do is look in the BMPs once they have been in service for awhile.

We will see ponds that are impacted with excessive sediments and are full of trash. Standing pools that are supposed to draw down in 24 to 72 hours will appear semi-permanent, full of algae, mosquitoes or invasive plants. They may be contaminated with hydrocarbons or other noxious materials. We may see a manufactured vault with an trash rack or clogged filters. Believe it or not, these observations are actually good news overall: the pond is capturing sediments and trash; nutrients and oil are being intercepted; a manufactured vault can have its contents removed and be like new. Almost all of these "failures" can be corrected by cleaning and maintenance. The bottom line is that if we simply look at what we are doing and diligently record the results, we will quickly see what works and what does not.

About the Author: John Moll is the CEO of CrystalStream Technologies and is an active researcher in the water quality field. He holds eight water quality-related patents, and has incorporated a maintenance company into CrystalStream that has provided invaluable data about actual field performance of BMPs.

More in Lab / Sampling / Analytical