 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 
|
Fish toxicity testing in Puget Sound Car Wash Runoff The impact of chemicals in car wash runoff was measured by using the runoff in fish toxicity tests by assessing the death rate in juvenile rainbow trout. This seemed to be a very practical way to show how harmful car wash runoff effluent can be in our streams. It should be noted that chronic effects (reduced fertility and lower than average weight or size) impacting the aquatic environment will occur at much lower concentrations than indicated in the acute test presented in this article. Brown Bear Car Wash of Seattle, WA, operates a chain of car washes in the Puget Sound vicinity. Environmental Partners Inc. (EPI) of Issaquah, WA, is the environmental consultant for the company. There are little, if any, reliable data to assess the stormwater loading of a typical driveway car wash event. This study was sponsored by Brown Bear Car Wash to develop a more reliable empirical data set to help evaluate this impact. Brown Bear Car Wash did not dictate the test procedures or otherwise influence the design or outcome of the study. Test Description The same detergent concentrate was used in water samples for both tests. Juvenile rainbow trout were used in both tests and both tests were conducted according to standard protocols specified in “Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms” (EPA-821-R-02-012). The tests were performed by an experienced, certified laboratory. The tests produced similar results. The first test indicated a percent concentration that was lethal to 50% of the test organisms (LC50) of 3.1%. The second test indicated an LC50 of 3.0%. There were significant differences in the way the stock water solutions for the two tests were prepared. For the first test, runoff water was collected from the parking lot of an automotive service facility during a fund-raising event. This water ran across approximately 30 feet of asphalt before collection and likely included contact with petroleum hydrocarbons and the grit and grime typically associated with a heavily traveled asphalt lot. Approximately 15 gallons of this water were sampled and delivered “as collected” to the laboratory. (Note: The youth organization used a car wash kit supplied by King County that prevented the effluent water from entering the storm drain. Effluent water was collected by a catch basin insert and pumped to a sanitary sewer drain.) For the second test, the same detergent concentrate that was used for the car wash event was used by the laboratory to prepare a simulated effluent for testing. This simulated effluent was mixed according to instructions on the product container and was further diluted to simulate addition of rinse water. All water used in the second test was potable. These tests are termed “practical” fish toxicity tests because the effluent solutions for both were collected or prepared such that each represented the actual runoff water that would be expected to enter into stormwater drains and, eventually, the streams and rivers of Puget Sound. The tests were not run simply to determine the lethal concentration of a pure chemical or to satisfy a discharge permit requirement. As such, the results of these tests represent one piece of evidence that points directly to the impact of wash water from residential driveway or fund-raiser car washes that enters storm drains emptying into water bodies containing threatened and endangered salmon. Car Wash Effluent Fish Toxicity Test Prior to test start, dissolved oxygen, pH, conductivity, and temperature of the test waters were measured in each test chamber to ensure parameters were within acceptable limits (prescribed by EPA method guidance). Water-quality measurements and survival observations were made daily.
The car wash effluent water caused 100% mortality in all concentration steps tested. Complete mortality occurred within 24 hours of test start. Survival of the laboratory control was 100%. Results are presented in Table 1. The calculated LC50, the concentration of sample that is expected to cause mortality in 50% of the select population of organisms, was 3.125% due to the complete mortality observed in the lowest concentration tested (6.25%) and the 100% survival observed in the laboratory control (0%). Another measure of toxicity is called Toxic Units (TU = 100/LC50). TU measurement is typically a specified criterion for discharge monitoring permits. For this case, the Acute Toxic Unit (TUa) result was calculated to be 32, meaning that the tested effluent is 32 times more toxic than an acceptable effluent.
The test was aerated at initiation due to low dissolved oxygen levels [4.3 milligrams per liter (mg/L)] in the received sample car wash water. Dissolved oxygen levels remained within protocol limits for the duration of the test. The results of an associated reference toxicant solution using copper sulfate fell outside the 95% confidence limits of the historical laboratory mean. This indicated that the organisms tested might have been less sensitive to concentrations of copper than typical populations. Since complete mortality was observed in all concentrations of car wash effluent, this reference toxicant deviation had no impact on test results.
Simulated Effluent Fish Toxicity Test Prior to test start, dissolved oxygen, pH, conductivity, and temperature of the test waters were measured in each test chamber to ensure parameters were within acceptable limits (prescribed by EPA method guidance). Water-quality measurements and survival observations were made daily.
The simulated effluent solution caused 100% mortality in the 10% concentration solution and 2.5% mortality in the 1% concentration solution. All mortality at the 10% concentration occurred within 24 hours. Survival rates were 100% for all other series concentrations. Survival of the laboratory control was 100%. Results are presented in Table 2. The calculated LC50 was 3.046%, which equates to a detergent concentrate concentration of approximately 1.6 parts per million (ppm). The test was aerated at initiation and during its duration due to low dissolved oxygen. Dissolved oxygen levels remained within protocol limits for the duration of the test. The results of an associated reference toxicant solution using copper sulfate fell within the test 95% confidence limits of the historical laboratory mean.
Toxicity Test Water Samples Cars were washed on an asphalt surface at an oil change service facility. The asphalt condition was typical of a parking lot; its surface had numerous dark spots indicating leaks of petroleum product. Wash and rinse water that dropped to the asphalt ran about 30 feet across the asphalt to a storm drain grate. The 30-foot traverse was across a driveway of the facility. The event was held on a sunny September day. The people running the event were using a King County–supplied car wash kit that consisted of an impervious plastic tub, small electric pump, and hose. The plastic tub fit into the storm drain opening and prevented water from going down the drain. It collected the wash water, which was pumped through a hose to an onsite sanitary sewer drain. The car wash effluent water sample was collected from the hose prior to discharge to the sewer. The sample was cooled to 4oC and delivered to the test laboratory. The simulated effluent solution for the second fish toxicity test used the same detergent that was used during the car wash event. The solution was prepared using directions printed on the product container and was further diluted to simulate the addition of rinse water. All water used in the second test was potable. Based on product label directions, approximately 16 milliliters of detergent concentrate were mixed with 4 gallons of water to make the wash solution. This wash solution was diluted by a factor of 20 to mimic the addition of rinse water to produce a concentration of approximately 53 ppm that was the simulated effluent solution used to prepare the dilution series for the second fish toxicity test (i.e., 10%, 1%, 0.5%, 0.1%, 0.05%, and 0.01% of the effluent sample). An analysis was made of summertime stream flows for several small creeks and streams in King County that flow into Puget Sound, Lake Washington, and Lake Sammamish. Although flows were highly variable depending on stream size and recent weather, a typical range of summertime flow was about 2 to 10 cubic feet per second (cfs), equivalent to 900 to 4,500 gallons per minute (gpm). This range of stream flow rates was compared to an assumed flow of water from two hoses running at 5 gpm each that was assumed to be typical of a fund-raiser car wash event. The ratio of car wash effluent to stream flow was about 1/100 (0.01 or 1%) to 1/1,000 (0.001 or 0.1%). This analysis was used to bracket the range of the dilution series performed by the laboratory for the second fish toxicity test. Thus, the concentration of the simulated effluent and the dilution series used for this toxicity test represent realistic conditions. Organisms living and swimming in small creeks and streams around northwest lakes and flowing into Puget Sound would likely be exposed to car wash detergent concentrations that were used in both fish toxicity tests reported here.
Fish Toxicity Test Results Because the car wash effluent used in the first toxicity test was generated in an uncontrolled manner, it is not possible to make conclusive remarks about the LC50 results of the toxicity test. This is because the amount of detergent and water used was not measured; hence, detergent concentrations in the dilution series were not known. Also, no chemical analyses were performed to determine petroleum hydrocarbon or metals concentrations in the effluent. Nevertheless, the effluent water sample was collected from an actual fund-raising car wash event and the effluent water represented an actual potential impact to a local stream. On the other hand, the laboratory-prepared simulated effluent solution used in the second fish toxicity test used measured quantities of detergent and water, which allowed exact calculation of detergent concentrations in the dilution series water. Uncertainties associated with this test include lack of exposure to a petroleum-contaminated asphalt parking lot and lack of exposure to grime from a dirty car. The similarity of LC50 results is unexpected. There is no way to know if this similarity indicates true replicability or is merely coincidental. The common feature between the two tests was the use of the same car wash detergent concentrate. This concentrate is a commercially available product marketed specifically as a car wash detergent. As indicated by the second test results, a detergent concentration of approximately 1.6 ppm is sufficient to kill one-half of a population of juvenile rainbow trout. In the first toxicity test the car wash effluent solution was fatal to all specimens tested within 24 hours down to the minimum dilution tested of 6.25%. Because the simulated effluent solution for the second test was prepared in the laboratory, it is reasonable to assume that the fish mortality was due solely to the effect of the chemicals in the car wash concentrate. The most likely chemical that could be found in such a product that would be toxic to fish is a surfactant or mix of surfactants. The exact physiological impact of a surfactant chemical on the fish is unknown in this case. The chemical could be toxic by simple ingestion, could affect the surface chemistry of fish gills and thereby asphyxiate fish, could disrupt or destroy cell membranes, or could produce some other lethal effect. Other research in this area has indicated that detergents as a rule will destroy fish mucus membranes and gills to varying degrees. Natural oils may be washed away, affecting oxygen uptake by the gills. The damaged mucus membranes make fish more susceptible to organic chemicals such as petroleum and pesticides and inorganic chemicals found in fertilizers. Thus, smaller concentrations than predicted of these chemicals may become chronically toxic to fish. Some surfactant chemicals in detergents have been shown to break down into more toxic compounds and to mimic natural hormones in fish, causing abnormal growth and development and therefore lowering survival rates. Material Safety Data Sheets (MSDSs) for the detergent concentrate were obtained but revealed little about the chemical constituents of the product. The MSDS for the product tested listed only the constituents “water” and “surfactant (mixture).” The surfactant was indicated to be at a concentration between 5% and 20%. No ecological information was presented in the MSDS. The only precautions listed were to avoid eye contact (“May Cause Eye Irritation”), likely due to a listed pH of 9. MSDSs for similar car wash products marketed by the same vendor indicated a few chemical compounds. Among those listed for similar products were the following:
Ecotoxicity information for the first of these chemicals indicates moderate toxicity to fish, high toxicity to nematodes and flatworms, and slight toxicity to crustaceans and zooplankton. The chemical use is listed as microbiocide, adjuvant, fungicide, and insecticide. Puget Sound Setting Puget Sound is the second largest estuary in the United States. It has 2,300 miles of shoreline. The Puget Sound watershed covers nearly 16,500 square miles and consists of more than 10,000 rivers and streams that drain into the sound. All but a tiny fraction of stormwater that falls on developed areas enters storm drains and flows untreated into the sound. Over 80% of the surface water flowing into Puget Sound comes from the following major river drainages: Cedar River (Lake Washington), Green/Duwamish, Elwha, Nisqually, Nooksack, Puyallup (White), Skagit, Skokomish, Snohomish, and Stillaguamish. In King County, the major river drainage systems are the White (Puyallup) River, the Green/Duwamish River, the Cedar River (Lake Washington), the Sammamish River, and the Skykomish/Snoqualmie rivers. As of 2006, the number of registered vehicles in Washington was approximately 5.6 million. There are approximately 3.7 million vehicles in the Puget Sound area, and about 1.7 million of those are in King County.
Test Result Hypothetical Implications Approximately 100,000 people are assumed to live in the watershed area. One percent of the cars of the population are washed in driveways during the time period. A consumer car wash detergent is used to wash the cars, and 75 gallons of water flows to the storm drain and, subsequently, to the small stream for each car washed. Calculations indicate that within this watershed approximately 1,000 vehicles will be washed in driveways during the weekend. The 75 gallons of car wash effluent per vehicle contain 53 ppm of detergent. A simple “bathtub” calculation was performed in which all of the stream flow and car wash effluent was pooled for the 48-hour period and the resulting detergent concentration calculated. The calculated detergent concentration ranged from 0.2 ppm to 1.5 ppm for high and low stream flow conditions, respectively. These detergent concentrations are similar to the 1.6-ppm value that was found to be lethal to 50% of juvenile rainbow trout tested. Thus, some fish in the stream could be killed and it would be likely that the detergent would wash protective mucus from the gills of some surviving fish. The surviving fish would thus be more susceptible to other contaminants that may exist or be introduced into the stream. It is also possible that oxygen uptake necessary for fish survival may be impaired and that other physiological impacts to fish survival may occur. Other freshwater organisms living in the stream would also likely be affected, depending on individual species’ sensitivities. Minor changes to the assumptions made in the above analysis drive the calculated detergent concentration to much higher values and make significant impacts to fish and other freshwater organisms more likely. For instance, increasing the percentage of cars washed from 1% to 1.5% increases the total amount of detergent flushed to the stream by 50% and raises the calculated detergent concentration in the stream to 2.2 ppm for the low-flow situation (i.e., 2 cfs). Dilution by the stream is the most important factor in the calculated detergent concentration. Conclusion Jeff Dengler, Ph.D., P.E., is a senior engineer and John Brasino, Ph.D., P.E., LG, LHG, is president and principal engineer with Environmental Partners Inc. in Issaquah, WA. SW October 2007 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
Home + About + Subscribe + News + Calendar + Glossary Distributed Energy | Erosion Control | Grading & Excavation Contractor | MSW Management | Onsite Water Treatment | Water Efficiency | StormCon | ForesterPress | Forester Media
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||
