Applying Whole Effluent Toxicity Testing to Aircraft Deicing Runoff (2015)

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.

5 S E C T I O N 2 To understand the potential implications of sampling airport discharges on WET testing and determine which site conditions are important to consider such that a sam- pling program that is representative of site conditions can be designed, it is important to have a basic understanding of the methods commonly used in WET testing programs. The fol- lowing section provides an overview of the regulatory context and the methods used in WET testing programs routinely used in monitoring airport discharges. Since the issuance of the “Policy for the Development of Water Quality-Based Permit Limitations for Toxic Pollutants” in March of 1984 and the subsequent publication of EPA’s “Technical Support Document for Water Quality-Based Tox- ics Control” in September of 1985 (revised in March 1991) (EPA 1991b), state agencies have been requiring aquatic tox- icity testing and developing permit limitations for municipal and industrial facilities including airports. The test objectives at airports have ranged from data collection (i.e., monitoring only without limitations) to compliance monitoring. In 1995, EPA published a final rule standardizing 17 WET test methods for use in NPDES monitoring and codifying the test methods into 40 CFR part 136 (60 FR 53529). Spe- cific test procedures for conducting the approved WET tests are included in the following test method manuals, which are periodically updated and are available on the Internet at http://water.epa.gov/scitech/methods/cwa/wet/#methods: EPA. 2002a. Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms, 5th ed., EPA 821-R-02-012. U.S. Environ- mental Protection Agency, Office of Water (4303T), Washington, D.C. EPA. 2002b. Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters to Freshwater Organisms, 4th ed., EPA 821-R-02-013. U.S. Environmen- tal Protection Agency, Office of Water (4303T), Washing- ton, D.C. Overview of WET Testing EPA. 2002c. Short-Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Waters to Marine and Estuarine Organisms, 3rd ed., EPA 821-R-02-014. U.S. Environmental Protection Agency, Office of Water (4303T), Washington, D.C. The EPA toxicity test methods describe specific steps in the conduct of an aquatic toxicity test, establish criteria for test acceptability, and are specific to each test species. The result of each test is a numeric value quantifying the toxicity of the sample. For chemical-specific permit monitoring require- ments [e.g., copper, ammonia or biochemical oxygen demand (BOD)], the test result is expressed as a measurable concentra- tion (i.e., mg/L). Because stormwater and municipal/industrial effluents are a complex mixture of compounds, toxicity test results are expressed either as a percentage of the wastewater or stormwater sample (i.e., 50% wastewater) or as toxic units (toxic units, TUs, are defined as the reciprocal of the LC50 expressed in terms of percent stormwater). Provided below is a summary of how WET tests are con- ducted, as well as important aspects of aquatic toxicity testing of airport discharges that require review. Further, a summary of the meaning and interpretation of the observed results is pro- vided. There are numerous documents and guides that provide detailed discussions on the conduct of aquatic toxicity tests and interpretation of data; however, the purpose of this section is to identify and discuss information pertinent to environmental managers at airport facilities responsible for stormwater com- pliance monitoring, reporting, and outfall permitting. 2.1 Summary of Toxicity Testing Conduct In basic terms, a toxicity test consists of the following: Preparation of Exposure Concentrations. A typical test consists of 5 exposure concentrations and a laboratory control.

6Exposure concentrations are prepared by diluting the orig- inal stormwater or effluent sample using receiving water or laboratory control water. The exposure concentrations typically range from 100% of the original stormwater to a low concentration established such that the range from high to low spans the expected exposure concentrations within the receiving water. Interim concentrations typi- cally form a geometric progression from high to low con- centrations (i.e., 100%, 50%, 25%, 12.5%, 6.25%, plus a laboratory control, see Figure 2-1). In some permits, only a single exposure concentration is required, which would consist of a specified test concentration (i.e., 100% or other concentration of stormwater) and a laboratory con- trol. Test solutions are placed in replicate test chambers, which are then placed in randomly pre-assigned positions in a test area that provides for stable test conditions. Once prepared, the test solutions are allowed to equilibrate to a specific test temperature prior to initiation of the test. Test Initiation. Test organisms are randomly placed into each exposure concentration such that each exposure cham- ber contains a known number of organisms. To improve statistical power of the test, each exposure concentration consists of multiple exposure containers. For example, the 100% exposure concentration will consist of between 2 and 4 replicate test chambers with each containing between 5 and 10 test organisms. Test Maintenance. During the test, daily observations of each test concentration and associated replicates are conducted to determine the number of organisms that are alive, dead, and/or missing. In addition, water quality conditions [pH, temperature, dissolved oxygen (DO)] are measured on a regular basis to ensure that conditions are within the range of test acceptance. Test Solution Renewal. Renewals of test concentrations, if required, are conducted at 24- or 48-hour intervals. Dur- ing renewal, subsamples of the originally collected sample or a new, freshly collected sample are used to make a new set of exposure concentrations. Once the new set of expo- sure concentrations have met water quality test conditions (i.e., temperature is within test range), the organisms are either transferred to the new solution or the solution is replaced by siphoning out the old solution and siphoning in the new solution. Test Completion. Final organism counts are conducted at the conclusion of the test. Depending on the test type, the test may be concluded after 24, 48, and 96 hours up to 168 hours (7-day test). Acute toxicity test endpoints generally include mortality or immobilization. Chronic toxicity tests also measure changes in growth or reproduction depending on the test species. If the test is a chronic test, then the numbers of young produced (reproduction) or changes in organism weight or size (growth) are calculated for each exposure concentration. Typical Freshwater Whole Effluent Toxicity Test Organisms Ceriodaphnia dubia (water flea) C. dubia (in test container) P. promelas (fathead minnow) P. promelas (in test chamber)

7 Figure 2-1. Example of test solution preparation. Acute End-Points: An LC50 (median-lethal concen- tration) or EC50 (median-effective concentration) is a statistically-based point estimate derived from the response of the test organisms to a series of exposure concentrations. The organism responses (typically death or immobility) can be represented by a dose-response curve. In general, there are three types of response as exhibited by the graphs below: Data Support Calculation of an LC50: Higher con- centrations have greater mortality than lower concentrations. At least one test concentration has >50% mortality. Sample “Non-Toxic and Data Do Not Support” Calcu- lation of an LC50: No mortality or observed mortality is less than 50% within each exposure concentration. Dose/Effect Curve Dose/Effect Curve Test Fails to Meet Quality Control Requirements: High Toxicity (60% mortality) in Control Sample and unexpected dose-response curve. Dose/Effect Curve

8required test conditions are outlined in the test protocols and consist of: – Sample holding time (must be <36 hours). – Test temperature range (must not deviate by more than 3°C during the test). Note: different species have differ- ent temperature requirements which are identified in the US EPA protocols. – Minimum levels for dissolved oxygen must be main- tained (>4.0 mg/L for warm-water species, >6.0 mg/L for cold-water species). • Test organisms. Because organism sensitivity may change with age, standard ages for test organisms have been established: – C. dubia <24 hours old. – P. promelas <14 days but all within 24-hour window. – Each exposure concentration should consist of at least 20 organisms. • Control survival. The use of a test control ensures that the organisms are healthy and test results are not affected by organism handling. Control survival must be at least 90% to consider the test valid. • Test solution. To minimize build-up of waste products in the test water, test solutions must be renewed every 48 hours. • Concentration–response relationship. In general, the higher the test concentration, the greater the response. The exact The test results may be expressed in several ways, depend- ing upon the purpose of the test and the requirements of the permitting agency. For acute toxicity tests (short-duration tests generally ranging from 24 to 96 hours), the most com- mon toxicity estimate is the median-lethal concentration (LC50) or median-effective concentration (EC50). The LC50 or EC50 point estimates represent the effluent concentra- tion that is estimated to result in lethality or measured effect to 50% of the exposed test population. Often per- mitted limits are expressed as TUs rather than as percent stormwater. Toxic units are a measure of effluent toxic- ity; acute toxic units (TUa) are defined as the reciprocal of the LC50: ( )= 1 %50TU LCa This convention is utilized such that a higher number of TUs indicates a greater level of toxicity and can be thought of in terms similar to chemical concentration. The higher the TUa number, the greater the toxicity; this is analogous to chemical concentrations such that the higher the concentra- tion, the more chemical that is present in solution. Chronic toxicity tests are longer-duration tests and reflect more sensitive endpoints such as growth and reproduction. As a result, statistical analyses associated with chronic tests include both tests for differences between the control sur- vival, growth, and reproduction and point estimates to deter- mine when survival, growth, and reproduction is inhibited by, for example, 25%. Chronic toxic units (TUc) are calculated similarly to TUa calculations as the reciprocal of the chronic endpoint [no observed effect concrentation (NOEC), chronic value (ChV), or other metric] such that the higher the TUc, the greater the chronic toxicity. 2.2 Required and Recommended Toxicity Test Conditions In the aquatic toxicity testing protocols, EPA has established a number of required and recommended test conditions and has developed a series of test review parameters. Although each laboratory should have an internal quality assurance and quality control (QA/QC) program, the following items should be reviewed by the airport environmental manager and/or person responsible for compliance. These items consist of the following: • Sampling and handling. Samples shall be stored between 0°C and 6°C until ready for use. Time from sample collection to test initiation shall not exceed 36 hours. • Test conditions. EPA has established specific test condi- tions or ranges. Some of the test conditions are noted as “required” while others are noted as “recommended.” The Discharges to Saltwater or Estuarine Environments For discharges to saltwater or estuarine receiv- ing waters, saltwater organisms are typically required for testing. However, the stormwater discharge is likely to have a very low salinity; thus, the salinity of the stormwater sample must be adjusted upwards through the addi- tion of commercially available sea salts to the sample to achieve the desired salinity. When this is required, it is recommended that a freshly prepared synthetic seawater control be utilized. Studies have demonstrated that the addition of synthetic sea salts may affect organism survival, growth, and reproduction [State Water and Resources Control Board (SWRCB) 2000, Pace and Arnold 1993]. Alternatively, hypersaline brine may be utilized to adjust the salinity of the stormwater discharge; however, this reduces the maximum test concentration due to dilution of the stormwater with the hypersaline brine.

9 monitored during the conduct of the test and to ensure that proper DO levels were maintained or that actions described by EPA in the Acute Toxicity Test Manual (EPA 2002a) were implemented. If low DO and mortality are concurrently observed in a test, it will be unclear whether the mortality was the result of the toxicant working directly on the organism, if the observed low DO contributed to the observed mortality, or if the low DO occurred due to the decomposition of the dead organisms. The purpose of reference toxicant testing is to evaluate the health and sensitivity of organisms over time and ensure that laboratory procedures do not affect the results. Since the con- trol limits that define the range of acceptable results are a statistical calculation, approximately 1 out of every 20 tests will fall outside of the acceptable control limits. Because of this, exceedance of these control limits is not a definite reason for rejecting test results. Should this occur, the extent of deviation should be considered (e.g., how much above/below the control limits was the result) and the recent trend in reference toxi- cant testing should be considered (e.g., were the last 3 tests all trending in the same direction). If the results of the reference toxicant test fall outside of the acceptable range, the results of the compliance test, when reported, should indicate that not all laboratory QA/QC requirements were met. 2.3 Test Interpretation Common objectives for aquatic toxicity testing on airport stormwater discharges are to 1) collect data on the toxicity and variability of the discharges, 2) confirm that stormwater does not have a reasonable potential to contribute to aquatic toxicity within the receiving water, and 3) comply with per- mit monitoring requirements. To facilitate test interpreta- tion, it is beneficial to collect the following information at the time of sample collection: Stormwater flow at the time of sampling. Receiving water flow at the time of sampling. Stormwater composition/quality (note, data such as ammo- nia, pH, and conductivity are typically collected as part of aquatic toxicity test procedures). Additional data that are not typically collected as part of an aquatic toxicity test but may provide insight as to the source or characteristics of the toxicity include: – Chemical oxygen demand (COD). – 5-day biochemical oxygen demand (BOD5). – Conductivity and ion concentration (calcium, sodium, potassium, chloride). – Total suspended solids. – pH. Over time, the collection of these data allows unusual events to be identified and potential correlations with stormwater shape of the concentration–response curve can be vari- able; however, the data should be reviewed to determine if higher exposure concentrations indicate higher toxicity. If the curve indicates otherwise, this may be indicative of an anomalous result and retesting is recommended. • Reference toxicant testing. Reference toxicant tests are conducted on a monthly basis using a single and consis- tent toxicant to demonstrate that test procedures and test organism populations provide consistent and repeatable results. The laboratory should report the results of the monthly reference toxicity test and the associated control/ acceptance limits to demonstrate that organism sensitivity has not changed over time and that the test protocols and associated results are consistent and reproducible. Upon receipt of the results of an aquatic toxicity testing report, the above information should be reviewed to deter- mine acceptability of the test. Specifically, test data should be reviewed to confirm that water quality conditions were monitored daily and fall within acceptable ranges (e.g., DO >4 mg/L, temperature in range and exhibits a range of less than 3°C, etc.), the dose–response relationship shows an increasing toxic response with increasing exposure concen- trations (assumes that toxicity is observed), control survival was greater than 90%, and a standard reference toxicant test was conducted for the test species and the results were within the acceptable range. An example of a toxicity test report with critical information identified is provided in Appendix B. Should any of the data fall outside of the required range, the results of the toxicity test can be considered suspect. For example, if mortality in the laboratory control exceeds 10%, this may be indicative of improper test conditions, contami- nation of test containers, poor organism handling technique, stressed/diseased test organisms, or many other issues. Thus, the resulting toxicity test value should be considered invalid. As noted in the EPA Methods for Measuring the Acute Toxic- ity of Effluents and Receiving Waters to Freshwater and Marine Organisms (EPA 2002a), “the DO in the test solution should not be permitted to fall below 4.0 mg/L for warm-water spe- cies and 6.0 mg/L for cold-water species.” Further, EPA notes that “samples with a potential DO problem generally show a downward trend in DO within 4 to 8 h after the test is started. Unless aeration is initiated during the first 8 h of the test, the DO may be exhausted during an unattended period, thereby invalidating the test.” Airport stormwater discharges from areas of pavement and aircraft deicing operations may contain elevated concentrations of oxygen demanding substances and are likely to exhibit decreasing DO concentrations over time as the constituents degrade. Based on the above discussion and EPA guidance, if the DO falls below 4.0 mg/L in the test, the test should be considered invalid. Thus, it is important to review the DO concentrations

10 utilized. The discharge flow rates used in the calculation of permit limits are state-specific and can range from the flow produced from a 2-yr, 24-hour storm event to those flows pro- duced from a 25-yr 24-hour event or greater. Similarly, receiv- ing water flow rates are based on hydrologically-based flow statistics such as the 7Q10 (the lowest consecutive 7 day aver- age flow which occurs once every 10 years) flow. These metrics are utilized in combination with state criteria to determine permit limits. Typically, acute limits for both aquatic toxicity and acute ambient water quality criteria are applied at the end of the pipe or at the edge of the zone of initial dilution (ZID). Chronic limits for toxicity and ambient water quality criteria are typically calculated assuming a certain level of dilution in the receiving water. Determination of actual flow conditions allows for the determination of stormwater concentrations at the edge of the ZID or mixing zone. These concentrations can be compared to the observed toxicity to determine if there is a potential for toxicity within the receiving water. For example, if a ZID is allowed for compliance with the acute toxicity discharge limit and a dilution of 1:10 is obtained at the edge of the ZID, the resulting stormwater concentra- tion at the edge of the ZID would be approximately 10%. If the observed LC50 of the stormwater discharge was measured at 75% effluent, then it could be concluded that there is a low probability for contributing to acute toxicity in the receiving water. toxicity to be established. For example, plots of historical BOD concentrations and aquatic toxicity data may indicate a relationship between toxicity and BOD concentration (Fig- ure 2-2). Specifically, this data shows that when BOD concen- trations exceed approximately 4,000 mg/L, effluent toxicity increases. However, the data do not indicate the source of the increased BOD. Typical sources of BOD at an airport during the deicing season consist of aircraft and pavement deicing fluids; however, other sources such as leaky or broken sanitary sewer pipes should also be considered. In addition, further inspection of the data below indicates that on one occasion, another source of toxicity may have been present. Specifically, one sample with a BOD concentration of <2,000 mg/L was observed to be toxic. This data point is inconsistent with his- torical monitoring which indicates that samples are generally non-toxic when the BOD is less than 4,000 mg/L. Based on this analysis, more detailed investigation of each sampling date can be conducted. This investigation should consider weather conditions, total usage of deicers at the facility during sample collection, and other site-specific con- ditions. This review will assist in making modifications to best management practices such that conditions leading to effluent toxicity can be directly addressed. In the determination if there is a potential to contribute to instream toxicity, information on the discharge flow rate, receiving water flow rate, and monitoring data results are Figure 2-2. Comparison of observed toxicity with propylene glycol concentration.

11 How Discharge Limits Are Derived The basis for the permit limitations or requirements with respect to whole effluent toxicity testing should be understood and inform the sampling program. In the calculation of permit limits, the following factors are typically considered: 1) presence of a mixing zone, 2) receiving water flow, and 3) discharge flow. Thus, the permit writer must make assumptions or calculations with respect to each of these parameters. State regulations may or may not allow mixing zones. In general, mixing zones allow the discharge of acutely or chronically toxic effluents as long as they are diluted to non-toxic concentrations within a short distance from the point of discharge after mixing with the receiving water. Understanding these state-specific requirements can assist the airport environmental manager in developing sampling programs and interpreting the resulting data. Receiving water flows are utilized to determine the total amount of water available for mixing with the storm- water discharge. Hydrologically-based flows such as the 7Q10 (the lowest consecutive seven day average flow, which occurs once every 10 years) or 1Q10 (the lowest single day flow, which occurs once every 10 years) are typically utilized. In contrast to continuous industrial or municipal discharges, airport stormwater discharges occur during and/or just after precipitation events. Thus, it is unlikely that the receiving water flows are at or below the 1Q10 or 7Q10 flow condition. Finally, like the receiving water flow estimate, the discharge flow is also utilized to determine the extent of dilu- tion within the receiving water and therefore the size and frequency of the storm event must be determined. Statistically-based storm event conditions, such as the 2-yr or 10-yr 24-hour storm depth may be utilized to cal- culate the estimated stormwater discharge volume. However, these event statistics do not consider the type of precipitation and may be skewed by more extreme, episodic events such as hurricanes. Winter and summer storm events are further differentiated by the form of precipitation. For example, a 10-inch snow event (approximately equivalent to 1-inch of wet precipitation) will result in a different hydrograph compared to a 1-inch rain event even though both events are the result of the same liquid volume of precipitation. Specifically, precipitation that falls in the form of snow is not immediately available for runoff and will be retained on the site until melting con- ditions occur. As a result, assumptions with respect to the co-occurrence of both receiving water flow and storm- water flow may result in a condition unlikely to be observed in the field (i.e., 7Q10 flow condition coupled with a 10-yr stormwater discharge event).