API PUBL 4751-2005 Evaluation of Water Quality Translators for Mercury《水银水质评价》.pdf

1、Evaluation of Water Quality Translators for MercuryRegulatory Analysis and Scientific AffairsPUBLICATION 4751DECEMBER 2005Evaluation of Water Quality Translators forMercuryPrepared for the American Petroleum Institute by: ARCADIS 24 Preble Street, Suite 100 Portland, ME ACKNOWLEDGMENTS THE FOLLOWING

2、 PEOPLE ARE RECOGNIZED FOR THEIR CONTRIBUTION OF TIME AND EXPERTISE DURING THIS STUDY AND IN THE PREPARATION OF THIS REPORT: API STAFF CONTACT Roger Claff, Regulatory Analysis and Scientific Affairs MEMBERS OF THE CLEAN WATER ISSUES TASK FORCE John Cruze, Chairman, ConocoPhillips John King, Vice Cha

3、irman, Marathon Ashland Petroleum LLC Jeffrey Adams, BP America Incorporated Gregory Biddinger, ExxonMobil Corporation Mickey Carter, ConocoPhillips Richard Cuhna, ExxonMobil Refining and Supply Peter Dahling, Chevron Corporation Rees Madsen, BP Refining Shared Services Sandy Martin, Shell Chemical

4、Company Incorporated David Pierce, Chevron Corporation Energy Technology Company Jeff Richardson, BP PLC Kim Wiseman, Chevron Corporation Jenny Yang, Marathon Oil Company David Zabcik, Shell Oil Products US ABSTRACT This report discusses the technical issues and constraints associated with translati

5、on of a mercury fish tissue concentration into a water quality criterion, in the use and implementation of the Environmental Protection Agencys fish-tissue-based criterion for methylmercury (0.3 mg methylmercury/Kg wet weight fish tissue). The report focuses on available analytical methods for evalu

6、ating mercury in fish and water; proposed methods for translating a fish tissue concentration for mercury into a concentration in water; and implementation of the mercury criterion in the development of Total Maximum Daily Loads (TMDLs) and water quality-based effluent limits (WQBELs). The approache

7、s to criteria translation are, in order of preference: (1) derive site-specific bioaccumulation factors (BAFs), (2) use a bioaccumulation model, or (3) use EPAs national default translators. The collection of site-specific data allows for the most accurate assessment of bioaccumulation; however, val

8、idation of methylmercury analytical techniques is necessary to increase the certainty of results. Models have the potential to account for environmental factors contributing to data variability, but at present the available models are limited to reservoirs and lakes in a few geographic regions. Impr

9、ovements in national default translators do not decrease the importance of site-specific translators. National default values are likely to be inaccurate on a site-specific basis, given the very high degree of variability observed in mercury bioaccumulation rates. Research is needed to improve the n

10、ational default translators currently proposed by EPA, and additional data would increase the effectiveness of the translator calculation methods by reducing variability and minimizing the uncertainty of the resulting default values. Given the many uncertainties associated with mercury translators,

11、their use should be limited to cases where site-specific fish tissue data reveal the tissue-based water quality criterion has been exceeded, and point sources make up a significant contribution of the total mercury loading to the water body. i CONTENTS Executive Summary . 1 1 Introduction . 3 2 Anal

12、ytical Methods 4 2.1 Methods for Evaluating Total Mercury in Fish Tissue and Water 4 2.1.1 Method 1631. 4 2.1.2 Method 7471B/SW-846 4 2.2 Methods for Evaluating Methylmercury in Fish Tissue and Water 4 2.2.1 Method 1630. 5 2.2.2 UW-Madison SOP 5 2.2.3 USGS Method 5 2.3 Clean Hands Sampling . 5 2.4 C

13、onsiderations in the Selection of an Analytical Method. 5 3 Translators Issues. 6 3.1 Overview of Translations Within and Between Media. 6 3.2 EPA Proposed Translation Methods. 10 3.2.1 Bioaccumulation Models for Mercury 10 3.2.2 National Default Translators. 11 3.3 Site-Specific Translators. 13 4 U

14、se of Translators in TMDL and Permit Limit Calculations. 15 4.1 Introduction to WQBELs and TMDLs . 15 4.2 Listing Issues that Trigger TMDLs 16 4.3 Evaluation and Comparison of Mercury TMDL Targets 17 4.4 Allocation Approaches . 18 4.5 Implementation Approaches. 18 4.5.1 Reasonable Potential. 19 4.5.

15、2 NPDES Implementation Procedures. 19 4.5.3 Development of WQBELs 20 4.5.4 Other Options for Implementation 21 5 Applicability to Other Metals 22 6 Summary and Recommendations 24 7 References . 25 Appendix A Bibliography of Selected Recent Studies Available to Update Default Translators for Mercury

16、A-1 Appendix B Example Mercury TMDLs. B-1 Figures 1 Conceptual Overview of Mercury Translations 7 2 Total Mercury versus Methylmercury in Stream Water Samples Collected Throughout the United States as Part of the NAWQA Program 8 ii Tables 1 Summary of Key Mercury Ratios within and Between Media, as

17、Compiled by EPA (2000a,b) 9 2 Estimated Laboratory Costs for Mercury SamplingScenario 1 14 3 Estimated Laboratory Costs for Mercury SamplingScenario 2 15 4 Proposed Approaches to Implementation of Fish-Tissue Based Criterion in Permits 20 iii LEGEND AVS acid volatile sulfide BAF bioaccumulation fact

18、or CMC Criteria Maximum Concentration CWA Clean Water Act DOC dissolved organic carbon EPA U.S. Environmental Protection Agency ELG effluent limitation guideline FAV Final Acute Value FCV Final Chronic Value fddissolved methylmercury as fraction of total mercury in water column KDpartition coefficie

19、nt, dissolved methylmercury to particulate mercury MS/MSD matrix spikes and matrix spike duplicates NAWQA national ambient water quality assessment NPDES national pollutant discharge elimination system NIST National Institute of Standards and Technology OPR ongoing precision and recovery PAH polycyc

20、lic aromatic hydrocarbon PBT persistent bioaccumulative toxic POTW publicly owned treatment works QA quality assurance QC quality control RPA reasonable potential analysis SOP standard operating procedure TMDL total maximum daily load TSS total suspended solids WLA wasteload allocation WQBEL water q

21、uality based effluent limit WQS water quality standard 1 Evaluation of Water Quality Translators for Mercury Executive Summary This report was prepared for the purpose of presenting an overview and discussion of the use of the ambient water quality criterion for methylmercury developed by the U.S. E

22、nvironmental Protection Agency (EPA) in 2001 (66 FR 1344). Unlike all other previous water quality criteria for the protection of human health, the criterion for methylmercury was issued as a fish tissue concentration (i.e., 0.3 mg methylmercury/Kg wet weight fish tissue), because the fish consumpti

23、on pathway is the main route of exposure to mercury in the environment. As a result, enforcement of water quality standards based upon this criterion requires either fish tissue sampling or the conversion of the criterion to a concentration in water, especially for the development of National Pollut

24、ant Discharge Elimination System (NPDES) permits. Because many permitted dischargers will be faced with meeting the methylmercury criterion, this report focuses on (1) available analytical methods for evaluating mercury in fish and water; (2) proposed methods for translating a fish tissue concentrat

25、ion for mercury into a concentration in water; and (3) implementation of the mercury criterion in the development of Total Maximum Daily Loads (TMDLs) and water quality-based effluent limits (WQBELs). EPA (2002a, 2004a) draft Guidance for Implementing the January 2001 Methylmercury Water Quality Cri

26、terion addresses these issues and is discussed where applicable. Analytical Methods Analytical methods to quantify mercury in environmental samples continue to be refined. EPA Method 1631 is currently used to measure total mercury in water samples, while EPA Method 747B/SW-846 can be used to analyze

27、 total mercury in solid or semi-solid samples. The basic method for quantifying methylmercury is EPA Method 1630. Modifications to this method are described in the UW-Madison Standard Operating Procedure (SOP) and the United States Geological Service (USGS) Method. The selection of a method to test

28、for either total mercury or methylmercury should be based on several factors, including method detection limit, validity of the method, and quality control procedures. All of these factors are important because they determine the quality and usability of the generated data. Mercury Translation The t

29、ranslation from methylmercury in fish to total mercury in water can be viewed as a single step (i.e., the direct method) or a multi-step process. In the direct method, total mercury in fish is directly translated to total mercury in water. In contrast, in the multi-step approach methylmercury in fis

30、h is converted to dissolved methylmercury in water and then to total mercury in water. The major challenge in developing mercury translators is the enormous variability in the site-specific potential for mercury methylation and bioaccumulation. Among the important factors that contribute to the vari

31、ability in mercury bioavailability and bioaccumulation are pH, dissolved organic carbon (DOC), salinity, water flow (e.g., stream flow, lake flushing rate, etc.), temperature, reduction-oxidation potential, sulfide and sulfate, suspended solids, nutrient loading, fish age and size, prevalence of wet

32、lands and forested land cover in the watershed, and concentration-dependent demethylation (Brumbaugh et al. 2001; Qian et al. 2001; Kamman et al. 2004; Marvin-DiPasquale et al. 2000; Sonesten 2003a; Rose et al. 1999; Ullrich et al. 2001; Watras et al. 1998). EPA (2002a, 2004a) draft guidance recomme

33、nds the following approaches to criteria translation, in order of preference: (1) derive site-specific bioaccumulation factors (BAFs), (2) use a bioaccumulation model, or (3) use EPAs national default translators (EPA 2004a). EPA (2004a) does not recommend any specific bioaccumulation models for mer

34、cury, although a modeling approach is preferred over the use of default translators. Despite extensive study, there are no accurate, nationally applicable models for predicting mercury bioaccumulation. As an alternative to process-based or mechanistic mathematical models, the draft EPA guidance sugg

35、ests that regression models incorporating such variables as pH, DOC, and fish age may be acceptable for criteria translation purposes. In addition, the draft EPA guidance includes empirically-derived default values to simplify the translation from the fish tissue criterion to a water quality criteri

36、on for total mercury. Two sets of values are provided: (1) BAFs describing the relationship between dissolved methylmercury in water and methylmercury in fish, and (2) fraction dissolved (fd) values describing the relationship between total mercury and dissolved methylmercury in water. 2 EVALUATION

37、OF WATER QUALITY TRANSLATORS FOR MERCURY Moreover, the guidance describes partition coefficients (KDvalues) that account for the role of total suspended solids (TSS) in determining dissolved mercury concentrations, although default values are not identified. Despite the identification of default tra

38、nslators, the preferred approach to criteria translation, as recommended by EPA, is the use of site-specific translators. Provided that the study design accommodates an expectation of high variability, it should be possible to effectively develop site-specific translators for mercury. Mercury Transl

39、ations in TMDL/Permits For many facilities, the major issue in developing mercury translators is incorporating the fish tissue mercury criterion into a discharge permit limit. Permits may be reopened and reassessed as part of a TMDL implementation. At least 45 states have fish consumption advisories

40、 due to mercury and over 1,000 water bodies are listed as being impaired due to mercury, thus triggering TMDLs. Several methods are generally used in developing mercury TMDLs: (1) the concentration in fish tissue; (2) the concentration in the water column; or (3) the concentration in sediment. Of th

41、e three methods, using a fish tissue concentration as the TMDL target is the most direct measure of the desired endpoint, protection of human health. Determining how to allocate loadings among point and nonpoint sources is the next major part of a TMDL. EPA (2002a, 2004a) offers three approaches to

42、allocation of loadings, depending upon the relative contributions of point source and nonpoint source loadings. First, where point source loadings dominate, the TMDL should specify reductions in these loadings, alone or together with nonpoint source loadings, to attain water quality standards (WQS).

43、 Second, where point source loadings are small, reductions in nonpoint sources are expected to achieve the TMDL. The third scenario also involves relatively small contributions from point sources, but reductions in nonpoint sources are not expected to be sufficient to attain WQS. The approaches used

44、 to implement a mercury TMDL will vary depending upon the allocations, as described above. Where there are waste load allocations (WLAs) to point sources, the traditional method to implement a TMDL is likely to be used (i.e., through the NPDES permit). The implementation of effluent limitations for

45、mercury and other pollutant parameters can be approached through several mechanisms including reasonable potential and development of WQBELs. Where a permittee anticipates mercury limits being incorporated into their permit, they should be prepared to proactively take steps to understand the concent

46、rations of mercury that are discharged from their facility (and the nature of these sources), the receiving stream characteristics (including sediment and aquatic species), and the policies of permitting and regulating mercury by the permitting authority. Despite the cost, collection of site-specifi

47、c data affords dischargers opportunities to develop more realistic bioaccumulation scenarios, potentially allowing the calculation of a more appropriate water quality criterion. Summary and Recommendations This report summarizes the issues associated with translation of a mercury fish tissue concent

48、ration into a water quality criterion. Section 2 presents an overview of the analytical methods available for analyzing total mercury and methylmercury in fish tissue, water, and sediment. Section 3 provides an analysis of the methods and models available to translate the fish tissue concentration i

49、nto a water quality criterion. Section 4 describes the impacts of the mercury water quality criteria on TMDL development and NPDES permits. Finally, Section 5 discusses the applicability of these issues for other metals. Although the report addresses many topics, it also identified several areas where additional research or study is necessary to fully understand mercury translation in the environment. The following bullets summarize the key findings and recommendations for additional study: Validation of methylmercury analytical techniques is necessary to increase the certainty of resul