Document vbbJe6O98Ge3nmqRgaDw7zzq

I CMA i EXHIBIT A COMMENTS ON WA1JER QUALITY CRITERIA EPA APPENDIX ^ AND APPENDIX B-l j (Fed. Reg. March j.5, 1979, p 15926f) I I I 049420 EPA APPENDIX B - Guidelines for Deriving Water Quality Criteria the Protection of Aquatic Life and its Uses i 1.0 GENERAL COMMENTS for We previously submitted comments on the proposed guide lines as they appeared in the Federal Register (43 FR 21506 May 18,. 1978 and 43 FR 29028 July 5, 1978)j and we are pleased to hear that these comments will be addressed fcjy EPA at some later date. ' However, we are concerned that the proposed water Quality Criteria (WQC) numbers as they appear in Appendix A are based on the guidelines in their original form. Since we ha^je serious reservations about the, statistical manipulation as developed in the May 18 document, (expressed in our letter of August 2, 1978) we have equally serious reservations about the-WQC numbers as they appear in Appendix A. We reaffirm the CMA position stated in the August 2, 1978 submission. Also attached as an addendum are general comments prepared by our consultant. Further, we support the position expressed by the Ameri can Petroleum Institute in its August 2, 1978 submission to the Agency (Attachments I-III of CMA Exhibit A). We trust that the Agency firnj ly believes in their statement that the Water Quality Criteria ar e not "cast in concrete". As the guidelines change, so will the re ultant numbers. Adjusted toxicity numbers shp uld not be used if the appropriate experimental numbers are availabl In addition, acute toxicity should not be used to derive a chi onic number. This reaffirms the statement in the guidelines that f chronic values are not available the chronic numbers cannot be der ived using the methodology. 2.0 EXAMPLE We will illustrate our concef n with the data manipulation, with a detailed examination of Table 8 in the May 18th Federal Register. This is a new comment on the proposed guidelines and the reason we were not able to develop this comment before is that the data was not available until recently. Table 8 is an attempt to combare acute toxicities of various chemicals to a range of invertebrate organisms in both freshwater and saltwater. Referring to Table 8 which is located on page 21516 of 43 FR, May 18, 1978, we notice that the first part of the Tabic deals with identified chemicals aijd a list of numbers. The numbers presumably are the comparable acute toxicity value for various species of invertebrates. Since no mention is made of the medium it is as sumed that the results are a mixture of values obtained for fresh and marine water. The data for T^xaphene (fourth from the bottom) will be examined. The acute toxipities of 16, 1.4 and 4.4 were taken from Schimmel et al., Arch. Environ. Contam. Toxicol., 5, 353 (1974). 1CMA 049421 The value of 16 ug/L represents the 96 hour EC50 for reduction of shell growth for oysters while thei 1.4 and 4.4 represent the 96 hour LC50 for pink shrimp and gras s shrimp. The values of 10,24 and 180 were taken from an EPA rep ort (EPA 600/3-77-069, June 1977) and represent a 48 hour EC50 for D aphnids and a 96 hour LC50 for Scud and Midge, respectively. In this one set of data the Agency has mixed freshwater and saltwater species as well as a combination of EC50's ranging from reduction o f shell growth in oysters to le thality. There is no mention of w|hat the 48 hour EC50 on Daphnids represents. Without further clarification of this Table it is im possible to believe the results that are generated. Certainly, the data on Toxaphene does not represent comparable acute values as the Table is represented. Continuing the examination of Table 8, we come to the bottom of p 21516 to the section called "^11 values from Reference 42". This looks like a series of random numbers and presumably is a series of> chemicals acting on a rarnge of invertebrate species and if we believe the reference they are all saltwater species. How many chemicals are represented? S ince 82 sets of data were summarized and there are 26 and 25 in the other two sections of Table 8, we would conclude that there are 3 1 sets of saltwater species. IIowever, the summary on p 21517 indie ates 34 sets, so once again we are confused and skeptical about the c onclusions from the analysis. Additionally, on page 21517 (t he top left hand column) there are< twenty-five sets of paired values taken from reference 37. The summary of these results has now b een made available and the chcmicals associated with each pair of numbers were identified along with the species tested. The first nun^ber in each pair was the 48 hour LC50 for the water flea (freshwate r species) while the second number is the 96 hour LC50 for Mysid Shrimp (a saltwater species). Thus, the Table represents two species f rom two different media. In a later version of this same data se t we have discovered the following errors: 1. The pair 487, 8.59 as reported in the May 18 guidelines is the data for 1,3 dichloropropane and should be 282, 10.3. 2. The pair 32.5, 0.0 4 is for fluoranthene and should be 325, 0.04 3. There are several other minor errors which should be corrected from the more recent data as reported by Bionomics in EP A Contract 68-01-4646. Finally, the report by Bion omics should be made available and subjected to p eer review before the data is used for establishing r egulatory guidelines. 2 CMA. 049422 The summary portion of Table 8 is the attempt by the Agency to use the above data to derive a fac tor that will protect 95% of the invertebrate species. There is no way of knowing what 25 sets of data were used to arrive at the fa ctor of 21.3 for freshwater species The only group of 25 is the data f rom Bionomics and since this represents both fresh and saltwater s pecies it would be invalid. 3.0 SPECIFIC COMMENTS In selecting final criteria i n Section XIV p 15973, reference to flavor impaired should be delet ed as it is covered under health guidelines. Furthermore, we reco:mm end that the word "avoidance" should be changed to behavioral ef fects." The formulation of water qua lity criteria for beryllium, cadmium and lead which take into acco unt the influence of water hardness on toxicity recognizes scient ific reality. The mathematical slopes detived by EPA and obtained from the relationship between hardness and acute toxicity cannot be used to represent the relationship between hardness and measures of chronic toxicity. This comment also applies to EPA's practice of using slopes calculated on the basi of fish toxicity data to determine relationships between hardness and invertebrate toxicity. EPA acknow ledges that ... "Since data of the right kind may not be available to determine whether such a statistically significant relationship e xists for both acute and chronic- ,, toxicity to both fish and inverteb rates, extrapolations from one species to another, assumptions an d judgments may have to be used instead of statistical tests."(p 15971-15972) However, these assumptions mi}st be tested before establishing water quality criteria. 4.0 SUMMARY Until the Tables in the May 18 Federal Register are clarified, the safety factors so derived raus be suspect, and should not be used in deriving a Water Quality riteria. in conclusion, while the concept being developed by th^ Agency for deriving aquatic WQC is most important, the methodology itself needs a great deal of work before it can be accepted. Finally, we recognize that tli e data base associated with some chemicals is much more complete th an others. Accordingly, a criterion number derived from the better dat; a is much more reliable. Some indication of the reliability should be mentioned in each document, without the appropriate confidenc interval the criteria could be badly misused in the standard sett ing operation. 3 CMA 049423 EPA APPEN'J X B-l Those are alternate procedures foe generating the proper numbers that are required in the guidelines. We will comment on them as they are listed on p 15974. 1. We do not feel that unpublished, data should be used in devcloping such an important number as a WQC. Acceptance of this pro-' cedure is bad science and can c nly lead to unacceptable regulations 2. A. We do not accept the assumption that a 24 hour average concentration can be estimated by multiplying the Pinal Acute Value by 0.44. It is unreasonable that a value based on 96 hour LC50 will be smaller than an experimentally determined chronic number. The chronic value should be the basis for a 24 hour average. B. A`'great deal of Structure Activity Analysis would have to be done before a procedure based on extrapolation could be used to generate missing data. While structure activity relations arc useful there are many examples where only a minor change in structure can and does lead to major changes in toxicity. C. Since we disagree with assumptions 2a and 21J, it follows that we do not believe the guidelines can be used with this alternate procedure. 3. A. If the final Fish Chronic value for a chemical is determined experimentally then the Same experimental procedure can be used on other chemicals. We do not believe that extrapola tions of interpolations to structurally similar chemicals can be made at this time. B. Once again, we believe that more structure activity studies must be made before the credibility of this assumption is accepted. c. Similar to 2c above. 4. This appears to be reasonable. 4 CMA 049424