Document gE3en1j28X7GjvMeyYxna4Y7L

) COMMENTS IN RESPONSE TO TWO ADVANCE ) NOTICES OF PROPOSED RULEMAKING RELATING TO THE MANUFACTURE OF PCBs ) BELOW SO PARTS PER MILLION PUBLISHED ) MAY 20, 1981, BY THE ENVIRONMENTAL PROTECTION AGENCY [46 FED REG 27617, ) 27619) EPA DOCKET NOS. OPTS - 62013 AND OPTS - 62014 Submitted to th* Environmental Protaction Agency by the Chemical Manufacturers Association Special Programs Panel on PCBs November 25, 1981 HONS 013429 f TABLE OF CONTENTS EXECUTIVE SUMMARY I. INTRODUCTION 2 8 II. BACKGROUND 10 III. HEALTH AND TOXICITY EFFECTS OF PCBs 17 IV. NATURE AND FATE OF THE INCIDENTAL MANUFACTURE, PROCESSING, DISTRIBUTION AND USB OF PCBa AT CONCENTRATIONS BELOW 50 PPM V. WORKER PROTECTION AND IN-PLANT EXPOSURES TO PCBa VI. ANALYSIS OF INCIDENTALLY GENERATED PCBa A. Problems, costs, and uncertainties in analysing chemical industry processes for low concentrations of PCB B. (Insert discussion of CMA-EPA Round Bobbin Survey] C. Regulations Directed toward control of PCB in products can probably be supported analytically at the current state-of-the-art D. The state-of-the-art in PCB detection and quantification limits the ability to study methods of controlling incidental generation 29 34 39 40 43 43 41 MONS 013430 VII. COST FACTORS IN THE ANALYSIS OF PCBa 49 A. Chlorobiphenyl analysis in procsss streams requires advanced equipment and methods 49 B. Factors involved in routine monitoring of process 52 VIII. PROBLEMS, RESOURCES AND COSTS TO REDUCE OR REMOVE INCIDENTALLY GENERATED PCBs 56 A. Introduction 56 B. Considerations in reducing generation 57 C. Example of efforts to reduce the formation of PCB D. Reducing PCB concentration after generation 62 1. Standard Separation Processes 63 (a) Distillation (b) Extraction (c) Adsorption (d) Absorption 2. Special Processes 64 (a) Examples of company efforts to remove PCBs after incidental generation IX. POTENTIAL IMPACT OF EXEMPTION PETITIONS OR INDUSTRY AND EPA 72 -- X. AMBIENT PCBs ' ,. I 92 XI. EPA's PROPOSED EXCLUSION FOR CLOSED MANUFACTURING PROCESSES/COSTROLLED WASTE PROCESS. 98 HONS 013431 XII. AN ANALYSIS OF THE LEGAL, PROCEDURAL AND POLICY ISSUES IN EPA'a REVIEW OF THE RECORD IN THIS PROCEEDING A. Introduction and Summary B. Tha legislative and ragulatory history of PCBs undar tha Toxic Substancas Control Act. C. Environmental Defense Fund v. Environmental Protection Agency: The 50ppm regulatory cut-off. D. CHA's agreement to conduct an intensive, expedited information gathering effort. E. EPA'a development of a regulatory program for incidentally generated PCBs. 1. The general nature of EPA'a authority and discretion to fashion a regulatory program for incidentally generated PCBs. 2. Sole reliance on the exemption petitions process will impose unreasonable administrative burdens on EPA. 3. Regulation of PCBs below SOppm would yield trivial or deminimus benefits. F. EPA must interpret Section 6(e) together with other provisions of the Act. G. EPA'a assessment of the benefits of regulating incidentally generated PCBs: The Health Effects of PCBs> Risks to be Addressed in Proposal Regulations. H. EPA'a obligation to consider the effects of other regulatory controls of PCBs. 1. Other provisions of TSCA 2. The Occupational Safety and Health Administation Exposure limits for PCBs 104 HONS 013432 3. The Clean Hater Act. 4. The Resource Conservation and Recovery Act Of 1976 (RCRA) 5. The United States Department of Agriculture - rood Establishment Controls. 6. The Food and Drug Administration Food and Drug Controls. 7. Agency proceedings held in abeyance I. EFA should review any proposed regulation of incidentally generated PCBa in accordance with Executive Order 12291 XIII. CONCLUSIONS APPENDICES REFERENCES ATTACHMENTS (not included) Previously submitted documents: 143 Analytical Dissertation Report on Incidental Generation Survey Ecology and Environment, inc. Report, and literature references MQNS 013433 EXECUTIVE SUMMARY When Congress enacted Section 6(e) of TSCA In 1976 it was concerned with the need to baa or severely limit coauierclally produced PCBs. Persistent and known to accumulate in body tissue, these PCBs were viewed as a relatively narrow group of products used as dielectric, heat transfer, cooling and hydraulic fluids. But neither Congress, industry or others concerned about PCBs at that time knew that a wide range of industry processes incidentally and unintentionally generated extremely low concentrations of PCBS* Consequently, they also did not know that many of these PCBs were in significant raspects different in chemical composition and Physical characteristics from, commercially produced PCBs, and that many of them were significantly different in the nature and extent of the potential health risks they posed. Accordingly, Congress could not have considered whether Section 6(e) was an appropriate means of regulating such KBs. Congress could not have addressed the practicality or feasibility of the narrow and limited provisions by which certain PCBs could be exempted or excluded from regulation. MOMS 013434 Information submitted to EPA by CMA demonstrates thatt 1. PC Be are not as toxic to humans as they were believed to be when Congress enacted TSCA or when EPA promulgated its PCB Ban Regulations. They do not present the carcinogenic risks once believed. Epidemiologic studies have not demonstrated a causal relationship between PCB exposure and excess in total mortality, or in mortality, due to canoer, cardiovascular disease or neurologic disorders. Certainly not every human exposure to PCBs can be deemed significant. The large number (thousands)^of newly published studies on the health effects of PCBs justifies a reassesmxent by EPA of its position that any exposure to PCBs presents a slgnfleant risk. 2. While a large number of chemical Industry processes generate low concentrations of PCBs, the vast majority of those PCBs J (a) are safely controlled in wastes regulated by EPAi HONS 013435 (b) are in proceaaea from which diachargea other than waatea are aubject to other controla which eliminate or aignlficantly minimize human expoauree; (c) are deatroyed within the proceaaea which generate them; or (d) are otherwiae in a form or atatua which effectively preventa aignificant human expoaure. 3. Becauae of the minimal opportunitlea for in plant expoaure to workera and the worker protectlona and practlcea employed by the chemical lnduatry, workera do not experience aignificant expoaure to Incidentally generated PCBa. 4. Becauae of the complex nature of the PCB family of chemicala, the limitationa and coata of avail able analytical equipment, and the exotic and little underatood acience and technology Involved in conducting analyaea of complex chemical atreaau, the analyaia of auch atreama to determine whether and what concentrationa of PCBa they may contain ia enormoualy expenaive and uncertain. HONS 013436 5. Similarly, the alteration of chemical process** to reduce or remove concentrations of incidentally generated PCBe is expensive, difficult and uncertain. Extensive Information supplied by CMA provides substan tial evidence to support the following conclusions! 1. There is no basis for concluding that inciden tally generated PCBs below 50 ppm should be regulated. In fact, evidence may exist to demonstrate that materials containing higher concentrations of PCBs do not merit regulation. 2. Reliance upon the exemption petition procedure provided in Section 6(e)(3) of TSCA as the means by which industry can obtain relief from regula tion of Incidentally generated PCBs is patently inappropriatei (a) Congress did not know of the existence of widespreadlow concentrations of incidentally generated PCBs and could not have Intended to subject either BPA or industry to the uncer tainties, costs and burdens of a caae-by- case, rule-by-rule, year-by-year petitions procedures. HONS 013637 (b) Because of the pervasiveness of low concentration incidentally generated PCBs, the petitions process will subject small manufacturers or any sired manufacturers of low volume products to uncertainties, costs and burdens which will very likely shift or eliminate innovation and investment in prod uct development. Because such PCBs exist in a range of processes which are common in the chemical industry, the uncertainties of the petitions process can be expected to adverse ly Impact the business plans of suppliers and customers of the generators of such PCBs. 3. The Agency's 50 ppm regulatory cutoff has already disadvantaged U.S. firms conpeting with foreign coapanles whose processes are not similarly regulated. 4. SPA should exclude from regulation all processes producing any concentrations of PCBs wherei a. all wastes or discharges are handled or disposed of in accordance with all applicable regulations; . HOMS 013438 b. the PCB concentration in product! of such proceases do not exceed 50 ppm (or such higher level as EPA may establish by rule); and. 5. Because of the physical, chemical and toxicological properties of monochloroblphenyl and dichlorobiphenyl, these compounds should not be classified as PCBs for regulatory purposes. In any event, monochloroblphenyl should not even be classified as a PCB because it has only one chlorine atom. HONS 013*39 X INTRODUCTION These comments *nd data ara submitted to EPA by tha Special Program Panal on PCBs of tha Chemical Manufacturers Association (CNA) in response toi o commitments which the Association made to volun tarily provide information to assist the Agency in developing new *PCB* regulations under the Toxic Substance Control Act, a project prompted by the fact that the DC Circuit Court of Appeals set aside certain of the Agency's PCB regulations in Environ mental Defense Fund v Environmental Protection Agency; and o EPA's advance notices of proposed rulemaking, published May 20, 1981, 46 Fed. Reg 27617, 27619. CMA is s nonprofit trade association whose approximately 165 United States member cospanies account for more than 90 percent of the total domestic production capacity for basic industrial chemicals. Many of our membera will be directly affected by EPA's regulations and policies under the Toxic Substances Control Act generally, and Section 6(e) of that Act specifically. MOMS 013440 CMA's submission contains information of a type and quantity not previously available to Congress or BPA -- information which will provide the Agency with substantial evidence to develop effective and not unreasonably burdenei regulations of PCBs. HONS 013441 XI. BACKGROUND A detailed deecription of the history of this natter is contained in Section_____ , infra. What follows ie a general deecription of the events leading up to CMA's submission to EPA. In enacting the Toxic Substances Control Act (TSCA) in 1976, Congress separately Identified by name only one chemical for special regulatory control -- polychlorinated biphenyls. In Section 6(e) of that Act, Congress established a phased program for the disposal of PCBs, the phase out of its manufacture, processing and distribution, and stringent limitation of the use of PCBs. That program was to be inplemented by a series of EPA rul sinkings. Certain of the Agency's subsequent rulemaking efforts were challenged by the Environmental Defense Fund (EOF) and on October 30, I960, the DC Circuit Court of Appeals sustained EDF's challenge that EPA had failed to support two of those regulations with substantial evidence. It directed the Agency to proceed seditiously to develop new regulations. The pertles to that litigation realised, however, that Emulating PCBs on the basis of the sweeping statutory language lone, and absent regulations which more precisely defined the "ture end scope of the control of PCBs, might be to the disadvantage of the general public interest, the regulated industries and the EPA. Accordingly, the litigants negotiated an Sedlted data collection and rulemaking schedule and jointly filed motions with the DC Circuit Court of Appeals to stay the andate of its opinion. A `stay of mandate* would leave the nous 013*42 challenged regulations In effect during that cooperative, industry-government effort. The Court subsequently granted those joint notions, after which EPA published advance notices of proposed rulemaking (ANPR) in order to solicit data froai the public to suppleaent information being collected by the industry litigants. CMA was not a party to the litigation. However, It voluntarily stepped forward and participated in the negotiations among the litigants. It then agreed to assist the Agency by expending considerable resources in an intense effort to obtain data from its members and to provide narrative technical dissertations on various issues facing the Agency in its Court-directed effort to develop PCB regulations. The Information submitted herewith is the result of CHA's efforts in meeting those commitments to the Agency, and in addition, contains information beyond that which the Association agreed it would provide. . Information is also provided in response to the additional questions posed by the Agency in ANPR*s published in the May 20, 1981, Federal Register, which called for the submission of information by November 16, 1981._/ CMA's submission contains a great deal of information -- in both type and quantity -- which was not available to EPA when it promulgated its challenged regulations. Obviously, this _/ An extension of time until November 25, 1981 was granted to CMA to collate portions of this submission other than the report on the questionnaire survey and review of toxicity. Those two documents were submitted to the Agency in final draft form on November 16, 1981. HONS 013643 IX I , information was also not available to Congress when It enacted TSCA. In order to provide an overview of the importance of this information it is necessary to provide a brief summary of the language and legislative history of Section 6(e) of TSCA, the nature and basis of EPA's challenged regulations, and the holding of the DC Circuit Court of Appeals while this information is detailed later in these comments, this overview may provide a helpful context in which to review the CMA's submission. When Congress enacted the PCB regulatory scheme set forth in Section 6(e) of TSCA and when EPA promulgated its regulation limited or preliminary information was available concerning important respects about PCBs. The focus of Congress's concern in adding S6(e) to TSCA during its passage in 1976 was on those PCBs that had been manufactured intentionally. Historically, PCBs had been sold for ; variety of applications amoung which were, in descending order, I capacitor or dielectrics, plasticiser applications, transformer coolants, hydraulic and lubricant fluids, and heat transfer media. | Curing 1971' and 1972, all sales except fluids for capacitors and transformers had been ended by voluntary industry action and total > dcmestic sales were about half of the peak reached a year earlier; j Yt the feeling in Congress was that additional controls were | desirable and that they should be of a mandatory nature. The focus of EPA's concerns in promulgating its 1979 emulation was also primarily these-same PCBs, but'the regulations Iso covered PCBs incidentally and unintentionally generated within industrial processes. Specifically, the Agency.proposed an HONS 013444 exclusion from its regulations for such Materials containing PCfc in concentrations below 50 ppm. However, the only chemical . processes incidentally significant generating low concentrations of PCBa which received any significant attention in the coarse of the rulemaking related to dye plgsents. The DC Circuit Court of Appeals, in setting aside BPA's *50 ppn* regulation as unsupported by substantial evidence, adopted BPA's view that Section 6(e) of TSCA encompassed incidentally generated PCBa, and observed that not only had BPA failed to show that its regulatory cutoff did not pose unreasonable risks for health or the environment, but failed to identify the amount of PCBa left unregulated by the 50 ppm cutoff CHA'a submission fills these important gape of information. It provides significant new information on the nature, poundage, fate and exposure characteristics associated with PCBa incidentally generated in concentrations below 50 ppm in chemical processes. In general, CMA's submission supports the following! a. While the total quantity of low concentrations, incidentally generated PCBa is extraordinarily Mall, the PCBa occur in types of processes which CMA believes are common with 8IC 28 and within Industry at large. b. The vast majority of these PCBs are either destroyed or controlled to the extant that there is little if any human or environmental exposure, including in-plant worker exposure. c. Analysing and quantifying low concentrations of incidentally generated PCBs in the complex matrices of MOWS 013445 chemical processes involve a range of uncertainties hitherto unidentified and a aignifleant resource burden on large aa well as small corporations. d. Reducing concentrations of PCBs to levela below SO ppai is difficult, uncertain, and expensive. Perhaps the single most iaportant issue facing the Agency in this Matter is the need for it to obtain the best scientific information regarding huaan effects of exposure to PCBs. When Congress passed Section 6(e) of TSCA, PCBs were viewed s a *nad dog chemical", a life threatening chemical and an animal end hixean carcinogen. When BPA promulgated its PCB regulations it held to a poeltlon which became a pivotal finding for the DC Circuit Court of Appeals -- a conclusion that any exposure to PCBs was significant. - Mew scientific information Indicates that not only was BPA's conclusion in error, but that PCBs are not the threat to huaan health as was previously believed; Furthermore, recent information Indicates that certain chemicals within the Chlorobiphenyl family do not exhibit the same potentially adverse offsets as others. the fact of this new information is Itself iaportant. In the ton years prior to the 1976 passage of TSCA, some 2,000 scientific ottloles had been published on PCBs. Since 1976, 5,800 additional scientific and technical papers.have been published. In an effort to help BPA evaluate the significance of this new wealth of HONS 013466 * information, CMA is submitting a number of critical reviews of significant studies in that body of published literature. CHA urges EPA to direct that its own scientists carefully review any significant new information on the huswn health effee of exposure to PCBs and that, in addition, it utilise its currea authority to solicit the views of respected scientists from outside the Agency on an expedited basis, regarding their latest views on the human health effects of PCB exposure. HONS 013**7 III. HEALTH AND TOXICITY EFFECTS OF PCBs PCBs have been singled out in the legislature and news sedia of this country as posing a general threat to human health. Environmentalists, legislators, and the sedia dubbed PCBa as 'deadly toxins,* 'human and animal cancer causing agents* and 'life threatening chemicals resulting in a general belief by the US population that the slightest presence of PCBs represents a significant potential haxard.* PCBs are the only chemical substance which Congress specifically mandated in TSCA as requiring immediate regulatory action by the BPA Administrator. The PCB regulations written thus far appear to have been premised on EPA's belief that any closure of PCBs to hiaians is significant and may present an unreasonable risk. In proaulgatlng PCB regulations, EPA must utilise substantial evidence in determining whether any unreasonable risk to health or the environment are associated *lth PCBs at various levels of exposure. Substantial evidence MOMS 013448 should comprise good scientific evidence relating to the effects of PCBs on human health. CMA believes that substantial evidence exists to support the contrary conclusion that low level exposures to PCBs do not present a risk to human health or the envlronaent. In developing regulations to implement Congress's statutory mandate, the EPA Administrator relied upon adverse effects informstlon contained within 4 reports identified in EPA's Support Document. All of these were published In either 1976 or 1977. Much new Information has been published on the health effects of PCBs since that time. A recent CMA search of the literature (TOXLINE, MEDLINE, and Chemical Abstracts) shows that over 5800 scientific and technical papers on PCBs hsve been published from 1976 to date. At most. Congress and EPA had available less than 2800 scientific articles which were published during the 10 year period prior to 1976. On this basis alone, CMA requests that EPA evaluate this new Information and reassess Its position that any exposure to PCBs is significant. The probability that a chemical will result in sn adverse effect is a function of exposure to the coq>ound end its inherent toxicity. As noted elsewhere, low vapor pressure minimises the potential for exposure even at concentrations of several thousand ppm. Animal toxicity literature documents that PCBs are not a highly toxic compound. Utilisation of standardised criteria for acute toxicity classifies PCBs as only slightly toxic to non-toxic (Table 1). This ranking is based in part on the fact HONS 013449 ! * that doses required to produce short tarn toxic effects in sons tast spacias ara so high that tha dose necessary to causa similar affects in humans would exceed X quart. A comparison of tha acuta toxicity of PCBs with other commonly known chanicals is shown in Table 2. Chlorinated biphenyls have been evaluated for mutagenic potential in several systems, ranging from bacterial test systems to intact animal studies. They have been found to be negative in the Ames test with the exception of one positive mutagenic response for the 4-chlorobiphanyl isomer. Repeated attempts to duplicate the results of this early study have not been able to confirm a mutagenic potential for 4-chlorobiphenyl in the Ames teat. This lack of reproducibility in the early findings reported by Wyndham at. al. (1976) can only lead to the conclusion that thsir findings are unfounded. Chlorinated biphenyls also exhibit negative responses in the dominant lethal test (a test for heritable damage leading to embryotoxiclty) and re not clastogenlc (capable of causing gross or morphological 4*age to chromosomes of blood or sperm cells of rats or cultured bwan blood cells). Collectively, the results of these studies indicates that ambient levels of chlorinated biphenyls do not Present a mutagenic risk. PCBs have produced embryotoxiclty and apontaneous abortions in laboratory animals but only at doses which are high enough to Produce maternal toxicity. PCBs are no more toxic to the fetus than most chemicals when given at toxic doses to the maternal MQNS 013450 i'l 4 animals. PCBs (at high levala?) produce ainlaal change* in newborns and are much less toxic than some common vitamins and hormones. CMA contracted with the Independent consulting firm Ecology and Environment, Inc. (EtE) to prepare a review of the latest health effects information on PCBs. Several newly collated studies were given to EtE and these and other significant published or readily available studies were reviewed. It is EtB's "independent professional opinion that 'any exposure to PCB's' does not pose a significant health risk to humans." EtE concluded "that PCBa do not represent a carcinogenic (genotoxic/initiator) risk and that, even if they possess oncogenic (spigenetlc/promoting) activity, the risk at low exposures is insignificant.^' In other words, PCBs share the characteristic of being active only at high, sustained doses, and up to a certain point, the lesion may ba reversible. Therefore, safe levels of exposure may be established by carrying out proper dose-response studies." Numerous studies Involving both mice and rats have been reported, but only one study exists to date that suggests PCBs may cause an increase in hepatocellular carcinoma. Two other chronic tests in rats, as well as tests in the mouse, did not reveal a statistically significant increase in liver cancer or any other cancer with PCB treatment. A recent bioassay, performed by the National Cancer Institute, did not demonstrate that PCBs produced cancer in the rat. (ref.) A comparison of the one positive study with the large number of negative tests. HONS 013451 nd consideration of other important contributing factors, such as the spontaneous rate of liver injury and liver regeneration which can lead to aberrant hepatocellular growth, leads to the conclusion that PCBs do not represent a cancer rich to people. At worst, PCBs at high or elevated doses meet some of the criteria for a chemical promoter, l.e., they enhance tumor growth through epigenetic mechanisms. PCBs do not induce new tumor growths. At most, PCBs have only been shown to possess a week promoting activity. Therefore, the thresholds both proposed in theory and observed in animal studies suggest that there are safe espoeure levels for PCBS. The most meaningful information for assessing a chemical's chronic health effects can be derived using human data from Properly designed epidemiology studies. Epidemiological studies *re often possible in situations where people have been exposed for many years. A recant review of twenty-one epidemiology studies found ^xt during occupational exposure, PCBs can cause temporary skin Problems, but that there are no clinically observable long-term *"**n health effects, including the occurrence of cancer (ref.). Judies of non-occupationally exposed populations found neither fcwatitia nor other clinically evidence of exposure-related f*tcta. Many of the recent human epidemiology studies include data on **rge numbers of industrial workers who were occupationally foeed to PCBs over many years. These occupational exposures. HONS 013452 In moat cases, involved chronic PCB exposures which are orders of Magnitude greater than any which can reasonably be foreseen under existing regulations. A comparative review of these occupational-exposure studies reveals that, like other chemicals, PCBe can cause adverse health effects, but in Met respects these have been Minimal. While dermatitis and chloracne, which were reversible after discontinuing the exposure, have been noted in some cases, no other significant findings were routinely made. In spite of over 50 years of use, no causal relationship has been eatablished for any specific type of cancer, nor has it been proved that the incidence of cancer Mortality has increased. In a recent HIOSH study involving about 2500 electrical equipaent Manufacturing workers, over half of the cohort was occupationally exposed in excess of 20 years. Some of the workers were exposed to PCBs for as long as 40 years. The study found no excess in total mortality, nor in Mortality due to cancer, cardiovaaular disease or neurological disorders. Since this study failed to denonstrate an excess cancer rate in a high exposure population, it provides soae reassurance that it is unlikely that future studies will show any increased risk of cancer from PCBs. As additional data have been accmnilated on the Yusho incident, it appears that it was WOT the PCBs which were responsible for the adverse health effects observed but the extraordinarily high levels of inpurltles in the used coamerlcal Japanese manufactured PCB mixture. (1) The conclusions from these epidemiologic studies are in distinct contradiction to statements appearing in the present HONS 013453 ; a of Coort of appeals Decision that "PCBs are extremely toxic to Amr kaMna>* (re ice) ana 'epidemiology data and experiments on llMrat0Cy ( la indicated that PCBa poae carcinogenic and la. ether risks jiumans." ess Although most recent acientiflc literature reviews have focused on human health effects studies, recent publications on n the effect of environmental contamination by PCBs do not support the conclusions that PCBs have exhibited a significant an Irreversible detrimental effect on the ecosystem. Early reports decemented that PCBa are both persistent in the environment and that they bioaccumulate in some species. CMA agrees that the higher PCB homologs are persistent and bioaccumulate to some Ifre*. however, at present available scientific, evidence does *ot support the alleged health and environmental consequences of ay this behavior. fear of dire ecological consequencies caused the a *dsialstrator of EP to find "that any release of PCBs into the *h ''iroemenl will eventually result in widespread exposure of Udllf * Including some of man's food sources, and husans and bi ,och *Posure may have adverse effects.* In 1976, EPA d data on PCB contaminated levels in food for the years (Table 3). These data clearly indicate a significant be ^ td trend in PCB levels, including the highly publicised al ^ ^***la. These downward trends in the data are the results *"ntary industry actions in 1971 to reduce usage and "ibi, ' *Posure of PCBs by eliminating all 'open* uses. These HONS 013454 i* h responsible industry actions were taken prior to promulgation of TSCA in 1976 and certainly much earlier than the actual PCB regulations which became effective in aid 1979. Very recent data on environsmntal contamination of fish and natural bodies of water (4,5,6,7)* have shown that environmental levels are still decreasing at a significant rate. Levels in fish in three different areas of the O.S. (Great Lakes, Mew York State, and Alabaaa-Georgia) were in the range of 1 to 5 ppm in 1980-81, as contrasted with the pre-1976 data which ranges from 60 ppm in 1973 to 18 ppm in 1975. In addition, at least two recent publications (5,7) demonstrate that the earlier prediction of extreme persistence of the higher chlorinated homologs in the environment is not of the order of a hundred years but is measuresble and is on the average in the range of 40-50 months. These deta are based upon half-life values of PC8s in the tlssoes of fish environmentally exposed, in addition, recent data published by the National Research Council (1979) have ehown a significant reduction of PCB concentrations in air over the North Atlantic Ocean. These data have demontrated a decrease in PCB concentrations in air ranging from 0.76*0.55 ng/m3 in 1973 to 0.074*0.099 ng/m3 in 1977. [should this be left in??, need to check with NAS). ... * Motet Numerous other references to individual studies in the same time frame are available in the literature. These studies were selected since they present summary and overview of various geographical regions and Include a cosparison of analyses of PCBa over a period of years. HONS 013455 1 ? I * t -------- 1 Of 1 1 r* f1 ork n oo ^lon tbs f* 1 ?! rtt KB Oncontrolled waste disposal practices of commercial mixtures tt TCU have resulted In human exposures and environmental aostwinatlon. Since some FCBs were widely dispersed and are psrslstent (especially the higher homologs), they became #lguitous In our enviroraeent. Concern over the adverse health effects were largely due to the possibility of effects from isaftera exposure to PCBs and not based upon scientific stservatlon and fact. Drastic regulatory measures were smwldered necessary in 1976 due to the then existing great scientific uncertainty concerning health effects. Compounding ttts uncertainty is the fact that the effects of PCBs vary Ifestly with the degree of chlorination and that coanaercial ***** PCBs, upon which most health effects studies are ****** were contaminated with incidental contaminants such as Klrchlorlnated dlbensofurans. ^ believes that a thorough evaluation of the toxicological terature substantiates the conclusion that under most lions of exposure PCBs have not been desnnstrated to be in the public and that intermittent, low exposures will to conseouences. CHA reauests that the ^ *trstor review the aoit recent scientific data and evidence and the previous determination that any exposure to 1 ||y _ p ssent an unreasonable risk of injury to health or the _^`fonmant ^.1 r Ibat any exposure to PCBs is significant. Ing such e^_ n *n assessment, the Agency should review its i t,n` nd establish new regulations based on an . * ri,k which are more consistent with the true * PCBs. HONS 013456 ** 1 1 1 1 1 1 1 1 I 1 1 j 1 1 1 1 ] 1 1 Tabla TOXICITY RATING CHART Ratine Probabla Oral Lathal Data (or RMau* 'SO txpsetad lin Data 1. Meataxle 2. Vaakly Toxic 2. Madaxataly Toxic 4. Toxic 5. txtrsaaly Toxic 4. Sapartoxlc 15,000 ag/kg S,000*15,000 ag/kg 300*3,000 ag/kg 50-500 ag/kg 5*50 ag/kg 5 ag/kg * Avarago Adult of 70 kg Adaptad frsa Deall at al. 1580. k 1 Qaarc 1 Plat-1 Qaarc 1 Oaaea-1 Plat 1 Taaapooa-1 Oaaca 1 Prapa-1 Taaapoeo laaa Than 7 Drops HONS 013457 Table JL ACTUAL TOXICITY AND THE TOXICITY RATING CHART OF SOME SELECTED CHEMICALS Agents '30 Ida Alaabsl Table salt Xxea Wt Itrythslna Mcatlaa ran letallnea tesla 14.000 ng/kg 10.000 ng/kg 4,000 ng/kg 1,500 ng/kg 100 2 1 0.001 0^.00001 * Average Adult of 70 kg Adapted Iron 3oull at al. 1SSO. Expected Honan Dose* 1 Quart 1 Plat-1 Quart 1 Flat 1 Ounce*1 Plat 1 Teaspoon-1 Ounce 4 Drops 1 Drop Loss Than 1 Drop __ Lass Than 1 Drop MQNS 013458 Tabla PCBa IN FOOD Parent of aaaplaa with PCJs Prodocta 1973 1974 riah Kill Ut* Ckaaaa Faad Cowpocaati taiat1 Faad Proeaaaad Fruit laby Food . Haata, Poultry 60.4 3.2 1.1 0.9 12.7 7.2 4.5 1.1 1.9 44.0 2.6 4.2 2.6 0.0 0.0 0.0 0.0 1.2 idaptad frou Jalinak and Cornalioaaan 1976. Haxiaal 1975 Coocaatratiaa 17.6 0.7 0.0 0.0 0.3 0.0 0.0 0.0 0.3 123.0 2.3 11.0 2.0 9.0 199.5 19.2 traca > 3 ppa HONS 013459 31 *t 1TTO8 AMP FATE OP THE IHCIDENTAL MANUFACTURE. PROCESSIMG. DISTRIBUTION AMD USE OF PCBS AT CONCENTRATIONS BELOW 50 PP- CMA surveyed Its membership to determine the nature, fate, poundage and other characteristics of the incidental manufacture, processing, distribution and use of chlorinated biphenyls at concentrations below SO ppm. The complete report is appended. The poundage of chlorobiphenyls reported represents data Mfcaltted by 65 firms, 2t of whom found 13 5 processes to report. These 65 firms represent between 36 to 37.6 percent of all U.S. chanieal industry sales in Standard Industrial Classification Cbde 28, based on 1978 sales data. Moreover, they represent a large percent of basic industrial chemical sales, since firms in this group are members of the Chemical Manufacturers Association. It is this segment of the industry where incidental generation of *CB is most apt to occur. Accurate projections to the chemical Industry as a whole eaonot be made without a high level of uncertainty. Attesting to project to the total chemical Industry by multiplying the r*Ported poundage by a factor of three would not be reasonable *nd result in vastly overstating the total produced. The reported data show that only about 13,600 pounds of chlop foblphenyls are Incidentally generated per year of which 13,100 Pounds or 95 percent are disposed of as controlled waste. Of the 13,100 pound disposed of as controlled waste, 9100 pounds (66 HONS 013460 percent) la incincerated, 1800 ppcfids (13 percent^ is in a aalt watar wah dischxrgad to a aalt water reservoir surface land!ill, 1600 pounda (12 percent) goes Ce^landfill, and^6iM5 pounda (2 percent) goaa to ground injection. The aurvey alao found that on the order of 660 net pounda per year of chloroblphenyla are contained In producte produced by the ehaaiical induetry. Of thia amount, approximately 500 pounda or 75 percent are encloaed in aolid producte such as paint files and polymers, or destroyed, thus minimising human contact and exposure in the environment. Wo chloroblphenyla were reported as going into consumer products. About 490 of the 660 pounda were reported to be in products which were listed as captive intermediates, intermediates, or industrial products. Commercial products were reported as containing 68 pounds per year. It is believed that the reported data represented the bulk of the chloroblphenyla incidentally generated in the chemical Industry. This is based on the coverage of the basic Industrial chemical production represented by the firms responding to the survey. The most Important feature to consider in evaluating the results of the survey are the demonstrated patterns of production and fate of the chloroblphenyla incidentally generated. Since 95 percent of the Incidentally generated chloroblphenyla are disposed of as controlled waste, even large errors in the estimates of quantities produced would mean that there is a minimal potential for human or environmental exposure. MONS 013401 i rurtheraore, of tba reaaining 5 percent found in products, about 4 parcant Is ancapsulatad in soaa solid Matrix auch aa a paint fils or polyaar. Tharefora lncldantally generated PCBa doaa not credent a raal potential for axposura. Nhila It is believed tba poundage reported la representative of tba total, the nuaber of processes reported ia probably aaall in relation to the total_aao<afTt o^'lncidantally generated tl chlorobiphenyls. The typaa of procaaaaa reported are widely and extenalvely uaad, and aany aaall aanufacturera could eng>loy thaa to produce low-voluae producta. While this would not add significantly to the total poundage.-TherqgSfe, the nixnbar of Individual procaaaaa incidentally generating ehlorobiphenyla nay he understated by a factor of ten or aore. The data provide no i basis for estiaating the exact nunber. The opportunity for workplace occupational exposure aaterial containing ehlorobiphenyla ia also extreaely low. At least 99 Percent of the poundage reported ia generated and processed in enclosed process systeas or process systeas where any release or fnsoval occur under carefully controlled conditions. The survey data show that only 8 of the 135 processes **ported had been cheaically altered to operate below a 50 ppa concentration level. The survey did not request infonaatlon on Physical changes which were necessary in order to operate Processes below 50 ppa. - , The data reported by aurvey respondents were based on hnalytleal tests for 44 percent of the poundage, on inforaed HONS 013462 1 I opinion based on sons analytical evaluation for 13 percent, and on best-theoretical-eetimages for 43 percent of the poundage. Several of the respondents reported the "firms's known or estimated detectable limits of PCBs." These data are tabulated in the survey report. The responses do not appear to be in complete concordance with the CMA Analytical Report previously submitted to EPA. The survey requested information on the limits of detection (LOD) of chlorobiphenyl in process streams. The principal purpose for requesting those data was to provide additional information on the accuracy of the poundage reported. In other words, lgw values of LOD would indicate that firms were able to estimate the quantity of chlorobiphenyl generated with greater certainty. The instructions in the survey did not provide scientific criteria for defining LOD, such as that currently offered by the proposed standard of the American Chemical Society that the LOD be defined as at least three times quarter than the instrument signal-to-noise ratio. In the absence of such criteria, it would be expected that reported LOD's would be lower than that established by the standard since (1) the proposed standard is new and not yet widely adapted, and (2) the purpose of the standard is to correct a problem in overstatement of detectable limits (i.e.low values) anJt V provide comparability between laboratories. The survey instructions requested that LOD be reported for the specific process stream where chlorobiphenyl was being re ported. The implication was that firms would report on the basis of their own analytical data. It is not possible to be sure that all firms reported on this basis. It is possible that some firms reported literature values for LOD rather their own. Reporting of literature values for analysis of PCB in wate.r would explain the very low LOD values reported by some firms. Nhile the LOD provides some information on the level of uncertainty regarding the presence of unobserved chlorobiphenyl, it is nota usefull concept for evaluation of analytical capability to measure chlorobiphenyl. The LOQ is greater than the LOD. The Proposed ACS standard is that the LOQ should be ten times the instruaient signal-to-noiae ratio, 'compared to three times the HONS 013403 i_* ignal-to-noi.se ratio for the LOD. The LOQ level is higher because of the accuracy errors introduced into the analysis by use of lower signal values. As the analytical paper provided to BPA by CMA states, regulation should be based on the ability to ateasure chlorobiphenyl, not the ability to detect. Further information on LOO and LOQ will be provided in CMA's report on its round robin analytical study. TThe supposition here is that the round robin survey results will be relevant and persuasive on this point! HONS 013464 WORKER PROTECTION AMD IN-PLANT EXPOSURES TO PCBs In two of ita three May 20, 1981 Federal Register notices on PCBs, ERA posed the following questions soliciting informa tion relating to worker protection requirements: II. What worker protection requirements might be appropriate with respect to closed manufacturing process? (46 Fed. Reg. at 27618) 5...What precautions are being taken now, or are planned to protect workers who may be exposed to PCBs? (46 Fed. Reg. at 27621) CMA believes that imposing additional protection measures on top of those now employed by industry would duplicate existing efforts and/or contribute little if anything to the goal of worker protection. First of all, as discussed above, it is important that EPA bear in mind the latest scientific informa tion on the law toxicity of PCBs, disucssed supra at , Secondly, the low vapor pressure of PCBs makes it highly unlikely that workers will be exposed to significant amounts of PCBs. The accepted scientific physical laws of solutions, especially Raoult's law, dictate extremely small levels of PCB inhalation exposure from solutions containing low levels of PCB. /frie vapor pressure of PCBs over a water solutions MONS 013465 50 ppm of PCBN will be reduced by a factor of about 0.000003 as compared to the PCB itself. In every situation where an OSHA standard exists for the primary chemical involved in incidental manufacture of PCB's, control of exposure to that primary chemical gives orders of magnitude of safety relative to the OSHA PCB standard if the PCB concentrations is below a few thousand parts per million. For example, chlorobenzene exposure is regulated by OSHA. . If a worker is exposed to vapors from a chlorobenzene solution containing 1000 ppm PCB's, and the chlorobenzene exposure is at the regulatory limit, the PCB exposure will be below the OSHA PCB standard by a factor of 160. (For further discussion see App. A.) Thirdly, before PCBs were recognized as a potential risk and before they were discovered to be incidentally generated in chemical manufacturing processes, the chemical industry employed proper protective clothing -and good management practices to protect workers from exposure to the primary chemicals with which PCBs have now become associated as incidental contamina tion. SPA recognized this fact in the Preamble to the Nay 31, 1979 Final Ban regulations. 44 Fed. Reg. at 31527 (1979). However, in that preamble EPA erroneously asserted that the use of protective clothing and good management practices to control worker exposure to the primary chemical often left the exposure to PCBs uncontrolled. Id. As demonstrated in these ii HONS 013466 comment* (see Appendix A), it is CMA's belief that control of the primary chemical insures control of the PCB contaminants. In this connection, it is important to note that the EPA has already proposed, and the DC Circuit Court of Appeals upheld various authorized uses of PCBs based on the Administrator's findings that protective clothing and good management practices employed by industry reduced certain PCB exposures to low levels, thus presenting no unreasonable risks. Specifically, the Court upheld EPA's authorization of continued servicing of PCB transformers- those containing PCB concentrations greater than 500ppm, and the servicing and rebuilding of contaminated transformers-those containing PCB concentrations between 50 and 500ppm. The potential for worker exposure to PCBs incidently generated at some point within a chemical process facility is much less than the potential for workers servicing and rebuild ing electrical equipment. In any event, there are a wide range of protection measures currently utilized by industry either becuase of a need to protect against toxic or corrosive primary chemicals or as a result of specialized mandatory federal or state requirements relating to those primary chemicals. It is important to note that "proper protective equipawnt" and "good management practices" are difficult to define in terms which would be uniformly applicable or appropriate. Particular definition will frequently depend upon the nature of the MOMS 013467 % i manufacturing process involved. - For example, safety glasses, gloves, and long-sleeved shirts are generally used in the chemical industry when unintentional contact with a toxic or corrosive chemical may occur. If the risk of contact is higher, for example, during sampling of a pressurised stream or drumming-off of waste, then additional protection is appropriate. This protection would generally include splash goggles or shields, gloves, and pro tective outer clothing such as rainsuits or aprons. Food, drink and tobacco products are frequently prohibited from work areasr employers may also be instructed to wash their hands and faces prior to eating or smoking. If eye contact with a toxic or corrosive chemical may be involved, chamical showers and eyewash facilities are provided. Equipment or process engineering and design considerations ay also be important, for instance, in minimizing the potential for inadvertent releases and to control Intentional removal of a chemical from a process. For instance, saaiple points axe generally designed to be easily accessible to purge systems. Drum filling stations generally provide for proper ventilation. Pump seals and other likely leak points are generally shielded and drained to an appropriate point for clean-up and disposal. Further, the chemical industry today deals routinely with the proper design and operation of spill prevention control and countermeasure (SFCC) systems for storage and handling of toxic HONS 013468 I 1 chemicals. Under other regulations, best management practices (see 40 CFR 125) and SPCC plans (see CFR 112 and proposed 40 CFR 151 at 43 F. R., 39276 (September 1, 1978) either are or will be required. These regulations currently address many of the primary chemicals with which incidentally generated PCBs are associated. Therefore, additional spill control require ments addressing incidentaly generated PCBs are unnecessary. As discussed in our legal, p^gcedural and policy discussion infra at , Section 9 o/tSCH^ requires the Administration, before taking regulatory action, to coordinate the Agency's various authorities and to consult and coordinate with other Agencies in order to avoid duplicative or insufficient regulatory action. Congress also indicated that the results of such review and coordination were to be published with a proposed regula tion.' Accordingly, before EPA proposes any additional worker protection requirements on the chemical industry it should review existing proposals, consult with other Agencies including the Occupational safety and Health Administration and carefully consider all applicable existing requirements and practices^ . In conclusion, because of recent information on the/law/ toxicity, of PCB, the chemical's inherently low vapor pressure. and the extensive^of a wide range of worker prevention and good management measures already in place within industry fox a wide range of reasons, and because of EPA and the DC Circuit's pre vious conclusions that such measures appropriately reduce risks to PCBs, it is not clear what purposes EPA would serve by layering another regime of protection measures on top of those which exist. HONS 013469 35 VI. ANALYSIS OF INCIDENTALLY GENERATED PCBs A. Problems. Cot And Uncertainties In Analyzing Chemical Industry Processes for Low Concentration* Of PCBs. Aarcng the laeues which CHA voluntarily agreed to addreaa aa a part of ita information gathering efforts for EPA was: *[A]n explanation of the problems, coats and state of the art of sampling and analysis for the identification of the pressure and concentrations of PCBs below 50 ppm in complex chemical streams^/ Aa CHA further agreed, that information was submitted within four months of the Court's approval of the February 20, 1980, joint motion filed by the iltigants_.in EPF v. EPA.- On ______ , CHA sumitted a 38 page single spaced report, with attachments, entitled The Analysis of Chlorinated Biphenyls, (hereinafter: CHA's Analytical Report). This document represents the first publicly available effort to identify and describe analytical problems associated with the type of chlorinated biphenyls incidentally generated within chemical processes. It presents a nwber of important conclusions: *y Letter from David Forlsyth Zoll to Ruth Greenspan Bell, February 20, 1981, attached to Joint Notion for Further Stay of Assurance of the Mandate, EPF v. EPA, No. 79-1580, (D.C. Cir. filed February 20, 1981). HONS 013470 T I < ! I A. Chlorinated biphenyl* are in reality some 209 individual^ indentiftable "isomers*; . " B. Virtually everyone of theee leoaer* hat differing chemical and phyaical properties; C. Analyst* of process streams requires the individual indentification and measurement of each of thesei O. The various lnstrueents which can be used to measure chlorobiphenyls have different sensitivities to these lsomersj E. The variability in the sensitivity of the instruments requires that each instrument be calibrated against the specific chemical co^ound to be measured or that the variations from the standards used to calibrate the instrument ere knownj P. The absence of comaerclal^avaliable standards for all isomers prevents instrument callbrstion or determination of the difference in ..instrument sensibility between a number of these iscemrs; ' The chemical composition of process streams', or the matrix," has a large iapact on the ability of available instruments to detect and 'ateasure low PCB HONS 013471 -- " Limits of detectability and quantification in general are lower in product atreaae than in sore coaplex contaminated waate product streams. Limits of measuresMnt can range from less than one ppm per Isomer in some product streams to several hundred ppm per Isomer in some cosplex waste Streams; B. Development of an analytical protocol to analyse the set of output streams from a process have been reported to cost on the order of $120,000 / a protocol for analysing a comparatively clean product stream would be materially . less expensive. ' CMA' analytical report describes, how the absolute level of chlorobipheyl in^ process could be large yet, becauae of the number of other Isomers present as well as the limits of . detectability of available instruments^those PCBe could still be imdeteetable. These findings, of course, are of great Importance for EPA's consideration of its regulatory approach to PCBs. As currently implemented, the agency's 50 ppm regulatory cut-off applies to each pact of a chemical process, and its waste streams. At or below 50 ppm the currently available technology cannot, in many Instances, provide reliable assurances to either the manufseturer HONS 013422 11 *b- or EPX as to whether PCB concentration In a particular procaaa or waste atraaa ara within a particular concentration range. The analytical problems and uncertainties caused by the mis, - or matrix of various chemicals associated with a chemical process and the limits of detectability of instrument methods may vary widaly. For instance, product streams are comparatively `clean*, although they may still contain on the order of 50 separate chemicals. On the other hand, waste streams can be extra ordinarily cobles, containing many substances which have not been thoroughly characterised, or who by their nature may absorb or `lack* the PCBa in tarry mixtures or solids. These physical ores****- characteriatlca Increase the chance that ths-j.asssese of PCBa in a stream will be masked, and/or that the level of detectability and quantification will be raised. - The extent of this masking can be significant. Data in cm's analytical report that in some situations the level of detection of a single Isomer can be raised to 100 ppm in some matrices. HONS 013473 Unfortunately, there is no published data on the precision (meaning the repeatability of an analytical procedure) or amiracy (meaning the variation between a true value and what is seen) of chloroblphenyl analysia in process streems as opposed to analysis ' h.'t*''' for aroclors, or lntentially manufacuted KBs. The few literat ture references indicate a precision on the order of 420 percent for a given protocol using gas chromatography/mass spectromett'f Again, It is tha aatrlx effect which can hava tha moat Impact on praclaion. Significant decreases in tha relative raaponaa of tha masa apactroaatar occur aa a raault of aanpla mix (citation]. Data raportad in CMA'a analytical diaaartation indicated recovery of spiked aaaplaa aa low aa 65 percent. In aua, given tha facts that (1) multiple laomera are known to'occur in many proceaa streams, (2) tha uncertainty in tha limit of detection (LOD) of a given isomer under the best of circumstances ia leas than one to five ppm* (3) the limit of quantification (LOQ) per laomer is on the order of five to ten times the limit of detection value, and (4) matrix effects confound and Increase both LOD and LOQ values, then the ability to observe. Identify, and measure accurately the concentration of chloroblphenyl present in a process or waste stream at the current state-of-the-art will often be far above a 50 ppm concentration level. --- A (Insert dicusalon of CMA-EFA Round Robbin project] f Regulation Directed Toward Control Ot PCB Tn Products dan Probably Bb Supported Analytically at the Current State-^C-`theArt. On a theoretical basis, if chemical processes had the potential for generation of all 209 chloroblphenyl lsoawra simultaneously, .then the analytical- methods used to determine PCM would have to be capable of measuring the quantity-present HONS 013474 to level of 50/209 or 0.239 ppm/isomer. Such a limit la wall balow the currant atata of tha art in analytical capability for detection alone in tha type of procaaa atreama generally analysed. Available information from the CMA incidental generation survey, and information reported on analytical methoda development and shown in Figure X of the CMA analytical report, both ahow that the niatoer of iaonera in a given procaaa and product are apt to be few in nwber. Three baaic types of reactlona can yield chlorinated biphenyli (1) coupling reaction starting with chlorinated aromatics, (2) chlorination reactions, and (3) degradation reactions where hydrocarbon radicals are genr*ted in the presence of chlorine radicals. In tha first type of reaction, the point of coupling is well defined and the number of chlorinated biphenyls determined by the structure of tha starting Mterlals(a). For example, 3,4-dichlorobenzoyl peroxide is deco^osed in bensene under certain conditions to 3,4,-dichlorobipbenyl. Inpurities in the starting material could lead to other chlorinated biphenyls and some selfcoupling of the peroxide and isomerisation could occur under the conditions of the reaction, but a single product would dominate all others. _- HONS 013475 VI Wien chlorinating aronatlca, the second basic reaction type, it is well known that a phenyl or a chlorine substitute will direct subsequent substituent to the ortho or para position, for example, the normal iron chloride catalysed chlorination of monochlorobenxene will yield about 5St paradichlorobenzene, 40% orthodlchlorobensene, and only 5% setadlcblorobenzene. Conditions can sodify the ratios slightly, but the basic selection will continue to exist. It is. also will known that chlorine will deactivate a benzene ring. That is, it is more difficult to get the second chlorine on tha ring than it was to to get the first chlorine. This means that the second chlorine will usually go into the second ring in a biphenyl molecule, father enhancing the selection process. Tire review by Hutsingerd) amply deawnstrates the degree of selectivity in the chlorination of biphenyl. One example is the major coaponents of Aroclor 1221. Major Coaponents of Aroclor 1221 Compound biphenyl 2-chloroblphenyl 4-chlorobipbenyl 2,2'-dichloroblphanyl (continued) t in Aroclor 1221 12.7 28.4 18.7 9.2 (1) Hutxlnger, O., Safe S., and Xifcfco V FC'a (1974). CRC Press, Inc, -- The Chemistry of HONS 0134?6 2,41-dichlorobipbenyl 2,4'-dichloroblphenyl 4,4'-dicbloroblphenyl 3.5 13.6 6.2 All of the chlorine are in tha o,p poaition (to tha phanyl-phanyl bond). Ninety percent of the dich'lorobiphenyla have one chlorine on each ring. Hutlinger alao notea that there are S major coaponenta of Kanechlor-400 (a tetra-penta chloroblphenyl mixture) and 7 major coaponenta of Fhenoclor 0PC (a hexa-hepta chloroblphenyl mixture). Theae data clearly Indicate that chemical reaction selectively will limit the total number of components and will severly restrict the nimber of major constituents. the third basic type of reaction leading to chlorinated byphenyl ia degradation. Such high teaperature processes will yield coaponenta baaed on thermodynamic stability and the relative concentration of reactants. Steric and electronic V effects dictate that different structure will have different thermodynamic stability and tha relative concentration of reactant will dictate the approximate degree of chlorination. The two factors will combine to determine that a few coaponenta will predominate. Therefore, the level of uncertainty in the estimation of total KB in a product is leas than it might otherwise have been if such processes produced large numbers of homologs and isomers. HONS 013477 Matrix affects are greatly reduced when analyzing clean product streaais. While such individual product streams may contain a variety of chemical substances defined as 'product,* they tend to be better characterised than waste streams and therefore are easier to analyse. This is not to say that they are free of interferences* they are not. In addition, soma products are solids, such as polyaers and pigments, and thesa aatrices can be quite difficult to analyze for PCBa. However limits of detection of 1 ppa or less per isoasr are achievable in clean streams. On aany instances, therefore, the available analytical technology is capable of providing greater assurance ' that a product did not contain in exceas of 50 ppm PCS. A regulatory cutoff at leas than 50 ppm for products, depending on the level, say or may not be supportable at current state-of-the- art of analytical technology, depending on the nature of the Interferences caused by product in the analysis as veil as the number and type of isomers present. ~~ ~ V* D. The 6tate-of-the-Art in PCB Detection Ahd Quantification Llmlts*the Ability Xo Study Methods &l Controlling Incidental Generation The thickness of a piece of paper cannot be measured with a yardstick. Similarly, if the operating conditions of a process ara changed, today's state-of-the-art in analytical methods cannot measure how this changes the minute concentrations of chloroblphenyls present. As noted in preceding parts of this HONS 013478 4 discussion, many chloroblphenyls nsy bs prsssnt which art not detected at all, and a process change could merely push or puli isomers above or below the limit of detection. Moreover, the available levels of accuracy and precision are not capable of tbs' accuracy required to correlate the changes which may be occurria) with changes in process conditions. ' As has been noted above, matrix effects create great uncertainty in the analytical determination of PCM in the complex chemical systems found in waste streams and reactors. Bowever, there is an even more important analytical problem which prevents study of factors for controlling Incidental generation in a number of Instances. For example, thenaare many mechanise for the formation of bensene which create PCBs in relatively high concentrations in the process reactor, and at measurable levels. On the other hand, some Incidental generation processes produce concentration in the process reactor which are well below Instrument detection limits, and are estimated to occur in the l. parts per billion range. In such processes, PCM are observed only at a point in the process where they are concentrated in waste stream, such as still bottoms, where matrix effects are often most severe. Thus, the problem of monitoring a process analytically is doubly compounded because the analytical methods are not capable of observing the changes in concentration at the point of PCB generation, only at a point later in the process where they are concentrated and more heavily, masked by matri* effects. HONS 013A79 A further problem that PCBa may net be generated by the primary production procaaa at all, and may occur through aacondarv affacta in tha procaaa. Exaaplaa include refluxing a chlorinated aromatic solvent, a'aide reaction with a solvent or with a catalyat. In auch inatancea the limits of the capability of analytical methods may prevent even the ideittlfleation of that portion of the process responsible for PCB generation. Ohtll sclenca produces a better `yardstick*, one cannot seriously consider attempts to modify the chemistry of a process to control chloroblphenyl Incidental generation, you cannot change or control that which you cannot measure. ^4 <COST PACtOM III THE AMALTSIS OP PCs. The discussion of the econoaUc `impact of exemption petition procedures* later in this document, presents rough, order of magnitude estimates of the costa associated with that approach to control incidental generation of FCSs. What follows is a discussion of the cost factors Involved in the analysis of PCSs. /k Chloroblphenyl Analysis In Process Streams Acquires Advanced Equipment and Methods. As has been shown in the CHA analytical report the only procedure for the effective quantitative analysis of HONS 013*80 * t chlorobipbenyls In process streams is tha combination of gas chromatography with mass spsctromstry. This atatasMnt is baaed on ths following facts. First, ths 1avals of concantration of intarast require tha aioat responsive datactor systeiss, and detector ayatsaw which raact spaclfieally to tha lsoawrs of chlorobiphanyl likely to ba prasant. This limits tha choicas to gas chroawtography with alactron captura or electrocondoctivity datactora, and to Bias spactromatry. , I Sacond, matrices in process or waste stream saaples are likely to contain substances which intssfere with the use of electron capture detectors. This detector will respond to any haloganated substance, as well as to many oxygenated and sulfur containing compounds. Many of tha process or waste streams where chlorobiphanyl may ba generated result from proeassas which contain these interfering substance, particularly organochlorine compounds. Such compounds tend to illuta with the same retention times as tha chlorobibpenyls, and therefore mask tha signal. Therefore, in order to ba certain that the material being awasured is a chlorobiphayl, a detector which responds specifically to the chlorobiphanyl in a masked signal must ba usedi in this ease a mass spectrometer. Further, the response of the electron capture detector to halogen coopounda is so intense that the ' HONS 013**1 detector la aaaily "saturated", and loaaa its ability to raapond or discriminate between halogen cospounds wban cancantratlon of ona compound la large with raspect to tha . othar. Third, there la now ample sclentific'evidenee that tha signal raaponaa of tha alactron captura datactor varlaa slgnlfleantly from ona chlorobiphanl laoaar to anothar, even within tha aaaa homolog. Tha aciantifie litaratura now atatas unequivocally that tha alactron captura datactor cannot ba uaad in quantatlve analysis uniats or until tha datactor can ba calibratad against the speciflc isomer being quantitated (citation). Thara are at `present only SO standards for 209 lsosMr, thus affactively eliminating this detector for serious use unless the isomer of interest is among tha 60 standards or until store standards are . available. This absence of standards is also a most serious problem when using mass spectroscopy, but the relative error level seams to ba lass (on the order of a factor or two, rather than a factor or several thousand with tha electron capture datactor.) fourth, the pattern recognition analytical methods developed for analysis of Aroclor fluids using the electron capture detector, are not applicalba to process stream HONS 013462 61- analyaia.* Chloroblphenyla In proceaa atraaaa do not exhibit tha familiar hroclor fluid pattarn of iaomera. Iaomer dlatzibutiona In procaaa atraaaa tend to ba unique for each procaaa in which they are obaerved. In conduaion, it ahould ba noted that the availability of aora atandarda could Make the electron capture detector ore applicable in thia area of analyaia i the Matrix la non-interfaring or the chlorobiphenyl can be removed fro* the matrix. Much additional original reaearch on Methoda of a atandard preparation euat be conducted before a full aet of atandarda can be produced. It la believed that auch an effort will require aeveral calendar-,yeara to execute. . unleaa a apecial effort la Initiated to accelerate the proceaa. - 0 Factora Involved In Routine Monitoring of Proceaaea there are no rulea-of-thumb which apply generally acroaa all chemical proceaaea. which tell one how often a proceaa auat be Monitored to determine if it la in `control*. Quality aaaurance Monitoring can occur hourly, per ahlft, or per day, or once per week, depending on the * It ahould be noted that pattern recognition Methoda are not appropriate for analyaia of environmental aaa^lea either, but they aey have eoae limitad application to analyaia of traneforaer olla. - MOMS 013483 S3 stability of tho process, its degree of sutoiMtion, and other factors. ' Hhen ln-procsss concentration is used as the control, echanlsa for Incidental generation, then the closer the control level is to the catlasted level of observed process streaa concentration, the greater the effort required to aonltor the process for chlorobiphenyl. Bow often one would have to aonltor to deteralne if chlorobiphenyl levels were in 'control,* given a specified concentration level, is unknown since the aeasureaent aethods available are not capable of reliably detecting variation as a function of process change. Moreover, since chlorobiphenyl concentrations have has been observed to vary in different parts of an overall production process, the average nuaber of staples which would toe required is also unknown. It is therefore difficult to predict the level of effort and cost of aonltoring with confidence for any change in the concentration level for a regulatory cutoff, riras who have had reaaonabla confidence that the current level of chlorobiphenyl generation is well below 50 ppa, probably have not (1) expended the tine and effort to develop analytical protocols or (2) purchased the additional lnatruaentation necessary to support the added laboratory workload that routine Monitoring caused by a lower regulating cutoff would generate. The CMA incidental HONS 013484 i . ..c generation aurvey haa shown that raports on S3 of tho 137 reported procasaas wara basad on othar than analytical ' tasting. This is supporting tha hypothasis that not all tins hava found it nacassary to aonitor procasaas routlnaly at tha currant 50 ppa cutoff level. Tha CNA analytical report haa shown that tha development of an analytical protocol for an entire proceas (i.a. all process streams) costs on tho order of *120,000. If tha currant regulatory cutoff ware to be reduced below $0 ppa, than aora flras would be required to develop such protocols for their processes, the number depending on the relationship between the new cutoff liait and the expected concentration levels of PCI in these processes. Alternatively, if only products containing PCS were to be regulated, the coat of analytical protocol development would be expected to be less. If only products containing PCBa were regulated at a 50 ppa concentration cutoff, then there aey be no additional significant industry analytical protocol cost impact. It should be noted, that irrespective of the regulatory approach selected by EPA, a reduction in the 50 ppa PCS concentration cutoff will create cost lapacts which will fell aost heavily on the saw 11 Manufacturer or the low voluae product lines. The coet of the gas chroastography/aass speetroaeter instrumentation is approximately *250,000, and requires *12,000 HONS 013485 Sb per year In maintenance support. Many small manufacturers do not havs such equipment, or ths trained staff to operate it. The" protocol development Is an additional significant cost burden. Soutine Monitoring expenses using the equipment and protocol is a further cost burden. The result would be that few snail firms or firms supporting low voluaie product lines would find it possible or profitable to operate those chemical processes subject to regulation. Given the fact that the CMA incidental generation survey has disclosed the wide breadth of the processing situtatlon whore PCB is incidentally generated, including processes widely used in the industry, small firms with small production operations could find themselves force out of entire lines of chemical processing, this could have a severe ispact on new chemical substance innovations by small firms. This loss in innovation would be felt in the chemical Industry, the related manufacturing industry served by these fins, and the general public. . HONS 013486 PROBLEMS, RESOURCES AND COSTS TO REDUCE OR REMOVE INCIDENTALLY GENERATED PCBS A. INTRODUCTION The purpose of this section is to present the technicsl end economic Aspects of developing and operating chemical processes in which the concentration of PCBs must be controlled to less thsn SO ppm. The discussion focuses on two specific examples; one, s process adjusted to reduce generation of PCBs to less thsn SO ppm; the other s process in which removal after generation is th* only known way to comply with existing regulstions. Further concentration reduction in either of these cases would require considerable additional research with little confidence of success. The limitations of current analytical technology poses a not Inconsiderable barrier to this research. In this section are included theoretical considerations.in the reduction or suppression of PCB formation, processes for removal of incidental PCBs and the two case histories. It should also be noted that limited, valuable manpower and financial resources must be diverted into extensive research and development programs, which, even if successful, do not benefit health or the environment. HONS 0134! la For purpose* of this section, it is presumed that the maximum PCS content permitted in any commercial product will be <$0 ppm. Any additional incidental PCBs produced will be disposed of in accordance with existing EPA regulations. B. COWS!DERATIONS IH REDUCING GENERATION Little Information is available about manufacturing processes or chemistries leading to the formation of low concentatlons of PCBs. It isjui area of trace chemistry which has not been eifestantlvely addresses Ui published literature--and for good reason. As thermally and^hemlcally stable coapounds, PCBs, evan at significant impurity concentrations in other chemicals, are not more than diluents exerting little if any influence on uses of the contaminated chemicals. The existence of a regulatory cutoff level at 50 ppm My be the only reason for concern about PCBs as impurities in chemical Mterlals. . Hutslnger, at. al., in The Chemistry of PCBs (CBC Press, 1974) illustrates 41 different reactions that can form PCBs. An engineer in EPA's OPTS reportedly said (Chemical Week. January 14, 1981) that PCBs can be forMd anytime a material containing hydrocarbons is chlorinated in the presence of heat. By EPA definition, PCBs are any level of chlorination to the biphenyl molecule (from one chlorine atom to ten). The incidental generation and contamination of chemical processes by HONS 013408 5* KBs csn coin from Mny diverse sources. It is necessary to examine the physical properties of KBs to understand their ppsaranee in chemical processes. The molecular weight of s KB isomer can vary from 188.5 to 499. Bolling points range from 273*C for the mono homolog to soluble In water, but are soluble In oils and other organic solvents. KBs, with the exclusion of the lower homologs, are not readily blodegradeable cot^ounds and, once formed, are extreawly persistent. These characteristics give an indication found in processes, i.e., frequently in high boiling waste streams. KBs can find their way into chemical processes through two main routes; chemical reaction and contamination. Since chemical reaction types that form KBs are already documented In the exemption petition files and the CHA survey, they will not be discussed hero. Suffice it to say, when chlorine exists in a reactive fora (CI2, BC1, Chloride salts, etc.) and carbon or hydrocarbon are present with sufficient temperature and/or catalyst (such as iron, copper, or sine), it is possible that KBs can form at very low concentrations. This commonly takes place in s reactor, furnace, distillation column, tarpot, phase separator etc. KBs can even be formed at extremely low concentrations by breaking down a chlorinated solvent in a high tm^erature heat exchanger. KBs can also find their way into processes by incidental contamination, feedstocks may be contaminated at relatively low concentrations. A manufacturing HONS 013489 process aay destroy the PCBe or they may paaa on through unchanged. Hater contaainated with PCBa can eaally Introduce & into a proceaa aince PCBa are aore soluble In organic ecapounds. Hater wash atepa In processes readily transfer trace aaounts of PCBa to the organic aaterial. PCB oontaainatlon of oheaieal procesaes due to incidental reactions or incidental contaainatlon usually occurs at wary low concentration. These concentrations aay be below the detectable Halts for clean cheaical streaas* however, the point of oontaainatlon will rarely be a clean cheaical streaa. The analytical probleas associated with detecting PCBa at low concentrations in interfering aatrices is docuaented in CMA's analytical report. Since PCBa boil at high temperatures. Moat processes concentrate PCBa into the waste streaaa. For energy and enwlronaental reasons, nost procesaes concentrate wastes to as great a degree as possible and have one or aore recycle streaas. This has the tendency of contributing to PCB buildup. Ibr example, if PCBa are generated at a level of 100 ppb in a chemical reaction process and are concentrated in the wastestreaa by a factor of 100 (It yield loss to high boiling waste), the concentration of PCBa in the waste streaa will be 10 ppm. Correspondingly, if a PCB free aaterial is washed with water containing 10 ppb PCB at a 10 to 1 water/organic ratio, then the organic product will contain 100 ppb PCB. Again,' using a concentration factor of 100 (It yield loss) the high boiling waste streaa will be contaminated at a level of 10 ppm. Both examples illustrate that by incidental generation or HONS 013490 LO 9 contamination, a waste atraaa (tha hardast place to analyse) can be concentrated to a level of 10 ppm PCB with an entry concentration of only 100 ppb. Once PCBs are Identified in a process, it ie a significant research undertaking to try to eliminate thee. Besides the analytical problems of finding PCBa at extreamly low levels, the reactions which fore thee are unknown. The reaction kinetics of trace contaminants is a relatively unexplored field of sclenc Interferences, coapeting reactions, etc., make it extremely difficult to determine the precise mechanism of PCB production. There are many interacting variables to consider such as reactant concentrations, order of reactant addition, process teaperature, pressure, hold-up time, catalytic effects, mixing, etc. for example, lowering the process texperature 10*C should theoretically halve the reaction rate. Thus, PCB production may be halved. However, the primary reaction rate may correspondingly be halved. To get the same amount of production, the throughput must be doubled. The ratio of PCBa per pound of primary product may remain the same or may Increase due to a larger recycle stream associated with more throughput. To date, there has been little success in reducing the formation of PCBa in extremely low concentrations. Bach process represents an entirely different research project with at time a negligible effect on the amount of PCB generated PCB generation. The lack of success in reducing PCB generation can be supported by examining the exemption petitions filed with EPh. Significant reduction in the formation of PCBa at concentrations greater than HONS 013491 f SO ppa have baen reported in very few instances. Most of the success in reducing PCB concentrations has been nde by asking (engineering) process Modifications. These do not reduce the total poundage of PCBs foraed but rather just the concentrations at various stages in the process. This has been accomplished st substantial capital lnvestaent and energy costs. In other words, coapanies often have the capability to aake Modifications to ccnply with the regulation but not necessarily to reduce the total aaount of PCBs foraed. I. Ba--Tie of efforts to reduce the formation of PCB During the development of a process for the Manufacture of a key specialty cheaical interaedlate, it was recognised that the potential existed for the foraatlon of by-product PCBs. This was confiraed by analysis, with up to 2000 ppa detected in a process streaa. Bellowing is a suauury of the technical efforts eaployed to define and ainiaise the guantity of incidental PCBs foraed, and the consequent adverse iapact on the coaaercial Manufacture of the required interaedlate. (1) The analytical research prograa needed to support this project was estlaated to have cost $200,000. The development of practical analytical procedures was facilitated by the availability of advanced analytical equlpennt at a central research facility. This equipment is not coaaonly found in most company analytical control laboratories. Before the coaaerlcal production of the required cheaical interaedlate could be established at a Major cheaical Manufacturing site, established, it was necessary to purchase and Install additional analytical equipment at the site for a cost of $50,000. . (2) Research programs were initiated to study ways to reduce incidental PCB formation. The coupany lnveatigatedi (a) the influence of reaction variables on the rate of incidental PCBs produced in the process, and, (b) alternative cheaical routes to the desired interaedlate. HONS 013492 0 After an effort which took 2 1/2 parson years of technical time and an expenditure of $500,000, the company was able to reduce the concentration of FCBa in the process stream by a change in reaction conditions fro* 2000 ppm to between 30 and 50 ppa. A pilot plant was constructed at an additional cost of $50,000 demonstrate the revised process. As noted below In Section (3) this reduction in incidental PCB formation was achieved only with a significant economic penalty. It is also emphasised that in this case, the degree of response to change in reaction conditions is extraordinary and one would not expect this solution to be broadly applicable to other processes. The specialty Intermediate described above described above is being produced commercially by the <50 ppm PCBe process at an lncramental cost penalty of 10-15% and at a reduced capacity penalty of 50% as compared with the preferred process. Moreover, although the less than 50 ppm process produces fewer pounds of PCBe, it does produce a larger quantity of total waste material for disposal which results in an additional penalty of 20% in waste disposal costs. The less than 50 ppm process offers no benefit to health or the environment since the PCB by-product from either process are separated for disposal in accordance with BPA regulations. It is believed that a responsible foreign coapetitor could operate the 2000 ppm process, without threat to health or the environment, and in compliance with local regulations, at an economic edvantage. At capacity, the 0. S. process would operate at an estimated penalty in excess of two million dollars per year (1981 dollars). In summery, this project has resulted in an expenditure of $750,000 (should this be $800,000) with extensive associated manpower resources expended to develop a process which complies with current regulations but which affords no additional benefit to health or the environment. Further, additional ongoing multiaillion dollar par year operating penalties will be Incurred, again without environmental benefit. D. Reducing PCB Concentration After Generation . Methods for removing PCBs or reducing their concentration after they have been Incidentally generated are known. These are more likely to succeed at minimum loss of capacity than efforts HONS 013493 7 a CBa In lone plant n tlon alty. M Of M raduca pcb generation. The following brief dlacuaaiona la leant to be only illuetratlve of approachea to renove PCBa and lot to be a coaprehenalve treatlae on chealcal unit operatlona apable of auch reaoval. ct-.^dard Separation Proceaaea ion .to (a) Dlatlllatlon ad xa ISt rad The aeparatlon of trace contaalnanta by dlatlllatlon enerally require* one or more of the following! (a) Widely differing boiling polnta between dealred KM, output and trace contaminants; Mil ban (b) Very high reflux ratio (recylce with attendant high energy coat penalty; lth (c) Reduction in yield of dealred output with a or to local r# aalty correeponding increaae in waate for dlapoaal. In addition to the operating coat penaltiee inherent in (b) nd (c), capacity reduction cauaed by high recycle and/or reduced laid require aubatantlally greater lnveataent to achieve the itim i * M lion * aan output reeulta. (b) Extraction btraction proceaaea depend on partition coefflcienta or eaaurea of lade of affinity of aolvent for product but with lgnlfleant affinity of aolvent for aolute. In the caae of PCBa, he ooat aucceaafull extractiona proceed in the wrong direction, e., fro* contaminated water or raw Material inputa to procaaa .ration treaaa. Thua, procaaa watera contaainated at the part per Theae arc mien level nay end up In the organic cheaical'proceaaea effort* oncentrated to part par el11Ion lavela. OKlih HONS 013494 v+ ft Thar* ar* no known successful extraction proeaaaaa for raoving trao* 1 avals of PCBa fro* n chemical proeaaa streams. (e) Adsorption . Adsorption proeaaaaa rely on tha surface charactarlatiea of solida to prafarantially rataln th* eontaainant while allowing product to pass unaffacted. Coapllcatad procedures are usually needed to liait yield loss. Adsorbents are usually regenerated in place by reaoval of concentrated contaainants for subsequent disposal. Substantial capital investment is usually required, and such processes ar* soastiaes successful in reducing trace contaainants to low levels. (d) Absorption Absorption processes ar* aost often applied to th* preferential removal of certain gases or vapors. Absorption has no known applications to the reduction of trace PCB contaalnftlon in cheaical aateriala. Absorption of PCBs from trace contaalnated alneral oil transformer fluid oil has been reported using a polyaerised absorbent._/(ref. needed) This co^letely non-polar Mineral oil transformer fluid haa a significantly different partition coefficient relative to KBs than do process streams and therefore the efficiency of separation ia lost. 2. Bpeclal Processes Cheaical processes for the reduction of KB concentration in transformer oils have recently be announced. These porcesaea appear to be very effective in reducing PDB concentration* in transformer oiler however, they will have few applications in HONS 013*95 nr:  operations. Because transformer coolant oils are essentially Inert saturated hydrocarbon oligomers, very powerfull reactants such as metallic sod1us or organo sodium compounds can be used to atrip the chlorine fros PCBs while forming essentially' harmless alts and phenylene polymers both of which can be filtered from the oil. Only s limited number of conMrclal chemical processes would find use for such treatments to decontaminate their products. In most cases, the use of powerful reducing agents such as sodlia would result in reactions destroying or significantly altering desired producte. (a) EXAMPLES or COMPACT EFFORTS TO REMOVE PCBs AFTER IHCIPBWTAL CBBBBATIOW CASK QBE Identifying the Problem of Incidental Chloroblphenvl Generation. 1. DISCOVERY In the ease described below, the existence of mono and dichlorinated biphenyls as impurities resulting from th reaction process was known from theoretical considerations, by physical separation, and identification of PCBs during the earliest days of process development., One of the key material Inputs is a chlorinated aromatic. The process described was developed during the late 1930's snd early 1940's. The fact that MOMS 013496 Bono and dichlorinated biphenyl* are unavoidably produced aa incidental by products in this process was published in the literature in the 1940's. 2. THE NEED FOR ACTION Action to reduce the concentration of these chlorobiphenyls in the intermediate manufactured in thi process was initiated because of early state legislation requiring that any product sold in that those state containing 100 or more ppm chlorobiphenyls be labeled. Analytical Development 1. EXISTING CAPABILITY When the time came for action to reduce the chlorobiphenyl concentration in the intermediate manufactured in this process, analytical capability was limited to identification of three of the fifteen possible mono and dichlorinatad biphenyl isomers. Ouantification was not sufficiently developed to meaaur less than 0.14 PCB by weight (1000 ppm). 2. DEVELOPMENT OF NEW ANALYTICAL TECHNIQUES In order to reduce the known PCB concentration, several sophisticated analytical procedures were developed to analyse intermediates, products and waste streams containing mono and dichlorinatad biphenyl lsomsrs formed in this process. Procedures were developed HONS 013497 to identify PCB Isomers at trace levels in river and process waters. The total effort to develop and implement these procedures as protocols required 2.4 person-years of professional (PhD level) tiae and 2.C person-years of technician tine. Development of Removal Process 1. REDUCTION VS REMOVAL The scientific assessment of this situation led to the conclusion that removal of the PCBs was the only feasible way to reduce the PCB concentration level to a low concentration. A simple experiment demonstrated that at essentially sero yield of useful output, the process still generated mono and dichlorinated biphenyl at concentrations in the 1000 ppm range. This result was also theoretically consistent. 2. RESOURCES REQUIRED a. PHASE Ii State Regulation - lOOppm cutoff In order to coeply with the early state law requirement, the coapany's initial goal was to reduce the chlorobiphenyl concentration in the intermediate to below 100 ppm. This was necessary to assure that PCBs in - concentrations greater than lOOppm would no be achieved in any product manufactured from this HONS 013498 Ut 12 intermediate. This Initial phase required the following colletions and resourcesi a) Analytical Developnent 1.) 2.4 professional (PhD) person-years 2.) 2.6 Technical person-years b) Development of a removal Process (Design/installation/atartup of pilot system) 1.) 5 professional (Bnglneer/Scientist) person-year . 2.) 6 technician person-years b. PHASE lit TSCA Regulatlon-SOppm Based on federal (BPA) proposed regulations lowerln the cutoff to 50 ppm, the full scale process was designed to routinely meet an analytically measured concentration of leas than 41 ppm. The "less than 41 ppm* goal was necessary to insure coapllance due to analytical uncertainty. Tile additional resources needed were as follcwst a) process development 1 professional person-years 1 technician person-years b) design/installation/ startup of system 2 professional person-years 3 technician person-years HONS 013499 1 w D. Coats Xf 1. Personnel Costs-Developeent. Professional Analytical 1.32* X 2.4 years X $55,000 Phase I 1.32 X 5.0 years X 35,000 Phase II 1.32 X 3.0 years X 35,000 Tetal Technician Analytical 1.32 X 2.6 years X $22,000 Phase I 1.32 X 6.0 years X 20,000 Phase II 1.32 X 4.0 years X 20,000 $174,000 231.000 139.000 $S44,000 $ 76,000 158.000 106.000 Tbtal Deealopaent Expense * This represents direct and overhead expenses 2. INVESTMENT (a) Pilot systea ' (b) Pull Scale Systea (c) Analytical Equipaient (d) Systsa Modifications Vent changes Additional storage Tank dikes Sever segregation (e) Incinerator Based on 10* use for chlorobiphenyls Haste disposal (Total cost $10,000) Total Investment $340,000 $884,000 $1,000,000 2,500,000 350.000 800.000 1,000,000 $5,650,000 MOMS 013500 Xf 3 OPERATING COSTS INCREASE Basisi Total cost at capacity In 1981 dollars (a) tabor, aatsrial, IME 360,000 IMB"Indirect Mfg. Expense (b) QC 120,000 (c) Incinerator fuel 100,000 (d) all other 100,000 totel 680,000 Nhen the total lnvestaent cost of $5,650,000 is aaortised on a ten year straight line basis and allocated per pound of useful output, the total Increase In cost per pound is 60*. CASE TWO PCB levels of 20ppa> were neaaured in the reactor afference of a chlorinated organic process. Through several distillations, the PCBs remained in the bottoe streaas and w eventually contained in the plant tars for latter lncinerati However, several atreaea contained PCBs at concentrations gr than 50ppn. In order to coaply with PCB regulations on in-process streaas, the manufacturer Invested in 8200,000 of equipment to concentrate the.PCBs in one step following the reactor. In addition, an operating expense of 850,000 per year is necessary to keep all non-waste process bottoe streams below the 50ppa regulatory cut off. HONS 013501 VI. WASTE DISPOSAL COST One important consideration of lowering the regulatory cut-off of PCBs would be the increased disposal cost. Current disposal costs are significantly greater than the $.08/pound for incineration that BPA's consultant Versar predicted. The current incineration price is $.$0/pound for PCB contaminated oils. (What about (Freightcost) as eaapsred to t.014-0.18 per pound for non-PCB containing chlorinated chemical waste. At a concentration of 50 ppm, the incineration cost per pound of PCB is $12,000. when the level of PCB contamination is reduced, the cost per poun of PCB increases as shown in the following tablet Level of Contamination 50 ppm PCB 25 ppm PCB 10 ppm PCB Cost/lb. of PCB Incinerated $12,000 $24,000 $60,000 By lowering the regulatory cutoff from 50, the required ssnunt of contaminated material to be disposed of would be multiplied by a large factor. This would overwhelm the waste disposal capacity for PCBs. HONS 013502  ^ lapact of the Exception procedure on Industry and EPA -> [THE ASSUMPTIONS AND CALCULATIONS IN THIS SECTION NEED TO BE CIARIPIED] A. Overview Congress authorised EPA to grant, on a case by case, rule by rule, year by year basis, exemptions from the Act's limitations on the manufacture, process or distribution of PCBa ( ). An exemption may be granted if the* Administrator of EPA finds thati (i) an unreasonable risk of Injury to health or the environsMnt would not result, and (ii)good faith efforts have been made to develop a chemical substance which does not present an unreasonable risk of Injury to health or the envlrorasent and which may be substituted for such . polychlorinated biphenyl. ( ) As is explained more fully in our legal and legislative analysis (infra at ) it is clear that Congress focused its attention on the deliberate use, manufacture and distribution of PCBa. It was not aware that some manufacturers produce PCBe as incidental by-products of their manufacturing processes. There is simply no indication whatsoever that Congress intended to subject Industry or EPA to a filings and rulemaking procedure on the great number of industrial processes which CMA's data indicate generate PCBa and are, therefore, candidates for exaaption petitions.- HONS 013503 -7^ 2 eyond the fact that ralianca on tha exenptlon petition procedure aa tha only aaans of relief froa the overbroad language of the statute would iapoae significant adsinlstrative burdens on OK, the economic consequences for an Industry of such reliance could wall be severe. In general, the petition procedures 1. would inject considerable uncertainty into the business planning of Manufacturers, as well as their custoaers and suppliers, thus adversely iapacting production, Innovation and growth, with iapacts in both doaeatlc and international Markets. 2. would very likely force the abandonaent of products, and new investaent and expansion plans slaply because the profit Margins for the output of small flras or for low voluae Products could not support the scientific, technological, resource and dollar coaaltaents necessary to successfully clear the exemption petition process. B. General Scope of the Potential Economic Iapact on Industry The economic iapact of EPA's reliance on the exemption procedure could be substantial. First of all, processes in which PCBa are'incidentally generated are basic and widely used in the chMical industry (see) as well as other segments of the economy. Many of the adverse economic iapacts would occur principally because of the way in which firas aake Investaent decisions. Industrial chaaical flras, like their counterparts in other aajor HONS 013504 capital intensive sectors, sake long-range planning forecasts of projected returns on investment (ROI), and then coapare those forecasts with their target investments. Those projects expected to meet or exceed target levels of profitability are comeneed. Those below profitability targets are rejected. It is important for EPA to recognize that these profitability projections are normally sade on an individual product line basis. Ssall firaa and the low volume product lines of large fins are both in the category of commercial activities whose profit marglnes say be significantly lapacted by even aaall cost increases [Consider citation to CHA's submission on the economic impacts of EPA'a PHH rules on innovation.] Reliance on the exemption procedure would almost certainly impose cost burdens, uncertainties and delays on firms along the economic chain of suppliers through manufacturers to customsrs. There could be a significant reduction in replacement or . expansion capital for existing processes, as well as in innovative new processes designed to offer cost-reducing process alternatives for existing products or serve new product manufacturing needs. The impacts could also translate into losses for the economy as a whole in the form of employment and foreign trade losses, as well as into inflation-feeding losses in productivity and efficiency. Of particular relevance is the fact that many chemicals that come out of the complex reactions where PCBS ~can occur are highly engineered for unique properties or for highly critical specialty HONS 013505 applications. Substitution or replacement of these aaterials la often impossible, or at best requires long-tera,~~expenslve development programs which could be expected to result In significant Inflationary coat increases for entire categories of affected products. Ibr example, PCBs are generated in the production of phenyl-containing silicone products, which are typically used in highly critical applications such as military and commercial aircraft, military equipment, space vehicles, and nuclear reactors. In many applications, specifications require a phenyl-containing silicone product. In those cases, even if lower performance aaterials could be substituted, such aitoatltutlon would require development programs estimated to range between one to ten years in duration._/ More critical applications would require longer development and evaluation times.* As another example, PCBs are generated in the production of bensene phosphorus dicblorlde, an intermediate in the production of a catalyst for nylon in carpets. No substitute has been found for this intermediate material and without it, carpet quality nylon cannot b manufactured.** Withdrawal of that intermediate from the marketplace would require the producer of nylon for carpets to shut down a large process line, in order to 17--------------------------------- * Cite to Rearing Transcript, July 10, 1979, pp. 83, 98, etc. Testimony of NASA, pp. 65-93* and Naval Air Command, pp. 93-107. ** Cite HONS 013506 n# 5 reestablish production, embark on a substantial expensive research program for a substitute catalyst or a substitute for carpet nylon. The carpet manufacturer would have to find a substitute fiber or go out of business. The constsaer in turn would be faced with reduced choice in carpet selection. Costs of the Exemption Procedure The direct financial cost of complying with an annual exemption requirement is an Important factor exerting substantial negative economic Impact on firm's investment and production plans. Initially, this would consist of the cost of preparing, asseaUftling, and coordinating the firm's presentation of its rationale for an exemption. This would include the cost of analytical testing, developing of other supporting materials, (use of experts and consultants,) as well as the costs involved in interacting with EPA and other officials. Included within these costs are incremental analytical teating coats the firm might feel necessary to gain exemption petition approval, over and above those which would normally be incurred. After initial approval, there would be additional annual costs of assembling and filing any supplemental materials needed by the agency, as well as additional annual costs of any necessary communication with the agency. Again, the significance of these costs should be evaluated in terms of their impact upon particular product lines. In addition, they may require, especially for small firms, a HONS 01350* -.1 * significant coamltaent of tha firas adainistrative, tachnleal and ragulatory resources. Tha CNA incidantal ganaration survay raport Indicates thara nay ba a larga nuaber of aaall firaa that ganarata soaa flnlta aaount of PCBs at lass than 50 ppa. With sola callanea upon tha exeaption procedure, tha aggragata burdan of tha reporting raqulraaant on thasa saall firas could ba of a larga aagnltude. Purtheraora, for those processes which aay have bean operating naturally at soaa concentration below 50 ppa, ellalnatlon of a regulatory cutoff aay aaan, as to thea, a first tlaa inquiry into cosplicated and costly analytical issues. The analysis of incidentally generated PCS is difficult, both because of tha low concentrations Involved, and tha technical liaitatlons of the available analytical tools. The published technical scientific literature does not contain one paper which deals with tha issue of analysis of process straaas for PCBs. Bence, firas aust independently develop their own analytical. procedures at great expense. It is laportant to note that detailed procedures for such analysis aany never Be published in the scientific literature becauset o The process-specific aspects of the probleas Involved severely llait the general interest in such papers and thus inhibits their publication; o The process aspects inherent in a useful discussion of such analytical probleas increases probleas of the potential disclosure or industrial secrets. Based on lnforaation provided in the CNA's analytical report and the survey of incidental aanufacture, it is possible to sake HONS 013508 km ordar of magnitude projections of the analytical coats involved If EPA choose to rely solely upon the exception procedure rather than a regulatory cut off. The following assuaptions are utilised as a basis for projecting anticipated analytical costst o The number of processes Involved would be substantially larger than the nueber of processes reported in the CMA survey as generating KBs as suggested by the [cite to Pittaway Report]. We will usiee that 700 processes would be the subject of an exception petition, o The nature of the exception petition process Is such that monitoring or sampling of the process will be asployed as a conservative and prudent approach to obtaining relevant process information, if a process is stable, less saepling aay be warranted. Furthermore, while the CMA survey indicated that almost all of the reporting . processes operated between eleven and twelve months per year, many small manufacturers aay have processes which operate for much shorter periods, such as a few weeks a year. Estimates of the total number of samples to the analysed must take these factors into account, o CHA's estimates of the costs associated with needed equipment, procedure development and monitoring costa are real world approximations. Given these assumptions, the cost for analytical work in support of exemption petition could be roughly estimated as described below> HONS 013509 The analysis of PCBa in process streams esquires the development of an analytical protocol. The development of a protocol for analysis of clean product streams is much less expensive than the development of a protocol for analysis of all process output streams (e.g. product and waste streams), because of, among other things, complex matrix problems associated with waste streams. The CMA analytical report has estimated the cost of full protocol development at $120,000. If one assosea that only the products of a process stream would be analysed then a protocol night be developed for one-tenth the cost, or $12,000. The number of such protocols to be developed is dependent on the number of different firm Involved with the estimated 700 different processes, and the number of different chemical aubatances produced by those processes. Since it is aasumd that most of these processes are operated by small manufacturers. It is arbitrarily assumed that the niaaber would be on the order of 3S0. On thia basis, the protocol development cost could bet 350 x 12,000 a $4,200,000. [Can we provide additional clarification of the basis of our assumptions] Once a protocol has been developed, the cost of running an analysis is a function of the number of processes, the length of time a process is in operation, the rate of sampling required to assure process stability, and the cost of single sa^le analysis, tor purposes of developing order of magnitude cost estimates, the following assumptions are made, since no objective data are available to provide better estimates. It is assiaedi HONS 013510 w I! 1 P o XO percent (70) of the processes operate 12 months a year and require weekly saaples, 1. e. 1 * 52 i 70 3840 saaples. o 40 percent (280) of the procesaes operate 3 months a year and require three aaaplea per week, 1. e. 3 x 12 a 280 10,080 samples. o 50 percent (350) of the processes operate 1 month a year and require daily aaaplea, 1. e. 1 x 30 x 350 10,500 aaaplea Based on data supplied in the analytical paper on the coat of a PC* analysis, it is estimated that the occaalonal weekly aaaple would coat 8200 each, that the sore frequent analyais of 3 per week would cost 8125 each, and that a daily analyais would ooet about 875 each. Therefore, the cost of analytical work could be on the order oft (3840 X 8200) (10,080 x 8125) + (10,500 x 875) - 82,775,500 Equipments cost are separate froa and in addition to the above analysis costs, and can be estimated froa the number of samples to be analysed, and the aaortlsed equipment and maintenance costs. Proa CMA's analytical report it is known that a PCB analysis requires two to four hours of instrosent tlae. Including the running of standard saaples and other instrument calibration procedures. At an equipment cost of 8250,000 aaortlsed over five years (1. e. 850,000 per year) plus an equipment maintenance cost of 812,000 per year, the Investment cost is approxisately 862,000 per year. Assuming that the equipment can be used eight hours per day five days per week, 52 HONS 013511 /a '7 ^**'*W`> /- V - /^f(*4^^ weeks par year, tha equipment and maintenance cost ia approximately $30 par hour. Assuming that a PCB analysis requires two hours of Instrument time, the cost per sa^le Is therefore about $60. The estimated total number of sables is 3640 * 10,080 + 10,500 or 24,220. The $24,220 X $60 $1,453,200. In summary, the rough estimate of the analytical costs Involved would be on the order of $8,430,000 for the first year, and $4,229,000 in subsequent year annual operating costs. If no processes were to be excluded from regulation, then the estimated number of procesees affected could more than double; these estimated costs would also at least double, if for no other reaaon than the Increased cost of analytical protocol development for waste stream analysis. It will be noted from CMA's analytical report that firms have had little experience with trace analysis for PCB, and that the analytical requirements press the capacity of the available instruswnt methods. It can be ejected that this situtation would improve over time under the impetus of exaaption petitions. In the short run (i.e., 2 years) this would increase the analytical expense considerably over the figures given above, as more money was spent in protocol development and improved instrument techniques. In the long run (i.e., 5 years) costs could fall as methods became more routine and automated, but better methods would probably lead to the discovery of SK>re instances of Insignificantly low levels of KBs. This iwould ' ' " ......... - HONS 013513 ...... _ - n. tfi result in additional processes requiring montorlng, thus maintaining the overall level of coat burdan to industry, in ordar to control aaaantially tha same aaount of PCS. Tha principal isaua to keep in aind in evaluating thasa costs is that thair total magnitude for the Industry is not nearly as iaportant as tha intact on a single product line. For the manufacturer of a low volume product to be faced with spending $12,000 to develop an analytical protocol, and $4,000 more for analysis of samples on a process operated 30 days per year, plus the cost and personnel time required to fils the assertion petition itself, means that production may well be abandoned. The profit margin would not justify such costs. Significant dislocations on the availability and costa of many tail volume chemicals would result. It is worth noting that while many processes have the potential to produce incidentally generated FOB, only analytical analysis can determine if in fact a process does produce FCB. The scale of manufacture and the equipment materials of construction are both determining factors in lncidantal generation, (need clarification] Consequently, a small manufacturer may not be able to determine in advance if an exemption is required until the analytical work is coapleted. Manufacturers will tend to avoid processes with the potential for KB generation in the face of an exaaption requirement, thus stifling innovation in the development of new and improved leal products. HONS 013514 (2) Effect of Uncertainty Whan a Ira files an exemption petition, it faces two major kinds of uncertainty, both of which would have aavere adverse impact on their economic decision. The first type is Inherent in the regulatory exemptions process itself. Becaue exceptions are issued for only one year at a time, uncertainty would existi (a) for eanufacturers who from year-to-year would not know whether the process would be allowed to continue; (b) for customers, who would not know whether the product would continue to be available; and (c) for suppliers, who would not know whether their markets would continue to exist. The year-to-year petition process would in effect become a sword of Damocles over the head of manufacturers and their suppliers and customers, creating and re-creating annually a new element of risk to be factored into each year's invsstment projection, without any.corresponding .benefit to the public. Firms would be put at the mercy of year-to-year contingencies, potential changes in enforcement priorities, state-of-the-art changes in m testing requiresMnts--factors which simply cannot be taken account of in the long investment planning cycle which characterises this capital-intensive segment of the economy. Such levels of risk will inevitably choke off investment. The second type of uncertainty arises from the requirement of Section 6(e)(3)(B) that firms make a 'good faith* effort to develop a substitute process under a general exemptions HONS 013515 requirement that; (a) doaa not praaant an unreasonable risk of Injury to health or the environaenti and (b) nay be substituted for the process which incidentally generates PCBs. It is not ' certain, of course, how the interpretation of *good faith* sight change over time. Second, the requireaent creates a particularly acute investaent dlleaaa for aany firas. Incidentally generated PCBs are frequently not easily controlled. The natural laws of eheaistry are such that they do not allow an absolute 100 percent reaction to fora a given product. Thus eoae processes will always generate KBs as an incidental by-product, tor firas with such processes, a 'good faith* effort to find a replaeeaent for the processes would ultiaately have to be directed at finding substitutes for products which theaselves are not KB-contalnlnq materials. Firas in such a situation could be faced with a de facto ban on a whole class of aaterlala generated froa aajor classes of process reactions which are basic to the chealeal industry. This was surely not intended by Congress. Firas faced with one or both of these types of uncertainty are likely to do one or aore of several things depending on the nature and specialisation of their processes, the level of economic risk they face in making investaent decisions, and the specific nature of their own particular incidental generation process sltutatlon. tor existing processes, firas are likely to take one of three stepei tare these the only options?] (a) liait investaent to ainiaua aaintenance of replaeeaent levels, with no U^roveaent or expansion; (b) gradually phase-out their relevant processes by not replacing existing* equipment as it wears outs MOHS 013516 and (c) look for ways to develop modifications to disperse KBs. K coemltaent of scarce investment resources to this last course is as unproductive as the deliberate capital diminution program associated with the first two courses of action. Several Industrial chemical manufacturers indicated that they have alreedy begun an investment retrenchment program for processes affected by a potential expansion of the exemption requirement. [True? or was this because of the current exemption requirement?] The impact is also likely to be significant for Investment in new processes. New process innovation Investments ere of two typesi (a) Investment in more efficient, cost-reducing alternative to existing processes; and (b) investment in completely new processes to serve new product needs. Both of these types of innovation investment are likely to be severely discouraged. Firms faced with the uncertainty of the exemption process would be unlikely to be interested in making the large capital Investment necessary for innovation of a process characterised by a coapletely certain annual regulatory requirement. In most cases, the risk would almost certainly outweigh the project return. 3. Effects of Delay The process of obtaining initial approval under an exemption petition would also impose cost burdens by delaying a firm's ability to realise its Investment objectives. Delay costs Involved in obtaining an approved exemption petition can add untold millions to plant costs, currently at the rate of 15-20* (?) per year. Thus,- even a few months delay can HONS 01351?  * add anoraous iiai to tha coat of tbaae projects. The lapact of this delay on innovation will coae in two fonw. In aoM cases it is not bean possible to determine whether a process generates incidental PCBa until a full scale plant is placed in operation. This ia true for several reasons. First of all. tha level of operation at bench ecale and pilot plant My operate at auch low levels that the PCBs are below the liait of detection of available analytical tools. Second, in the changes between pilot plant scale and full scale, auch iteaa as Mterlals of construction, airing velocity inthe reactor, etc., can lead to PCB generation in full acale not present in pilot plant scale. Third, there are inatancea where two full scale production processes which are believed to be identical are observed in one case to produce PCBs and in the other not to produce PCBS. For these reasons, a firm My erect a full scale plant with a aultiailllon dollar investaent, only to find that an exasptlon la required for operation. At this point, delay in approval is the cost of the investaent at today's cost of aoney plus the lose in product on the Mterlal to be produced. In addition, product is sold in advance of production and delay can cause a chain reaction of siallar costa to custoMrs. Firaa will be reluctant to chance this kind of event for all cheaical processing situations where PCBa have been observed to occur. A second type of Innovation iapact will fall aost heavily on SMll Mnufacturing operations, whether by larger or t SMll flrM. These flrM .will have to file an exeaption petition [ HONS 013518 I (or operation of vary bhII and/or one-tlae nanafactura of chemical subatancaa. Thaaa patitiona will have to ba (Had In dvanca, knowing that tha procaaa haa tha potantlal for ganaratlon of PCBa. Thua manufacturers of auch items aa apaclalty organic chemicals will hava to toll customers that tha aatorial roquaated cannot be produced until EPA approves an exemption petition for tha production. Such a aituation would have a far more davaatating Impact on innovation than the Preaanufacture Notification Rule (PNN) of TSCA Section 5, bacauaa tha prohibition la not United to 90 days, and tha filing eoata could be at laaat aa great aa thoae of PHN rule. [Cite to CHA's coausenta to EPA on the economic lapact of EPA'a propoaal PHN program.] in auanary, dalaya aaaoclatad with auch conpllance procadurea can ba substantial, and can apall the difference between proceeding with a aigniflcant procaaa inveataent and cancelling it. Such delaya involve potentially large dollar anounta in terns of the increased cost of both facilities and equipment (due to inflation inpact), high Interest and finance cost, ultimately lost sales, loss of competitive advantage, and can involve as well, lost opportunity costs when the flra has alternative developaental choices which were foregone. International Trade Impact Reluctance of U.S. businesses to produce products covered by the exenptione process would open the door to foreign competition, giving foreign firms an undue cospetitive advantage. HONS 013519 For example, In Canada, incidental generation of PCBe la not Halted per ee. Limits are placed on releasee to the environment and In concentrations of PCBs In the product. Thus, a process which would require an annual exemption in the United states bears no restriction in Canada. American industry could lose Its competitive position with respect to certain materials to those foreign producers who are free to develop, produce, and export to the U.S., and elsewhere, products derived from processes which generate PCBs as an unwanted by-product. Several examples of this have already occurred. One Cha ember coapany reported that they were unable to manufacture a valuable new industrial product because the PCB concentration in the manufacture process streams would be greater than SO ppm. The new product was a 50iSO mixture of components A and B. Manufacturing component B would be relatively easy and contains PCBs in concentrations much lower than SO ppm. However, its. manufacturing process would contain between 70 and 100 ppm PCB. It was therefore necessary to purchase component B from a Japanese chemical company. This will potentially result in an annual purchase cost of $2,000,000 versus an estimated manufacturing cost in the U.S. of $1,500,000. Moreover O.S. jobs have besn lost to Japan. The situation may also arise where manufacturing industries which use PCB contaminated product would have to relocated outside the United states. Though it was beyond the scope of this effort to estimate the precise economic significance of processes affected by incidental generation, indicationa are that these basic and HONS 013520 rl * widely used processes esploy substantial numbers of workers whose jobs could be significantly reduced or eliminated by the threat of curtailing investments resulting from exemptions requirement. In addition, loss or disruption of customer markets by feedstock supplisrs, and loss, reduction, or higher cost of products resulting from curtailed or eliminated processes for which off-setting substitutes are not found, would make their own significant adverse Incremental contribution to unemployment. Productivity and Efficiency Loss of innovation of coat-saving or new production procaaaes because of a incidental generation exemptions petitions process, could result in substantial adverse impacts on efficiency and productivity. These effects might go well beyond the lost opportunities for cost reduction or innovation product development for the individual firm by extending to the economy as a whole. . Such an impact would occur because the benefits from a particular innovation are not confined solely to cost savings or new revenues for the originating firm, but rather, may be passed along to eoapetlng firms who may achieve new efficiencies or manufacture new products by imitating the innovating process, or using it as a basis for improvement. An innovation nay be passed along to customers and suppliers, to whom it may becoaw the market stimulus for their own innovative activity, and/or the besla of lower-priced outputs. One study, based on a limited number of chemical industry innovations, showed that the social rate of return on an innovation (counting the added benefits to HONS 013521 V evitoMtt and competitors) amounted to more than twice the private rate of return for the individual firms.* Workload Impact on BPA Staff Though not a specific burden on fires, there is yet another significant administrative cost associated with a uniform exemption provision that merits discussion here. This is the burden on the coapliance and enforcement resources of the government. As indicated above, EPA received over 70 applications for exesptlon at the end of 1978. Bach applicant for an exemption is entitled to a hearing and an opportunity for cross-examination on disputed issues of fact. (TSCA Section 6(e)(3)(b); 6(c) (2), (3).) Even before the EOF decision, two years had passed without EPA's having taken final action on the exemption petitions. Under an exemption petitions process, such a petition would have to be renewed annually, with potentially a new hearing each year. Given the potentially hundreds of . processes which would be potential candidates for exemption proceedings, this process would constitute a significant burden for an agency already under substantial budgetary and resource constraints, and with many more significant enforcement responsibilities in areas posing much greater potential for unreasonable risk. Any unnecessary regulation that results In exemption applications should be avoided. Without the exclusion of unintended PCB by-products in controlled systems, the Mansfield, Edwin, et. al. The Production and Application of Mew Industrial Technology pp. 144-158. HONS 013522 aj administrative burdan would ba groat, particularly if the 50 ppa cutoff were lowered. For exaaple, for 1000 exanptlone, EPA would have to hold five public hearings per day, each working day or the year, just to allow public conwnt. HONS 013523 X. THE EFFECT OF A REGULATORY CUTOFF ON THE LEVEL OF AMBIENT PCBS (TECHNICAL PORTIONS OF THIS SECTION MILL BE CLARIFIED AND SUMMARIZED IN LAY LANGUAGE] Tha Agency aeke the question (46 f.r. 27621) Could a regulatory cutoff be established such that uncontrolled PCBs below that cutoff would not significantly increase ambient PCB concentrations? Nhat cutoff would be appropriate and why? It can be shown that (1) the anbient PCB concentration will decrease regardless of the regulatory cutoff and (2) that the rate of decrease will not have such relation to Incidental manufacture. The National Research Council^) has calculated that as of 1975, there were 68 million Kilograms of PCBs that were free In the environment. Of this quantity, the MRC estimated that 38t of the total quantity was comprised by the lower chlorinated blphenyla. There is ample evidence In the Support Document of the May 31, 1979 rule and in recent literature!?) to show that the lower chlorinated biphenyls degrade with a half life of five years. Hence, 36.84 of the 38% lower chlorinated blphenyla would degrade between 1975 and 2000. This represents over 55,000,000 pounds of degradation in spite of (?) Beaton, A. M., et. al. (1979) Polychlorinated Biphenyls, National Acadesqr of Sciences, Nashington, D.C. (?) (2) Tabak, H. H., Quave, S. A.r Maskin, C. I. and Barth, E. F.. Journal of WPCF, Vol. 53, No. 10. p. 1503 (October, 1981) HONS 013524  Ignoring the slower, but significant, degradation of the higher chloroblphenyls and using a degradation half life that is probably overly conservative by a factor of 5-10. The Agency, in the preamble to the Nay 31, 1979 rule, estimated that 100,000 to 500,000 pounds of PCB's were generated by incidental manufacture. Exemption petitions have been filed for Incidental manufacture in concentrations greater than 50 ppm as required by the rule. The Agency can revise the estimate baaed on that data and the Agency can estimate the percentage of such PCBa that will reach the mobile environment. In addition, CMA presents with this submission data relative to incremental manufacture of PCBs and the fate of such >CBa for processes not regulated in the Nay 31, 1979 rule. These data indicate very low generation of such PCB's and of those almost all pre diverted to controlled disposal site. The total quantity of chlorinated material incidentally . generated between today and the year 2000, as represented in the CNA survey, would be less than O.St of the quantity degraded as defined above. This quantity is infintesimal in relation to the amount of PCBs which will still be mobile in the environment. It may be useful to examine a purely theoretical calculation in order to appreciate the degree of impact on the ambient environment from PCBs incidentally manufactured. HONS 013525 Theoretical Calculation Assumption lit 10,000 lba. of incidental PCB'a released par year. Assumption |2i Such PCB'a degrade with a five-year half life. A - *o (l-e-*t) X~ A - Hew PCB'a, 1980-2000 k Rate of addition - 10,000 lba./year k * Rate eonatant for dissipation A 67,000 pounds A 0.126 of the above calculated degradation. A 0.076 of the undegraded PCB'a from above calculationa. Thia vaatly over-aiaplified calculation ignorea many factora but pute in perspective the lapact of incidental PCB'a relative to the amount of PCB'a now abile in the environment. Biodegradation Degradation ratea have been measured on some chlorlanted biphenyls. Environmental measurements, although preliminary, appear to confirm a decrease in ambient concentrations already MOHS 013526  occurring.(h) it it Impossible to get sn exact half life for each isosMr since each homolog will vary as will the air, wits and sediment degradation rates. Order-of-magnitude estimates can be made and degradation pot into perspective. Furukawa et al () investigated the effect of chlorine substitution on the biodegradability of polychlorinated biphenyls. They used two pure strainsi Alcaliqene Y42 and Aclnetobactsr P6. They came to the following conclusions: A. Degrsdation decreased as chlorine substitution increased. I. PCBe containing two chlorines on either the ortho position of a single ring, i.e. 2,6 or on both rings, i.e., 22' showed poor degradability. C. PCBo containing all chlorine atoms on only a single ring were generally degraded faster than when the same number of chlorines were substituted on both rings. . D. Ring fission of the molecular occurred preferentially with nonchlorinated or lesser chlorinated rings. Since these authors Indicated the microbial cell count in cella/mL, it is possible ot convert their data to a second order rate constant and subsequently estimate the lifetime of the chemicals in a typical lake or river. For this purpose, the data suamarlsed in Table I for the Aclnetobacter strains will be used. TB) National Academy of Science, p. 21. ~ <b> Great Lakes Fish - Ck. Craddock 'Furukawa, K. Tonomura, K. and Kamebayashi, A. (1978). 'Effects of Chlorine Substitution on the Biodegradability of PCBs", Appl. Bnvlron. Microbial. 35:223-227. HONS 01352 7 In order to convert this rate data into a half-life, a knowledge of the concentration of Microorganisms is required for representative water bodies. Baugham and Lassiter!*) gave such concentrations and these are shown in Table II. They also assumed that only 10* of the natural population of organism would be active in any degradation. Accepting this assumption, the half-life for a lake and the bottom sediments are shown in Table III. These data are in order of magnitude agreement with specific degradation experiments and with observed decreases in environmental concentrations. One such degradation experiment is reported by Tabak et al. Photodeqradation Using the correlation analysis of Hendry and Kenley(b) a rate constant for photodegration in the atmosphere can be estmated for the various chlorinated biphenyls. The correlation is based on the Haunt sigma constants for the chlorinated phenyl species. Assuming that all the chlorinated phenyls have a sigma similar to phenyl, an estimate was ude of the rate constant for gas phase photodegradation and atarespherlc half-life for chlorinated biphenyl and the results are shown in Table IV. Due to this assumption all half-lives reported in Table IV for the higher biphenyls are probably too short. However, it does ^Baughman, G. L. and Lassiter, R.R. (1978). 'Prediction of Environmental Pollution Concentration Estimating the Hasard of Chemical Substances to Aquatic Life*, ASTM STP 657, John Cairns, Jr., K. L. Dickson and W. Makl, Eds. ASTM p. 35-54. ') Hendry, P.G. and Kenley, R.A. (1979). 'Atmospheric Reaction Products of Organic Conpounds*, EPA-560/12-79-001. 5 HONS 013528 Indicate that chlorinated blphenyla are photoreactlve. Hence thla becomes an important cleansing reaction for removing PCB frosi the envlronswnt. Conclusion The data clearly Indicate the eoaplexity of calculating envlronsMntal degradation rate for a real world situation. However, It is obvious that degradation does occur at Its possible to Illustrate the inpact of aoblle PCBs fros incidental aanufacture using order of magnitude estimate, such order of magnitude estimates.are clearly simplistic but serve to show the major pointsi PCBs in the ambient environment will decrease and the rate of decrease will not be influenced by release of PCBe from incidental Mnufacture. One can find many issues to debate relative to any specific calculation but the overall result will be the sane. Arguments could fill pages but the facts remain that well over 100,000,000 lba. of PCBs are in the environment, that Incidental PCB generation would require centuries to develop quantities of cosparable magnitude and that degradation, even slow degradation, of the material currently in the environment will result in a net decrease. HONS 013529 DA'S PROPOSED EXCLUSION POE CLOSED NANDPACTORIRS PROCESSES/ cnwTfit.Tjn WASTE PROCESSES In ita Nay 20, 1981 ANFR, SPA daacribes and aaka for comment! on a propoaad exclusion frou regulation for 'closed* manufacturing processes or procaaaaa Iron which all waataa ara 'controlled* (46 Pad. Rag. 27617, Nay 20, 1981). That propoaal, in part, ia a product of diacuaaiona batwaan CNA and EPA. During CNA'a negotiation with EPA on tha nature of tha data collection Effort tha Aaaociation could undertake, CNA indicated believed that there were a significant auafear of cheaiical induatry procaaaaa containing PCBa which had one or both of two charactariatica - tha PCBa were either `deatroyad by chemical or other reactioaa within the proceea or ware aaaentially ancloaed within the proceaa. CNA commented that, while acme matber of solocales of PCBa might at aome point escape from the proceaa, for all practical purpoaea there would bo no aignificant exposure to thoae PCBa from outaide tha proceea system raqaxdlaaa of tha concentration of PCBa within tha aystem. CNA and EPA agreed that if what CNA believed to be the case proved true, a aeparate category of excluaions from PCS ragulationa might be juatified. CNA agreed to collect data from ita ammbara regarding the exiatenca and prevalence of thaaa "" ..... ' ayatama and naked EPA to propose euch an HONS 011510 -i- 71 eceluaion la bb advanced notice of proposed rulaaaklag. SPA agreed. " ' Oafortunately, 1b two raapaeta, spa*a proposed version of thla exclusion does not conport with tha details of tha seclusion which CMA nrgad SPA to propoaa. First of all. It aada bo aaaaa at all to propoaa auch aa exclusion for only thoaa procaaaaaa which had a PC* concentration of 50pea or laaa. Ctteh aa axclaaloa would add nothing which a raprcaulgatad aaaaral regulatory cut-off of SO ppa would aot achlara. IYA ataff aaaorad CMA that consideration would be given to axelading qualified procaaaaa ragardlaaa of their PC* eoaeantratloa. Tha AVPK doaa aot by It* word* propoaa that ` option. lowerar, wa aaauaa that SPA will evaluate tha merits of ita final pcopoaal oa tha baala of that aana logic it pravloualy aharad with CMA. Secondly, CMA reslatad tha SPA'a uaa of tha term "cloted ayatae*. In tha courae of aagotiatloaa both CMA and SPA ataff undaratood that aa a technical nattar it waa probably iapoaalbla to prevent at laaat ace* amber of PCS aoleculea or ace* nlalacula eoaeantratloa of PCS free escaping a given proeaaa. HONS 013531 -2- /OO Xt that time CNX favored aaa of tha tan "essentially enclosed" with a request for views and data as to how that term Bight ba dafinad. Tha majority of tha DA staff involved favored nsa of tha tarn "closed system" with a request for vlaws and . data on how it Bight ba dafinad. t CKX's mambarship survay provides evidence of only 4 proeassas oat of tha total of 13S raporting an incidantal ganaration of FCBs which tha respondents dafinad as "oloaad*. Tbara is low question as to wbathar aran thasa 4 proeassas ara in fact eoaplataly ciosad. However,' basad on CMX's report to DA on tha aaesrtaintlss involved in tha analysis of process streams for low concentrations of PCBs, it is possible that this data merely reflects a limitation of available analytical tools. As CNX has consistently maintained, there exists no chemical process dbieh incidentally generates SCBs or which uses as an input PCS contaminated material from which soma miniscule amount of PCS will escape. "' HONS 013532 3 1-01 . EPA's proposed excluaion would operate only If DA were eeeured that no PCBs entered the produet. However, such a requirement ie obviously impractical. More importantly, we believe that the Agency should focus its Attention on . the nature of PCBs and the potential for exposure to them rather than on what eight constitute a technical definition of a dosed chemical, process. -Accordingly, insofar as the data collected by CMA to date is concerned, there does not appeal to be a basis for developing an exclusion for regulations of the type proposed in the Agency's AKPA... However, it does appear that sub stantial evidence exists to support an exclusion frost regu lation of certain types of-processes which, on the one band, contain various doncentrations of PCBs, including eoncentra tions above 50ppm, but where virtually all PCBs exiting the ' process are directly controlled or are in a state which demonstrably do not pose an unreasonable risk to health or the environment. HONS 013533 / ox. To illustrate what type of an exclusion night be utilised, assume firat that a particular chanleal proceaa containa any eoncantration levels of PCBa. Aa to any diacharge from that procaaa which ara to navigable water, or publicly owned treatment works, the atringent diacharge prohibitiona of the Clean Mater Act are applicable. Othe waatea or diachargea fron the proceaa would have to be diapoaed of or handled according to applicable DA PCB dlapoaal regulations. [See discussion of those regulators infra at______ .1 As to employee exposure within the workplace, several laportant sets of restrictions would apply. Pirat of all, OSHA baa established a PCB workplace threshold Unit value for PCBa is designed to assure a safe workplace (supra pp._____ to ____). Purthnore, threshold Unit values applicable to primary chemicals in a proceaa which incidentally contains PCBa result in PCB exposure ordera of magnitude below the OSBA tlv for PCBa. (supra pp. _ to____). A partial examination of EPA's files containing pending industry petitions for execution from the Agency's PCB rules reveals a number of procesa streams incidentally generated PCBa which also primary chemical which are subject to various EPA and OSHA regulations (see infra at _____ ). Nbrkplace exposures are theoretically of three types--Inhalation, ingestion and dermal. The potential for inhalation exposures from PCBa contained in chemicals which have accidentally spilled or leaked is infinitesimal. As the PCB HONS 013534 * /o3 vapor pressure calculations set forth at pp___ to ____ above illustrate, the possibility of exposure by inhalation can be disaiissed for all practical purposes. There are two important points to observe with respect to the potential for workplace exposure to PCBs by dermal routes and by ingestion routes as a result of dernal exposure. First of all, the worker protections are especially effective in view of the recent health effects infonaation concerning the low toxicity of PCBs exposure (see supra pp__ ). Secondly, worker exposures which night occur as a result of accidental spills or leaks would be Minimised or eliminated by routine instituted worker protection and good management practices. These potential closures are Miniscule compared with the potential exposures in the authorised non-enclosed uses of PCBs which have already been sustained by the D.C. Circuit Court of Appeals. See discussion infra. HONS 013535 On connection with- those authorized usee, the Adaiiniatrator found that proper protective clothing and good management practices should reduce PCS exposure to vary low levels the regulations permit routine servicing of PCS transformers and electromagnets". Furthermore, insofar as spills or leaks of PCBs in concentrations above 50ppm occur in plants, the existing PCB disposal regulations would again be applicable. Finally, as explained above, CMA believes that PCB concentrations below SOppm in products do not pose an unreason able risk to human health or environment (supra pp___ to___ ). Accordingly, CMA believes substantial evidence exists to "a*. support an exclusion from regulation fort Any chemical process containing any concentration of PCBs where .* 1. all wastes or discharges are handled or disposed of in accordance with applicable EPA or OSHA PCB regulations; 2. The PCB concentration in products of such process do not exceed SOppm (or such highsr level as . EPA suy establish by rule). HONS 013536 4 -yrr AH ANALYSIS OF THE LEGAL, PROCEDURAL AND POLICY ISSUES IH EFA'S REVIEW OF THE RECORD IN THIS PROCEEDING Introduction and Summary I? The purpose of this section of CMA's aubeiaaion to EPA Is toi 1* review the legislative and regulatory history of PCHa under the Toale Substences Control Act; 2. review the decision In Environmental Defense Fund v Environmental Protection Agency. 63* F.2d 1267, (O.C. Circuit 1980) In setting aside certain of BFA's PCB regulation 3. detail the purpose, nature and scope of CHA's voluntary agreement to assist EPA In collecting Information relating to PCBs; 4. summarlae other significant regulatory controls of PCIs ( 5. suggest the legal, policy and procedural Issues which EPA should take Into consideration in reviewing the record In this proceeding. ,6^es>a & THE LEGISLATIVE AMP REGULATORY HISTORY OP PCBS ORDER THE TOXIC SUBSTANCES CONTROL ACT r 4- in general. Congress established In Section 6(e) of TSCA a detailed, phased scheme for -the disposal of PCB's and HONS 013537 K the llaltation of lta Manufacture, procaaalng, diatributlon, and uae. Section 6(e) directa the Agency (by January of 1977) to have prescribed PCB disposal Methods and to require that PCB containers to be Marked with appropriate warnings. By July of 1977, PCB's were to be Manufactured, processed, distributed and used only In a "totally enclosed" Manner. By July of 1976 all Manufacture of PCB's was to cease. Sis Months later, all processing and distribution of PCB's in ecauasree was to be prohibited. The Act contains two narrow exceptions to its regulatory scheae for PCBs. First, by rule the Adadnlstrator aay authorise the continued use of PCB's In a non-totally enclosed Manner if he finds that such activity "will not present an unreasonable risk of injury to health or the envlronaent.* (] Second, the Adainlstrator nay grant ease-by-case, year-by-year exasptiona to the prohibitions on . Manufacture, processing end distribution, on such conditions as he My set, and on the basis of his finding that "an unreasonable risk of Injury to health or envlronaent would not result, snd ...good faith efforts have been Mde to develop a chemical (substitute] which does not present an unreasonable risk of injury to health or the envlronaent."! ) On February 17, 1978, EPA proaulgated a PCB Disposal and Marking Rule pursuant to Section 6(e)(1) of TSCA (43 Fed. Reg. 7150]. That rule was subsequently clarified by awndMnt (43 Fed. Reg. 33918, August 2, 1978] and slightly \9< HONS 013538 i aodifled In the publication of the ban rule challenged in tor v, EPA. On June 17, 1978 EPA proposed (43 Fed Reg. 24802) and on Hay 31, 1979 finallted (44 Fed. Reg. 1514) lta ao called 8an Regulations which were the subject of litigation in EOF V EPA. The Host significant aspects of the Ban Regulations for present purposes were the Agency'si 1. exclusion froa regulation of any aaterlals Containing less than 50 ppm of PCBs; 2. definition-of statutorily exempted "totally enclosed uses" (in general, nonrailroad electrical transforaers, capacitors and electroaagnets) as uses which were "intact and nonleaking*i 3. determination to allow the continued use of eleven non-totally enclosed uses: servicing of non-railroad transforaers: use and servicing of railroad transformers; use and servicing of aining equipment; use in heat transfer systems; use in hydraulic systems> use in existing stacks of carbonless copy paper; use in pigments; ervlelng of electroaagnets; use in natural gas pipeline cospressorsi use in saall quantities for research and development; and use as a aicroscopy mounting aediua. (See *4 Pad. Reg. at 31549-51 (1979), to be codified in 40 CFR *741.31.) The first two regulations - the 50 ppm regulatory cut *f and the definition of "totally enclosed uses", were set I&t/ HONS 013539 I Id* by the D.C. Circuits the third set of regulation* *re subetalned by the Court. Unfortunately, whefia Congreaa enacted Section 6(e) of 78CA neither Congress, industry or other interests concerned about PCBe knew that in addition to PCBs purposely sianufactursd for various cossaercial uaes there also were low levels of PCBs unintentionally generated or concentrated as by-products in a wide range of cheaical processes. They certainly knew nothing of the risks of hunan or environmental exposures for these PCBs and they did not know that these PCBs varied significantly in cheaical, physical and toxicological characteristics from purposely aanufactured PCBs. 8*ctlon 6(e) was added to the Senate and Bouse T8CA bills as a floor anendawnt in both the Senate and Bouse debates. (See Bouse Coaaittee on Interstate and Foreign Coaaerce, Legislative Bistory of the Toxic Substances Control Act, 94th Cong., 2d 8*ss. at 213, 580 (Com. Print. 1976) (hereinafter cited as "TSCA Leg Is. Blst.*]. Thus, it was not the subject of the thorough and coaplete coaadttee analysis which other TSCA natters were accorded. Bowever, that the focus of Congress' concern was on intentionally nanufacturad PCBs of the so-called Arnohler type, which were produced, when TSCA was passed,/only by Monsanto, clear from a reading of various studies SBB rtpoTta referred in the legislative history. HONS 013540 For Instance, report prepared by the Legislative Research Service of the Library of Congress (tsca Leg la. History at 235) refers to Monsanto as "the sole United States producer of PCBs* and that the eoipany 'limited PCB sale to a few companies for use In electrical closed systems.* The Council on Environmental Quality reported that KBs are 'also known as 'Aroclors* (TSCA Legis. Hist, at 776). Senator Tunney, further evidencing the focus of Congress' concern on highly concentrated mixtures of KBs of the type included in electrical equipment stated: Despite assurances from the sole domestic manufacturer of KB's, they are still escaping into the environment in awesome quantities, in fact some 10 million pounds per year. Although an agreement several years ago by the Monsanto Co. supposedly restricted PCB's use to closed systems, Monsanto obviously has no control over those to whom it sella KB's nor does Monsanto have any control over those who iaport KB's. Id. at 240. Even a^re vivid evidence that Congress was targeting Aroohlor type RCBs is the statement by Congressman Dlngell, the House sponsor^! the Section 6(e) amendment, that the restrictions it contained would have no economic iapact because it was consistent with Monsanto's plans to phase out its production of KBs. The CongresamajK^tated: I spoke, on the way ovp^here. to a representative of one of the majormanufacturers of KB's, and he advised me his company does not object to this amendment. Mis^doipany felt that it is a good one and it wouldjimlp his coapany to be citlsens. The HONS 013541 amendment^&qes not call, I want my colleagues to know, for lMM^at^prohibltlon but, rather, a gradual phase ouJ/T to assure action within a reasonable period of tlse. TSCA tegis. Hist. at 582. " Environmental Defense fund v. Environmental Protection Acencvi -the 50 dpw Regulatory Cutoff As part of its regulatory scheae for PCBs, EPA Halted application of its disposal and Ban Regulations to articles and materials containing concentrations of at least 50 ppa of PCBs. In alaost all cases, naterials with lower concentrations reaained unregulated. EDP contended that this lialtatlon contravened the statutory coaaand in Section 6(e)(2)(A) and Section 6(e)(3)(A) to regulate *any polychlorinated byphenyl.* While the Court did not adopt all of EOF*a reasoning, it held that, under the applicable standard for judicial review, there was a lack of substantive evidence in the record to support the Administrator's decision to establish a regulatory cut-off at SO ppm. In the proposed Ban Regulations, EPA had listed four reasons for setting the regulatory cut-off at 50 ppm limit would 'exclude from the rule municipal sludges and other mixtures containing low, (less than 50 ppm) levels of PCB's whose presence is due to ambient levels of PCB presnt in the air or water.* 43 Fed. Reg. at 24804. Second, EPA believed that some industrial chemical processes inevitably produce traces of FCBs, and that careful control could reduce the concentration of PCBa only to 50 ppm. Third, EPA felt that it was iiqpractlcal to regulate the 'diffuse HONS 013542  - Il*> and extremely numerous PCB sources* with concentrations below SO ppm. Id. BPA believed that the proposed cut-off would ensure maximum effectiveness of the regulation by focusing agency attention under TSCA upon the aost significant and controllable sources of PCB exposure. id. Fourth, the Agency believed that other statutes were available to regulate low concentrations of KBs, particularly In municipal sledges and dredge soils. The Court agreed with BPA that the legislative Intention was not to require regulation of ambient concentrations of KBs. Bowever, It determined that, while some cut-off light be appropriate, BPA had not explained why the regulation could not be designed expressly to exclude ambient sources, thus directly fulfilling the congressional Intent, rather than achieving that goal Indirectly with a cut-off of SO ppm. Nor did the Court find persuasive EPA's justification for the regulatory cut-off on the basis of the serious i^act a lower court would have on industries that inadvertantly produce KBS during the manufacturing process. Ths Court emphsslsed that, as the Agency itself had conceded, inadvertent commercial production of PCBe was intended to be regulated under Section 6(e). By providing a blanket exemption for concentrations below 50 ppm, the Court found the the Administrator had circumvented the authorisations and exemptions requirements provided In the statute. The Agency had made no finding that the cut-off posed no unreasonable risk to health or the environment, which is the finding required by Section 6(e). Consequently, the Court held MONS 013563 II' I the burden* faced by industries could not be the sole besis for the 50 ppm cut-off. The Court also rejected the argument that the 50 ppm cut-off was justified on the basis of an Agency's inherent authority to create exemptions to a statute based on administrative necessity, or upon an Inherent authority to exempt de minimis concerns from regulation*. EOF v. EPA, supra, slip op. at pp. 25-38 It should be emphasised that the Court did not Independently tO* ,-<! determine that incidentally generated were covered by Section^" - fi-Jl cK-nJt U /*.euS ao a C(a) of TSCAjj^ A careful and fair reading of footnote 37 discloses that the Court recognised that Congress did nbt even know about such PCBs. However, it thus proceeded to adopt as "reasonable* EPAs conclusion that Incidentally generated PCBs were covered by PCBs and that ...Congress's express concern with widespread . contamination and worker exposure, tne chemical's toxicity at extremelT low levels and the atatuatory language of "any polychlorinated biphenyl in any manner lead us to conclude tbat Congress intended to include all commercial sources within the EPA - regulations...(emphasis added.) But CMA's data demonstrate that the concern which the Court supplied to Congress was misplaced! there is no widespread worker exposure to these PCBs, they are not toxic at extremely low levela, and there are chemical, physical and toxicological dlfference*between the two categories of PCBs. In addition, as will be explained below, CMA's submission contains substantial evidence to support a regulatory cutoff of MOMS 013544 not loan than 50 pp* by demonstrating that such a cutoff does nto posa unraasonabla risks to huaan health or tha environment, is naeassary in ordar to avoid imposing unraasonabla administrative burdans on EPA; regulations aora stringent will achieve banafits of da alnlala or trivial value. fi CHA'S AOKEEHBHT TO COWDOCT AH HITCNS1VE. EXPEPITBD mrOWIATIOW CATHBBIHO EFFORT ' CHA was not a party to EPF v. EPA. Navarthelass, it voluntarily antarad tha negotiations among tha partlaa in tha wnka of that decision which had as its goal tha structuring of an expedited data collection effort to assist the Agency in developing jla new regulatory prograa for PCBs. All of tha participants recognised that if the Court's Opinion setting aside tha 50 ppm regulatory cutoff want into effect, tha regulation of PCBs would be governed by the litoral language of the Act. Ho PCBs could then be anufaetured, distributed or used without tha grant of a pacific exaaptlon by EPA. Thera was by this tlae a recognition that low concentrations of PCBs sight wall exist In a great aany Industrial processes. Accordingly, tha Protpaeti ware that EPA and industry would be unreasonably burdened by an extensive petitions process. The participants also recognised that a large aagxwnt of industry utilising transformers, capacitors and lectroasgnets would be out of coi*>llance by virtue of tha HONS 013545 |V3 10 Court having aat aaida the Agency's definition of what constitutes statutorily exempted 'totally - enclosed* uses Of PCBs. Agreement was rsachad that if the details of how industry and EPA might cooperate in a data collection effort could be agreed upon the parties would file joint notions with the court seeking a stay of the mandate of the Courts opinion during that data collection effort. CMA was one of three trade association which agreed to undertake specific data collection efforts._/ The Agency aaked CMA to collect, on an expedited basis, 'scoping* data, _/ The Edison Electric Institute and Dry Color Manufacturers Association, parties to the litigation, also agreed to assist the Agency, information on hand, reasonable estimates or narrative dissertations, which would help the Agency go 'back to the drawing boards* and plan its reconsideration of the regulation of low concentration PCB's. CMA agreed to provide all three types of data on certain conditional (1) because rapid production of complex, competitively sensitive data was required, the confidentiality of coapany data sources had to be maintained; (2) only aggregate, non-confidential data would be provided to the Agency although CMA's consultant would describe to EPA the procedures utilised in collecting and analysing data which had not been agreed to in advance; (3) the Association would have to be given adequate time to collect the data; (4) de novo data collection would not be HONS 013546  required given the expedited schedule the Agency, proposed: (5) based on CMA's belief that most of the incidentally generated PCB poundage exists in essentially closed processing systems with controlled waste, the EPA would propose an exclusion froa its regulations such PCB's in its AMPR on the 50 ppa issue from its regulations: and (6) because of the aany 'unknowns* connected with the 50 ppa Issue, sgreeaent on all substantive details was necessary before the joint notion on this issue was filed with the Court including agreeaent on CMA's survey questionnaire, survey instructions and protocols for the analysis of the data. CMA also presented the curricula vita for its consultants to the litigants. After an Intense schedule of cordial but detailed negotiations between CMA, EPA and EOF, the following prograa was agreed to: 1. CMA's entire submission would be given to EPA within 310 days of Court approval of the request for a stay on the 50 ppm issue; 2. PCB data would be collected and reported to EPA in terms of its relationship to certain manufacture or process situations (chlorination; chlorinated interxndiates; Chlorinated solvent; contaminated input; and other) and, as to each of those situations, in terms of its relationship to certain structural classifications (alkanes, alkenes, alkynes, cyclo, aromatics). HONS 0135*7 (I* vS The manufacture or procaaa situations identified were selected because they were recurrent factors in Industry petitions to EPA for exemptions for stream with PCB concentrations above SO ppn. Logic dictated that they eight also be the situs of low concentration PCB'a. 3. PCB data for any concentrations below 50 ppn for each of the general process/structure categories would, in general, relate toi (a) PCB poundage per year; _ ... (b) the general fate of the PCB's; (e) the general exposure characteristics associated with the PCB's> <d) whether PCB controls had been applied to the stream, and (e) general use characteristics of the product. 4. Narrative dissertations would be provided on the following topics: (a) the problem and costs associated with the analysis of complex chemical stream for the Identification of the presence and concentration of PCB's.* (b) problem and costa associated with the reduction of PCB concentrations in complex chemical atresias. ' (c) recent information on the toxicity of PCB's.* 5. Attempts will be made to relate the data collected to a larger aegmnt of industry. HONS 013548 9 Additional dataila of CMA'a undertaking are aet forth in the report on oour incidental generation aurvey. In addition to responding to EPA's request for data on incidentally generated PCBa CMA also indicated that it would conduct a survey of its members on the so-called 'totally enclosed uses*. That Information is being submitted separately, on DecasUser 7, 1981, in response to the Agency's advanced notice of proposed rulemaking on that issue._/ J Cite HONS 013549  (<&*u** Q++A-- I WMHi BBQCgaUBAL-Mm-QLICY TfiBUg Jli BPA'S DSVgLQMUrt Or A REGULATORY PKOCRAH FOB IMCIDEMTALLI GgHgKATED PCB The major tasks which EPA faces in determining what its regulatory program for Incidentally generated PCSa should be arei 1. To utilise its discretion and expertise in implement ing TSCA in light of the latest information available to the Agency and to support its conclusions with substantial evidence. v<" y? ^ jl. 'Bo determine, on the basis of the latest scientific informatloni the nature of the health or environ- | mental risks posed by PCBs. To implement the Act's individual provisions while at the same time giving effect to the intent of Congress that the Act be implemented in a reasonable and pru dent manner, and that consideration be given to envi ronmental , economic and social impacts of Agency actions. 4. To propose regulations which are efficiently coordina ted with the Agency's other statutory responsibilities and the responsibilities of other agencies. HONS 0135M L Ill * 5. To nviax the costs and benefits of any proposed regu latory program and alternatives to that program on the basis of Executive Order 12291. Fashion * Regulatory Program for Incidentally Generated PCBa. In delegating authority to an agency. Congress is often forced to leave to the agency the power to Identify problems issues and solutions. This policy simply recognises that Congress and its staff work under demanding time constraints, and may lack an agency's scientific and technical expertise. Sunshine Anthracite Coal Co. v Adkins, 310 u.S. 381, 398, (1940). An agency la often the better vehicle by which to assure that government regulatory activities are economically, practically, scientifically, and administratively feasible. Consequently, an adminlatrative agency is often entrusted with wide latitude in interpreting its governing statute, especially when extremely complex economic, scientific, or technical judgments are essential to administer a particular law. HONS 013551 < In such situations, courts usually permit an agency's interpretation of its authorizing statute to be controlling, Weyerhaeuser Co. v. Costle, 590 F.2d 1011, 1025 (D.C. Cir. 1978); Lubrlzol Corp. v. E.P.A., 562 F,2d 807, 816 fn. 23 (D.C. Cir. 1977) f E.I. du Pont de Nemours t Co. y Collins, 432 U.S. 46, 54-55 (1977). This is especially applied to the initial, contemporaneous construction of a statute which requires an agency to act "under the gun" or under a tight statutorily mandated schedule. Norwegian Nitrogen Co. v. 0.8,. 288 O.S. 294, (1977); Udall v. Tallman, 380 U.S. 1, 16 (1968), and especially when scientific matters are concerned, Bercules Inc, v. E.P.A., 598 F.2d 109, 116,117 (D.C. Cir. 1978). When Congress provides an administrative agency with a specific directive, however, the agency cannot redefine or broaden the scope of its delegated powers by going beyond an unambiguous Congressional mandate. Alabama Power Co. v. Costle. 636 F.2d 323 (D.C. Cir. 1979); Asarco, Inc, v E.P.A., 578 F.2d 319 (D.C. Cir. 1978); Association of American Railroads v Costle, 562 F.2d 1310, 1315, 1318 (1977). Under no circum stances can an agency exceed authority which Congress has delegated to it, Trenton Chemical v U.S., 201 F.2d 776 (6th Cir. 1953), cert, denied, 345 U.S. 994, (1953), or act at variance with the statutory language. Miller v Laird, 349 F. Supp. 1034 (D.O.C. 1972). While an administrative agency may select the best means of implementing regulatory actions when HONS 013552 I they art net detailed in the governing legislation, Publlc Service Co--iaaion of the State of New York v r.P.C., 327 F.2d 893, 897 (D.C. Cir. 1964), agency diacretion is available only upon the agency's showing that Congress has allotted that diacretion to the agency. Citizens to Preserve Overton Park v Volpe, 401 U.S. 402, 416 (1971). . It is clear that theae principles apply equally to the E.P.A. Ethvl Oorp. V E.P.A.. 541 F.2d 1 (D.C. Cir. 1974), cert, denied, 426 0.8. 941 (1976). The courts have accorded the EPA flexibility in interpreting its statutory obligations, so long as the agency exercises its regulatory and enforcement powers to achieve the best protection of human health and the environment. Ethyl Oorp. v E.P.A.. 54f F.2d_at 24. JL- ' - ' In view of the fact that EPA new has data which was clearly ' uiwa not available to Congress eh it enacted Section 6(e) of TSCA, EPA should utilise its expertise and discretion and implement regulatory controls osad incidentally generated PCBs in a fashion which comports with all relevant scientific economic and administrative facts. HONS 013553 y rs3*- 0* J )***** vTh Ex--pPttlloon Petiittiioonn!! Proceee Will Impose Administrative ^ Burdens iffiwh BPA Ha-Msa Be Able lu Heel. Sole reliance upon the exception petitions' process as a mean of affording industry relief from inappropriate regulation npmx. .>nwun4n a^wumhVix mifM. of PCBs. Approximately 460 manufacturing, processing, distribution in com merce and use exception petitions^/ have been filed with BPA. None have been acted on. If these petitions undergo/the longer administrative procedure of a full rulemaking, that procedure will take an inordinate amount of time, and regulre greater personnel needs at the agency. As noted in our incidental generation survey, over 1300 exemption petitions could be filed with the Agency. Under the proposed rulemaking procedures for each exemption petition, each applicant for an exemption would be entitled to a hearing and the opportunity for cross-examination on disputed issues of fact. TSCA Section 6(e)(3)(b)>6(c)(2),(3). Bach . petition would have to be renewed annually, with potentially hundreds of processes which would be potential candidates for exemption proceedings. This process would constitute a significant burden for The Agency. For exanple, if 1,000 exemption petitions were filed, CMA has calculated that BPA would be required to hold five (5) hearings per day, each working day of the year, to allow a full hearing of public comments. As the court recognised in B.P.F v B.P.A., citing Alabama Power Co. v Costie, 636 P. 2d 323 (D.C. Cir. 1979), HONS 013559 ...n agency nay depart froa tha requirenent of a regulatory statute. 8-- Id. at 3S7-360 (opinion for tha court by Leventhal, J.). Nhlla tha court In Alabana Powar emphasised that *(c)ategorica exesptloni~f roa tha clear coaaanda of a regulatory statute, though aoaetlaes peraitted, are not favored,* id. at 358, it also noted thawt there Is `substantive authority (for an agency] to take appropriate action to cope with the adainistratlve impossibility of applying the coaaanda of the substantive statute.* Id. at 358-359. However, *(t]he agency's burden of justification in such a case Is especially heavy.* Id. at 359. Considerations such as the availability of enforceaent resources are relevant to the adainistratlve necessity exeaption. The theory of adainistratlve convenience is not new. It Is in fact a longstanding rule that an Adainistratlve `agency confronted with a coaplex task aay rationally turn to slapllclty in ground rules, and adainistratlve convenience, at least where no fundaaental Injustice is wrought.* Culf Oil Core, v. Hlckel, 435 F. 2d 440, 446, (D.C. Cir. 1970), citing Caralchael v Southern Coal Co., 301 O.S. 495, 511, (1937) See also, Persian Basin Area Kate Cases, 390 O.S. 747 , 777 (1968). If regulatory action results in an undue burden on an adainistratlve agency, courts have allowed the agency to aiaplify its procedures. The court in Alabaaa power Co. v. Coatle applied the theory of adainistratlve convenience to the EPA, once the agency daaonstrates the Infeasibility or lapossiblllty of nesting the adainistratlve deaands involved! HONS 013555 Court* frequently uphold streamlined Agency approaches or procedures where the conventional course, typical case by case determination*, would as a practical natter, prevent the agency froa carrying out the mission assigned to it by Congress. 636 P. 2d at 358. EM cannot be expected to perform impossible tasks M.K.D.C. v. Train. 510 P. 2d 692 (D.C. Cir. 1975). The exemption petitions process as contemplated by the agency in its ANPR would constitute or approach such an impossible task. Nothing in TSCA requires the agency to embark upon such a rigorous and administratively infeasible for incidentally generated PCBs in closed manufacturing processes. The principle of de minimis non curat lex also applies in this proceeding. The PCB regulatory proceedings are particulary suited to the de minimis exception to statuatory commands. The court in B.P.P. v E.P.A., supra.'addressed the applicability of this theory to the agency's regulatory cutoff level of 50 ppmi ...In Alabama Power, this court found that an agency has tbe power, "inherent in most statuatory schemes, to overlook circumstances that in context may fairly be considered de minimis." Id. at 360. That power "is not ability to depart from the statute, but rather a tool to be used in implementing the legislative design." id. at 360. As the Alabama Power court esphasised, de minimis authority may be available "when the burdens of regulatory yield a gain of trivial or no value." Id, It is not HONS 013556 4 sufficient that tha agency nay believe that the costs outweigh the benefits, for Congress has already made the judgement that the benefits of regulation are sufficient. 636 F. 2d at 1283. As further explained by the Alabama Power v Costle Court, IT]he "de minimis* doctrine that was developed to prevent trivial items for draing the time of the courts has room for sould application to adminsitratlon by the Government of its regulatory programs.... The ability ... to exempt de minimis situtations from a statuatory command is not an ability to depart from the statute, but rather a tool to be used in iaplementlng the legislative design. Determination of when matters are truly de minimis naturally will turn on the assessment of particular circumstances, snd the sgency will besr the burden of Mklng the required showing. But we think most regulatory statutes...permit such agency showing In appropriate cases. 636 F. 2d at 360 The court in E.D.F. v E.P.A. did not permit the Agency to avail itself of the de minimis exception because the record was devoid of sny indication that the Administrator relied on this principle. . As indicated by the results of CHA's survey, amounts of incidentally generated pcbs below the 50 ppm regulatory cutoff level are clearly de minimis amounts. Because Section 6(e) does not address Incidental generation specifically, and because of the de minimis standard to closed -manufacturing processes and to maintain a cutoff level of SO ppm, at a minimum. See. Volkswagen v. F.H.C., 390 O.S. 261, 276-277 (1968); Marine space Enclosures, xnc. v. F.H.C., 420 F. 2d 577 (D.C. Cir. 1>69)| District of Columbis v. Orleans, 406 F. 2d 957 (D.C. Cir. 1368. MOWS 013557 { ' K EPA MUST INTERPRET SECTIOH 6(e) TOGETHER WITH OTHER PKOVISIOKS OF THE ACT The U.S. Court of Appeals for the D.C. Circuit ha* ruled that an agency must, when undertaking a regulatory action or investigation pursuant to an Act of Congress, construe that statute In the context of the entire regulatory scheme. An agency Must interpret each provision of a statute In light of the entire statute, rather than looking only to one specific provision as the basis for its action, r.T.C. v Manager. Retail Credit Company, Miami Branch Office, 515 r.2d 988, (D.c. Cir. 1975)) See also, Weinberger v Hynson, Westcott and Punning, Inc.. 412 D.S. 609, (1973). In Section 6(e) of TSCA, Congress designated on* specific chemical, PCBs, as the target for regulatory action. In devising any regulatory scheme with regard to PCBS, however, EPA must also conply with the intent of Congress as expressed in Section 2(c) of the Act which provides) (c) Intent of Congress. - It is the Intent of Congress that the Administrator shall carry out this Act in a reasonable and prudent manner, and that the Administrator shall consider the environmental, economic, and social Impact of any action the Administrator takes or proposes to take under this Act. As the Court in EPA v EDf noted this section of the Act anticipates that the Agency will consider a range of benefits and costs of proposed regulatlons._/ HONS 013558 the Information in CMA's aubalasion addresses all of the factors involved in Section 2(c), and aore. BM's Assessment of the Benefits of Regulating Incidentially Generated PCBat The Health Effects of FCBai Elsies To Be Addressed In Proposed Regulations Accurate estixwtes of the benefits to be derived from proposed regulations are essential. Such evaluations are especially necessary for environmental health and safety regulations in general, which confront agencies with a potentially intimidating array of scientific and technical evidence. Perhaps the most important task facing BPA in this entire proceeding is the need to adequately address the information on the health risks posed by PCBs, and whether any risks posed are unreasonable. These analyses will permit the Agency to identify any benefits which will be achelved by regulating incidentially generated PCBs. In conducting this analysis BPA of course, will want to analyse the record in this proceeding in light of the unreasonable risk criteria set forth in Section 6(c)(1) of the Acti (A) the effects of such substance or mixture on health and the magnitude of the exposure of human beings to such substance or mixture, (B) the effects of aueh substance or mixture on the environawnt and the magnitude of the exposure of the environawnt to such substance or mixture. HONS 013559 i (C) the benefit* of such substance or mixture for various uses and the availability of substitutes for such uses, and (D) the reasonably ascertainable econoailc consequences of the rule, after consideration of the effect on the national economy, small business, technological innovation, the environment, and public health. 1* One of the pivotal aspects of the litigation over EPA's PCS regulations was BPA's finding that no level of exposure to PCM can be considered safe._/ While the Court In EDP v EPA resisted EDP's attempt to equate a risk of exposure to an unreasonable risk, CPA's policy position combined with its failure to present an assesmaent of the risks Involved was clearly a fatal flaw in the Agency's defense of its regulations. In view of the Congressional, government and public misunderstanding of the risks which PCBa pose we urge EPA to give careful consideration to utilising scientific expertise outside the agency to supplement its staff scientists in obtaining the best judgments available on the health effects of PCBs. The National acadesw of Sciences, the General Accounting Office and other groups have stressed the value of independent review in ensuring that agencies draw accurate conclusions from scientific data.5 A number of agencies are making increased use of outside scientific review. EPA, itself, of course, plans to make Increase the use of the Science Advisory Board.59 We applaud these efforts and urge EPA, at an early date, to establish procedures which will provide the Agency, on an HONS 0135* expeditious basis, with ths scientific judgments it needs to conduct a thorough analysis of the health risks of PCBs. 5*See, a.g.. General Accounting Office, Xmprovina the gclentlflc and Technical Information AvsTIsbleto the Environmental ProtectlonAgen^ In its Decisionmaking Process '26 (September Academy ot sciences, MClslonmaklna In the Environmental Protection Agency 46-51 t77). Clean Air Scientific Advisory Committee, BetEing Ambient Air Quality standard, SEImnlTpviriroovn*inmgentThtae.l IPPrrooc.tees*cstio1n"'Ag(SeenpcytemRebseerarlcfhllAJliutChoonrifsearetinocne Report, for foproprlafclons for riacal tear 1978. H.R.. hep. No 722, 95th Cong., 1-s- t -B--e--s--s-.---16- -1-7 (1977) ( --Ser-i-a--t-e- --C- -o-m---m--, on EErnvironment and Public *"*' Environmental Research and Development Authorisation Act of 1*77, S. Asp. No. 188, 95th Cong., 1st Sess. 1-4 (19771.------------------ *See Memorandum from Deputy Administrator John Hernandes, proving the Scientific Adequacy of Agency Regulations and *28*,*"1*98**1).' *0, 1981), reported in -I-n--s--i-d---e----E--R--A- 11-12 (Auauat HONS 013561 ..... tam SPA'S OBLIGATION TO CPUSIDEB THE erncrs or other regulatory controls or pcbs Certain provisions of TSCA require that the Administrator of SPA consider the totality of the requlatory eontrola of chemical aubatancea. It was Conqreaa' qeneral purpose to avoid overlapping and duplicative regulations._/ ftor instance. Section 9(b) of TSCA specifically requires the Adainistrator to coordinate Its actions under TSCA with actions taken under other laws administered by the Agency. Legislative history of this section of the Act Bakes clear that the Adainistrator, in determining to utilise TSCA is to publish s review of other authorities in the Agency's jurisdiction and publish that review at the tiae the Adainistrator takes actions under TSCA.(cite) Furthermore, Section 9(d) of TSCA requires the Administrator of SPA to "consult and coordinate" with other agencies for the purpose of acahleving maximum enforcement of TSCA while imposing the least burdens of duplicative requirements on persons subject to the Act, and for other purpoaes. The following is a review of other regulatory authorities relevant to the control of FCBs. HONS 0135*2 3> 9 In its PCBs ban regulation* issued on May 31, 1979, CPA noted that it eould integrate its PCB rules with the regulations issued pursuant to the Resource Conservation and Recovery Act of 197C, 42 C.S.C. 3251 at sag. (RCRA). (See, 44 Fed. Reg. at 31539.) The saae notation was Bade in CPA's RCRA Regulations. (See, 45 Fed. Reg. at 33118.) To date, the agency has not so integrated these regulations. In the RCRA regulations, PCBs are Hated as *Hasardous Constituents* (See, Appendix vm to 49 CFR fart 261, 45 Fed. Reg. 33084, 33133, May 19, 1980), but are not regulated per se under the RCRA regulations. Hazardous constituents are those chemicals which caused the CPA to list a waste as a CP toxic waste or toxic waste in Sections 261.31 and 261.32, as stated in 49 CFR Section 261.30(b), 45 Fed. Reg. at 33122. One court has ruled that PCBs are not currently regulated under RCRA, U.S. v Burns. 15 B.R.C. 2120 (W.D. Pa. 1981). The Burns court specifically deterained (l) that the federal govenwwnt cannot bring an action concerning PCBs under RCRA, since PCBs are regulated under TSCA, and (2) that the federal govsrnswnt cannot recover its clean-up costs under TSCA, but aay recover such costs under Section 311 of the Clean Hater Act. However, CMA believes that as a practical effect, RCRA does regulate sany tfMh haxardous waste mixtures which contain incidentally generated PCBs in concentrations at less than 50 PPm.1 The CMA survey indicates, for exaaple, that the majority of Incidentally generated PCB* are associated with process involving HONS 013563 I 1*3 the chlorination of hydrocarbon* or th use of chlorinated solvents. Kt least 17 wastes are listed as 'hazardous* under 40 CFR 261.31 and 261.32 can reasonably be expected to contain toea level of incidentally generated PCBs because they are generated froa the types of processes noted above.2 Finally, the Cm survey demonstrates that over sixty percent (60) of these wastes are incinerated. . 2. PW> nireetlv neeulslts PeBr Bmlerlthe Clean Hater Act Polychlorinated biphenyls are directly regulated in two ways under the Clean Mater Act, 33 u.S.C. 466 et seq.t (1) point source discharges from manufacturers are regulated by effluent standards pursuant to Section 402; and (2) discharges from any facility are regulated under Sections 311 and SOI. Other materials which my be reasonably expected to have low-level FCB contamination, e.g., dredge spoila, are directly regulated under the Act thereby, in effect, regulating the introduction of PCBs into the environment. The effluent standards of 5307 of the Act currently reguire that no detectable PCbs be *in any discharge from any PCB manufacturer.* - 40 CFR 129.105. The standard is directly enforceable against any mnufacturer of PCBs through the provisions of Paragraph 309(a)(3) of the Act. Further, the standard sets a water quality criteria of 0.001 mlcrogram/liter for PCBs in navigable waters (see, .40 CFR S129.10S). when ' HONS 013564 A implemented by the states, these requirements regulate FCBs which may be discharged at very low concentration levels. Section 311 of the Act is general in its applicablility snd regulates the discharge of FCBs into navigable water from any source. Regulations promulgated under this section provide for the designation of FCBs as a hazardous substance, the establishment of the maximum quantity of ten (10) pounds which may be discharged without cosplying with the reporting raqulrssMnts or the becoming subject to penalties for failure to report such discharges, (see, 40 CFR 116 and 117). *9- JEA_turUimLJQmm4uLiis-FeB under Provisions St TSCA >3 ` other--this Beaties 6(a-). BPA should recognize that TSCA regulations already place established adequate control of FCBs generated as a process waste. Detailed recordkeeping and monitoring activities are currently required under the PCB Ban Rule, 40. C.F.R. 761.45; Red. Reg. at 31S27. The regulation applies to process wastes or contaminated primary chemicals containing FCBs in concentrations of 50 ppm or greater. In general, the current requirements included the maintenance of inventories of FCBs and PCB equipment as well as the monitoring of incineration process and chemical waste landfill facilities. In addition, whenever an uncontrolled release of FCBs or a PCB mixture into the environment poses a substantial risk to human health or to the requirement. HONS 013565 i i $ notification of (PA would ba required within the parameter* of 8ectlon 8(e) of T8CA 15 U.S.C. 2607 (3). (See, 43 Fed. Reg. 11110.1111-11112 (March 16, 1978), CPA Statement of Interpretation and Enforcement Policy.) c) The United Statee Department of Agriculture hemCemewer*fced ite Authority to Ountrol Incidental PCT Qontasdamfclpn In Food Eatahliafamsntm. The United STatee Department of Agriculture (U.S.D.A.) under lte proceedng entitled Prohibition of PCB-Contalnlnq Equipment In Machinery, and Liquid PCBa In Federally Inspected Meat Establishments, Poultry Product Establishment* and Ego Product Plants proposed an amendsMnt to the Federal meat Inspection regulations, and the Federal poultry products inspection regulations governing the inspection of eggs and egg products prohibiting the use of FCB-contalning electrical equipaMnt in concentrations exceeding 50 ppm PCS. 45 Fed. Reg. 30980 (May 9, 1980). The purpose of these proposed amendments were to reduce the potential of accidental PCB contamination of food products. Id. However, in response to the court's decision in E.D.P. v E.P.A., supra, the U.8.D.A. issued a notice of abeyance of the ruled proposal in 46 Fed. Reg. 15512 (March 6, 1981). The Department noted *the desirability of further FSQS [Food Safety and Quality Service) consideration of relevant technical data* and *(l]n the meantime,...encourage the submission of additional eoanaents by anyone who has information that would assist U.S.D.A in making its decision.* 46 Fed. Reg. 15512. HONS 013566 w .1If ^ it ,, 'tn+**'a** The Food and Drug Administration Iwi Peuwna Luted if----Authevlty to Central Incidental FOP Ouii Lanina LlUII In The Pood and Drug Administration (PDA) has pravioualy aetad to control accidental FCB contamination of food, faods and food packaging materials. 38 Pad. Rag. 18096 (July 6, 1973). in addition, the PDA has proposed further amendments to its earlier KB regulation in its proceeding entitled Current Pood Manufacturing Practice Relating to Poisons and Deleterious Substances in Pood, Peed, and Pood Packaging Material Plants. Docket No. B0N-0128, 45 Ped. Reg. 30984 (May 9, 1980). However, this proposal is also being held in abeyance pending the outcome of this Instance regulatory action, 46 Ped. Reg. 15518 (March 6, 1981). In PDA Docket No. 77N-0080, entitled Polychlorinated Biphenyls (PCBs in Pish and Bhellflshi Reduction of Tolerances (Notice of Proposed Rulemaking, 42 Ped. Reg. 17487 (April 1, 1977), a final rule was issued reducing toleranace for unavoidable residues of the KBs in certain foods (44 Ped. 57389 (October 5, 1979), and stayed its rule regarding fish and shallflsh. On May 1, 1981, 46 Ped. Reg. 24551, FDA announced that it would hold a formal avidentiary rulemaking hearing on the tolerance of PCBe in flah and shellfiah to admit into the record HONS 013567 7 toxicity data and information previously unavailable on the iesuet ia the magnitude of the human food loss (in terms of dollars, poundage, percentage of catch, etc.) that would result from lowering the tolerance of PCbs in fish from 5 ppm to 2 ppm? Bearing are now in progress before an Administrative Law Judge. a) The Occupational Safety and Health Administration --m--f Hamper Exposure Limits for pcbs. EPA may be permitted, under Section 9(c) of TSCA, to prescribe and enforce standards or regulations affecting occupational safety and health, so long as EPA coordinates its regulatory program with Occupational Safety and Bealth Administration (OSBA), aa required by Section 9(d) of TSCA. Existing OSHA regulation (29 C.r.R. Part 1910.1000) establish stringent exposure limits for PCBs upon degree of chlorination, and assign a lower exposure limit [expressed as an 8-hr. time-weighted averate (TWA) to the more hightly chlorinated category, as followsi 8 hour TWA Chlorodiphenyl, 42% chlorine (skin) 1 mg/M3 Chlorodiphenyl, S4t Chlorine (skin) 0.5 mg/M3 HONS 013568 The akin notation indicates that skin contact is the aost probable route of entry into the body, but provision is also aade for entry via inhalation. The OSHA regulations provide that if respiratory protection is deeaed necessary, respirators aust be selected and used in compliance with provision of 29 C.F.R. 1910.134. EPA should recognise the degree of protection that the cited 06HA regulations provide to-chealcal plant personnel, especially in the matters of incidental generation of PCBs in various chealeal processes, and refrain froa promulgating additional, unneeded regulations in this area. 1 SetaraJ- federal AnenciafcJtrn.iMeldtee Their Pseeeedinas (in Abeyancefading Cellseeieiu Meed Beienfcttie MeideaeeL The U.8.D.A., P.D.A., and, as will be explained later, the BIA, are currently holding their FCB-related proceedings in abeyance for the specific purposes of (1) collecting additional scientific data for support their respective regulations, and (2) allowing SPA to gather data pursuant to its advanced notices of proposed rulemakings in the 'regulatory cutoff,* `closed manufacturing processes* and `totally enclosed* PCB proceeding. These actions ara in keeping with the Administration's policy to Mploy good scientific data to substantiate regulatory action. HONS 013569 In conclusion, any regulation of Incidentally regulated FCBs should be baaed upon a thorough assessment of the extent to which existing federal regulations already address health and environmental protections against Incidentally generated PCBa. EPA Should Review Any Proposed Regulation Of Incidentally Generated PCBs In Accordance With executive Order 12291 On February 17, 1981, President Ronald Reagan lesu< xecutlve Order Mo. 12291, entitled *rederal Regulation.- Ithe Stated paspasee, Increased agency accountability for regulatory aations, minimizing duplicative and conflicting regulations, and Insuring well-reasoned regulations, the President Issued the following mandate) Sec. 2. General Requirements. In promulgating new regulations, reviewing existing regulations, snd developing legislative proposals concerning regulation, all agencies, to the extent permitted by lew, shall adhere to the following requireswntsi (a) Adsinlsttrative decisions shall be based on adequate information concerning the need for an consequences of proposed governamnt action) (b) Regulatory action shall not be undertaken unless the potential benefits tor society from the regulation HONS 013570 outweigh the potential coats to society; (c) Regulatory objectives sahll be chosen to Malaise the net benefits to society; 'i (d) Hong alternative approaches to any given regulatory objective, the alternative involving the least net cost to society shall be chosen; and (e) Agencies shall set regulatory priorities with the ala of Mxiaising the aggregate net benefits to society, taking into account the condition of the particular industries affected by regulations, the condition of the national econoay, and other regulatory actions conte^lated for the future. Sec. 3. Regulatory lapset Analysis an Review (a) In order to lapleaent Section 2 of this Order, each agency shall, in connection with every aajor rule, prepare, and to the extent permitted by law consider, a Regulatory Iapact Analysis. Such Analyses My be combined with any.Regulatory Flexibility Analyses performed under 5 U.S.C. 603 and 604. Although Section 8(a)(2) of the Executive Order (EO) exeapts from the requirement of Section 3 (a-) the preparation of 013571 RONS n i* Regulatory inpact Analysis in any regulations subject to deadlines laposed by a judicial order, such as the regulations at issue, the President required that any such regulation shall be reported to the Director of the Office of Hanageaent and Budget together with a brief explanation of the conflict, the agency shall publish in the federal Register a stateaent of tha reasons why it si impracticable for the agency to follow the procedures of this Order with respect to such a rule, and the agency, in consultation with the Director, shall adhere to the requlreaents of this Order to the extent permitted by...judicial deadlines (section 8(a)(2). CMA does not believe that it will be lapractlcable for the agency to prepare a Regulatory lapact Analysis as part of its decisionmaking procass in this natter. The Agency has data relating tot 1. The benefits and costs of regulating incidentally generated PCBs. 2. Alternative approaches to consider, e.g., the regulatory cutoff, the exemptions procedure alone. 3. Information on the benefits of regulatory PCBs based upon the latest Intonation on the health affects of PCBs and the economic consequences of regulation. HONS 013572 Again, CMA strongly urges SPA to solicit the views of known experts outside the Agency in the fields of toxicology and analytical chemistry to add to the views of its staff scientists or questions relating to the health effects of PCBs and currently available analytical capability for the identification and Masurament of low concentration PCBs in coaplex streams. These views can be obtained expeditiously and perhaps at minimal costa by tapping resources in the academic and research communities. CONCLUSIONS ( Baaed upon the lnforaation and legal analysis being submitted to the Agency, CMA concludes: 1. Congress did not intend to ban incidentally generated PCBs under Section 6(e) of TSCA. Such FCBs were not even known to Congress when it passed the Act. Congress' assuaption concerning the econoaic iapact of its actions and concerning the few hardships which the provisions of Secton 6(e) aight iapose siaply do not address the nature, scope and prevalence of incidentally generated PCBs. fjWHJrtna ffifC nfiBaaiaew f"Hof 3 A There is no basis for concluding that such regulations should apply to aaterials containing less than 50 ppa of PCBs; evidence may exist to deaonstrate that a higher level would be appropriate; and. HONS 013574 2 b. Substantial evidence exists to support an additional exclusion from regulation of chemical processes generating any concentration of PCBa wherei o the PCB concentration in its product does not exceed 50 ppm [or such higher level as EPA may establish by rule]; 4* tr^^^iUrnintil T| nrd~tng *Vi fffi's PCB regulation; and, o al^| discharges comply with any applicable BPA or OSHA regulations. n HONS 013575 'D'E.AfC T Appendix Perspective of FCB Exposure Associated with Primsrv Chemicals Containing Incidental Quantities of PCBa The higher homologs of PCBa, trichlorobiphanyl to dacachlorobiphany1, which CMA believes ara tha compounds of concern for potential exposure, are high molecular weight (MW) compounds with low vapor pressures (VP). See Table I. This fact and the low concentrations of incidentally generated PCBa dictate that PCBs cannot be present in the mwohr at other than very low concentrations when the exposure is due to chemicals with incidental PCBs. The calculations below demonstrate this and indicates why CMA believes that control of the primary chemical inherently controls potential exposure to incidentally generated PCB contaminants. ' The calculations are based-on Raoult's Law, a basic principle in the properties of solutions. Raoult' s Law states that the vapor pressure of a component of a solution (in this case, PCBs are a component of the solution consisting of the primary chemical and the PCBs) is equal to the vapor pressure of that component in its pure form multiplied by the mole fraction (i.e., concentration) of that component in the solution. Mathematically, ^Afsol) - VPA x MP Where, V?A(soi) m Vapor pressure of component A in atmosphere above solution (sol). VPA m vapor pressure of component A in its pure form. MP * Mole fraction of component A in the solution. Once this vapor pressure of PCBs above the solution has been caleulated, the atmospheric concentration of PCBa can be calculated using the partial pressure law, a basic principle in the properties of gases, which states that tha volumeric concentration of a gaseous * Vapor pressure can be viewed as the tendency of a liquid or solid to escape into the gas phase. A low vapor pressure indicates a low, tendency to vaporise. '` HONS 01357b . * component 1* directly proportional to the ratio of that components partial praasura to tha total pressure of the system. In thia casa, we will be dealing with tha partial pressure W^a(soI)* of PCBs and atmospheric preaaure. Daing che above principle*, calculationa can be made to deawastrate the lavala of potential PCS expoaura in the work place. Tor example, if the PCB concentration in the primary chemical ia 1000 ppm and the primary chemical haa a molecular weight of 1/2 that of the PCB (the caae if the primary chemical ia trichlorobenzene , and the PCB ia hexachlorobiphenyl) , then the mole -fraction ia calculatedi *- - MF - Hole* PCB Molea total aolution 1000(parts PCB) HP 361.o7mW PCS) ,999,OOP (parta trichlorobenzene) + 1000 (parta PCB) 191.5 (MW trichlorobenzene) + 361.0 (MW PCB) f - 5.03 x 10 Bepeeting theae calculationa for varioaa aaawptlona of com- . centrationa and molecular weight ratioa allow* ua to conatruct Graph I, which illuatratee the rang* of mol* fraction* for the moat common aituatlon -- PCB. concentration below 1000 ppm and PCBt primary chemical molecular weiaht ratio Mm> between ltl and 2il. As can be 'calculated, the mole fraction variea between 5x10-7 (,t lppm) and 1 x 10~3 (at lOOOppm). Theae are ob viously email numbers. Applying Bacult's and noting tha low vapor pressures (Table I) of PCBs, it is easily seen that the vapor pressures of PCBs I^Afsol)^ *r* extremely low sine* w* ere multiplying two small nuzbers. . To illustrate the practical effect of these calculations on atmospheric concentrations of PCBs, we give the example below of 50 ppm of trichlorobiphenyl (the moat volatile of the PCBs which CXA believes should be considered in the rulemaking) ia monochlorobenaene, a typical primary chemical with which PCBs may be associated. - HONS 013577 The vapor prasaura of trieblorobiphanyl (TCB) is^bout O.OOX m at 2S*C. Tharafora, assuming SO ppm of TCB in chlorobenzene: (1) MF - __________ 50/257.5 ____ " 2.2 X 10-5 999,950/112.3 + 50/257.5 (2) VP at equilibrium 0.001 nra x 2.2 x 10"5 - 2.2 x 10-8 and tha atmospheric concantration at aquilibrium and 25*C ia: 2.2 x 10~8 mm x 257.5 g/molo x 1000 l/3 x 1000 mq/q 3,1 x 10_* mg/m3 760 Bn 24.45 1/mola ^ ^ Tharafora, tha maximum PCB vapor inhalation axpoaura (25*0 that can occur from axpoaura to chlorobenzene containing 50 ppm of trichlorobiphanyl ia 0.034 of tha OSHA TLV (1.0 mg/m3) for lower chlorinated biphanyla. . . This is a specific example. Tha ranee of molecular weight ratios between.ehamicala involved in thasa processea ia generally within tha range of . Ill to 2il'. Table IX ahowa the resulta of assuming various concentrations of trichlorobiphanyl in tha primary chemical for thasa assumed ratios. ' ~ Table XX also shows tha calculated atmospheric concentrations as a percentage of tha OSHA TLV for lower chlorinated biphenyls. Conversely, control of the primary chemical to meet OSHA limits for that chamieal will result la control of potential exposure to FCBs. Continuing our example above, the OSHA TLV for monochloro benzene is 350 mg/m3. The vapor pressure f 25*C of monochlorobenzene is 12.1 emHg. We calculated above that the partial pressure of FCBs above e solution of 50 ppm TCB in monochlorobenzene is 2.2 x 10*8 amBg. Tharsfore, the amle fraction of PCB in the vapor is: _ HP - 2.2 x 10~8 mi 7 in-9 V |- A Therefore, using the sane principles above, the equivalent concentration of PCBa in the vapor if monochlorobenzene is controlled below the OSHA TLV is calculated:, HONS 013578 Concentration - 350 me MCB x 1.7 x 10-9 x 257.5 (xw TCB) --HT3-------- TT5.1 )mw Mtt! 1 Concentration - 1.4 x lo"S Mg TCB n ' Thua, controlling monochlorobanzana at tha OSHA limit raaulta in a PCI ataoapharic concantration that is 0.0002* of tha OSHA Unit for PCBs and, avan morn importantly, is so low as to b not BMasurabla. - Thus, tha potantial axposurs to PCIs raprasants far lass risk than axposura to tha primary chanical of which tha PCB is a contaminant. Tha sama principla would apply to skin contact whara axposura to tha primary chamical is 1000 tol.OOO.OOOtimas greatar (1000 to 1 ppm PCBs) than tha axposura to PCBs. Tha abova calculations shows - (1) Vapor axposura . consonants of a solution is .dapandant on concantration. (2) Low concentrations must result in low axposura. (3) tf volatility (vapor prassura) of tha componant is -' low, and tha concantration is low, tha axposura must ba vary low. (4) SO ppm raprasants a waight ratio of 1/20,000. If tha MW of tba ccmponants is equal, tha mol* fraction At 50 ppm is 1/20,000. If tha PCB MW is twice as high as tha primary chamical, tha mola fraction is 1/40,000. (5) Vapor prassuras for pura PCBs ara' vary low and whan multipllad by 1/20,000 to 1/40,000 waight ratios shown abova, tha vapor is not a significant axposura problem. MOWS 013579 -mi Graph I Mole Fraction* of PCS in Solution a* a Function of PCB concentration ' w/ molecular weight ratio (PCBs Primary Chmaical) a* a Parameter HONS 013580