Document o9KpJzEE4agL4BwoO0kex5zw3

BEFORE THE UNITED STATES ENVIRONMENTAL PROTECTION AGENCY ++ + (tearing for tha Closad and Controllad vtaata Proeass Rula ++ + Rood 3906 Mataraida Mall 401 M Straat, S.H. Washington, D.C. Monday, July 26, 1982 Tha maating convanad at 9t07 a.n., Richard J. Guimond presiding. EPA PANEL Richard J. Guimond Alan Carpian John Smith William Guntar David Radford Laura Ctmpball Daniaa Kaahnar Amy Moll For Dittribtirien by CMA SPECIAL PROGRAMS DIVISION From f.T. ~fLt Ref. No. TC.*-/7~r Dte___ 7/A EXECimvi COURT REPORTERS (XI) M14M4 ^ MONS 216120 t 2 3 4 STATEMENT OTt Dr. Kannath Burgoo*, Chairman, Chamical Manufactures Assn.- 0 Dr. Robart Kalay, mambar CMA, 7 PCB Analytical Task Group 0 Dr. Garaldina Cox, CMA 9 Jaegualina Warran, NRDC 10 Elian Silbargald, NRDC It Richard L. Bluma, Ganaral Elactrie, Division Counsal 12 Dr. Danial E. Sliva, Ganaral Elactrie, 13 Ganaral Managar, Silicona 14 t 4 Av 17 10 10 20 21 PAGE NO. ~li} 29 - s 112 H9 144 144 2 24 20 EXECUTIVE COURT REPORTERS (i) HONS 216121 3 1 L2eedings MR. GUIMOND: Welcome to our hearing this morning. My nan* la Richard Guimond. I'm Chief of the Chemical Regulation Branch in the Office of Toxic Substances, i'll be the chairman of this hearing panel. The other members of the panel and myself would like to give you a welcome to the hearing. As you know, epa proposed a rule to control the release of FCBs enclosed in control waste processes. This action was taken in response to e decision from the U.S. Court of Appeals for the District of Columbia Circuit. Our goal for the hearing is to learn sore about PCBs manufactured in low concentrations, the impact that the proposed exclusion from the statutory ban may have, learn about the analytical methods for monitoring PCBs and processes affected by the rule, learn any problems associated with this. Asnng the issues in which the agency has requested comment are the appropriateness of the exclusion, the need X to specify criteria for determining the absence of PCBs, the appropriateness of using limits of quantification versus 21 limits of detection, and the suitability of allowing 22 manufacturers to use best theoretical analysis in lieu 23 M of actual monitoring. As most of you are aware, the Federal Register 3# EXECUTIVE COURT REPORTERS (m) w mi MOMS 216122 4 Notice of June 8th included misprint concerning the dates for the closing of the mein comment period end for this hearing. This misprint resulted in en error hy the steff of the Government Printing Office thet wes unfortunetely out of our control. ._ He apologize for any inconveniences that this may have caused end you will now have until August the 9th for the close of reply comments. As with our actions to the extent feasible, the agency will consider other coaments provided to it beyond that date within this constraints of our overall rulemaking. If, during the course of the hearing, any of you have any questions that you would like the hearing panel to ask the people that are testifying, I'd request that you write them down on a card or a sheet of paper and give them to the hearing clerk that's located up here in the front and she will pass them to the hearing panel to ask. In addition, if there ars any other people that are not on the agenda for the hearing today that believe they have information that is relevant to this rulemaking and would like to present it, we will consider that to the extent feasible within our time constraints of the hearing today. So if anyone does want to provide some additional information, I request that you see me sometime during the executive court reporters (m) m mi MONS 216123 dp s I day and I will talk with you and aaa if we can maka a appropriate arrangement!. 3 There are agendas for the hearing located on the 4. table just as you come in the door -- excuse me. it'a S outside the door. . At this time. I'd like to introduce the remainder 7 of the hearing panel for this informal hearing, way down at the end on my left is Amy Moll. Amy is one of the economists with the Economics and Technology Division 10 in the Office of Toxic Substances. 11 Next to her is Laura Campbell. Laura is with the 13 Office of Pesticides and Toxic Substances Enforcement and 13 has been working with us with respect to enforcement issues 14 on the rule. 11 Next to Laura is Bill Gunter. Bill is the team 1$ leader for the PCB Regulations Teem in the Office of Toxic 17 Substances. It To my ianediate left is Alan Carpien. Alan is It the attorney with the Office of General Counsel who has served as a legal advisor on these PCB rulemaking. To my right is John Smith with the Exposure Evaluation Division of the Office of Toxic Substances. John has bean the principal chsmiat in tha development of the guidance on the methods for sample collection. To his right ia Dave Bedford, also of the Exposure EXECUTIVE COURT REPORTERS (JU) MMMI HONS 216124 6 Evaluation Division. D*vs is s project manager of the Hidwsst Research Institute contract that is assisting us in tha development of analytical methodology for th rulemaking. to my far right is Danisa Ktahnar who is biologist on tha PCB Regulations Team and is tha laad staff manbar for this rulemaking. At this time, I'll turn tha microphona ovar to Alan Carpian who has sou remarks regarding tha procadural aspacts of tha rulamaking in this haaring. NR. CARPXENi I'd lika to straae. first of all. that this is an informal haazing and we're going to try to remain as flexible as possible as a haaring panal. However, aftar saying that. I also want to point out that wa hava procadural rulas undar Volume 40 of tha Coda of Federal Regulations, Part 750, Subpart A, and thara ara copies of thosa regulations on tha table as you cone in. 1 will be available for any questions about any of tha procedures involved in thasa rulas during any breaks. I believe Rich said that reply comments will be duo two weeks from tha day the haaring ands. Undar our current sahadula, wa expect tha haaring to and today and that would aeon that reply coaswnts would ba due by August 9th. Let me comment briefly on how we are going to EXECUTIVE COURT REPORTERS (Ml) MMM* NONS 216125 7 proceed. We ask all participants to first identify themselves and their affiliation to tha court reporter before they begin speaking. This is in order to help us later as we review the transcript. Any additional material you hav,~any additional written material or any other material that the panel might request from you should be given to the hearing clerk for insertion in the record. Although participants are not sworn, I would like to remind everyone they are subject to 18 USC 10001 which is the False Reports to the Government. Following each speaker's presentation, each member of the panel will have an opportunity to ask questions and as Rich said before, we will ask any reasonable questions submitted by the audience. If anybody on the panel should ask a question or if any speaker has anything to say that is of a confidential business nature in the speaker's opinion, please note that. You can later present your response in a close session if that's necessary, but present the information in writing on a confidential basis under agency rules for the submission of confidential material, the agency regulations for submission of confidential material will found in Part 2 of Volume 40 of the Code of Federal EXECUTIVE COURT REPORTERS (Ml) NI4NI HONS 21612b 8 Regulation*. That's all I hava. MR. GUIMOND: Than I guess wa'll begin with CMA? STATEMENT OF GERALDINE COX, CMA MS. COX: My name is Geraldine Cox. I an Vice President and Technical -- Director of the Chemical Manufacturers Association. The Chemical Manufacturers Association is a trad* association representing the manufacturers of more than 90 percent of the basic industrial chemicals in the United States. CMA and its special programs panel on PCBs welcome the opportunity to discuss EPA's proposed rule for closed and controlled waste manufacturing processes with you this morning. CMA shares EPA's concern that PCBs do not pose unreasonable risk to man or to the environment. As you know, in this country, intentional manufacture of PCBs ceased more than five years ago and w* do not advocate such manufacture be reinstated. The chemical industry has worked diligently to develop substitutes for these uses for which PCBs were invaluable in the past. CMA further agrees with EPA that if the current uses of praviously intentionally generated PC -- I'm sorry, if there are currant uses of previously intentionally generated PCBs that pose unreasonable risk, adequate controls should be developed and if there are any specific circumstances where unreasonable risks are presented by EXECUTIVE COURT REPORTERS (Mt) M14HI HONS 216127 9 1 mall quantity of unintentionally generally PCBs found as 2 impurities in e wide variety of chemical!, controls to 3 reduce human and environmental exposures would be 4. appropriate. S However, CMA cannot support EPA's present proposal for what it terms "closed and controlled waste manufacturing 7 processes." That proposal is premised on ths scientifically 9 and leqally flawed assumption that any exposure to PCB is 9 significant. 10 Not only is this premise wrong, as a matter of 11 science for any chemical, it also mischaraoterises what we 12 know about the toxicity of PCBs. Contrary to what might 13 have been believed by soma in the past, including Congress, 14 PCBs are not uniquely toxic. 10 As the recent epidemiology studies of the - incident and the heavily exposed capacitor worker have II demonstrated significant health risks will not be posed at 17 II the low levels from inadvertent generation that EPA is now 11 considering regulating. Once EPA reviews the health evidence in light of 90 the exposure levels at issue, we are confident that the 21 agency will agree with us that inadvertent generation poses 22 22 no significant risks to sun or the environment. CMA's special program panel has proposed a reasonable regulatory 24 21 program for inadvertent generation of PCBs, although we do EXECUTIVE COURT REPORTERS (M) IU4HI MOWS 216128 not believe that likely -- believe tha likely expoeure* 10 that would occur if PCBs were below 50 parte per million in product end waste streams would pose any significant risks. we would support a regulatory seheir-j -baaed on such a quantified cutoff. By employing this cutoff, EPA could assure the public that no reasonable risks win be posed and industry, that it will bs allowed to continue to produce the many valuable chemicals under reasonable regulations. Here today to spell out in detail CMA's positions are Dr. Kenneth Burgess, Chairman of our Special Program Panel on PCBs, and Dr. Robert Kaley of the Panel's Analytical Task Force. I encourage you to listen closely to their presentations and to ask them questions in order to assure that EPA clearly understands CMA's position or. the issues raised by its closed and controlled proposal. Dr. Burgess? STATEMENT OF DR. KENNETH BURGESS, CMA DR. BURGESS Thank you. Good morning. My name is Kenneth Burgess and 1 am chairman of the Chemical Manufacturers Association Special Program Panel on PCBs. With me today is Dr. Robert Kaley of the Program Panel's Analytical Task Force who will also be presenting testimony and at the conclusion of Dr. Kaley's testimony, I will be EXECUTIVE COURT REPORTERS (wt) >*y*tn MQNS 216129 li glad to answer your questions and refer them to Dr. Kaley, as appropriate. Also with us today are Tin Hardy and John Jackrison of ths firm of Kirkland Ellis, counssl to the PCB Special Panel. ___ CMA has repeatedly urged that tPA conduct a thorough review of potential exposures to substances containing inadvertently generated PCBs and the relative consequent health and environmental effects of the PCB content in response to the Court of Appeals remand of EPA's 1979 PCB regulations. With respect to both intentionally manufactured PCBs still in use in electrical equipment and inadvertently generated PCBs in a number of basic chemicals, CMA acknowledges any unreasonable risk as defined by TSCA Section 6(c) should be eliminated pursuant to EPA's authority under the Toxic Substance Control Act. However, EPA to date in both the electrical equipment proposal and its most recent closed and controlled waste proposal, failed to review the exposure data or the health and environmental evidence. Both EPA proposals suffer from the absence of a straightforward description of the record evidence demonstrating that the trace quantities of PCB now entering the environment pose no significant, let alone unreasonable risks. Executive court reporters (jm) ut-m* HONS 216130 12 In particular, that# proposals fail to consider several recently released studies of PCB health effects that dsmonstrate no significant risk will be posed by the trivial exposures the agency is now considering regulating. Such studies include, among others, the Rashimoto,- at al. study, "Role of Polychlorinated Dibenrofurans in Yusho PCB Poisoning" published in Volume 36, Ho. 6 of the Archives of Environmental Health, and Brown and Jones "tortality and Industrial Hygiene Study of Workers Exposed to 1i Chlorinated Biphenyls," also in Volume 36, Ho. 3 of the Archives of Environmental Health, EPA'a failure to address the de minimis risks created by unintentional PCB generation explains in large pert why CMA has urged the agency to withdraw its closed and controlled waste proposal. As our July 8th comments detailed, that proposal is of little value to anyone. It provides no assurance that public health will be protected. It never answers the court's questions in EDP v. SPA of whether any unreasonable risks are posed by ao inadvertent generation of PCBS and it will not provide 21 industry any assurance in inadvertently generated PCBs 2* in closed processes will be excluded from TSCA regulation, 2* even though Any release is trivial. 24 The analytical guidelines are so uncertain that 25 although we doubt any process will ba excluded, this is EXECUTIVE COURT REPORTERS (Ml) MI4M4 HONS 216131 13 1 peculation, -- not fact. In fact, tha proposed rule may 2 create considerable opportunity for supplier-customer controversy as to whether PCBs in customer plants or plant 4 waste were generated by the customer or received as part of ft raw material from the supplier. ft We have thus urged EPA to publish a new proposal 7 that conforms with congressional intent in TSCA and the ft mandate of the Court of Appeals to address Unreasonable risks. That proposal would recognise that no significant 10 risks are posed when PCBs, in products or waste streams, 11 are in concentrations below 50 ppm. 12 Such a proposal would appropriately address the 13 TSCA Section 6(c) criteria, health effects and exposure to 14 humans and the environment, benefit of the regulations and IS cost of regulating substances containing inadvertently 1ft generated PCBs, and determine that inadvertently generated 17 PCBs pose no unreasonable risks. 1ft A 50 ppm cutoff has been the law of the land for 1ft three years, although the exposure and health evidence would 20 support a higher regulatory cutoff, CMA would not object to 21 the continuance of such a regulation. 21 EPA's closed and controlled waste proposal begins 23 with tha appropriate initial premise, namely, that there is 34 no reasonable rationale for any regulatory concern about as levels of PCS insider manufacturing equipment. As long as EXECUTIVE COURT REPORTERS (Ml) W40M MOMS 216132 inadvertently generated PCBe are within chemical plant H equipment, no environmental or human exposure occurs. The concentrations of concern are those to which humans and the environment might be exposed, in air, water, or products. However, the proposal's nonquantillable approach is not a rational approach to the correctly perceived regulatory issue. . This morning, I would like to provide some background information to explain further why CMA urges that EFA withdraw its nonquantifiable proposal. First, and most importantly, EPA'a proposal bears no articulated relationship to protection of the public health as required under TSCA Section , and ae requested by the court in EOF v. EPA. Although EPA justifies its proposal as being ae close to aero release as practically possible, the agency hes not in feet sought to determine that levels -- what levels of PCS its proposal will allow in the various media. Moreover, the agency has made no attempt to determine to what extant any such concentration will lead to human or anvironmantal exposures, not to mention whether tba likely trivial axpoauraa pose any significant risks or outweigh the cost of regulation that must be balance against risk before making an unreasonable risk determination. The proposal cannot withstand ravlaw under the EXECUTIVE COURT REPORTERS (Jtl) MMM4 HOMS 216133 15 1 statutory standard* of TSCA which call for clos* acrutiny of 2 risk and which ware in the forefront of the court's reversal 3 of SFA's '79 PCB regulations. 4 As CMA demonstrated in its comments, current inadvertent generation of PCB does not pose~~unfeasonable risks. The questions asked by the Court f Appeals whether 7 regulation of inadvertent PCBs has other than trivial health and environmental benefits or whether potential risks are unreasonable, can and should be answered in the negative. O CPA's current proposal, however, has not snswered t these questions and because it is not based on a definite 3 limit on PCB concentrations, it is incapable of doing so. 3 Second, the nonquantiflable approach is 4 analytically unsound and administratively impractical. FDA considered use of this approach to implement the "no-residue" requirement of th# Delaney clause for animal reeds, but 7 rejected it as "unreasonable, unmanageable, and making no a sense from the perspective of public health." CMA has for a long time recognised the importance 20 of analytical chemistry issues -- developing reasonable PCS 21 regulatione. Nearly a year ago, we submitted to CPA a report on "Analysis of Chlorinated Biphenyls" and have attempted to assist CPA in developing reasonable analytical approaches to 74 low-level PCB quantification in chemical SMtrices. we were thus quite dismayed by the serious Executive court reporters (Ml) MMM* HONS 216134 _ ' I 16 1 analytical inadequacies of the current proposal. we are 2 alto quit* concarnad about tha continuing failure of PA 3 to make available tha essential guidelines docuaents 4 explaining its analytical proposals. The two docuaents we have received were not made * available to the public until the week after comments on the 7 proposal were due and the proaised guidelines-for theoretical assessment has not been released to this day. Dr. Xaley will further discuss the many scientific 10 flaws in the analytical section of the current proposal. 11 . Third, the proposal's new requirement that any 13 quantifiable PCB in a waste stream from closed and controlled 13 processes, even if concentrations below SO ppm be disposed of 14 in an EFA-epproved PCB disposal facility, is without justification. ii The existing SO ppm cutoff for waste disposal has IT never been challenged and no justification exists for changing 10 it now. To require any waste with less than 50 ppm but 10 quantifiable PCBs to be handled by PCB disposal techniques 30 would be tremendously expensive with no consequent meaningful 31 reduction in risk. As much as a billion dollars annually 22 could be required for such disposal. 23 In promulgating the Hay 31st, '79 PCB rules, the 24 agency required PCB-approved disposal for substances 30 containing 50 ppm PCBs, rather than the previously proposed EXECUTIVE COURT REPORTERS (Ml) M Ml HONS 216135 1 1 17 SOO ppm PCBa However, in taking this stsp, tha agency recognized tha additional economic burdan and alleviated it by allowing laaa atringant mathoda of dispoaal for aubatancaa containing nora than 50 but laaa than 500 ppm PCBa. . Tha agancy ahould now recognire that a major part of that relief, high intanaity boilers, will not ba availabla for moat waata products containing laaa than 50 ppm PCBs. High intanaity boilara ara oparatad to ganarata ataam for a variaty of uaaa. Minaral oil, containing traca quantltiaa of PCBa, may ba a auitabla fual for oparating auch units, but chemical waata atraama will not normally ba accaptabla ainca thay do not rapraaant a conalatant fual and may hava componanta othar than PCBa that could laad to unacceptable boiler operations. Before turning to Dr, Kaley and tha issues related to the analytical guidelines which ara at tha core of EPA's proposal and of tha inadequacy of the guidelines, I would like to provide tha panel with a batter perspective on why PCBa are inadvertently generated, soma aspects of tha fata of auch PCBa, and how tha chemicals in which thay ara contained are handled. Traca polychlorinated biphenyls are likely to be formed in most chemical processes whara hydrocarbons and EXECUTIVE COURT REPORTERS (mi) ill am HOMS 216136 13 t chlorine exist *t high temperatures or if a catalyst such as 3 iron chloride are present. Three basic types of reactions j can yield chlorinated biphenyls. 4 One, coupling reactions with chlorinated aromatics; 5 two, chlorination reactions where aromatic chemicals are . present and temperature catalyst and/or chlorination is T sufficiently intensive to cause aromatiration; and three, 1 degradation reactions where conditions are such that 2 hydrocarbon radicals are generated in the presence of l0 chlorine radicals. In the first type of reaction, the point of , coupling is well-defined and the number of chlorinated 2 biphenyls are determined by the structure of the starting 4 materials. For example, bis-2,4-dichlorobenzoyl peroxide is decomposed under certain conditions to yisld trace quantities of 2,2*,4.4'-tetrachlorobiphenyl. Many variables will influence the relative yield t of PCS versue the deaired products, impuritiss, temperature, . catalyst, -- affects, pressure, time, concentration, and ^ other variables cea change the relative kenetice. Destruction or isomerization could occur undar the conditions of various reactions. However, in coupling a reactions, regardless of the concentration of PCBs, a single M or lizdted number of congeners will dominate all others. a In the second basic type when chlorinating executive court reporters (Ml) HI4M HONS 21*137 19 1 aromatics, a tubatitution of chlorina for hydrogen on th 2 aromatic will follow a aalactivity pattern that ia well- 2 taiovn. For example, when normal iron chloride catalyzed 4 chlorination of monochlorobenzene will yield about 55 percent ft paradichlorobenzene, 40 percent orthodichlorobenzene, and only 5 percent metadichlorobensene. 7 Condition! may modify the ratio alightly, but the baaic aelaction will continue to exiftt. Thus, aeveral congenere may exiat but a few will dominate. 0 To review the analysia -- a review of the analyaia 1 of commercial PCBs produced by the chlorination of aromatic 2 -- de -- aromatic -- biphenyl -- indicate that the moat 3 prevalent congener ranges from 32 percent of the total 4 mixture in lower chlorinated biphenyls to 9 percent of 5 the total mixture in higher chlorinated products. PCBs generated during incidental chlorination 7 should have similar selectivity in the ratio of total PCB to the most prevalent congener should vary from 3 to 1 to 11 to 1. From this, it is obvious that a 1 ppm per congener sensitivity limit means that the undetectable limit for 30 31 total PCS is 3 to 11 ppm. i n Another chlorination of aromatic material yielded a PCB contamination with 2$ identifiable congeners, but a 24 ratio of 150 between the most and least prevalent congener. The relative concentration of PCB versus desired chemical is EXECUTIVE COURT REPORTERS (Ml) MMM4 HONS 216138 ! 20 subject to all the sane variables as listed above and any single change will normally affect yield of both the impurity and the product and the relative ratio must then be redetermined. The third basic type of reaction leading to chlorinated biphenyl is degradation. Catalyst* high temperature processes, or hot spots with any process will yield a large number of chemicals and perhaps some PCBs. These chemicals in turn degrade based on thermo dynamic stability since PCBs are quite stable, some may survive. Steric and electronic effects dictate that different structures will have different thermodynamic stability. ' The relative concentration of reactants will dictate the approximate degree of chlorination. Such generation may yield many congeners but thermodynamic stability will determine that a few of these congeners will be prevalent probably in about the same ratio as is present in the chlorination reactions. If a chemical reaction yields trace quantities of PCBs, those impurities may end up in product, waste, or 21 recycled stream and the ratio of such distribution will 22 23 depend on the separation technique. No separation technique 24 is 100 percent effective, thus there will ^always be some molecules in every stream. EXECUTIVE COURT REPORTERS (Ml) MMW HONE 216139 21 Th physic*! property differencaa, namely, vapor pressure end water solubility, will determine the ease of separation and to some extent, the ratio of the constituents between the streams, whereas it is technically feasible to get good separation of decachlorobiphenyl from low boiling products, the same degree of separation may not H feasible for trichlorobiphenyl in products that boil between 150 and 200 degrees centigrade. ' Since these properties vary across the series of PCBs, separation techniques may change the congener distribution as well as the congener concentration. Irrespective of separation, a few congeners will still predominate. . Regardless of which of these three basic reaction types occur, the levels of FCB in final products are released to air and water during processing are in almost all substances -- instances -- very low, usually because some degree of separation has occurred and most of the PCBs are concentrated in the waste streams. As CMA has indicated in its contents, a regulatory cutoff of 50 ppm for PCBs in products is frequently achievable, meaning that products will typically contain less than 25 ppm. I am not sure many people recognize what a small quantity of PCBs are being discussed when I say 25 ppm. A 25 ppm concentration represents only one part EXECUTIVE COURT REPORTERS (Ml) HONS 216140 out of 40,000 or 0,0025 percent. In order to dramatize the triviality of such concentrations, I would like to put the numbers in terms more easily comprehended, ror example, if 660 pounds of PCBs representing a total annual quantity found.in commercial products in the CMA survey were contained in a chemical at 50 -- at 25 ppm concentration and were dumped on someone's doorstep, it might be considered by some that that person would have a PCB problem. That problem, however, would be trivial compared to the problem posed by the other 26,389,020 pounds of chemical on the doorstep, enough to fill ten olympic-sized swimming pools. Not only is 660 pounds trivial compared to the more than 26 million, but it would be almost impossible to get actual exposure to all of those 660 pounds given the other 26 million that you would be faced with. Imagine for a moment that you were standing alone just outside an athletic stadium filled with 80,000 blindfolded fans. All 80,000 are trying desperately to get out of the stadium but they don't know where the exits are and two fans are ruffians, that's 25 parts per million, while the other 79,998 members of the mob ere just trying to . get out. You can see that the ruffians' probability of a fast exit are minimal, but even if they did get out, the EXECUTIVE COURT REPORTERS (Ml) JU4M4 HONS 216141 23 chance that the ruffian would hava an advaraa effect on you ralativ* to tha probability of bain? trampled by tha other 79,998 ie neglible. So it is with PCBs containad at such low levels in unintantionally -- in intantionally manufacture chemicals. . Carrying our example of tha tan swimming pools further, where those other 26 million-plus pounds of non-PCB material water, there might be reason for concern about the 660 pounds of impurities. That is not the case. He are dealing with chemicals other than water. Yet, EPA totally ignores the fact that chemicals containing inadvertently generated PCBs have limits for handling because they are chemicals and often are also strictly regulated by the government. For example, carbon tetrachloride is one chemical for which PCB -- manufacturing ban exemption petitions have been filed. Carbon tetrachloride is already regulated by a myriad of agencies; the CPSC has banned -- its use in consumer products; contaminant levels have been set by EPA for drinking water; disposal is regulated under RCBA; the Department of Transportation regulates its movement; OSHA has set workplace -- standards; and the Mine Safety Bazard Act regulates the chemical's use in mines. If PCBs exist in very low concentrations in carbon tetrachloride, human and environmental exposures to EXECUTIVE COURT REPORTERS ()*i) ;mm HONS 216142 t the small amount of PCB will alraady ba strictly controlled 2 without a single PCB regulation. The carbon tetrachloride 2 regulations represent an important point ignored by CPA in 4 its proposal. The issue in the current CPA proceeding is entirely a different than that addressed by Congress in TSCA Section ? 6(e). PCBs were at one time intentionally manufactured and a millions of pounds per year entered the asaeasible a environment. 10 The absence of adequate control on such entry was 11 a problem worthy of concern that should have been and in 12 fact, in the last 15 years, has been addressed, inadvertent 13 generation is an issue of an entirely different stripe. 14 The issue of whether small amounts of inadvertently ia generated PCBS contained at very low levels in already TO controlled chemicals is worthy of further regulation -- 17 of further regulatory concern and an entirely different If answer is warrented. ia The potential for worker exposure to PCBs 20 inadvertently generated at some point within a chemical 21 process facility is much less than the potential for workers i 22 servicing and rebuilding electrical transformers. The low 23 concentrations of such chemicals in process facilities 24 automatically reduces exposure. 2B A wide range or protection measures currently are EXECUTIVE COURT REPORTERS (Ml) WWi MOWS 216143 23 1 utilized by industry because of a need to protect against toxic or corroaiva primary chemical*. Food, drink, and tobacco product* are frequently prohibited by the work areas, employee* may also be instructed to wash their hands and faces prior to eating or smoking. . .. If eye contact with a toxic or corrosive chemicals can occur, chemical showers and eye wash facilities are usually provided. Equipment or process engineering and design considerations are also important in minimizing the 18 potential for inadvertent releases in controlling intentional removal of a chemical from a process. 11 For instance, sample points are generally designed to be easily astestible and have appropriate 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. The chemical industry routinely employs spill prevention control and countermeasure systems for storage and handling of toxic chemicals. These measures already 30 address many of the primary chemicals with which inadvertently 21 generated PCS* are associated. 22 21 Moreover, the OSHA Act imposes a general duty on employers to provide safe workplaces regardless of other 34 regulations. Therefore, additional requirements addressing 21 EXECUTIVE COURT REPORTERS (Mg watt MOMS 216144 26 the low levels of inadvertently generated PCBs are unneeessar In addition, if there be any doubt that the trivial quantities of inadvertsntly generated PCBs pose any risk, several PCB specific regulations already exist. Under the Clean Hater Act, strict effluent limitations' exist for PCB discharges. 05HA regulates exposure to PCB in the workplace. PDA has set tolerance levels for PCB in food, and of course, EPA has established under TSCA 5(e) special disposal requlresMnts for PCB concentrations of 50 ppm or above. The case for additional regulation of inadvertently generated PCBs thus cannot be made. The risks posed, if any, are trivial and not worthy of regulatory concern. It is also worth pointing out that ths substantial economic cost -- that substantial economic cost would be posed were EPA to seek -- to control unduly inadvertently generated -- inadvertent generation of PCBs. I have already mentioned the tremendous potential oost of SPA* s new proposed requirement that any waste stream with quantifiable PCB disposed in an EPA-approved manner. More significantly, it is important that the agency recognise the infeasibility of producing of many important chemicals without inadvertent PCB generation. The risks posed by inadvertent generation are so trivial as to require no balancing of economic cost of EXECUTIVE COURT REPORTERS (Ml) MJ4SM HONS 216145 27 regulation against the risks to justify the absence of regulation. Nonetheless, it is worth noting that the potential economic impact of unreasonable regulations in light of TSCA's direction to determine unreasonable risk through such a balancing. Many chemicals from the complex reactions for PCBs can be generated are engineered for unique properties or for highly critical specialty applications. Substitution or replacement of these materials is often impossible or at best requires long-term expansive developetent programs. For example, PCBs are generated in the production of phenyl-containing silicone products typically used in highly critical military and commercial aircraft, military equipment, space vehicles, nuclear reactor applications. Even if lower performance materials could be substituted, such substitution would require long development programs. PCBs are also inadvertently generated in production of benzene phosphorus dichloride, an intermediate used in the production of a catalyst for nylon and carpeting. No substitute has been found for this intermediate material and without it, carpet quality nylon cannot be manufactured, withdrawal of that intermediate from the marketplace would require the producer of carpet nylong to shut down a large process line. In order to reestablish production, a producer EXECUTIVE COURT REPORTERS () KMOM HONS 216146 1 would be forc'd to embark on aubatantial and expenaive reeearch program for a aubatituta catalyat or a aubatituta for carpet nylon, a carpat manufaeturar would in turn -- in turn would aithar ba forcad to find a aubatituta fibar or go out of buainaaa. Tha conauoar in turn would ba facad with raducad choica in carpat aalaction. Similar example* axiat for many othar chamicala that contain traca PCS impuritiaa, in each caaa, tha high coat of finding aubatitutaa not containing thia unintentional impurity cannot ba juatifiad by tha minimal riaka poaad by ita preaence. EPA'a failure to review tha expoaure and affacta data that demonatrate no riaka worthy of regulatory concam -- that demonatrate that no riaka worthy of regulatory concern axiat from inadvertantly generated PC9a haa lad tha agency to propoae a rule that neither asiurss tha public that health and tha environment will ba protected nor aaaurea industry that draatic regulationa will not have a prohibitive impact on production of many important chamicala. CMA thua urge* tha cloaa and controlled waata propoaal ba withdraw and that EPA focua ita attention on a an achievable 50 ppm regulatory cutoff that will ba more 23 than adequate to aafeguard public health. 24 I would now like to turn tha floor over to Dr. a Kaley who will diacuaa tha important acientific flawa in EXECUTIVE COURT REPORTERS (101) 1(14004 HONS 216147 29 the analytical guidelines lying behind EPA's proposed rule. STATEMENT OF ROBERT G. KALEIf, II, MONSANTO COMPANY DR. RALEY: Good morning and thank you for this opportunity to testify. My name is Dr. Robert G. Kaley. I am currently a Senior Research_SpeciaIi*t~with Monsanto Company and a member of CMA's PCB Analytical Task Group. I received my Ph.D. in analytical chemistry from the University of Illinois in 1974. Since that time, I have been employed at Monsanto in various analytical capacities. In particular, I have had over eight years experience in the analysis for PCBs. I have co-authored several journal articles and presentations dealing with analyses for PCBs. In addition, I have served as Chairman of the PCB Task Group for ASTM Coasaittee D-19 on Water. At Monsanto, I was initially involved in developing analytical methods for Monsanto's PCB Trademark Aroclor products in a variety of environmental and experimental matrices. More recently, I have been involved both actively 20 21 and in a supervisory capacity in the development and i 22 implementation of analytical methods for the determination a of Inadvertently generated PCBs in various madia, 24 As my testimony today should make clear, method a development and analysis of inadvertently generated PCBa is EXECUTIVE COURT REPORTERS (**> NONE 216148 30 a totally different and Bore arduous time-consuming and analytically exacting task than the analysis for eoranercially produced PCB mixtures in environmental samples. In addition to developing these in-house methods, our laboratory has participated in smveral inter-laboratory tests of PCB methods, including the recent CM* round-robin. CMA has long considered sound analytical principles to be central to the development of appropriate PCB regulations, particularly in the context of inadvertently generated PCBs. for this reason, the CMA PCB Analytical Task Group was formed and has developed and submitted substantial analytical comments on several occasions in conjunction with EPA's rulemaking. We have appreciated these opportunities and look forward to providing additional analytical comments where appropriate. While 1 have included some comments on MRi's guideline protocols here, because of their complexity and because full comments are not possible until the methods have been run on actual samples, we hope to be able to make ore detailed constants later. Consistent with its past interest, CMA's PCB Analytical Task Group has carefully evaluated EPA's proposed closed and controlled waste rule. i#e conclude that EPA's proposal is analytically unsound and inappropriate. A nonguantifiable cutoff using any analytical EXECUTIVE COURT REPORTERS (Ml) W4M4 MQNS 216149 Ji method will rtiult in arbitrary and inconsistent regulatory results involving a moving regulatory target. Moreover, EPA's proposal does not establish a validated, workable method that can be followed by the analytical chemist since it leaves critical analytical procedures to his discretion. The only way to develop a rule consistent with the principles of analytical chemistry ie to establish a previously determined regulatory cutoff add develop analytics:, methodology to enforce it. With unlimited financial and analytical resources, analytical techniques can be refined to quantify lower and lower PCB amounts. Consequently, absent a previously specified cutoff, the analytical chemist has not way of determining when to stop seeking PCBs. Zn fact, EPA's proposed rule and guidelines essentially specify an analytical search for 2ero. Using an analytical method to identify the absence of a chemical as EPA's proposal in reality is intended to do is an unsound use of analytical principles. The analytical chemist can never warrant that a substance is not present in a sample but only that its concentration does not exceed a given level. As a result, a nonquantifiable cutoff constitutaa no more than an indirect regulatory cutoff defined by the chemist himself. It is imperative that EFA establish this cutoff EXECUTIVE COURT REPORTERS (mi) wen MOMS 216150 directly presumably by examining the relevant hazard 32 information and not indirectly through the development of analytical methods which bear no relationship to that hazard evidence. EPA's proposed nonquantifiable cutoff reflects a lack of appreciation of both analytical chemistry's limits and the complexity of the analytical problem posed by inadvertently generated PCBs. For a variety of reesons which Z will explore in more detail later in my testimony, a nonquantifiable cutoff provides no target for the analytical chemist. With enough analytical resources, a chemist would ultimately be able to quantify extremely low levels of PCBs which may be present in matrices associated with processes unintentionally generating PCBs. Accordingly, the principles of analytical chemistry do not and cannot tell the analytical chemist how extensively to analyse his samples as EPA seems to expect. But even if they could, because of the analytical complexity of inadvertently generated PCBs, a nonquantifiable approach will have inconsistent, arbitrary, and unfair regulatory results. Substances having similar PCS level and similar exposure potentials will be regulated differently simply because of analytical peculiarities associated with EXECUTIVE court REPORTERS (Ml) MMW MOMS 216151 dp 1 inadvertently generated PCBe. 33 2 Moreover, a nonquantifiable approach will result } in a moving regulatory cutoff even with the MRI's protocol 4 guidelInas. I Our July 8th comments highlighted aroma of the * irrational results of a nonquantifiable rule. For example, 7 due to the isotopic properties of chlorine in an electron impact ion source, monochlorinatad biphenyls will be more consistently quantifiable than decachlorinated biphenyl and 10 other higher chlorinated biphenyls. II Upon electron impact, monochlorobiphenyls generate II two predominant ions, while decachlorobiphenyl generate six II ions of 20 percent or greater relative intensity. This, in 14 effect, dilutes the analytical signal for decachlorobiphenyl. 10 As a result, an amount of mono will generate a to sere intense signal than the same amount or decs and is 17 consequently more likely to be quantifiable and thus 10 regulated. Other analytical problems exist under a non 10 quantif iable system. 10 None of the chromatographic systems described by 21 MX! or yet developed is capable of separating all possible 21 PCB congeners. Columns with different operational properties 23 thus resolve congeners differently. 34 For example, two different pentachlorinated 9 biphenyl isomers will show two peaks on the recommended OB-5 EXECUTIVE COURT REPORTERS (m) itt mi HONS 216152 34 capillary column, but only on* larger peak on a dexsil 410 capillkxy. column which ia also permitted undar Mftl's guideline*. Under some circumstance*, the larger cumulative __pe_ak will be goantlf < ehl a, whtrm the em.i^ iiwiiinf PCfej divided between the two peaks will not be leading to a different regulatory result. - Even if PCBs themselves were analytically less complex, the nonquantifiabl* approach will inevitably lead to inconsistent results. Sine* the limit of quantitation is a function of the analytical noise level of the total method which is directly dependent on the properties of the matrix, it would be purely fortuitous analytically if the limit of quantitation turned out to be the same in any two different matrices. Given the wide range of matrices, liquid to solid, simple to complex, halogenated to non-halogenated, limits of connotations will stand several orders of magnitude undar EPA's closed and control rule. As a result, if EPA's rule excludes any processes at all, it will still include widely varying quantities of PCBs for regulation. Not only will the limits of quantitation vary widely among substances, it will elso vary over time for any givan substance. That is, the approach establishes a moving regulatory cutoff. Significant improvements in EXECUTIVE COURT REPORTERS (*M) JM-CM4 MONS 216153 35 analytical hardware hava occurred over the past aavaral yasrs. Taka, for instanca, tha mats apactromatar. Soma older nodal* tuned for optimal operation a* specified by tha MAJ protocol exhibit more than an order of magnitude higher minimum detectable quantities than a newer instrument similarly tuned. Similar differences exist between older capillary columns and new ones with different bonded phases. Continued improvements are inevitable, thus even if the protocols were precisely defined and there were no other source of analytics] variability, the regulatory cutoff would decrease as this analytical hardware is improved. 13 Moreover, as methods and instruments are further 14 improved, E?A's rule provides no assurance that the protocols will not be modified repeatedly to incorporate new methods II _ H with lower limits of quantitation. Such changes will n necessitate costly revalidation with each incorporated i? is improvement. IS The uncertainty and inavitable inconsistency of a nonquantifieble approach is greatly magnified under EPA's 30 proposal since the proposal fails to establish a validatad, 31 workabla, analytical mathod for tha analytical chemist 33 leaving him to determine the limit of quantitation on his 23 own. 34 EPA's proposal in tha Fadaral Ragistar, of coursa. EXECUTIVE COURT REPORTERS (Ml) MJ-CBM MOMS 216154 ! 36 does not more than specify the measurement tachniqua to ba used, capillary GC alactron impact# man spectrometry, leaving tha full analytical method a* an unvalidated aeries of nebulous options. Evan EPA's guideline as described dn-MRI's interim protocols leave tha most critical decisions to the individual analytical chemists and are largely unworkable and still unvalidated. Ironically, in contrast to EPA's original proposal# these guidelines offer a choice of separation techniques# either packed column or capillary column gas chroowtography. While the CMA approves this choice as scientifically justified, this new choice highlights the great range of decisions left to the individual analyst. This choice, like most of those left to the chemist, will affect ultimate quantiflability of certain isomer* or isomer pairs. In addition to offering this choice of separation techniques, MRX's guidelines leave totally open the question of extraction, clean-up# or concentration procedures. MRI's I guidelines permit dilution or direct injection, that is, no sample pretreatment at all or a choice of numerous extraction methods. The guidelines explicitly state in describing liquid-liquid extraction, the solvent -- number of extractions# solvent-to-sample ratio# and other parameters EXECUTIVE COURT REPORTERS (Ml) HMW HONS 216155 uniquly dependent on the nature of the original matrix, Tha othar chamieala it may contain at various lavels and similar factors and tha limit of quantitation can only be specified after proper method validation. indeed, in the analytical -task group's experience, many of the techniques described in Mm's guidelines have proven unworkable or impractical for inadvertently generated PCBs in organic matrices with analytical properties similar to those of PCBs themselves. The extraction and clean-up techniques suggestsd in the guidelines were designed for use on environmental or other samples containing cosuercial PC8 mixtures such as Aroclor, Kaneclor, and other fluids. Such methods are not readily transferable to the types of analyses required under this rule. For example, an environmental matrices which normally contain a relatively small quantity of chemicals with properties similar to those of PCBs, the amount of absorbent and solvent needed in the recommended clean-ups to separate interferences is manageable, but in product or waste matrices in which the ratio of interferences to PCBs is orders of magnitude larger, the amount of needed absorbent and solvent is unmanagebly high. Similarly, none of the proposed extraction-clean-up techniques ia applicable to tarry matrices which sometimes EXECUTIVE COURT REPORTERS (j*l) 1CJ-4M4 MONS 216156 37 I I 1 are chosen at the analyst'a discretion. Similarly for a clean-up, the guidelines describe eight alternative 3 techniques but state only that a surrogate spiked sample is to be cleaned up at the discretion of the analyst. s The degree end rigor or extraction and clean-up are central both to the ultimate performance and limits of 7 quantitation of any PCB analytical method and to its cost. In the experience of the members of the task group, 0 these steps in a procedure affect ultimate quantiflability 10 by several orders of magnitude. Indeed, limits of detection 11 and quantitation for PCBs in environmental samples have II fallen several orders of magnitude over the past 10 to 15 13 years because of improvements in sample clean-up which ' 14 removed interfering pesticide residues. If Similar differences in limits of quantitation are " If likely to exist between those samples under this rule which 17 are subjected to clean-up and those which are not. It is If not surprising that EPA has been unable to specify these If method steps. 20 No capillary GC electron impact mass spec method 11 has yet been validated for use on the types of samples to t 33 which the method pertains and many of the suggested alternatives have never been tried with analyses for 24 Inadvertently generated PCBs. 'S Proper extraction and clean-up techniques are EXECUTIVE COURT REPORTERS (Ml) H14M4 HONS 216157 i ! iti i i ! ! i 39 1 will have to be analyzed. Even worse, some of the matrices 2 which contain inadvertently generated PCBs are themselves 3 the beet solvents for PCBs making any extraction or clean-up 4 virtually impossible. 5 Xn short, EPA's proposed nonquantifiable cutoff will have arbitrary end inconsistent results. A non7 quantifiable cutoff will result in substances with similar PCB levels being regulated differently, both because of the 9 substantially differing matrices and because of the 10 complexity of analyzing all 209 chlorobiphenyl congeners. 11 This inconsistency is compounded by EPA's failure 12 and ultimate inability to specify the required analytical 13 procedures. Most importantly, the nonquantifiable cutoff 14 provides no analytical target at all thus requiring the 10 analytical chemist himself to define the regulatory cutoff 10 instead of the regulatory authority. 17 The only analytically sensible way to define 10 excluded processes is to specify in advance the regulatory 10 cutoff, that is, the performance standard or criteria the 20 ansdyet suet meet. Only a cutoff, determined primarily on 21 the basis of appropriate health and exposure related factors 22 gives the analytical chemist the guidance necessary to make 23 decisions concerning necessary analytical procedures. 24 Moreover, only a cutoff set at an analytically 20 feasible level can eliminate the irrational and EXECUTIVE COURT REPORTERS (mi) iv+m MOMS 216158 40 1 irroconcilobl* rooulto of EPA's nonquantifiabl* approach. 2 Of course, avon a cutoff must bo accompanied by appropriate 3 analytical guidelines to guide compliance and enforcement. 4 As I suggested at the output -- outset -- a nonquantifiable cutoff with no lower limit presents the analytical chemist with an impossible dilesna. As analytical techniques have improved, analytical chemists have concluded 7 I that it is virtually impossible to state that any chemical is absent from a particular matrix. In the case of PCBs, by investing greater and 10 II greater resources into the analysis of any given sample, lower and lower amounts of PCB congeners will become 13 13 quantifiable almost without limit. For example, by using 14 successive exhaustive and highly expensive chromatographic clean-up procedures, some samples could be cleaned up and 15 interferences removed sufficiently to aahe low part per billion or part per trillion concentrations of some PCB 17 10 congeners quantifiable. (1* Since only a single congener need be quantifiable to eliminate the regulatory exclusion, the limit of 20 quantitation in this light is an extremely low threshold. 21. Similarly, adjustments in the mass spectrometer can be 22 devised to further reduce the limit of quantitation. 23 Of course, the coot in terms of time and resources 24 20 become prohibitive for routine usage of such procedures. EXECUTIVE COURT REPORTERS (Ml) MMtM HONS 216159 41 1 In light Of these possibilities, whan can the 3 analyst atop analysing his samplas and honastly say that 3 PCBs ara not guantifiabla? Analytical chanistry itsalf 4 providas no answer. indaad, EPA's guidalinas designed to . answar this quastion suggest that the chemist must g investigate available procedures and analyse his samples until he quantifies PCBs without any limit. I For example, MRI specified GC-MS operating . parameters are described as minimum suggesting that a more |0 sensitive parameter may, or perhaps should be used. MRl's guidelines on Page B-13. The guidelines specify that simple . volumetric dilution rather than extraction may be used where . the PCB, concentration is high implying that alternate and continually more rigorous procedures must be chosen when low concentrations are involved. Host significantly, in a somewhat ambiguous reference, the protocol states that upon an injection of a t sample aliquot, if the responses of any PCB ion are below the suss spectrometer's working range, recheck system performance, concentrate the sample, emphasised, and reanalyze. 21 It is difficult to escape the conclusion that a samples are to be analysed using whatever procedures are a necessary to detect and measure PCBs. Indeed, EPA's M proposal itself suggests the same thing. t is apparently EXECUTIVE COURT REPORTERS I (JWJ HMW HONS 216160 42 designed to exclude from regulation only those processes emitting zero PCBs or as close to zero as can be measured under the most exacting procedures. ha our comments point out, FDA considered and rejected an identical regulatory search.for zero regime as unreasonable and unmanageable, EPA should do the same. In short, analytical chemistry cannot answer the fundamental question of the level of PCBs the analytical chemist should seek in his samples. Moreover, were EPA to attempt to define precisely and unambiguously the required analytical procedure, it would face the same questions faced by the analytical chemist under the rulei what procedures are sufficient? Specification of a procedure will thus constitute a de factor, albeit variable, regulatory cutoff. Only an independent determination of the acceptable PCS levels can begin to provide the information necessary to answer the question of what the cutoff ahould be. Moreover, only a cutoff can eliminate the irrationality and uncertainty which are inevitable under a nonquantifiable approach. A known specified cutoff provides the chemist a target toward which he can calibrate and develop hie methods. Knowing the general composition of his samples and performance of his analytical techniques on those samples, EXECUTIVE COURT REPORTERS (M) MMM4 MOMS 216161 43 t he csn develop techniques sensitive enough to assure that the 2 sum of all possible congeners cannot exceed the cutoff if it 2 is set at an analytically feasible level. 4 If his product contains lower chlorinated 9 biphenyls which generate fewer but more Intense peaks than 5 the higher chlorinated congeners, he may be able to use 7 different clean-up techniques to assure that he would be able to measure these congeners at the cutoff level. t Similarly, if two congeners co-allude on his 10 column, his calculation procedure would account for this and 11 would be adequate to determine whether he meets the cutoff. 12 Of course, a regulatory cutoff can only be enforced using 13 analytical methods. 14 Accordingly, a cutoff must be accompanied by appropriate analytioal guidelines both to promote uniform IB compliance with the rule and to guide enforcement. 10 CMA's PCB Analytical Task Group will gladly assist 17 15 the agency in developing appropriate guidelines. He feel IS such guidelines should contain, first, a description of the ao analytical method to be used for enforcement purposes. GC-MS techniques are the preferred measurement techniques, 21 though as out previous comments show, extraction and clean-up 22 techniques are inevitably unique to the matrix under study. 22 Secondly, a description of appropriate validation 24 n criteria. Every method must be validated according to such EXECUTIVE COURT REPORTERS (Hi) JU4N4 HONS 216162 44 criteria before ita use for regulatory purposes to ensure data reliability. Validation must be accomplished on the matrix to which the method is to be applied and be performed over a range of concentration above and below the cutoff level. . .... '---------- Thirdly, a description of appropriate Quality assurance procedures to assure data reliability in ongoing analyses. Quality assurance procedures should be adopted to assure purity of standards, to establish the performance of laboratory equipment, and to assess method performance criteria. Host importantly, guidelines accompanying any cutoff must recognize the uncertainty inherent in any analytical methodology. Validated methods include a measure of variability which must be acknowledged in making regulatory decisions. Recent commentators have reemphasized this important point and I'd like to quote the following from a statement which has appeared in Analytical Chemistry by h. 8. Rogers, et al. 'Analytical chemical data are central to governsmntal and industrial decisions relevant to regulations that can have far-reaching impacts on society. Mien the results of chemical analysis are used for regulatory purpose, the information must be reliable and creditable from EXECUTIVE COURT REPORTERS (m) m*m HONS 216163 45 1 4 scientific point of view." ... 2 "Analytical chemists must always emphasize to the 3 public thet the single most important characteristic of any 4 analytical chemical result obtained is an adequate statement 1 of its uncertainty interval. It is critical that the public appreciate the existence and nature of uncertainty in 7 scientific measurement.* . In other words, every method has a definable . S precision which delineates how the results of repetitive 10 analyses of a given sample will vary around the measured tt value. This range of variability is statistically 12 characterized as the method standard deviation often 13 expressed as relative standard deviation or confidence 14 interval. It Because of this variation, one can never have "* 11 1 perfect confidence that he has identified the true value 17 contained in a sample. Every measurement is accompanied It 1 by a confidence interval. Guidelines accompanying a It regulatory cutoff should recognize this variability and 20 indicate that analyses must provide at least 95 percent 21 confidence that the cutoff has been exceeded before 4 22 regulatory action is taken. 22 In addition to those points discussed above, EPA's 24 proposal has certain technical deficiencies in other aspects '.. 3t which deserve comment. EXECUTIVE COURT REPORTERS (Ml) NMN4 MOMS 216164 <46 First, *s noted in our conznsnts, EPA's rule appesn to be based on en incorrect concept of limit of quantitation. To summarize, the proposal appears to base quantifiability on the noise level generated by the instrument itself in the absence of the matrix to be analyzed. . ~-- . Such a definition is not based on sound analytical principles and cannot be justified. The American Chemical Society, in discussing method development and validation, has confirmed this point recognizing that for some samples, relative variation due to matrix effects is substantial. The ACS cormittee concluded that, and I quote, "If the field blank, actual or simulated, is not properly defined in the protocol, then the reeults are invalid." EPA's protocol clearly must base the limit of quantitation on the ratio of the signal to the noise from the total method run on the actual matrix. Second, I wish to emphasise that EPA's proposal will impose substantial analytical coats that have not b*n considered in EPA's assessment. These costs are likely to be especially onerous for small businesses with processes potentially generating incidental PCBa because they often do not have the sophisticated analytical capability or sufficient analytical resources. Our consents mention that EPA's assessment omitted consideration of significant method developswnt costs. In EXECUTIVE COURT REPORTERS (JM) AQNS 216165 47 addition, we stated that MRX's method was not routine and that therefore, EPA'a estimate of monitoring costs may be underestimated substantially. Turther review of MRI'a guidelines confirms that its protocol does not constitute a routine method and win be extremely costly to operate. In particular, the guidelinen impose quality assurance criteria that far exceed routine monitoring requirements,' particularly where one is monitoring a manufactured substance not subject to wide variability. Most significantly, the method requires the laboretory demonstrate recoveries from 10 to 120 percent with relative standard deviation of plus or adnus 10 percent. For some matrices, these are exacting costly standards if they are attainable at all. In addition, the method contemplates daily documentation of performance criteria and requires that 10 percent of the qualitative determinations and 10 percent of the quantitative measurements be checked by a second mass spectrometer. Indeed, many of the industries affected by this proposed rule do not even have one mass spectrometer. These and other requirements will add dramatically to coats incurred by both industrial and regulatory laboratories for analyses which already average about $1,000 a sample. Finally and most important, EFA's claim concerning EXECUTIVE COURT REPORTERS (Ml) HMW HOMS 216166 the validation of mrx's protocol deserves comment. e?A 48 suggests that MRI's preliminary validation teats indicate that the proposed method is applicable and useful for the .analyses of PCBs in the matrices study. The available data do not demonstrate-the adequacy of MRI's method. Indeed, based on those data, it is premature to conclude that MRI's protocol is workable and reliable for the matrices of concern here, until very recently, MRI's methods had never been tested on any matrices like those to which it must be applied under the rule. MRI attempted to validate selected clean-up steps by analysing a mixture of 11 calibration standard congeners in an appropriate solvent. Aa we have repeatedly stated, validation must take place on actual matrices since it is the similarity of matrices to PCS that makes analysis and clean-up for inadvertently generated PCBs so difficult. Accordingly, MRI's standard solution validation says little about the performance of the method in actual matrices. Indeed, MRI indicated that the experiment should be repeated. In late May, CMA sent MRI aliquots of the samples used in CMA's round-robin for further validation work. While five aaisplee were sent, analyses of only two are reported. The extraction and clean-up procedures used EXECUTIVE COURT REPORTERS (Ml) MJ4M4 HONS 216167 are not specified. 43 Racovary valuaa for aach analysis are not reported either. MRI raports the average PCS laval it found in samples CMA-A aa 400 parta par million. Thia average, during CMA'a round-robin -- axcuaa raa, the. yerage during CMA's round-robin of 25 determinations in nina laboratories, eight of which used GC-MS on that sample, waa 280 parts per million. ~ The highaat value reported in the round-robin from a single laboratory was 412 parts par million using the relatively nonspecific electron capture detector. MRI'a validation results for sample CMA-E ware similar. MRI's average was about 18 parts per million, twice that of the round-robin. These results raise significant concerns about the applicability of MRX's methods as currently written to these types of samples, especially since these are not the most difficult matrices encountered by the CMA task group laboratories. Whatever the procedures used, they resulted in values up to 100 percent higher than those of laboratories more experienced in these types of analyses. Consequently, these data surely do not provide a basis for supporting the applicability of Kill's methods and suggest that the methods as written have potential substantial high bias whieh is EXECUTIVE COURT REPORTERS (m) MMW HONS 216166 50 1 significant whan dealing with a nonquantifiable cutoff. 2 Finally, MRX's validation fails to addrasa perhaps 9 the nost important consideration for this ruls. Zt does not 4 specify or even estimate the limit of quantitation for these s or any other samples. Limit of quantitation Cs not even addressed in MRZ's interim guidelines. 7 Moreover, the guidelines provide no; practical information how to determine LOO for any particular sample. t Since the limit of quantitation is clearly the most critical 10 performance parameter for this proposed rule, it must be 11 experimentally determined in a statistically reliable way 12 for each individual matrix. 13 MRI has failed to provide information on how to 14 determine LOO or how to validate LOO determination in any 10 specific samples. Accordingly, EPA's proposed method is fi far from validated and is not practically usable to comply 17 with or enforce EPA's proposed rules. It Mill's protocols are, by and large, technically It competent though EPA cannot assume that they are suitable 20 for use with its rule. But this overall technical competence should not obscure the fundemental point 21 that EPA's whole nonquantifiable approach la analytically 22 23 unsound. Zt will lead to inconsistent and irrational 24 a results. It does not provide a validated, workable, EXECUTIVE COURT REPORTERS (Ml) MMW MOMS 216169 51 enforcement or compliance method and in asssncs, requires each individual analytical chamiat to dafina the laval of PCBs that ia quantifiable. Only a specified regulatory cutoff xan remedy thaaa problems. Indaad, only a prior datarnination of tha permissible PCS laval allows aanaibla daciaiona concerning nacaaaary analytical procaduraa. CPA mat make thia datarnination diractly. It ia unfair and auraly inappropriata to ahift thia daciaion to tha analytical chaniet in tha regulated induatry aa tha currant rula doaa. I appraciata thia opportunity to taatify and wiah to thank tha members of tha CMA Analytical Taak Group for thair aaaiatanca. Kan will ba happy now to addraaa any quaationa. DA. BURGESSt Hr. Guimond, or. Cox did hava to laava but I think Or. Xalay and myself will at laaat try to anawar whatever queationa you hava. HR. GUIMOND: Fine, thank you very much. I'm X aura wa'll have a number of queationa froa the various members of tha panel. I'd like to start off on my left a with Alan carpien. a HR. CARRIERi I only have two brief statements, quaationa, auch of what we'll ba discussing hara obviously 34 will ba of a technical nature and I'm aura --by tha way, EXECUTIVE COURT REPORTERS (Ml) HMM HONS 216170 52 1 thank you vary much for your highly compatant tachnical ciantific tastimony. z want to maka ona clarification. Savaral tinas 4 during tha taatimony, tha statamonts wars atada aaa if I can find tha axact paga -- that this ragulation is doslgnod C to axcluda fro ragulation only thosa procaaaaa andtting zaro 7 PCSs or as closa to taro as can ba maaaurad undar tha most axacting procaduras. I'd lika to point out that whilo EPA has statad 10 that wa'ra axcluding thaaa procassaa, a class of procaasas 11 within thia rula, I don't baliava -- it's not a quaation of not ballaving -- st no point in tha ragulation or at any tiaa during this proeadura hava wa avar said that only thass procassaa will ba axcludad from ragulation. EPA has always maintalnad that wa hava a third rula -- you know, wa'va talkad about tha alactrical usa rula, wa talk about this rula, and wa know thara's a third rula. Whila this rula doaa axcluda eartain procassaa, wa hava by no ataans avar statad that it is tha only, thass ara tha only procaaaaa, or that this rula will astablish a ragulatory cutoff. t a My only quastion raally, I guass, that I could ask a that -- it's sort of ralatad to that and that would bo 94 supposo this rula want out as is or substantially as is a with ehangos basad upon many of tha coanants and in tha EXECUTIVE COURT REPORTERS (Wl) JO Ml HONS 216171 [ iI i i 1 third rule ws were to propose or consider * so part per 2 million cutoff, would CMA, as an association, hava objections 3 to that? 4 OR. BURGESS: I think -- wa'va got to_J.ook at tha I two things quite separately. Wa have heard about potential t Rule 3, we've seen nothing nor heard about the specifics of Rule 3. Aa we have said, we think a 50 ppm cutoff is a 7 reasonable, justifiable cutoff, and whet resemblance it bears to this rule or why continue working on this rule if such a thing is to be a part of Rule 3 is a little hard 10 It to eomprehend. ' 12 MR. CARPIEN: Again, thank you very much for your 13 testimony and let's get on with the technical aspects of this. Thank you. 14 ` MR. GUXMONDt Bill? II Ml. oUHTERi Dr. Burgess, I'd like to follow-up on II that question- It seems to me there art two possibilities 17 with respect to this rule that ERA could go forward with it II II or they could do as cma suggests, to withdraw it, and there are two possibilities with respect to Rule No. 3, that epa 20 could pursue the course that CMA recomnends by promulgating 21 a cutoff or we could pursue some other regulatory approach a a end I'd like to look et the various combinations of those approaches. 24 a Case No. 1 would be if we went ahead and issued EXECUTIVE COURT REPORTERS (Ml) J4J-MM HQNS 216172 54 1 this rul* and than can# out with a cutoff in Rula 3. Cat# No. 2 would ba if wa did not procaad with a 3 thia rula and cana out with a cutoff in Rula 3. 4 Can you axplain to me how anybody would ba any I battar off undar Caaa No. 2 than undar Cata No. 1? OR. BURGESS: I'm not <ura what you mean by anyona. 7 No. 1, wa could gat at tha raal issuaa aarliar if wa < discontinuad Casa No. 1. If Casa NO. 1, as wa hava pramined, and I must say again, as I did in ay comments, that wa think 10 thara is no banafit to anyona from Rula 2 or whatavar you II want to call this rula, but tha analytical thing is so 13 dafinita -- so indafinita -- that mayba thara's banafits 13 thara that I haven't saan. I'va heard analytical chemists, pratty intelligent people, arguing at fairly great length as to just what analytical procedure is and what limit of quantification relative to what has bean proposed is going to ba and I have to admit that as a non-analytical chamists, 1 can sit back and say that tha order of magnitude that Dr. Raley was talking about frightens me but from what I haard, it's not an an order of magnitude, it's several orders of magnitude. HR. GUIMOND: Could you speak closer to the microphone? 24 DR. BURGESS: I'll aim it at my -- perhaps that EXECUTIVE COURT REPORTERS (Ml) JO-4M, will -- is it batter in tha back? 55 Tasting one, two, three, four -- can you hear me now? It would depend on the timing, on the.sequence in what is Kule 3, You're asking a question to which really there is no definitive answer, MR, GUNTERi If Rule 3 were cut off, would anybody be batter of by E7A not promulgating Rule 2 as a result? DR. BURGESSi It depanda on what the cutoff is, it depends on many factors. 1 guess I can turn it around, mould anyone -- would the effort that's going into Rule 2 have produced any results? You know, you're spending time, we're spending time. MR. GUNTERt Let me ^o back to some of the origins of this rule. This rule was suggested originally by cma in discussions that followed the decision of the Court of Appeals and now CMA seems to be almost totally reversing themselves. Is it because you've rethought the matter or is there some particular aspect of the way SPA ia implementing this that troublaa you? 1 think you hava identified two of tboae iaauaa today, ona being the nonquantifiable approach versus aatting soma limit and the other being the disposel cost. la it thet theee aspects art so diffarant from Executive Court reporters (Ml) MMth HONS 216174 56 1 whit you originally conciivid that you no longar support the 3 idaa or hava you had lacond thoughts about tha concapt in 3 ganaral? 4 DR. BURGESSi Tha concapt as 1 identified in my I testimony, vs still agraa with. Z think it's probably a cats I of poor communications batwain ouraalvas and tha ;paopla in 7 tha agancy whan wa originally proposed that thing. I think t avaryona in tha industry raeognised that a limit of detection or limit of quantification would be totally unacceptable. 10 That was tha worst thing in my estimation and I expressed it to many people many times, that tha worst thing would be 11 a limit of detection or limit of quantification. 13 To go with a no-regulated quantity released, and 13 by that in ay mind, I was talking 50 ppm, was what we were 14 talking -- I think it was a case of very bad miscoonunication II between the two groups. Certainly anyone who has ever worked II with FDA and the Delaney clause recognised the problem of 17 II taro. II MR. GUNTER* Would CMA's objection be at least partially satisfied if EPA specified for purposes of this 30 rule only numerical values that would be approximately equal 31 a to currently achievable LOQ'a? a DR. burgesst z guess I would have to review that in soma detail. Number one, as Dr. Relay has testified, 34 a numerical values limited to LOQ's is going to be.considerably EXECUTIVE COURT REPORTERS (MI) HMW HONS 216175 different for different congeners and different matrices- So I'a not sure how you're going to do that. In one case, you may be at 10 parts per billion and another case at 100 parts per million. - MR. CUSTERi Dr. Raley, your testimony indicated that this regulatory approach put quite a burden on the analytical chemist and -- much better from his point of view if there were some value established and then a procedure designed to meet that value and that's the point I'm trying to get at. DR. BURGESSt Hell, Mr. Gunter, I have used an analogy -- I'm getting a reputation for analogies, I guess -- I have used an analogy that I think is appropriate here for the analytical chemist. It's like telling an engineer design a car to go 100 miles an hour efficiently or design csr to go as fast as it will go efficienty. Those are two different things, orders of magnitude difference in what you're asking the engineer or in this case, the analytical chemist to do, and I think her* when you say limit of quantification, you're asking the IQ analytical chemist to design a process that will do the 21 n best it can. a MR. GUNTER*. But my question was that if EFA specified numbers that would give the analytical chemist a u a specific target to shoot for such as the 100 miles per hour EXECUTIVE COURT REPORTERS (Ml) >*1-00*4 MOMS 216176 AM/3? 38-39 i in your analogy, would that maat at laaat some of cma'b objections? DR. KALEY: Wall, first of all, I guass it's obvious to both of us that that depends on what-those-numbers night be. I think the answer to the rest of the question it in certain cases, 1 think that would be a benefit. The advantage to us would be that It would allow us to target our resources to those matrices which we do not feel we are at present capable of reaching a cutoff analysis in. So if we had a reasonable analytical cutoff or a regulatory cutoff, then we would have a feel for where those resources needed to be aimed in order to cleer up the questions we still have. Xn the current proposed rule, I we have to assume that all of our resources have to be applied to all of the samples in order to determine our I limit of quantitation in every particular sample and it'* , just an economic and analytical impossibility. End Tape 2) 30 at 33 a 14 as EXECUTIVE COURT REPORTERS (JC1) MOWS 216177 60 MR. GUNTER: One more follow up question. you don'j have to reipond today, but you could give it some thought j and perhaps give u* your answer in reply coranents. In your opinion, whet should that analytical base value be for the various media? I would like to turn finally to one.last point, and that is the disposal costs that you cited, Or. Burgess, for the waste generated from these processes. And, you pointed out that when EPA lowered the concentration cut off for the disposal from 500 to 50 parts per million, we specified certain alternate, less stringent disposal methods. Could you suggest any alternate, less stringent disposal methods for these wastes that have under SO parts per million PCB's? OR. BURGESS: Of course, many of those wastes contain, as do many chemical plant wastes, contain hatardous substances other than PCB's and are currently being processed as such. To go into the limit -- the available PCS approved disposal, whereas many -- it would be a problem because as l say, many of tham are already going to incine rttore as the CMA document indicated and if you ara talking, as your and numbars cama up with 100,000 pounds of materiel, if all of that wara going to an ineinarator, which ware 99.9 HONS 216170 i ! 3 I t 3 1} 61 versus the 99.999 on an Amex On* incinerator that would be required if we went to PCB approved incineration, you are talking the difference between on* pound and 99 pounds j I | or something such as that. 1 didn't -- I'm not. sure, of my numbers, but it's not very significantly different.. And, yet, the 99.9 incinverator is in general not going to be available to us. Many of those wastes should be incinerated. There is little doubt about that. Many of those wastes should be incinerated. Most of the incinerators where they are now going are perfectly competent to handle them, are probably over 99.9. Some of them, in this case it would not make the same degree of difference, could go to landfill and b* handled that way. MR. GUNTER*. Would it make much of a dent in that one million dollar figure if EPA defined incineration to be in any RCRA approved facility for those that do fall in the category you mentioned of being haiardous wastes? DR. BURGESSt Certainly, if some of the RCRA definitions were included in that and those incinerators, I'm not that familiar with RCRA. Z know some of those are called process units, rather than incinerators, and there are a lot of different definitions, But, that type of thing would certainly help. MR. GUNTER; That's all I have. Thank you. HONS 216179 62 I 1 DR. BURGESS: Before I leave today, I should point j I out that much of that wasta, which is going to contribute significantly to the large number and we are trying to gat a more definite fix on this number and hope to have it by the time oi the reply consnenta, one of the big problems is water. And the incineration of water is en expensive operation, it can be done, but it is expensive. MS. CAMPBELL*. My first question relates to the issue of water which you Just brought up, PCS contaminated water. I was wondering how the costs were increased by this regulation, since the affluent guidelines limitations ar< already lower than anything we are proposing in this rule? DR. BURGESS: But you're requiring that waste be disposed of by a PCB approved disposal facility. And those facilities are different than RCRA or affluent guidelines or any other facility. MS- CAMPBELL: Okay. The other question we asked in the regulation, but I was wondering if you have any date related to the number of accidental or unplanned releases of PCB*s and, if so, what amount of PCB's would be released 31 during those events? a DR. BURGESS: I think the data that was submitted a in the Verear report is as accurate as anything that we 34 would have, relative to that. Certainly, spills -- we don't a MONS 216180 63 have any hard data, but -- and, I think they treat that quite fairly. They didn*t really deal with explosions. Those , j I .-`re euch infrequent things that we doit*t have good ' statistics on it. Certainly, it's somethin? we try to | avoid rsther vigorously and when we have an explosion, the problem is usually not the PCB's, the exploeion of PCB's,- if you have an explosion, you have a number of elements that you have to take care of, people injured by shrapnel, people burned, something such as that, and you have a clean up procedure that you have to go through, whether there are PCS's there or not. But, the frequency of such things is so da minimis that I don't have good data on that- MB. CAMPBELL: 1 would also be interested in the type of protective clothing and the kind of material in the clothing used during the times of worker exposure that were indicated in the CMA submission- DR. BURGESS: It will vary with different plants and different chemicals. If -- and, it is primarily aimed at the primary chemical. The protection will be aimed at the primary chemicals. Most plants will have procedures for so-called hazardous operations as line opening. If you are going to do a maintenance job and have to open the line. | HONS 216101 64 The normal requirement there would be face shield, *s wall as, in most, just common working, you would I have ay* protaction and probably gloves. j If you are handling waste streams', normally you are going to hava glovaa on. Versar, avan though they found the expoaure quite email, did assume that that axposura came from the fact that anyone handling waste had both hands constantly vat on both sides during the entire operation of handling vastae. I would certainly hope no one would practice that, whether you ware doing it yourself or asking someone else. That type of thing would not be dona. Certainly, some nonex clothing in some cases, if there is hazard from fire from the primary chemicals, but very little of that would be aimed at the PCB. It would be aimed at the primary chemicals. MR. GUNTERi Amy Moll? MS. MOLLi You continually cite in your analysis the fact that EPA should call for a method validation and also that EPA underestimated its costs because it did not include the cost of method development and method validation. You probably don't have figures on the tip of your tongue, but if you could provide us with some estimates of the costs of method development and method MOMS 216102 I validation, chat would bs halpful to us. DR. BURGESS: Wall, obviously, ths mors validations you do on ths mors strsims, ths sscond ons is not going to bs ss sxpsnsivs ss ths first, but it is still an sxpansiva procsdurs. Rob, did you want to havs any commont ;or should ws just commant in raply cotnmsnts? DR. RALEYi Wsll, I think primarily anything mors spacific that ws might havs to say will havs to bs in ths rsply conmants. 1 think ws did maks sons limitsd information avallabIs in conjunction with both our first submission, "Ths Analysis of Chlorinatsd Biphanyls* and in ths rsport on ths round robin sxpsrimsnt. And, I'm not sura vs'va rsally got anything bsttsr than that. Ws havs discusssd thsss issuss at our various msstings and it bscomss difficult bscausa, frankly, ws don't know what it is going to taka to validats a mathod such as propossd in ths varisty of matrixss that ws will havs to sneountsr. And, ths limitsd quantitation issua just makss avan mors critical, bscausa our validation, instsad of bsing soma rslativsly workabls lsval, liks 50 parts par million, has to -- aasantially, to bs dons right, axtand balow our limit of quantitation and ws'rs rsally not sura how to MONS 216103 address that at this point. 6f I will taka down your quaation. Wa will do our bast to get soma more -- gat aoraa firm antwars to you in tha raply commants. , MS. MOLL: Oo you faal that tha aathod.will have to ba validatad on avary single diffarant process that vs ancountar? DR. RALEY: Yes, wa definitely do faal that. MS. MOLL: Okay. DR, RALEY That'a ona of tha primary iaauas of my testimony, ia that validation muat ba on tha matrix of intaraat. So, for avary ragulatad or axcludad matrix, wa will hava to hava a validatad mathod. DR. BURGESS*. Your quaation thara, you uaa tha tarm procaaa; wa uaa tha tarn matrix. Thay may ba analogous, but thara may ba situations whore thay aran't also. MS. MOLL: Okay. You citad, Dr. Raley, you cited tha potentially large burden on small businesses, that eould result from this rule. Do you hava any idea as to tha number of small businesses that might ba affected? DR. BURGESS: I think it would ba pure speculation. Versar has listed tha processes that, although I'm not sura that that is comprehensive, but it MOMS 216184 67 would be pur* speculation to 90 beyond that list that v*rsar has already conpilad. MS. MOLL; Okay. And, I hays on* mors follow up on Bill's qusstion. Hs asksd you about suggestions for dsstructlon of PCB's in concsntrations bslow 50 parts par million. - Ones you'vs com* up with that suggestion, can you give us any idsa as to ths differences in costs bstwasn tht methods that discharge each of these in concsntrations greater than 50 parts par million and those that you might suggest ars less than 50 parts per million? OR. BURGESSt Hell, I think ths main thing that wa ar* looking at, ths last figur* that I heard for destruction of PCB's in an Amex Ons incinerator was around 60 cants a pound and that is what is quoted in our coimnents. That certainly is a figure that is much higher and whether it's 5 times higher or 10 times higher than incineration of normal wastes, but I think we're talking in that ballpark, that most wastes can be incinerated for -- maybe it's somewhere between 5 and 20 cents a pound, although water is unique again. Whereas, the Amex One commercial incinerators are, the last I knew, ware quoting about 60 cents a pound. MS. MOLLI Okay. That's all 1 have. MR. GUlMONht Okay. I have a few questions. HONS 216165 / - t i * 3 ! j I 1 68 1 If I understand your testimony today and soma of ths 3 consents that you mada previously, you'd prafar that v 3 withdraw this and than want out and said that inadvertent 4 PCS's ara not covered and, therefore, axclude-evsrything on S that basis. And, in the alternative, withdraw this and put a SO pon cut off. Am I correct? 7 Oh. BURGESSi Yes, I think w* feel that this proposal is -- you are takine your resources and ours awav from the real issue of what is the risk tn incidental 10 generated PCB's11 MR. GUIMOND: Okav. If I take that first option 13 or proposal of yours and withdraw this and propose that 13 Inadvertent PCB's not covered by TOSCA, how would vou 14 define Inadvertent PCB's? | DR. BURGESS* I haven't civen SDeeific thought to '8 19 1 that. We are having a oanel meeting tomorrow. 17 Basically, it i those PCB's generated -- maybe vou nou'd ue something 14he th FDA has in their constituent policv that have no function in the oroduct. ' They are not there fnr any preferred properties. They are * 21 unintentional, thev are unwanted in ceneral. a we would be glad to sit down and work out some a kind of -34 MR. nuIMOWIt would you envision something that was regardless of concentration or amount or something, 'ust HONS 216186 69 1 i as long as you don't want it, it would oualifv as uninten 2 tional? 9 DR. BURGESS: 1 think that thos* cases that wo are 4 aware of *ody, concentrations are extremely low. i know S | -- at ieat that t know of. * 8 I know of no situation where we're tel>in 7 percents of pcB's. Certainly, in that situation, if someone cones along and is going to sell a product that is 49 percent and call that inadvertent, I guess probably we ought to 10 define that out obviously. 11 MR. GUIMONDt Okay. That, 1 think, is one 12 potential problem, is a definitional issue. But, in any 13 event, -- okay. Then, if I -- your second alternative then, which - 1# was, okav, withdraw this, out a 50 ppm cut off or somehow put 10 one 50 ppn cut off, wherever possible, then on pace 6. in l i " vour ccements. I noted the last paraoranh on the pace, it says, "Due to unique circumstances of this proceeding. !" i " however, another course is open to EPA. CMA would be ! prepared to support an approach permitting PCB's and closed i " processes and in waste streams, disposed of under EPA's ! PCS disposal requirement, providing that there were a 23 regulatory cut off of at least 50 ppm for products in waste 24 I streams'. 2B Am I to interpret that, that you mean-that you -- HONS 216187 70 is this ths suns alternative ss your overall 50 ppm or is | i it s third alternative in which you would sty if vs wsrs to dsfins clossd systems so that it was dsfinsd st 50 ppm and products and vasts strsams qualified you, that that would bs accsptabls to you? DR. BURGESSt Basically, it's going back to ths original concspt which was discussad with with obvious misunderstanding. Ths currant nils, if you have In a recycle stream, a thousand parts per million, you must get an exemption for that process. It is still, I think -- as long as that's in a recycle stream, it's going around a pips, it never gets out. So, if the product and the waste are below 50, you don't have any real concern about what's in that recycle stream. He maintain no, that there isn't. So that we are still interested in getting that recycle stream or any at the bottom of a still or seme -- MR. GUXMDHDi X guess what I'm driving at here is that if you recall the proposal that now sits, it sits that it is non-quantifiable for effluents, emissions, for waste products, waste streams to be closed and controlled. And, then, for control, you can dispose of the waste, as you have said in there, and then you could qualify HONS 216188 71 A* I read this, if you're Saying that if we change j I the products and vasts* portions of that definition, to seats that SO ppm number for those two portions of the definition, that would be acceptable to you, is that correct? . DR. burgessi Basically, we are saying no, if the emissions are not above a regulated level, whether they are in product or waste of air or water; the emissions are not above a regulated level. That is what we have thought we were talking about a year and a half ago. MR. GUIMONOi So, waste streams, too, means more than just waste; it means effluents and emissions, too, is that correct? DR. BURGESS) I think we've got to look at the entire -- whatever is coming out, relative to exposure- MR. GUIMOND: Okay, okay. So, you mean SO ppm all over, this is net a subset, what I was trying to determine was whether you were talking about a subset product* and wastes as opposed to effluents and emissiona. You mean the whole shebang? DR. BURGESS; I'm talking about regulated quantities being dissipated into the environment or for exposure. . MR. GUIMOND) Okay. Thank you. I guess I've bean i a little bit puzzled and trying to answer the question HONS 216189 12 myself, * to this misunderstanding that we apparently have, and I went back and took a look at some of the historical information that we've had here relative to this, and if you take a look at the SPA submission to the Court back in January, I guess, of '*1, on page Id of that, there is a section that states, this was done subsequent to discussions we had with yourselves and a variety of other groups, "The industry representatives participating in the discussions of the EPA and EOF believe that most of the PCB's which are generated in concentrations below SO ppm are generated eithet in closed chemical processes in which the PCB's are both generated and destroyed, to the present levels of detectability, within the processesv without ever leaving I the closed process (closed system) or in chemical processes where any PCB's are removed from the process as wastes which are either incinerated or disposed of in EPA approved landfills or stored for such disposal (controlled wastes)." "Seme chemical processes may produce low I concentrations of PCB's which ars rslsased to tha environment, appear in the final product, ere disposed of in non-EPA approved landfills or by other disposal means (un controlled releasee). These are thought to account for a much smaller portion of the total generation of PCB's below SO ppm". You know, we have sort of been operating for a lonj HONS 16190 71 period of time that that's tht way ve thought the world was ' i and so, from ths standpoint of what I hear you saying in your cements today, you are indicating to ua that, in fact, tht reverse is true, very few processes would-qualify for closed or controlled wastes as presented there and most would fit into the other category, Is that correct?- DR. BURGESSt I think two things have happened; we have learned more about our processes. Also, the fact that you have been working on the analytical methods and we have been working on the analytical methods, and as Dr. Kaley has indicated, analytical methods move, I think we knew a lot more how to analyse for things today than we did IS months ago when we were talking. I don't know whether the limit of quantification has been cut by a factor of 2 or 10 or 100, but I'm sure that the limit of quantification has gone down and we certainly knew how to quantify at lower levels today. we also learned more about our processes and I guess, again, I think there is seme misunderstanding and I can't argue with what you have written because I haven't read it. But, there certainly was considerable emphasis that we could not go to a detectable limit. At the time one of the orders was to be written or was being written for the Court, the word detectable PCB showed up in the thing and HONS 16191 wa had a eonsidarable amtrgtncy in trying to get torn* of the CMA peopls to go talk with the people, your people who were writing that to try and get the word detectable out of the document* _ I think, aa I say, it'a an unfortunate case of miaunderstanding, ; MR. GUIMONDi Okay, if, again* I underatand your teatimony, you indicated that you felt very few -- this excluaion proposed rule would have very little good since few, if any, people would likely qualify. OR, BURGESSi If we understand the analytical, I think it is technology forcing, if we understand it correctly and that the limit of quantification is going to be considerably different. As Dr. Raley aaid, you give those analytical chemiats enough money and enough time and they can change the limit of quantification. MR. GUIMOND> Me have received aone comments, that same people have felt that they believe they may qualify for this. But, to pursue your question, then you did indicate further that the costs associetsd with the proposed rule would be very high. If few people ere going to qualify, obviously, then, would be no reason for them to try to take advantage of the exclusion, so thty would hava no coat in doing that, would thi HONS 216192 be corrtct? 75 DR. BURGESS: Excuse me. I don't - MR. GUIMOND: Well, if you don't think you qualify for this, you're not going to undertake an extensive monitoring program or theoretical assessment or whet have you, where you anticipate the coet are going to be, to show that you qualify for this. So, in fact, there would not be high costs if, as you indicate, that there would be few people qualifying for the thing. DR. BURGESS: Well, if you are ready to say we will file for an annual exemption, there is great incentive to not have to file for an annual exemption. Certainly, if you were ready to say we are just going to file for an annual exemption, then all your costs are involved with filing for the annual exemption. There again, we get back to what is Rule 3 going to say? And, I have no idea. I hope at some point we find out. MR. GUIMOND: So, you're saying a lot of people would really like to be able to qualify for this and so the costa would be involved in trying to see if they can possibly determine whether they could qualify? DR. BURGESSi An annual exemption for an on-going chemical process is very onerous. MR. GUIMONDI Okay. So, again, if I interpret MOMS 216193 76 that, you would indicate that a lot of people, because of the onerousnesa of the annual exemption and the uncertainty aaaociated with Rule 3 at thie stage, would very much like to try to qualify for this exclusion, if they feasibly could, and, therefore, would undertake a considerable monitoring effort to assure themselves or to try to qualify? DR. BURGESSt Right. Again, it's dependent on what happens with Rule 3. you eould get halfway through it and find Rule 3 took care of you. MR. GUIMOND! The -- this is a question for either you or Dr. Raley, I'm not quite sure which one of you would want to take it. Right now, you had indicated some problem associates with -- throughout your testimony, problems associated with the detection of the PCB's in difficult media, tar, solvents, what have you, and what I'm wondering right now is what techniques are people using to insure compliance with 50 ppm? Are they having any problems with detecting the SO ppm detectable everywhere? DR. RALEY: I can answar the last part of your question, and then I will let Ren address the second part, because I, frankly, don't know the answer. The true is that no, 50 parts per million are not detectable in some matrices. The submission we gave ERA HONS 216194 11 eoncarning the "Analysis of Chlorinated Biphenyls' showed an example of actual chromatograms where, I believe it was one hundred parts per million spike into gne of these tarry, crumby matrices that in essence showed ho ^detectable peaks above baseline for the PCB's. This was 100 parts per million total PCB's. I don't frankly know what the individual congener concentration might have been. There certainly are matrices where these problems exist, and what is done with these matrices, frankly, I don't know, after the numbers are reported. But, certainly, we still report numbers, nondetected, less than 100 parts per million. That type of thing. Okay. Let me -- our plea for a cut off is that let ue have our resources to go after those hard problems, eo that we can try different ways of cleaning up or separation or whatever for those difficult problems, so that we can hel our engineers or our environmental people make those decisions wisely. Kh. GUIMOND: Okay. Now, -- so, 50 ppm is not " you cannot detect it in certain difficult media at the moment? DR. BURGESS: t think that's true. DR. KAJLSYi That's true. You are basically dealing with waste streams when you are -~ MR. GUXMOND: Processes, too? . MON 5 216195 *8 DR. BURGESS I There may be some "I'm not familiar with thaw, but there may ba soma fairly complex j j procaaaas whara it's inside the aquipmant. Certainly, the 5< ppm cut off has baan technology forcing and continues to be. HR, GUIMONDt Than, but if z undarstand you corractly, you ara willing to accapt that? . DR. BURGESSi That'* what wm have propoaad, yes. MR. GUIMONDt In which case, you - what would be the net result, either now you would have to develop method ology to ba able to achieve that or,secondly, you would Jusi: wall, wa can't find it,it's not there, 50 ppm? DR. KALEY: Hall, there is another alternative in which you know, hopefully -- I don't want to speak for anybody, but at certain points, you would assume that thu PCS'a might ba at a higher level and treat that material as if they were PCB contaminated. Z don't know. i MR. GUIMONDi Okay. On some kind of a theoretical basis or something? ! DR. XALEYi Hall, if you would expect that PCB's : I might be in that type of matrix. i s MR. GUIMONDt Okay. I DR. KALEYi I mean, in general, we signals. If we cannot define whether that signal is or is not a PCB, then, you know, we have to almost assume that the PCB's could be in there, up to that limit of quantification. MQNS 216196 OR. BURGESS: I think I would like to just add to what Dr. Kaley has said. I think tha product streams and anythin? that is bain? distributed has been pretty well worked out. We maybe dealing in the question^of--50 ppm which is fairly complex waste streams, most of which are being handled according and to the best of my knowledge -- obviously they are all complying with the law, so they would be handled, whether they are 50 ppm or not. They would be handled as a PCS stream. Stuff in the process, for those people who have not -- who have filed for an exemption petition, they may have filed not knowing whether they were at 40 or 400 and still file for the exemption petition. Again, I can't apeak to that type of thing. But, as Dr. Kaley has indicated, that is an option that you have, rather than trying to pin down are you at 40 or 50 or 60. You could certainly file for the exemption petition, rather than -- because -- perhaps the likelihood is high that you are going to be there anyhow. HA. GUXMOND: Okay. You indicated in your testimony today that you did not believe that the proposal we had provided any additional public health protection. I guess my question is how would your proposed solutions provide additional public health protection than ours does now? HONS 216197 eo OR. BURGESSt I think our comment, at least what I meant to say, is that wa have not justified -- this proposal doesn't do any calculation as to whether the public health is protected to a greater extent or not. what we need to do is to do a risk assessment on the incidental manufactured material and what happens to it, what the exposures are, what the health effects from such environmental effects are, and determine what is an unreasonable risk. HR. GUZMOND* Okay. Also, in comments, written comments, you indicated that you did not believe anyone would take advantage of the theoretical analysis that we have made optionable. Everybody would go out and monitor. In fact, today, aren't a lot of people talcing advantage of the theoretical analysis as opposed to so ppm? Everybody going out and measuring to see if they are under SO ppm? It was my impression from talking to a lot ol industry people that a lot of people are saying 1 or SO ppm. So, in essence, they are doing that theoretical type of thing as opposed to anything more than that. I mean, as an example, your own data --an example of your own data from the survey that you sent to us, much of the data in there was not data reported from monitoring results, but was data reported from theoretical calculations that HONS 216198 SI people did indicating that they, in fact, fait that they j nay vary wall generate aoraa PCS'a, but thay did a calculation and it waa, you know, wall undar SO or whatever. I'm trying to aaa what -- - DR. BURGESSi Calculating aa to whathar you are meeting 50 ppm raquiramanta or whathar you are.maating a quantifiable limit, la two different thinge. X auapact alao that your definition of a theoratical calculation may bo diffarant than oura and wo havo not aaan youra, ao wa can't compara thorn. Aa to whathar or not anyone would taka advantage of it, ia certainly going to depend on what era the limit* of auch a theoretical calculation. MR. GUIMOKDt I would anticipate that the individual proceaoea would probably bo ao noted in your particular data you aubmitted to ua, ara in the bast poaition to run their own atychiometry and their own calculationo out -- DR. BURGESSi Whan you ara dealing with quantifiable level* in the part par million range, tychiomatry isn't of much help. You know, I can account for 990,000 pounds, 999,999 pounds and I'v* still got a hundred pounds, a hundred plus million out there. You can't even pump liquids with that kind of accuracy. MR. GUIMOND: But, wouldn't that give you the | MGNS 216199 aarne problem at 50 ppm than the other? OR. BURGESS; But, we have 4 tool then, we have the analytical chemiat hat said with some effort, wa can gat these analytical thinga worked down to where-wa can determine 50 part per million. MR. GUZMONDt okay. Then, to pursue juat a little bit more one point that Bill vaa talking about earlier, ia that you do make a very significant puah for quantifying aone thing, getting you a target, aoaie number to ahoot at here, hnd, do I hear that aa, you know, 50 or not, aome particular number other than a real embarraaaaMnt in trying to go and find a apecific number that wa can give you that ia going to give you not neceaeerily a 100 percent aaaurance that you're going to beat it all the time, but a target to ahoot at frequently, aince even 50 ppai doean't give you the 100 percent aaaurance? DR. BURGESS: From an analytical atandpoint, if you have -- if the analytical chemiet ia aaaigned the job of coming up with a method which la eeneitiva to aome limit, that ia vaatly different than being aaaigned the job of defining what ia quantifiable. MR. GUXMOND: You don't mind working hard to find the atuff, you juat like to make aure how hard you've got to look or exactly what you're got to look for, ia that right? DR. BURGESS: Well, X think if we look at the HONS 216200 e3 definition of significance, perhaps that's -- we hate to keep looking at something that is insignificant. Obviously, at som level, PCB's, like most other chemicals, the exposure is insignificant. -' Let's find out what is significant and what isn't. Let's find out what represents an unreasonable risk and aim our efforts at controlling those unreasonable risks. MR. GUIMONDt Okay. X think Z asked you this question during our lunch period, but I'll ask it again relative to this. you Just mentioned significance here, with respect to the type of things we're talking about# what are -- are there any significant exposures? OR. BURGSSSi Of the chemicals that X'm aware of, X don't know of any that I consider to be a significant exposure, to the PCS. Again, you've got -- in this situation, you have two separate risk factors involved. One is the exposure to PCB, does it exist in the toxicity and that typi of thing# but you also have the relative hazards of exposure to the primary chemical. And, you know, drinking chlorobenzene with or without PCB is not my choice regard less. MR. GUXMOKDt So, you're saying that the fact than PCB is in most of the stuff# doesn't change the hazard MONS 216201 S-i associated with on* iota and, therefore, forget the pcb's. DR. BURGESS; I'm aura that in moat of the , situation*, PCB* ara wall below 50 ppm. But, chloro- bentene, carbon tetrachloride, the chemical* that we're talking about, at 500 ppm, PCB does not represent a risk. Tha chemical -- wa'va got all kind* of "handling technique* for those chemical* and the chemical* are the problem, tha relative risk from the PCB is extremely small, de minimis, whatever you want to call those vary small numbers. MR. GUIMOND: Okay. Thank you very much. John? MR. SMITH: No one has really spoken that much about sampling here. My questions are related to sampling and since no on* has really spoken about them, you needn't answer now, but perhaps in the reply comments. EPA has proposed a sequential sampling schema which involves sample collection from randomly selected locations, at randomly selected times, within the nrocas* cycle. Could you describe the procedure you would use to select samples to represent a process? Include the maximum number of samples that you would select to represent a proeeas and why you selected this number, exactly how you would select the sites and times for collecting these samples, and the sample volume or waste you would collect for air, water and other matrices. DR. BURGESS: Could -- in late lata week, in tryinc MOHS 216202 85 to discuss the sampling documents that we received, there was obviously soms misunderstandings and questions as to what it really meant, and we decided we could not address it at this time because we just didn't have it resolved. Your question sounds rather complex to me, at least. We can certainly attempt to answer that-question: if we could have it in writing, it would be very helpful. MR, SMITH Fine. That's all I have. MR. GUNTER: Okay. Dave? MR. REDFORD: Before X attempt to verbalise my question, let me clarify an issue you brought up before, thJt in the Federal Register, it spoke about noise being determined just by using carrier gas. That that was a misprint, I would like to let you know. Neither myself nor MIR would agree with what was stated in the Federal Register. It has to be done with the matrix. We egree with you. Okay. In your testimony, you pointed out that aone of the MXR protocols are a little too loose and you spoke about clean up and you said it was a little loose and it was left up to the analysts, a lot of the things. later on, in your conclusion, you recommend that we develop an analytical method and leave the clean up up to the analysts. That, Z think, was your final recommendation, since it's media specific, metric specific. MONS 216203 t 86 1 I Could you clarify what you maan by thasa two 2 statements? Thay don't seem to agree with one another. 11 3 1 DR. XALY; Those two statamants are made in what wa ; 4 feel are different contexts. The first .statement is made in 1 the context that a non-quantifiabla limit in which the analytical chemist is asked to define at What limit his 7. chemical or his PCB's are going to be regulated. So that we feel that if we are given a series of options, we have no guarantee that soeteone in an enforcement to position won't make a different series of options that come up tt with s different quantifiable limit and, thus, we will be found 12 in violation. 13 The second recommendstion for leaving clean up - u mnd separation options open to the chemist is made in the - I 1 1 11 context of regulatory cut offs, at which point we know the target level that we must be able to measure our PCB's and in " that case, we can design our experiments in the way we find ' most effective to clean up our particular matrix and still be tc assured that somebody 1h using a different methodology will 2q jstill find us in compliance at that cut off. 2, I MR. BIOrORD: So, if you're -- you're stying that a mf wa usa -- stick with ths limit of quantification, you 22 want us to eome up with soma kind of a vary strict protocol j4 that would naad to ha followed. Do you think that sort of 2S thing is faasibla? -* MOHS 216204 e7 DR. KALEY: I think that for us to be surs of compliance, w* need ths sgsncy to com* up with vary strict protocol thst thsy will follow. MR. REDFORD: You yoursslf have stated that, in a diffarant matrix, though, that that's not possibla. DR. RALEY: Z agraa. That is tha basis of our concarn with this non-quantifiabl* approach, wa don't know whan to stop locking and va don't know whan anybody alsa Is going to stop looking. DR. BURGESSt Z think it's important to nota that the difference between enforcement and the analytical group is also going to be a difference between supplier and customer. And, it can lead to all kinds of problems if tha customer analyses his product and finds quantifiable pcb's. His first reaction is going to ba, it must have bean supplied to him from tha supplier, who said it did not have quantifiable PCB's. And, ha's going to go back and wich a little work he may well find it in tha supplier's company. You are going to have all kinds of customer/suppliar relationships. MR. BEDFORD: All right. Wa came out with a number that was 50 and we only cam* out with, say, an analytical method to clean up and wa left all that loose, wouldn't the saa* situation exist as far as whan ERA comas in to chack you? Zf they want in and used a different method, it would MOMS 216205 S3 b the same situation as if it was limit of quantification. Thsy could do vhatavar they fait and you could have dona something complataly diffarant. DR. KA1EY: I know. But, at that point, we're \ talking about axeaading a previously established limit, if wa do the bast job that wa have dona, we are confident that wa have dona a sufficient job to determine ihat the PCB's in any given particular sample are less than SO part* par million. Wa can be confidant that EFA or anyone else coming in, a supplier, competitor, whatever, will also find that the FOB'S in that particular sample are lass than 50 parts par million. Okay? In the case of non-quantifiables, wa may determine that our limit of quantification is 10 parts par million. Somebody else may coma in, do something diffarant with the sample, spend 6 months trying to lower their limit of quantification, and, say, ah, but, now, I've found that my limit of quantification is one part per million on those samples and look, you've got 5. Okay? So, we've got an order of magnitude different. I say Z lower my limit of quantification to 10, somebody else says, you're above aiy limit of quantification at 1 and they are both telling the truth. There is no disputing of facts. But, if the cut off is set at 50 or whatever number, then I think there would be agreement that both are MONS 216206 balow that point. 89 j MR. RCorOROi But, thay both could coma out j with diffarant numbara avan if tharawas a limit? DR. KALEYi I think our round robin axparianca, our individual laboratory axparianca, has shown that claarly two paopla analyzing tha sama samplas of tha typas wa'ra talking about, will gat diffarant nuabars. Tha analytical variability wa'ra talking about, I think, our rourld robin rasults hava shown variabilitias of plus or minus 70 pareant. But, wa naad -- avan with that, wa shoulr ba sura wa ara still balow 50 parts par million, basad on our analytical variability and what tha trua valua was, DR, BURGESS < Ms hava a targat which is "vary difficult to hit and if you start swinging it around and moving it all ovar tha plaea or don't tall us what tha targat is, it makas it avan mora difficult to hit. MR. REDrORDt Okay. All right. Thara has baan a lot of ecanants mada from yoursalf and othars in tha writtan comments wa'va racaivad talking about tha MR1 methoc not baing validatad. If -- I'm going to hava to go to two diffarant possibilitias. If wa aithar stuck with tha limit of quantitation or if wa cams out with a numbsr, what do you think would ba an accaptabla laval of validation work baforn you would say that tha method was validatad? _ How much MONS 21620? 90 effort would you think would be needed before we could j say that the method wee validated? I realise you're saying that we can't validate the clean up methods because we have to do.it for a particular matrix. What about just-the analytical, just the instrumental? What level do you think would be needa< just for that part? - DR. RALEY: z don't think -- MR. REDFORDi Where are you going to draw the line, and stop saying that EFA hasn't validate the method? DR, RALEYt okay. This 90ts back to our basic question and Z'm not sure. Z think it does matter whether it's a cut off or a non-quantifiable limit. Z think validation must include the total method run on an actual sample. And, Z know that you agree with that. At that point, Z think that the validation burden rests with the industry on their particular samples. Z think the ageney can give guidelines on what is suitable validation, so that we have some idea of what kind of statistics the ageney feels we need to develop for our samples. Okay, This gets back to the question, your very first question. In terms -- you know, we want the MONS 216208 91 iptiona for clean up and separation and measurtmont in terms 1 >f a cut off type regulation, and we, speaking for myself, 2] :ertainly in our laboratory and I know that at least some 2 nambars of the group were willing to accept thoie'burdens of 4 validation, we're going to have to do them anyway. Okay. If we talk in terms of a limit of quantitatio^ lystem, frankly, I don't have an answer to your question, 7 rhat's one of our concerns. Z don't see how we can validate 8 iveryona of our methods as a limit of quantitation for all the 8 aatrices given a specific protocol and I don't see how the 10 igency can either. 11 He both, as analytical chemists, we all have the 12 iae problems with this rule. 13 OR. BURGESSt Your ... analytical chemists, if they 14 go to enforcement, aren't going to have the same kind of 18 problem of whatever method you're using is going to hava to be 18 validated for that specific media matrix also. 17 MR. REDEORDt Sure, okay. Zf we did go to tha 18 established limit, to the number, do you think that the 18 guidelines, the analytical guidelines that MRZ had in their 20 documents, do you think they are acceptable to that sort of 21 a method? OR. KA1EY: Okay. Z can't apaak in total, but I think Z mada a statenant, I cartainly agraa with it, that I think MRZ people heve done a fine job as far as thay hava MOHS 216209 92 gone. They have certainly studied the field of anajysis of 4 PCB'a and whatever matrice* were available and literature references. I think they have identified soma option* which have worked in the paat. Our experience i* that soma of these option* do not work in the metric** we're talking about, but MRI has no way of knowing that beeauae they haven't tried them on thoae metric**. I think the MAI document* atanda a* a fin* starting point for any laboratory which need* to comply with the regulatory limit. However, 1 don't think that the documents should necessarily be considered complete by the ZFA. I think the industrial laboratories* who have more experience in these types of systems, should be allowed to modify and/or use alternative techniques to reach the regulatory cut off as long as thoae teohniques are documented and the quality of assurance and method performance criteria are documented in that industrial laboratory. In other words, Z don't think, because something is mentioned in the MU -- is not mentioned in the MAI documents, it should be excluded from useag* for methods to comply with the regulatory cut off. However, I think the KAI documents does serva as a very fine starting point for somebody that is going to have to comply with this cut off. HONS 216210 93 MR. REOFORPi AH right. In -- if EPA has to 1 aetabllah a sot Analytical procedure, aa you are recommending, analytically, not claan up, do you think we ahould *tay with tha GC MS or do you think va ahould laava that epw lika the way tha MRI documents ia, that you can uaa packad or capillary column, or -- I maan, if you want ua to narrow it down, wa hava to narrow it down. Juat how looaa ara you saying? OR. RALEYi Okay. I think ~ MR. REOFOROi I'm asking you to ba axplicit aa to what mathod you think EPA ahould follow. OR. RALEY: I think everything wa hava auppliad to tha agency, axpraaaaa our prafaranea for alaetronic impact, maaa apactromatry aa tha prafarrad enforcement or eomplianea mathod. Ha faal that in tha matrieaa wa'ra talking about, that ia, in many caaaa, tha only inatrumant which will allow you to datasiaina PCS'a at 50 parta par million, if you can do tham at all, at that point, wa would -- wa faal that tha agaacy ahould hava a rafaranca mathod which thay faal will work on tha aamplaa, abaant claan up procaduraa, ao that wa know what atandard wa ara going to ba judgad againat. a But, wa alao faal that tha agancy ought to allow tha a uaa of aquivalant mathoda if tha induatry, tha ragulatad 24 induetry, can davalop information ahowing tha equivalency a HONS 216211 of those methods to ths primary method. 94 In othsr words, ws don't particularly object to havin to analyze a set number of samples by gcms, but certain of the regulated laboratories, for certain of the products, can do an excellent job with a short packed column and an electron, captured detector. We don't feel that those groups, who now have validated methods and are doing that to meet the 50 part per million cut off, ought to be -- ought to have to go out and buy a mass spectrometer or send all of their samples to a contract laboratory. Does that answer your question? HR. BEDFORD: Yes. All right. In one of the comments that we had received, it wes suggested that we 4 acquire a specific minimum extraction recovery percentage 5 efficiency, and that was something like 90 percent. In whichever way we go, do you think that specifying 7 what we require to be the extraction efficiency is a good 9 idea, of the surrogates that MRI has prepared, that we require if someone to have a specific extraction efficiency? 30 DR. RALEY: I would hate to see it in terms of a 31 requirement. Certainly, we feel that guidelines are useful in 13 giving us a target, but there are matrices in which 60 to 70 23 percent recoveries must be considered good. 24 There are matrices in which plus or minus 50 percent IS relative standard deviations must be considered good. MQNS 216212 9; So, certainly, we don't went to be told that we have to do something which is going to be a virtual impossibility. 1 think one of the points we made, plus or minus 10 relative standard deviation on some of ehese-matrice*, we work with is an extremely exacting thing. As long as we know our analytical variability and know the projected value given our recovery and precision, then 1 feel we can operate effectively below a cut off. MR. BEDFORD: No matter which way EFA should go, what forms of QA and QC and what Intensity QA and QC do you think EFA should require? Such as interlaboratory studies, laboratory certification, the use of surrogates, blanks, dupes? OR. KALEYi Could we address that -- could1 I address that in the reply comments, in the group's reply comments? j MR. REOFOROi Sure, I think Z'm about questioned out. I MR. GUNTER: You're done? Okay. Denise? i I MS. KEXXTNER: I'd like to direct this to Dr. i s' Burgess. In your testimony here today, you contented that ! you doubt that any processes will qualify for this exclusion. Do you have any Information on what criteria under the definitions of closed and controlled waste HONS 216213 ?s processes will be least likely mat? In other words, ire releases in air moat likely to disqualify people from excluding or releaaea in water or the releasea in producta? OR. BURGESS: Again, it would be speculation on my part. 1 think aa the induatry haa become more familiar with theae thing*, there are unique situations that have come up, vacuum pump problems which would give some release in air. Very low levels, very low quantities, but fairly high concentrations would be detectable in air, because a vacuum pump doesn't move that much air. Hater, per ae, I don't -- I'm not sure. It would depend a great deal on which PCS you were talking of because the partition coefficient is such that decachlorobiphanyl is exceedingly insolube, has very low aolubility in water. So, I can't visualize vary much dacachloro getting in water if the water is in contact with some organic stream. But, there again, in some of the other materials that have some water aolubility, you could get problems. A year ago, we were looking monthly at the produet consideration. I think aa we have sioved through this, we have realised there are some eir end water concerna and the extent of them or the ratio of them, I don't know. MS. XSEITNER: One of the comments received in response to the proposal suggested that we define process. HONS 216214 97 release, product and waste more carefully in tha proposal so j that paopla know whara exactly the sampling should occur. j For example, do we mean affluents from soma type of water treatment, on-site water treatment, or~do .we mean water effluent directly from the process. Do you have any suggestions on how that should he defined or some recommendations, on thoss four? DR. BURGESSi I've debated some on that within our group. Is heat exchanger water included or not included? Is a water aspirator included or not included? And, I think we would have to sit down and work on those definitions in order to get anything specific. He can try tc address scsm of that in our reply comment. MS, XEEITNERt In your main comments, you state that the quantities of PCB's inadvertently produced in the United States today is miniscule. You use that information to a great degree to support your contention thst the risks proposed are de minimis, The data on the PCS levels are the amounts of PCB's produced as reported in the survey, were gathered by different analytical methods and by beet theoretical estimates by each CMA member firm that responded to the survey. The survey did not require actual analysis, did not specify analytical schemes as a standard and did not HONS 216215 i 96 1 | define sample six*, extraction procedure, the clean us 2 I procedure. Ze it possible that the data reported in the 9 | survey are inaccurate and the actual amounts of PCB's 4 present could vary by several orders of magnitude? ~ ~~ 9 DR. BURftESS: I don't think by orders of 8 magnitude. Are they inaccurate Certainly, they are ? estimatee and any estimate is subject to consideration, is i * a factor of two. Certainly, some people were reporting high, * some people were reporting low. I don't know what the inaccuracy figure is. Is it 50 percent, 100 percent, 500 percent? I don't think we're orders of magnitude and I think the work that you hava done it looking at the exemption petition, where we are dealing with materials over 50 ppm, would confirm that, that we are in the right ballpark, whether it's 13,600, I'm -- I like to use statistically significant numbers, even to say 13,600 makes me shudder a little bit, but that's what we need to say in order to get the things to add up. I i '* Now, somewhere -- I guess I'd say somewhere around I 15,000, plus or minus some number, and I worry about saying i 13,600, but it -- but, that is, I think, the limit of our 1a1 teohnology at the present time in saying that. MS, XEEITNERi Are you aware of the analytical w method that ERA compliance monitoring personnel are currently a using to enforce the 50 part per million cut off it has set RONS 216216 on manufacture? 99 | OR. BURGESS: I'm not, but I'm sura I can find j an analytical chemist who is. MS. KEEITNER .* Aren't you concerned about not being in compliance? DR. BURGESS: Am 1 concerned about not being in compliance? For three years, 1 have been concerned about not being in compliance. A great deal -- but, as far as -- if that is being derived from your other question about the analytical chemistry, irrespective of how concerned I was about being in compliance, X would not spend much time on the analytical method because X have analytical chemists who are much more profound at reading those things and doing it than X am. MS. KEEITNER: Thank you. That's all the questioni X have. MR. GUXMOND: Thank you, Denise. There are follow up questions. Xf there are any from the panel, I've got a couple myself X would like to ask, and then move on. On your written testimony, page 35 and then again on page 52, you note that all these processes that we're talking about -- I guess if we don't accept any of the other alternatives, all these processes should be considered as totally enclosed uses of PCB's as opposed to manufacturing, and I guess X would like to expand a little HONS 216217 i bit on that, as to why you believe this is a totall enclosed use. DR. Bl'RGESS : ,, Well, certainly, we have not 1 intended to manufacture, process or distribute- PCS's. This is not a commercial venture. It is something which the PC3 is an inadvertent thing that occurs in the use of the chemical itself. I think we can develop those arguments further, if you are interested, certainly I would like to get a legal counsel more vitally involved in it. MR. GUIMONDi Yes, I -- that was an interesting concept to me, too. Let me just expound on another question here that you were talking with Dave Redford. If I understood you correctly. Dr. Kaley, you indicated that if we specified a specific number, you felt that it would be up to individual companies to accept the burden for individual validation? DR. KAUY: Well, in the climate today of good laboratory practices and things like that, I think that is a perfectly reasonable expectation and that, certainly in our laboratory, I know of many others, feel that we cannot use methods for regulatory purposes or for any purposes at all unless they have been fully validated in our laboratory and we know at what level they are able to perform. MONS 216218 MR. CUIMONDi Okay. f OR. BURGESS: But, simultaneously, we would aspect i you to have your method validated. j DR. KALEYs Right. He would lika -- ____ MR. CUTMONDi But, if i hoard you correctly, not ! nacaaaarily to tha axtant that you would, you would not oxpact us to validate our mathod for each and ovary possi bility. DR. RALEYt if it's going to bo uaod for enforcement on that matrix, z think -- yoa, sir. Cortainly, it would nood to bo demonstrated that it gave reliable results by some statistically significant manner. MR. GUIMOND: Even with respect to enforcement action, but I gather from the standpoint of promulgation of the regulation, 4n the first place, you would not antiaipate that we would have to validate it for each and every possible -- DR. RALEYt He realise that's an impossibility. MR. GUIMOND1 Okay. DR. RALEY< That's one of our points with non- quantifiable being a problem. MR. GUXMONDt I see. Any other follow up questions? Bill? -- MR. GUNTERt I'd like to follow up on a point that Denise raised about the data in the CMA survey and it also relates to an exchange that you had with Rich about the HONS 216219 102 1 information that was given to tha Court over a year ago. Zn raading tha instruction* that wara mo ilad out with tha quastionnaira that wa* usad for that aurv*y, in dafining what a closed system was, you indicated "this is a special case and requires estimating pounds produced in manufacturing operations without detectable PCI concentrator in any output strain". Now, your discussion related to Rich's question, you indicated that you might not have communicated the 10 concept very well, but what you really had in mind was 11 something less than the 50 part per million regulatory cut 11 off or -- is that correct? 13 OR. BURGESS: Yes, it is. 14 HR. GUNTER: Do you think that there were error* II made in responding to the survey based on that same misII communication, with this kind of language in the instructions'! 17 OR. BURGESS: I think most of the people respondin' II to the survey had telephone communication which would hav It straightened that out. Is it possible there were errors? Yes, it *1 certainly is possible, but I think, again, this kind of a survey that we were doing, we were ballpark numbering. But, most people were basically looking at is it a 7CB by rule, 50 part per million. MR. GUNTER: Even with the language like without HONS 216220 i: j ii detectable PCS concentration in the instructions, do you i think that's what happened? OR. BURGESS: I certainly -- Z guess I -- it's 1 j I somewhat speculation as I try to remember backj but I think most of the people did call in and that we ware talkinj pretty much 50 part per million. MR. GUNTER: That's all Z have. " MR. GUiMONDt Any additional questions? MR. BEDFORD: Yes, Z have another one. This is alonq the same vein as one of my questions before. Zf you had a new product, a new matrix, could you explain what types of things you would do with that product to validate your analytical method to say -- to mak you comfortable now, say yes or no, this is what the concentration of PCB'a is? Could you just explain what kind of validation procedures you would 90 through? i DR. KALEYt well. Z can't -- z cannot speak for i the panel on this point, but let me just give an example of ! whet -- maybe how Z would start. Okay. 3D A sample comes in and the guy says Z want to know 21 if you've got 50 ppm PCB's in it. So, I look at it, you 23 know, pick a matrix. Well, say, it's pretty bad. if I 23 say well, first thing Z'm going to try to do is dilute it >4 and shoot it. Okay. So, we dilute it. okay. Let's say 28 MONS 216221 ; :-i that work*, w* can dilute it by a factor of 10 to 50 or something and it looks like the baseline is going to be clean enough, we can do PCB's. At that point, we establish the working range ve arc, going to need for standards to quantify SO parts per million or less PCB's in that system. We run a series of at least three standards, in replicates of 5, to determine our linear range, the reproduceability of our stnadard. We try to then spike -- first of all, hopefully, ve can have a blank sample. Somehow, either we've got generate one or we've got to hope that we can somehow -- one of the samples is truly clean, gets no PCB signal or a very small PCB signal under those circumstances. At that point, we spike the sample at three different levels in repliactes of 5 determinations each to determine our recovery and precision of the PCB's in that matrix at those various levels. And, then, that presumably i would give us at least a baseline validation. I Then, from that point on, on-going checks of standard responses and spiked samples and replieata samples nut and compared against those previously established \ performance criteria in either a controlled chart system or at least same sort of documented system to tell when we were exceeding our specified limit of precision and accuracy. Basically, we're talking about three level* in MOMS 216222 1 OS- 111 replicates of 5 to establish our statistics in our laboratory. Tha othar members may agree or disagree. MR. REDFORD: Okay. Thank you. MR. GUIMOND no mora questions? Thank, you vary much, Dr. Burgess, Dr. Relay. Axa thara any questions from tha floor that va didn't gat? 1 don't think va got any yat. (No response) ' Okay. Thank you vary much. At thia tima, rathar than starting on NRDC, although va hava fini shad up a littla aarliar than Z had anticipatad, I think va vill braak for lunch and plan on coming back -- I think va'll plan on coming back at 1 o'clock rathar than 1:30 and gat on a bit aarly. (Whereupon, at 11:40 o'clock a.m., tha hearing was racessed, to reconvene this same day, Monday, July 26, 1982, at 1:00 p.m.) HONS 216223 n: 1 AFTERNOON SESSION 3 1:10 P.M. 3 MR. GUIMOND: Tha first testimony this afternoon 4 will be by the Environmental Defense Fund and- NRDC, Jackie. 9 STATEMENT OF JACQUELINE WARREN, t ELLEN SILBERGELD, ENVIRONMENTAL DEFENSE FUND, NRDC e MS. WARREN: Thank you, we don't have extensive 7 testimony prepared for this afternoon. MR. GUIMOND: Would you identify yourself, please? MS. WARREN: I'm sorry. Jacqueline Warren with 10 Natural Resources Defense Council and Ellen silbargeld 11 who is a acientiat with Environmental Defense Fund. 13 I wrote in the request to testify at the haaring 1) that we would address five particular subjects. I really 14 think we are only qoinq to speak to four of them because IS with the proposed guidelines, there seems still sons confusion It with -- 17 VOICE: Louder, please. It MS. WARREN: The change to now saying that packed ft column, GS Mass Spec is acceptable as capillary wise. We 30 have been talking about, I received a very small guidelines 31 document in the suil and I didn't aaa any particular 33 difficulties with it. But I am not going to spaak to it S this aftamoon. 34 Tha firat thing, in our taatimony wa startad out at by saying that wa didn't think tha LOQ was an. appropriata EXECUTIVE COURT REPORTERS (m) mmn MOMS 216224 113 point to differentiate between PCS, which would b. in category for further consideration in the next rulemaking, as opposed to PCBs that wouldn't be. And after considering the various problems, many of which were discussed this morning end which tieve been in the written testimony, we reelly have come to the point to agree that it is e moving terget end it would be better to specify e number. Z have never been opposed to a number in the | first piece. The objections thet we hed with the 50 parts per million limitation hed everything to do with the apparently very arbitrary choice of it, the applicability Iof it to all PCBs regardless of the medium and the pathway to the environment and any consideration of what the health effects would be. 10 What we would really like to see is the agency 21 starting from the point that the statute was Intended to a address, which is the prevention of the addition of PCBs a to the environment. The court opinion spoke about reaching M point sources of PCBs and we think that you can do that by a setting a nwfeer which is in the range of what the limit of detection would be. __ EXECUTOR COURT REPORTERS Since the statute(*ls)aMyMs Wno one may manufature any PCBs, obviously you are going to be concernMeOdMS abo2u1t62th2e5 presence of any PCBs. Once you find then, you have another 1 problem tnd that, obviously, la whether you are going to do 2 anything about that or do Something about soma aapacts of 2 thoat PCBa. 4 But tha first question you hava to ask la ara < thara any, which ia why va said start with tha LOO. But if It la aaslar for paopla to look for, in tha product and , 7 waata stream, a number which ia approximately in tha range of where tha LOO would be, I think that might clarify soma of tha difficulties. to X realise tha range can differ, but it doesn't 11 differ by orders of magnitude in what you are looking for, 12 presence. He would like the number to be at the low end of 13 the spectrum. Obviously, because we are concerned about the 14 health effects. Sven though there has been a lot Introduced It in this record purporting to show that PCBa are net hazardous, It since they are persistent bio*accumulating compound!, tha 17 question la not what risk does thla molecule of PCS pose, It but what risk Is going to ba posed to a parson at the point It of actual axposura whara that PC* is no longer ona molacula. 20 Raving said that, we do stand behind the position 21 we took in the written testimony, with the exception of \ 22 saying, rathar than specifying LOD, wa would rather see you 22 earns up with a number that everybody can than attempt to 24 l gear their analytical procedures to finding. I X think you would be heading in that direction EXBCUTIVB COUSr REPORTERS (m) iiijwi HONS 216226 115 4t goma point anyway bacauaa of tha variability of what LOD or IOQ maana to tha varloua haarara and It waa vary claar to ma hara that, ainca CMA'a aurvay waa aaklng paopla for no datactabla PCBa and paopla wara intarprating that aa SO parta par million, tha opportunitiaa for mlatakaa, miaundar- atandinga and variability ara graat unlaaa you do aat a number. So z think you ahould. With raapact to tha damonatratlon of allglblllty for an axcaptlon, I hava a graat daal of eoncarn about tha approach that la takan In tha propoaal bacauaa, flrat of all, thla la an axcaptlon from a atatutory prohibition. Wharaver you draw tha lina, you ara atlll 90109 to ba -- If you ara goln9 to ba accaptlag paopla aa da minimia riek, paopla who qualify for aaying that thalr ralaaaaa ara da minlmla, tha aganey, I think, haa to undaratand tha baala for that claim of allglblllty and If a manufacturar aimply announcaa to himaalf that ha la aligibla and filaa tha papar aignad on tha bottom lina in tha drawar, you don't know about it. Ilia full burdan la on tha aganey to find out who thosa paopla may ba, which out of tha group daaarvaa tha ao axpaaditura of your vary limitad, and avary lncraaaing 21 limitad, anforcamant raaourcaa, 23 I think It would ba much mora appropriata for tha 22 paopla who ara claiming allglblllty to coma forward and show 24 you thay ara aligibla and how thay got thara. And v will 21 EXECUTIVE COURT REPORTERS (Ml) HIM HONS 216227 . i. a be speaking in a sonant to the question of the appropriateness I of tha thaoratieal calculation to get thara. But I think at tha vary minimum tha agancy should hava tha information praaantad to it as to who .those paepla ara who claim to ba allgibla and what that elain is basad on. Thara wasn't any dascription of what anforcamant strategy tha aganey plans to follow. Z don't knew whether you are avar going to publiciia what that anforcamant strategy is, but I find it practically an unenforceable proposal ths way it is worded now. Not only that, but tha statute does allow for citizen enforcement actions and that kind of proposal where tha relevant information remains in tha file of the company means that tha public is completely excluded from access to that information, which, ware it in your files, would ba available under the Freedom of Information Act. It complatal> defeats the ability of anyone to enforce -- you or a member of the public, who I believe is entitled to know that somebody is still producing PCSs, even though tha statutory intent was that that be, either entirely eliminated or 10 21 only permitted whan EPA was aware of it and could make a 23 finding as to no unreasonable risk. 23 Z don't think tha no unreasonable risk finding 24 is applicable to da minimis exceptions from tha law. if 21 you read tha court's opinion, they weren't talking about EXECUTIVE couxt mpoateas (Ml) MMM HONS 216226 127 balancing risks and benefits* having found that tha benefit la there aa a natter of law. You would have to find that the regulation waa trivial or of no value* but it ia not strictly an unreaaonable risk test* although the unreasonable^ risk test certainly comes into play with the exemption process. I fully understand the difficulty of an annual application from thousands of producers and 1 aa not advocating that you get involved in that. Z think that a de minimis risk approach is an appropriate way to be proceeding. But Z do have a lot of problem with starting from 50 parts per million as has been advocated. The 50 parts per million level* as far as I can see* has no magic about it. Zt was an arbitrary choice to begin with. Zt was never justified by an rationale that 1 could understand and it is being promoted here as if it has some rationale inherent in it* which Z don't believe I to be the case. A previous witness testified that the fact that 50 parte per million was drawn as the lisiitatlon has been a technology forcing -- a fact of life and many companies hsvs changed their processes and fine-tuned their analytical procedures to be able to comply with that. 9 If that is the ease* it is within the intent of the statute for you to be forcing further action on it. By analogy to tha axampla of dioxin of 2,4,5T where dioxin EXECUTIVE COURT REPORTERS (Ml) JU-M* HONS 216229 ;ia I is in inadvertent eontiminint in the minuficturi of 3,4,5T 3 because of the concern about the toxicological properties 3 of dioxin, companies have learned to look for it at lover 4 levels than they ever thought they would be able to and 5 manufacturing processes have been changed in order to minimise the presence of the contaminant. And I think we 7 should be looking for the same result with this. For that reason, if and when you do set a number for this, I think you should be bearing in mind that it is 10 desirable for companies to be changing their processes II in order to minimise the presence of the contaminant. Considering how many processes there are and how shaky the data are on the number of FCBs that are actually being genera ted and being released into the environment, I think it behooves all of us to be working toward a reduction in the production of that quantity. I also think it would be appropriate for you no have a mnber of different limits so that if you are concerned about the release of FCBs into the air, it would be appropri ao ate to have one number for that and the releaae of FCBs into at the water, another number for that, if the ability of the a analytical methodolgles to get down to dlffereit numbers a depending on the medium is ss variable sa it is. And the M hardest part is dealing with the complicated matrices such as a EXECUTIVE COURT REPORTERS (mi) mam HONS 216230 tarry matrices that paopla war* talking about. But ona number, I think, has inharant difficulties with it and you i j ought to ba looking at variatias so that thay ara madia spaciflc and always kaap to tha haalth affacts that you found as a mattar of findings of fact in tha pravious procaading and in tha affluant discharga procaading that tha TDA has also found and which ramain tha currant paar reviewed and unrebutted findings of fact with raspact to tha haalth affacta of PCBs. Elian will apeak to tha quastlon of thaoratical analysas. MS. SILBERGELD: Ha undarstaad that soma burdan has baaa placad on tha usa of thaoratical rathar than actual analysas of matarials in ordar to datarmina axamptions, Ha hava soma concarn about this as a procsss to ba raliad upon, Tor tha most part in anvlronmantal and other typas of ragulation, thaoratical analysas ara usually invokad whan it is difficult, if not impossible, to do empirical analysis. For example, in air pollution whan calculations based on models of dispersion, entrainment and othar phenomena ara obviously appropriate because tha cost of doing adequata empirical studies would ba enormous and of dubious reliability. In this instance wa ara not convinced, however, that tha barriers to actual ampirical chamical analysis of EXECUTIVE COURT REPORTERS (Ml) M140M HONS 216231 120 product* process, streams or waste streams are so great. 2 the barriers to doing this are so great as to eneouraga 3 theoretical analysis. 4 We are, moreover, concerned that thaagency and 5 others be satisfied of the errors inherent in theoretical analysis ao as to determina, really, the bounds to which T these types of analyses can be used in regulation. Tor example, I would assume that the way thasa S analyses are going to be conducted will be to set up a 10 chemical engineering flow diagram of the various reactions II that are thought to occur in producing a certain product and, therefore, demonstrate how many molecules of PCBs 13 would be produced per molecule of product or other such 4 factors. 14 However, in most chemical engineering streams, II chemical processing streams, one is usually dealing with 10 IT great excesses of precursors or catalysts because yields 10 are considerably below 100 percent, even under optimum 10 conditions. Small changes in the conditions under which those 30 processes occur, such as temperature, aerobic or anaerobic 31 conditions and the like, can, of course, drive chemical 23 13 reactions in differant ways by considerable factors. Therefore, we would have to obtain a vary elaar 34 understanding of what tha likaly ranga of arror in tha final a EXECUTIVE COURT REPORTERS (Ml) IU4M4 moms 216232 i 21 calculation baaad on thaoratical analyaia of PCBa. indeed, if ona wara to do such analyaea thoroughly and carafully, va wondar if much affort or time would ba savad ovar doing actual ampirical analyaaa, and, indeed,, whether the raliability would ba auch that both companiea and the.public would ba aatiafiad by that approach. - MS. WARREN j I wanted to apeak to ona othar iasua. In our writtan commanta for July Ith wa addraaaad a number of atatamanta that EPA made with raapaet to tha riaka of inadvartantly manufacturad PCBa. On paga 11 of tha teatimony I quotad ona atatement in tha Praaabla about axpoaura lavala during routina oparatlona ara not expected to raault in aignifleant axpoaura, praaumably to workara, baeauaa of maaauraa alraady inatitutad in tha induatry to raduca expoaurea.- With raapaet to that, I wantad to call your attention to a document that NIOSH publiahed in January of 1981 where they ware evaluating tha permeation of protective garment material by liquid halogenated athanaa and KM. With raapaet to PCBa, with tha exception of ona material, tha PCBa permeated the protective material in under three minutea and moat of them in laaa than a minute. So there ia a real quaation about how affective thoa protective materiala really ara. I will aubmit that EXECUTIVE COURT REPORTERS (XII) MOMS 216233 122 for the record with the reply comment*, but i wnt*d you to be aware of it. W* have hoard a lot of testimony, representations about how protective thoaa protaetiva. garments actually ara and hava not haard a lot from tha unions on it. I asauma and understand that soma taatimony is going to ba submitted by them, but I wanted to gat that into tha record because that's a vary significant difference from what we have haard if it is in faet true that PCBs coma through in lass than a minuta through most of tha matarials. That's all. MR. GUIMOND: Thank you, Jackia, Elian. . Why don't wa start tha quastioninc at my far right. Danise. MS. KEEHNEAt In your taatimony you spoka of limited detection. What do you believe is limited detection for PCBs? MS. WARRENt I think it depends on the medium you are looking at. ms. XEEKNBki in products. MS. WARREN: in products? I think what wa've heard today is that it depends on the particular matrix that it's in. But I think a range of what the likely PCB concentrations are is something that could ba identified. EXECUTIVE COURT REPORTERS (Ml) NMNi HONS 216234 1 I'm not in a position to aay what that limitad 2 dataction would ba. 3 MS. KZEHNERj If EPA war# to withdraw thia 4 proposal and not procaad with tha third rulamaking^and lat i tha TSCA Saetion 6~E ban 90 into affact, what typa of analytical taathod would you propoaa that EPA would uaa to 1 anforca tha total ban? MS. WARREN 1 I don't think it's raaliatic for tha agancy to raally conaidar ttat aa an altarnativa, aftar 10 avarything that haa baan haard about tha nunbar of it procassas. Otharwiaa you juat wouldn't ba anforcing it. 12 I can't imagina that tha agancy la going to taka a atap ,, that would ba aa aeonomlcally diaruptiva aa that. I 14 MS. KEEBNER1 That'a all tha quaationa 1 hava. 10 MR. GUIMONDi Dava. 10 MR. REOrOROt I guaaa I hava, raally, ona quaation. IT If wa vara to pick a nunbar naar tha dataction limit or 10 whatavar, do-you think wa ahould juat uaa dataction limit It aa aomathlng to find it, to dafina that numbar? 20 Lika you ara aaying that it would ba a good idaa I 21 to hava dlffarant nunbara for diffarant madia, air, watar. 22 ate. What do you think wa ahould uaa to datarmina that 22 numbar? 24 MS. SIUERGELOt Hall, obvioualy, if you go to aatting a numbar, and I think for many raaaona wa andoraa 21 " J r Executive court reporters (MI) MI-MM HONS 216235 12-1 that as a procadura rathar than th# rathar murky concepts that ara baaad on currant tachnologia# which may or may not raprasant tha bast availabla technologies, or currant average practices in analytic laboratorios which' may or may not raprasant tha best possible practices using a'specific technology, once you go to a number several things inform tha choice of that number. One is, obviously, tha ability to detect tha substance. But tha other, and which most be an overriding factor, ara knowledge and estimates of tha adverse health and environmental effects of any given amount or presence of the compound. It may well be, as in the case of dioxins and other substances, that there may be reason to believe that in some instances the levels of a toxic substance known or suspected to have adverse health effects may, indeed, be lower than a generally established level of detection currently available. in those instancee, then, the health concern is overriding and really drives the analytic technology to develop methods sensitive enough to reflect that concern. That certainly has been the case with dioxins and probably is also the case with the dibensofurans. MB. REDFORD: Thank you. MR. SMITH: I have no questions. EXECUTIVE COURT REPORTERS (Wt) Ml-MM ROMS 21623b 125 MR. GUIMOND: Okay. I hava just a coupla. In your written testimony you indicate that you don't think the control wut should be excluded. Could you expand on that a little bit more? - I MS. WARREN: Yes, th* major thrust of the ssction 6-E thAt dAls with manufacturing is thst manufacturing of PCBs was to And. Zf you are going to Allow unlimited KiAnufActurA of PCBs as long as thA wssts is going to soma 10 sort of controllsd disposal source, than I think you 11 ara just ignoring ths initial statutory requirement. I am not persuaded that any of ths disposal tAchnologias availabla ara all that Adequate in daaling I with tha PCBs. Cartainly landfilling and things lika that ara not and thara ara a lot of questions about ineinaration, about whathar it is baing dona right and what tha by* products and combustion products of it ara. Just to allow any amount of PCBs to go out as long ss it is what you ara eslling controllad wssta saams 20 I to ma to ba gatting around tha basic point of Saetion 6E. *1 MR. GUXMOND: You would not saa, than, sny a a 24 advantaga to tha anvironmant or public haslth from, parhsps, a tha incantiva that might crasta. Thst is somaona thst altamataly might haEvXaECwUhTaIVtEyoCuOUwRoTurldepcoratlelrsuncotrollad PCBs, but could conoeivably, via th(Mal) ItUa4sHtI of an axamption, not hava an unraasonabla risk, that thara would MbaOMaSn-in2c1a6n2t3iv7a 125 1 for thorn to try to qualify for on exclusion by taking their 2 wstea and getting rid of then in thia manner as opposed 3 to some other way and thus actually presenting leas risk 4 than might otherwise be done if you handled --" 1 NS. WAARBNt There are a lot of ifa and maybes in that and it is clear to me that the final rule haa any 7 guarantee whatsoever that the waste is going to have to go a to an appropriate facility. There was a question raised this morning about, 10 what if it is a RCXA facility as opposed to a PCS facility. 11 Nall, RCRA leaves the control of PCB disposal to TSCA so 12 I don't know what sort of RCRA requirements would apply 13 that would be appropriate for PCBs. 14 With that uncertainty, I don't like the idea II that the quantity of PCBs could be increasing in the waste II or in a process and that that would be all right as long 17 las it goes to the right facility. t X would rather have a cap on the amount, II tut. GUIMOWDi Okay, let me see if I can understand 20 your proposal. You would suggest that we go with a closed 21 system exclusion, with the definition, basically, being that !a we would establish some kind of numerical value, probably a near the general level of detector limit -- limit of 24 detection that might be applicable to these things for air, 21 water, products, waste, is that correct? EXECUTIVE COURT REPORTERS (W) HS-Ntt HONS 216230 12? MS. SILBERCEOr Not neceaaarily, if I didn't make 3 that clear. whan I said if you 90 toward* lotting a specific 3 number, than tha ovarriding concarn mult ba tha evaluation 4 of tha actual ptential and advarsa affacta to human ha#1th and environment. You cannot than aoley raly upon level of detection or level of quantification aa tha criterion for 7 aatting a number. Xnaofar aa what you atata about varying level for varying aourcea of fCB'a, that may wall ba appropriate. MR. CUIMONOi obey. Aa you know, there probably 10 i* aema diaaanaion among a variety of people relative to how 11 aerloua tha health effect* ar*. environmental effeota >aaabeiatad 13 with PCB'a. 13 Z guaaa I am trying to tranalata if I wa* to 14 practically try to taka your reoonmendationa today and translate II thoae into .aooa kind of regulatory requirement*, how would I II go about changing what I currently have in tha rule aa 17 propoaad. II And if I tranalata them right, one thing Z would do 11 ia that Z would either eliminate the control waate 30 excluaion. or. if Z waa going to include it, z would have to 31 include it with acme kind of a total quantity amount and. 33 X gather, and/or a concentration value for it, too. a And then for the cloee waate excluaion, I would 34 put in a numerical limit for air, water, product, waate and a Z would aelect that limit by looking at aooetltig around tha EXECUTIVE COURT REPORTERS (mi) mj-oo*. HONS 216239 123 1 level of detection, assuming that I convince mysslf that 2 at that laval I an not going to hava sarious haalth affects. 3 An I correct, is that what you ara advocating? 4 MS. WARRENt Z think if you go baek~ to the $ statutory statement that no one should be doing any, you are ( going to want to be somewhere near the limit of detection 7 in setting the number. That doesn't mean that you are then going to be regulating and imposing a ban on what you measured. But it means that you have now differentiated the category for purposes of this rulemaking that you are not going to worry about any more or that you are going to worry about further. Tor purposes of making that' kind of differentiation I think the number would have to be eomewhere around the limit of detection. MR. GUXMONDi As it exists today presently? MS. warren t I think you can updata it as appropriate. I don't think the limits of detection change that frequently. 21 Or you could leave it as it is. There are other a regulatory procedures where numbers are set. They tend to 22 get engraved in stone, even though the ability to deal with 24 them better. IS MR. CUIMONDi Well, let me use as an example, if- EXECUTIVi COURT REPORTERS (M) JU4M* HONS 216240 119 and I an just using this as an example. If, for axanpla, a Unit of dataction in products -- va looked at a bunch of products and wa found that it lookad lika it was 50 parts par million for claarly a majority of tha products, 75 parcant or more, and wa did tha sana thing for wastas. Ha lookad at- air and watar and found it was, mayba, a factor of 100 lass for air and watar, mayba avan mora. You could liva with that kind of approach? MS* WAMUENi Z would lika to know vhara tha products ara going, what routs into tha snvironmant ara thosa PCBs raally taking. MR.- GOXMONDi So you ara saying, not only limit of dataction, but limit of dataction with scan kind of risk assassmant, if you will, that assures us that thasa ara not interacting in tha snvironmant diractly or exposing humans directly MS. WARRENt Aid what quantity tha product is being produced every year. MR. GCXMONDi Do you sea an ovsrall quantity limit astablished to this whola thing, too? MS. WARRENi I think so because, first of all, just drawing the line at any one number doesn't taka into account the kind of concerns that we ware expressing all tha way along with tha 50 parts par million cutoff- EXECUTIVE COURT REPORTERS (Ml) I4JMM MONS 216241 : 1 If tha route into tha environment ia getting out 2 into tha water, gatting into fish which ia tha major route 2 into tha diet, tha numbar should ba a lot lowar than 50. e it ought to ba a numbar that corraaponds to .what we know 9 about PCBs and how much of the percentage of diat is really a involvad with tha aating of fish, that kind of thing. 7 you can do tha sama thing with exposure relaasas into tha air, but to do that you hava to know whara tha I* product is going or how much of it ia coning out, not only in tarns of parts par million or parts par billion, but ovar n tha course of a yaar what ia tha total quantity aa wall. t2 I'm just saying that it ia a difficult problam is bscausa you don't want to just say tha sama thing goes for ia any PCI aa long aa it ia below a certain level. I have tha H same difficulty with any level that you could coma up with, it That'a why Z auggastad that you hava differant onas that are gaared mora to the medium that you are concerned about. Z am lesa concerned about PCBs found within a matrix where they are not vary likely to gat out and, 20 depending on what the use of tha product is, or if it is 21 in the waystream and the waste is going to a controlled 22 disposal, tha fact that the number is higher than a number | that would be going out into the water or going into the 24 air doesn't bother me so much. IS MB. CUIMONDi obviously, wa can't havs an infinite EXECUTIVE COURT REPORTERS (Ml) HUN) HONS 216242 131 1 number of limits, vs have to have sons finite numbs: of 2 limits. 2 HS. HiRRENs Kow about one for air, one for watit a for starters. _' MR. GUIMONDt Okay, one for products, all products -- - 7 MS. WARRZNt Nell, there are different kinds of -- veil, if you can categorise the products into separate, deecret subsets so that you vouldn't have too many of tham. o MR. GUIMONDt same thing for vasts, perhaps, depending on vhat happens to the vesta. 2 MS. WARRZNt Z think probably the vasts you don't 3 need to differentiate the same vay. t MR. GUIMONDt X'd like to talk just a little about ! your points relative to eligibility, hov somebody determines t eligibility, hov they should certify or self-certify. 7 Again, if I interpret vhay you are saying, you I- believe that vhatever -- vhether this vas done on an analytical basis or vhether it vas done on a theoretical basis, that the agency should require reporting of that information, is that correct? MS. WARRZNt Initially, yes. n MR. GUIMONDt For example, in the proposed rule 24 ve indicsta that one must self-certify each time your IS self - certification becomes no longer -- . EXECUTIVE COURT REPORTERS (mi) >u-aou MONS 216243 MS. WARREN: It'* * qualitative difference. MR. GUIMONO: Right. MS. WARREN: But they're not tailing you that, thay'ra tailing thamaalvai that. That's the"problem that I have with it, I think EPA should know. MR. GUXMONDt So you would say wa should hsva then cartify to us each time that occurs. MS. WARRENi Yas. MR. GUIMOND: And -- MS. WARREN: That's an altarnativa to an annual exemption petition so, unless their process changes more frequently than that, I would imagine it is not going to be as often as annually the first time and then whenever the process changes qualitatively enough that the quantity of PCBs that's being generated changes. MR. GUIMOND: I guess somewhat on a philosophical plane telativa to this, if you crests this thing as an exclusion, it is supposedly an exclusion, as you pointed out, from the court's language, is because there is da 20 minimis risk or little benefit associated with regulation of the thing. Then, via that same thing, you are then a saying that, basically, these things I am excluding because a Z am not concerned about them, I don't have that ouch 34 Interest in them. So, philosophically, wouldn't you also think that EXECUTIVE COURT REPORTERS (Ml) M)4Ml MONS 216244 133 you would not want to create a major effort on these people'a part or our part, for axampla, to determine that you, in fact ara eligible for something that the agency doesn't care that much because of the health risk.- - MS. warreni Hell I just wanted you to know that you don't care. I don't trust them because they don't think it's hazardous. So I think you can get a lot of theoretical calculations that are done by just putting your hands over your eyes and theorizing about it. Where the msabers could be quite high and, unless you happened to discover them through enforcement which X also don't think is that likely, you are not going to know about it. MR. GtJIMONDi Hell, as far as enforcement and everything, do you have any suggestions for a strategy? For example, when we approached -- when you take a look at some of this, we'v# got the information from the exemption petitions that we have and, as we note in the proposed rule, we utilize that as one effort in developing 30 an enforcement strategy. 31 You would see this other information as sort of, a in addition to that, it would help us focus on people that a may have, in fact, not even provided an exemption petition? 34 MS. WARRENi I don't understand the second part. a MR. GUIMONDi z am trying to get at what extra EXECUTIVE COURT REPORTERS (Ml) 1U4M4 HONS 216245 134 piaces of information you think you might got from all of these people. Potentially we could have a lot of paper coming in here, right, if we asked for a lot of the people who qualify -- they actually qualify for this xclusion how they did it. w# could have a lot of paper in here and I'm generally somewhat concerned about having a lot of paper coming to us from reports that stay or may not be useful for us in doing somelthing. Then there are the burdens associated with having that paper created and, of course, the burdens of us doing something with it, too. I'm not 100 percent clear -- ms. WARREM* How many pieces of paper are we | talking about? MR. GUIMOND* Well, if you made everybody -- MS, WARREN< you already have the group that was I producing inadvertently about 50 parts per million. If you confined yourself to the group that was balow that -- MR. GUIMONDt I don't know how many piaces of papar we may have. I have had astimates coning in from various people ranging from e few hundred to thousands. EXECUTIVE COURT REPORTERS (MS) US4M4 MONS /vh l . 135 HS. WARREN; The alternative is to say w just trust 2 the industry to do it and given thair concarns or lack of 3 concarns about PCBs, it doasn't saan to ma that it is reason- 4 abla to just assume that it's going to be adequately done. t HR. guxmond: Okay, so this is basically a safeguard You feel it would put more pressure on than to do -it right and 7 to do it in the first place, is that correct? HS. WARRN: Yes. HR. GUIHOND: Thank you. I don't have any more 10 questions. Alan? 11 HR. CARPXENi X don't have any questions. 12 HR. GUNTERi I'd like to follow up on Rich's last 13 one. Xn your comments you indicated that ERA should than. 14 after these pieces of paper eome in, do something to verify II the information. Reading your description of what EPA should 11 do, X got the idea that we would be creating another kind of 17 exemption procedure, albeit on a one tine basis. It What did you really have in mind here that we should 11 I do with the papers when they come in? ! 20 HE. WARRENi X thought that you would look through 2t them and then see which out of the group of people that are 23 inadvertently producing PCBs look like candidates for enforce ' a ament. The likeliest ones to be producing levele that oughtn't 24 be excluded. At least it will tell you who they are. you don't even really know who they are. EXECUTIVE COURT REPORTERS (M)) 14MM4 HONS 216247 13 6 t MR. GUNTER: So the verification you really had in 3 nind was really enforcement of the rule? 3 MS. WARRENi Yea. I don't see how you are going to 4 nforce this by -- with no information about what processes 9 and what companies, except for the ones -- you know about the ones over fifty who have already applied to you for exemptions 7 You have no idea really who the ones under fifty are. ~ MR. GUNTERi with regard to the establishment of a S series of cutoffs for varioue media, you have suggested that to these should be in the neighborhood of ths LOO. Wouldn't thes^ It numbers of necessity have to be around the LOQs, since other t2 wise you wouldn't be able to make quantatative measurements, t3 it would be below what you would be able to quantify. 14 MS. WARRENi My understanding is that there are ways tl to figure out how to quantify once you have detected them. to That might not be correot. but Z understood that it's possible to develop a procedure so that you could measure vhst you 17 to really have. But it's sore than just measuring what you have. to The other questions are, how much of this product contains 30 whatever it is that you have measured that you have, and where 31 a is it going? Which are the things that X am more concerned a about. I am really less concerned about whatever point you pick and what the ultimate environmental fate of it would be 24 a and what the exposures involved in it are. EXECUTIVE COURT REPORTERS (NO IU4N4 HONS 216248 137 1 MR. GUNTER: Turning to the question of controlled 2 wsst* processes, when this concept wss originally introduced 3 by CMA end submitted to the court in paperwork that we init 4 iated, EOF did not object to that concept then and now you ara I raising some objections. What *bout it has changed your mind? MS. WARRENi Z think we didn't realise that the 7 amount of PCBs could really keep increasing and that' pro 9 duction could -- as you have even acknowledged in your Prsambli, 9 that production could increase significantly as long as it goai 10 to a controlled disposal site. II I didn't envision any conclusion to this proceeding 12 that allowed the amount of PCBs that are being produced to 13 increase. That's really the main reason I changed my mind 14 about it. MR. GUNTtRi Turning now to the disposal of wasta 19 from controlled waste processes, it's true that ?CBs era not 19 a listed waste under RCRA but would there be any legal problem 17 19 or other problem with specifying in the FCB regulations that I15 disposal of wastes be dona in a facility approved under the RCRA regulations as an incinerator of organic wasta? 30 31 22 a 24 MB. WARREN: The problem is that if PCBs ara not a incinerated at the right temperature for the right amount of time, you get Dibenzofuran and other toxic intermediate com bustion products. You still get them as long as there are PCBs, and Z think if you are not going to control the EXECUTIVE COURT REPORTERS (Ml) MJ4M* MOMS 216249 1 incineration of tham properly that'* going to be a problem. 2 Zf that could be taken care of by some specifications 3 in the RCRA regs then I have less problem with it. But I am , 4 concerned about the PCBs still getting out into^the environ I ment and to the extant that they go into landfills that aren't appropriate for them or are incinerated under inproper 7 conditions for then, you are not solving the problem that you 5 set out to solve. MR. GUNTERi That's all the questions I have, thank to you. it MR. GUIMOND; Laura. 13 MS. CAMPBELL: I don't have any questions. 13 MR. GUIMOND: Any. 14 MS. MOLL: I don't have any either, II MR. GUIMOND: I have a few questions from the ia floor, let me take care of. One of then is regarding, I K 17 think this is a coament, and Jackie, you can respond to it if it you would like, regarding Ms. Warren's statement on the 10 protection -- worker protection and clothing barriers. NIOSH 3b document by Or. Weeks has since published an errata statement. 31 Are you aware of this? The errata shows Dr. weeks was in 23 error in his PCBs -- in that his PCB sample was not PCBs, but I contained over fifty percent of a highly chlorinated aromatic 33 34 compound known to permeate rubber. 21 MS. WARPEN: Wall, unless that condition could never EXECUTIVE COURT REPORTERS (W) MJ40M MONS 216250 139 occur in the manufacturing processes in which the protective 2 clothing wes being used, whara PCBs were present as a contamin 3 ate, I think what he said was still relevant. 4 MR. GUIMOND: Okay. Second one, if an~axpesure and I health effsets assessment demonstrates no significant risks from PCBe with a one hundred PPM cutoff, would you be satisfied 7 with a cutoff of one hundred PPM? MS. WARREN: Depending on how you reach the deter mination that no significant exposure could occur. If you 10 change your definition of what is a significant exposure, II probably I will have a problem with that. tl MR, GUIMONDi Do you mean a BPA change from no 13 detectable? ' . H MS. warren: no, from what your determination of .13 what constitutes a significant exposure, which you are pro 10 posing to keep the same definition of significant exposure 17 that you had in the previous rulemaking. 10 MR. GUIMONDi Okay. For the record, would I10 Ms. Silbargeld please state her education and possible exper 30 iences with industry practices or alternatively, your qual 31 ifications to give testimony today on engineering and analy 33 tical technology. a MS. SXLBERGELD: Okay. I am the chief toxic 14 scientist of the Environmental Defense Fund. I have a Ph.D. a in environmental engineering from Johns Hopkins University, EXECUTIVE COURT REPORTERS (M1J MI4M* HONS 216251 ! 143 1 and poat-doctural training in toxicology and biochemistry. For 2 thraa yaara I was tha Kannady fallow in neuro-aciencea at 3 Johns Hopkins Madical School and for aavan yaara a scientist 4 at tha National inatitutaa of Health, ending upas chief of 5 tha Naurotoxicology section. I hava alao bean trained in GCMS and other aapacta of chemical engineering-and analysis. 7 MR. GUIMONDi Thank you. This la another point relative to tha errata ahaat that I just mentioned in the S other comment from NIOSH. Thla ia gueation that I will read j 0 and if you would like to comment on it you can. 1 am not 1 quite aura exactly how it would fit in. 2 Tha question la, doaan't FWPCA preempt tha regulation 1 of PCBa by TSCA in tha control of liquid waate atreama or 4 how will tha propoaed rule affect tha regulation of liquid waate atreama by FWPCA? MS, WARRZMt I don't think that tha Clean Water Act 7 doaa preempt it. In fact, in the Clean Water proceeding back in 19??, where EPA was reaching to direct chargea of PCS manufacture and had a very narrow category of aources, at that 20 time it waa juat PCS manufacturer, deliberate PCS manufactur 21 er, traneformer and capacitor manufacturer, some parties 1 22 in the proceeding Including the EOF, for which 1 was employed 21 at the time, argued that there were many other aources of 24 PCS discharge which weren't being included and the answer was 21 we win reach them under other statutes and under tha permit EXECUTIVE COURT REPORTERS (Ml) WNM MOMS 216252 141 program. So, at that point, the agency wasn't considering the Clean Hatar Act to praampt it. Thi* atatuta allows tha administrator to deal with any problem under tha statute or another statute if--on* or tha othar seams more appropriate and it's entirely discre tionary. I don't think that it is preempted. . MR. GUIMQNDt Okay. Do you really believe, as CPA said in the past -- MS. WARAZNt Yes. MR. GUIMONOi Okay. Do you really believe, as CPA said in the past, that any exposure to PCSs is significant? Is not such a statement incorrect as a matter of science? MS. SILBERGZLDi I will be glad to answer it ss a 4 matter of science. Jackie, perhaps, can answer it in its 5 historical oontext. I think one of the major problems with a a substance like PCBa, as Jackie mentioned, is that it doesn't 7 occur in a vacuum. Both its entrance into the environment is a into a context that is already considerably contaminated by s its presence, and certainly ite uptake and residence time in 90 huaans or any other organism is for a very long period of 21 29 So, it's very difficult to say would one part per 23 billion of PCBs in my tuna fish sandwich today represent a 24 significant event. It would depend very largely what I was 20 doing physiologically m my life at that time. For example, EXECUTIVE COURT REPORTERS (Ml) IU-4M4 MOMS 216253 14 2 i if I were involved in reproduction, it might be very signifi- j cant. It would also depend very much a to my historical 3 encounter* with PCBs. 4 MR. COlMONOt Anything to add? ~~ f MS. WARREN; I already answered the question. MR. GUIMOND) Are there any other comments from the j floor? Questions? MR. REDFORD: I have two. MR. GUIMOND; Go ahead. MR. REDFORDi in regard to the setting of limits using the LOD, Dr. Kailey, earlier this morning, spoke about the LOD for some pruducts being far greater than fifty parts per mill.ion. What would be your reaction to a rule that re flected this type of a limit, well above the fifty? ii MS. WARREN; I think that we would find that un acceptable but on the basis of the potential health effects I of it and the LOD Itself is just not enough, you need to know more about where the product is going, how much of it -- If you know there are PCBS there ** They know there era pcbs there even though the limit of detection because of the nature of what they are trying to detect it in, doesn't permit them a to find it except at relatively high levels but they know it's a there. 3* It seems to me that that Information plus the a answers to those other questions about where it it going, EXECUTIVE COURT REPORTERS (Ml) MI4W* MONS 216254 143 i should determine what the regulatory retponi# is going to be. LOO by itself* or a numerical equivalent of lod, la not adequate all by Itself. HR. BEDFORD: My other question la'on back to the theoretical analysis part. There are a lot of products out there that, I think, will fall into the category that we are talklnq about, and we will have to, the way x understand it with your suggestion, Is send In something saying that they have looked at It and have determined that It doesn't have 10 PCBs. 11 What about products that *- I mean, how far are we 12 going to go? How far do you suggest we go? Should DS Steel 12 have to check steel parts, steel X beams -- X mean, where do 14 we draw our line on the amount of paper work that we are going 10 to accept? Virtually anybody out there who makes a product 10 could end up having to send us in something saying we don't 17 think we have PCBs. Where do we draw our line? It MS. wjuuuSN: The drawing of lines is something -- 15 it's necessarily arbitrary and it's something that regulatory 20 agencies have to do and that is what they are in business to 21 do. X am not going to be able to sit here and say to you this 22 is where you do it. You have some Inherent authority on the 22 basis of administrative necessity to draw the line at a point 24 that I might be unhappy with, but which you, for justifiable 20 reasons, can't deal with beyond that point. ~ Executive court reporters (Ml) M>4N MOMS 216255 1 That's somathing that you aro going to have to do 2 on your own. But it's impossible -- it's something that you 3 fact in everyday -- 4 MR. REDFORDi What would you suggest?' I mean, what s could va uaa to avan attempt to cona to that kind of a line? I hava no idaa. You don't hava to anawar if you don't want 7 to. Kava you got a suggastion? M5. WARREN: I don't think I hava an answer off the a top of ay haad right now. 10 | MR. GUIMONDi Any othar followup questions? Okay, 11 thank you vary much. 12 I Electric here? 11 MR. BLUMEt Ara tha rapraaantativas from Ganaral Wa ara hara today to rapraaant General 14 Elactric Company's Silicone Products Division. I am Richard It Blume, our Division Counsal. To my right is Dr. Dan Silva, j i ia Ganaral Managar of our Tachnology Davalopmant Department. To 17 Dr. Silva's right is Mika Scarbal, our Managar of Quality 10 Assuranca. To my laft is JaffrayCarar, our outside litigation It counsel* with tha Washington law firm of Squlrers, Sandars 20 and Deapeey. Dr. 5ilva will load off our praaantation today. 21 DR. SILVAi Thank you for tha opportunity to part- 22 icipata in this informal haaring. Tha Ganaral Elactric company 23 raquastad this opportunity so that wa maka axplain the effect 24 that era's recant proposal for the regulation of PCBa enclosed and controlled waste manufacturing systems will have on G.E., EXECUTIVE COURT REPORTERS (Ml) MI4N* MOMS 216256 its customers, end the public. 14 5 | G.E. has invested a considerable amount of time end expense In efforts to respond to the PCS Issue. Individually, end as a member of the Chemical Manufacturers Association, we recently submitted comments on EPA's June 9, 1962 proposal for the regulation of PCBs and inadvertently generated enclosed manufacturing systems with controlled waste processes. I direct your attention to those commenst as a statement of G.E.'s concerns with the overall regulatory ap proach that EPA has chosen to follow respecting incldential PCBs. Those consents explain that as currently formulated, the EPA proposal does not provide relief from the ban for G.E.'s carefully controlled process. Today, I would like to concentrate on a number of aspects of this subject. Number one, the nature of the G.E. process within which incidental PCBs are generated. Number two, the Importance of G.E.'s manufactured products to u.s. Industry and national security. Three, G.E.'s efforts to eliminate Incidental generated PCBs. Four, General Electric's efforte to develop alternative processes and reformulate its products. And five, the economic impact of the EPA proposal. I G.E.'s Involvement in this proceeding stems from our phenylchlorosilane production process at General Electric's Silicone Products Division in Waterford, New York. An unwanted blproduct, monochloroblphenyl and dlchlorobiphenyl, unavoidably EXECUTIVE COURT REPORTERS (Ml) MI4M4 MONS 216257 146 1 result from the chemical reaction by which phenylchlorosilanes 2 art produced, 3 These unwanted by-product occur within General 4 Electric's procaaa equipment end ace removed.Joe destruction | s in an incineration facility approved by EPA, Region xx, in j . ' i . compliance with the PCB dispoaal and marking regulations. This 7 integrated process releaeea PCBa well below the EPA permitted e values to the air and water and only trace amounts of mono , and diehlorobiphenyl in concentrations below the fifty part . per million level end up in the products. 1 Phenylchlorosllanes are essential precursors in the . production of certain highly engineered silicone products due . to their superior high and low tamperatue performance, radia4 tlon resistance, clarity and flexibility. Silicone products are used extensively in the aerospace, automotive, construc tion, defense equipment, electrical, electronic, and energy industries. Typical uses for silicone products containing phenylchloroailanea include seals, hoses, insulation, adhesives, coatings, lubricants and fluids where properties of high and low temperature performance are required. Phenylchlorosilane based silicone products are used for thermal Insulation and sealants in aircraft, spacecraft and communication satellites. Our Verallube F-50 silicone fluid is the only fluid available for use in the constant speed drive of the A-4 aircraft flown EXECUTIVE COURT REPORTERS (Ml) JOWQM HONS 216258 147 by the Navy. 2 G.. products in the RTV-SOO soriss srs ussd in 3 rockat motors snd re-entry body ssssmbliss of tha Tridant 4 flaat ballistic missis, snd ara tha onlymaterialsvhich are 5 qualified for such use. Phsnylchlorosilana based silicone e products are used as an adhesive for tha thermal tiles of the 7 spec* shuttle due to their ability to withstand extreme tem peratures upon re-entry to the atmosphere. They are also an essential part of our daily life 10 ss saalants and as coolants in car engines as trail as weather 1J proof coatings for construction materials and iniulators for 12 electrical television components. 12 Historically, in 1975, the General Electric Silicone 14 products Department concluded that tha presence of mono10 and dichlorobiphenyl products in silicone products derived 10 from phenylchlorosilanes, even at low levels, was likely to 17 bacoma a public concern. to Without waiting for legislative or regulatory action. 10 General Electric voluntarily undertook an extensive and com 2 prehensive technical program with the objective of ensuring 21 that PCI materials were not distributed in commerce and did 22 not enter the environment's a result of phenylchlorosilane 23 manufacturing 24 In lg?S, no one for saw that unintentional forma 2 tion of PCB byproducts enclosed in controlled systems would ba EXECUTIVE COURT REPORTERS (Ml) MOWS 216259 1 subject to a regulatory ban. Between 1975 and 1979, General Electric devoted eubatantial, highly sophisticated scientific facilities and human resources to the effort to reduce PCB generation. , -- Over those four years we attempted to find a sub stitute process for phenylchlorosilana production that would not produce by-product PCBe, and to modify our current direct process to reduce the percentage of by-product PCBs produced, to develop substitute products for phenylchlorosilanes and to develop processes which would remove and destroy unwanted by product PCBs from the phenylchlorosilane process. Since 1979, General Electric has had an ongoing program to identify substitute products and to attempt to discover a new process that would not involve by-product PCBs. The specifics of that program were eupplied in a confidential submittal to EPA on July 31, 1979. to date, this program has resulted in only limited improvement in General I Electric's ability to substitute products or to develop a new process. X We have been successful, however, in removing PCBs 31 from our- current process and destroying them. As early as a 197, General Electric's Silicone Products Division reduced a the PCBs in the process to a sufficient degree that by-product M PCB concentrations in products are below fifty parts per a million. Moreover, our on-site disposal facility, which we Executive court reporters (M) >U4M4 HONS 216260 installed at a cost of tans of millions of dollars, assuras tha safa and afficiant incinaration of PCBs. EPa's ban on incidantal byproducts in 1979 came as an unplaasant surprisa to General Electric, particularly because it applied to PCBs generated inside process -equipment, where there is no exposure. EPA must have assumed that the number of processes generating an incidental byproduct was small and intended to use the annual exemptions to deal with the phenomenon. G.E. filed a request for a manufacturing exemption in order to continue operating the phenylchlorosilanas process, Because the 1979 regulation defined PCBs to include only those substances that contain PCBs in concentrations of fifty parts per million or greater, the processors, distributors, and end users of our products did not require exemptions. I General Electric filed a law suit challenging the regulation on the basis that Section 6E of the Toxic Substances Control Act did not apply to incidental byproducts. However, when the EOF court decision necessitated a new rulemaking. General Electric chose to postpone its law suit in the belief that EPA would formulate an appropriate, regulatory solution for incidental byproducts in that rulemaking. One final avenue of relief General Electric has sought is the exclusion of mono* and dichlorobiphenyl from the definition of PCBs. Based on our belief that there is EXECUTIVE COURT REPORTERS 1 (JMJ HMM HONS 216261 150 1 1 ample 3unification for that action, Ganaral Electric peti- 2 tioned EPA on July 14, 1982, to amend tha PCB regulation in 3 that respect. 4 This is what the picture looks like-in- 1982 for ! t General Electric Company. Our exemption request, originally I filed in 1978, has never been acted upon by EVA. On j August 2, General Electric will file, at EVA's request, a renewal application for exemption. The fifty part par million cutoff in now in doubt, meaning that even if General Electric gets a one year manufacturing exemption, tha processing, distribution and use of our phenylchlorosilane based silicone products may still be prohibited. What General Electric is hoping for is that EPA will recognize the strong justification for excluding our system and products from the ban. However, EPA has been working since November of 1980 on developing a rule to deal with the incidental FCBs and other PCBs in low concentrations, and no relief is yet in sight. Given the history of this rulemaking, it is clear that EVA's notice of proposed rulemaking of June 8, 1982, is an attempt to coma to grips with incidental byproducts. In that proposal, EPA announced its conclusion that excluding from the ban incidental PCBs produced in closed and controlled waste manufacturing systems would not pose unreasonable risk to health or the environment. EXECUTIVE COURT REPORTERS (M) JU-MM HONS 216262 However, EPA failed to establish a meaningful, 151 2 di minimi* standard by which incidentally generated PCS* can 2 be regulated. The current proposal therefore offer* no relief 4 to General Electric and others who can quantify PCS* of their 5 end proouct*. As the June 8, 1982 proposal is structured, regardless of whether trace amounts of PCSs found -in end7 products of these processes cause minimal or no risk to health 8 or the environment, all processes and end products in which trace PCBs can be found, will be banned. 0 We at General Electric are very concerned. The 1 date which the court's mandate goes back into effect is less 2 than three months away. EPA wa* to have completed work on a 2 di minimis standard by then, yet EPA has given no indication 4 when, if at all, it will propoee a di minimis standard by 8 which companies in our position continue to operate. 8 | unless the court once more extends the stay of its 7 mandate, companies such as General Electric, who can quantify s trace amounts of PCB impurities in their end product will see le their products banned. Even if EFA publishes the regulation as currently proposed, industrial users and supplies of these 20 21 products will also be prohibited from using or distributing 2X products containing trace amounts of PCBs after this date. 23 If this ban continues. General Electric will have 24 only one elternative to ceasing production of phenylchloro28 silane based silicone products. It will have to rely upon epa executive COURT REPORTERS (Ml) MJ4M4 HONS 216263 action on its exemption request. 152 2 Aa SiMrtl Else trie has alraady pointsd out in its comment of November 16, 1981, submitted in response to tha Hay 20, 1981 advanca notica of proposad rulemaking in this proceeding, tha exemption procasa is an inadequate solution to tha problem, dus to tha unjustifiad expense and administrativa burdan that it placas upon industry and EPA alike. If Ganaral Electric's only hopa is to rely upon submitting its exemption request to EPA each year in order to continue producing phenylehlorosilanes, then Ganaral Electric must consider tha likelihood that it will have to discontinue production of phenylchlorosilane based silicone products. Tha loss of its phanylchlorosilane based product lina will not topple G.E. However, tha impact of this event upon Ganaral Electric, its employees, consumers, and indeed our national defense, will be severe. Currently General Electric sells hundreds of pro ducts derived from the precursor phenylchlorosilanes to thou- I sands of direct customers. The immediate economic effects to General Electric of discontinuing its phanylchlorosilane based products include tha immediate loss of tens of millions of 21 dollars per year in sales and the estimated direct loss of 22 more than 200 jobs at its Waterford, Sew York facility. 22 24 In addition, this would represent a loss of our 21 substantial investment in developing a process for removal cf EXECUTIVE COURT REPORTERS (Ml) 153 i by product PCBs from phenylchlorosilane -- from the phenyl- 1 chlorosllsns procsss, along with ths loss of prassnt facilities 3 for ths production of phonylchlorosilana and removal of by products, mono- and dichlorosiphsnyls. Ths combinsd vtlus of $ thssa facilitiss is estimated In ths tsns of millions of 8 dollars. Ths majority of ths capacity of these facilities can 7 not be used for any othar applications. Further, ths unavailability of phonylchlorosilana 8 based silicons products would have a savers, adverse scon* 10 omic impact on direct customers and other users. As indicated 11 above, phenyl containing silicones are typically usad in the 13 highly critical applications such as military, commercial 13 aircraft, military equipment, and our nations' highly visible 14 I space shuttle. 18 I In many applications, phenylchlorosilane containing 15 I silicone products are dpdcifically required by customer 17 Rspecifications. The consequences to commercial customers 18 would be substantial, both in terms of lost profits and sales. IS Hand the inability to find substitutes. 80 Even if lower performance materials for substitution 31 eould be found, this would require development programs esti 23 mated to range in duration from one to ten years. The most 23 critical applications would tend to require longer development 34 and evaluation time. Numerous General Electric customers have written to EPA to emphasize the importance of our producti to EXECUTIVE COURT REPORTERS (Ml) IU-W4* HONS 216265 i them. 4 Mot recently, on July 12, 1982, the Department of the Navy wrote to EPA thet a ben would ground the A-4 aircraft end would neceeeitete e eeverel year program to replace the constant speed drives on all effected aircraft at a cost of more then $16 million. Other adverse effects of the discontinuance of General Electric's phenylchlorosilane based silicone products, include the substantial loss of sales and jobs for suppliers Iof goods and service to the Waterford plant. The discon tinuance of General Electric's phenylchlorosilane process will have an impact, not only on the Waterford, New York area, but on the national economy as well. Use of the annual exemption process to deal with the large number of processes and products involving inci- I dental PCB by products would be most unfortunate as a matter of public policy. Such a solution to the problem would be illusionary. Not only the administrative burdens that it places upon EFA and industry alike, but because it fails to ao address the real issue posed by incidental FCBs. The real issue in this entire proceeding is whether si I regulating PCBs as a defined regulatory cutoff would adequate 22 ly balance the interests of industry, government, and con 22 24 sumers. The exemption process is cumbersome and time-consuming as and is designed to deal with case by case issues, it is not EXECUTIVE COURT REPORTERS (Ml) H144M HONS 216266 L5S I th proper forum to Address major issues that cut across all 1 the interests involved. 3 Such an ad hoc approach will necessarily make a 4 shem of the exemption mechanism. Either the annual-exemption s will be pending for years, as has already happened, or EPA will rush through applications with a cursory view.that does 7 not address anybody's Interests. Meanwhile* the applicants t are left hanging in the balance wondering when an adverse 5 decision will put them out of business. 10 X have attempted to emphasise to you today the 11 efforts that have been made by General Electric to solve the 12 PCS problem. So far, these efforts have not guaranteed us 13 the right to operate our process nor has the substantial 14 amount of time and money invested in the administrative pro1B cess by private industry and government achieved such a ig guarantee. X have outlined the severe economic impaet EPA's most recent proposal will have on General Electric and its customers. We have no doubt that the proposal equally effects 30 many other companies. EPA itself recognised in its June 8, . 21 1912 proposal, that a ban on Incidental by products would cost 22 billions of dollars and cause the elimination of a wide variety 23 of products with great eocial value. 24 If this is the ultimate result of EPA's efforts on 2$ PCBs, we submit a major public policy mistake will have been EXECUTIVE COURT REPORTERS (Ml) M14M4 HONS 216267 1 I I 1 I I I 1 1 l 1 1 1 1- I 1 1 1 I 1 1 I 1 136 , mad*. w* must reach 4 solution to th* dilemma 0f incidental by product* * soon 4* possible. October is already close *t 2 3 hand for PA to be equivical on this issue. If EPA is not 4 prepared to agree that the statute doe* not berrincidental 5 PCB by products, then it must achieve a practical compromise, setting a meaningful regulatory cutoff for the manufacture. 7 processing, distribution and end us* of incidental PCSs, can adequately balance the interest involved in this proceeding. Until EPA takes action to resolve the current 10 dilema, the uncertainty and confusion will continue. Thank 11 you. HR. GUIMOND: Thank you. If X understand your pre 13 13 sentation here you, of course, are particularly concerned 14 that G .. is not going to make th* proposal as it is now IB structured, right? nVDile C0iTTVVnlfl fwAOriPaVdCrfbi '* n||Bil a uuiTAlluOnllwA ai 9Uv iyfoAutl a 1 aa Wumv1 idim"ifvA *uWiteb6 %yaou < r mp*roaQd uctsfl 17 11 _p___n. siot 4n1 ctit nuubn& tAxwpeeoeiuii rwv*r ans vhnd MoAneniecmnil cSsA1i eixi,p mobi>u. rtrg ' 1 30 correct. bDbR* Sm IL*Vi>As l. eT>OA etM1. b kDBtbtbt Ob r ob%ut br kKbAbO. .V1 lbtaA9b tb , ltnw ab bt P b9 nUBil* fuiullTinliuftWnbfl *i 8 04Vaaviui knsuvav fadvAfBivB *vAvi#u|ev bfviiiu 8wHiRif1liVj Cn1*8 " and you feel that people are not coming in eontaet directly or 33 they don't get into th* environment or whatever, is that 33 34 correct? AOtKts f9i1trLiViaA,i Un%b krvbtv>bt Am uWi*QJb# *V(>iab JivBr4y\f aw#i BpBroAafduitcAbbee ( enu 33 EXECUTIVE COURT REPORTERS (Ml) III4M4 HONS 216268 1*7 1 product*. Many of those materials have trace level* of PCB* 2 in * matrix which do** not allow movement of that material 3 from th* matrix. I could give you a mor* definitive anawer on that if you would *ubmlt that question to ua in writing, w* possibly could give you a spectrum on some of our products. MB. GUIMOND: Okay. Let me ask a hypothetical question. You have already identified your concern that you right now do not fall into this proposed exclusion, therefore are up in the air as to what might happen in th* future, whether you have to just rely on th* exemption petition pro cess , if subsequent to this rulemaking, if EFA went ahead with this current rulemaking and th* way it's set up now, went final with it, and you did not qualify, and in th* following rulemaking which w* are currently planning to undertake, we then proposed to exclude products that presented little risk, and assuming that your's fell into that, would there be any economic impact or any problem to you at that stage? Would that be reasonable solution for you? Could you live with that? OR. SILVAi If you are talking about, again, an 30 annual exemption procedure, and we would have to verify the 31 levels -- 33 MR. GUIMONDi I *m not talking about an annual ex 33 emption procedure, I *m talking about -- 34 DR. SILVA! You are talking about a permanent -- 30 MR. GUIMONO: Let's just hypothetically assume that EXECUTIVE COURT REPORTERS (Ml) MMW HONS 216269 158 1 we came up with a third rulemaking and tha third rulemaking 1 consisted of, part of it in any avant, w proposed to exclude 3 cartain procaaaas that prasantad di minimi* risk or small risk, that ware ovar and above, that were ,different kinds of 5 processes from closed and controlled waste processes. Let's hypothetically assume that your's fell into it T What would be your problem with this rulemaking in that event? DR. SILVAi My opinion is that it would be of minimal impact, assuming that we had guantative limits on whtt you are talking about in tarns of risk. We would want a quantatlve sat of specifications. a MR. BLUKE) I am not cartain to what you are refer 3 ring with respect to tha next phase of the rulemaking rela 4 tive to a process other than that which is being referred to 6 today. MR. GUIMOND) Alan Carpien referred to it earlier ? when we talked this morning, is that we are doing this closed and controlled rulemaking response to the court and then we I have schedules to do a third rulemaking that deals with un 30 controlled PCBs, and it has always been our intent in the 31 closed and controlled one to deal with a category which was 33 described from -- started from original discussions with 33 industry and then was ultimately -- a deal, if you will, was 34 cut in an order with the court. 23 What I am saying is that that does not preclude us EXECUTIVE COURT REPORTERS (Ml) MI4DM HONS 216270 1'9 from excluding additional processes because they present di 1 minimi* risk although thay may not ba cloaad and controlled 3 west* proca**a*. What I am just trying to find out ia that 4 if you faal -- Lat'* assume you don't qualify for. this., and I w* fix this so you still don't qualify, in th* third rule making you still conceivably might qualify for an exclusion depending upon the risk associated with what you are doing 7 and how that gets crafted. If, in fact, you did qualify in that third rule 10 making, would there be any particular impact or negative 11 problems created for General Electric by this current rule 13 making? What I am trying to get at, it seems that you are 13 calling this on* a large impediment, this is a problem because 14 you are not included in it. 1 Th* response, I guess, or the thing is that the 1* rule may not have been crafted for Inclusion of your kinds of things. It does not mean that your kinds of processes would 17 10 be banned. I am trying to find out, given that as an assump 10 tion, do you have any additional problems with this rule making? K KR, BLUMEt At the risk of reducing the question to 21 a basics, I gather the question is, that if a further phase of a the rulemaking were to come up with parametere be fashioned along lines which would exclude our process, would we be warm 24 and happy? EXECUTIVE COURT REPORTERS (Ml) >U<4M HONS 216271 1 MR. GUXMOND: Yes, you have got it. 150 MR. BLUME: The answer to that i definitely yes. MR. GUXMONDs A real furry feeling, MR. BLUNE; Jeff, did you have something? MR. CXAAAt Yes. I'm Jeff Cerer. I em counsel to G.G. on this rulemaicing. The thing that troubles me about the question is that it was precisely our kind of process that this deal that was struck was supposed to be addressing. That is why we have difficulty with it. When we listened to your questions this morning of CMA, where you read frost a January 1961 submission to the court, X think it's a mistake to characterise our positions today as inconsistent with what was said back then. For example, the closed system, as defined in that submission, and defined in some of our discussions, we had trouble with the concept of any detectible PCBs, but as a matter of fact, the closed system was supposed to be -- the attempt was being made there to define something that was almost a theoretical possibility. That was a system in which SO PCBs come and go and they are created and disappear and never ai go anywhere. aI Nobody has ever believed, the CMA people have never a believed that there would be very many systems falling into 34 that category. The controlled waste system, as you have called it, and as was part of the discussion back in 1961, 1980, is EXECUTIVE COURT REPORTERS (Ml) M14M4 HONS 216272 161 i really the on* that w* ar* talking about and it'a tha one that ; G.E. would hop* to qualify for. That'* on* whr* th*r* ar* 1 PCBs that com* out of th* ayataa in th* products and in the waste. . s Th* problem with th* mechanism that you fashioned a hr* is that it goes down to l*v*ls that ar* so low that no- j body can qualify for it. If there ar* paopl* who can qualify S for it. th*n our surmis* is in error, but G.E. can quantify at such low l*v*ls that th* system, as st up, is of no M> |practical value, Sur*, it would make G.E. vary happy to b* axcludad on* way or th* other from this, and in that sans*, this proposal wouldn't pos* an impediment. But, this proposal was I I supposed to d*al with precisely the kind of problem that w* have brought to your attention. MR. GCIMOND: Okay, thank you. X am, unfortunately. | going to have to leava for a few minutes to go to a meeting I can't avoid. In my absence. Bill Gunter will be tha Chairman of th* panel. MR. GUNTERi Amy Moll. at MS. MOLL: Would you say that most of the processes 21 that you say would not be excluded under this rulemaking, 33 would not be excluded on th* basis of tha concentration of 34 FCBs in their products, like in G.E.'s case? DR. SILVA: If I can repeat the question a little EXECUTIVE COURT REPORTERS (Ml) J414M4 MQNS 216273 IS? and if I understood what you ara saying, you are saying the exclusion would not be granted based on the levels in the final products, is that correct? MS. MOLL: Yes. ------ DR. SILVA: I. speaking for our own process. I believe that's correct and I think for many othe.r processes that I an familiar with I would also agree. MS. MOLL: You mentioned that C.E. has filed ex emption petitions in the past. Can you give us any idea as to what the coats of filing those axemptions has been for your company? Or if you could submit that information. DR. SILVA: Well, the cost is -- if you want to try to calculate the direct costs, you might want to try to look at xeroxing. I am trying to be a little facetious here, but it's the indirect costs that gat you on these proposals. For instance, the exhaustive analysis of our product line and the Iprecursors in many of our product lines are the type of in direct costs that are built into these exemption petitions. Having to do this yearly would swan we would have to update our anaylsis every year and, in fact, I would put the cost in the range of several -- tens of thousands of dollars -- certainly over one hundred thousand dollars cumu lative for preparing just our petition. IMS. MOLL: Does that include doing all the research to find alternative - EXECUTIVE COURT REPORHTOERNSS 216274 (m) NMoa If 3 1 DR. SILVA; No, no. That's why I **y, if y0U want 2 to include tha actual totting, at CAtara, that want into 3 being abla to prapara tha patition, it*a in tha ranga of 4 hundrede of thouaanda of dollara. Aa wa gat .into the effort* s I outlinad ragarding altarnativa procaaaaa, wa ara talking milliont of dollara, and that did not ineluda thoaa afforta. J MS. MOLL: What would you aay ara tha cloaaat 9 aubatitutaa for phanylchlorotilanaa at thia point? 9 DR. SILVA I with l know. I think wa hava lookad 10 at bom material* which coma cloaa, but having tha right mix It of ovarall propartiat ia tha iaaua, flaxibility, low tempar12 atura, high tanparatura, alactrical raaiatanca, radiation 13 raaiatanea, having all thoaa thinga in tha right molacula ia r 14 pratty tough to coma by. 1 It`a many of thoaa problama that hava causad many of our cuatomara to coma to ua for thaaa apacific material*. '' I The apace ahuttla ia a good example. Thera ia no material 10 that hae been abla to achieve tha type of adhaaion and thermal 1* raaiatanca to data. K MS. MOLL I That1a all I hava. 21 MR. GUNTERI Laura. 22 MS. CAMPBELL; My firat quaation ralataa to worker 23 jjexpoeure to PCSa. What maaauraa hae your company taken to 34 avoid worker expoeure or deal with thia problem to PCBr during IB tha production of phenylchlororilaner? _ EXECUTIVE COURT REPORTERS (Mi) J4J.4M4 HONS 216275 164 1 DR. SILVR: Okay. Phenylchlorosilanes are produced 3 in 4 closed system. The work pl4C monitoring is-done routine-: 1 ly 4t W4trford tnd h4i looked at vary low level* of PCBs in 4 the air. All those anayliea have been, if you-will, analyzed I and processed and are dona routinely. I am not exactly sure of the time frame in which we are doing them npw, but it is 7 essentially an air sampling type of technique. Some procedures involve tha operators wearing II actual sampling devices which are then analysed as part of 10 tha whole site survey. II Mike has commented that some of that data is prsssnt 13 in tha 1979 exemption patition. 13 MS. CAMPBELL: He also asked for information in tha < 14 notice of proposed rulemaking and Z was wondaring if you have II any data indicating both the number of accidental or unplanned II releases of PCBs, and if so. tha amount of PCBs releaaed 17 during such events? 10 DR. 5ILVAi I am not familiar with what information II wa have subeiitted to date. Dick, would you like to - K MR. BtOMB: He eertainly don't have that information 31 with us today, but to the extent that such incidents are 33 raquired to be reported, they of course are, and that inform33 I ation is available to you. 34 MS. CAMPBELL: Okay, I am interasted particularly in if those incidents have occurred, what measures have been EXECUTIVE COURT REPORTERS (M) J4MM4 HONS 216276 taken to protect the worker* or to prevent environmental exposure, elen-up procedures. DR. SILVA: We have standard procedure* for all the processes that involve phenylchlorosilanes which contain pcb*. Those are documented procedures in terms of how operator* and various other workers can-- must, if you will, handle the materials which are spilled or whatever. Dick pointed out that there is a notification procedure involving pcb releases, and to my knowledge we have been following that procedure. I do not have the specific information relating to how much or if there have been any releases. MS. CAMPBELL! Thank you. MR. GUNTERi Dr. Silva, in your prepared statement, you indicated that the date on which -- on page 7, you communicated that the date on which the court's mandate goes back into effect is less than three months away. Are you not aware that that mandate has been extended -- that the mandate has been extended to December 17 DR. 8XLVA! No, X was -- I MR. CARRIER: it is stated in the Preamble. It ha* 9 been requested in the Preamble and the court granted the request and gave the date. It is April 15 or 9, I can't 8 a remember exactly the date. 34 DR. SILVA: Okay, *o we have an additional month. a It doesn't help us -- it would be nice for a Christmas present EXECUTIVE COURT REPORTERS (XI) MI4M MOMS 216277 166 MR. GUNTERi I would like to address this next question to the two attorney*, if epa i* to go back to the court, a* we are currently scheduled to do on November 1, with a plan a* to how we are going to address the third part of the rulemaking, wouldn't our credibility with the court be im proved if we had gone forward with the promulgation of a final rule on closed and controlled systeM as the court had ordered us to do, rather than to withdraw that and report to the court that we are simply dropping beck to rethink the whole situation? hr. CERARi I agree with your basic promise that it is always better to show some action when you are going back to the court# particularly action that you said you were going to take. But# if we are correct that it doesn't do anything, that isn't a very satisfactory kind of action. Maybe at this point we ought to say that we haven't asked you to withdraw this proposal. General Electric has not. What General Electric would like would be this proSposal with an acceptable cutoff substituted for the level of detection criteria and that is doable within the period directed by the court. MR. GUNTERi You would like to see an acceptable cutoff based on the highest di minimis standard which could be supported, is that correct? MR. SLUMSi That's correct. EXECUTIVE COURT REPORTERS (mi) tii mi HONS 216278 167 MR. GUNTER: How would EPA manage to get ill of that 1 don* bifore Die unfair 1? 1 MR. CERAR: You havi got thi information in the 4 ricord. There hal been a good deal of health effects inform 5 ation submitted to you. The main problem with the current schedule is that even with an extension until December l, 7 what you are going to be doing at that point is going back to the court and asking for a further extension on the basis of I a plan that you will be laying before it at that time. 10 MR. GUNTER*. That was precisely the point of my II question on credibility. Wouldn't we be in much better shape 12 in getting that plan and further state of mandate approved if 12 we had done exactly what we had told the court earlier that we 14 were going to do then if we hadn't? 8 MR. CERAR: I think that's quite true. It's unfor II tunate that you can't accomplish the whole thing by doing 17 something maaningful at the seme time. I'm not trying to be II |facetious there. It seems to me it must be -- I mean. I hope II it's possible to anaylze all of that information by the time-- 20 by December 1, ideally. 21 ut the thing that makes a company sitting here 22 welting for all of this so insecure is that the court may or 22 may not grant the extension. They have granted -- the court 24 issued an opinion over two years ago. or almost two years ago 21 now, and they have allowed their mandate not to go into effect, EXECUTIVE COURT REPORTERS (m) HONS 216279 168 and to kaap coming back and asking Cor mora time is a bit precarious - MR. BLUME: I gather that EPA has no assurances on that score. _'~ MR. GUNTERi Not from the court. MR. CARPIEN: Obviously the court doesn't say what it is going to do, but in the original order, the court broka up the incidental manufacture of PCBs into two different rule making. One, of course, is the one we are doing now, and th* other one was the so-called uncontrolled PCB regulation, it seems to Bta the court wouldn't have made that statament, wouldn't have allowed the rule to be broken up into that other rulemaking if it did not intend to grant some kind of stay of its mandate, given the fact that we could make the appropriate showing. It is certainly implicit in that stay of the mandate would be, I would imagine, at least a good chance of being able to have that mandate extended. The question would be how long, X think, would be our main problem. tat. GUNTERi I have one last question. Back in tl 1975, you said you became avara of the possibility, the a likelihood of public concern over the presence of mono- and a dichlorophenyl in your products. Row did you arrive et that 34 conclueion? n DR. SILVAt Well, from a quality assurance prospective EXECUTIVE COURT REPORTERS W MONS 216280 16a 1 t and some, if you will, anaylsis of our upstream processes, we 2 concluded that tha levels we were currently running with in 3 our finished products could be reduced and should be reduced. 4 I think it was a culmination of somg studies we had 5 going on internally and X think certainly some public reports concerning the effects of PCBs on a larger scale. ;We took the 7 initiative because we felt it was appropriats to remove the |mono- and dichlorobiphenyls that were present in our process. X would like to point out that we do not have any 10 trichloro or higher chloronated PCBs in our phenyclchloro- 11 silanes process and that the mono- and dichlorobiphenyl 12 materials are not as long lived in the environment as sms of 13 the very highly chloronated biphenyls are. 14 However, wo felt that we could do something about >19 the problem and we took that action in 1976. ia MB. GUNTER: It has been G.E.'a position, X believe 17 consistently throughout all these proceedings, that the 19 incidentel manufacture of PCB should not be subject to I1* Section 6(e) of TSCA. If you were aware of this problem back 30 as early as 1975 when TSCA was being considered, why didn't 21 you try and get Congress to make clarification that would have 22 helped that position? 23 DR. SILVA I X think X will ask Jeff here to coimnent 24 on that. Let me just say that X quoted in the text, that we i 29 always viewed incidental by products where we had a closed 1 EXECUTIVE COURT REPORTERS (Ml) MJ4M4 MOWS 216281 no yitim and v* do have in integrated cloud system involving 2 |both incineration and aome absorption type of processes, we 3 did not anticipata a ban based on incident*! by products. * HR. BLUMEi I can answer that question-* the extent I we are able. Neither Hr. Cerar nor myself were around in that era and we can only speculate at this juncture,- so I don't 7 know that we can give you any kind of meaningful response. Mft. GUNTER: If there are others around who were I around during that time, perhaps you could give something to during the reply consents. That's all I have now. Alan? II HR. CARPIRNt Z have no question*. 13 HR. SMITH: Since Z am involved with sempling, 13 sampling relies quite a bit on the variability of the system 14 and I have heard statements this morning about folks feel that II you get relatively the sum isomers in patterns. My question II is, from your analytical experience with your system, do you 17 see homogehiety'in your, if you will call it contamination II or aide reactions, or as you get slugs, what I would call II sluge of PCSs? DR. SILVA: Well, as Z mentioned earlier, we have 31 pent a considerable amount of time and money understanding 33 the chemical reaction with the intent of trying to reduce the 23 amount of PCBs generated. We have obtained information which 34 indicates that in our particular system it's essentially a n steady rate of production and that the isomer content does not EXECUTIVE COURT REPORTERS (Ml) HMM HONS 216282 1 chang appreciably. 2 As I mentioned, we have moncchloro end dichloro- 3 byphenyls. There Is no trichloro material. The monochloro- 4 and dichlorobiphenyla I am sure you could get out of a standard textbook, would therefore give you nine possible isomers. We see all of those nine isomers. He understand the. ratios of 7 each one to the other. He routinely analyte for each one of those nine isomers. hr. REDFORDi I would just like to get your reaction ' 10 to the analytical methods that EPA Is proposing as in the KRI 11 document. 12 DR. SILVA: You are talking about CC Mass Spec, is 13 that correct? 14 MR. REDFORD: Yes. 10 DR. SILVA: He currently use GC Mass Spec to do many 10 of our analyses. I think it's what I would feel comfortable 17 Jvith in terms of soma of the analyses that we do in our 10 facility. 1 think, contrary to some of the comments that were 10 made earlier, that the level of detection or, if you will, the 30 LOQ, or whatever you want to call it is going to change 21 continually over the next decade. It has over the past decade. 22 I think GC Mass Spec gives us some very good answers 23 now and I expect to see more techniques coming along that will 34 take that even further. I think that's the real concern on LOO or level of detection. In a sensa, if you go by level of 20 EXECUTIVE COURT REPORTERS I (Ml) HM4M HONS 216283 172 ) Idetection, you are 901119 down even lower than loo, because 2 moat analytical chemists will tell you they need a little bit 3 acre level of detection to be able to quantify what you have. 4 I think we are askln9 a lot of our chemical engin- earing people as we move down an* become more sophisticated in our analytical techniques. J MB. REDrORD. Thank you. ' f 10 1 *1 a a 2 a II EXECUTIVE COURT REPORTERS (Mi) MMH HONS 21628* ;73 MS* KEEHNER: Xf EPA modified it* definition of a controlled waste proceee end aet the cut-off for product! et 50 parts per million, could you qualify for the exclusion? MR. SLUMS: Repeat the question one more time, please? MS. KEEHNER: Xf EPA modified its definition of a controllsd waste manufacturing process and set the level for no PCBs at 50 parts per million in the product, could you qualify for exclusion under this rule? MR. BLUME: it's still sero quantifiable with respect to the manufacturing process and waste streams? MS. KEEHNERi Zero in air and zero in -- well, no quantifiable in air and no quantifiable in water releases. DR. SLIVA: X think the definition of no quantifiable would get us into a discussion immediately, X think that's, again, the problem. Zero is a very low number and maybe that's the intent. However, from a reality position, X think we need a definition of what sero means. MR. BLUMS: And X gathered. Dr. Silva, you are referring not only to the GX system, but virtually any systma if you are going to test it in terms of zero quantifiable. DR. SLXVA; That's correct. I believe that EXECUTIVE COURT REPORTERS (Ml) MSMt HONS 216285 174 whanavar you can talk about on* molacula aacaping from a ystaa, you'va violatad zaro. I think thara ara paople who try to intarprat tha law in that aanaa, avan though it night not ha tha intant of tha law to count tha tnelaculaa that laava fron a procaaa, I think that would ba tha difficulty that induatxy and Ganaral ElaCtrlc would hava with a atatamant of no quantifiabla fron a cloaad ayatnn, MS. KEZHNERt Sow would you dafina manufacturing procaaa, a product and a ralaaaa for purpoaaa of your aya tan? DR. SLXVA Row would I dafina -- lat na taka tha firat ona, a product. MS, KEERHIRt X procaaa, could wa try that first? DR. SLXVA1 A procaaa. MS. KKEHHERi Right. DR. SLXVAt How would I dafina a procaaa which would ba inadvartantly producing PCRa, ia that -- MS. KEERMBRi Yaa, an ooncarnad about, for inatanoa, nayba you hava nora than oha building on a aita or a plant and you ara moving things fron building to building. What is tha procaaa? Xa tha procaaa tha firat atap in a nultl-staga, or would you conaidar avarything -- DR. SLIVAt Moat chanical procaaaaa do not taka EXECUTIVE COURT REPORTERS (mi) mm* HONS 216286 175 1 taX* plac* in on* particular v**s*l and ar* eonnactad by pipes. In too* caa*a they ar* connected by tha movement of tank wagon*. Our proc*aa do*a conalat of several at*p*. it doaa take place in on* or more building* and la connected by a piping ayatem. NS. KSEHNERt So you would consider that to ba your proceaa. DR. SLIVAt X would call that the Integrated process for the manufacture and control of phenylchlo roa1lane. MS. KSEHNERt How about releaaea auch *a to water and air? Do you consider releaaea -- DR. SLIVAt All our vents and all our points of release ar* categorlied and have undergone both a theoretical analysis, and i'll content on that In a minute, and an actual empirical analysis of what thoaa emissions ar* over a period of time. There are protocols for those test procedures. I do not know what the results are, other than that we did prfsent in a 1979 exemption petition some data relating 1 to workplace monitoring. tn regards to the theoretical calculations, one can calculate -- can quantitatively measure with GC mass spec what the concentration of PCIa ar* in the materials EXECUTIVE COURT REPORTERS (WlJ MUM* HONS 216267 176 going to tha vanta. Wa call that thaoratlcal bacauaa wa bacX-calculata baaad on what wa'v* found in tha anvironmant and aak ouraalvaa la that raaaonabla. In moat eaaaa tha matarlal balanca did ehack. MS. XCEHNZRt I hava no furthar quaatlona. MR. GUNTIRi Follow, up anyona?-' thank you vary much, gantlanan. Thla conplataa tha achadulad aganda of partieipanta and, alnca thara la tlaa ramaining bafora tha and of tha day, la anyona alaa praaant who would Ilka to laaka a atataaant for tha raeord? Ha hava an opportunity to hava that at thla tlaa. (No raaponaa.) MR. GUNTERi Thla haaring ia cloaad. (Kharaupon, at 2t52 p.n. tha haarlng In tha abova-antitlad nattar waa concludad.) ++++ EXECUTIVE COURT REPORTERS (M) MMW HONS 216288 ripoetia i certificate DOCKET NUMBER) CAfS TITLE) Cloted And Controlled Nut* Proeeae Rule HtAAINC DATSi July 2, 12 LOCATIONi Weehlnpton, D.C. . I hereby certify that the proceedings end evidence herein ere eenteined fully end eeeurately on the tepee end netee reported by M et the hearing in the above eeee before the United ft*tee environmental Protection Agency . end that thii ie e true end correct trentcrlpt of the earn#. Detei July 27, 1P02 flL Official Reporter 21 Cxeeutlve Court Reportere S2S Coleeville Road 29 Silver Spring, Md. 20110 29 94 H HONS 216289