Document w8LYKM8YLwBeR313Q2bNaVY6

S T E P H E N B H A M IL T O N JR OCR ' ( V ' H O N M t N T A L I C K M C t AMO T lC ~ M O l. O O * G E N E R A L ELECTRIC GENERAL ELECTRIC COMPANY rAmrido. Connecticut oe3i Ju ly 2, 1985 Dr. J . G. Wirth General Manager P la s t ic s Technology Department 1 P la s t ic s Avenue P it t s f i e l d , MA 01201 Dear Joe: As you can see from the attached, I have been asked to comment on EPA's d ra ft of proposed co n tro ls on the inadvertent manufacture o f halogenated dibenzofurans and d io xin s as contaminants during the manufacture of other chem icals. I was requested to review th is because EPA is aware o f the development work done at CRD on q u a n tita tiv e a n a ly tic a l techniques fo r chlorinated dibenzofuran congeners. I am enclosing a copy of the questions on which EPA requests advice, the proposed l i s t i n g , the g u id e lin e s fo r sampling and the q u a lity assurance plan. I have sent the g u id elin es fo r the a n a ly tic a l te stin g to CRD fo r t h e ir comments. I would appreciate your comments on how important an issue t h is i s to PBG and any other comments you consider relevant. Please note the request th at comments be sent to EPA by Ju ly 19, 1985 in order to be used in the proposal. Sincerely SBH :1s S. B. Hamilton 783320 GENP 01 fKnn U SB/ UNITED STATES ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 I Dr. Steve Hamilton General Electric Corporation W-1B2 3135 Easton Turnpike Fairfield, Connecticut 06431 ornee or PESTICIDES ANO TOXIC SUBSTA N CES /c ei Je 2 4 tifi 3 &. V Dear Dr. Hamilton: The Environmental Defense Fund (EDF) and the National Wildlife Federation (N W F ) petitioned the Environmental Protection Agency (EPA) under Section 21 of the Toxic] Substances Control Act (TSCA) to immediately impose controls on the inadvertent manufacture of dibenzofurans and dioxins as contaminants during the manufacture of other chemicals. EPA, in its response to the petition, agreed to initiate an administrative proceeding to determine whether additional regulatory action was warranted. EDF/NWF have subsequently' filed suit challenging EPA's decision to initiate an administrative proceeding in lieu of immediately issuing controls on the inadvertent manufacture of dibenzofurans and dioxins. 1 The Office of Pesticides and Toxic Substances (OPTS), in co operation with the Office of Solid Waste arid Emergency Response, has developed a list of chemicals which, based primarily on theoretical assessments, may have the potential to produce chlorinated and/or braninated dibenzofurans and dioxins as contaminants. OPTS intends to require analysis of the chemicals on the list (1) which are prioritized as 1 2 or 3, and (2) with routes of contamination shown as S or FS,. (and Fj in those cases where the feedstock is not listed), and (3) which show uses other than as a pesticide only. The analysis will| be used to determine whether dioxins and dibenzof urans are prese'nt, ajt what levels, and under what process conditions dioxins and furans are formed in these chemicals. Mike^Dellarco, Program Manager for Environmental Monitoring Systems Division, Research and Development, |and Dr. Don Barnes, Senior Science Advisor to the Assistant Administrator for i 783321 GENP 010510 i Pesticides and Toxic Substances, suggested that we ask you to review the preliminary list of chemicals t'o be tested, the rationale for selection of chemicals to be1 tested and the preliminary analytical guidelines for tesding. j We have included a copy of th list, a detailed description of how it was compiled, and a preliminary copy of the analytical guidelines, along with a copy of the preliminary sampling plan and QA plan. To help direct your review, we have attached a page of questions and issues we have identified thus far. We would very much appreciate any comments you can give us onj these issues, and on any problems you see that we may have missed. We are soliciting these comments prior to proposing the rulel, so that we need your comments by July 19, 1985, to use them in the proposal. I have enclosed a postage-paid, self-addressed envelope for your use in sending us your comments. Let me thank you in advance for your time 'in responding to this request. Sincerely, Enclosures cc: Don Barnes Mike Dellarco Suzanne RudzinsJ:i, Chief Chemical Regulation Branch 783322 GET4P 0105U Q Questions/Issues on the Dioxin/Furan iiist and Methodology 1. Is the rationale for developing the list of chemicals to be tested a reasonable one? What probiems, if any , do you see with it? 2. Are there other candidate chemicals th'at fit the rationale but are not on the list? If so, what are Ithey? 3. What types of problems are there with jthe scheme of providing only a skeleton of the analytical method and relyinq on industry to fill in the gaps? - Can they develop analytical standards? ijlow easily? - Can they do t h e appropriate cleanup? How easily? - will we be able to get reliable data and compare results? - If the answer to any of the above is no, what more do we need to provide? 4. Is 1 part pet billion a practicable limit of detection usir.q this method? If not, why? What limit is more appropriate? Why? 5- Are there laboratories available to perform these tests? If so, please provide a qualitative sense of tjow many, 6. Is the sampling plan and the Quality Assurance plan adeguate? If not, what needs to be added? ' 1 7. Are any available methodologies not addressed? If so, please identify them and their strengths and weaknesses. 8. Do you know of any available biological test inq systems which could serve as a screen for the chemicals to be tested? If so, please identify them and provide informatiojn you may have concerning sensitivity of the test and the rate of false positives/false negatives found. 783323 GENP 010512 l e> Table 1. Summary of Current Commercial Status and Use PRIOR" ITT CAS NUMBER CHEMICAL SAME CURRENT COMMERCIAL USE STATUS 'I PEST PEST PROD IMPT QNLT INTM CR L 58-39-9 'l,2,3,4,5,6-Hexachlorocyclo- hexaae, gamma Isomer I 70-30-4 /2 ,2'-Methy lenebis( 3,4,6-tri** chlorophenol) FS 87-36-5 Pencachlorophenol and sales S 93-72-L 2.4.5- Trichlorophenoxypro- panoic acid, all escers & sales FS 93- 76-5 2.4.5- Trichlorophenoxyacecic acid, all escers 4 sales FS 94- 75-7 (2,4-0ichloropnenoxy)acetic acid, all sales 6 eseers 94- 82-6 2.4- Oichloropheaoxybueyric acid, all sales & escers 95^77-2 3.4- Dichlorophenol 95- 95-4 2.4.5- Trichlorophenol y n n nn nny FS FS S I ILS-74-1 Hexachlorobenzene n nan I 120-36-5 2-[2,4-(Oichlorophenoxy)Jpropidnic acid, all sales & escers y yyn S 120-83-2 131-52-2 /2,4-Dichlorophenol Sodium pehtachlorophenaca y y a y y aya 136-25-4 2,2-Dlchloropropaaolc acid, 2-< 2,4,5-trichlorophenoxy)echyl ester n ny S FS FS 1 136-78-7 2-(2,4-01chlorophenoxyeehyl n n yn FS- sulfate), sodium sale 1 299-84-3 Phosphorochioic acid, 0,0,- n n yn FS dimethyl 0-(2,4,5-trichloro- phenyl) ester 783324 1 GENP 010513 > I Table 1* Summary of Currenc Commercial Status and Use PRIOR ITY CAS NUMBER 1 576-24-9 L 583-78-8 l 25167-83-3 Za -- CHEMICAL NAME '2,3-Dlchlorophenol H .,5-Dichloropheno 1 2,3,4,6-Tetrachloroptienol 'Tecrabromabisphenol-A dlacrylate (V) CURRENT COMMERCIAL USE STATUS '1 PEST PEST PROD IMPT ONLY INTM nnn 1 Y1 n n y Q1 Q y n CR S S S y nan S 2a -- Tetrabromobisphenol-B n n nn S 2a 79-94-7 /Tecrabromoblsphenol-A 2a 79-95-8 Tetrachlorabisphenol-A y yn n s n nnn s 2a 87-L0-5 3,4,5-Tribroaosaiicylanilide n i n n FS 2a 87-65-0 '2,6-Dichloroptienol Jyj n y S 2a 95-94-3 v1,2,4,5-Tetrachlorobeazene * y a y I 2a 99-28-5 2,6-0ibromo-4-nltropheaol n n n n FS 2a HS-79-6 '2,4,6-Tribrooopheaol 2a 344-07-0 Chloropencafluotobenzene 2a 392-56-3 Hexafluorobenzene 2a 488-47-1 * Tetrabromocatechol (V) 2a 608-71-9 'Pentabromophenol y nnn n nan n na n1 a a a 1 y nnn FS I- I S FS 2a 615-58-7 2,4-Dibroaophenol 1 y e a a a FS 2a 632-79-1 ^Tetrabroaophthalic anhydride y nn n P 2a 771-60-8 Pencafluo roaniline rLi a n a I 2a 1689-84-5 /3 ,5-01bromo-4-hydroxybenzo- nitrlle y1i y a y F 2a 2577-72-2 3,5-Dibromosallcylanilide n n n a FS 783325 GrENP 010514 Table 1. Summary of Current Commercial Statua and Use PRIOR- ITY CAS NUMBER * CHEMICAL NAME 2a 21850-44-2 4Tetrabrooobisphenol-A-bie- 2,3-dibromopropylecher (V) 2a ' 25327-89-3 'Allyl ether of tecrabroaobisphenol-A 2a 37853-61-5 BlamethyLether of tetrabroaobisphenol-A CURRENT COMMERCIAL USE STATUS i| PEST PEST PROP IMPT ONLY y nnn g CH S a ynn S n nnn S 2a 55553-62-3 1Tetrabromobisphenol-A-bts- y n n n S ethoxylate (V) 2b 1163-19-5 2b 32534-81-9 2b 32536-52-0 2b 42576-02-3 2b 42874-03-3 2c 2c -- *Decabromophenoxybensene * Pencabromodiphenyloxide * Occabroaodipheayloxide y ynn y ynn y i an a Methyl-5-(2,4-dichlorophenoxy)-2-nlcroben2 oate yp a y n 2-Chloro-l-(3-ethoxy-4nitropheaoxy)-4-(trifluoroaechyl)benzene a n y i n ALKYLAMINE TETRACHL0&0FHENATS N N ---- rwm 2-Chloro-i,4-diechoxy-5nitrobenzene n a -- -- S S s F I FS F 2c 82-68-8 4Pentachloronitrobenzene n nyn I 2c 88-06-2 2,4,6-Trichlor ophenol a ayn S 2c 92-04-6 2-CbIoro-4-iphenylphenol (V) a any P 2c 95-88-5 A-Chlororesorcinol y nn ,* P 2c 97-17-6 0-(2,4-Dichlorophenyl) 0,0- a * a y n F di-ethylphosphorochloace i 2c 97-50-7 '5-Chloro-2,4-dimethoxyeniline y 1 n n n P 2c 116-29-0 4-Chlorophenyl-2,4,5-tri- i n n y n P chlorophenyl sulfone 783326 p E N P 010515 I l Table 1* Summary of Current Commercial Status and U9e PRIOR- ITC CAS NUMBER 2c 299-85-4 2c 320-72-9 CHEMICAL NAME 0-(2,4-Dichlorop'henyi)-yme thylisopropylphosphoramido- thiaate CURRENT COMMERCI AL USE STATUS '] PEST PEST PROD IMPT ONLY INTM R n -n y a F . 3,5-Oichlorosalicyclic acid n n S 2c 470-90-6 2-Chloro-l-(2 ,4-dichloro- a yya F phenyl) vinyl diethyl phosphate 2c 500-28-7 0-(3-Chloro-4-nitrophenyl)- a nya F 0,0-dimechylphosphorochioace 2c 6 L3-67-3 Chlorohydroqulnane n nan P 2c 933-75-5 2,3,6-Trichloropheaol n 2c 961-11-5 2-Chloro-l-(2,4l5-trichloro- y phenyl)vinyl dimethyl phosphate S' F 2c 1757-18-2 0-(2-Chloro-l-(2,5-dichloro- n a y n F phenyl)vinyl) 0,0-diethyl phosphorochioate 2c 1836-75-5 2,4-Oichlorophenyl-pnlcropnenyl ether n nyn F 2c 1940-42-7 4-Bromo-2,5-dichlorophenol n n a y FS 2c 1982-47-4 3-(4-(4-Chloropheaoxy) phenyl)-!,1-dimethylurea ni n y a P 2c 2104-96-3 0-(4-Bro*b-2,5-dichlorophenyl) 0,0-dimethyl phosphorochioate n n ya F 2c 2300-66-5 3 ,6-Dichloro-o-anisic acid, y dimechylamine sait (V) n y n FS 2c 2463-84-5 0-(2-Chloro-4-nitrophenyl) a 0,0-dimethylphosphorothioate nyn p 2c 2675-77-6 1,4-Dichloro-2,5-dimechoxybenzene yn p ii : GENP 010516 783327 Table 1. Summary of Current Commercial Status and Use - PRIOR ITY CAS NUMBER 2c 3380-34-5 CHEMICAL NAME J5-Chloro-2-(2,4-dichloro- phanoxy)phenol 2c 3772-94-9 Pentachlorophenyl laurate CURRENT COMMERCIAL USE STATUS PEST PEST PROO IMPT ONLY IMTM y ann CR FS n nnn FS 2c 5736-15-2 2,2 '-Me thylenebls(3 ,4 ,6-tri- n n y n chlorophenol), monosodium salt FS 2c 19666-30-9 2-terc-3utyl-4-(2,4-dichloro- n 5-tsopropoxyphenyl) delta1,3,4-oxadiazolin-5-one y yn I 2c 37853-59-1 '1,2-8is(tribromophenoxy)ethane y 2c 41193-08-7 0-(4-3romo-2-chlorophenyl) 0-ethyi S-propyl phosphorothioate n nnn yy n FS F 2c 43121-43-3 l-(4-Chlorophenoxy)-3,3- n dimethyl-l-( 1H-1,2 ,4-triazol- l-yl)-2-butanone y yn P -- 21923-23-9 Chlorthlophos a nyn 3-- 3-- 3-Chloro-4-fluoronitrobenzene y 2-Chlor0-4-fluorophenol a nnn n ---- I F 3-- 3-Chloro-4-fluoropheool n n---- F 3 N-2-Chloro-4-trlfluoromethyl y n yn I phany1-DL-vs line(+)-cyano- 0 (3-phenoxyphenyl)methyl ester 3 Class Phthalocysnine dyes and y ynn I pigments 3 50-31-7 2,3,6-Trichlorobenzoic acid n n ny I 3 56-23-5 Carbon tetrachloride y y nn I 3 75-01-4 Vinyl chloride y y nn 3 76-13-1 1,1 ,2-Trichloro-l,2,2-tri- y n n n I fluoroethane 3 78-88-6 2,3-Olchloropropene C y nny I 783328 I 1 i Table 1. Summary of Currene Comoercial Status and Uae i - PRIOR ITY CAS NUMBER 3 79-00-5 3 79-01-6 CHEMICAL AME 1,1,2-Trichloroechana 1,1,2-Trichloroachyleae 3 79-27-6 l,L,2,2-Tetrabromochae 3 79-28-7 1,1,2,2-Tecrabromoethylene 3 79-34-5 Ttitrachlocoechane CURRE NT COMME rcial! USE STATUS PEST PEST PROD IMPT! ONLY INTM yin' n i n y y 'n y y n n .n y n nn i n ynn CR I I _ I I 3 85-22-3 Pencabromoechylbeozene y nnn I 3 87-6L-6 L,2,3-Trichlorobnzene (V) y y n n I 3 87-68-3 Hexachlo robutadiene n nnn I 3 87-83-2 Pencabromotolueno y n nn I' 3 87-84-3 Pencahromochlorocyclohexane y nan P 3 89-6L-2 l,4-Dlchloro-2-nicrobanzene y nn I 3 89-64-5 4-Chloro-2-nitrophenol n ynn P 3 89-69-0 2,4,5-Trichloronicrobenzene n n ---- I 3 91-08-7 Tolucn-2, 6-diiaocyanate y yaa _ i 3 93-65-2 2-(4-Chloro-2-achylptanoxy) y yny F proplonifc acid 3 94-74-6 4-Chloro-o-toloxy aeacie aeld y yay P 3 94-81-5 4-(2-Mthyl-4-chlorophanoxy) n n- a y P bucyric acid 3 95-50-1 ^Dichlarobanzana y yay I 3 95-56-7 o-Bromophenol y n n a P 3 95-57-8 o-Chlorophenal n il y n y P 3 95-76-1 3,4-Oichloroaniline y iy n y s I 3 96-18-4 l, 2, J-Tr ichloropropane y 1* n i n I 3 96-19-5 1 ,2,3-Trichloro-l-propene y i n n1 n * I 6 7833PQ ncMP Table 1* Summary of Current Commercial Status and Use PRIOR" ITY CAS NUMBER* 3 97-23-4 3 99-30-9 CHEMICAL NAME 2,2'-Methylaaebts< 4-chloropheaoi) 2,6-Dlchloro-4-nltroaalline CURRENT I COMMERCIAL USE STATUS 1 PEST PEST PROD IM?,T ONLY INTM yj y " n n i y ynn CR I P 3 99-54-7 1,2-Dichloro-4-nitrobemen* y yny X 3 102-36-3 3,4-Dichlorophenylisocyanate y any I 3 106-37-6 p-Qibromobenzeue 3 106-46-7 p-Diehlorobeamene y ynn i y Any I I 3 107-05-L 3-Chlorapropeae y any I 3 LQ7-06-2 1,2-Dichloroethane y yny 3 108-70-3 1,3,5-Trichlorobanzene (V) f y, n n I 3 108-90-7 Chlorobenzene y n' n y I 3 LI7-08-8 Tetraehlorophthelic anhydride y nn I ( i 3 L17-18-0 1,2,4,5-Tacrachloro-3-nitro y n nn I benzene 3 117-80-6 2,3-Qichloro-l,4-naphthalena- a yya P dlone 3 118-75-2 * 2,3,5,6-Tetrachloro-2,5- n na 5 cyclohexadiene-1,4-dione iy 3 120-82-1 1,2,4-Triehlorobenzene y yny I i 3 127-18-4 1,1,2,2-Tecraehloroethylene y yan -1 3 147-82-0 2,4,6-Tribroaoeniline (V) y an I i 3 320-60-5 2,4-0iehloroben2otrlfluorlde y j y n y I 3 328-84-7 3,4-Di chlorobeozo trifluorlde a i y n y I i 3 330-54-1 3-(3,4-Oichlocophenyl)-l,1- y i y y n I dimechylurea i i 783330 i nTJTvTP 010*1 o t I 1 li i l Table 1. Summary of Curreac Commercial Status and Use i - PRIORITY CAS NUMBER 3 348-51-6 # CHEMICAL NAME o-Chlorofluorobenzene CURREJrr COMMERCIAL USE STATUS PEST PEST PRODjIMPT ONLY INTM y ;n n n 3 367-12-4 o-Fluorophenol y ann 3 513-31-5 2 ,3-Dibromopropylene n n-- 3 527-20-8 Pencachloroaailine a a 1-- CR I P I I 3 542-75-6 1,3-Dlchloropropene y ny i n i 3 555-37-3 l-Sucyl-3-(3, 4-dichloro- a yy* phenyl)-l-echylurea I -- 3 558-13-4 Carbon cecrabromlde y nan I 3 584-84-9 Toluene-2,4-diisocyanate y ynn 3 608-93-5 Pencachlorabenzene n nny --- I 3 626-39-1 Tribrooobenzene y .a a 3 709-98-8 3 ,4-Dichloropropioanilide y nya I I 3 827-94-1 2 >6-01broao-4-nlcroeniline y n n n | 1 3 1L94-65-6 2v6-Dichlorobeozonitrile a y y ri 1 3 1344-32-7 Trichlorobenzyl chloride n n 1y 3 1435-53-6 2,4-Oibronofluorobenzene aa -- P I I I 3 . 1861-32-1 3 1897-45-6 3 13041-24-9 3 13463-67-7 3 15086-94-9 Dimethyl tecrachloroterephehalace t Te crachloroisophtbeloni trile 3 ,4-Dichlorobenzocrichloride y y n Titanium dioxide Tetrabroaofluoreeeela and disodium salt y y aya e1 \ 1 yy a 1 1 n y nn ii* \ 11 y ci n ii I I I I S 3 20925-85-3 Pntachlorobenzonitrlle n n-- I 3 35915-19-6 Dichlorobenzoic acid n nnn I GF/NTP nm*TM 783331 Development of the List of Chemicals to be Proposed for ^estira OPTS began bv compiling existina lisjts of chemicals potentially contaminated with chlorinated) and/or brominated dibenzo-p-dipxins and/or d iber.zofur ans (DRDs/nj^Fs) . The "workina" list was created from several published and unpublished ex istina lists. They are: - Dioxins (Rsposito et.al. 1990) F pA 600/2-Rp-l7. - Dioxin Rtrateay (TISERA 1993). | ,| - Informal list prepared bv Office of Pesticide Programs (o d d ) Literature searches were made for al!. known reaction conditions uder which DRDs/DRFs have been formed. These reaction conditions were checked aqainst chemicals wi_t_h]known contamination, and mechanisms of formation were described as diagrammed on the following oaqe. Rased on its assessment of reaction conditions and mechanisms of reaction for the formation of haioaera ted ur ans and dioxins for chemicals knownm to be contaminated with these materials, it became clear that DRDs/DBps share most the same "nr ecur sor" compounds. The exceptions are: crilou ina ted/ brominated 2-hydroxyphenyl ethers which form onlv dioxins , and chlorinated/ brominated biphenyls which form only diber.zo' furans. The other precursor compounds identified are: or tbo Chlorophenols, Chlorinated benzenes, Polychlorinated oher.vi ethers, Chlorinated phenoxvphenols and Polychlorinated hydroxydibenzofurans. Although brominated compounds are no t listed or specifically discussed, the same type of chenic al reactions, qenerally reauirina less^ severe reaction cor.d itions , will occur with brominated compounds. tJsinq these theories on the mechanism of formation haloqenated furans and dioxins, OPTS reviewed all chemicals on the 11work ina" list for which actual monitor ina data do mot exist to determine whether the theories would apply ar.d indicate that there is a potential for formation of nRDs'/DRFs The theories were also used to review each chemical listed in the Directory of Chemical Producers (SRT 1994), and additional chem reals potentially contaminated with DBDs/DRFs we're added to the "workinq" list. Chemicals bel ieved to be precursors to DRO/DBF formation were also added. To prioritize the list of chemicals according to probability of DBD/D3F contamination, those chemicals with actual monitor inn data showing contamination were designated group 1. Group 2 chemicals are listed because the theories apply. Generally these chemicals have molecular structures similar to the chemicals known to be contaminated andjare ;nown to be processed under conditions most favorable to dioxin/furan formation. The favorable conditions include (lj chlorination or brominatior. at elevated temperatures (i.e.i >100C) or pressures 783332 nXlKTP 0 1 0 ^ 9 1 ni 2 and (2) subseauent alkaline treatment o F the hemica1 intermediate formed (e.g., for product wa shina this treatment results in the addition of oxygen to the phenyl molecule which in turn leads to the formation of dioxins/furans) This larne group of chemicals has been subdivided as follows: 2a Polyhromir.ated nher.o.ls and dec.ivatijves. These chemicals ace prepared under the same conditions as the chloronhenols where dioxin contamination has been documented. 2b ^olvbtom in a ted nhenvl ethers and d e r c ivatives. These chemicals are nreoared bv bromine tip r. of dinhenvl ether under vigorous conditions. 1 2c Chloronhenol pesticide derivatives similar in structure to di- and tr ichloroohenols and derivatives', and d i- and tr ichl oronhenvl ethers and derivatives f(pr which dioxin contamination has been documented. |Thes<p chemicals are preoared under conditions verv similar to those used for the chlotophenois for which dioxin contamination is d o c u m e n t e 1. 2d Compounds produced from hexachlorobenzene which is known to be contaminated and other fully haloqer.ated benzenes and any chemical that nay contain hexachloro- or hexabronober.zene as an impurity. Group 2 chemicals are those ootentially contaminated or that may lead to contamination according to theories on the formation of nBDs/ORFs. They have structures similar to |chemicals known to be contaminated, hut without sianificantly more process data the probability of contamination cannot be ascertained. For aroun 4 chemicals the probability of contamination is highly speculative and for many chemicals,1 improbable. These chemicals appeared on other published lists; however, existing theories do not apply. Borne of these chemicals are structurally similar to chemicals known to be contaminated, Ihut process information indicates that formation of npios/nBFs is unlikely. for chemicals in groups 1, 2 and 3, and where possihle in group the route of contamination (CP) was idjentified. nr.like the contamination pathway, which is a chemical 'reaction sequence, the route of contamination is the physical! mechanism or source of contamination. The route of contamination! is designated and defined as follows: - p This chemical is a precursor to DRD/OBF contamination; the chemical itself, as produced, is noti contaminated . Industrial use .of the compound, under the proper DRn/nRF formation conditions, could resu It in the formation of DRDs/DBFs in. final products. I 783334 I1 GENP 010523 I - This chemical Is contaminated only as a result of a contaminated feedstock chemical (i e., either the raw starting material or an unisolated intermediate). The chemical's synthesis does not lead to further o r d /o b f formation. T h i s chemical is contaminated as a result of ORD/o r f synthesis durir.c the chemical's production. This chemical may in turn he a feedstock for further chemical, production. pS This chemical is contaminated as a |result of both a contaminated feedstock and further synthesis of rmns/^BPs dur ihg nroduction. - I This chemical contains an impurity hat may he contaminated or lead to contamination with DBDs/OBFs. The noss 1hie level of co n tamination is expected to he verv low. I I The complete list of chemicals was reviewed within and chemicals ir. qrouos 1 and 4 were eliminated from consideration because OPTS believes they have a very low probability o f contamination. Prom the tema in ino group of chemicals, those believed to he contaminated via a contaminated impurity (T) were carefully examined. As discussed above in the definition of impuritv (I) , the contamination of these chemicals is expected to be veLv low compared to chemicals contaminated via othIer ro1utes. There fore the chemicals marked X were dropped from consideration. The oroun of chemicals marked (for precursor to dioxir./ furan contamination) were further examined. As discussed under the definition of precursor (p ), these chemical|s are not themselves contaminated, but may lead to dioxin/furan contamination. Therefore they were removed from the list designated for testing, but were placed on a separate list for which process and use information will be Ireaui r e d . I Coordination with OSW followed-and cHemica Is listed bv OSW as "good" or "possible" candidates for wastestr earns contaminated with DBDs/ORFs were added to the OPTS list', and duplicates were eliminated. This consolidated list contains chemicals known to he contaminated with nRDs/ORFs (l), chemicals' with molecular structures and process conditions favorable to fiRD/ORF formation, but without actual monitoring data (2), and chemicals with molecular structures favorable to DRD/DRP formation, but not enough process data to make a determination (3). This list was examined for those chemicals used only as pesticides (PO), since TSCA has no jurisdiction over these. The chemicals marked with a "v" in the "PO" column are used only as pesticides. They will he 783335 GENP 010524 4 eliminated from the OTS list and nicked up bv O p p in their data call-in program. Some chemicals or. the list are not currently in production, but could come into oroduc tion at any time! While they have not been removed from the list , manufacturers will ipot be reauired to test a chemical not curren tlv in production, at the time the rule is promulgated. He will, however, be reauired o test that chemical should he ever re sume production. i i i i I 783336 GENP 010525 DRAFT MAY 5 1985 GUIDANCE FOR A SEQUENTIAL APPROACH TO THE SAMPLING OF DIBENZODIOXINS AND DIBENZODIOXINS INTRODUCTION ' This is the sequential sampling scheme referred to in the Proposed Rule for Processing, Distribution, and Use of Certain Designated Chemical Substances Suspected to Produce and/or Be Contaminated with Dioxins and Furans (49 FR 281721 28182, July 10, 1984). For this proposed rule, dioxins and furans are in reality four classes of compounds: brominated dibenzofurans, brominated dibenzodioxins, chlorinated dibenzofurans, and chlorinated dibenzodioxins, and specifically within these classes the compounds substituted with from four to seven chlorine and/or bromine atoms per molecule. Hereafter these[particular compounds will collectively be referred to as polyhalgenatecl dibenzofurans tPHDFs) and polyhalogenated dibenzodioxins (PHDDs). Specific subgroups within the collective group addressed by the regulation will be more specifically named. j . This document contains a statement of the sequential decision scheme, the rationale for its use, the details of sample selection (how much, when and whe r e ) , and the statistical properties of the sequential decis ion scheme! This scheme is part of an overall activity to provide guidance to those who are in a- position to respond to the proposed regulations. 1 PROPOSED TRUNCATED SEQUENTIAL DECISION SCHEME For a given material at the production/formulation site, select a saimmpnl1e aannrd? mm en aa sc un rr eo tt-hhe concent rr*aat iion of PDHTDnFPse aa nn Hd PHDDs The sampling scheme must be used for Iall ests and analytical evaluations for determination of PHDFs and PHDDs. A measurement may be obtained by a screening method which has been demonstrated by the tester to a documented degree of sensitivity and specificity. The confirmatory method for identifying the PHDFs and PHDDs and the concentrations of these compounds is high resolution gas chromatography / high resolution mass spectrometry (H R G C / H R M S ) j The product sample selection procedure is repeated until either a product sample is found to contain PHDFs |and/or PHDDs or seven samples have been processed, in which case declare the production of the sampled material to be free from PHDFs and PHDDs. If any of the concentrations of PHDDs' in a single product sample exceeds 0.1 parts per billion (ppb) and/or Ithe concentra tions of PHDFs in a single sample exceeds 1.0 parts per billion (PPb)# the product will be deemed to contain |pHDDs' and PHDFs respectively. In HRGC/HRMS a resolvable chromatographic peak is DRAFT 3 E N P nirKo/r 783337 -*- Dr a ft considered a single brominatcd dibenzofuran jcompound, a single chlorinated dibenzofuran compound, an single brominated dibenzodioxin compound, or a single dibenzodioxinj compound. The number of samples, N, required is not to exceed seven. REASON FOR NOT USING A SINGLE SAMPLE TO DETERMINE COMPLIANCE i ] Analysis of a material is subject to two mai!n sources of PHDFs and PHDDs being generated as byproducts or jimpurities may vary over time and/or location at the site. [ Secondly, the measured concentration of PHDFs and PHDDs will vary among repeated measurements on the same sample. Therefore, the measured concentration of a particular single screening test or single gas chromatographic peak will vary among samples selected randomly in time and space so that a single |sampl!e would not be representative of the underlying process except for the extreme cases. In tne following paragraphs the variable nature of observations will be expressed by the symbol the probability that a sample will have a measured concentration below 0.1 ppb for PHDDs and 1.0 ppb for PHDFs in a particular cas chronato- graphic peak. JUSTIFICATION FOR USING SEQUENTIAL SAMPLING Given the analytical costs associated with HRGC/HRMS, it is important to utilize statistical procedures which minimize tne number of samples which need to be analyzed, Sequential monitoring of a material allows a determination of the presence of PHDFsand/or PHDDs to be based on fewer samples, on the average, than would be required under an equivalent fixed Isample size decision rule because of the large number of samples required to descibe a production having undemonstrated variability of PHDF and PHDD content. For each such screening test |or HRGC/HRMS analysis this efficiency increases with the number of potential-PHDFs and PHDDs in the material and decreases as the magnitude of e increases. For materials containing sufficijentlyj high concentrations of even a single PHDF or PHDD compound (i.e., e is close to zero for its associated screening est or peak) only one sample will be required to declare the presence of PHDFs or PHDDs. For most multiple congener materials, the! use of the proposed scheme would probably result in the analysis of only 4 or 5 samples to identify materials containing PHDFs at 1.0 ppb and PHDDs at 0.1 ppb. Even for materials containing a single congener at levels above the 0.1 ppb, the application of the scheme would probably result in the analysis of fewer than 7 samples to declare the presence of PHDFs and! PHDDs when e is no larger than 0.7. 783338 DRAFT GENP0J0527 ' DRAFT TIMES, LOCATIONS, AND QUANTITIES | A material may be produced through a group jof activities, within one or more locations at a site. The production of the material is' assumed to be cyclic in nature [and that within each cycle the distribution and concentration ra'nge o'f brominated and chlorinated dibenzofurans and dibenzodioxin's is [consistent. Each cycle, then, is a time interval representing a period within which the activity mix of the production of the .'material is essentially repeated, such as a batch, shift, or day. Samples under the proposed sequential sampling plan will be selected from separate cycles of the production process. Further description of sample selection appears in ("Guidance for Sampling Brominated anti Chlorinated Dibenzofurans and Dibjenzodioxins." It is important to randomly select the sampling time and location independently for each cycle. Any deviations from this random selection must be documented and justified, AS many as seven cycles will be required under tne proposed procedure. .Changes in the production process, especially changes judged to potentially affect brominated and chlorinated dibenzofuran and dibenzodioxin levels, require .that__the entire sequential testing effort be redone. If a single production process is the potential source of brominated and chlorinated dibenzofuran and t^ibenzodioxin exposure through several media, each medium should be separately tested by the sequential sampling. The size of a sample to be chemically analyzed depends on the medium of the sample and the intended analytical method. Sample sizes are recommended in the "Guidance for Sampling Brominaced and Chlorinated Dibenzofurans and Dilienzodioxins." STATISTICAL PROPERTIES OF THE PROPOSED DECISION SCHEME In general, two dcision errors are possible: (1) declaring materials in compliance to be not in compliance,! and (2) declaring processes which are not in compliance to be in compliance. Choice of the 0.1 ppb as the test threshold and requiring two samples to exceed it eliminates any significant livelihood jof committing an error of the first type. The probability of committing an error of the second type decreases with the number of {potential bromi nated and chlorinated dibenzofurans and dibenzodioxins congeners present in the process and the maximum number of required samples, and increases with the magnitude of e for each such peak. The proposed maximum number of seven samples was chosen because it results in an acceptable probabilityjof the second error without the requirement of an excessive amount of samples to be analyzed to determine the presence of PHDFs and PHDDs. For, the 3even samples, the error of the secondtype will be less than! 0.25 for a typical process (i.e., one having 5 or more peaks) provided the e-values for the process are all no larger than 0.92 (assuming unconditional independence or complete gas chromato graphic resolution of individual PHDF and PHDD concentrations). nDACT O T~?~VT-I-V A- ^ _ 783339 e DRAFT GUIDANCE FOR SAMPLING BROMIHATED AND CHLORINATED DIBENZOFURAHS AND DIBENZODIOXINS' The following are recommended steps for collecting samples the following four classes of compounds: brominated dibenzo furans, brominated dibenzodioxins, chlorinated dibenzofurans, and chlorinated dibenzodioxins. Of specific interest! within these classes are the compounds substituted with from four to seven chlorine and/or bromine atoms per molecule, j Hereafter these compounds will collectively be referred to as poiyhalgenated dibenzofurans (PHDFs) and polyhalogenated dibenzodioxins (PHDDs). Specific subgroups within the collective group addressed by the regulation will be more specifically named. All of the following steps should be carefully recorded. 1. Determine the Objectives of the Sampling Exercise. In the case of this proposed rule, some objectives might be (1) to measure PHDF and PHDD levels occuring' as byproducts or impurities in a product to determine compliance with the rule, and (2) to know the limitations of these measurements. 2. Describe the System or Production Process from Which the Sample Materials Will Be Collected. Tne samples collected can only provide information about the product from which they have been collected and dhe completeness of this information depends on how, at what ltime( s ), where, and for wnat duration the samples are collected .j Of particular importance in determining a sampling plan for the process are the continuity, homogeneity, and cyclic characteristics of the production process. Compositing of samples |for chemical analysis nas benefits in reducing analytical costs but ris'ks in requiring lower instrument quantification levels and reducing the ability to identify whether PHDFs and PHDDs are from uniform concentra tions in all members of a composite or high jconcentrations in one member and no concentrations in the other members! of the composite. Each of the seven samples described in the "Guidance for a Sequential Approach to the Sampling of Dibenzofurans and Dibenzodixins" may be a randomly selected composite, but the seven may not be composited unless the levelj of concern is revised to be one seventh of the currently proposed levels, i.e. 0.1/7 or 0.014 ppb fo PHDDs and 1.0/7 or O.lk ppb for PHDFs. 3. Prepare for Sample Collection Operations. For this section and the following section t is important to first read the appendices from the document "Analytical Methods f o r .Brominated and Chlorinated Dibenzofurans and Di benzod ioxins" 1 783340 DRAFT G NP 010529 ' DRAFT ( a. Methods and Equipment Sample collection methods, to be applied here, depend on the nature of the material to be screened or! chemically analyzed and the screening or chemical [analytical procedure, which for high resolution mass spectrometer [detector (HRGC/HRMS), here is assumed to include extraction. Proper guidance, such as sample preparation, extraction, and brominated and chlorinated dibenzofurari and jdibenzodioxin measurement by high resolution gas chromatography with a HRGC/HRMS, for collected samples appears in ["Analytical Methods for Brominated and Chlorinated Dibenzofurans and Dibenzodioxins" b. Accounting for Potential Sources of Contamination or Loss During Sampling. j Sample collection containers, sample collection equipment, and cnemical analytical equipment! are common sources of contamination. Rigorous cleaning of containers and equipment coming in contact with samples] and sample extracts decreases the possibility of contamination. The use of method blanks and field blanks evaluates the entire collection and extraction procedures for contamination. Field spikes should be used to determine the! possibility of losses of brominated and chlorinated dibenzofurans and dibenzodioxins between collection and actual screening or chemical analysis. Coordination among the sampling plan designers, the sample collectors, and the screening testers' or chemical analysts is a useful way to obtain information on accounting for potential contamination or loss during sampling. c. Preparing an Operational Plan Which Describes the Details of the Proposed Sample Collection Procedures. The actual sample collection operational procedures may differ from the proposal and the differences] and reasons for the differences must be recorded. A Sampling Plan whicn describes the sample size, location selection and time selection procedures, and quality assurance guidance, appears in the next section. | 4. Prepare a Sampling Plan and Collect Samples* Professional judgement is very important in sampling Frequently using such judgement can isolate ja worst-case situa tion, sample the situation, and determine that th!e magnitude of the worst-case situation is below the stated level of concern. Professional judgement may provide information to! more completely sample a system by focusing attention on steps in a process which potentially produces PHDFs and PHDDs For t'hose complex systems m A1 r\^ ^ 0 783341 d r a ft professional judgement must be assisted by proven sampling methods. To make a complete analysis, sampling should also determine the presence or absence of PHDFs and PHDDs in locations or situations not necessarily -judged to be most important. The proposed rule requires random sampling. A random sample selection system includes all possibilities and systematically chooses from among a list or compilation of those possibili ties. The way of choosing from among the possibilities is the use of a random number table or through an automated random number generator. A haphazard selection system, the result of selection through convenience and/or purposeful neglect, includes only a portion of all possibilities, has very little estimating power, and should not be considered a random selection procedure. For the purpose of incorporating this sampling plan with the CPA "Guidance for a Sequential Approach to the Sampling of Brominated and Chlorinated Dibenzofurans and Dibenzodioxins" the sampling unit for each process will be a cycle. A cycle is defined as the time interval representing a period within which the activity mix of a process is essentially repeated. This cycle may be a batch, a shift, a day, or some other time period. It is important to use random selection procedures in deter mining the sampling time and location for each individual cycle (the details for these determinations are' given below in parts b and c ) . Cycles should be sampled, one at a time, until as many as seven cycles have been sampled. If two samples from any of the seven or fewer cycles are analyzed and show measurement levels of PHDFs and PHDDs above the limit of quantitation, the material is not in compliance with the rule. If none of the seven cycles show PHDDs above the 0.1 parts per billion (ppD), or PHDFs above 1.0 ppm, no further sampling is necessary until there are changes in the process as described below. If there are changes known or suspected to change the generation of PHDFs and PHDDs as byproducts or impurities in a process or a cycle, the seven cycles sampling exercise must be done again. For each defined process and independent cycle: a. The size of a sample of a material to be analyzed or sample will be determined by the kind of material to be sampled and the sensitivity of the analytical method. If the material is gaseous, a sample size of ten cubic meters is usually sufficient for analysis. If the material is aqueous or liquid, the sample size of one liter is usually sufficient for analysis. If the material is a solid, the sample size of one hundred grams is usually sufficient for analysis. Any sample size must have proper documentation demonstrating capability to measure PHDDs at 0.1 ppo and PHDFs at 1.0 ppb. b. Location(s) of sample collection are defined as a subset of all possible locations of storage of a given material. p m a _ G E N P 010531 783342 DRAFT Physically possible sample collection sites are listed and chosen through random selection procedures. c. In processes where over seven batches of a material are produced in a year, sample collection times must be selected from a random time period within each individual batch production cycle. The cycle time is completely segmented into intervals long enough to collect a sample and one interval is randomly selected for the actual collection. For processes where under seven batches of material are produced, random selection of seven samples from a list of production time segments for all batches. d. Quality assurance of the sampling and actual sample collection exercise includes documentation of: the sampling plan, sample collection procedures, and any chain of custody records. Special notation should be made of any deviations from orginal plans and p r o m o t i o n s recorded at an earlier stage of the sampling exercise. Especially: i. dates and times of sample collection and chemical analysis of samples. ii. exact location and time of sample collection. iii. process or product batch or lot identification iv. chain of custody records employed in the sample collection and ultimate disposition of the samples. DRAFT GENP 010532 783343 DRAFT QUALITY ASSURANCE PLAN FOR MEASUREMENT OF BROMINATED AND CHLORINATED DIBENZOFURANS AND DIBENZODIOXINS For quality assurance (QA) to be an effective method of mea surement validation of qualification, a Quality Assurance Plan (QAP) must be prepared and should include the following: history and disposition of samples; sampling and sample collection procedures? and extraction and instrumental analysis procedures. The QAP documents how a laboratory intends to demonstrate its capability to produce data of acceptable quality. The proposed rule is concerned with the following four classes of compounds: brominated dibenzofurans, brominated dibenzodioxins, chlorinated dibenzofurans, and chlorinated dibenzodioxins. Of specific interest within these classes are the compounds substituted with from four to seven chlorine and/or bromine atoms per molecule. Hereafter these compounds will collectively be referred to as polyhalgenated dibenzofurans (PHDFs) and polyhalogenated dibenzodioxins (PHDDs) An accurate trace or history of the life of a sample (to be chemically analyzed for PHuFs and PHDDs must be assembled and should contain information beginning with a description of the system, scheme, or survey design for sample collection. A written record, sometimes called a chain of custody form, shall follow the sample. Necessary contents of the record are written general descriptions of what happens to the sample, schedules and timetables, disposition, and handling. Following final chemical analysis, the last entry in this trace should be the disposition of the sample. Any further use (particularly for another activity), movement, or examination of the sample should be added to the history. Details of the sampling or sample collection procedure are the second section of the QAP. Since the history describes handling disposition, schedules, and general descriptions of what happens in sampling, the requirement here is a detailed descrip tion of sampling and sample collection. Reasons for using a specific or general sample selection process and reasons for not using other processes must be included here. Estimates of how well the selected samples represent the material to be characterized is an essential part of quality assurance. The greater the variability of composition of a material, the more frequently sampling is required for estimation and characteriza tion purposes. In determining or estimating errors generated in sample collection, control samples or blanks, and PHDFs and PHDDs (native compounds or, preferably isotopically labelled compounds) reinforced controls or standards, must be sent to the collection site(s) and returned with the samples for identical handling' and treatment. Duplicate samples, which are collected, documented, and handled the same as other samples, are necessary for recovery, precision, and accuracy determinations. The third section of the QAP is a description of the extrac- GENP 010533 783344 DRAFT tion and chemical analysis or screening test procedures. To determine both extraction efficiency and measurement efficiency, it is necessary to use chemical standards of PHDF and PHDD mixtures o r _individual PHOFs and PHDDs comp'oundsj in control samples. Once capabilities have been established, the operator must determine precision and accuracy must be determined and use those determination to designate acceptable! bounds for analytical performance. The operator must keep control chart records to assure that instrument readings of standards fall within the range of acceptable performance. The operator must establish and describe the quantitative range of an instrument and analytical procedure. Specific requirements are as follows: For chemical analysis, these procedures must be denontrated capable to reproducibly and repeatedly quantitate PHDDs at 0.1 parts per billion (ppb) and 1.0 ppb of PHDFs. Quality control check sample] analysis begins the determination of system capabilities. For PHDDs, at least two analyses of the same spiked standard (instrument standard) at 0.1 ppb must be quantifiable to within +10% of 0.1 ppb. |The recovery of standard spiked into a product material at 0.1 ppb and run through the entire chemical analysis must be within 80-120% of the amount spiked, with reference to the instrument standard, |ninety-nine percent (99%) of the time. For PHDFs, at least two analyses of the same spiked standard (instrument standard) t 1. ppb must be quantifiable to within +10% of 1.0 ppb. |The recovery of standard spikecT into a product material at 1.0 ppb and run through the entire chemical analysis must be within 80-120% of the amount spiked, with reference to the instrument standard, ninety-nine percent (99%) of the time. Qualitative requirements include: response factors for PHDFs and PHDDs to be measured, instrument hardware and operating conditions (including type and source of columns I carrier gas and flow rate, operating temperature range, and|ion source tempera ture) , and structural assignment and/or library matching of mass spectra. For both qualitative and quantitative measurements, the instrument operator should be blind to the nature or source of samples, particularly to duplicates, blanks! and|brominated and chlorinated dibenzofuran and dibenzodioxin reinforced samples. The limit of detection (LOD) and limit of quantitation (LOQ) shall be described for each material. Tentative|LOD and LOQ definitions are greater than or equal to three times background noise and greater than or equal to ten times background noise, respectively. Details of quantitative caliiration procedures for the known and/or expected range of the PHDF and HDD levels in actual samples complete this section of the QAP. nDACT (t F A T P 783345 DRAFT The last section of the QAP must include the results of laboratory participation in round robin analytical progra n s , the results of performance audits, the results of systems audits, analytical result of performance audit samples, persons responsible for all aspects of sampling, chemical analysis, data analysis, corrective actions, and quality assurance/quality control. This section also must include a description of how problems are handled and documented, and how corrections in working level notebooks are indicated and explained. DRAFT 783346