Document omXJvy2EEGmzRweQxJ1Zj7BDw

AR226-2556 June 23, 2000 Dr. Charles M. Auer, Director U.S. Environmental Protection Agency Office of Pollution Prevention and Toxics Chemical Control Division 401 M Street NW, Room 403 Washington, D.C. 20460 Dear Dr. Auer: As you requested in your April 19, 2000 letter and during our subsequent meeting on May 1,2000, attached is a summary ofDuPont's U.S. uses of Ammonium Perfluorooctanoate (APFO, CAS# 3825-26-1) as a fluoropolymer reaction aid including releases from DuPont site and the fate of APFO in fluoropolymer dispersion products, a summary of industrial hygiene data collected at our U.S. fluoropolymer manufacturing site, and a summary of the employee blood data from a site in the U.S. A summary of the toxicology available to DuPont was sent under separate cover on May 26, 2000, It is important to emphasize the following points: > DuPont does not manufacture APFO. All APFO used in our processes as a reaction aid is purchased from an outside supplier. > Most of the APFO used is removed from the fluoropolymer products before they are sold to outside customers. A relatively small amount of APFO (worldwide, in the U.S.) leaves DuPont facilities in fluoropolymer dispersion products, > Of the APFO in the products sold. most (>97%) is destroyed during customer processing to a non-carboxylated hydrofluorocarbon, > All of the U.S. DuPont operations that use APFO with significant exposure potential are concentrated at one location; Washington Works in Washington. WV. Therefore, most of the industrial hygiene data and blood serum data presented in this document are from that location. EID156211 Dr. Charles M. Auer, Director U.S. Bnvironmental Protection Agency page 2 > Extensive industrial hygiene data collected on workers potentially exposed to APFO show airborne exposures to be significantly below the ACGM TLV of O-Olmg/m38 hr. TWA Exposure levels of plant workers have dropped significantly since the conversion to an APFO solution from a dry powder. > As part of the ongoing surveillance of workers potentially exposed to APFO, in March and April of this year a series of blood sampleswere taken from workers in the U.S., The Netherlands and Japan to be analyzed for serum APFO concentration. DuPont has not received the results from our contract laboratory at this time. DuPont will submit a summary of the results when they become available. The format of the information in the attached is a modified UEIP format. If you wish to discuss the information contained in the attachment, please contact Robert F. Pinchot at (302) 999-4074 or e-mail at Robert.F.Pinchotfg.usadupont.com or me at (302) 366-5259. Very truly yours, Gerald L. Kennedy Director, Applied Toxicology and Health EID156212 Voluntary UEIP, Aniinonhnn Perfluorooctanoate Voluntary Use and Exposure Information Profile Ammonium Perfluorooctanoate (APFO) L CHEMICAL IDENTIFICATION Chemical Name: CAS Number: Ammonium Perfluorooctanoate 3825-26-1 H. COMPANY IDENTIFICATION Company Name: B, I. du Pont de Nemours and Company Site Locations: Site where APFO is used as a reaction aid: Washington Works Route 892 Washington. WV 26181 Sites where APFO containing products made at Washington Works are processed; Parlin Plant Cheesequake Road Parlin, NJ 08859 Spruance Plant 5401 Jefferson Davis Hwy, Richmond, VA 23234 She which disposesof waste containing APFO: Chambers Works Rte. 130 Deepwater.NJ 08023 Technical Contact: Robert F. Pinchot (302) 999-4074 DuPont Fluoroproducts Chestnut Run Plaza Bidg. 711/2210 Centre Boulevard Wilmington, DE 19805-0711 EID156213 Voluntary UEIP, Ammonium Perfluorooctanoate ffl. DUPONT AND CUSTOMER ACTIVITIES Narrative Description ofAPFO Use The block diagram on the back page titled "DuPont US APFO Balance" describes the processes discussed below. DuPont uses APFO as a reaction aid in the production ofpolytetrafluoroethylene (PTFE) and tetrafluoroethylene (TFE) co-polymers- The process utilized at DuPont's Washington Works for making PTFE and co-polymers consists of polymerizing TFE (and other co-monomers if desired) in an aqueous media with a small amount of APFO to aid in the reaction. Following the polymerization step, the polymer dispersion is either dried to remove water and APFO or concentrated (removing some of the APFO), stabilized and sold as an aqueous dispersion. The dried polymer contains very little, if any, APFO, The APFO removed from the polymer is recovered for recycle, captured and destroyed off site in an incinerator, captured and sent to an offsite industrial landfill, and/or emitted to air or water at the Washington Works, The stabilized polymer dispersions are sold by DuPont to industrial customers (both in the US and outside the US) for a variety of uses, internally transferred to the DuPont Spruance Plant for the production of Teflon fibers and PTFE coated synthetic fibers, or internally transferred to the DuPont Pariin Plant for the production of Teflon Finishes. A small amount of non-hazardous waste polymer, water, APFO and other additives generated at Washington Works is treated in a wastewater treatment facility at DuPont's Chambers Works. This material is either emitted in the Chambers Works water discharge or captured on carbon and landfilled in a secure landfill. The internal process at the DuPont Spurance Plant to produce Teflon fibers involves,. for most of the product, a "sintering" step in which the APFO contained in the product is destroyed by the following reaction;1 CF3(CF2)6COO'NH4~ "> CF3(CF?)sCF2H+ COa +NH:3 This reaction goes to completion at 350<'C and 0.2s residence time. A small amount of product processed at DuPont's Spruance plant does not get sintered and thus contains a small amount of residual APFO. These products are used for industrial pump, valve and compressor packing materials. P.J. Knisic, D.C. Roc, 'Tlicnnal decompositionof C8 fluorinatcd surtactants and related materials studied by high temperature gas-phase "F NMR. A new Alternative to thermal gravimetric analysis, DuPont Internal Report. EID156214 Voluntary UEIP, Ammonium Periluorooetaaoate The process for making Teflon finishes at the DuPont Parlin Plant involves a blending operation of fluoropolymer dispersions with other additives includingsolvents, binders, and pigments. The small amount ofAPFO emissions to water from this facility is due to waste generated during product changeovers. Some of the fluoropolymer dispersionis processed at contract facilities where the material is dried at temperatures >350C thus destroying the APFO according to the reaction above. This dried material is then incorporated into finishes products, The final product produced is then sold to applicators that apply the product to a substrate (such as cookware) via automated spraying or rollercoating. Emissions of APFO from these operations consist ofoverspray that is either captured on filters and landfilled or absorbed into water resulting in a water emission. Product that is appliedto the substrate is then typically "sintered" at temperatures approaching SWF resulting in the removal of the APFO from the substrate and subsequent destruction according to the reaction above. Customers of dispersion products use the material for a variety of applications. However, most applications involve a "sintering" step where the APFO is destroyed. There are a small number of applications where the customer heats the dispersion products to temperatures that allow the APFO to sublime resulting in air emissions. There are also a small number of applications where the customer's product is not heated resulting in the APFO staying with the product. These applications include industrial packings, and industrial filter fabrics, IV. SITE RELEASE AND TRANSFER INFORMATION FOR TRI CHEMICALS Not applicable- APFO is not listed on the TRI V. SITE RELEASE AND TRANSFER INFORMATION FOR NON.TRI CHEMICALS A. On-site Air Releases Fugitive Stack (Point Source) Estimated Total Annual Releases (lbs.1999) Washington Works Negligible Parlin 0 Soruance 0 Chambers Works 0 24000 0 0 0 Comments Air emissions are estimated using engineering calculations and judgements and limited measurements of specific point sources conducted in the past. EID156215 Voluntary UEIP, Anunoiumn Perfluorooctanoate B. On-site Water Releases Point Source Estimated Total Annual Releases f Ibs 1999) Washington Works PsaM Spruance Chambers Works 55000 300 150 9500 Water emissions are estimated using engineering calculations and judgements and limited measurements of specific sources conducted in the past. Washington Works emissions occur for approximately 350 days/yr while the other sites' emissions occur for 10-100 days/yr. Releases of APFO to the Ohio River from the DuPont Washington Works Plant were modeled using the Probabilistic Dilution Model (PDM Beta Version 4,0 Beta June 11, 1999, US EPA Office of Pollution Prevention and Toxics) and a constructed Microsoft Excel spreadsheet model. APFO release data for 1996 were used in both modeling exercises.2 The PDM indicated that APFO concentrations of 1.0 ug C8/L would be exceeded about 50% of the time during the year. APFO concentrations of in the river would exceed 0.1 ug APFO/L 90% of the time during the year and 10 ug APFO/L about 2.2% of the time during the year. Average annual APPO concentrations in the Ohio River calculated by using a Microsoft Excel spreadsheet was 0.423 ug APFO/L. Modeled AFPO concentrations in the river ranged from a low of 0.199 ug APFO/L in March to a high of 0,965 ug APFOC-8/L in September, which correspond to high and low river flows, respectively. Average Ohio River flows and volume data calculated from the US Geological Survey was collected at the Belleville Dam and used in the spreadsheet model. The Belleville Dam is on the Ohio River 13 miles downstream of the Washington Works Plant. This river flow data is the closest location downstream from the plant where this type of information is available.' In 1999 a drinking water sample obtained from GE plastics, Washington WV, immediately downstream on the Ohio River from DuPont Washington Works showed 0.552ug/l APPO. In addition samples obtained in January 2000 from three different wells at the Lubeck Public Service District, downstream of Washington Works on the Ohio River, showed 0.8ug/l, 0.44ng/l and 0.313 u.g/1. APFO. 2 W.R.Berti, Modeling releases of ammonium perfluorooctanoale into the Ohio River, DuPont Internal Report EMSE-054-00. EID156216 Voluntary UEIP, Ammonium Periluorooctanoate C. On-Site Land Releases Chambers Works treats APFO containing waste in a wastewater treatment system. Engineering calculations and measurements indicate that approximately 30% of the APFO in the wastewater treated is absorbed on to a carbon media that is landfllled on site. These land releases are estimated to be 39001b in 1999. Prior operations have resulted in measurable APFO concentrations in three landfills operated by the Washington Works in West Wginia. At Letart3 landfill surface water measurements in 1999 and 2000ytd range from 2,23^g/l to 3240|Ag/l with an average of 1392ug/l. Groundwater measurements taken during the same time period at Letart landfill range from 60.3|j.g/l to 17400p,g/l with an average of2537ng/l. At the "local landfill" the groundwater concentrations range from 0.046|.ig/l to 39ug/l with an average of8.83|.ig/l. Surface water samples at the "local landfill" range from 0.54ug/l to 87|Ag/lwith an average of 18.5ng/l. At Dry Run landfill there are limited measurements of groundwater and surface water with maximum concentrations in groundwater of l5ng/l and the maximum concentration in the permitted outfall has been 33^ig/l. In 1999 a RCRA Facility Investigation was completed for Washington Works and was submitted to EPA Region BQ in June 19994. The report contains data on groundwater concentrations of APFO at Washington Works. D. Transfers to Off-site Locations Washington Works: Incineration Wastewater treatment Underground Injection Hazardous Waste Landfill Other landfill Recycle or recovery Estimated Tot at Aimua[R.eleas<esor Transfersdb.im 16000 134003 0 2600 0 IV. ON-SITE WORKPLACE EXPOSURE A. Information on the Number of Employees Potentially Exposed 3 Maps of the landfiU locations and specific monitoring locations and results are available upon request. '' Report vms submitted to Martin. T. Kotsch. Remedial Program Manager, EPA Region III, Philadelphia. 5 This is the same material that was described above in paragraph I of section V.D. EID156217 Voluntary UEIP, Animonhan Perfluorooctanoate The tables below describe the number of workers that may be exposed to APFO during their normal work activities for each of the three sites where APFO is used or APFO containing product is processed. Hours/Day <l0 <0.25 0.25-1 1-8 >8 Washington Works Days/vr 10-100 100-250 242 >250 Routine worker activities that have potential for exposure: > Handling raw material APFO > Handling raw dispersionscontaining APFO > Maintenance of polymerization reaction systems > Polymer dryer operation and maintenance > Packout ofPTFE and co-polymer dispersion products > Operation and maintenance of APFO recovery systems E1D156218 Voluntary UEIP, AnunoniumPeifluorooctanoate Hours/Day <10 <0.25 0,25-1 1-8 >8 Parlin Plant 10-100 Days/yr 100-250 18 >250 Routine worker activities that have potential for exposure: > Handling ofPTFE and Co-polymer dispersion products > Operation and maintenance of blending facilities > Packout of finished product Note that at no time is the material handled at the Parlin Plant at an elevated temperature where the APFO could sublime. Therefore there is little potential for exposure to airborne APFO at this facility. All exposure potential is through skin contact during handling of the polymer dispersion materials all of which contain <l% APFO with most containing <0.25% APFO. Hours/Day <10 <0.25 0,25-1 1-8 >8 Spruance Plant 10-100 Days/yr 100-250 <10 >250 Routine worker activities that have potential for exposure: Handling ofPTFE and Co-polymer dispersion products Operation and maintenance of fiber coating facilities Operation and maintenance of sintering rolls Packaging of non-sintered product, Note that the PTFE and co-polymer dispersion products used at the Spruance site contain <0,9% APFO with most containing approximately 0.3% APFO. B. Information on the Exposure Levels of Washington Works Employees Since most of the processing done in the US with APFO and APFO containing intermediates and products is done at Washington Works, DuPont's airborne industrial hygiene data is concentrated at that site. The limited measurements of airborne APFO concentrations at the other sites where APFO containing products are used have shown much lower levels (mostly non-detectable) levels of APFO. The data in the table below EID156219 &WWlW.SSW9W^^ii{'!Sis Voluntaty UEIP, Anuiioimim Periluorooctanoate reflect monitoring done over the last 5 years at Washington Works. The sample results are a combination of chemical operator and maintenance worker personal samples. Year Sample Type 1999 1998 1997 1996 1995 Partial Shift (mostly 6-8 hours) #of Samples 100 83 100 73 32 Minimum Concentration (mpb6) <0.01 .001 <0,0l N/D N/D Maximum Concentration (mpb) 0.58 0.78 2.4 0.29 0.16 Mean Standard (mpb) Deviation 0.061 0.103 0.146 0,055 0.067 0.151 0.145 0.378 0.069 0.063 Partial shift air samples are taken at the rate of 200 mL/min using a Tenax collection tube that has been pretreated with sodium hydroxide/ethyleneglycol/methanol. The APFO is desorbed from the tubes using methanolic hydrogen chloride, which also serves as a derivatizing reagent, converting the APFO to its methyl ester. After workup, the methyl ester is quantified using a gas chromatograph equippedwith an electron capture detector. The methyl ester ofperfluorodecanoic acid is used as an internal standard, and at least three calibration samples are prepared to cover the concentration range of interest. Precision is estimated to be +/-10% relative. The data above show averages consistently below the AGCffl TLV ofO.Olmg/m3 with only a very few samples above the TLV. Where results are above or near to the TLV, the event is investigated and corrective action (additional personal protective equipment or engineering controls) to reduce the exposure levels is undertaken. Older data from the 1980's show higher levels of exposure. In the early 1990's Washington Works switched from receiving the APFO as a powder to receiving it as an aqueous solution. This change was done to reduce the potential for exposure during handling of the dry powder. It should be noted that in the 1997 time period, the site was starting up new APFO recovery facilities. Operating and maintenance difficulties associated with the start-up of these facilities may have contributed to the higher levels of APFO in the personal samples during that year. Task specific monitoring data and wipe monitoring data exist. However these data are not indicative of employee exposure and are not presented here. These samples are taken to identify areas where additional exposure controls may be necessary. Engineering controls to reduce exposure consist of the following: > Reaction systems are closed systems with continuous ambient monitoring for monomer concentrations > Ventilation systems are installed where airborne concentrations are significant > The polymer dryers operate under negative pressure to contain APFO and other materials. > Recovery systems are in place to reduce airborne emissions. ' mpb"' moles per billion. 0.56mpb is equivalent to the ACG1H TLV ofO.01 mg/m3 EID156220 Voluntary UEtP, Anuiionium Perfluorooctanoate Personal protective equipment that workers regularly wear consist of the following: > Safety shoes and side-shield safety glasses in all areas. > Impervious gloves when handling APFO solutions or aqueous dispersion products. > Chemical protective coveralls and goggles or face shields when the possibility of splashes of APFO containing solutions is present. > Airline respirators or cartridge respirators where monitoring has shown to have high exposure potential. At Washington Works, blood serum levels of APFO have been measured since 1981, Prior measurements of blood fluoride levels have been taken prior to 1981 but are of limited value in assessing exposure to APFO. A summary of results of employees with identified APFO exposure potential the 1995, 1989.90, 1985,and 1984 volunteer sampling events is in the table below. Due to significant job assignment movement during this period of time, analysis of trends of data are difficult. The data in the table below prior to 1995 are for employees included in the 1995 sampling data so that comparisons of relative levels of APFO in blood serum can be compared. The entire data set of blood concentrations is available upon request, Year 1995 1989-90 1985 1984 #of Samples 73 23 21 19 Minimum Concentration (ppm) 0.12 0.4 0.067 0.077 Maximum Concentration (Ppm) 4.5 8,5 18s 2^ Mean Concentration (Ppm) 1,57 3.13 2.44 3.82 7 This individual was working in a job that has APFO exposure potential at the time of the sample, 8 This individual consistently has had the highestblood concentration of APFO since APFO specific samples were (aken. This employee ten an APFO exposure potential assignment in 1991. In 1995 Uus employee's blood scrum level was 4.4ppm. E101S6221