PFOS CHROMIUM ELECTROPLATER STUDY U.S. ENVIRONMENTAL PROTECTION AGENCY-REGION 5 September 2009 Final Report ------- 2 TABLE OF CONTENTS ACRONYMS... 3 EXECUTIVE SUMMARY 4 INTRODUCTION............. 5 Background 5 Purpose of Study 6 PROJECT DESCRIPTION 7 Site Selection, , 7 Figure 1: Location of Electroplating Facilities in Cleveland, Ohio 7 Figure 2: Location ofElectroplaters Sampled in Chicago, Illinois 8 Facility Inspections 8 Sampling 9 RESULTS 10 Figure 3: PFC Concentrations at Sampled Electroplater Facilities (reported in ppt) 10 DISCUSSION 11 Figure 4: PFC Concentrations in Effluent of Electroplaters Using Chemical Fume Suppressants 11 Figure 5: Proportion of Total PFCs 12 CONCLUSIONS 14 REFERENCES 15 APPENDIXES 16 Appendix A - Data Quality Assessment Report 17 Appendix B. Facility Operations 25 Appendix C. Rinsing Practices, Pretreatmenl and Wastewater Discharged 27 Appendix D. Hexayalent Chromium Controls 30 Cover: View of Chromium Electroplating Line Photo: Mark Conti & David Barna U.S. EPA-Region 5-Cleveland Office ------- 3 Acronyms 3M Minnesota Mining, and Manufacturing Corporation ANPRM Advance Notice of Proposed Rulemaking CAA Clean Air Act CARB California Air Resources Board Cr(VI) hexavalent chromium DQA data quality assessment dynes/cm dynes/centimeter LC/MS/MS liquid chromatography/mass spectrometry LCS laboratory control samples MACT Maximum Achievable Control Technology MDH Minnesota Department of Health MDL method detection limit MPCA Minnesota Pollution Control Agency MS matrix spike MSD matrix spike duplicates NEORSD Northeast Ohio Regional Sewer District ng/1 nanograms per liter OAQPS U.S. EPA - Office of Air Quality Planning and Standards OSHA Occupational Safety and Health Administration PFBA perfluorobutanoate PFBS perfluorobutane sulfonate PFCs perfluorinated chemicals PFDA perfluorodecanoate PFDoA perfluorododecanoate PFHpA perfluoroheptanoate PFHxA perfluorohexanoate PFHxS perfluorohexane sulfonate PFOA perfluorooctanoate PFOS perfluorooctane sulfonate PFOSA perfluorooctane sulfonamide PFNA perfluorononanoate PFPeA perfluoropentanoate PFUnA perfluoroundecanoate POTW publicly Owned Treatment Works ppt parts per trillion PQL practical quantification limit QAPP Quality Asssurance Project Plan R5 U.S. EPA - Region 5 office RPD relative percent difference U.S. EPA United States Environmental Protection Agency WWTP wastewater treatment plant ------- 4 EXECUTIVE SUMMARY In June 2008, R5 conducted a study to examine whether detectable levels of PFOS may be in the effluent of decorative chromium electroplating facilities that discharged to WWTPs. A year earlier, the State of Minnesota found high levels of PFOS at the Brainerd, Minnesota WWTP, and identified a chromium electroplating facility (Keystone Automotive) as the source JJLJ. Based on the State of Minnesota's findings, R5 initiated this study to investigate whether releases from chromium electroplating facilities could be a widespread source of PFOS in the environment. Along with other data, R5's study will be considered by the OAQPS to evaluate the use of PFOS in suppressing Cr(VI) emissions under air standards for this industry. Samples were taken from seven Chicago, Illinois (Chicago), and four Cleveland, Ohio (Cleveland) facilities. R5 tested for thirteen PFCs, including PFOS, and data showed the following: • PFCs were discharged from all eleven facilities' waste streams at quantifiable levels above background. • "Background" was defined by the rinse water measurements. All eleven facilities used municipal tap water for their rinse water. Therefore, one rinse water (background) sample was taken in each city as a measurement of background PFC levels. The background PFOS level for Cleveland was 5.75 ppt. The background PFOS level for Chicago was 2.52 ppt. • Ten out of the eleven facilities had PFOS detected in their wastewater in concentrations ranging from 31.4-39,000 ppt. • Of the ten facilities with PFOS detections, none had effluent levels higher than those found at Keystone Automotive facility located near Brainerd, Minnesota. ------- 5 INTRODUCTION In 2007, the MPCA tested the influent, effluent, and sludge at WWTPs across the State for PFCs, including PFOS. This investigation was done in order to determine if PFCs were present in these waste streams, and could therefore be a source of PFCs to the broader environment £1], PFCs had been manufactured in Minnesota by 3M since the 1950s. 3M phased out the manufacturing of PFOS-related products in 2002 because of the growing research findings that PFOS was toxic to animals, persistent in humans, and widespread in the environment. In 2004, PFCs were detected in drinking water supplies in several eastern Twin Cities communities and traced to the legal disposal of 3M waste [2], Through broader investigations, MPCA found widespread PFC contamination in various environmental media, including places with no known PFC sources. Through testing at WWTPs, MPCA found relatively high levels of PFOS at the WWTP in Brainerd, Minnesota. The city of Brainerd is located about 135 miles northwest of St. Paul, along the Mississippi River. The initial 2007 sampling results at Brainerd were: • Influent: 811 pptPFOS; • Effluent: 1500 pptPFOS; • Sludge: 861,000 ppt PFOS LQ. WWTP effluent may be a significant entry of PFCs to the environment [3], and several studies have concluded that conventional wastewater treatment may not be effective in removing these compounds [4] [51. MPCA traced the PFOS in Brainerd's WWTP to a local chromium electroplating facility, Keystone Automotive (Keystone). Keystone was reportedly one of the largest chrome bumper repair and plating facilities in the United States. Since 1995, Keystone had been applying a commonly used PFOS-containing mist suppressant (Fumetrol 140®) in order to comply with the CAA's Cr(VI) MACT standard. As a result of MPCA's findings, the company switched to an alternate non-PFOS containing mist suppressant in early September 2007 [11. Background Cr(VI) electroplating is the electrical application of a coating of chromium onto a surface for decoration, corrosion protection, or durability. An electrical charge is applied to a tank (bath) containing an electrolytic salt solution. The electrical charge causes the chromium metal in the bath to fall out of solution and deposit onto objects placed into the plating bath. In an anodizing process, an oxide film is formed on the surface of the part. These electrolytic processes cause mist and bubbles containing Cr(VI) to be ejected from the bath, released into the work place, and eventually dispersed into outdoor ambient air unless controlled with add-on air pollution control equipment or chemical fume suppressants. Chemical fume suppressants reduce surface tension and thereby, control Cr(VI) emissions. Surface tension is the force that keeps a fluid together at the air/fluid interface, and typically is expressed in force per unit of width, such as dynes/cm. By reducing surface tension in the plating/anodizing bath, gas bubbles become smaller, and rise more slowly than larger bubbles. Slower bubbles have reduced kinetic energy so that when the bubbles do burst at the surface, the ------- 6 Cr(VI) is less likely to be emitted into the air, and the droplets fall back onto the surface of the bath [6], Ideally, chromium plating baths should have surface tension values between 45-55 dynes/cm [71. Cr(VI) is a human carcinogen. Therefore, the U.S. EPA regulates Cr(VI) electroplating or Cr(VI) anodizing tank operations by applying the CAA MACT limits. The MACT limits require control of Cr(VI) emissions to the atmosphere by either limiting the amount of Cr(VI) through use of add-on air pollution control devices or utilizing a chemical fume suppressant [8], These facilities are also regulated by OSHA under 29 CFR Part 1910.1026 to protect workers from occupational Cr(VI) exposure. Employers are required to use engineering and work practice controls to reduce and maintain employee exposure to Cr(VI) at or below the permissible exposure level of 5 micrograms per cubic meter of air, calculated as an 8 hour time weighted average [6] [8], Purpose of Study After the release of MPCA's findings, R5 examined whether the release of PFOS through normal electroplating operations to WWTPs was a widespread or isolated event. Conversations between R5 staff and the Metal Refinishers Association indicated that PFOS use had become the industry standard as the most economic method of complying with the MACT rule [9], Additionally, a 2003 survey conducted by the CARB, found that 190 of the 222 Cr(VI) electroplating operations in California used a fume suppressant, either in part or solely, to control Cr(VI) emissions. Almost all of the 190 operations used a chemical fume suppressant with PFOS as the active ingredient, and 124 reported using the same suppressant (Fumetrol 140®) that Keystone used. M- R5 provided this information to OAQPS. OAQPS was preparing to conduct a residual risk assessment for Cr(VI) electroplating, and to collect data through the ANPRM. Typically, releases of PFOS compounds would not be considered during a residual risk review since it is not one of the listed 188 hazardous air pollutants as defined by the Clean Air Act. However, OAQPS agreed to use the ANPRM to review data on the extent of PFOS mist suppressant use in Cr(VI) electroplating facilities and the potential release to WWTPs. Because available data were likely to be limited, R5 also decided to gather data for OAQPS through a study to evaluate whether detectable levels of PFOS were present in the effluent of decorative Cr(VI) electroplating facilities that discharged to publicly owned WWTPs. ------- 7 PROJECT DESCRIPTION During June 2008, R5 collected samples of discharged process wastewaters at eleven decorative Cr(VI) electroplating facilities. The study was confined to facilities in the Chicago and Cleveland areas. Only decorative, rather than hard, chromium electroplaters were selected because of their higher likelihood to generate wastewaters that would then be discharged to a WWTP [101. Site Selection Prior to sampling, field investigators conducted telephone surveys of chromium electroplating and chromic acid anodizing facilities to determine the best candidates. The Cleveland area list of potential study candidates was assembled by identifying facilities with chromium emissions reported in the Aerometric Information Retrieval System, and facilities that were subject to the Electroplating Point Source Category at 40 CFR 413. The latter group of facilities was provided by the NEORSD, which operates the three area POTWs, and is the control authority for indirect dischargers in the Cleveland area. The combined list contained sixty-four potential study candidates. Twenty-two candidates were contacted by telephone and asked if they: (1) performed Cr(VI) electroplating or anodizing; (2) discharged process wastewater; and (3) used chemical fume suppressants. If a company met these criteria, it was given a brief description of the project and told that the Cleveland team may sample their wastewater discharge as part of the project. Seven of the twenty-two facilities screened by telephone met all three criteria, and due to funding limitations, only four facilities that were furthest along in arrangements were selected (see Figure 1). Figure 1: Four Chromium Electroplaters Sampled by Cleveland Team (Facilities #1-4) Mttntor Lake Erie Faclllty,#4 Facility #2 NEORSD Westerly Cleveland'Metro Area Facility #1 Source I \ Base map frahires via StreelMap USA; WRP locators wa Permit Ccmptance System < PCSf Figure 1: Location of Electroplating Facilities in Cleveland, Ohio ------- 8 The Chicago area list of potential candidates was assembled by identifying facilities with existing air permits issued by the State of Illinois for operation of Cr(VI) electroplating tanks. The initial list contained twenty-six potential study candidates. Each of the twenty-six facilities were telephoned and asked the same questions as the Cleveland facilities. In addition, pre- sampling site visits were conducted following the telephone surveys. Seven Chicago facilities were picked as final candidates (see Figure 2). Figure 2: Seven Chromium Electroplaters Sampled by Chicago Team (Facilities #5-11) Flgin Facility #9 Lake Michigan '"Schaumburg CKh Metro Area Facility B11 Figure 2: Location of Electroplaters Sampled in Chicago, Illinois. All of the Chicago and Cleveland facilities selected performed decorative Cr(VI) electroplating on metal and/or plastic. Each facility used chemical fume suppressants (wetting agent and/or foam blanket) in its chromic acid bath tank to comply with the Cr(VI) electroplating MACT. In addition to chemical fume suppressants, two Cleveland facilities also employed add-on air pollution control devices. It was noted that facilities plating on plastic also used wetting agents in their chrome etch tanks for process control (i .e. to prevent voids in corners and creases of parts). Facility Inspections Both field investigation teams conducted cursory inspections at the facilities in conjunction with the sampling. During the inspections, the plating process, wastewater treatment, water usage, and usage of chemical fume suppressants were reviewed. Details of the inspections are summarized in Appendix B-D. ------- 9 Sampling The Cleveland field investigators collected samples between June 9 and 12, 2008, and the Chicago field investigator collected samples on June 9, 2008. All samples were taken during normal plating operations. At least one sample of discharged process wastewater from each facility was collected immediately prior to entry into the public sewerage system. Discharged process wastewater was comprised of the treated rinse waters from the plating operations, not including sanitary wastewater. In addition to the single effluent sample collected at each facility, the field investigators collected additional quality control samples at one select facility. These additional samples included the rinse water (background sample), field blank, and effluent duplicate. Samples were collected directly into laboratory-provided containers using standard operating procedures. The field blank was obtained by pouring reagent grade water into a laboratory-provided container while adjacent to the facility's discharge location. Samples were placed into iced coolers, refrigerated under custody until shipment to the laboratory, and cooled with blue ice packs during shipment. The samples were subsequently analyzed by AXYS Analytical Services Ltd. of Sidney, British Columbia, Canada, for thirteen PFCs, including PFOS (see Figure 3 for results). The analytical method used was solid phase extraction with High Performance Liquid Chromatography, tandem mass spectrometry (LC/MS/MS). ------- 10 RESULTS Figure 3: PFC Concentrations at Sampled Electroplater Facilities (reported in ppt) Facility ID#: Fume Sum of PFOS/ Suppressant(s) PFBA PFPeA PFHxA PFHpA PFOA PFNA PFDA PFUnA PFDoA PFBS PFHxS PFOS PFOSA PFCs PFC Facility #1: Mist Suppressant A, B, C 9.06 42.6 90.7 56.2 83.3 ND ND ND ND 9,160 67.8 11 100 ND 40,610 77% Facility #2: Mist Suppressant D, E 48.3 30.9 ND ND ND ND ND ND ND 41,800 306 708 ND 42,893 2% Facility #3: Mist Suppressant B, F ND ND 177 175 650 13,100 27.1 44.1 ND 75.5 ND ND ND 14,249 <0.26% Facility #4: Mist Suppressant D ND ND ND ND ND ND ND ND ND 15,600 ND 39,000 ND 54,600 71% Facility #5: Mist Suppressant G ND ND ND ND ND ND ND ND ND 1,010 ND 2,320 ND 3,330 70% Facility #6: Mist Suppressant Unknown ND ND ND ND 4.02 ND ND ND ND 1,570 16.3 1,380 ND 2,970 46% Facility #7: Mist Suppressant H ND 1.08 ND ND 3.11 ND ND ND ND ND ND 301 ND 305 99% Facility #8: Mist Suppressant H ND ND 2.3 1.17 3.17 ND ND ND ND 311 993 1,770 ND 3,081 57% Facility #9: Mist Suppressant Unknown ND ND ND ND 1.73 ND ND ND ND 2,250 163 4,460 ND 6,875 65% Facility #10: Mist Suppressant Unknown 1.54 1.29 1.82 ND 3.32 ND ND ND ND ND 3.53 31,4 ND 42.9 73% Facility #11: Mist Suppressant Unknown 14.3 ND ND ND ND ND ND ND ND 1,510 9,430 1,260 ND 12,214 10% Number of Detects 4 4 4 3 7 1 1 1 0 9 7 10 0 11 Minimum ND ND ND ND ND ND ND ND ND ND ND 31,4 ND 42.9 Maximum 48.3 42.6 177 175 650 13,100 27.1 44.1 ND 41,800 9,430 39,000 ND 54,600 Cleveland Background Sample 1.42 1.58 ND 1.74 2.19 ND ND ND ND ND ND 5.75 ND Chicago Background Sample ND ND ND ND 1.37 ND ND ND ND ND ND 2 ^2 ND Notes ND means that the analyte was not detected at the method detection limit. ND ranged from <1.00 to < 45.3 ppt depending on analyte. For total PFCs, a value of zero was used in the sum of PFCs calculation. ------- 11 DISCUSSION Data results summarized in Figure 3 showed the following. • PFCs were discharged from all eleven facilities' waste streams at quantifiable levels above background. Ten out of the eleven facilities had PFOS above background detected in their waste discharge streams (Figure 4). 60,000 50,000 parts per trillion 40,000 30,000 20,000 10,000 PFBA PFPeA BpFHxA PFHpA ¦ PFOA ¦ PFNA PF DA 1 PFUnA 1 PFBS BPFHxS W PFOS ¦ ¦ ¦ : _ . 1 Fac. #1: Fac. #2: Fac. #3: Fac. #4: Fac. #5: Fac. #s Fac. #7: Fac. #8: #Fac. #9: Fac. #10: Fac. #11: Mjst Mist Suppressant Mist Mist Mist Mist Mist Mist Mist Mist Mist Suppressant^ anc' ^ Suppressant Suppressant Suppressant SuppressantSuppressant Suppressant Suppressant Suppressant Suppressant ,A B and C ®anc' F D G Unknown H Unknown Unknown Unknown Unknown Figure 4: PFC Concentrations in Effluent of Electroplaters Using Chemical Fume Suppressants • "Background" was defined by the rinse water measurements. All eleven facilities used municipal tap water for their rinse water. Therefore, one rinse water (background) sample was taken in each city as a measurement of background PFC levels. The background PFOS level for Cleveland was 5.75 ppt. The background PFOS level for Chicago was 2.52 ppt. In addition to PFOS, four other PFCs were detected in the Cleveland background sample, and one other PFC was detected in the Chicago background sample. ------- 12 At least eight different mist suppressants or mixture of suppressants were used at the various facilities and are as followed (several facilities did not provide information). Of the ten facilities with PFOS detections, none had effluent levels higher than those found at Keystone. In a sample dated December 2007, Keystone had a PFOS result of 278,000 ppt. [1], The highest effluent PFOS result in this study was 39,000 ppt. The averages of the four highest concentrated compounds were: PFOS at 7680 ppt; PFBS at 6580 ppb; PFNA at 1190 ppt; and PFHxS at 1100 ppt (these averages were calculated using zero for the nondetects). These four chemicals made up over 99% of all compounds (Figure 5). PFOS, PFBS, PFHxS, and PFOA were the most commonly detected PFCs. Figure 5: Proportion of Total PFCs Another PFC compound of general interest, PFOA, was detected at seven of the eleven facilities, ranging from 1.73 - 650 ppt. PFCs were found in one field blank and in the background samples. Field blanks consisted of reagent grade bottled water exposed to the atmosphere at the designated facility. The field blank with PFHxS detection was exposed at facility #11 whose effluent samples contained PFHxS concentrations at the highest levels detected in this study. We attributed PFCs in the background samples to trace background levels found Benchmark Benchbrite STX AB (custom-made) Benchmark Benchbrite STX Benchmark CFS MacDermid Proquel B MacDermid Macuplex STR Plating Process Systems PMS-R Fumetrol-140 Brite Guard AF-1 fume control. PFOSA ^PFBA PFOA PFHpA PFDA PFHxA PFDoA PFNA 7% , PFUnA PFHxi 7% Proportion of each compound of the Total PFCs measured in this study - all location results summed together ------- 13 in municipal tap water. Lake Michigan is the source water for the Chicago Municipal Utilities, and Lake Erie is the source water for the Cleveland Municipal Utilities. During the time of this study, both utilities were in compliance with all federal and state drinking water standards. • Although not applicable to this industry, we compared our results to state and federal PFC guidance levels. Nine of the ten facilities tested above the U.S. EPA provisional health advisory for PFOS in drinking water set at 200 ppt. Ten of the eleven facilities tested above the Minnesota water quality criteria for PFOS in the Mississippi River (6 ppt). ------- 14 Conclusions The data clearly indicated that decorative chromium electroplaters discharge PFOS and other PFCs to WWTPs in concentrations higher than background levels. Data also indicated that mist suppressants have very specific PFC mixtures, which may be found in the resulting electroplater effluent. The concentrations vary widely which is most likely due to the inherit design of study. Therefore, care should be taken when comparing results from one facility to another, as the study included facilities of different operational sizes and production schedules. Facilities also varied widely in the brand of mist suppressant used, and amount added to the plating baths. We would like to emphasize the nexus between the PFOS emissions and the Chromium MACT rule. To comply with the MACT rule, many facilities have chosen PFOS-containing mist suppressants as the best available technology to achieve Cr(VI) risk reduction in lieu of adding control technology. EPA believes that the PFOS emissions (as well as other PFC emissions reported in this survey) should provide target areas for improved pollution prevention performance including: (1) the development of alternative PFC free mist suppressants; (2) the improved procedures to reduce and capture downstream PFC levels in the wastewater prior to release into the waste water treatment facility; and (3) enhancing operating processes that limit the amount of PFC added to plating baths to efficaciously promote plating while reducing PFC total consumption. ------- 15 References [1], Minnesota Department of Health. Health Consultation: PFOS Detections In the City of Brainerd, Minnesota. Minnesota Department of Health, St. Paul, Minnesota. August 13, 2008. http://www.health.state.mn.us/divs/eh/hazardous/topics/pfcs/pfosdetectbrainerd.pdf [2], Minnesota Pollution Control Agency. PFCs in Minnesota's Ambient Environment: 2008 Progress Report, http://www.pca.state.mn.us/publications/c-pfcl-02.pdf [3], Schultz, M.; Higgins, C.; Huset, C.; Luthy, R.; Barofsky, D.; Field, J. Fluorochemical Mass Flows in a Municipal Wastewater Treatment Facility. Environmental Science Technology. December 1, 2006. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2556954 [4], Sinclair, EL; Kantian, K. Mass Loading and Fate of Perfluoroalkyl Surfactants in Wastewater Treatment Plants. Environ. Sci. Technology. January 21, 2006. http://pubs.acs.org/doi/full/10.1021/es051798v [5], Clara, M.; Scheffknecht, C.; Scharf, S.; Weiss, S.; Gans, O. Emissions of perfluorinated alkylated substances (PFAS) from point sources—identification of relevant branches. Water Science Technology. 2008. http://www.iwaponline.com/wst/05801/0059/058010059.pdf [6], Air Resource Board. Proposed Amendments to the Hexavalent Chromium Airboure Toxic Control Measure for Chrome Plating and Chromic Acid Anodizing Operations. 2006. http://www.arb.ca.gov/regact/chrom06/cpisor.pdf [7], http://hyperphysics.phy-astr.gsu.edu/Hbase/surten.html [8], U.S. Environmental Protection Agency. National Emission Standards for Chromium Emissions from Hard and Decorative Chromium Electroplating and Chromium Anodizing Tanks. 40 CFRPart63, Subpart N. 1995. [9] Metal Finishers Association. Telephone Conversation. 2008. (U.S. EPA-R5 Interviewer). [10], P.J. Paine. Telephone Conversation. 2008. (USEPA-R5, Interviewer). ------- 16 APPENDIXES ------- 17 Appendix A - Data Quality Assessment Report INTRODUCTION This report was developed as a scientific evaluation designed to determine if the PFOS data obtained from the 2008 R5 Electroplater PFOS study were appropriate to meet the study objectives, and were of the right type, quality, and quantity to support the intended use. This assessment also estimated the level of confidence attributable to the data set. In brief, our analyses showed that some decisions and conclusions associated with these data could be made with a high degree of confidence, while other decisions had significant limitations associated with them. The data were used to evaluate wastewater contamination associated with average industrial chromium decorative electroplaters. This PFOS study did not include a statistical sample design, and as such, rigorous statistical evaluations were not used. The data were assessed using the following criteria. 1. Review the data quality obj ectives 2. Conduct a preliminary data review 3. Perform an analysis of the data 4. Verify the assumption of the analysis 5. Draw conclusion from the data 1. REVIEW OF DATA QUALITY OBJECTIVES Statement of the problem: The overall objective of this study was to evaluate potential wastewater PFOS release associated with decorative chromium electroplaters. Study question: 1) Was there PFOS in the wastewater discharge to WWTPs from decorative chromium electroplaters? 2) Were these discharges quantifiable? Identification of the decision: Decision statement - If PFOS discharges from decorative chrome electroplaters are present at facilities using an approved MACT standard technology for suppressing Cr(VI) emissions, then these results may be useful in informing the OAQPS rulemaking process. ------- 18 Identification of inputs to the decision: Facilities were selected based on: • Use of PFOS containing surfactant; and • Potential for rinse stream/waste water contamination. Definitions of the boundaries for the study: This study was confined to chromium electroplating facilities in the Chicago and Cleveland areas. Samples were taken during normal plant operating conditions, and sampling locations were representative of discharged wastewater to POTWs. Documented decision rules: • PFOS was present when the analyzed concentration was above the laboratory MDL. • PFOS discharges were quantifiable when effluent concentrations observed were above the laboratory PQL. • PFOS discharges were believed attributable to the use of MACT complaint Cr(VI) suppressants, when effluent concentrations were above background PFOS levels. Optimize the design for obtaining data: This analysis may be useful in addressing future studies of PFOS and other PFCs as related to wastewater discharges. 2 PRELIMINARY DATA REVIEW Completeness All samples identified in the QAPP were collected and analyzed. Holding Time All samples were analyzed within the required holding times. Sample Preservation All samples were collected and iced for shipment to lab. Sample Receipt All samples were received on ice within 24 hours of shipping. The samples were all refrigerated at 4 degree C prior to extraction and analysis. Sample Extraction and Analysis Samples were analyzed in three batches. Sample extraction, instrumental analysis, and analyte quantification procedures were in accordance with the lab's standard operating procedures. Samples were spiked withl3C-labelled quantification standards and extracted and cleaned up using SPE cartridges. Extracts were instrumentally analyzed using liquid chromatography/mass spectrometry (LC/MS/MS). Analyte concentrations were determined by isotope dilution/ internal standard quantification. Reporting limits were defined as the concentration equivalent to the lowest calibration standard or the sample specific detection limit, whichever was greater. ------- 19 Instrument Calibration All initial calibrations specifications were reported as met. Initial calibration percent recoveries and retention times demonstrate ongoing precision and accuracy. Continuing Calibration All continuing calibration and verification specification were met. Internal standards All ongoing precision and recovery specifications were met. Target Compound Results No apparent matrix interferences were noted in the analysis of the target compounds. Sample analyte concentrations were not blank corrected and results should be evaluated with consideration of the procedural blank results. 3. Data Analysis In this study, the MDL was used to determine if an analyte was present in a sample and the PQL was used to make a quantitative determination of the amount of analyte in the sample. The U.S. EPA uses the term MDL and PQL to describe the specific approaches of estimating the detection and quantification limits, respectively. If comparing concentration directly to a standard, it must be greater than the quantification limit in order to provide a reliable estimate whether or not the standard has actually been exceeded. To determine whether or not an analyte is present or absent in a sample, a result will be above the detection limit. Measurements above the quantification limit can be used directly. Measurements below the quantification limit are considered censored and must be appropriately adjusted. The blank data from this experiment showed rather conclusively that PFOS concentrations found in the electroplater effluents were tied to PFOS-containing suppressants. It is also unlikely that PFOS or other PFCs were introduced at significant levels through other means. Laboratory blank samples were free of contamination. Field blanks, consisting of commercially available reagent water, were also free of significant contamination. Facility source waters were analyzed to evaluate potential contamination coming into the facilities. These samples were also free of significant PFOS concentrations. Surrogate recoveries for the lab, field, and source water blanks ranged from 69 to 148% and averaged 105%, demonstrating that sample preparation and analyses were free of contamination. ------- 20 Analysis of Quality Control Background PFOS Sampling (in ppt) Blanks Analyses Min Max Laboratory 3 0 0 Field 2 0 0 Source Water 2 2.52 5.75 Note: Lab blanks were free of contamination. One field blank contained trace level of PFHxS. Source water samples contained trace levels of PFOS and other PFCs. Laboratory Control Samples Three laboratory control samples were analyzed for the Chicago and Cleveland batch of samples as well as the re-analysis of selected samples. PFOS recoveries ranged from 89 to 105%, and had an average recovery of 96% (see below). Sample Name A-Ohio B-Illinois C-Illinois Analyte % Recovery % Recovery % Recovery AVG %RPD PFBA 77.3 107 85.3 90 33 PFPeA 87.1 109 94 97 23 PFHxA 84 119 101 101 35 PFHpA 76.8 117 102 99 41 PFOA 84.4 112 78.1 92 37 PFNA 89.9 96.6 99.2 95 10 PFDA 104 98.8 118 107 18 PFUnA 107 94.2 118 106 22 PFDoA 86.1 119 95.5 100 33 PFBS 102 110 113 108 10 PFHxS 85 119 101 102 33 PFOS 89.2 105 94.2 96 16 PFOSA 94.7 100 109 101 14 Matrix Spike Duplicates Matrix spike and spike duplicate analyses were performed to evaluate the potential for sample interferences. Matrix interferences are also referred to as matrix effects. Matrix spike interferences are those chemical and/or physical interferences that impede the analytical instrumentation in detecting the true value concentration of a target analyte within a sample. One possible source of matrix interferences may be caused by contaminants that are co-extracted from the sample and result in a positive or negative bias. The extent of matrix interferences will vary considerably from source to source, depending upon the nature and diversity of the sample matrix. ------- 21 MATRIX SPIKE MATRIX SPIKE DUPLICATE Illinois Illinois ID Matrix Spike I Matrix Spike Duplicate I Sample Size 0.0349 L 0.0318 L Analyte % Recovery % Recovery AVG %RPD PFBA 101 89.6 95 12 PFPeA 92.4 120 106 26 PFHxA 101 114 108 12 PFHpA 99 109 104 10 PFOA 81.8 122 102 39 PFNA 116 112 114 4 PFDA 99.1 90.1 95 9 PFUnA 76.9 88.7 83 14 PFDoA 93.5 87 90 7 PFBS 104 109 107 5 PFHxS 98.2 101 100 3 PFOS 95.8 105 100 9 PFOSA 105 114 110 8 Spike recoveries for PFOS ranged from 96 to 105 averaging 100% with a 9% RPD. While these recoveries were within laboratory specifications, the spiking concentrations were well above the sample concentration. This practice did not allow an appropriate assessment of the impacts from the sample [see comparison table of laboratory control samples (LCS) and matrix spike(MS) /matrix spike duplicates (MSD) %RPDs below]. Comparison of Precision between LCS and MS/MSD MS/MSD LCS Analyte %RPD %RPD PFBA 12 33 PFPeA 26 23 PFHxA 12 35 PFHpA 10 41 PFOA 39 37 PFNA 4 10 PFDA 9 18 PFUnA 14 22 PFDoA 7 33 PFBS 5 10 PFHxS 3 33 PFOS 9 16 PFOSA 8 14 Field Duplicates A field duplicate is a duplicate sample collected by the same team or by another sampler or team at the same place, at the same time. It is used to estimate sampling and laboratory analysis precision. PFOS duplicate analyses ranged from 40 to 96% RPD and demonstrated variable ------- 22 precision for the selected samples. Values for RPD appeared dependent on the concentrations found in the sample. Analysis of sample with low concentration of PFOS appeared more precise (i.e., lower RPD). Samples with higher concentration of PFOS appear less precise (i.e., higher RPD). The laboratory narrative report from AYXS attributes these variable recoveries to the presence of particulate matter in the subject samples. Given the limited number of samples collected and the general expectation that PFOS will attach to particulate matter, future analyses should ensure greater homogenization of samples or collection of sufficient samples such that statistical evaluations may be conducted. Based on the particulate-free LCS results and their acceptable precision and accuracy, we believe the variability in wastewater sample recoveries do not impinge our study conclusions. Field Duplicate Results Cleveland & Chicago Dup Sampling ID Sample size 0.0163 L 0.0154 L 0.0658 L 0.0727 L 0.0718 L Analytes PPt PPt %RPD PPt PPt PPt %RPD PFBA 48.3 45.6 6 <7.60 8.6 17.1 66 PFPeA 30.9 33.4 8 8.29 9.93 7.4 30 PFHxA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFHpA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFOA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFNA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFDA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFUnA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFDoA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 PFBS 41800 39900 5 1410 1580 1820 26 PFHxS 306 227 30 8900 11400 12600 34 PFOS 708 470 40 2040 6180 4680 96 PFOSA <30.6 <32.4 0 <7.60 <6.88 <6.97 0 Surrogate Spikes A surrogate is a pure analyte that is extremely unlikely to be found in any sample. It is added to a sample aliquot in known amounts before extraction and is measured with the same procedure used to measure other sample components. A surrogate behaves similarly to the target analyte and most often used with organic analytical procedures. The purpose of a surrogate analyte is to monitor method performance with each sample. This study used 7- C 13 substituted isotopes. For all samples collected, surrogates recoveries ranged from 25 to 148% and averaged 91%. These recoveries were within historical laboratory specifications and analyses were generally within control. For the analyte specific surrogate, 13-C4 PFOS recoveries ranged from 53.1 to 132% and averaged 90%. These recoveries demonstrate acceptable precision and accuracies for evaluating the target compound. ------- 23 Surrogate Percent Recoveries Sample ID A B C D E F G H I J K LABELED COMPOUND 13C4-PFBA 115 119 118 79.6 24.7 28 84.6 47 78.2 46.9 105 13C2-PFHxA 107 124 114 99.2 94.2 40 95.7 80.5 91.4 61.5 115 13C2-PFOA 138 113 116 101 109 106 85.9 109 113 93.4 104 13C5-PFNA 77.8 147 66 137 126 66.4 96.6 78.2 85.7 69.9 111 13C2-PFDA 75.4 124 108 69.5 90.3 43.7 84.9 80.9 66 82.2 78.8 13C2-PFDoA 111 148 83.7 84.7 114 69.7 60.2 89.2 74.5 73.4 86.8 13C4-PFOS (80) 100 132 132 83 82.7 53.1 93.8 70.6 55.7 72.6 103 PFOS/PFC Suppressant Analysis Eleven facilities were sampled and analyzed for the presence of PFOS, and other PFCs. A review of facility records showed that at least eight different suppressants or mixtures were used at the various facilities. Several facilities did not provide information on the suppressant used. Ten facilities had PFOS wastewater sample results above the MDL. For this study, the PQL was defined as 5 times the MDL. The positive PFOS results ranged from 231 to 2976 % of the calculated sample PQLs. Facility Results Compared To Quantitative Definitions Facility AVG Sample size PFOS (ppt) Sample specific MDL PQL (5XMDL) % >PQL #1 0.0619 L 31100 209 1045 2977 #2 0.0163 L 708 61.2 306 231 #3 0.0269 L U 37.1 185 0 #4 0.0601 L 39000 423 2115 1844 #5 0.200 L 2320 42.9 214.5 1082 #6 0.201 L 1380 11.8 59 2339 #7 0.498 L 301 9.13 45.7 659 #8 0.494 L 1770 32.1 161 1099 #9 0.350 L 4460 97.5 488 914 #10 .497 L 31.4 2.01 10.05 312 #11 .0718 L 4680 33 165 2836 4. VERIFICATION OF ASSUMPTIONS We have verified the following assumption in evaluating our study question, "Is PFOS discharged from decorative chrome plating operations?" • Various Cr(VI) control methods are available; • PFC-containing mist suppressants are in common industry use; ------- 24 • Composition of suppressants may include various PFOS formulations; • Active suppressant ingredients contain other PFCs beyond PFOS; • Suppressant application is monitored; • Electroplating discharges are amenable to PFOS analysis; and • PFOS analyses are reasonably precise, accurate, recoverable, and reproducible. 5. SUMMARY OF CONCLUSIONS Several conclusions can be made from the data collected: •S The small sample size limits the ability to draw conclusions beyond the observation that PFOS as well as other PFCs appear to be discharged from decorative chromium electroplating facilities through wastewater discharge; •S These discharges are quantifiable; •S Composition of PFOS containing mist suppressant vary widely; •S Variability in wastewater sample recoveries do not impinge our study conclusions; and •S PFOS data obtained from this study were appropriate to meet the study objective, and were of the right type, quality, and quantity to support the intended use. ------- 25 Appendix B. Facility Operations Facility #1 Facility #2 Facility # 3 Facility # 4 Plating and metal finishing operations Two lines of decorative chrome electroplating with chromic acid on plastic parts. A third plating line applies a gold-colored finish to nickel-plated parts. Decorative chrome electroplating with chromic acid on metal and plastic parts. One line uses plastic parts and one uses metal parts. Decorative chrome electroplating with chromic acid on plastic parts. Decorative chrome electroplating with chromic acid on metal parts (brass, steel, and stainless steel), cadmium plating, and chromate conversion coating. Operating schedule 24 hours/day, 5 days/week Plastic substrate parts line: 24 hours/day, 5 days/week Metal substrate parts line: 8 hours/day, 5 days/week 24 hours/day, 5 days/week 10 hours/day, 4 days/week Chromium electroplating process description Parts are prepared by dipping in a chromic acid etch bath, neutralization tank, palladium-tin activator bath, and accelerator to remove tin. The plating process includes copper strike, bright acid copper, electroless nickel (and semi-bright, high sulfur and bright nickel for exterior use parts), and chromium electroplating with chromic acid. Each step is followed with rinsing. Plastic substrate parts line: Parts are prepared by dipping in a chromic acid etch bath, neutralization tank, activator, and accelerator. The plating process includes electroless copper, copper strike, bright acid copper, nickel, and chromium electroplating with chromic acid. Each step is followed by rinsing. Metal substrate parts line: Parts are prepared by dipping in a cleaner and then an acid tank. The plating process includes copper strike, bright acid copper, nickel, and chromium electroplating with chromic acid. Each step is followed with rinsing. Parts are prepared by dipping in a chromic acid etch bath, neutralization tank, activator, and accelerator. The plating process includes electroless copper, copper strike, bright acid copper, semi- bright or satin nickel, bright nickel, microporous nickel, and chromium electroplating with chromic acid. Each step is followed with rinsing. Parts are prepared by dipping in a soak cleaner, electro- cleaner, and then sulfuric acid (with or without current). The plating process includes nickel strike, bright nickel, and chromium electroplating with chromic acid. Each step is followed with rinsing. ------- 26 Facility #5 Facility #6 Facility #7 Facility #8 Facility #9 Facility #10 Facility #11 Plating and metal finishing operations This facility operates two chrome tank lines which are 225- gallons each. This facility operates a single 350-gallon chrome tank. This facility operates a single 400- gallon chrome tank. This facility operates a 500- gallon chrome tank. This company operates a 500- gallon chrome tank. This facility operates a large 4000-gallon chrome tank. The facility decorates a variety of metal parts including shopping carts and other pieces. Operating schedule This facility has not used its chrome tank for over sixty days and rarely chromes metal pieces. 2 hour/day, maximum of 15 hours/week. 1 day/week. 2-4 hours/day. 8 hours/day, 5 days/week. Chromium electroplating process description ------- 27 Appendix C. Rinsing Practices, Pretreatment, and Wastewater Discharge Facility #1 Facility #2 Facility #3 Facility #4 Rinsing practice for chromium electroplated parts The interior parts line has four city water rinses followed by one deionized water rinse. The exterior parts line has three city water rinses followed by one deionized water rinse. Rinsewaters flow counter-currently. The metal substrate parts line has three city water rinses. Rinsewaters flow countercurrently. The plastic substrate parts line has four city water rinses followed by one deionized water rinse. Rinsewaters flow countercurrently. Two city water rinses followed by three deionized water rinses. Rinsewaters flow countercurrently. One city water static rinse, three countercurrent city water rinses, then one deionized water static rinse. The final deionized water static rinse is emptied daily. Rinse water pretreatment Acid, chromium electroplating, electroless nickel, and copper-nickel rinsewaters are received in separate tanks in the pretreatment plant. Nickel is recovered by ion exchange before nickel rinsewaters are pumped to the pretreatment facilities. At the pretreatment plant, acid rinsewaters are combined with chrome-bearing rinsewaters. Chrome is reduced with sodium metabisulfite. Dissolved metals in electroless nickel rinsewaters are precipitated with calcium chloride. Chemically treated chromium and nickel wastewaters are pumped to the copper-nickel tank, where pH is adjusted with acid or caustic. Solids are settled in three clarifiers in series with the addition of a flocculent. Clarified wastewater is pumped to a storage tank with level control. When the tank level reaches a set point, wastewater is pumped to a sand filter. Filtered wastewater flows to a discharge tank. Clarifier sludge is pressed. Filtrate is returned to the copper-nickel tank. Chrome-bearing rinsewaters are reduced by lowering pH with sulfuric acid and adding sodium metabisulfite. Reduced chrome rinsewaters are combined with other metal-bearing rinsewaters in the acid/alkali tank and neutralized with caustic. Solids are settled in a clarifier with the addition of flocculent polymer. Clarified wastewater flows through a surge tank and equalization tank. Wastewater then flows through a sand filter prior to discharge. Clarifier sludge is pressed. Filtrate is returned to the acid/alkali tank. Chrome-bearing rinsewaters are reduced in a 4-stage tank by lowering pH to 2.5 s.u. with sulfuric acid and adding sodium metabisulfite. Other rinsewaters and calcium chloride are mixed with reduced chrome rinsewaters to raise the pH. Solids are settled in a clarifier with the addition of polymer. Clarified wastewater discharges from a flow-through final effluent tank. Clarifier sludge is thickened and pressed. Filtrate is returned to the clarifier. Electroless copper rinse water is treated separately by plating on steel wool. Chrome-bearing rinsewaters (from chrome electroplating and chromate conversion) are reduced by lowering pH with sulfuric acid and adding sodium metabisulfite. Other rinsewaters are combined with reduced chrome rinsewaters and the pH is raised with sodium hydroxide. Solids are settled in a clarifier with the addition of polymer. Clarified wastewater is discharged. Clarifier sludge is dewatered in a filter press. Filtrate is returned to the clarifier. Cyanide plating rinse water is batch-treated with sodium hypochlorite, then combined with other rinsewaters. Average wastewater discharge for operating days during June 20081 97,000 gal/day 29,400 gal/day 146,000 gal/day 6,700 gallons/day 1 Discharge flow data were provided by the supervisor of enforcement, water quality & industrial surveillance, Northeast Ohio Regional Sewer District. The companies' permits require that self-monitoring data be reported to NEORSD. ------- 28 Facility #1 Facility #2 Facility #3 Facility #4 POTW that receives wastewater from facility Northeast Ohio Regional Sewer District Westerly WWTP Northeast Ohio Regional Sewer District Easterly WWTP Northeast Ohio Regional Sewer District Easterly WWTP Northeast Ohio Regional Sewer District Southerly WWTP POTW design flow 35 million gal/day 155 million gal/day 155 million gal/day 175 million gal/day ------- 29 Facility #5 Facility #6 Facility #7 Facility #8 Facility #9 Facility #10 Facility #11 Rinsing practice for chromium electroplated parts The chrome pieces are dipped into a single rinse tank. Rinse tank flows are approximately 1- 2 gpm. Discharge flow through chrome rinse tanks are 1-2 gpm. Rinse waters at the chrome tank flow between 2-4 gpm. After metal pieces are chromed they pass thru two dead rinse tanks, followed by a running tank at 1-2 gpm. After metal pieces are chromed they are dipped into two static rinse tanks, followed by three flowing rinse tanks 0.5 gpm. Rinse waters flow are 5 gpm. They operate three rinse tanks (one counterflow). The make up rinse rate is approximately 2 gpm. Rinse water pretreatment Pretreatment consists of a series of oxidations tank, followed by a flocculation tank, followed by a clarifier, then discharged to a sewer. The facility's pretreatment system consists of an oxidation tank, flocculation tank, clarifier holding tank, then finally to discharge. Flow thru entire pretreatment system is approximately 20 gpm. All rinse waters flow to a pretreatment system. Entire flow thru the pretreatment system is 50 -60 gpm. This facility operates a complete pretreatment system consisting of a pH adjust with flocculation agent, clarifier, final filtration then discharge. Flows thru the pretreatment system are typically 50 -55 gpm. This facility operates a complete pretreatment system consisting of a pH adjust with flocculation agent, clarifier, final filtration, then discharge. Flow thru the pretreatment system is about 55-55 gpm. The facility operates a complete pretreatment system consisting of chrome reduction tank, flocculation tank, clarifier, sand filters and pH adjust tank. Flow thru the pretreatment system is about 80 gpm. The electroplating shop operates a complete pretreatment system. It includes a chrome reduction tank, an equalization tank, a pH adjust tank, a clarifier, and finally, an effluent discharge pipe. Typical flows thru the pretreatment system varies between 50 - 55 gpm Average wastewater discharge POTW that receives wastewater from facility POTW design flow ------- 30 Appendix D. Hexavalent Chromium Controls Facility #1 Facility #2 Facility #3 Facility #4 Chemical fume suppressant and add-on air pollution control devices used at facility Mist Suppressant A, B and C Mist Suppressant D and E Mist Suppressant B and F Mist Suppressant D Tensiometer readings Surface tension not greater than 35 dynes/cm as measured by a tensiometer Surface tension not greater than 45 dynes/cm as measured by a stalagmometer Surface tension not greater than 35 dynes/cm as measured by a tensiometer Surface tension not greater than 45 dynes/cm as measured by a stalagmometer Amount of chemical fume suppressant used ~2.6 gal/week of Mist Suppressant A, ~1.5 gal/week of suppressant B, and ~0.9 gal/week of suppressant C.2 1.2 gal/week of mist suppressant D and 3.5 gal/week of mist suppressant E. (Mist suppressant E usage is about 1.1 gal/week for the chromic acid tanks and 2.4 gal/week for the chrome etch tank.)3 8-9 gal/week of mist suppressant F.4 The usage of mist suppressant B was not determined. 0.06 gal/week of mist suppressant D.5 2 The company will cease using mist suppressant C once it depletes its inventory. 3 These are average values based on the following usage - 16 gallons, 15 gallons, and 15 gallons of mist suppressant D during the third and fourth quarters of 2007 and the first quarter of 2008; 47 gallons, 44 gallons, and 46 gallons of mist suppressant E during the third and fourth quarters of 2007 and the first quarter of 2008. 300 mL/day of mist suppressant E (0.4 gallons/week) are added to the chromic acid tank of the metal substrate line and 500 mL/day of Mist Suppressant E (0.7 gallons/week) are added to the chromic acid tank of the plastic substrate line based on surface tension logs. The remainder of suppressant E's usage is for the plastic substrate line chromic acid etch tank. 4 This value was given verbally by Facility #3 Director of Engineering during the inspection. 5 This is an average value calculated by summing the amount of chemical fume suppressant recorded on the company's surface tension log from February 19, 2007 through June 12, 2008 and dividing by the number of calendar weeks during that period. ------- 31 Facility #5 Facility #6 Facility #7 Facility #8 Facility #9 Facility #10 Facility #11 Chemical fume suppressant and add-on air pollution control devices used at facility Mist Suppressant G Unknown Mist Suppressant H Mist Suppressant H Unknown Unknown Unknown Tensiometer readings Recent tensiometer value of 44 dynes/cm2 was measured on June 7, 2008. Latest tensiometer reading at chrome tank was measures at 43dynes/cm2 on April 10, 2008. Latest tensiometer values of 23.6 dynes/cm2 were measured on June 2, 2008. Latest tensiometer reading of 41dynes/cm2 was measured on June 9, 2008. Last tensiometer reading was 27.3 dynes/cm2 as measured on June 3, 2008. The last tensiometer value recorded at the chrome tank was 33.96 dynes/cm2 on May 29, 2008. Amount of chemical fume suppressant used ------- |