United States Environmental Protection Agency National Exposure Research Laboratory Research Triangle Park, NC 27711 Research and Development EPA/600/SR-97/037 June 1997 4vEPA Project Summary Field Evaluation at an Agricultural Chemical Manufacturing Facility of VOST and SemiVOST Methods for Selected CAAA Organic Compounds Joan T. Bursey, James F. McGaughey, and Raymond G. Merrill Laboratory testing and one field evalu- ation study had previously been per- formed to assess the performance of nonhalogenated volatile and semivolatile organic analytes from Title III of the Clean Air Act Amendments (CAAA) of 1990 in the Volatile Organic Sampling Train (VOST) and Semivolatile Organic Sam- pling Traing (SemiVOST) methods. For this work assignment, a second field evaluation study was performed at a different source to demonstrate that the methodology is not source-specific. At an agricultural chemical manufac- turing facility, an incinerator that burned chemical waste was selected for the second field method evaluation test site. The field test was designed according to the guidelines of the En- vironmental Protection Agency (EPA) Method 301, using gaseous and liquid dynamic spiking with three spiking schemes. Volatile organic compounds were spiked into two of four quadruple VOST trains as a gaseous spike; semivolatile organic compounds were spiked as a liquid spike into two of four quadruple SemiVOST trains either as a solution of Acid/Neutral com- pounds or Base/Neutral compounds. These two solutions were spiked in separate sampling runs to avoid compound losses due to known acid/base chemical reactions. A minimum of ten quadruple sampling runs each were performed for VOST, Acid/Neutral SemiVOST and Base/Neu- tral SemiVOST. Each quadruple run used four collocated sampling probes into four similar sampling trains, with two spiked trains and two unspiked trains. Statistical analysis of the results was performed according to the guide- lines of EPA Method 301, as well as EPA's Handbook of Quality Assurance/ Quality Control (QA/QC) Procedures for Hazardous Waste Incineration. Bias and precision were good for the Neutrals, poor for the Acidic and Basic com- pounds. The field test and statistical analysis of the results are discussed in this report. Using the EPA Method 301 criteria for acceptable performance (correction factor between 0.70 and 1.30, with rela- tive standard deviation of 50% or less), the VOST methodology showed accept- able performance in a chemical waste incinerator emissions matrix for the fol- lowing compounds: benzene, n-hexane, 2,2,4-trimethylpentane, and toluene. These VOST results are consistent with the first field test, where the carbon disulfide correction factor was also un- acceptable. Using the EPA Method 301 criteria for acceptable performance (cor- rection factor between 0.70 and 1.30. with relative standard deviation of 50% or less), the SemiVOST methodology showed acceptable performance in a chemical waste incinerator for the fol- lowing compounds: Acids: di-n-butyl phthalate, dimethyl phthalate; Bases: carbaryl, N-nitrosodimethylamine, N- ------- nitrosomorpholine; Neutrals: 1,4-diox- ane, 2,4-dinitrotoluene, 4-nitrobiphenyl, acetophenone, biphenyl, chlordane, cumene, DDE, dibenzofuran, ethylbenzene, isophorone, lindane, mVp-xylene, methyl isobutyl ketone, naphthalene, nitrobenzene, o-xylene, and toluene. In the first field test, the following compounds showed ac- ceptable performance: Acids: 2,4-dini- trophenol, 4-nitrophenol, 4,6-dinitro-o- cresol, o-cresol, phenol; Bases: 3,3'-dimethylbenzidine, 4,4'-methylene bis (o-chloroaniline), 4,4'-methylenedianiline, caprolactam, carbaryl, ethyl carbamate, N,N-diethylaniline, N-nitrosodimethylamine, N-nitrosomorpholine, propoxur, quinoline; Neutrals: 1,4-dioxane, 2,4-dinitrotoluene, 4- nitro-biphenyl, acetophenone, biphenyl, chlordane, cumene, DDE, dibenzofuran, dichlorvos, ethylbenzene, heptachlor, isophorone, lindane, m-/R-xylene, me- thyl isobutyl ketone, naphthalene, ni- trobenzene, o-xylene, parathion, and styrene. In general, far fewer of the polar semivolatile compounds (Acids and Bases) showed acceptable correc- tion factor and precision in the second field test than in the first field test be- cause of the challenging source condi- tions. Chemical interactions of polar compounds with a high moisture and chemically reactive background source matrix would be expected to be signifi- cant. The methodology is therefore not source-specific. If the methodology were source-specific, no compounds would have met acceptance criteria at the second source. However, some Ac- ids, some Bases, and most Neutrals met acceptance criteria at the second source, so the methodology is not source-specific. However, the perfor- mance of individual compounds is in- fluenced by conditions at a particular source. This Project Summary was developed by National Exposure Research Laboratory's Air Measurements Re- search Division, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering infor- mation at back). Introduction A field method evaluation test at an incinerator burning chemical wastes was performed for the VOST and the SemiVOST. The EPA methods were ap- plied exactly as written, with no deviation from the written methodology allowed. The objective of the field test was to establish the bias and precision of the VOST and SemiVOST specific analytes using Method 3011 criteria. Method 301, a method vali- dation protocol, served as the basis for the design and execution of this method validation study. The specific analytes were pesticides and nonhalogenated organic com- pounds listed in Title III of the CAAA of 1990, which had previously been tested at a coal-fired power plant. The chemical waste incinerator test site was chosen to demonstrate that the test methodology is not source-specific. The VOST methodology (sampling and analytical) consists of a combination of the following EPA methods: • For sampling, SW-846 Method 00302 Volatile Organic Sampling Train • For analysis, SW-846 Method 50413 Protocol for Analysis of Sorbent Cartridges from Volatile Organic Sampling Train: Wide-bore Cap- illary Column Technique. Method 0030 describes the collection of volatile principal organic hazardous con- stituents (POHCs) from the stack gas ef- fluents of hazardous waste incinerators. The method defines volatile POHCs as having boiling points less than 100°C (212°F). Method 0030 states that many compounds that boil above 100°C (212°F) may also be efficiently collected and ana- lyzed. The SemiVOST methodology consists of a combination of the following EPA methods: • For sampling, SW-846 Method 0010": Modified Method 5 Sampling Train • For sample preparation, SW-846 Pro- posed Method 3542s: Extraction of Semivolatile Analytes Collected Using Modified Method 5 (Method 0010) Sampling Train • For analysis, SW-846 Method 82 706: Gas Chromatography/Mass Spec trometry for Semivolatile Organics: Capillary Column Technique The SemiVOST methodology defines semivolatile organic compounds as com- pounds with boilinq points above 100°C (212°F). To be a candidate for the SemiVOST or VOST methodologies, an analyte must be able to be analyzed by gas chromatogra- phy/mass spectrometry (GC/MS). Labora- tory studies7 identified volatile and semivolatile organic compounds and pes- ticides listed in Title III of the CAAA of 1990 that could be analyzed by GC/MS. The candidate analytes identified for the VOST methodology were benzene, car- bon disulfide, n-hexane, 2,2,4-trimethylpen- tane, and toluene. Since the boiling point of toluene is 111 °C, toluene was also tested using SemiVOST. After laboratory confir- mation of the composition and concentra- tion of certified cylinders, the cylinders were used to perform dynamic spiking in the field method evaluation study. The candidate analytes identified for the SemiVOST methodology included 10 Ac- ids, 20 Bases, and 24 Neutral semivolatile organic compounds from Title III of the CAAA. These analytes were dynamically spiked in the field as an Acid/Neutral solu- tion or as a Base/Neutral solution. EPA Method 301 guidelines were fol- lowed for experimental design, number of samples collected, and statistical evalua- tion. A methodology (sampling and ana- lytical) is determined to be valid when it meets the acceptance criteria for bias and precision outlined in Method 301. Method 301 provides guidance on the experimen- tal design, the number of samples to be collected, and the calculations to deter- mine bias and precision. The bias and precision of an analyte must fall within a specified range (correction factor between 0.70 and 1.30, with relative standard de- viation of 50% or less). Method 301 re- quires either testing the candidate method side-by-side with a validated method, or introducing a known amount of the target analyte(s) into the sampling train without interrupting normal sampling procedures. For this field evaluation, the analytes were dynamically spiked in the field because comparable validated reference methods do not exist. Dynamic spiking is the ac- cepted means of introducing a known analyte. During dynamic spiking, analytes are introduced into the sampling train, as close as possible to the end of the probe, for the duration of the sampling run. EPA Method 301 provides guidelines for design of the sampling scheme to en- sure that a sufficient number of valid samples are collected to statistically evalu- ate precision and bias. For dynamic spik- ing in the field, four similar sampling trains are operated from four collocated probes. Two of the trains are dynamically spiked; two trains are unspiked. Method 301 re- quires at least six complete sampling runs (twelve paired spiked trains, twelve paired unspiked trains) in order to statistically assess the data. For this field test, eleven complete quad VOST sampling runs were made; eight complete Acid/Neutral sam- pling runs were made; and eleven com- plete Base/Neutral sampling runs were made. Data collected for each sampling run were statistically analyzed according to Method 301 and the QA/QC Handbook for Hazardous Waste Incineration.6 2 ------- Experimental Approach A chemical waste incinerator was se- lected as the test site for this field method evaluation study. A site presurvey estab- lished that no high levels of the com- pounds of interest were present in the background, and that the moisture level of source was high (approximately 55%). The dynamic spiking level projected for the VOST methodology was 250 ng per analyte; for the SemiVOST methodoloy, 500 |ig per analyte. To collect samples, stack gas was drawn from a single port in the stack through a quad probe. The stack gas was then di- rected to four similar VOST sampling trains or to four similar SemiVOST sampling trains. The quad probe contains four simi- lar heated sampling probes that can be inserted into the stack as one unit. The front end of the quad probe was posi- tioned in the center of the stack and re- mained in that location during each day of testing. The true concentration of the com- ponents of the stack gas was of no inter- est to this program, so traversing the stack was not required. For both VOST and SemiVOST methodologies, two of the qua- druple sampling trains were spiked and two were unspiked, according to the guide- lines of EPA Method 301. VOST dynamic spiking was performed by introduction of the compounds of inter- est from a certified gas cylinder at a flow rate of approximately 3 mL/min for the duration of the 20-min VOST sampling run. Method 0030 sampling procedures were followed. SemiVOST dynamic spik- ing was performed by introduction of the compounds of interest from a methylene chloride solution (either Acid/Neutrals or Base/Neutrals) through a heated glass el- bow equipped with a spiking injection port positioned between the end of the probe and prior to the heated filter. Approxi- mately 20 ml_ of the solution was intro- duced into the spiking injection port using a syringe pump, which operated at a rate appropriate for maintaining a hanging drop of the spiking solution throughout the one- hour sampling run. Method 0010 sam- pling procedures were followed. VOST field samples were analyzed ac- cording to the analytical procedures de- scribed in Method 5041. The VOST tubes were analyzed individually, with the back tube first, in order to determine the distri- bution of the analytes on the VOST tubes and to demonstrate that breakthrough had not occurred. SemiVOST field samples were prepared according to the proce- dures of Proposed Method 3542 and ana- lyzed according to the procedures of EPA Method 8270. Three samples were ana- lyzed for each SemiVOST train: the filter/ front half rinse, the XAD-2® and condenser rinse, and the condensate/condensate rinse. For both VOST and SemiVOST, field blanks, reagent blanks, method blanks, and method spikes were prepared and analyzed. Results and Discussion The SemiVOST method is most effec- tive for the Neutral compounds at the source tested. Because of the high mois- ture levels of the source, the polar and water-soluble Acidic and Basic compounds were washed through the sorbent into the condensate of the sampling train, where the pH-adjusted extraction required by the SemiVOST method did not result in quan- titative recoveries. The results of the statistical calculations are summarized below: • Four of five VOST analytes met EPA Method 301 acceptance criteria for bias and precision; • Nineteen of 24 Neutral compounds met EPA Method 301 acceptance cri- teria for bias and precision; • Two of 10 Acidic compounds met EPA Method 301 acceptance criteria for bias and precision; • Three of 20 Basic compounds met EPA Method 301 acceptance criteria for bias and precision. Comparable results are obtained when the statistical calculations are performed according to the guidelines of the EPA QA/QC Handbook. During the dynamic spiking experiments for the SemiVOST in the field, the Base/ Neutral spiking solution changed color from yellow to a dark green k proximately one9Qalf hour after the solution was poured into the syringe for spiking. Three aliquots of the spiking solution were ana- lyzed in the laboratory: an aliquot of the solution that was not sent to the field, an aliquot of unused solution that had been taken to the field but not opened, and remaining spiking solution from an aliquot of the solution that had changed color. The color change was due to photochemi- cal reaction of benzidine, 4,4'-methylene- dianiline, 3,3'-dimethylbenzidine, and 3,3'-dimethoxybenzidine, photoreactive dyes. The natural ultraviolet light in the field was sufficient to cause the photoreactive com- pounds to react and change the color of the solution. The aliquot of the solution remaining in the laboratory did not change color under the artificial ultraviolet light in the laboratory. Upon analysis of both VOST and SemiVOST, the presence of molecular io- dine (l2) in the background matrix was established by a major l2 chromatographic peak. The reactivity of the l2 was demon- strated by the disappearance of acenaphthene-d10 (a Method 8270 inter- nal standard added immediately before analysis) from several of the XAD-2® ex- tracts and by the appearance of iodinated compounds in the chromatograms. The XAD-2® extracts were quantified by using an alternative internal standard (Method 8270 uses six internal standards so an alternative is available). However, the chemical reactivity of the background source matrix did affect compound recov- eries, especially for the Acids and Bases. Reaction could occur in the sampling train, on the sorbent or in the condensate, dur- ing the extraction/concentration process, or in the heated injection port of the gas chromatograph. When the analyte distribution of the VOST samples was determined, all of the VOST analytes showed >90% recovery of spiked analyte from the front (Tenax®) tube. When the analyte distribution of the SemiVOST samples was determined, the Neutral compounds, as expected, showed primary recovery from the XAD-2® sor- bent. The major exception was 1,4-diox- ane, a polar water-soluble Neutral com- pound recovered mostly from the conden- sate. Polar water-soluble compounds washed through the sorbent. The Acid and Basic analytes showed primary re- covery from the condensate, with the least volatile analytes recovered from the filter. Low recoveries for the Acidic and Basic compounds can be attributed to the high moisture level of the source: the com- pounds washed through the XAD-2® and into the condensate, where they were ex- tracted with poor recoveries. Laboratory experience has demonstrated that polar water-soluble analytes present in an aque- ous solution (the condensate) are recov- ered poorly (30-60% recovery) using the pH-adjusted extraction technique specified by the SemiVOST method. Three factors in this field method evalu- ation study acted to reduce the recoveries of polar semivolatile analytes from the SemiVOST train: • A high level of moisture in the source washed the polar (Acidic and Basic) analytes through the SemiVOST train to be retained in the condensate. Recovery of polar water-soluble semivolatile organic analytes from aqueous media tends to be poor un- der standard conditions of pH-adjusted extraction. • The high level of molecular l2 present in the source reacted more readily with functionalized molecules (Acids and Bases) than with Neutral com- 3 ------- pounds. Chemical reaction with the background stationary source matrix reduced recoveries of Acidic and Ba- sic compounds. • Photochemical reactivity reduced the recovery of photochemically active compounds such as dyes or dye in- termediates. Conclusions and Recommendations The following conclusions may be drawn from the results of the second VOST and SemiVOST method evaluation field test for selected Clean Air Act analytes: • The method is not source-specific. If the method were source-specific, the methodology would succeed or fail completely, depending upon the source. There would be only one source or only one type of source where any compounds meet accep- tance criteria. At all other sources or source types, the methods would fail completely and no compounds would meet acceptance criteria. Since both VOST/SemiVOST field tests demon- strate some compounds that perform acceptably at both source types, the method is not specific to the source. However, factors characteristic of a specific source affect the performance of specific compounds, and an over- all evaluation of the performance of the methodology at the two sources says that both the VOST and SemiVOST methods perform better and more consistently with Neutral (i.e., non-polar) compounds. • Using the EPA Method 301 criteria for acceptable performance (correction fac- tor between 0.70 and 1.30, with relative standard deviation of 50% or less), the VOST methodology showed acceptable performance in a chemical waste incin- erator emissions matrix for the following compounds: benzene,n-hexane, 2,2,4-tri- methylpentane, and toluene. • Using the EPA Method 301 criteria for acceptable performance (correc- tion factor between 0.70 and 1.30, with relative standard deviation of 50% or less), the SemiVOST methodology showed acceptable performance in a chemical waste incinerator for the fol- lowing compounds. Acids: di-n-butyl phthalate, dimethyl phthalate; Bases: carbaryl, N-nitrosodimethylamine, N- nitrosomorpholine; Neutrals: 1,4-diox- ane, 2,4-dinitrotoluene, 4-nitrobiphenyl, acetophenone, biphenyl, chlordane, cumene, DDE, dibenzofuran, ethylben- zene, isophorone, lindane, mYp-xylene, methyl isobutyl ketone, naphthalene, ni- trobenzene, o-xylene, and toluene. • The following compounds did not meet Method 301 acceptance criteria: Ac- ids: 2,4-dinitrophenol, 4-nitrophenol, bis(2-ethylhexyl) phthalate, 4,6-dinitro- o-cresol, m-/p-cresol, o-cresol, phe- nol; Bases: 2-acetylaminofluorene, 3,3'-dimethoxybenzidine, 3,3'-di- methylbenzidine, 4,4'-methylene bis (o-chloroaniline), 4,4'-methylene- dianiline, 4-aminobiphenyl, aniline, benzidine, caprolactam, dimethyl- aminoazobenzene, ethyl carbamate, N,N-diethylaniline, N,N-dimethyl- aniline, o-anisidine, o-toluidine, propoxur; Neutrals: dichlorvos, hep- tachlor, methoxychlor, parathion, sty- rene. • The chemical composition of the back- ground source matrix is a significant factor in the success or failure of the sampling and analytical methodology. • In a reactive background emissions matrix, reactive organic compounds (such as acids and bases) show the effects of interaction with the back- ground matrix far more strongly than less reactive compounds such as hy- drocarbons. • Interaction of the spiking solution with natural ultraviolet light can be a sig- nificant factor in determining the re- coveries of spiked compounds. The Base/Neutral spiking solution con- tained photoreactive dyes (members of the benzidine family), which re- acted with the natural ultraviolet ra- diation at the test site as the Base/ Neutral solution was being poured into the syringe for spiking. • Application of the SemiVOST to polar reactive semivolatile compounds pro- duces widely variable results at dif- ferent sources, depending upon the reactivity of the background matrix. Laboratory tests will demonstrate that the methodology can in general be applied to a particular analyte and can predict certain failure of the meth- odology under field conditions. That is, if an analyte cannot be quantita- tively extracted from the XAD-2® and analyzed reproducibly, failure under field conditions is highly probable. However, successful performance un- der laboratory conditions does not guarantee that field testing at a given source will be successful. The char- acteristics of the particular source must be considered. • When a source such as the chemical waste incinerator has a high moisture content, it is essential to monitor the desorption temperature of the VOST tubes to ensure that the tubes be- come sufficiently hot for quantitative desorption of the collected analytes. The high level of moisture collected on the tubes during sampling (up to several mL of collected water) slowed the heating of the tubes being des- orbed for analysis. If the temperature does not reach the desorption tem- perature specified by Method 5041 for the period of time required by the method, analyte recoveries will not be quantitative. On the basis of the results of this field method evaluation study, the following recommendations can be made: • When polar water-soluble semivolatile organic compounds are sampled by the SemiVOST at a source with high moisture, the polar compounds tend to wash through the sampling train to be collected in the condensate. When polar water-soluble semivolatile com- pounds are dissolved in the conden- sate, these compounds are recovered poorly by the pH-adjusted extraction techniques required by the SemiVOST. Two possible solutions to this prob- lem should be explored: - Investigate the use of sorbents which retain polar water-soluble semivolatile organic compounds better than XAD-2® does to pre vent the compounds from wash- ing through the SemiVOST train to the condensate. - Because the semivolatile organic compounds dissolved in the con- densate are typically present at low concentrations in a large vol- ume of condensate, direct analysis of the condensate (direct aqueous injection into GC/MS or High Per- formance Liquid Chromatograph [HPLC]) will probably not be sen- sitive enough. Some technique such as solid phase extraction for con- centrating the organic compounds for analysis will be needed in order to perform successful analysis within the detection limits of the instrumen- tation. • Faced with a candidate analyte for either VOST or SemiVOST for which no method validation information is available, laboratory experimentation can aid in establishing the validity of assigning the analyte to a particular methodology: - Determine whether the compound can be analyzed by either liquid in- 4 ------- jection (GC/MS) or purge and trap • GC/MS. This information is frequently available from the literature. If the compound cannot be analyzed by GC/MS, neither VOST nor SemiVOST can be applied. Determine whether the compound can be recovered quantitatively from the sorbent. If the compound can- not survive thermal desorption or extraction/concentration, neither VOST nor SemiVOST can be ap- plied. If the compound is recovered poorly or erratically, an informed decision regarding the applicability of the methods as screening meth- ods can be made. Determine whether the compound can be recovered quantitatively from a dynamically-spiked sampling train. If the compound cannot be recov- ered after passage through a heated sampling train, the sampling meth- odology will not collect the com- pound quantitatively and is not ap- propriate. Use of some basic labo- ratory experimentation can prevent .the erroneous assumption that be- cause an analyte was not observed when VOST or SemiVOST was ap- plied in the field, the analyte is not present at the test site. If neither method is appropriate for the analyte, the organic compound could not possibly be observed, even if present. Careful consideration of the chemical properties of candidate analytes is essen- tial in predicting success or failure in the application of VOST or SemiVOST sam- pling and analytical methodology. For compounds with marginal or unac- ceptable performance in the VOST or SemiVOST methods, a detailed study of the chemical properties of these com- pounds may provide guidance for the modi- fication of existing methods to optimize the methodology for these compounds. References 1. Appendix A to Part 63 — Test Meth- ods. Method 301— Field Validation of Pollutant Measurement Methods from Various Waste Media, Federal Regis- ter Vol. 57, No. 250, December 29, 1992, pp 61998-62002. 2. U. S. Environmental Protection Agency. Method 0030, in Test Methods for Evaluating Solid Waste, Physical/ Chemical Methods, SW-846 Manual, 3rd Ed. Document No. 955-001- 0000001. Available from Superinten- dent of Documents. U. S. Government Printing Office, Washington, DC, No- vember 1986. 3. U. S. Environmental Protection Agency. Method 5041, in SW-846 Manual, Third Update to 3rd Ed. U. S. Government Printing Office, Washington, DC, May 1995. 4. U. S. Environmental Protection Agency. Method 0010, in Test Methods for Evaluating Solid Waste, Physical/ Chemical Methods, SW-846 Manual, 3rd Ed. Document No. 955-001- 0000001. Available from Superinten- dent of Documents. U. S. Government Printing Office, Washington, DC, No- vember 1986. 5. U. S. Environmental Protection Agency. Proposed Method 3542. Proposed for Inclusion in Test Methods for Evaluat- ing Solid Waste, Physical/Chemical Methods, SW-846 Manual, 3rd Ed. U. S. Government Printing Office, Wash- ington, DC, January 1995. 6. U. S. Environmental Protection Agency. Method 8270, in Test Methods for Evaluating Solid Waste, Physical/ Chemical Methods, SW-846 Manual, 3rd Ed. Document No. 955-001- 0000001. Available from Superinten- dent of Documents. U. S. Government Printing Office, Washington, DC, No- vember 1986. 7. U. S. Environmental Protection Agency Contract No. 68-D1-0010, Work As- signment 74 to Radian Corporation, Laboratory Evaluation of VOST and SemiVOST for Selected CAAA Organic Compounds. September 30. 1994. 8. Handbook. Quality Assurance/Quality Control (QA/QC). Procedures for Haz- ardous Waste Incineration, EPA-625/ 6-89-023, Cincinnati, OH. 1990. 5 ------- Joan T. Bursey, James F. McGaughey, and Raymond G. Merrill are with Eastern Research Group, In., Morrisville, NC 27560. Merrill D. Jackson is the EPA Project Officer (see below). The complete report, entitled "Field Evaluation at an Agricultural Chemical Manu- facturing Facility of VOST and SemiVOST Methods for Selected CAAA Organic Compounds, "(Order No. PB97-174585; Cost: $85.00, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air Measurements Research Division National Exposure Research Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 United States Environmental Protection Agency National Risk Management Research Laboratory (G-72) Cincinnati, OH 45268 Official Business Penalty for Private Use $300 BULK RATE POSTAGE & FEES PAID EPA PERMIT No. G-35 EPA/600/SR-97/037 ------- |