DRAFT C0701-1TOC/UST 2/5/88 1 CHARACTERISTIC WASTE DESIGNATION OF SOILS CONTAMINATED WITH PETROLEUM PRODUCTS DRAFT REPORT FEBRUARY 5, 1988 EPA Contract No. 68-01-7381 MRI Project No. 8900-07-54 Prepared for: Mr. Michael R. Ka11nosk1 Policy and Standards Division Office of Underground Storage Tanks U. S. Environmental Protection Agency 401 M Street SW Washington, O.C. 20460 ------- ' L--~ / Rl$ MIDWEST RESEARCH INSTITUTE Suite 35! 401 Harrison Oaks Boulevari Carv N C 275' April 6, 1988 Mr. James DeLong Third Floor 1029 Vermont Avenue, N.W. Washington, D.C. 20005 Dear Mr. DeLong, Mr. Michael Kal1nosk1 of the U.S. EPA Office of Underground Storage Tanks requested that I send to you a copy of the report our office prepared entitled "Characteristic Waste Designation of Soils Contaminated With Petroleum Products." This draft report has not been through complete peer review. This review process 1s currently being conducted and the revised final report 1s anticipated to be 1n the OUST docket on April 15, 1988. Should you have any questions about the content of this report, please call me at (919)467-5215 or Michael Kal1nosk1 at (202)382-7989. Sincerely iwnet H. Dean Environmental Scientist ------- DRAFT C0701-1T0C/UST 2/5/88 11 PREFACE This report describes work performed by Midwest Research Institute (MRI) under EPA Contract No. 68-01-7383, MRI Project Number 8900-57-L. This work Involved the performance of the Toxicity Characteristic Leaching Procedure on two son types that had been contaminated with petroleum products, as well as a discussion of the regulatory impacts of the results. This report has been prepared 1n accordance with directions provided by the Office of Underground Storage Tanks, U. S. Environmental Protection Agency. APPROVED BY: MIDWEST RESEARCH INSTITUTE Lloyd T. Taylor Director 11 ------- DRAFT C0701-1T0C/UST 2/5/88 111 TABLE OF CONTENTS Page LIST OF FIGURES LIST OF TABLES EXECUTIVE SUMMARY 1 SECTION 1.0 INTRODUCTION 3 SECTION 2.0 DATA PRESENTATION 5 2.1 SOIL CHARACTERIZATION 5 2.2 CONTAMINATED SOIL PREPARATION 5 2.3 FUEL CHARACTERIZATION 5 2.4 CONTAMINATED SOIL CHARACTERIZATION 6 2.5 IGNITABILITY CHARACTERISTIC TESTS 7 2.6 TOXICITY CHARACTERISTIC LEACHING PROCEDURES (TCLP) TEST 7 SECTION 3.0 REGULATORY ANALYSIS 9 3.1 IGNITABILITY CHARACTERISTIC 9 3.2 TOXICITY CHARACTERISTIC 9 3.3 IMPACTS OF THE PROPOSED TC REGULATION 9 3.4 IMPACT OF TOXICITY CHARACTERISTIC ON THE HAZARDOUS SUBSTANCE TANKS 10 3.5 APPENDIX VIII IMPACTS 11 3.6 CERCLA REPORTABLE QUANTITY IMPACTS 12 SECTION 5.0 REFERENCES 14 APPENDIX A. TCLP ANALYSIS RESULTS AND QA/QC PROCEDURES A-l A.l MATERIALS AND METHODS A-l A.1.1 Soil Characterization A-l A. 1.2 Sqo1l/Fuel Products Mixing A-2 A.1.3 Determination of Characteristic of Ign1tab1l1ty of Fuel-Contaminated Soils A-2 A.1.4 Toxic Characteristic Leaching Procedure (TCLP) A-3 A.1.5 Analysis A-3 111 ------- DRAFT C0701-1T0C/UST 2/5/88 1v TABLE OF CONTENTS Page A.2 RESULTS AND DISCUSSION A-5 A.2.1 Soil Characterization A-5 A.2.2 Soil/Fuel Product Mixing Results A-6 A.2.3 Characteristic of Ign1tab1l1ty Results... A-6 A.2.4 Analytical Results A-6 A.3 QUALITY ASSURANCE/QUALITY CONTROL A-14 A.3.1 Volatile Organic QA/QC A-15 A.3.2 SemlvolUe QA/QC A-17 A.4 REFERENCES FOR APPENDIX A A-21 1 v ------- DRAFT C0701-1T0C/UST 2/5/88 v LIST OF FIGURES Page Figure A-l. Leachate/nomlnal soil concentration plots of volatile aromatic analytes 1n Action topsoll/ diesel mixtures A- Flgure A-2. Leachate/nomlnal soil concentration plots of semivolatile aromatic analytes 1n Action topsoll/ diesel mixtures A- Flgure A-3. Leachate/nomlnal soil concentration plots of semlvolatlle aromatic analytes 1n Action topsoll/ diesel mixtures A- Flgure A-4. Leachate/nomlnal soil concentration plots of volatile aromatic analytes 1n Action topsoll/ gasoline mixtures A- Flgure A-5. Leachate/nomlnal soil concentration plots of semlvolatlle aromatic analytes 1n Action topsoll/ gasoline mixtures A- Flgure A-6. Leachate/nomlnal soil concentration plots of volatile aromatic analytes 1n Kaw River soil/ diesel mixtures A- Flgure A-7. Leachate/nomlnal soil concentration plots of semlvolatlle aromatic analytes 1n Kaw River soil/ diesel mixtures A- Flgure A-8. Leachate/nomlnal soil concentration plots of semlvolatlle aromatic analytes 1n Kaw River soil/ diesel mixtures A- Flgure A-9. Leachate/nomlnal soil concentration plots of volatile aromatic analytes in Kaw River soil/ gasoline mixtures A- Flgure A-10. Leachate/nomlnal soil concentration plots of semlvolatlle aromatic analytes 1n Kaw River soil/ gasoline mixtures A- v ------- DRAFT C0701-1T0C/UST 2/5/88 v1 LIST OF TABLES Page TABLE 1. SOIL CHARACTERIZATION RESULTS FOR KAW RIVER SOIL AND ACTION TOPSOIL TABLE 2. CONCENTRATIONS OF THE SOIL/FUEL PRODUCT MIXTURES TABLE 3. VOLATILE AND SEMIVOLATILE ORGANIC TARGET ANALYTE CONCENTRATIONS IN UNLEADED GASOLINE, DIESEL FUEL, AND NO. 6 FUEL OIL TABLE 4. IGNITABILITY TEST RESULTS FOR FUEL CONTAMINATED SOILS TABLE 5. CONCENTRATIONS OF PROPOSED TCLP COMPOUNDS IN GASOLINE CONTAMINATED SOIL SAMPLES AND LEACHATE TABLE 6. PROJECTED HAZARDOUS SOIL CONCENTRATIONS OF PURE TOLUENE AND BENZENE TABLE 7. HAZARDOUS SUBSTANCES STORED IN UST's BY NUMBER OF TANKS TABLE 8. CONCENTRATIONS AND REPORTABLE QUANTITIES OF HAZARDOUS SUBSTANCES FOUND IN PETROLEUM PRODUCTS TABLE A-l. SOIL CHARACTERIZATION RESULTS FOR KAW RIVER SOIL AND.. A- ACTION TOPSOIL TABLE A-2. CONCENTRATIONS OF SOIL/FUEL PRODUCT MIXTURES A- TABLE A-3. IGNITABILITY TEST RESULTS FOR FUEL CONTAMINATED SOILS A- TABLE A-4. VOLATILE AND SEMIVOLATILE ORGANIC TARGET ANALYTE CONCENTRATIONS IN UNLEADED GASOLINE, DIESEL FUEL, AND NO. 6 FUEL OIL A- TABLE A-5. THEORETICAL CONCENTRATIONS OF SELECTED TARGET ANALYTES IN ACTION TOPSOIL/FUEL MIXTURES AT THE VARIOUS NOMINAL CONCENTRATIONS A- TABLE A-6. THEORETICAL CONCENTRATIONS OF SELECTED TARGET ANALYTES IN KAW RIVER SOIL/FUEL MIXTURES AT THE VARIOUS NOMINAL CONCENTRATIONS A- TABLE A-7. CONCENTRATIONS OF CONTAMINATED ACTION TOPSOIL LEACHATE A- v1 ------- DRAFT C0701-1T0C/UST 2/5/88 v11 LIST OF TABLES (continued) Page TABLE A-8. CONCENTRATIONS OF CONTAMINATED KAW RIVER SOIL LEACHATE A- TABLE A-9. RESULTS OF RECOVERY CALCULATIONS FOR TCLP OF FUEL- CONTAMINATED ACTION TOPSOIL A- TABLE A-10. RESULTS OF RECOVERY CALCULATIONS FOR TCLP OF FUEL- CONTAMINATED KAW RIVER SOIL A- v i 1 ------- DRAFT C0701-1/UST 2/5/88 EXECUTIVE SUMMARY BACKGROUND The quantity of soil contaminated by releases of petroleum products froai underground storage tanks 1s potentially very large. These soils could be classified as hazardous waste based on the "conta1ned-1n" policy defined by OSW 1f they exhibit a hazardous characteristic. Under existing regulations, the characteristic that soil contaminated with fuel products 1s likely to exhibit 1s that of 1gn1tab1l1ty. However, 1n June 1986, EPA proposed amending Its hazardous waste Identification regulations by expanding the Toxicity Characteristic to Include additional chemicals, several of which are constituents of fuel products. OBJECTIVE The objectives of this study were to determine 1f soils contaminated with petroleum products would be classified as hazardous under the 1gn1tab1l1ty characteristic or the proposed toxicity characteristic, to examine the Implications of regulating these soils as hazardous waste, and to develop recommendations for regulating these soils. To determine whether soils were hazardous, sandy and clayey soil samples were contaminated with up to 10,000 ppm of unleaded gasoline, dlesel fuel, and No. 6 fuel oil. The samples were subjected to both the 1gn1tab111ty and toxicity characteristic test procedures. Toxicity characteristic leachates were analyzed for selected RCRA hazardous products, RCRA Appendix VIII compounds, and compounds subject to proposed CERCLA reportable quantity requirements. RESULTS OF CHARACTERISTIC TESTS • None of the contaminated soils exhibited the 1gn1tab1l1ty characteristic. • Benzene concentrations 1n leachates from some gasoline contaminated soils exceeded the proposed TCLP regulatory level of 70 ug/L. Hazardous leachate was produced from sandy soil contaminated with 1,000 to 10,000 ppm of gasoline. No other gasoline constituents exceeded proposed regulatory levels. However, for the most contaminated sandy soil sample, the toluene concentration was within 10 percent of the proposed regulatory level of 14,400 ug/L. 1 ------- DRAFT C0701-1/UST 2/5/88 • Leachates from soil contaminated with up to 10,000 ppm of dlesel fuel and No. 6 fuel oil did not exceed any of the proposed TCLP regulatory levels. • Napthalene was the only Appendix VIII compound (other than benzene and toluene) that was detected 1n TCLP leachates. It was detected 1n leachates from both sandy and clayey soils contaminated with each of the fuels. EFFECT OF REGULATIONS ON CONTAMINATED SOIL MANAGEMENT • Soil contaminated up to Its saturation point by a fuel product release would not be hazardous based on the 1gnitab1l1ty characteristic. • The quantity of contaminated soil that could be classified as hazardous under the toxicity characteristic 1s not known, but 1t 1s estimated that 33,000 UST systems containing gasoline may be leaking. Thus, 1t might be necessary to manage a very large quantity of hazardous soil 1f the proposed TCLP amendment 1s promulgated. • Because benzene, a volatile compound, was the only constituent detected 1n the TCLP leachates above proposed regulatory levels, treatment techniques such as air stripping and vacuum extraction could be viable alternatives to disposal of the hazardous soil 1n Subtitle C landfills. • Releases from UST's containing benzene or toluene as pure product may produce a larger quantity of contaminated soil than an equivalent volumetric release of gasoline. • Because napthalene, an Appendix VIII constituent, was detected 1n the leachates from soils contaminated with each of the fuels, an Implementing Agency could classify as hazardous those soils that would not be hazardous under the toxicity characteristic. • Fuel releases that did not generate hazardous soils could still be subject to proposed CERCLA reportable quantity requirements. Gasoline releases of about 100 gallons would contain a reportable quantity of benzene. Diesel fuel and No. 6 fuel oil releases of 27,000 gallons and 125,000 gallons, respectively, would contain reportable quantities of napthalene. 2 ------- DRAFT C0701-1/UST 2/5/88 1.0 INTRODUCTION The U. S. Environmental Protection Agency (EPA) Office of Underground Storage Tanks (OUST) 1s responsible for developing programs for regulating underground storage tanks that contain petroleum products and hazardous substances. Under Subtitle I of the Resource Conservation and Recovery Act (RCRA), the EPA must promulgate standards for release detection, recordkeeping, reporting, corrective action, and closure for new and existing tanks. They must also promulgate standards for design, construc- tion, Installation, and compatibility of new tanks. To develop these standards 1t 1s Important to know what concentrations of contaminants 1n soils constitute a hazard and what the Implications are of regulating such soils as hazardous waste. Currently, soils contaminated with petroleum products could be classified as hazardous 1f they were to fall the criteria for the 1gn1tab1l1ty characteristic (40 CFR 261.21). Also, these soils could be classified as hazardous 1f they were to fall existing toxicity characteristics. The current toxicity characteristic leaching procedure or the EP toxicity procedure 1s based on levels of metals and pesticides. The only metal of concern to petroleum product tank owners would be lead and 1t 1s not expected to be present 1n significant enough concentrations to result 1n test failure. However, the toxicity characteristic proposed by the Office of Solid Waste (OSW) (51 FR 21648, June 13, 1986) Includes a number of new organic compounds. Some of these compounds, such as benzene and toluene, are commonly found 1n petroleum products. The proposed criteria for defining hazardous wastes (51 FR 21648, June 13, 1986) also Include a test for determining the mobility (by leaching) of toxic organic and Inorganic constituents 1n wastes: the toxicity characteristic leaching procedure (TCLP). Regulatory levels for selected constituents 1n TCLP extracts are stipulated as part of the criteria for defining hazardous wastes. There currently are no data available on the concentration of toxicity characteristic contaminants 1n petroleum-contaminated soils therefore, 1t 1s unclear whether soils contaminated by leaking UST's will pass or fall a characteristics test. Other Issues that remain are the quantity of soils requiring disposal at a Subtitle C facility under a 3 ------- DRAFT C0701-1/UST 2/5/88 Toxicity Characteristic (TC) classification or other regulatory classification mechanism, and the environmental and economic impacts of both disposal and alternatives to disposal. The EPA initiated this study to help resolve these Issues. Specific objectives of this study were to determine whether soils contaminated with petroleum products would be defined as hazardous waste under the 1gn1tab1l1ty characteristic or the TC criteria proposed by OSW and to evaluate the Impact of the regulation through estimation of the potential volume of soil affected by both petroleum products and hazardous substances. Two different types of soil were contaminated with three different petroleum products (unleaded gasoline, No. 2 dlesel fuel, and No. 6 fuel oil) at several concentration levels and subjected to both the 1gn1tab1l1ty and TCLP tests. The TCLP extracts were analyzed and the results were examined to determine which contaminated soils would be defined as hazardous. This report presents the results of the 1gn1tab1l1ty characteristic test and the analytical results for the petroleum products and the TCLP extracts. The report also discusses the regulatory Implications of the results and presents recommendations for resolving related Issues. The experimental procedures used to generate the data, the quality assurance/quality control (QA/QC) procedures followed, and the results of QA/QC performed during sample preparation and analysis are presented 1n the Appendix. 4 ------- DRAFT C0701-1/UST 2/5/88 2.0 DATA PRESENTATION 2.1 SOIL CHARACTERIZATION Two different types of soils were used to determine the mobility of petroleum products mixed 1n soil. The first soil was a light-colored, f1ne-gra1ned, sandy river soil obtained from Kaw Sand Company. The second soil was dark and loamy with a higher apparent clay content than the Kaw soil. This soil was obtained from Action Topsoll. Each soil was charac- terized for Its water content. The Action Topsoll and Kaw River soil were 13.5 and 10.1 percent water by weight, respectively. A rough particle size characterization (percent sand, silt, and clay) was also performed on each soil using a method based on an American Society of Testing and Materials method. Both soils were approximately 50 percent silt, but their sand and clay contents were different. The Action Topsoll was 3.7 percent sand and 41 percent clay while the Kaw River soil was 45 percent sand and 4 percent clay. Table 1 presents the results of these characterization tests. Appendix A contains more detail on the analytical procedures used for the characterization tests. 2.2 CONTAMINATED SOIL PREPARATION Three different types of fuel products (unleaded gasoline, dlesel oil, and No. 6 fuel oil) were mixed with the two soils described above. Each soil was spiked with each of the fuel products at five different nominal concentration levels (10,000, 5,000, 1,000, 100, and 10 ppm). The actual fuel concentrations 1n each of the mixtures are presented In Table 2. These values were used to estimate the pre-TCLP, or theoretical, concentration of each analyte 1n the soil and to perform subsequent comparisons with the TCLP extract concentrations. A detailed discussion of the product mixing procedures 1s presented 1n Appendix A. 2.3 FUEL CHARACTERIZATION The data obtained from the analysis of the three fuels for their volatile and semlvolatlle organic composition are presented 1n Table 3. The compounds are on either the proposed TCLP 11st (51 FR 21648), Appendix VIII compounds 11st, and/or CERCLA 11st of hazardous*substances for which there are reportable quantity requirements. The results for analyses of additional semlvolatlles are presented 1n Appendix A. 5 ------- DRAFT C0701-1/UST 2/5/88 The unleaded gasoline was, as expected, very rich 1n volatile aromatic hydrocarbons. The most abundant was toluene, followed by m-xylene, o- and p-xylene, benzene, and ethylbenzene. In contrast to very high levels of volatile organic compounds 1n gasoline, relatively low levels of semlvolatlles were observed 1n this fuel. Only naphthalene (198 ug/g) was detected at a level significantly higher than the quantitation limit. Diesel fuel had lower levels of volatile aromatic compounds than the gasoline; m-xylene and o- and p-xylene were the most abundant volatile aromatlcs at approximately 650 ug/g. Toluene and ethylbenzene were detected at between 310 and 360 ug/g, and benzene was detected, but at a concentration lower than the quantitation limit. The most abundant semlvolatlle aromatic compound 1n dlesel fuel also was napthalene (658 ug/g). No. 6 fuel oil had lower concentrations of volatile aromatic hydrocarbons than gasoline and dlesel fuel. The most concentrated analytes were m-xylene (69 ug/g), o- and p-xylene (65 ug/g), and toluene (60 ug/g). Ethylbenzene and benzene were detected at less than 25 ug/g. The most abundant semlvolatlle organlcs were large four-ring aromatlcs (benzo[a]anthracene and chrysene), with significantly lower concentrations of naphthalene. The latter 1s expected to Impact TCLP extracts more than the more abundant large aromatlcs due to significant differences 1n their aqueous solubilities. 2.4 CONTAMINATED SOIL CHARACTERIZATION The theoretical concentrations of each target analyte 1n Action Topsoll and Kaw River soil were obtained by multiplying the concentration of every analyte found 1n each fuel times the actual concentration of that fuel 1n each soil. For example, 1f the concentration of benzene 1n unleaded gasoline was 13,800 ug/g, and 1f the concentration of gasoline 1n Kaw River soil (nominal concentration of 100 ppm) was 0.103 g/kg, then the theoretical concentration of benzene 1n this contaminated soil would be: 13,800 ug/gx0.103 g/kg = 1,421 ug/kg In the absence of data validating these theoretical.concentrations, 1t will be assumed that minimal evaporative losses of all analytes occurred. This may not be a valid assumption, especially for the volatile 6 ------- DRAFT C0701-1/UST 2/5/88 organic compounds. However, every possible measure was taken to minimize any such possible losses. 2.5 IGNITABILITY CHARACTERISTIC TESTS To determine whether fuel-contaminated soil would be 1gn1table, both soil types were tested using the two most concentrated gasoline contaminated samples and the most concentrated samples contaminated with dlesel fuel and No. 6 fuel oil. The samples were contacted with the flame of a low-flame bunsen burner and observed during and after flame contact for any flashing or sustained combustion. This procedure 1s based on guidance provided 1n SW 846, Volume 1C, Chapter 7, Section 7.1.2.2. A summary of the results 1s presented 1n Table 4. Flashing or sustained combustion was not observed 1n any of the samples. Thus, none of the samples exhibited the characteristic of 1gn1tab111ty. 2.6 TOXICITY CHARACTERISTIC LEACHING PROCEDURES (TCLP) TEST Samples of the fuel-contaminated soils were subjected to the TCLP protocol described 1n the Federal Register (Volume 51, No. 216, November 7, 1986). An acetate buffer solution was used as the leaching medium. The results of the analyses of the TCLP leachates obtained from gasoline-contaminated Kaw River soil samples are presented 1n Table 5. The actual fuel concentrations 1n the soil, the theoretical analyte concentrations, and the proposed regulatory levels also are presented 1n Table 5. The leachates from Kaw River soil containing 1,000, 5,000, and 10,000 ppm of gasoline and from Action topsoll containing 10,000 ppm of gasoline exceeded the proposed regulatory level for benzene of 70 ug/L. Toluene leachate concentrations were high 1n the same samples, but only the leachate from the most contaminated sample approached the regulatory level of 14,400 ug/L. The only semlvolatlles on the proposed TCLP 11st that were detectable 1n gasoline were 3-and 4-methylphenol. However, these compounds were not detected 1n the leachates. The results of the analyses of the TCLP leachates obtained from Action topsoll samples are also presented 1n Table 5. As with the Kaw River samples, benzene was the only analyte that exceeded the proposed TCLP regulatory levels 1n any of the samples. The leachate from the Action topsoll sample contaminated with 10,000 ppm of gasoline contained 109 ug/L and 1t was the only leachate to exceed the proposed regulatory level. 7 ------- DRAFT C0701-1/UST 2/5/88 The difference 1n volatlles concentrations 1n the leachates (shown 1n Table 5 and Tables A-7 and A-8) for similar concentrations 1n the soil Indicates that volatlles bind with Action topsoll (a clayey soil) much more effectively than they bind with Kaw River soil (a sandy soil). This effect 1s represented by the extraction efficiency which 1s a measurement of the percentage of the analyte 1n the soil sample that 1s extracted to the leachate. As the concentration of the analyte Increases 1n the Action topsoll, the extraction efficiency Increases. This Indicates that the capacity of the soil to bind with the analytes 1s reduced as the concen- tration of the analyte 1n the soil increases. However, the extraction efficiency of volatlles from Kaw River soil remains constant or decreases which may Indicate that the binding effect is not a function of the concentration 1n this soil and that some of the analytes approach satura- tion 1n the leaching fluid at the higher concentrations 1n the soil. The leachates from soils contaminated with dlesel fuel and No. 6 fuel oil did not exceed any of the proposed TCLP regulatory levels, except for three questionable data points. The benzene concentration of 42.9 yg/L 1n the leachate from the Kaw River soil sample contaminated with 10,000 ppm of dlesel fuel came the closest to the proposed regulatory level of 70 yg/L. Toluene leachate concentrations were at least three orders of magnitude below the proposed regulatory level of 14,400 yg/L for all samples. With the exception of the questionable data, the extraction efficiencies of volatlles were much higher than those from the gasoline- contaminated samples, but they followed the same trends. A complete presentation of all of the data is presented in Appendix A. 8 ------- DRAFT C0701-1/UST 2/5/88 3.0 REGULATORY ANALYSIS Soils contaminated with petroleum products that exhibit a hazardous characteristic could be classified as hazardous wastes based on the "contalned-ln" policy defined by OSW.1 This study was conducted to determine whether soils contaminated with fuel products would be hazardous based on the 1gn1tab1l1ty characteristic or the proposed TCLP characteristic. In addition, fuels contain compounds that are listed hazardous wastes 1f discarded, RCRA Appendix VIII hazardous compounds, and hazardous substances subject to CERCLA reportable quantity requirements. The Impacts of these regulations on the management of soil contaminated with fuel products and the test results of this study are discussed below. 3.1 IGNITABILITY CHARACTERISTIC Based on the results of this study, soils contaminated with up to 10,000 ppm of gasoline, dlesel fuel, or No. 6 fuel oil would not be Ignltable. Higher concentrations were not tested, but 10,000 ppm 1s close to the saturation point. Thus, most soil contaminated by a fuel product release would not be hazardous based on the 1gn1tab1l1ty characteristic. 3.2 TOXICITY CHARACTERISTIC The results of this study Indicate that some soils contaminated with gasoline, would be considered hazardous waste because they fall the proposed TCLP test. At the same concentrations, however, soils contaminated with dlesel fuel and No. 6 fuel oil did not exhibit the toxicity characteristic. The acetate leaching medium used 1n the proposed TCLP test 1s representative of that found 1n a landfill. The actual leaching potential for the TC compounds 1n petroleum products could differ from the results of the TCLP test, because the actual leaching medium at a release location 1s likely to be rainwater. 3.3 IMPACTS OF THE PROPOSED TC REGULATION The experiments conducted 1n this study also Indicated that the concentrations of gasoline that would be hazardous varied with the type of soil tested. Thus, the volume of hazardous soil generated from a leaking UST depends on the national distribution of soil types as well as the number of releases and the average quantity of a release. At this time, only the number of releases can be estimated. Based on survey results 9 ------- DRAFT C0701-1/UST 2/5/88 presented 1n the proposed underground storage tank (UST) rules, 1t was estimated that releases could exist at about 10 percent of the 193,000 retail gasoline facilities nationally (52 FR 12662). If 1t 1s assumed that gasoline also 1s stored 1n 50 percent of the 651,000 petro- leum storage UST's 1n other Industry sectors and that 10 percent of these tanks are leaking, a total of about 33,000 UST systems containing gasoline may be leaking nationwide. Because the soil type distribution and the quantity of gasoline released per site are not known, 1t 1s difficult to estimate the quantity of hazardous soil nationwide. However, it 1s clear that the quantity of hazardous soil may be extremely large. The land ban on disposal of hazardous waste and the potential volume of hazardous soil make excavation and disposal in a Subtitle C landfill an unlikely soil management option. However, because benzene was the only gasoline constituent that made the soil hazardous, other treatment options (e.g., vacuum extraction and aeration) are available and could be used to eliminate the hazard by reducing the benzene concentration. Because the treated soil will no longer exhibit the toxic characteristic, 1t can be managed 1n any manner consistent with State regulations. Most States currently use subjective criteria (such as odor) to establish disposal options. These criteria may be more stringent than RCRA regulations. Thus, 1f the State criteria are retained, disposal of soil that 1s nonhazardous according to RCRA may still be restricted by State regulations. 3.4 IMPACT OF TOXICITY CHARACTERISTIC ON THE HAZARDOUS SUBSTANCE TANKS Toluene and benzene are hazardous substances on the 11st of proposed toxicity characteristic contaminants that are derived from petroleum and also are stored as pure products 1n UST's. Notification reports from New York, California, and Iowa and a report by the Chemical Manufacturers Association Indicated that the number of tanks storing toluene 1s greater than that for any other hazardous compound and 1s more than an order of 2 magnitude greater than the number storing benzene. The proposed rules for UST's Indicate that there are about 54,000 tanks storing hazardous substances nationwide (52 FR 12662). Although the total number of tanks storing toluene and benzene 1s not known, 1t 1s likely that several thousand contain toluene. The concentrations of pure toluene and benzene 10 ------- DRAFT C0701-1/UST 2/5/88 1n the soil that would result 1n a hazardous waste designation based on the proposed TCLP procedure could be estimated using the results of this study. Table 6 shows the toluene and benzene soil concentrations, TCLP concentrations, proposed TCLP hazardous concentrations, and projected hazardous soil concentrations for soils contaminated with gasoline. The ratios of leachate concentrations from gasoline-contaminated soils to the proposed hazardous TCLP concentrations were equated with the corresponding soil concentration ratios to estimate the hazardous soil concentrations. Gasoline tests were used because they contained more of both toluene and benzene than the other fuels. As shown 1n Table 6, soil contaminated with toluene concentrations of 600 to 3,200 ppm or with benzene concentrations of 4 to 89 ppm would be characterized as hazardous waste. The range 1s presented because the hazardous substance concentration 1s a function of the soil type. A number of other TC substances are stored 1n UST's so that releases from these tanks, depending on the volume of the release and the type of the soil, have the potential of contaminating significant volumes of soil that would then necessitate treatment and/or disposal as hazardous waste. Table 7 presents a 11st of these hazardous substances ranked by numbers of tanks. 3.5 APPENDIX VIII IMPACTS Fuel products contain several substances that are not on the proposed TCLP or CERCLA reportable quantity lists, but that are listed 1n Appendix VIII of RCRA (see Table 3). Wastes that contain these substances could be considered hazardous 1f 1t 1s determined that the substance poses a substantial threat to human health and the environment when Improperly treated, stored, transported, disposed, or otherwise 'managed' (40 CFR 261.10). The determination of what 1s hazardous 1s based on a considera- tion of many factors, one of which the potential for constituents of the waste to migrate Into the environment under plausible types of Improper management. Napthalene was the only Appendix VIII compound (other than benzene and toluene) contained 1n fuels that was detected 1n the TCLP leachates. In many cases, 1t was detected 1n leachates that did not exceed the proposed TCLP regulatory levels for benzene. Thus, an Implementing Agency could classify additional soils as hazardous based on the migration potential of napthalene. 11 ------- DRAFT C0701-1/UST 2/5/88 3.6 CERCLA REPORTABLE QUANTITY IMPACTS Petroleum product releases that would not result 1n hazardous soils may still be large enough to be subject to proposed CERCLA reportable quantity requirements. Table 8 shows the concentration of the volatHes and napthalene (the only semlvolatlle 1n petroleum on the CERCLA 11st) 1n gasoline, dlesel fuel, and Ho. 6 fuel oil. The reportable quantity of each constituent and the release volume of each fuel that would contain the reportable quantity of each constituent also are shown 1n Table 7. For gasoline, benzene 1s the constituent that would exceed the limit first; the reportable quantity would be contained 1n a little over 100 gallons. Reportable quantities of dlesel fuel and No. 6 fuel oil would not be exceeded until about 27,000 gallons and 125,000 gallons were released, respectively. These release volumes are based on the quantity of napthalene 1n these products because napthalene would exceed the reportable quantity first. 12 ------- DRAFT C0701-1/UST 2/5/88 5.0 REFERENCES 1. Memorandum from M. Williams, EPA:0SW, to P. Tobln, EPA:WHD, Region IV. November 13, 1986. RCRA regulatory status of contaminated ground water. 2. Camp Dresser and McKee, Inc. Draft Evaluation of Hazardous Substances Most Frequently Stored 1n Underground Tanks. Prepared for the Office of Underground Storage Tanks, U. S. Environmental Protection Agency. June 1987. 14 ------- DRAFT c0701-lal/UST 2/05/88 1 TABLE 1. SOIL CHARACTERIZATION RESULTS AND ACTION TOPSOIL FOR KAW RIVER SOIL Component concentrations, percent4 Action topsoll Kaw River soil Water, wt/wt 13.5 10.1 Sand 3.7 45 S1lt 55 51 Clay 41 4.0 aSand, silt, and clay concentrations reported on a dry basis. ------- DRAFT c0701-la3/UST 2/03/87 1 TABLE 2. CONCENTRATIONS OF THE SOIL/FUEL PRODUCT MIXTURES Saaple No. Soi 1 Samp 10 description Nominal fuel Fuel concentration type in soil, jig/g Amount of fuel added, U9 Total amount of fuel* soi1, g Final con- centration, yg/g 7035/36 Act 1on Diesel 10 1,960 200 9.8 7033/34 Action Diesel 100 20,610 200 103 7029/30 Action Diesel 1,000 200,000 200 1,000 7051/52 Action Oiesel 1,000 200,000 202 991 7027/28 Action D1ese1 5,000 1,000,000 200 5,000 7025/26 Action Diesel 10,000 2,010,000 200 10,000 7059/60 Action Fuel oi1 10 1,600 200 8.0 7023/24 Action Fuel oi1 100 20,400 213 96 7057/58 Action Fuel oi1 1,000 210,000 205 1,030 7055/56 Action Fuel oi1 5,000 1,070,000 201 5,330 7053/54 Action Fuel oi1 10,000 2,040,000 200 10,200 7011/12 Action Gas 10 2,040 200 10.2 7009/10 Action Gas 100 19,910 200 100 7005/6 Action Gas 1,000 200,000 200 1,000 7007/8 Action Gas 1,000 200,000 200 1,000 7003/4 Action Gas 5,000 1,020,000 200 5,100 7001/2 Action Gas 10,000 2,030,000 200 10,100 7045/46 Ksm Diesel 10 2,100 200 10.5 7037/38 Kaw Oiesel 100 20,120 200 101 7043/44 Kaw Diesel 1,000 190,000 200 950 7041/42 Kaw Diesel 5,000 980,000 200 4,890 7039/40 Kaw Diesel 10,000 2,010,000 200 10,100 7069/70 Kaw Fuel oi1 10 2,100 200 10.5 7021/22 Kaw Fuel oi1 100 20,600 200 103 7067/68 Kaw Fuel oi1 1,000 220,000 201 1,100 7065/66 Kaw Fuel oi1 5,000 1,060,000 200 5,300 7063/64 Kaw Fuel oil 10,000 2,040,000 200 10,200 7019/20 Kaw Gas 10 2,600 200 130 7049/50 Kaw Gas 100 20,600 200 103 7017/18 Kaw Gas 100 21,700 201 108 7031/32 Kaw Gas 1,000 230,000 200 1,150 7073/74 Kaw Gas 1,000 210,000 200 1,050 7015/16 Kaw Gas 5,000 980,000 200 4,900 7047/48 Kaw Gas 10,000 2,000,000 200 10,000 7013/14 Kaw Gas 10,000 2,010,000 200 10,000 ------- DRAFT c0701-la3/UST 2/03/87 2 TABLE 3. VOLATILE AND SEMIVOLATILE ORGANIC TARGET ANALYTE CONCENTRATIONS IN UNLEADED GASOLINE, DIESEL FUEL, AND NO. 6 FUEL OIL GasolIne, ug/g Diesel, yg/g No. 6 fuel oi yg/g Volatile organic compounds Carbon dlsulfIdea Benzene8 c Toluene8 c Ethyl benzene8 ni-Xy lene® o~ and p-Xylenee hdP 13,800 58,000 12,500 33,100 32,100 NO TR 359 312 669 638 NO 10.0 60.4 21.7 69.0 65.2 Quantitation limit 500 125 6.00 Sew IvoI atIle organic compounds Phenol 2-Methylphenola 3- and 4-Methylphenola Nitrobenzene8 PyrIdIne8 FIuoranthenec Benzo(a)pyrenec BenzoIb J~Ik)fIuoranthenec BenzoIa]anthracene0 Chrysenec 2,4-0IoethyIphenoIc Indenod ,2,3-cdlpyrene° Naphthalene0 01 benz(a,h1 anthracene0 W) NO 34.6 NO NO 2.36 W) 4.55 2.20 2.01 4.69 NO 198 NO NO 5.47 3.49 NO NO NO NO NO NO 168 8.42 NO 658 HD NO NO NO W NO W) 775 663 9,810 21,200 NO 237 142 NO Quant I tat i on Unit (base/neutraIs) Quantitation limit (phenols) 2.00 2.00 113 2.00 453 2.00 ?Analyte is listed in the proposed TCLP rules (51 FR 21648, June 13, 1986). 3 analyte not detected. °AnaIyte is IIsted In Append I * VIII. a trace; analyte was detected, but at a concentration below the quantitation limit. aAnalyte is a CERCIA hazardous substance subject to reportable quantity requirements. ------- DRAFT c0701-lal/UST 2/05/88 2 TABLE 4. IGNITABILITY TEST RESULTS FOR FUEL CONTAMINATED SOILS son Fuel type Fuel concentra- tion, ppm Flash point Sustained combustion Action Gasoline 10,100 None None Action Gasoline 5,100 None None Kaw Gasoline 10,000 None None Kaw Gasoline 4,900 None None Action Diesel 10,000 None None Kaw Diesel 10,100 None None Action Fuel oil 10,200 None None Kaw Fuel oil 10,200 None None ------- I T cU/01-la6/UST 12/23/87 2 TABLE 5. CONCENTRATIONS OF PROPOSED TCLP COMPOUNDS IN GASOLINE-CONTAMINATED SOIL SAMPLES AND LEACHATE Analytes Actual concentration of gasoline in sol 1, )g/g Theoret i ca1 concentration of analyte In contaminated soil, )g/ga Concentration of analyte In TCLP leachate, )g/L Extraction efticlency, percent recovery Proposed TCLP regu1atory 1 eve 1, ) g/L Kaw Benzene 103 1,050 4,900 10,000 1.42 14.5 67.6 138 31.6 241 1,660 2,380 44.5 33.3 49.3 34.5 70 Toluene 4,900 10,000 284 580 8,420 13,100 59.6 45.2 14,400 3- and 4-methylphenol 10,000 0.35 NDd 0 10,000 Action topsoi1 Benzene 5,100 10,100 70.4 140 TRC 109 0 1.56 70 Toluene 5,100 10,100 296 589 417 2,620 2.83 8.90 14,400 3- and 4-methylphenol 10,100 0.35 NO 0 10,000 Assumes that minimal losses from volatilization occurred. This may not be valid, especially for the volatile organic compounds. However, every attempt was made to minimize volatilization. ND = not detected. CTR = trace; analyte was detected, but the concentration was less than the 2.0 )g/L quantitation limit. ------- DRAFT c0701-la5/UST 1/4/88 2 TABLE 6. PROJECTED HAZARDOUS SOIL CONCENTRATIONS OF PURE TOLUENE AND BENZENE Noainal Toluene Benzene Soil type gasolIne concen- tration In the toil, ppa Actual TCLP concen- tration. ppb Hazardous TCLP concen- tration. ppb Actual concentration in soil, ppa Projected hazardous (oil concen- tration, ppa Actual TCLP concen- tration. ppb Hazardous TCLP concen- tration, ppb Actual concentration In soil, ppa Projected hazardous soil concen- tration, ppa Km River 1.000 1.480 14.400 60.9 S92 241 70 14.5 4.2 S.000 8.420 14.400 284 486 1.660 70 67.6 2.9 10.000 13.100 14,400 580 638 2,380 70 118 4.1 Action Topsoll 1.000 7.62 14.400 58 109,606 5.000 417 14.400 296 10.221 -- -- - 10.000 2.620 14,400 586 3,220 109 70 139 89 ------- DRAFT c0701-lal/UST 2/05/88 3 TABLE 7. HAZARDOUS SUBSTANCES STORED IN UST'S BY NUMBER OF TANKS® No. of Rank Compound tanksD 1 Toluene 450 2 Methyl ethyl ketone 261 3 Methylene chloride 112 4 1t1,1-Tr1ch1oroethane 86 5 T etrach1oroethy1ene 36 6 Trlchloroethylene 24 7 Benzene 20 8 Chlorobenzene 16 9 Acrylon1tr1le 10 10 Vinyl chloride 6 11 Pentachlorophenol 3 12 Carbon disulfide 2 13 Lead-containing water 1 14 Water containing mercury 1 15 Chromate rinse 1 ^Reference 2. ^Numbers represent results of surveys In California, Iowa, and New York, and some additional Information from the Chemical Manufacturer's Association. ------- FT w„/01-la4/UST 2/03/88 2 TABLE 8. CONCENTRATIONS AND REPORTABLE QUANTITIES OF HAZARDOUS SUBSTANCES FOUND IN PETROLEUM PRODUCTS Analyto Concentration In fuel, ppm Gasoline Diesel Fuel oil No. 6 Reportable quantity, lb Quantity of fuel containing reportable quantity of analyte. gallons8 Gasoline Diesel t-uei olI no. 6 Vol at Iles Benzene 13,800 Trace Toluene 56,000 3S9 Ethyl benzene 12,500 312 Xylene 65,200 1,307 10.0 60.4 21.7 134.2 10 1,000 1,000 1,000 129 13,070 14,250 2,730 500,000 570,000 140,000 180,000 2,900,000 8,200,000 1,300,000 Semlvolatlles Napthalene 198 658 142 100 89,950 27,000 Assumes that the densities of dlesel fuel and No. 6 fuel oil are 42 lb/ft , the same as that of gasoline. 125,000 ------- DRAFT C0701-1A/UST 2/3/88 APPENDIX A. TCLP ANALYSIS RESULTS AND QA/QC PROCEDURES A.l MATERIALS AND METHODS A.1.1 Soil Characterization Two different types of soils were used to determine the mobility of petroleum products mixed 1n soil. The first soil was a light-colored, fine-grained sandy soil obtained from Kaw Sand Company (Lawrence, Kansas). A second soil was obtained from Action Topsoll (Olathe, Kansas). The Action soil was dark and loamy with a higher apparent clay content than the Kaw soil. Both soils are commercially available. Each soil was characterized for its water content. Approximately 10 g of each soil was transferred to tared beakers and placed 1n an oven (100°C). The beakers were weighed at 2-h Intervals until two consecutive weighings gave the same result. A rough particle size characterization (percent sand, silt, and clay) was performed on each soil using a method based loosely on ASTM Standard Method D422. Ten grams of each soil were mixed with 15 mL of a 0.2-g/mL solution of Calgon detergent to promote clay dispersion. The soll/Calgon mix was diluted to 125 mL with delonlzed water. The mixture was shaken for approximately 1 h on a rotary shaker and wet-s1eved through a 230-mesh sieve to separate sand-size particles from s1It- and clay-sized particles. The liquid was drained through a large funnel Into a 1-L graduated cylinder. The sand-size particles remaining on the sieve were washed with delonlzed water, with the fine particle-conta1n1ng washings drained into the graduated cylinder to a maximum volume of 1 L. The sand- sized particles were oven-dried and weighed. The suspended particles 1n the graduated cylinder were mixed thoroughly using a perforated rubber stopper on a stainless steel rod. After 20 s, a volumetric plpet was Inserted to a depth of 20 cm and a 20-mL aliquot was collected and A-l ------- DRAFT C0701-1A/UST 2/3/88 dispensed Into a 50-mL tared beaker. The silt- and clay-sized particles thus collected were dried 1n an oven and weighed. The same procedure was repeated for clay-sized particles after 123 m1n, but at a depth of 10 on. The weights for each of the f1ner-than-sand sized particles were corrected for the Calgon and the total volume from which they were sampled, and the s11t content corrected for its clay content. A.1.2 Soil/Fuel Products Mixing Three different types of fuel products were mixed with the two soils described above. Unleaded gasoline and dlesel oil were obtained from a Fairway 011 gasoline station (Shawnee, Kansas) 1n June 1987. Fuel oil No. 6 was obtained from Consolidated Fuel 011 Co. (Riverside, Missouri) 1n May 1987. Each soil was spiked with each of the fuel products at five different nominal concentration levels. These concentrations were 10,000, 5,000, 1,000, 100, and 10 ppm. A small amount (10-20 g) of the soil was supple- mented with the appropriate amount of the fuel product (2, 20, 200, 1,000, and 2,000 mg) so that when the weight of the mix was made up to 200 g, the actual concentration would be close to the nominal concentration. The mixing was done 1n precleaned, 32-oz laboratory jars which were modified by a glassblower so that portions of the glass wall were pushed into glass fingers extending Into the containers' cavity. These glass fingers promoted more efficient mixing of the soil and the fuel when the sealed jars were rotated 1n a tumbling table (1 h). At the end of the mixing period, they were stored at -10°C. This procedure worked well for the 1ow-v1scos1ty liquid fuels (gasoline and dlesel), but the viscous fuel oil did not mix well and produced clumps. As an alternative, the soil and fuel oil were dispensed Into unmodified 32-oz laboratory jars and placed directly Into a freezer (-10°C) for several hours. At the end of this time, three ceramic balls (l-1n diameter) were placed 1n the jar, and the jar was rotated for 3 to 4 h. This procedure broke the solidified clumps of fuel oil and effectively mixed them into the soil. A.1.3 Determination of Characteristic of Iqn1tab111ty of Fuel- Contamlnated Soils The fuel-contaminated soils were allowed to come to room temperature, and 10 to 20 g was transferred to a clean watch glass. The flame of a A-2 ------- DRAFT C0701-1A/UST 2/3/88 low-flame Bunsen burner was allowed to contact the samples for 2 s and then removed. The sample was observed during and after flame contact for any flashing or sustained combustion of the samples. If no flashing or sustained combustion occurred, the above procedure was repeated three times. This procedure 1s based on the guidance provided 1n SW 846, Vol. 1C, Chapter 7, Section 7.1.2.2. A.1.4 Toxic Characteristic Leaching Procedure (TCLP) Samples were subjected to the TCLP protocol according to the procedures described In the Federal Register (Volume 51, No. 216, November 7, 1986). Acetate was used as the leaching fluid. The TCLP for volatile organlcs was performed using a 90-mm zero headspace extractor (ZHE). Two of the devices were purchased from MllUpore Corporation, and four other similar devices were constructed by a local machine shop. The samples were added to the ZHE cold and allowed to warm up inside the device (with zero headspace) before commencing filtration. The fuel-contaminated soils were previously determined to have no free liquid, and thus no filtrates were collected. Leaching solution was pumped Into the ZHE from a graduated cylinder using a piston pump (Cole Parmer R-7115-60). After the 18-h extraction period, the leachate was collected 1n glass, gastlght syringes (50 mL); the first portion collected was discarded, and the second portion was dispensed Into vials with no headspace. These 1eachate-conta1n1ng vials were stored 1n charcoal-filled paint cans. The V1ton 0-r1ngs on the ZHE were cleaned with methanol and oven-dried. A portion of the fuel-contaminated soils was Introduced Into flint- glass 1-gal containers which had been previously treated 1n a 50 percent HN03 bath (12 h minimum). The appropriate volume of leaching fluid was then added. Leachates were allowed to settle for 1 h after the TCLP extraction and then filtered using the 142-mm high-pressure filtration device. A.1.5 Analysis A.1.5.1 Volatile Organic Analysis. Volatile organic compounds 1n the samples were analyzed according to the procedures 1n Methods 5030 and 8240 from SW-846. Nonaqueous samples (soil blanks and fuels) were dispersed 1n tetraglyme prior to analysis. The samples were analyzed by purge-trap- A-3 ------- DRAFT C0701-1A/UST 2/3/88 desorb 6C/MS using a Flnnigan/MAT CH4 magnetic sector mass spectrometer controlled by an INCOS data system. The following minor method deviations were necessary to analyze these samples: 1. D„-l,2-D1bromoethane was used as the second Internal standard rather than the recommended l,4-d1fluorobenzene. Du-1,2-D1bromoethane 1s stable, available 1n pure form, demonstrates good purging efficiency, yields fragmentation patterns with Intense 1ons at m/z 111 and 113, and elutes near the middle of the chromatographic run. 2. O10-Ethylbenzene was used as a surrogate 1n addition to the suggested D„-l,2-d1chloroethane, D8-toluene, and BFB (bromofluorobenzene). 3. The response factor for the SPCC bromoform was calculated using the Internal standard 0^-1,2-dlbromoethane. Method 8240 uses l,4-d1- fluorobenzene as the corresponding Internal standard. The minimum RF criterion for bromoform (RF >0.250) was achieved during analysis. 4. The matrix spiking solution consisted of all of the volatile target compounds. 5. The GC/MS operating conditions were changed so that the final column temperature was 225°C rather than 220°C. This allowed the xylenes to elute with better chromatography. 6. The samples were not screened by GC/FIO. Rather, the samples were reanalyzed by PTD-GC/MS 1f necessary using an appropriate dilution. 7. Mass spectra of target compound "hits" from the INCOS target compound analysis (TCA) routine results were visually checked and the Identification of the compound confirmed. For a compound to be a target compound candidate In the TCA routine, 1t had to elute at the correct retention time, and have characteristic 1on response. Mass lists for the 1ons with greater than 10 percent relative abundance were not printed or checked; 1t would not have provided any better compound Identification nor would 1t have been an effective use of time. A.1.5.2 SemlvolatHe Organic Analyses. Sem1volat1le organlcs were solvent-extracted from the fuels, soils, and TCLP extracts using SW-846-3580 (Waste Dilution, for fuels), SW-846-3520 (Continuous Liquid/ liquid Extraction, for leachates), or SW-846-3540 (Soxhlet Extraction, for solid samples). Prior to subjecting the samples to solvent extraction, A-4 ------- DRAFT C0701-1A/UST 2/3/88 they were spiked with surrogate standards. All of the fuel sample extracts were cleaned up using SW-846-3650 (Ac1d/Base-Neutral Cleanup). The base neutral extract of each sample was fractionated using SW-846-3611; the resulting aromatic fraction was combined with the addle extract. The sample extracts were concentrated using Kuderna-Danlsh (KD) glassware. Each sample extract and dilutions thereof were screened using GC/FID to determine the optimal concentration for GC/MS analysis. The GC/MS analyses were performed on a F1nn1gan/MAT 5100 quadrupole mass spectrometer controlled by an INCOS data system. The analyses were as per SW-846-8270 with the following exceptions: 1. The fused silica capillary column used had a 0.25-ym film thickness rather than the specified 1.0-ym thickness. The latter column may cause excessive "bleeding" of the stationary phase at the temperatures required to elute polycycllc aromatic hydrocarbons (PAHs), adversely affecting chromatographic quality. 2. The final temperature of the GC temperature program was 300°C rather than the specified 270°C. This temperature was held until benzo- Ig,h,1]perylene eluted. No problem was noted 1n the separation of d1 benz[a,h]anthracene, 1ndeno[l,2,3,cd]pyrene, and benzo[g,h,1jperylene. 3. Mass spectra of target compound "hits" from the INCOS target compound analysis (TCA) routine results were visually checked and the Identification of the compound confirmed. For a compound to be a target compound candidate 1n the TCA routine, 1t had to elute at the correct retention time, and have characteristic 1on response. Mass lists for the 1ons with greater than 10 percent relative abundance were not printed or checked; 1t would not have provided any better compound Identification nor would It have been an effective use of time. 4. Two sets of components were poorly resolved under the chromatographic conditions employed for the analysis of these samples and are reported as a combined value. These analytes were m- and p-cresol (3- and 4-methylphenol), and benzo[b]- and benzo[k]f1uoranthene. Since these analytes were not found 1n the samples, this Impacts only on calibration results. A-5 ------- DRAFT C0701-1A/UST 2/3/88 A.2 RESULTS AND DISCUSSION A.2.1 Soil Characterization Table A-l presents the results of the soil characterization studies performed on the two soils used during this study. The data indicate that Kaw River soil was characterized by having a very low clay content, with sand and silt 1n approximately equal proportions. The Action topsoll had a very low sand content, with the bulk of the material split approximately equally between silt- and clay-sized particles. Both soils had about 13.5 percent (wt/wt) water content; however, the Kaw River soil had clumping problems during mixing with the fuels, and 1t was dried 1n an oven to a 10 percent (wt/wt) water content, which effectively avoided the clumping problem. A.2.2 Soil/Fuel Product Mixing Results The actual and nominal concentrations of the soil/fuel product mixes are presented 1n Table A-2. The actual values were used to estimate the pre-TCLP concentration of each analyte 1n the soil, and to perform subsequent comparisons with the TCLP extract concentrations. A.2.3 Characteristic of Iqn1tab1l1ty Results The results for the 1gn1tab1l1ty test are presented 1n Table A-3; none of the fuel product-contaminated soils were 1gn1table. A limited number of samples were used for this test: the two most concentrated gasol1ne/so1l mixtures, and the most concentrated dlesel and fuel o1l/so1l mixtures for both soils. If any of the above had been determined to be 1gn1tab1e, additional samples would have been analyzed. A.2.4 Analytical Results A.2.4.1 Quantitation Limits. The reliable detection limits- hereafter referred to as quantitation limits—for volatile organic compounds not detected 1n the samples were estimated from the sample preparation procedures and from the Instrumental detection 11m1t. For aqueous samples, the quantitation limit of a compound (10 ng on-column from a 5.0-mL sample aliquot) 1s: . 2 ng/mL - 2 Ug/L A-6 ------- DRAFT C0701-1A/UST 2/3/88 For tetraglyme extracts, 1f a 10.0-g sample Is dispersed 1n tetraglyme to a final volume of 25 mL (25,000 uL), and a 20-yL portion 1s taken for purge-trap-desorb (PTD) analysis, the quantitation limit for a compound (10 ng on-column) would be: (10 ng/20QHL)x 25tOOO HL a lf250 ng/g = L25 vg/g The quantitation limits of target semlvolatlle compounds not detected 1n the samples were estimated from the sample preparation procedures and the Instrumental detection limit. For example, suppose that 10 g of sample were extracted and the extract concentrated to 5 mL. The smallest amount of any target analyte that could be accurately quantified using the average response factor from the calibration curve would need to have a concentration no lower than the lowest calibration standard 1n the curve (1.0 ug/mL). The estimated sample quantitation limit was calculated as follows: I"? ;^pi° extracted* = " ^ These values were corrected for any required dilutions that occurred during sample preparation as determined by screening with gas chromatography/flame Ionization detection (GC/FID). A.2.4.2 Results of Fuels Analysis. The data obtained from the analysis of the three fuels (unleaded gasoline, dlesel, No. 6 fuel oil) for their volatile and semlvolatlle organic composition are presented 1n Table A-4. The unleaded gasoline was, as expected, very rich 1n volatile aromatic hydrocarbons. The most abundant was toluene, followed by m-xylene, o- and p-xylene, benzene and ethylbenzene. All of the concentrations were above 10,000 ug/g, which was significantly higher than the quantitation limit for the analysis of this fuel. Carbon disulfide was not detected 1n gasoline. In contrast to very high levels of volatile organic compounds 1n gasoline, relatively low levels of semlvolatlles were observed 1n this fuel. Only two target analytes were detected at levels significantly higher than the quantitation 11mlt—naphthalene (198 yg/g) and 2-methylnaphthalene (259 ug/g). A-7 ------- DRAFT C0701-1A/UST 2/3/88 Diesel fuel had lower levels of volatile aromatic compounds than the gasoline; m-xylene and o- and p-xylene were the most abundant volatile aromatlcs at approximately 650 ug/g. Toluene and ethylbenzene were detected at between 310 and 360 ug/g. and benzene was detected, but at a concentration lower than the quantitation limit. The semlvolatHe aromatic compounds were more abundant 1n dlesel fuel than 1n gasoline, especially two- and three-ring aromatlcs (Mphenyl, naphthalene, 2-methyl- naphthalene, fluorene, and phenanthrene), and the most concentrated were 2-methylnaphthalene (2,800 ug/g), phenanthrene (1,250 ug/g) and naphthalene (658 ug/g). No. 6 fuel oil had lower concentrations of volatile aromatic hydrocarbons than gasoline and dlesel fuel. The most concentrated analytes were m-xylene (69 ug/g). o- and p-xylene (65 ug/g). and toluene (60 ug/g). Ethylbenzene and benzene were detected at less than 25 ug/g. The most abundant semlvolatlle organlcs were four-ring aromatlcs (benzola]anthracene, chrysene, pyrene), with significantly lower concentrations of naphthalene and 2-methylnaphthalene. The latter two are expected to Impact TCLP extracts more than the more abundant large aromatlcs due to significant differences 1n their aqueous solubilities. For example, methyl naphthalene has an approximate solubility (1n water) of 25 mg/L compared to pyrene, which has an approximate solubility of 0.15 mg/L.1"3 The theoretical concentrations for each target analyte 1n each fuel- contaminated soil are presented 1n Table A-5 for Action topsoll and 1n Table A-6 for Kaw River soil. These values were obtained by multiplying the concentration of every analyte found 1n each fuel times the actual concentration of that fuel In each soil. For example, 1f the concentration of benzene 1n unleaded gasoline was 13,800 ug/g. and 1f the concentration of gasoline 1n Kaw River soil (nominal concentration of 100 ppra) was 0.103 g/kg, then the theoretical concentration of benzene 1n this contaminated soil would be: 13,800 ug/gx0.103 g/kg = 1,421 ug/kg A-8 ------- DRAFT C0701-1A/UST 2/3/88 Only those analytes that were detected 1n all the fuels above the quantitation limit are reported 1n this table. Appendix VIII compounds and those compounds Identified as toxicity characteristic contaminants 1n Table C-l of the TCLP proposed rules (51 CFR 21648, June 13, 1986) are flagged as such 1n the table. In the absence of data validating these theoretical concentrations, 1t will be assumed that minimal losses of all analytes occurred. This may not be a valid assumption, especially for the volatile organic compounds. Every possible measure was taken to minimize any such possible losses. A.2.4.3 TCLP Extracts. A.2.4.3.1 Action topsoll. The results of the analyses of the TCLP leachates obtained from fuel-contaminated Action topsoll samples are presented 1n Table A-7. For convenience, only those analytes that were consistently detected 1n more than one TCLP extract of the same soil/fuel combination are reported 1n Table A-7. A complete table with all of the target analytes 1s Included 1n Appendix A. For some of the fuel/soil mixtures, not all of the leachates produced were analyzed. Whether to analyze for semivolatile organlcs was determined after screening the extracts by GC/FID. In the case of volatile organlcs, the rapid turnaround of GC/MS reports permitted real-time evaluation of the trends, and whether or not 1t was necessary to analyze the next lower concentration leachate. The Action topsoll/dlesel leachates indicated that, for volatile aromatlcs, the soil with the highest fuel concentration produced the most concentrated leachates. Thus, m-xylene and o- and p-xylene showed their highest concentrations 1n the leachate resulting from the 10,000-ppm fuel/soil mixture. There was at least a two-fold decrease 1n concen- tration to the 5,000-ppm leachate, and at least a ten-fold decrease 1n concentration from the 5,000- to the 1,000-ppm leachate. These differ- ences closely parallel the differences between the levels of fuel contamination 1n each soil and are Interpreted as being significant. This pattern was not observed for the semivolatile aromatic compounds studied. The difference between the 10,000-ppm and the 5,000-ppm leachates for all of the reported semivolatile organlcs was not significant. However, the difference between most of the semlvolatile A-9 ------- DRAFT C0701-1A/UST 2/3/88 analytes was significant between the 5,000- and the 1,000-ppm leachates. Only one analyte was detected 1n the 100-ppm leachate, and 1t too shows a significant decrease 1n concentration from the next most concentrated leachate. These data can be visualized better by plotting leachate concentrations against the nominal fuel/soil concentrations. The volatile organlcs from the so1l/d1esel mixture (Figure A-l) suggest that the leaching fluid will extract increasing amounts of these compounds 1f they are used to test more highly contaminated soils than used for this study. On the other hand, the semlvolatlles from this fuel/soil mix (Figures A-2 and A-3) Indicate that these leaching solutions are approaching (or have approached) a saturation point 1n their capacity to extract higher levels of these compounds from Increasingly more concentrated fuel/soil mixtures. For example, an Increase 1n the fuel/so11 concentration above the 10,000«ppm level would probably result 1n Increased leachate concentrations for toluene (Figure A-l), but an Increase above 5,000 ppm would not produce an Increased leachate concentration for 2-methylnaphthalene (Figure A-3). The results for the Action topsoll/fuel oil No. 6 combination do not provide much trend information because only two leachates were analyzed based on observations made during GC screening of the extracts. Some of the same trends discussed above for semlvolatlle aromatlcs were seen, with the 5,000- and 10,000-ppm leachates exhibiting nonsignificant differ- ences. Most of the volatlles exhibited the same patterns discussed observed for the Action topsoll/dlesel combination. However, benzene and toluene lend some confusion to the pattern. The patterns that were observed for the dlesel/Actlon topsoll mixture are apparent for the volatlles 1n the Action topsoll/unleaded gasoline leachates (Figure A-4). There was at least a two-fold difference 1n concentrations between the 10,000- and the 5,000-ppm leachates, and no less than a 45-fold difference between the 5,000- and the 1,000-ppm leachates. The 100-ppm leachates had no detectable analytes. Only two semlvolatlle aromatlcs were present 1n the leachates: naphthalene and 2-methylnaphthalene (Figure A-5). There was a less than two-fold differ- ence 1n concentration between the two most concentrated leachates, but at least a four-fold difference between the 5,000- and 1,000-ppm leachates. A-10 ------- DRAFT C0701-1A/UST 2/3/88 These patterns are not unexpected given the differences 1n solubility between various aromatic hydrocarbons. For example, the solubility of toluene (538 mg/L) 1s significantly higher than that of naphthalene (31.5 mg/L), which 1n turn 1s significantly higher than that of blphenyl (5.9 mg/L).1'2 Higher aromatic hydrocarbons (I.e., three or more rings) have solubilities of less than 1 mg/L.3 A.2.4.3.2 Kaw River soil. The results of the analyses of TCLP leachates obtained from fuel-contaminated Kaw River soils are presented in Table A-8. Only those analytes that were consistently detected 1n more than one leachate of the same soil/fuel combination are presented. A complete table with all of the target analytes detected 1n all of the leachates 1s presented 1n Appendix B. For some of the fuel/soil mixtures, not all of the leachates produced were analyzed. For semlvolatlle organlcs, whether to analyze the extracts was determined after screening the extracts by GC/FID. In the case of volatile organlcs, the rapid turnaround of GC/MS reports permitted real-time evaluation of the trends, and whether or not 1t was necessary to analyze the next lower concentration leachate. The Kaw River so1l/d1esel leachates exhibited similar patterns to those previously discussed for the Action topsoll (Figures A-6, A-7, and A-8). There were discrepancies between the trends observed for ethylbenzene and the xylenes at most concentrations and those for benzene and toluene. First, the 1,000 ppm leachate results for toluene and, most noticeably, for benzene appear to break the pattern of the other volatile analytes. Second, the 10 ppm leachate concentrations are totally out of line with the established patterns, and may be attributed to analytical problems (e.g., contamination) with this leachate. The remaining values for benzene and toluene, and all values for the other volatlles support the observations made for Action topsoll. Only three leachates were analyzed for the Kaw River soil/fuel oil No. 6 mixture. To a large extent, these leachates support the patterns observed for other soil/fuel mixtures. All of the leachates analyzed for the Kaw River so1l/gasol1ne mixtures (Figures A-9 and A-10) also support the arguments presented above. A—11 ------- DRAFT C0701-1A/UST 2/3/88 The Action topsoll and Kaw River soil leachates exhibit a potentially Important difference. The slope of the leachate/soll plots decreased for Kaw River leachates at lower nominal soil concentrations than for the Action topsoll leachates. Even though the numerical difference between the Kaw and Action soil leachates at most nominal concentrations was small, the trend was consistent for most of the analytes studied. This suggests that the sandy Kaw River soil allows leaching of contaminants more readily than the Action topsoll. A.2.4.3.3 Regulatory levels and leachate concentrations of contaminated soils. Table C-l of the TCLP proposed rules (51 CFR 21648, June 13, 1986) Indicates that, of all of the compounds targeted and detected 1n the leachates during this study, only benzene and toluene affect the classification of contaminated soils as hazardous waste. The regulatory levels are 70 yg/L for benzene and 14,400 yg/L for toluene. These levels were exceeded 1n the 10,000-ppm Action topsoll/gasoline leachate (Table A-7) and 1n the 1,000-, 5,000-, and 10,000-ppm Kaw river soil/gasoline leachates. A.2.4.3.4 Efficiency of TCLP extraction. TCLP 1s an extraction procedure, and 1f the Initial waste concentrations for each analyte are known, the percent recovery of each analyte can be calculated. This recovery can be a useful Indicator of the efficiency of TCLP 1n mobilizing the various contaminants from a waste. If the calculated theoretical concentration of m-xylene 1n the Action topsoll/dlesel fuel 5,000-ppm mix- ture was 3,345 yg/kg (this calculation was described 1n Section 3.4.2), and 1f 9.95 g of this soil/fuel mixture was used for TCLP, then the Initial amount of m-xylene 1n the system 1s given by: 3,345 yg/kgxO.00995 kg = 33.3 yg of m-xylene Similarly, 1f 200 mL of leachate had a concentration of 47.6 yg/L, the total amount of m-xylene extracted by TCLP would be given by: 47.6 yg/Lx0.200 L = 9.52 yg of m-xylene 1n total leachate A-12 ------- DRAFT C0701-1A/UST 2/3/88 The leaching efficiency of m-xylene from this particular system 1s calculated as follows: 9.52 ug of m-xylene 1n leachate = 28 g Dercent recoverv 33.3 pg of m-xylene 1n waste Percent recovery These values also Incorporate the efficiency and accuracy of the sample preparation and analysis. Table A-9 shows the TCLP leaching efficiency for the Action topsoll/fuel mixtures. Some analytes were not expected to be in the contaminated soil because they were not detected 1n the neat fuel. In these cases, the recovery calculation would have required a division by zero, which 1s not processable (HP). If the analyte was expected to be 1n the contaminated soil, but was not found 1n the leachates, then the recovery would have been zero. The results Indicate that, for volatile aromatlcs, the leaching (extraction) efficiency increased with Increasing Initial soil concentration. The most efficient extractions were those involving No. 6 fuel oil-contaminated soils, possibly because the Initially lower analyte concentrations 1n the soil did not allow saturation of the leaching fluid. The least efficient TCLP extractions were those which Involved gasoline contaminated soils, probably because of the high Initial concentrations of the analytes which may have permitted relatively rapid saturation of the leaching fluid. For semivolatHe aromatlcs, the maximum extraction efficiency was seen at lower levels of fuel/soil concentra- tions. For example, the dlesel contaminated Action topsoll leachates had maximum extraction efficiency 1n the 1,000 ppm soil. Some recovery values 1n Table A-9 appear to be unusually high. These will be discussed below. Table A-10 presents the recovery results for the fuel-contaminated Kaw River soils. The volatile aronatlcs 1n the dlesel contaminated soil show that the maximum extraction efficiency occurred 1n the 1,000 ppm leachates, 1n which nearly all of the material in the soil was extracted. With the exception of the 1,000 ppm No. 6 fuel oil leachate, the No. 6 fuel oil contaminated Kaw soil showed almost complete extraction of the volatile analytes. The gasoline contaminated soils showed that the 10 ppm and the 100 ppm leachates were the most efficient in extracting A-13 ------- DRAFT C0701-1A/UST 2/3/88 volatile analytes from the soil. For semlvolatlles, the highest recoveries were seen 1n the 1,000 ppm leachates for dlesel, and 1n the 1,000 ppm leachate for No. 6 fuel oil. Some of the analytes presented 1n Table A-9 and A-10 exhibited extremely high recoveries, Indicating analytical problems (e.g., sample contamination, matrix effects); these values are flagged 1n the tables. For some of the vol atlies, these problems were probably related to the analysis of the leachates, because the unusually high recoveries affect only some of the analytes or some samples of a soil/fuel set. In the case of gasoline/soil semlvolatlles and fuel o1l/so1l volatlles, 1n which high recoveries were observed for all of the analytes for both of the soil types, 1t 1s possible that the results of the analysis of the neat fuel may be underestimated. These observations are supported by QC results (see section 4.0) for the analysis of the neat fuels. Surrogate recoveries for volatlles analysis of No. 6 fuel oil were low, as were surrogate recoveries for semlvolatlles analysis of gasoline. These observations may clarify some of the apparent outliers discussed previously for leachate concentrations. Benzene and toluene concentrations for both nominal concentrations of Action topso1l/No. 6 fuel oil (Table A-7) were probably Inaccurate. Similarly, the Kaw River so11/No. 6 fuel oil 1,000-ppm leachate and benzene 1n the 1,000-ppm Kaw River so1l/d1esel leachate may be Inaccurate (Table A-8). The leaching (extraction) efficiencies discussed above may highlight the differences mentioned above 1n the ability of contaminants to be leached from the different types of soil studied. In general, extraction efficiencies were higher at lower nominal concentrations 1n Kaw River soil than 1n Action topsoll. For example, 1n the Action so1l/d1esel leachates, a maximum of 30 percent of the volatile contaminants were leached from 10,000 ppm soil, whereas the corresponding Kaw River soil leachates showed approximately 100 percent recovery from the 1,000 ppm leachates. This trend was observed 1n most of the other leachates. A.3 QUALITY ASSURANCE/QUALITY CONTROL This section reports the QA/QC procedures followed during this study and the QA/QC results obtained during the course of sample preparation and analysis. The QA/QC data assist 1n establishing the quality and validity A-14 ------- DRAFT C0701-1A/UST 2/3/88 of the sample analysis data. The data quality objectives (DQOs) for this study were as follows: 1. Analytical precision—30 percent relative percent difference (RPD) for duplicate analyses. 2. Analytical accuracy«70 to 130 percent of true value for QA check samples and 70 to 130 percent recovery for surrogates and matrix spikes. The QA activities performed during this study Include: 1. Surrogate recoveries—used to evaluate analytical accuracy and extraction efficiency. 2. Matrix spikes—to establish accuracy of the analysis procedures as well as matrix effects. 3. Duplicates and matrix spike duplicates—to establish analytical precision. 4. Method blanks—to evaluate potential contamination during sample preparation and analysis. 5. QA check samples—a blind sample prepared independently of the analysis team to check analytical Instrumentation and data Interpretation. Specific QA/QC procedures, deviations, and analysis results are described 1n the following sections. A.3.1 Volatile Organic QA/QC A.3.1.1 GC/MS Considerations. The GC/MS performance criteria for the analysis of volatlles specified 1n SW-846-8240 were used. The 1on abundance ratios of BFB (bromofluorobenzene) were compared on a dally basis with the 1on abundance criteria from Method 8240, and the Instrument was tuned 1f necessary. The BFB had to be analyzed by direct Injection rather than by purge-trap-desorb because the BFB coeluted with a silicon peak apparently originating from the 3 percent 0V-1 packing 1n the analytical adsorbent trap. When the peaks coeluted, the Interference caused the criteria not to be met. The analyses were performed on a F1nn1gan MAT CH4 magnetic sector GC/MS. A five-point calibration curve was constructed using mixed calibration standard solutions at 2, 10, 40, 70, and 115 yg/mL. Calibration standards were prepared from EPA repository chemicals and from the highest purity commercially available chemicals. All standards were assigned a unique number for Identification and tracking purposes. The A-15 ------- DRAFT C0701-1A/UST 2/3/88 calibration standard mixes were prepared In methanol according to SW-846-8240. Five microliters of these solutions were added to 5.0 mL of reagent water. The average response factors (RFs) obtained for the calibration check compounds (CCCs), target analytes, and surrogate recovery compounds deviated by less than 30 percent (relative standard deviation) of the average RF. The RFs for the system performance check compounds (SPCCs) exceeded the minimum RF. A mid-level (40 ug/mL) standard was analyzed before and after analyzing samples each day, and at the end of every 12-h period. This standard contained the target analytes, surrogates standards, BFB, SPCCs, and CCCs. The BFB 1on abundance ratios were compared on a dally basis with the 1on abundance criteria from Method 8240, and the Instrument was tuned 1f necessary. The SPCCs were checked to assure that the minimum RF had been met. The response factors of the CCCs were examined to determine 1f they were within 25 percent of the average RF per Method 8240. All criteria specified 1n Method 8240 were met on every day of analysis. The extracted 1on current profile (EICP) areas for the Internal standards were monitored as the samples were analyzed. Internal standard area counts were consistent (-50 to +100 percent) between standard checks. A.3.1.2 Quality Control Volatile Analysis Results. A.3.1.2.1 Surrogate recoveries. Volatile organic surrogate recoveries for all the samples analyzed are presented 1n Table A-ll. More than 95 percent of the surrogate data points met recovery DQOs (70 to 130 percent). However, the five surrogates that did not meet DQOs were 1n two samples. A.3.1.2.2 Duplicate precision and matrix spike recovery. All of the 1,000 ppm leachates were analyzed for volatile organlcs 1n duplicate and as a matrix spike. The results are presented 1n Tables A-12 through A-16. Each one of these leachates was part of a separate batch of samples extracted by TCLP. More than 95 percent of the analytes met precision DQOs established for this study (30 percent RPD), and all of the analytes met accuracy DQOs (70 to 130 percent recovery). A.3.1.2.3 Method blanks. Ho volatile organic target compounds were detected 1n the Instrument (system) blanks or 1n the method blanks. The A-16 ------- DRAFT C0701-1A/UST 2/3/88 system blanks consisted of organic-free purge water spiked with Internal standards. A.3.1.2.4 MRI QA check sample. The results of the analysis of an MRI QA check sample, typically used to check the analytical instrumentation and the analyst's data Interpretation are summarized 1n Table A-17. All of the analytes were within DQOs for check samples (70 to 130 percent of true values). A.3.2 SemlvolatHe QA/QC A.3.2.1 GC/MS Considerations. GC/MS performance criteria for the analysis of semlvolatlles established in SW-846-8370 were used. The Instrument was tuned on a dally basis with a solution of DFTPP (50 yg/mL) per the Key Ion and Abundance Criteria set forth 1n Method 8270. A mixture of the SPCCs and the tuning standards (4,4'-DDT, benzidine, pentachlorophenol) was analyzed dally to monitor the RFs of the SPCCs (minimum RF = 0.05), to monitor GC column performance (benzidine and pentachlorophenol), and to verify Injection port Inertness (4,4'-0DT). A s1x-po1nt calibration curve was constructed using mixed calibration standard solutions which contained the CCCs, target analyses not Included 1n the CCCs, and the surrogate standards; the concentrations ranged from 1.0 to 150 yg/mL. The curve was accepted 1f the percent relative standard deviation (percent RSD) did not deviate by more than 30 percent from the average RF of each Individual CCC. A mid-level standard (50 yg/mL) was run every day of operation, and the RFs of the CCCs checked to determine 1f they were within 30 percent of the average RF per Method 8270. If such was not the case, the GC/MS system was checked, and 1f necessary, a new calibration curve was constructed. No more than 12 h of continuous GC/MS analyses were performed without recallbratlon of the Instrument (DFTPP, SPCC, and dally standard). Before Injection, the GC/MS analyst spiked 1.0 mL of each sample with 10 yL of a 4,000-yg/mL mixed Internal standard solution as specified 1n Method 8270. Specific deviations from this present protocol were occasionally necessary as described below: 1. The most concentrated calibration standard (150 yg/mL) of the s1x-po1nt curve for 2-fluorophenol, D5-pheno1, 2-methylphenol, 3- and 4-methylphenol, anthracene, l,4-d1chlorobenzene (CCC), phenol and A-17 ------- DRAFT C0701-1A/UST 2/3/88 2-fluorob1phenyl slightly saturated the SC/MS detector,, resulting 1n lower relative response factors than for the other standard?. If the high point had been removed from the s1x-po1nt curves, tixe average RFs for the resulting f1ve-po1nt curves (1.0 to 50 pg/mL) would fcave decreased by less than 5 percent. However, 1f any of these analytes had appeared 1n the sables above 50 ug/mL, the concentration would have been outside the linear calibration range of the Instrument. It was therefore deemed best to accept the lower relative response factors; the reported sample concentrations may be overestimated by no more than 5 percent. 2. On a number of occasions, the response factor for pyridine 1n the dally standard was more than 30 percent of the average response factor of the curve. This 1s not expected to impact the data because pyridine was not detected 1n any sample. 3. The minimum response factor criteria for 2,4-d1n1trophenol (SPCC) were not met on 3 out of 9 GC/MS analysis days. On all occasions that 2,4-din1trophenol failed to meet criteria, the RF was more than 85 percent of the minimum RF. 4. A number of calibration check compounds cause difficulty achieving consistent response factor data. During the analysis of these sables, two phenolic and three nonphenollc components of the CCC mix were found to be extremely difficult to bring into compliance with the established criteria. These problem compounds were: All the target analytes were checked on a dally basis, and with the exception noted above, they were always within 30 percent of the average RF. A.3.2.2 Surrogate Recoveries. Sem1volat1le organic surrogate recoveries are summarized 1n Table A-18. A large number of the surrogate recoveries were outside project DQOs (70 to 130 percent recovery). CCC Failure frequency (percent of analysis days) Pentachlorophenol 2-N1trophenol D1phenylam1ne D1-n-octylphthalate Hexach1orobutad1ene 11 11 56 78 11 A-18 ------- DRAFT C0701-1A/UST 2/3/88 However, a number of factors need to be considered when evaluating these data. First, some of the base/neutral surrogates employed (and required by SW-846) are poor choices for surrogate standards, especially when alumina cleanup of the extracts 1s required. For example, both D5-n1trobenzene and 2-fluorob1phenyl have, because of their functional groups, a larger charge associated to their molecules. When passing through the fractionation column necessary to separate aliphatic Interferences from the aromatic compounds, these two surrogates Interact more strongly with the alumina, and may be more strongly retained than the less charged PAHs. This results 1n low and variable recoveries of these compounds. These two compounds are not structurally related to PAHs and do not behave like them. Second, project DQOs for semlvolatlle surrogate recoveries were extremely tight; a more reasonable range would have been between 50 to 130 percent. The phenolic surrogates (2-fluorophenol, D5-phenol, and 2,4,6-trlbromophenol) showed that 61, 64, and 27 percent of the data points, respectively, did not meet surrogate recovery DQOs (Table A-18). The average recovery was between 58 to 81 percent, and only 2-fluorophenol had an unacceptable spread 1n the surrogate recoveries (-50 percent RSD). If a wider range of 50 to 130 percent recovery had been used for DQOs, only 9 to 20 percent of the data points for surrogate recoveries would have been outside DQOs. This suggests that a large proportion of the surrogates that missed DQOs were between 50 and 70 percent recovery. The base/neutral surrogates (D5-n1trobenzene, 2-fluorob1phenyl, Dl0-pyrene, and Dllt-p-terphenyl) present a different case because of the additional fractionation step to which the base/neutral extracts were sub- jected. As was discussed above, 05-n1trobenzene and 2-fluorob1phenyl are not truly amenable to fractionation, and this 1s reflected 1n the surrogate recoveries. Project DQOs for surrogate recoveries were missed for more than 80 percent of the surrogate data points for D5-n1trobenzene and 2-fluorob1phenyl. The average recoveries were low and the spread of the data was wide, especially for Ds-n1trobenzene (as Indicated by percent RSD). The use of a wider surrogate recovery range (50 to 130 percent) for DQOs would not have Improved the number of data points meeting DQOs significantly. The other base/neutral surrogates employed (D10-pyrene and A-19 ------- DRAFT C0701-1A/UST 2/3/88 Dllf-p-terphenyl) had acceptable average recoveries, although the spread for Dllf-p-terphenyl was wide. Approximately 23 to 32 percent of the data points for these compounds did not meet project DQOs. However, an increase of the DQO range to between 50 to 130 percent recovery would have resulted in fewer than 15 percent of the data points to miss DQOs. A.3.2.3 Matrix Spike Recoveries and Duplicate Precision. The matrix spike recovery and duplicate precision results for the analysis of the neat fuels used 1n this study are presented 1n Tables A-19 and A-20. Analysis of dlesel fuel 1n duplicate (Table A-19) Indicates that more than 95 percent of the analytes met precision DQOs. However, the differences between the two replicates suggests lower precision than this because a number of analytes detected in one replicate were not detected 1n the other. The difference between the quantitation limits for these samples Indicates that most of the analytes should not have been seen 1n the second replicate. Only 2-methylnaphthalene was above the quantitation limit for the second replicate, and thus should have been detected. Therefore, only 2 out of 23 analytes (phenanthrene and 2-methyl- naphthalene) did not meet met precision DQOs. Table A-20 summarizes matrix spike recovery of analytes spiked Into gasoline. More than 60 percent of the analytes did not meet recovery DQOs, suggesting matrix interferences. Tables A-21 through A-25 summarize the matrix spike recovery and duplicate precision results for the TCLP leachates analyzed. The 1,000 ppm leachate of each fuel/soil mixture was selected as the QC sample for each TCLP batch, and was analyzed as a matrix spike and 1n some cases matrix spike duplicate. Approximately 31 percent of the analytes did not meet matrix spike recovery DQOs, and about 23 percent of the duplicates did not meet precision DQOs. A.3.2.4 Method Blanks. Table A-26 presents the concentrations of the semlvolatlle organic target analytes found 1n the method blanks. All the reported values for the samples were corrected for the concentration found 1n the blanks. A-20 ------- DRAFT C0701-1A/UST 2/3/88 A.3.2.5 MRI QA Check Sample. The results of the analysis of an MRI QA check sample, which 1s used to check the analytical Instrumentation and data Interpretation, are summarized 1n Table A-27. All of the analytes met accuracy DQOs. A.4 REFERENCES FOR APPENDIX A 1. McAullffe, C. 1963. Solubility of C1-C9 hydrocarbons 1n water. Nature. 200:1092-1093. 2. Bohon, R. L.# and W. F. Clausen. 1951. The solubility of aromatic hydrocarbons 1n water. J. Amer. Chem. Soc. 73:1571-1578. 3. Tlssot, B. P., and D. H. Welte. 1984. Petroleum Formation and Occurrence. Sprlnger-Verlag. New York. p. 315. A-21 ------- DRAFT C0701-1A1/UST 12/22/87 1 TABLE A-l. SOIL CHARACTERIZATION RESULTS FOR KAW RIVER SOIL AND ACTION TOPSOIL Component concentrations, percent Action topsoll Kaw River soil Water, wt/wt 13.5 10.1 Sand 3.7 45 S1lt 55 51 Clay 41 4.0 ------- DRAFT C0701-1A3/UST 12/2/87 1 TABLE A-2. CONCENTRATIONS OF THE SOIL/FUEL PRODUCT MIXTURES Sataple No. Description (soil/ fuel noalnal concen- tration in ppm) Amount of fuel added, ug Total amount of fuel and soil, g Final concentration, g/kg* 7035/36 Action/diesel 10 1,960 200 0.00980 7033/34 Action/diesel 100 20,610 200 0.103 7029/30 Actlon/dlesel 1,000 200,000 200 1.00 705 J/52 Actlon/dlesel 1,000 200,000 202 0.991 7027/28 Action/diesel 5,000 1,000,000 200 5.00 7025/26 Action/diesel 10,000 2,010,000 200 10.0 7059/60 Action/fuel oil 10 1,600 200 0.00800 7023/24 Action/fuel oil 100 20,400 213 0.0960 7057/58 Action/fuel oil 1,000 210,000 205 1.03 7055/56 Action/fuel oil 5,000 1,070,000 201 5.33 7053/54 Action/fuel oil 10,000 2,040,000 200 10.2 7011/12 Action/gas 10 2,040 200 0.0102 7009/10 Action/gas 100 19,910 200 0.100 7005/6 Action/gas 1,000 200,000 200 1.00 7007/8 Action/gas 1,000 200,000 200 1.00 7003/4 Action/gas 5,000 1,020,000 200 5.10 7001/2 Action/gas 10,000 2,030,000 200 10.1 7045/46 Kaw/diesel 10 2,100 200 0.0105 7037/38 Kaw/diesel 100 20,120 200 0.101 7043/44 Kaw/dlesel 1,000 190,000 200 0.950 7041/42 Kaw/diesel 5,000 980,000 200 4.89 7039/40 KaM/Diesel 10,000 2,010,000 200 10.1 7069/70 KaN/fuel oil 10 2,100 200 0.0105 7021/22 Kaw/fuel oil 100 20,600 200 0.103 7067/68 Kaw/fuel oil 1,000 220,000 201 1.10 7065/66 Kaw/fuel oil 5,000 1,060,000 200 5.30 7063/64 Kaw/fuel oil 10,000 2,040,000 200 10.2 7019/20 Kaw/gas 10 2,600 200 0.0130 7049/50 Kaw/gas 100 20,600 200 0.103 7017/18 Kaw/gas 100 21,700 201 0.108 7031/32 Kaw/gas 1,000 230,000 200 1.15 7073/74 Kaw/gas 1,000 210,000 200 1.05 7015/16 Kaw/gas 5,000 980,000 200 4.90 7047/48 Kaw/gas 10,000 2,000,000 200 10.0 7013/14 Kaw/gas 10,000 2,010,000 200 10.0 "Final concentration (gAg) * Amo"n''' of Amount of fuel added (ug) n fuel+soil (g) 1 1.000 gAq 1,000,000 ug/g ------- DRAFT C0701-1A1/UST 12/2Z/87 2 TABLE A-3. IGNITABILITY TEST RESULTS FOR FUEL CONTAMINATED SOILS Fuel Fuel concentra- Flash Sustained Soil type tlon, ppm point combustion Action Gasoline 10,100 None None Action Gasoline 5,100 None None Kaw Gasoline 10,000 None None Kaw Gasoline 4,900 None None Action Diesel 10,000 None None Kaw Diesel 10,100 None None Action Fuel oil 10,200 None None Kaw Fuel oil 10,200 None None ------- DRAFT C0701-1A3/UST 12/2/87 2 TABLE A-4. VOLATILE AND SEMIVOLATILE ORGANIC TARGET ANALYTE CONCENTRATIONS IN UNLEADED GASOLINE, DIESEL FUEL, AND NO. 6 FUEL OIL GasolIne, ug/g Diesel, ug/g No. 6 fuel oi yg/g Volatile organic compounds Carbon dIsuIfI do Benzene ToIuene Ethylbenzene m-Xylene o- and p-Xylene Quantitation limit ND 13,800 58,000 12,500 33,100 32,100 500 NO TR 359 312 669 638 125 NO 10.0 60.4 21.7 69.0 65.2 6.00 SemivoI at Ile organic compounds Phenol Acenaphthene FIuoranthene 8enzofa)pyrene Acenaphthylene Anthracene Benzo(b1~{k J fIuoranthene Benzolg.h,I]fIuoranthene Benzo[aI anthracene Blphenyl Chrysene Olbenz[a,hi anthracene 2,4-0 i aethy I p heno I FIuorene Indeno(1,2,3-cd J pyrene 2-MethyI phenol 3- and 4-MethyI phenol 2-MethyI naphtha Iene Naphthalene Nitrobenzene Phenanthrene Pyrene Pyridine ND 2.36 ND m TRb 4.55 UD 2.20 5.28 2.01 4.69 2.92 ND NO 34.6 259 198 M) 6.57 NO W) W) f© NO NO NO TR NO NO NO 249 168 W> 8.42 467 ND 5.47 3.49 2,800 658 ND 1,250 205 W) ND f© NO 775 ND ND 663 M 9,810 NO 21,200 ND NO NO 237 ND W) 461 142 ND 506 8,710 NO Quantitation limit (base/neutrals) Quantitation limit (phenols) 2.00 2.00 113 2.00 453 2.00 ^ND = analyte not detected. TR = analyte detected, but at a concentration lower than the quantitation limit. ------- T J1-1A9/UST 2/3/88 2 TABLE A-5. THEORETICAL CONCENTRATIONS OF SELECTED TARGET ANALYTES IN ACTION TOPSOIL/FUEL MIXTURES AT THE VARIOUS NOMINAL CONCENTRATIONS® Soil/fuel mixture Actlon/dlesel Action/fuel oil Action/gasoline Noatnal concentration, ppa loo 1.000 5.000 10.000 5.000 10.000 100 1,000 5.000 10,000 Volatile oraanic coapounds. llg/kg Benzene'1 N/Ac N/A K/A N/A 53.3 102 1.374 13.797 70.369 140.063 Toluene6 37.0 356 1,795 3,608 322 616 5.774 57,986 295,756 568,671 Ethylbenzene 32.1 309 1,560 3.135 116 221 1.244 12,497 63.740 126.869 ¦-Xylene 6B.9 663 3,345 6.723 368 704 3,295 33,092 168,785 335,948 o- and p-Xylene 65. 7 632 3,190 6,412 348 665 3,196 32,092 163.685 325,799 Seal vol at lie oraanic coapounds. llq/kg Fluoranthene*' Beiuo[a)pyrene<' K/A N/A N/A N/A N/A N/A 0.235 2.36 12.0 23.9 N/A N/A N/A N/A 4.169 7,976 NO NO NO NO Be «o[b]*[k] fluoranthene"1 Benzolajanthracene1' N/A N/A N/A N/A 3.53S 6,762 0.453 4.55 23.2 46.2 M/A N/A N/A N/A 52.300 100.052 0.219 2.20 11.2 22.3 Bi phenyl 25.6 247 1,245 2,502 N/A N/A 0.525 5.28 26.9 53.6 Chrysened 17.3 166 840 1.66B 113.024 216.218 0.200 2.01 10.2 20.4 2.4-01atthylphenold 0.867 8.35 42.1 84.6 N/A N/A 0.466 4.68 23.9 47.6 Fluorene 48.1 463 2.335 4,693 N/A N/A 0.291 2.92 14.9 29.7 Indenol1,2,3-cd]pyrened N/A N/A N/A N/A 1,264 2.417 N/Ac N/A N/A N/A 2-Nethylphenol® 0.563 5.42 27.3 55.0 N/A N/A N/A N/A N/A N/A 3- and 4-Methylphenolb 0.359 3.46 17.4 35.1 N/A K/A 3.44 34.5 176 351 2-Nethylnaphthalene 288 2,774 14.000 28,139 2.458 4.702 25.7 259 1,319 2.625 Naphthalene*' 67.8 652 3,290 6,613 757 1,448 19.7 198 1,010 2,010 Phenanthrene 129 1,239 6.250 12,562 2,698 5,161 0.654 6.57 33.5 66.7 Pyrene 21.1 203 1.025 2,060 46,436 88,833 N/A N/A N/A N/A jjAnalyte theoretical concentration (llg/kg) • analyte concentration In neat fuel (llg/g)xconcentration of fuel In toll (a/kg). Analyte is lilted In Table C-l of the TCIP proposed rules (51 CFR 21648, June 13, 1986) °N/A • not applicable; analyte was not detected In the neat fuel. ''Analyte Is listed In Appendix VIII. ------- DRAFT C0701-1A9/UST 2/3/88 3 TABLE A-6. THEORETICAL CONCENTRATIONS OF SELECTED TARGET ANALYTES IN KAW RIVER SOIL/FUEL MIXTURES AT THE VARIOUS NOMINAL CONCENTRATIONS® Soil/fuel alxture Kaw/dlesel Kaw/fuel oil Kaw/qasol 1ne Noalnal concentration, ppa 10 100 1.000 5,000 10.000 1.000 S.000 10,000 10 100 1.000 5.000 10.000 Itolatlle organic ccapounifa. Ua/ka Benzene1* N/Ac N/A N/A N/A N/A 11.0 53.0 102 142 1.421 14 .489 67.613 137,966 Toluene1* 3.77 36.1 341 1.757 3,608 66.2 320 616 597 5,972 60 .897 284.172 579,855 Ethylbenzene 3.28 31.4 296 1.527 3,136 23.8 115 221 129 1.287 13 .124 61.244 124.969 •-Xylene 7.02 67.3 635 3.275 6.723 75.7 366 704 341 3,408 34 .753 162.174 330,917 0- and p-Xylene 6.70 64.2 606 3.123 6,412 71.5 346 665 331 3.305 33 .703 157.274 320.920 >¦1 volatile organic coapounds. Uq/kq Fluoranthene'' Beruo(a]pyrened N/A N/A N/A N/A N/A N/A N/A N/A 0.0243 0. 243 2.48 11.6 23.6 N/A N/A N/A N/A N/A 857 4.145 7,977 NO NO NO NO NO Benzo[bl+|k)f1uoranthened N/A tyA N/A N/A N/A 727 3,514 6.763 0.0468 0. 468 4.78 22.3 45.5 Benzol a) anthracene*' N/A N/A N/A N/A N/A 10.756 51.996 100.072 0.0227 0.227 2.31 10.8 22.0 B1phenyl 2.61 25.0 237 1,219 2,502 N/A N/A N/A 0.0543 0. 543 5.54 25.9 52.8 Chrysened 1.76 16.9 160 822 1.688 23.244 112.366 216.262 0.0207 0.207 2.11 9.85 20.1 2,4-01aethy1phenold 0.0884 0.847 8.00 41.2 84.6 N/A N/A N/A 0.0482 0. 482 4.92 23.0 46. B Fluorene Indeno[1,2,3-cd]pyrened 4.90 47.0 444 2.286 4.693 N/A N/A N/A 0.0301 0.301 3.07 14.3 29.2 N/A N/A N/A N/A N/A 260 1.256 2.418 N/Ac N/A N/A N/A N/A 2-Hethyl phenol'* 0.0574 0.550 5.19 26.8 55.0 N/A N/A N/A N/A N/A N/A N/A N/A 3- and 4-Hethylphenol'* 0.0366 0.351 3.31 17.1 35.1 N/A N/A N/A 0.356 3.56 36.3 169 345.4S 2-Methylnaphthalene 29.4 282 2.660 13.706 28,140 505 2,443 4.703 2.66 26.6 272 1,267 2.586 Naphthalene 6.91 66.2 625 3,221 6,613 156 753 1.449 2.04 20.4 208 970 1.980 Phenanthrene 13.1 126 1.187 6.119 12.563 555 2,682 5.162 0.0677 0.677 6.90 32.2 65.7 Pyrene 2.15 20.6 195 1.003 2.060 9,550 46.165 88.851 N/A N/A N/A N/A N/A jjAnalyte theoretical concentration (P9/kg) • analyte concentration In neat fuel (P9/9) « concentration of fuel In soil (9/kg). Analyte is listed In Table C-l of the TCLP proposed rules (51 CFR 21648, June 13, 1986) °N/A ¦ not applicable; analyte was not detected In the neat fuel. dAnalyte is lilted In Appendix VIII. ------- *FT wd701-lA6/UST 12/1/87 2 TABLE A-7. CONCENTRATIONS OF CONTAMINATED ACTION TOPSOIL LEACHATES Act I On/d I ase I Actloil/f U6l Ol I ActlQn/imcAl I ha Nominal concentration, ppra 100 1 ,000 5,000 10,000 5,000 10,000 100 TJOOO——5(00Q 10,000 Volatile organic compounds. M/l Benzene Toluene Ethyl benzene m-Xylene o- and p-Xylene NAa NA NA NA NA NO TR TR 3.57 4.30 TRC 11.7 20.8 47.6 51.0 4.04 62.2 52.0 106 110 115* 52.9d 5.44 14.4 13.4 17.3 47.9 13.8 39.6 40.0 NO TR NO NO TR TR 7.62 6.38 22.2 28.6 TR 417 402 1,220 1,340 109 2,620 1,210 3,460 3,670 Quantitation limit NA 2.00 2.00 2.00 2.00 2.00 2.00 5.00 50.0 50.0 Semi vol at 1le orqanic compounds, ya/i Acenaphthene Bi phenyl F1uorene 2-Methylnaphthalene Naphthalene Phenanthrene ND NO NO 3.00 NO NO 3.33 6.66 4.45 84.6 27.1 3.81 6.14 11.7 9.02 269 112 11.1 6.93 15.6 11.9 251 174 13.7 NO TR NO 11.6 10.1 2.79 NO 2.76 TR 19.1 16.7 3.01 NO NO NO ND NO NO NO TR ND 54.2 63.2 NO ND TR ND 246 253 ND ND TR ND 350 448 NO Quantitation limit 2.00 2.41 2.00 2.04 2.22 2.04 2.05 2.25 3.09 10.1 cND = analyto not detected. dTR = analyte detected, but at a concentration lower than the quantitation limit. Potentially inaccurate results; see text for discussion. ------- DRAFT C0701-1A9/UST 2/3/88 4 TABLE A-8. CONCENTRATIONS OF CONTAMINATED KAM RIVER SOIL LEACHATE Soil/fuel «1xture Kaw/dlesel Haw/fuel oil Kaw/oasolIne Ncmlnal concentration, ppa 10 100 1.000 5.000 10.000 1.000 5.000 10.000 10 100 1,000 5.000 10,000 Volatile organic compounds. Ug/L Benzene 60.9s 2.04 51.8® 7.45 42. < 8.96 6.26 11.1 TR 31.6 241 1.660 2,380 Toluene 17.7® 2.09 51.6a 36.7 66.4 11.7 13.4 38.0 15.9 209 1,480 8,420 13,100 Ethylbenzene TR6 TR 15.0 31.3 47.4 2.60 5.57 10.2 4.88 46.9 329 1,800 2,530 ¦-Xylene 8.54® TR 32.6 64.5 99.1 6.80 16.7 30.3 13.5 129 876 4,520 6.560 o- and p-Xylene 3.01s TR 33.8 66.5 104 6.66 17.0 31.3 12.7 131 899 4,600 6.720 Quantitation 1Init 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 4.00 50.0 400 400 Seel volatile organic coapounds. llq/L Acenaphthene NAC NDd 5.35 8.40 7.39 NO TR NO NO ND ND ND ND 81 phenyl NA TR 10.6 18.6 16.3 TR 3.08 3.17 ND ND TR TR TR Fluorene HA NO 7.10 11.7 9.42 TR TR TR NO ND NO ND NO 2-Methylnaphthalene NA 8.32 123 251 203 10.5 20.9 34.0 HO 10.7 106 343 439 Naphthalene NA 4.47 40.6 131 125 6.38 13.6 16.5 TR 7.36 99.1 351 625 Phenanthrene NA TR 9.53 14.6 12.3 2.54 3.22 3.41 ND NO NO NO ND Quantitation Halt HA 2.00 2.04 2.04 2.00 2.04 2.00 2.00 2.02 2.00 2.01 4.04 10.1 'Potentially Inaccurate results; see text for discussion. **TR • analyte detected, but at a concentration lower than the quantitation liolt. CNA ¦ saople not analyzed. dND » analyte not detected. ------- T . J1-1A6/UST 12/1/87 3 TABLE A-9. RESULTS OF RECOVERY CALCULATIONS FOR TCLP OF FUEL-CONTAMINATED ACTION TOPSOIL® Soil/fuel mixture Actlon/dlesol Action/fuel oil Actlon/qasolIne Nominal concentration, ppm 100 HOOD TWO 10,000 57000 10,000 "TOO 1,000 5,000 10,000 VoI at 11e organ Ic compounds. percent Benzene NA NPC NP NP 4,310d 339d 0 0 0 Toluene NA O 13.1 34.6 329 156 0 0.268 2.83 Ethyl benzene NA 0 26.8 33.3 94.0 125 0 1.04 12.6 m-Xylene NA 10.8 28.6 31.6 78.3 113 0 1.37 14.5 o- and p-Xylene NA 13.6 32.1 34.4 77.1 120 0 1.82 16.4 Semi vol at lie organic compounds, percent Biphenyl FIuorene NP NP NP NP NP NP NP NP 0 54.0 18.8 12.5 NP NP 0 0 0 19.2 7.73 5.07 W NP 0 0 20.8 61.0 38.4 17.8 9.43 8.12 0 419' 0 83.2 68.1 52.6 26.7 23.1 0 639' 0 6.15 3.55 2.18 2.07 1.17 0 0 1.56 8.90 19.1 20.6 22.5 Acenaphthene NP NP NP NP NP NP NP NP NP NP 0 0 2-MethyI naphthalene 20.8 61.0 38.4 17.8 9.43 8.12 0 4l9d 373^ 278d Naphthalene 0 83.2 68.1 52.6 26.7 23.1 0 639d 501d 446d Phenanthrene * " "" ~ ~~ TCLP recovery percent = , .total g^unt oj analyte In leachate (ug) )0Q b... TOTai amounT ot anaiyTe in conTaminaTea soil (ugj CNA = sample not analyzed. (jNP = recovery calculation not processed due to a contaminated soil concentration of 0. Unusually high recovery; see text for details. ------- DRAFT C0701-1A9/UST 2/3/88 S TABLE A-10. RESULTS OF RECOVERY CALCULATIONS FOR TCLP OF FUEL-CONTAMINATED KAM RIVER SOILa Soil/fuel alxture Kaw/dlesel Ka Fluorene NA 0 32.1 10.3 4.02 HP HP NP 0 NP 0 0 HP 2-Hethylnaphthalene MA 59.1 92.8 36.7 14.4 41.6 17.1 14.5 0 HP 781d 541d ff> Haphthalene NA 135 130 81.5 37.8 82.0 36.2 22.8 0 NP 9S4d 724d NP Phenanthrene MA 0 16.1 4.78 1.96 9.16 2.40 1.32 0 HP 0 0 W> a total aoount of analyte In leachate (lla) TaP recovery, percent ¦ xl00 total aoount of analyte 1n contaminated soil (VIg) bNA • saaple not analyzed. SlP ¦ recovery calculation not processed due to a contaminated soil concentration of 0. ''Unusually high recovery; see text for details. ------- DRAFT C0701-1A3/UST 12/2/87 3 TABLE A-12. QUALITY CONTROL RESULTS FOR VOLATILE ORGANIC ANALYSES: KAW RIVER SOIL 1,000 ppm UNLEAOED GASOLINE LEACHATE QA type (dup.. MS, MS dup.) Matrix Matrix spike spike Sample, Dup., Average, RPO, a level, recovery >> Analytes pg/l pg/L pg/L percent ug/L percent Carbon d1su1f1de NO NO 25C NA 1,340 1,250 105 Benzene 258 224 241 14 1,610 1,250 110 Toluene 1,640 1,310 1,475 22 2,550 1,250 86 Ethyl benzene 373 285 329 27 1,590 1,250 101 m-Xylene 975 777 876 23 1,980 1,250 88 o- and p-Xylene 1,010 788 899 25 3,380 2,500 99 d Quantitation limit: 50 50 NA NA 50 NA NA NA = not applicable; NO 3 not detected. ^RPD (relative percent difference) = ((Rep. I - Rep. 2)/Avg. of Rep. 1 and Rep. 2)xlOO. cPercent recovery » ((amount found in spike - native level avg.)/amount spiked)x100. The native level of this analyte was estimated as 50 percent of the quantitation limit for ^percent recovery calculations. Quantitation limit * (ng of lowest standard x dilution factor)/ml of sample analyzed. ------- DRAFT C0701-1A3/UST 12/2/87 4 TABLE A-13. QUALITY CONTROL RESULTS FOR VOLATILE ORGANIC ANALYSES: KAW RIVER SOIL 1,000 ppm DIESEL LEACHATE QA type (dup., MS. MS dup.) Matr i x Matr i x spike spike Sample, Oup., Average, RPO, a MS, level, recovery^, Analy+es yg/L yg/L yg/L percent wg/,|~ percent Carbon d i su1fIde NO NO 1C NA 53 50 104 Benzene 51.9 51.8 52 0.2 106 50 108 Toluene 53.4 49.8 52 7 92 50 81 Ethyl benzene 16.7 13.2 15 23 60.7 50 92 tn-Xy lene 39 26.3 33 39 79.9 50 95 o- and p-Xylene 41.9 25.8 34 48 122 100 88 Quantitation limit:'6 2 2 NA NA 2 NA NA NA 3 not applicable; NO * not detected (no response for this analyte). ^RPD (relative percent difference) = ((Rep. 1 - Rep. 2)/Avg. of Rep. 1 and Rep. 2)x100. percent recovery 3 ((amount found In spike - native level avg.)/amount spiked)x100. ihe native level of this analyte was estimated as 50 percent of the quantitation limit for ^percent recovery calculations. eOutside data quality objective limits. Quantitation limit (aqueous samples) * (ng of lowest standard x dilution factor)/ml of sample analyzed. ------- DRAFT C0701-1A3/UST 12/2/87 5 TABLE A-14. QUALITY CONTROL RESULTS FOR VOLATILE ORGANIC ANALYSES: KAW RIVER SOIL 1,000 ppm NO. 6 FUEL OIL LEACHATE QA type (dup.. MS, MS dup.) Matrix Matrix spike spike Sample, Dup., Average, RPD, a MS, level, recovery^ Analytes ug/L Ug/L ug/L percent Ug/L percent Carbon disulfide f© NO 1C NA 45.2 50 88 Benzene 8.44 9.47 9 12 55.8 50 94 Toluene 11.6 11.8 12 2 59.7 50 96 Ethyl benzene 2.57 2.62 3 2 50.8 50 96 m-Xylene 6.87 6.72 3 2 52.6 50 92 o- and p-Xylene 6.46 6.85 7 7 103 100 96 Quantitation limit:"' 2 2 NA NA 2 NA NA NA = not applicable; NO = not detected (no response for this analyte). ^RPO (relative percent difference) » ((Rep. I - Rep. 2)/Avg. of Rep. t and Rep. 2)x100. cPercent recovery = ((amount found in spike - native level avg.)/amount spiked)xlOO. The native level of this analyte Mas estimated as 50 percent of the quantitation limit for ^percent recovery calculations. Quantitation limit (aqueous samples) = (ng of lowest standard x dilution factor)/mL of sample analyzed. ------- DRAFT c0701-1A3/UST 12/2/87 6 TABLE 15. QUALITY CONTROL RESULTS FOR VOLATILE ORGANIC ANALYSES: ACTION TOPSOIL 1,000 ppm UNLEADED GASOLINE LEACHATE QA type (dup., MS, MS dup.) Matrix Matrix spike spike Sample, Dup., Average, RPD, a MS, level, recovery^ Analytes wg/L yg/L yg/L percent lJ9/'L y9/L percent Carbon disulfide NO NO 2.5C NA 126 125 99 Benzene TR TR 2.5C NA 136 125 107 Toluene 7.21 8.04 8 11 126 125 95 Ethyl benzene 6.19 6.56 6 6 127 125 97 m-Xylene 21.8 22.5 22 3 139 125 93 o- and p-Xylene 28.2 29.1 29 3 267 250 95 Quantitation limit:*1 5 5 NA NA 5 NA NA NA » not applicable; NO = not detected; TR = detected, but at a level lower than the quantitation limit. ?RPO (relative percent difference) = ((Rep. 1 - Rep. 2)/Avg. of Rep. 1 and Rep. 2)x100. Percent recovery = ((amount found in spike - native level avg.)/amount spiked)x100. The native level of this analyte was estimated as 50t of the quantitation Iimit for t ^recovery calculations. Quantitation limit (aqueous samples) 3 (ng of lowest standard x dilution factor)/mL of sample analyzed. ------- DRAFT C0701-1A3/UST 12/2/87 7 TABLE A-16. QUALITY CONTROL RESULTS FOR VOLATILE ORGANIC ANALYSES: ACTION TOPSOIL 1,000 ppm DIESEL LEACHATE QA type (dup.. MS. MS dup.) Ana 1ytes Sample, ug/L Dup., ug/L Average, ug/L RPD. a percent MS, ug/i Matrix spike level, ug/i Matr 1 x spike recovery ^ percent Carbon disulfide ND NO 1C NA 53.2 50 104 Benzene NO NO 1c NA 54.3 50 107 Toluene TR TR 1c NA 52.9 50 104 Ethyl benzene 1.91 1.65 2 15 54.5 50 105 m-Xy1ene 3.73 3.4 4 9 55.8 50 104 0- and p-Xylene 4.69 3.9 4 18 109 ' .0 105 Quantitation limlt:d 2 2 NA NA 2 NA NA NA 3 not applicable; NO = not detected; TR = detected, but at a concentration below the quantitation limit. bRPD (relative percent difference) » ((Rep. 1 - Rep. 2)/Avg. of Rep. 1 and Rep. 2)x100. cPercent recovery » ((amount found in spike - native level avg.)/amount spiked)xlOO. The native level of this analyte was estimated as 50 percent of the quantitation limit for ^percent recovery calculations. Quantitation limit (aqueous samples) » (ng of lowest standard x dilution factor)/mL of sample analyzed. Quantitation limit (nonaqueous samples) » (ng of lowest standardxdilutlon factorxtotal mL extract)/(|iL injection volumexg or mL samp I ex 1,000). ------- DRAFT C0701-1A2/UST 12/1/87 2 TABLE A-18. QUALITY CONTROL RESULTS FOR SEMIVOLATILE ORGANIC ANALYSES: SURROGATE RECOVERIES Sample1 Recovery of surroaates 2-FPn Ds-Ph TBPh DS-NB 2-FB1 Dio~Py Dllt-Ter 80 79 115 53c 14C 54c 56c 53c 61c 81 45c 9.5C 41c 0C 86 78 120 74 51c 92 80 92 92 106 132c 44c 171° 27c 73 70 104 0C 86 95 82 75 44c 51c 35c 34c 86 84 65c 66c 69c 0C 67c 82 81 0C 88 162c 50c 87 87 62c 51c 82 90 78 83 92 60c 77 68c 75 57° 52c 30c 77 88 67c 57c 0C 56c 72 67c 94 53c 77 19c 35c 67c 47c 55c 55c 61c 0C 54c 56c 58c 0C 33c 86 54c 45c 70 72 0C 71 83 33c 60c 97 99 87 78 83 24c 53c 78 56c 69c 66c 79 25c 60c 67c 67c 64c 59c 85 57° 48c 77 73 59c 55c 82 39c 45c 76 74 0C 61c 46c 62c 23c 84 78 77 32c 84 73 52c 91 86 64c 49c 60c 0C 42c 88 74 78 70 78 0C 71 75 74 47c 35c 69c 53c 62c 67c 70 Unleaded gasoline Unleaded gasoline (MS) Diesel fuel Diesel fuel - duplicate No. 6 fuel oil Act1on/gasol1ne 100 Act1on/gasol1ne l,000d Act1on/gasol1ne 5,000 Act1on/gasol1ne 10,000 Act1on/d1esel 100 Act1on/d1ese1 1,000 Act1on/d1esel 5.000 Act1on/d1esel 10,000 Action/fuel oil 1,000-MS Action/fuel oil 1,000-MSdup Action/fuel oil 5,000 Action/fuel oil 10,000 Kaw/gasol1ne 10 Kaw/gasol1ne 100 Kaw/gasol1ne l,000d (continued) ------- JMFT W01-1A2/UST 12/1/87 3 TABLE A-18. (continued) Sample3 Recovery of surrogates'3 2-FPh Ds-Ph TBPh DS-NB 2-FB1 Oio-Py Dm-Ter Blanks 58c 58c 60c 0C 63^ 79 72 L 73 7°_ 94 54c 47 80 81 73_ 65 76 0C 35c 82 61c 0C 77 137 °r 14c 34c 32c 72 52c 127 172 106 79 77 0C 81 83 4C 42° 84 74 61c S6r 56c 55c 56c 75 73 56c 57 53c 56c 69c 89 88 67c 62c 88 58c 26c 61° 0C V Average recovery 58 62 81 44 54 79 67 Percent RSD 49 27 31 91 37 26 34 Minimum value 0 17 40 0 9.5 30 0 Maximum value 98 92 162 172 106 171 99 Percent outside DQOs 61 64 27 93 82 23 32 aSample descriptions for TCLP extracts are 1n the following format: soil type/petroleuro product and .nominal concentration (ppm). 2-FPh = 2-fluorophenol; D5-Ph = Ds-phenol; TBPh = 2,4,6-trlbromophenol; DS-NB = Ds-n1trobenzene; 2-FB1 ¦ 2-fluoroblphenyl; D10-Py = D10-pyrene; D1It-Ter = Dj„-terphenyl. Rvalue 1s outside recovery DQOs. Includes sample and two matrix spike duplicates. ------- DRAFT c0701-lA6/UST 12/1/67 4 TABLE A-21. QUALITY CONTROL RESULfS TOR S£fU VOLATILE ORGANIC ANALYSES: KAW RIVER SOIL 1,000-ppm GASOLINE LEACHATE _ OA type (dup,. KS. HSdup) Analytes vg/v Matrix spike MS. lever, JJJ/L Matrix spike rscowrvti percenr MS dup, ugA MatrIx spike level, wg/C Matrix sp i ke recovery j> fwrcenr Matrix spffce dupIicate WO, c percent Acenaphtftene Acenaphthy lene Anthracene Benzol alpyrene Benzolb)*IkIfIuoranthene Benzo(g,h,IIperyIene Benz(alanthrecene Biphenyi Chrysene Di benzla,h]anthracene 2,4-Olaethy Ipiienol Fluoranthene FIuorene Indenoll ,2,3-cd|pyrM« 2-Met hy I f\aph tha I ene 2-ttethyI phenol 3- and 4-Metbylphenol Naphthalene N1trobenzene Phenanthrene Phenol Pyreue Pyridine Quant i tat i an limit:- 5.35 1.01 1.01 1.01 1.01 1.01 1.01 10.6 1.01 1.01 1.01 1.01 7.1 1.01 123 1.01 1.01 40.6 1.01 9.53 1.01 1.01 1 .01 2.01 201 300 65" 245 300 80 20 218 300 79 278 300 92,1 16 229 500 76 195 300 Md 26 i 300 a? 379 300 126,. "S 574 600 95 838 600 139 j 3*2 263 300 87 152 300 50 53d 222 300 74 296 300 96 29ri 9.3 NS HAd 12.9 NS NA 32 206 300 68 274 300 28 262 300 87 NO 300 0* NA NO NS NA NO NS NA NA 252 500 84 273 300 9! 8 2*4 300 76 290 300 94 21 243 300 81 328 300 109 30 125 MS MA. 160 NS NA„ 25 j 137 600 79.6 600 53d 188 1200 16d 140 1200 '2d 29 231 300 60 248 300 69 12 226 300 76 228 300 76 0 222 300 71 245 300 10 241 600 40 210 600 35 14 238 300 7% 229 500 4 SD 300 or1 ND 300 0 NA 3.15 NA NA 3.49 NA NA NA NS = dot spiked; MA = rot applicable, a Anajytes with concentrations belon the quantitation limit were estimated to have a native level of 50 percent of the fluantftatjew .1iralt. percent recovery = demount found In spike - native level evg]/amount spiked)*!00. (relative percent difference) = (IRep I - Rep 21/Avg of ftep I and Hep 2)xl00. ^Outside data quality objective Hafts. Quantitation limit * equivalent niLA for total sample x lowest calibration standard (|ig/mL). ------- r 1-1A6/UST 12/1/87 5 TABLE A-23. QUALITY CONTROL RESULTS FOR SEMIVOLATILE ORGANIC ANALYSES: KAW RIVER SOIL 1,000-ppm DIESEL LEACHATE Analytes QA type (dup.. MS. MSdup) Sampl%, wg/L MS. yg^L Matrix spike level, ug/L Matrlx spike recovery b percent MS dup.. ygA Matrix spike level, pg/f. Matrix spike recovery j) percen-r Matrix spike dupIIcate RPO. c percent Acenaphthene 5.35 201 300 Acenaphthylene 1.02 238 300 Anthracene 1.02 229 300 Benzol ajpyrene 1.02 261 300 Benzolb1~1k1f1uoranthene 1.02 574 600 Benzo l'g,h,l Iperyiene 1.02 263 300 Benz(a 1 anthracene 1.02 222 300 Bl phenyl 10.6 9.3 NS Chrysene 1.02 206 300 D1benz[a,h)anthracene 1.02 262 300 2,4-D1 methy1pheno1 1.02 W) NS Fluoranthene 1.02 252 300 F1uorene 7.1 234 300 Indenol1,2,3-cdlpyrene 1.02 243 300 2-Methy1 naphtha 1ene 123 125 NS 2-Methy1 phenol 1.02 137 600 3- and 4-Methylphenol 1.02 188 1200 Naphthalene 40.6 221 300 Nitrobenzene 1.02 228 300 Phenanthrene 9.53 222 300 Phenol 1.02 241 600 Pyrene 1.02 238 300 Pyridine 1.02 ND 300 Quantitation limit:0 2.04 4.04 NA 65d 245 300 80 20 79 278 300 92d 16 76 195 300 65 '®d 87 379 300 126d 37d 95 838 600 139 37d 87 152 300 50 53d 74 296 300 98 29d NAd 12.9 NS NA 32 68 274 300 91 28 87 NO 300 NA NA NA NO NS NA NA 84 273 300 91 8 76 290 300 94 21 81 328 300 109 30 NAd 160 NS NAd 25d 23d 79.6 600 ,3S 53d 16° 140 1200 29 60 248 300 69 12 76 228 300 76 0 71 245 300 78d 10 40 210 600 35 14 7% 229 300 76d 4 0 NO 300 oa NA NA 2.03 NA NA NA £|S = not spiked; NA = not applicable. ^Analytes with concentrations below the quantitation limit were estimated to have a native level of 50 of the quantitation limit. cPercent recovery = ([amount found in spike - native level avgl/amount splked)x100. dRPD (relative percent difference) = ([Rep 1 - Rep 2)/Avg of Rep 1 and Rep 2)x100. ^Outside data quality objective limits. Quantitation limit = equivalent mL/L for total sample x lowest calibration standard (iig/mL). ------- |