Acr ? iC^UNiTtD STATES ENVIRONMENTAL PROTECTION AGENCY hrt\ _ -' i.w*. . Guidance on Petroleum Refinery Waste Analyses for Land >u ' Treatment Permit Applications John Skinner, Director^/) ( Office of Solid Waste /l*^ T0 Hazardous Waste Fermi tL/Branch Chiefs, Regions I-X Introduction The purpose of this memo is to provide permit writers guidance on evaluating petroleum refinery waste analyses submitted in land treatment permit applications. A list of Appendix VIII hazardous constituents suspected to be present in petroleum refinery wastes and a special analytical method for refinery wastes are provided. Background The general Part B information requirements specified under $270.14(b) require the submittal of (1) chemical and physical analysis data on the hazardous wastes to be handled at the facility including all data that must be known to treat, store, or dispose of wastes properly in accordance with Part 264, and (2) a copy of the waste analysis plan. In addition, the specific information requirements under §270.20 require an owner/operator of any facility that includes a land treatment unit to submit "a list of hazardous constituents reasonably expected to be in, or derived from, the wastes to be land treated based on waste analyses performed pursuant to §264.13." Also, S270.20(a) stipulates that the description of the treatment demonstration plan must include a list of potential hazardous constituents in the waste. Because the design and management of a land treatment unit is based on the goal of attaining treatment of hazardous constituents (i.e., constituents listed in Appendix viii), it is very important that the presence of these constituents in the land treated wastes be accurately identified and quantified. This is best achieved through a comprehensive waste analysis for all Appendix VIII constituents. However, due to the cost and analytical difficulties associated with these analyses, many applicants have submitted requests to conduct analyses for some subset of Appendix VIII, which are "reasonably expected to be in or derived from the wastes to be land treated." To date, the majority of wastes proposed for land treatment have been petroleum refinery wastes, specifically the listed wastes K048-K052. If* f~ ------- The evaluation of these Appendix VIII subsets for each land treatment application has been difficult due to the lack of published information on specific organic compounds In refinery wastes, and also due to the variability of waste characteristics within the refinery industry. However, OSW has gathered sufficient information from EPA research studies, in—house waste studies and analyses, and refinery process evaluations to develop a conservative list of hazardous constituents that are suspected to be present in petroleum refinery wastes. This list is proviced in Attachment 1. This list should be used by permit writers as a guide in determining which constituents may and may not be eliminated from consideration when completing waste analyses for a land treatment permit application. Additional explanation of the derivation and use of this list is provided below. Derivation and Use of List The list of hazardous constituents suspected to be present in refinery wastes was derived from a review of data on petroleum refinery wastewater and sludge characteristics from the following sources: (1) literature, particularly EPA research reports: (2) in—house waste analyses completed by EPA research laboratories; (3) preliminary data from the OSW refinery waste study; and (4) an evaluation of petroleum refinery processes. Although these four sources were used, the data base on specific hazardous organic constituents in sludges was still limited. Considerable weight was placed on wastewater data as indicators of sludge characteristics (e.g., API separator sludge). Also, the list in Attachment 1 is a generic list developed by combining waste analysis data on all five listed refinery wastes (X049—K052). Due to the lack of extensive data, no attempt was made to differentiate between the characteristics of these five refinery wastes. Until sufficient information is available to allow development of separate lists for each waste, the attached list should be considered applicable to dissolved air flotation float (1 (048), slop oil emulsion solids (K049), heat exchanger bundle cleaning sludge (1(050), API separator sludge (1(051), and leaded tank bottoms (1(052). To compensate for the limited data base and variability among refineries, the attached list is purposely comprehensive. It includes a total of 89 hazardous constituents or groups of constituents (e.g., trlchlorobenzenes). All of these con- stituents have been identified as possibly being present in the above referenced wastes. Many of the compounds on the list may be present at low concentrations and others may not be present at all in certain wastes at some refineries. The permit writer should use the attached list as a guide to the Appendix VIII constituents that should be addressed in ------- the up—front waste analyses and waste analysis plans for Part B applications that propose land treatment of petroleum refinery wastes. A permit applicant may further refine this list by providing detailed evidence that certain hazardous constituents cannot be present in the listed wastes at that particular refinery. In most cases, i owever, waste analysis data on the constituents listed in Attachment 1 will be necessary to make this showing. Analytical Methods To assist In the analysis for specific organic constituents in petroleum refinery wastes, OSw has developed a column cleanup procedure which is provided in Attachment 2. This draft method is used specifically to separate semivolatile a.liphatic, aromatic, and polar compounds in the waste matrix. The method should be used only by experienced residue analysts. Volatile compounds are determined using method 8240 with PEG (tetraglyme) Extraction. Test method 3050 should, be used for all metal analyses. These methods are described in SW—846. Relationship to Delisting and Listing Efforts Finally, the attached list is consistent with the waste analysis information that EPA has requested from delisting petitioners. Many petroleum refinery operators who are preparing Part B applications for land treatment facilities also have submitted delisting petitions to the Agency for one or more of their wastes.• It is important that the waste analysis data requested by the Agency for permitting and delisting be consistent, although there may be differences in the extent of data necessary in certain cases. Therefore, the list of Appendix VIII constituents provided in Attachment I is also being used in refinery delisting actions. Additional information on non—Appendix VIII constituents, however, is being collected as part of OSW’s new waste assessment and listing efforts for petroleum refineries. These compounds, which are listed at the end of Attachment 1 for your Information, may be added to Appendix VIII in the future. Although it Is not required at this time, permit applicants should be encouraged to provide information on these waste constituents. If you have any questions on the listing of specific hazardous constituents in Attachment 1 or on the recommended test methods, please contact Ben Smith (382—4791) of the Waste Identification Branch. Other questions pertaining to the use of the above guidance In permitting land treatment facilities should be directed to Mike Flynn (382—4489) of the Land Disposal Branch. Attachments cc: Jack Lehman 1att Straus Fred Lindsey Bruce weddle Ken Shuster Peter Guerrero Eileen Claussen ------- ATTACHMENT 1 Appendix VIII Hazardous Constituents Suspected to be Present in Refinery Wastes **Acetonjtrjle (Ethanenitrile) **Acrolein (2—Propenal) **AcryTonjtrlle (2—Propenenitrile) Aniline (Benzenamine) Antimony - Arsenic Baium Benz (c) acridine (3,4—Benzacridine) Benz (a) anthracene (1,2-Benzanthracene) **Benzene (Cyclohexatriene) Benzenethiol (Thiophenol) Benzidine (1,1-Biphenyl—4,4diarnine) Benzo(b)fluoranthene (2,3-Ben,zofluoranthene) Benzo(j )fluoranthene (7 ,8-Benzofluoranthene) Benzo(a)pyrene (3 ,4-Benzopyrene) •*Benzyl chloride (Benzene, (chioromethyl)-) Beryllium Bis (2—chioroethyl) ether (Ethane, 1,1—oxybis (2—chioro—)) Bis (2—chioroisopropyl) ether (Propane, 2,2—okybis (2—chioro—)) •eBis (chioromethyl) ether (Methane, oxybis (chloro)) Bis (2—ethyihexyl) phthaIate (1,2—Benzenedicarboxylic acid, bis (2-ethyihexyl) ester) Butyl benzyl phthalate (1,2-Benzenedicarboxylic acid, butyl phenyimethyl ester) Cadmium Carbon disulfide (Carbon bisulfide) p-CM oro-m—cresol e*Chlorobenzene (Benzene, chioro—) **Chloroform (Methane, trichloro-) **Chloromethane (Methyl chloride) 2— Chloronapthalene (Naphthalene, beta—chioro—) 2-Chiorophenol (Phenol, o-chloro-) Chromium Chrysene (1 ,2-Benzphenanthrene) Cresols (Cresylic acid) (Phenol, methyl—) **crotonaIdehyde (2-Butenal) Cyanide .Oibenz(a ,h)acridine (1,2,5,6—Oibenzacridine) Dibenz(a,j)acrldine (1,2,7,8—Dibenzacr ldine) Dibenz(a,h)anthracene (1,2,S,6—Dibenzanthracene) 7H-Dibenzo(c,g)carbazole (3,4,5,6-Dibenzcarbazole) Dibenzo(a ,e)pyrene (1,2,4 ,5—Di benzpyrene) Dibenzo(a,h)pyrene (1,25,6—Dibenzpyrene) Dibenzo(a,i )pyrene (1,2,7,8-Dibenzpyrene) 1 ,2-Dibromoethane (Ethylene dibromide) Di—n-butyl phthalate (1,2—BenzenedicarboxylIc acid, dibutyl ester) *Dj chi orobenzenes *t ,2-Dichloroethane (Ethylene dichioride) *ttrans_1,2_Djchloroethene (1,2-Dichiorethylene) 1 ,1—Dichloroethylene (Ethene, 1 ,1—dichloro—) **Djchloromethane (Methylene chloride) ------- 2 t Di chioropropane Dichioropropanol Diethyti phthalate (1,2 BenzenedicarboxyliC acid, diethyl ester) 7 ,12_Dimethyl —beflz(a)aflthracefle 2,4_DimethyiPheflOl (Phenol, 2,4-dimethyl-) Dimethyl phthalate (1,2—Benzeriedicarboxylic acid, dirnethyl ester) 4 ,6_Dinitro -o —cresol 2,4 DinitrOPheflO1 (phenol, 2,4-nitro-) 2fra_Dinitrotoluene (Benzene, 1 —methyl—2,4—din;trO—) Di—n—oCtyl phthalate (1,2 —Benzenedicarboxylic acid, dioctyl ester) **1,4_Dioxane (1,4-Diethylene oxide) 1,2 Diphenyihydra2ifle (Hydrazine, 1,2-diphenyl—) thyleneimifle (Azridine) etEthylene oxide (Oxirane) Fluoranthene (Benzo (j,k) fluorene) **Formal dehyde Hydrogen sulfide (Sulfur hyd 9de) Indeno (1,2,3-cd)pyrene (1 10(1,2_phenylene)pyrefle) Lead Mercury Methanethiol (Thiomethanol) 3—Methyichiolanthrefle (Benz(j)aceanthrylene, l,2_dihydro_3 methyl—) Methyl ethyl ketone (MEK) (2-Butanone) NaphthaIene Nickel p-Nitroaniline (Benzenamine, 4—nitro-) Nitrobenzene (Benzene, nitro—) 4—Nitrophenol (Phenol ,pentachI cr0-) PentachiorophenOl (Phenol, pentachioro—) Phenol (Benzene, hydroxy—) Pyridine Selenium * ,**Tetrachloroethafles **Tetrachloroethylene (Ethene, lj,2,2—tetra chioro-) **Toluene (Benzene, methyl—) Tri chiorobenreneS *,**Tri chioroethaneS **Trlchloroethefle (Trichioroethylefle) *Irjchlorophenol s Vanadium * If any of these groups of compounds are found, the specific Isomers listed in Appendix VIII should be identified. ** Use Test Method 8240 for these volatile compounds. Use Test Method 3050 in SW-846 for all metals; see Attachment 2 for semivolatile organic compounds. ------- Non—Appendix VIII Constituents of Concern (may be added to App. VIII ) Cobalt Indene 1-Methylnapthalene 5-Nitro acenaphthene Styrene Quinoline Hydroqulnone Phenanthrene Anthracene Pyrene ------- ATTACHMENT 2 Column Cleanup of Petroleum Wastes I ntroduct ion The following procedure is intended for application to the analysis of semivolatile organic compounds in oily waste samples. Its application is necessary in those cases where Lhe conventional cleanup procedures (Methods 3510, 3520, 3540, 3550) fail, to provide suitable detection limits (approx-’ imately loppm) for the semivolatile compounds specified in Attachment 1. Analysis of the cleaned—up extracts should be performed according to Method 8270, a capillary GC/MS technique. It should be noted that this procedure is in draft form. It may be modified as more experience is gained. Cleanup Techniques It is anticipated that after a sample is subjected to conventional extraction procedures (Methods 3510,3520, 3540, and 3550) or after dilution, a cleanup step may be required to remove matrix interferences and yield acceptable detection limits for compounds of interest. Determination as to whether an extract needs to be cleaned can usually be provided by either examination of the sample itself or by knowledge of the particular waste stream that was sampled. It is also possible to estimate whether or not the extract is suitably clean for GC/MS analysis. An aliquot of the methylene chloride extract can be evaporated to dryness and the total amount of material in the aliquot weighed. In general, if the extract contains less than a few milligrams of material per millilitre of solvent, .it is probably clean enough for capillary CC/MS. If it contains more materials, it will likely require additional preparation. In most instances, son e type of cleanup technique will be necessary in order to achieve suitably low detection limits for the target compounds. If much aliphatic material exists in the sample it will mask the compounds of interest. Mere dilution will not remedy the situation as detection limits are raised by the dilution. If acidic compounds such as phenols are suspected of being present in the sample, a separate fraction containing these acids can be created using the organic extract obtained above. Methods 3530, a base/neutral acid cleanup extraction technique, may be applicable to the cleanup of certain sample types. Modifications to Methods 3530 are as follows: ------- a) In Section 7.6, the organic and aqueous phases are both treated as containing compounds; and b) Section 7.15 will not be necessary. The aqueous phase, when transferred to organic solvent after Section 7.13, will contain acidic compounds. The orç,anic phase contains basic and neutral compounds. In most instances, the acidic fraction will be clean enough for CC/MS ar.alysis. The base/neutral extract, however, may require further cleanup. Thus, a cleanup procedure ha been devised for base/neutral extracts that minimizes the interferences caused by high concentrations of aliphatic and polymeric materials. Although the cleanup procedure is thoroughly described in the next section, one generally proceeds as follows. The sample is subjected to cieanup by placing a representative aliquot of the sample on an al 9 mina column and successively eluting with hexane, methylene chloride, and diethyl ether to yield 3 fractions containing the aliphatic (hexane fraction), aromatic (methylene chloride fraction) and polar compounds (ether fraction). The methylene cloride fraction is then concentrated to about 1 ml. and then is analyzed by CC/MS for the compounds of Interest. The hexane concentrate can be screened by GC/MS to determine if compounds were eluted into the hexane fraction. However, this usually will not be required. If polar compounds are of interest, the ether fraction is also analyzed. Quantitation of the semivo]atile constituents in Attachment 1 is to be performed using the reverse search technique. Additionally, tentative identification should be attempted for the ten organic compounds detected at the highest concen- trations. Identifications should be made via a forward search of the EPA/NIH mass spectral library. Concentrations should be approximated by comparison of the compound response to that of the closest eluted internal standard. A procedural blank, matrix spike, and duplicate should be analyzed for every batch of samples. Accuracy and precision control charts should be maintained for indicator constituents. The percent recoveries of spiked surrogate standards for a given sample type should be plotted versus sample identification number. Table 1 contains a list of the surrogate compounds to be employed for the analysis of semivolatile organic compounds, and recovery limits. Recovery limits are based upon obtaining a final extract sufficiently clean, such that the surrogate compounds should be present at 50 ppm or higher in the extract. If dilution of the sample is still required, detection of the surrogates may be difficult and the associated recoveries imprecise or non—existant. Such samples should be spiked with higher surrogate levels and resubjected to the cleanup procedure. ------- Table 1. Surrogate Standards for Semivolatile Organic Compound Analysis Recovery Limits Acid surrogates phenol—d 5 40—115% 2—f luorophenol 2,4 ,6—tribrornophenol Base/neutral surrogates nit robenzene—d 5 5—fluorobiphenyl 50—120% terpheny l—d14 acridine—d 9 pyrene—d 10 The precision control chart should consist of the percent difference for indicator constituent concentration determined in duplicate samples of a given sample type versus sample identification numbers. Column Clean Up of Petroleum Wastes Scope and Application This method is used to cleanup samples containing high levels of aliphatic hydrocarbons, such as wastes from petroleum refining. It is used specifically to separate aliphatics, aromatics, and polar compounds in the waste matrix. This method is applicable to API separator sludges, rag oils, slop oil emulsion, and other oily wastes derived from petroleum refining. This method is recommended for use only by or ; under close supervision of experienced analysts. Summary of Method Take a 200 mg aliquot of the waste/methylene chloride concentrate from step 7.13 of Method 3530. Dissolve the aliquot in r exane and spike with 10mg each of d 9 —acridine, d 5 —nitrobenzene, d 5 —phenol, 2—fluorobiphenyl, tribromophenol, d14—terphenyl, 2—fluorophenol, and dlO—pyrene. Apply the mixture directly to the alumina column. The column is eluted sequentially with hexane, methylene chloride, and diethyl either and the corresponding three fractions are collected. An aliquot of the CH 2 C I 2 fraction is evaporated under a gentle stream of nitrogen and weighed to determine the appropriate concentration factors prior to ------- GC/MS. If pyrene or terphenyl is recovered at.less than 50%, the procedure should be repeated. Interferences Matrix interferences will likely be coextracted from the sample. The extent of these interferences will vary considerably from waste to waste depending on the nature and diversity of the particular waste being analyzed. The use of additional cleanup extractions can be used as necessary for specific compound identification and quantitation. Apparatus Glass Column: 30 cm long x 1 cm I.D. with glass frit or glas5 wool and stop clock. Aluminum weighing boats: Approximately 2 in. in diameter. Analytical Balance: Capable of weighing to +0.5 mg. Concentrator Tube, KD, 10 ml Evaporative Flask, KD, 250 ml Snyder Column, KD, three—ball micro Snyder Column, KD, two—ball micro Steam Bath Boiling Chips: 10—40 mesh carbarundum. Heat to 450°C for 5— 10 hours. Syringe: 1 ml glass 50 ml beaker 250 ml beaker Reagents Hexane: Distilled in glass (B&J) or equivalent Methylene Chloride: Distilled In glass (B&3) or equivalent Diethyl Ether: Distilled in glass (B&J) or equivalent Alumina: Dried overnight at 130°C, neutral 80—325 MCB chromatographic grade Sodium Sulfate: Washed with CH 2 C1 2 and heated to 150°c for 4 hours ------- Procedure weigh Out 10.0 gm of alumina and add to the chromatographic column that is filled to about 20 rnL with hexane. Allow the alumina to settle and then add 0.5 gm sodium sulfate. Let the solvent flow such that the head of liquid in the column is ahout 1 cm above the sodium sulfate layer. Stop the flow. Add the aliquot equivalent to 100—200 ing of material. Start the flow and elute with 13 ml of hexane. Collect the effluent in a 50 beaker. Label this fraction ualiphaticsfl Elute the column with 100 ml of methylene chloride and collect the effluent in a 250 ñ l beaker. Label Waromaticsll Elute the column with 100 ml of diethyl ether and collect the effluent in a 250 ml beaker. Label wpolarsw. Weigh three sample boats to the nearest 0.5 mg. Reduce the volume of each fraction using the KDs to between 1 and 5 ml. Record the volume of each and place 1/2 of each sample in the respective boat. Evaporate the liquid in each boat under a gentle stream of nitrogen. Reweigh each boat and record the weight of each fraction. Calculate the weight of each fraction as a proportion of the total sample. For example, fraction I is 56.3 mg, fraction 2 Is 25.4 ing, and fraction 3 is 85.0 mg. Calculate the amount of sample in the fractions and adjust the volumes so in)ection will permit determination of various components on scale 12.7 mg/2500 ul 5.1 ug/ul Dilute each of the three fractions obtained by a ratio so that the sample entering the capillary column does not exceed 2.5 ug. For example, if the calculated weight of the fraction as a proportion of the total sample is 12.7, and the amount of sample in the fractions is 5.1 ug/ul as in the above example, dilute the sample 1:1 with methylene chloride. Quality Control Before processing any samples, the analyst should demonstrate through the analysis of a distilled water method blank that all glassware and reagents are interference—free. Each time a set of samples is extracted or there is a change In reagents. a method blank should be processed as a safeguard against ------- chronic laboratory contamination. The blank sample should be carried through all stages of the sample preparation and measure- ment. Standard quality assurance practices should be used with this method. Laboratory replicates 3hould be analyzed to validate the precision of the analysis. Fortified samples should be carried through all stages of sample preparation and measurement; they should be anlayzed to validate the sensitivity and accuracy of the analysis. ------- SKINNER LIST Metals Semivolatile Base/Neutral Extractable ComoourlCS Antimony Arsenic Anthracefle Barium Berizo (a) arnhracene Beryllium Berizo (b) fluorarnhene Cadmium Benzo (k) fluoranthene Chromium Benzo (a) pyrene Cobalt Bus (2-ethylhexyl) phthalate Lead Butyl benzyl phthalate Mercury Chrysene ‘J uckel Diberiz (aM) acridine Selenium Dibenz (aM) anthracene Vanadium DichlorobenZefles Diethyl phthalate loI atijg 7,1 2.Dimethylbenz (a) anthracene 3enzene Dimethyl phthal ate Carbon disulfide Di (n) butyl phthalate ChIorobenZer e Di (n) octyl phthalate Chloroform Fluoranthefle 1 2.Dichloroethafle Indene 4-Dioxane Methyl chryserie Ethyl benzene 1-Methyl naphthalene Ethylene dibromide Naphthalene Methyl ethyl ketone Phenanthrer ie Styrene Pyrene Toluene Pyridine Xylene Quinoline Semivolattle Acid- Extractable Comooufl Benzer iethiOl Cresols 2.4-DumethylPheflol 2 .4 DinitrOPheflOl 4-Ni trophenol Phenol ------- Office of Research and Development Office of Solid Waste and Emergency Response EPAI54O/4-9 1/O0i February 1991 P EPA Ground-Water Issue SOIL SAMPLING AND ANALYSIS FOR VOLATILE ORGANIC COMPOUNDS 1. E. Lewis, A. B. Crockett, R. L Siegrist, and K. ZaITabi The RegionaJ Superfund Ground Water Fo- rum is a group of ground-water scientista that represents EPA’s Regional Superfund Of- fices The forum was organized to exchange up-to-date information related to ground- water remediation at Superfund sites. Sam- pling of soils for volatile organic compounds (VOCs) is an issue identified by the Ground Water Forum as a conoem of Superfurid de- osion makers. A group of saentis actively engaged in method development research on soil sam- pling and analysis for VOCs gathered at the Enwonmentai Monrtonng Systems Labora- tory in Las Vegas to examine this issue. Members of the committee were R. E. Cameron (LESC). A. B. Crockett (EG&G), C. L Geriach (LESC), T. E. Lewis (LESC), M. P. Maskarinec (ORNL), B J. Mason (ERC), C. L Mayer (LESC), C Ramsey (NEIC), S. R. Schroedl (LESC), R L Siegnst (ORNL), C. G. Urchin (Rutgers University). L. G. Wilson (University of Arizona), and K. Zarrabi (ERC). This paper was prepared by The Committee for EMSL. LV’s Monitoring and Site Charactenzation Techmcal Support Center, under the direction of T. E. Lewis, with the support of the Superfund Technical Support Project For further informaton contact Ken Brown, Center Director at EMSL-LV, FTS 545-2270, or T. E. Lewis at (702) 734-3400. PURPOSE AND SCOPE Concerns over data quality have raised many questions related to sampling soils for VOCs. This paper was prepared in response to some of these questions and concems expressed by Remedial Proiect Managers (RPMs) and On-Scene Coordinators (OSCs). The follow- ing questions are frequently asked: 1 Is there a specific device suggested for sampling soils for VOCs? 2. Are there significant losses of VOCS when transfernng a soil sample from a sampling device (e.g., split spoon) into the sample container? 3. What is the best method for gettIng the sample from the split spoon (or other device) into the sample container? 4. Are there smaller devices such as suboore samplers available for coIle ing aliquots from the Larger core and effi- ciently transfernng the sample into the sample container? 5. Are certain containers better than others for shipping and storing soul samples for VOC analysis? 6. Are there any reliable rleservaxn proce- dures for reducing VOC losses from sod samples and for eztendng holdIng tImes? Thle p er La intended to famdlartze RPMs, OSCa, and field personnel with the ainern stale of the sdence and the current thinking concerning sampling soule for VOC analysis. Guidance is provided for seledlng the most effectIve sampling devIce for collectIng Superfisid T.chnology Support Center for Uonftortng and Site C*wactertratlon Envfronm.ntal MonItoring Systems Laboratory Las Vegas, NV United States Environmentaj Protection Agency /\ = Pnritso on R*,, ’rled Paper ------- samples from soil mati ’IceS. The techniques for sample collec- tion, sample han ifl9. oontelneflDflg, SUPfl Oflt, and storage desa’.bed paper reduce VOC losses and generally provide more repreSefl UVe samples for volatile organic analy- ses (VOA) than techniques in current use. For a discussion on the proper use of sampling equipment the reader should refer to other sources (Adiar, 1974; U.S. EPA. 1983: U.S. EPA. 1986a). Sod, as referred to in this report, encompasses the mess (surf and &Esurt ) of unconsolidated mantle of weath- ered rod and loose matenai tying above solid mdc. Further, a distinction must be made as to what fr on of the unconsoli- dated material is soil and what fra ion is not. The soil compo- nent here is defined as all minerai and naturally o .zmng organic matenai that is 2 mm or less in we. This is the size normally used to differentiate between soils (consisting of sands, ailta, and days) and gravels. Althou i numerous sampling situations may be encountered, this paper focuses on three broad categories of sites that might be sampled for VOCs: 1. en test pit orU’endi 2. Swfane soils (c5ftin depth) 3. Subsurface soils (>5 ft in depth) INTRODUCTION VOCs are the dass of compounds most commonly encoun- tered at Superfund and other hazardous waste sites (M oy, 1985; Plumb and Pjtchford, 1985; Plumb, 1987; Ameth at al., 1988). Table 1 ranks the compounds most commonly encoun- tered at Superfund sites. Many VOCs are considered hazard- ous because they are mutagensc, carcinogenic. or teratogenic, arid they are commonly the controlling contaminants in site restoration protects. Decisions regarding the extent of contaim- nation and the degree of deanup have far-reaching effects; therefore, it is essential that they be based on urate mea- surements of the VOC concentrations present VOCs. how- ever, present sampling, sample handling, and analytical diffi- culties, especially when encountered in sods and other solid matrices. Methods used for sampling soils for volatile orgaruc analysis (VOA) vary widely within and between EPA Regions, and the recovery of VOCs from soils has been highly variable. The source of variation in analyte recovery may be associated with any single step in the process or all steps, induding sample collection, transfer from the sampling device to the sample container, sample stuprnent. sample preparation for analysis, and sample analysis. The strength of the sampling chain is only as strong as rts weakest link sod sampling and transfer to the container are often fl e weakest links. Sample colle on and handling activities have large sources of random and systematic errors compared to the analysis itself (Barcelona, 1989). Negative bias (i.e., measured value less than true value) is perhaps the most significant and most dtffiaitt to delineate and control. This error is caused primarily by loss through volatlhzation dunng soil sample collection, storage, and handling. TABLE 1. RAM(BIG OF GROUP WA1 CONTA 4ANTS BASED ON FREQU9 ICY OF DETECTiON AT HAZARDOUS WASTE I 9 5 Con te Uo Tnddo m e f iene(V) 51.3 Tead oro ’iene (V) 36.0 12-Uw d methene(V) 29.1 ct &4m( ) 284 1,1- il c ioethene(V) 252 Meth 4enedLndo(V) . 192 1,1,1-Tii . ie(V) 18.9 1,1- ue(hane (V) 17.9 12-Do e fw (V) 142 Phenol (A) 13.6 Aoib ie(V) 12.4 ToIiene(V) 11.6 b i s-(2- y1)pttha l (B) 113 Beraene(V) 11.2 Wi 1 ói 8.7 V - e, A • d . J , B • Sciz PUTt d P al (198 There are currently no standard procedures for sampling souls for VOC analyses. Several types of samplers are available for collecting in t (undisturbed) samples and bulk (disturbed) samples. The selection of a particular device is site-specific. Samples are usually removed from the sampler and we placed in glass tars or vials that are then sealed with Teflon-lined caps. Practical experience and recent field and laboratory research, however, suggest that procedures such as these may lead to significant VOC losses (losses that would affect the utility of the data). Hanisch and McDevitt (1984) reported that any heac [ cpece present in the sample container will lead to desorp- tion of VOCs from the soil partides into the hear peoe and wifl cause loss of VOCs upon opening of the container. Siegnst and Jennsen (1990) found that 81% of the VOCs were lost from samples containerized in glass tars sealed with Teflon-lined caps compared to samples immersed in methanol in ars. FACTORS AFFECTING VOC RETENTION AND CONCENTRATION IN SOIL SYSTEMS Volatile organic compounds in sod may coexist in three phases: gaseous, liquid (dissolved), and solid (sorbed). (Note: Sorbed is used throughout this paper to encompass physical and chemical adsorption and phase partitioning.] The sampling, identification, and quantitation of VOCs in soil matrices are complicated because VOC molecules can coexist in these 2 ------- — -. three phases The interactions between these phases are illustrated in Figure 1 The phase distribution IS Controlled by VOC physicocherrilCal properties (e g., solubilty, Henrys constant), soil properties, and environmental vanables (e g, soil temperature, water content, organic carbon content) 0l a- I Oi ( 1)1 I. Figure 1 Equulubnum relationships for phase partitioning of VOCs in soil systems See Table 2 for definitions of abbreviations The factors that affect the concentration and retention of VOCs in soils can be divided into five categones: VOC chemical properties, soil chemical properties, sod physical properties, envirorinientaj factors, and biological factors. A bnef summary of VOC, soil, and environmental factors is presented in Table 2, which provides an overview of the factors that interact to control VOCs in the sod environment at the time a sample is collected The cited references provide a more detailed discussion The chemical and physical properties of selected VOCs are further descnbed in Table 3 Note that many of these properties have been determined in the laboratory under conditions (e.g, temperature, pressure) that may differ from those encountered in the field Devm et aJ. (1 987) offers a more exhaustive list. Many VOCs exhibit extreme mobilrties, particulaily in the vapor phase, where their gas diffusion coefficients can be four times greater than their liquid drffusion coefficients. The vapor phase migration is influenced by the moisture content of the soil which altars the air-filled to water-filled pore volume ratio. The reten- tion of VOCs by sod is largely controlled by reactions with the solid phase. This retention is especially true for the finer particles of silts and days The fine-grained particles (. mm) have a large surface-to-volume ratio, a large number of reactive sites, and high sorption capacities (Richardson and Epstein, 1971; Boucher and Lee, 1972, L.otse et aL, 1968). Some investigators athibute the greater sorption of VOCs onto fine- grained particles to the greater organic carbon content of smaller particles (Karickhoff at aJ., 1979). Soil-moisture content affects the relative contnbudons of min- eral and organic soil fractions to the retention of VOCs (Smith at aJ. . 1990) Mineral day surfaces Largely control sorption when soil moisture is extremely low (<1%), and organic carbon (Conbnued on page 7) TABLE 2. FACTORS AFFEC1tNG VOC CONC94TRATIONS l SOILS Fader Common Ab&. Unite Eflects on VOC Con ratIons i So flM.. ,.. VOC Chen cat Prop.rU.e So l ty C, mg/i. Affects fate arid transport m wa , efte wate a%p , fiuences o thn pa Roy arid Gil i (1985) Heniys Constant K (atimm )mde Constant of proponiona ty bmween the water a pf n aI ,ns; tenVec se Mid presase depondart Shin ind Sewal (19 ) Spenca . (1988) Vapor prassiae v.p. rim Hg AI rate of k,es from soi. Shin wid Sewal (1 9 ) Organic carbon pall. e& K , mg VOCI 9 C AdaorpLn coef&ient noni r sc ix Fwnw’ t (1980) OdenoL water part. eff. K,, n VOC/ mg o E iitimr r m for Wt on of VOC t in w r aid in orgar (odenol) pfl e. Giiiies atinitI of w organic frwton of L Voi Wehir (1983) . Bofng pou b.p. C C Aib -evaporabon of VOC aid water from aol aw . Voi id Wilier (1983) SoiL’water i str on coethoent Kd [ 11 Ec aberi ainsiaru for b on of airta 1 betuwen d arid U id phise& Vo end Wilier (1983) (Conthiueo9 PHASE 3 ------- TABLE 2. (CONTVIUED) o n Fsclor . OIS On VOC Co .i Mi i Ii Sal Sal Q u uq4rOu EC meqtl Dog the rwlet Of neg y dwged s s on si — wf ere dwped VOC may soib; pH de 1 Jert bi ronsor n pH -1c W] rfiuensos a rar er of sod pro ft ( vtty) nonneu orgerec p UUur 9 ; CftOatS CEC end r4n4 ty of some VOCs Tood OI9W1C sobon ‘d8flt •TOC tT C d J’l fl O( 1t P6ItoOflIIJQ me6tin b non-pdar hy&ophob 1 c Cha u at ai. (1988) ( Pu K , ) VOCs; sorpoon Of VOCe ii V ma cun may be Feniw at ai. (1980) I .• , re . Sal Phy uperUe . P 1 e we or sextire A % and, AI sif babon, poiiatr l , r rthon , sotpbon, end F ofw n and oft, day m ty of VOC6. flJeflsos hydraui as wod as es - Epstein (1971) eree-to . &rne ratio (s.&.cKd) s a. m 1 /g Me e soption of VOCe from v or pI se; aI soil Ke fl 61 aJ. (1979) rosdy end her te r propa es &ik der ty p g L ed r e una g mob dy and retention of VOCa ‘n sods, encer at aJ. (1988) w odtuenso sod sanUmg deviso seIe on Porosity n Void vkmne to total ‘ di.sne r o. Masts vofluiie, Femeret ai. (1980) 1dLdburi, ivt i , and rmgraflon of VOCs &n sod .vsds Shan end Se fI (19 ) Peros i mo6tnse 0 % (w/w) A1$e hy*aniuc sondu vnty of sod and sorpbon of VOCs. Fexner at af (1980) Def imt s the daso& n and mobibty of VOCs v sod Ctnmu w SP (1985) Water pstar nai p F m Relates to the rate, n ildy, and anoeratation of VOCs rw ror nidd w n l a. HØailic ron cMy K m d Masts .isson ow of VOCs rn sod water dapen ng on - degree of satiration, gr 6r , and other physisoi te rs En*o F. s Re uiidñy RH. vend Stu .ç (1985) ConM d the imyemer , and of Ten eran.we T cc VOCe; i*e. ,lat ,d s; be site speak end depexMnt ___ icon si s 1we u rf a 6fterariiafs 8a UK pre $e irni - Relevant to speed, nmemerl, and of VOCe remo d, or dth nng from sirtaso t ne y, r se, end bid, arid datb on of attest n ment, th on ra , end OOAooilribon Of VOCs 4 ------- TABLE 3. CHEMICAL PROPER11ES OF SELECTED VOLATiLE ORGANiC COMPOuplDst From VI be 19 , .My 1964 • 0 no *. r Qiut i o it • —I L n ra 20 . m.w. Compound (glmo le) SokLbilrdes (mg /i. @ 20°C) log K Vapor P euwe log K K,; (mm @ 20°C) Benze e 78 1780 -0.24 270 (@30°) Broniodichioramethane 164 191 211 0.22 76 Bromotorm 253 7500 3190 2.18 210 50 Bromomethajie 95 (@30°) 6 (@25°) 2-Butarione 72 900 270000 1 34 119 1 50 1250 Carbon disul de 76 2300 1 56 026 76 Carbon tetrachlonde 154 800 1 80 260 Chlorobenzene 113 500 2 04 2.64 094 90 Chloroethaj - ie 65 5740 218 1 40 284 016 9 2 Chloroethytviny1 ether 107 1.54 061 1000 Ch omtcrm 120 8000 Ch oromethane 51 8348 1 46 1 97 012 160 Dibromoch oromet,ane 208 3300 0.78 0 91 162 3800 l,2-Dichlorobenzene 147 100 245 2.24 15 (@105°) 1 3-Dc orobenzene 147 123 2.62 3.38 1 l4-Dichlorobenzene 147 (@25°) 49 3.38 1,1-Dichioroethane 99 (@22°) 5500 3.39 1 1 ,2 .Dichdomethane 99 8690 1 66 1 79 018 180 1,1 -Di oroethene 97 134 1.48 004 61 traris-1,2-Dsch loroethene 97 600 500 1.2-D ichloropropane 113 2700 1 56 2.06 200 (@14°) l3D ropcopene 110 2700 1.99 42 trans-1.3.-DichIorop -oper ie 111 2800 34 (@25°) Ethylbenzene 106 152 43 (@25°) 2-Hexanone 100 3500 2.60 3.15 7 Methyfene onde 85 20000 138 2 Methyflscbutylketone 100 17000 140 125 349 Perth roethyiene 166 150 1 34 1.46 0002 6 Styrene 104 300 2.60 2.60 0.85 14 11,22 .Tetr hloroethape 168 2.61 2.95 5 Teoroethene 166 2.07 2.60 5 Toluene 92 515 2.78 340 18 (@25°) 1,1,1 -Tnchloroethane 133 2.18 2.69 027 22 1,1 ,2-Tnchloroethane 133 4500 2.19 2.50 1 48 100 Tndlloroethyfene 132 700 2.14 2.07 19 Tnfiuoromethane 137 2.09 2.29 0.37 60 Vin y f acetate Vinyf cthionde 86 63 (@25°) 25000 1100 (@25°) 2.68 1.59 2.60 0.73 687 115 (@25°) - Total xy lenes 106 196 1.38 97.0 2660 (@25°) 5 ------- Various sormaobeS tev’a-, U ,-, di-, arid m-Chlorophenol (KobayasSu and RstUiiar i, 1982) 2,6-, 2,3Oid orobenzene. 2,4- and 2, tilorobenzene, CO 2 (Kobayash and R n, 1982) Redu ive dehafogenation under anoxc ndeons, (I e , < 0.35 V) (Xobayastv and Rittinan, 1982) Fungi Pset.ó merias sp Ac inetoba efsp Mrry ijc sp Acetate-grown biofitm Anaerobic Redu ive dehalogenaflcn to mth ene,d oroethene, arid — d ’ onde, and finally CO 2 (Vogel and McCaily, 1985) An robc aloger on to I ,2 .dithforoethene and not 1,1 -dic omethene (lGeop retal,1985) Aerobic *wai ed to CO 2 in the presenro of a mutsa of natsal gas and air o b ic Various particle breakdown products rrvneraDzed by other morgan i (Lechevaiier and Led aJier, 1976) Conlilete nineraiization in 10-14 days (JoN sen, 1976) Oi’garusms were cepable of sustaining growth in these compounds with 100%bodegradatlon (Jamison olaf., 1975) TABLE 4. MICROBiOLOGICAL FACTORS AFFECTING VOC IN SOIL SYSTEMS Compoun s) CoMtlons Pentath ropher i ol Aerobic 12,3- and 1 2,4-Tnchlocobenzene Aerobic Vanous soil bactena Anaerobic Tnth roethane, tnchbromelhar ie, meth 1c onde, th i ’oethane, dlthloroeihane. vin lxliene th nde, U, UoWethene, oroethene. meth 1ene thknde, Various soil miaobes TeU doroe1hene Various so miaobes ‘ C-4abe4ed U,chloroethene Various sod b ena Tnd ,broethene Adinon ycetes chionnated and non-d ,k,nnated Aerobic Aerobic DDT kom Chkmnaled afiphalics Ci-donnated and nonchbnnated aromatics Aerobic Methanogenic Aerobic Methanogenic Aerobic Blue-green algae Oil wastes (cyariobactena) No bodegr on deser wd (Boirwer, 1984) Nearly 100% biodegradation obsemed (Boirwer. 1984) Nearly 100% biodegradation (Boijwer, 1984) No biodegr on observed (Bot er, 1984) BiodegradaUon of automobile oil wastes, ank se oil, etc (Cameron, 1963) 6 ------- partitioning is favored when moisture content is higher (Chiou and Shoup. 1985). Biological factors affecting VOC retention in soil systems can be divided into microbiological and macrobiological factors. On the microbiological level, the indigenous microbial populations present in soil systems can alter VOC concentrations. Although plants and animals metabolize a diversity of chemicals, the ac1ivitie of the higher organisms are often minor compared to the transformations affected by heterotrophic bactena and fungi residing in the same habitat. The interactions between environ- mental factors, such as dissolved oxygen, oxidation-reduction potential (Eh), temperature, pH, availability of other compounds, salinity, particulate matter, and competing organisms, often control biodegradation The physical and chemical charactens- tics of the VOC, such as solubility, volatility, hydrophobicity, and K . also influence the ability of the compound to biodegrade. Table 4 illustrates some examples of the microbiological alter- ations of some commonly encountered soil VOCs. In general, the halogenated alkanes arid alkenes are metabolized by soil microbes under anaerobic conditions (Kobayashi and Rrttman, 1982, Bouwer, 1984), whereas the halogenated aromatJcs are metabolized underaerobicconditions To avoid biodegradation and oxidation of VOCs in soils, scientists at the U.S. EPA Robert S. Kerr Environmental Research Laboratory in Ada, OK, extrude the sample in a glove box. On a macro scale, biological factors can influence the migration of VOCs in the saturated, vadose, and surface zones (Table 5). BiollIms may accumulate in the saturated zone and may biode- grade and bioaccumulate VOCs from the ground water. The biofilm. depending on its thidcness, r iay impede ground-water flow Plant roots have a complex mucrofiora associated with TABLE 5. MACROBIOLOQCAL FACTORS A ECTPIG VOCs (P4 S L SYSTEMS ,-. Zone -1- B filrr Sat rated Biode alation, bio jn4ation, more or less c a, parent — , thsk frioã n y rete sat:ated few Plant roi CapBary fringe vatose fi.m moy bode ade or bsoa aniia VOC. root d rele may sane as ler VOCm Animal bwmws holes Vadose May as aliy pail for aid n onofsafwe aid sane as mn&utler .VWWd VOC nbg n Vegeteti ve So s Save as bamer tI fvwi from aid reted inMr kAI of ast spas them known as mycorrhizae. The mycorlThzae may enhance VOC retention in the soil by biodegradation or oaccumulation The root channels may act as condurte for increasing the migration of VOCs through the soil. Similarly, animal burrows and holes may serve as paths of least resistance for the movement of VOCs through soil. These holes may range from capillary-size openings, created by worms and nematodes, to large-diameter tunnels excavated by burrowing animals. These openings may increase the depth to which surface spills pen- etrate the soil. A surface covenng consisting of assorted vegeta- tion is a signrficant barrier to volatilization of VOCs into the atmosphere. Some ground-water and vadose-zone models (e g., RUSTIC) undude subroutines to account for a vegetative cover (Dean et aL, 1989). SOIL SAMPUNG AND ANALYSIS DESIGN Pnor to any sampling effort, the RPM or OSC must es blush the intended purpose of the remedial invesbganonJfeasi lrty study (RVFS) The goals of collecting samples for VOA may uridude source identification, spill delineation, fate and transport, nsk assessment, enforcement, remediat,on, or post-rernedlation confirmation. The intended purpose of the sampl i n g effortdnves the selection of the appcopnate sampling approach and the devices to be used in the investigation. The phase part itioning of the VOC can also inNuence which sampling device should be employed. Computer models gener- ally are used only at the final stages of a RVFS. However, modeling techniques can be used throughout the RL/FS process to assLst in samplmg device selection by estimating the phase parimoning of VOCs. The RPM is the primasy data userfor a AL’ FS led by a federal agency. As such, the RPM must se4e the sampling methodology to be employed at the site. Figure 2 illustrates the sequence of events used to plan aVOC sarr ltng and analysis activity. The domains of Interest also must be determIned. The target domains may indude surface (two dimensions) or subsurface (three dimensions) enwonments, hot apota, a concentration greater or less than as, on lImit, or the area above a lealong underground storage tank. Statistics that may be generated from the target domái date must be considered before a sample and analysis desigo Is developed. Poesdle statistics of interest may undude average_analyte concentration and the variance about the mean ( statistics that coni ie whether the observed level is signrficsntiy above or below an ellen level) as wed as temporal and sp l4al trends Data must be of sulfioentiy high quality to meet the goals of the seniØnq elMty. The level of date quality is defined by the date quality ob4e vee (DOOs). In RVFS vTtls$, sites are so Jffo nt and Information on overall measurement error (samplIng plus analydoal error) Is so limited that It Is not practical to set universal or generic precelon, uracy, representathieness, completeness, arid comparabil- ity (PARCC) goals The reeder Is referred to a use e grthde on quality assurance In aoü samping (Barth at aL, 1989) aid a gu ance doaiment for the development of data qua ty ob eo- fives for remedial response activities (US. EPA, 1987). 000s are qualitative arid quantitabve mente of the level of uncertainty a decision malcer is wi ng to pl in malW’.g decisions on the basis of environmental date. ft Is liT oita1t to realize that if the error ciatad with the sample coI on or 7 ------- “ Confidence level • Bias • Preasion • Action level • Analyte level DESIGN CHARACTERIZE SITE • History • Process • Soil properties • Soil conditions • E dsting data • Environmentai factors I SE LE C.T TOOLS. • Spilt spoon • Piston samplers • Zero contam. sampler • Shelby tube • Veihrneyer tube • Shovels * DRAFT S&A PLAN . . • S S . S S S Tools AnalyncaJ methods Holding times No. of samples Sample mass Decontamination QNQC Field analysis Handling Random] systematic design Figure 2. Flowchart for planning arid implementation of a sod sampling and analysis activity DEFINE- GOAL • Soil population • Location • cs •Trend • x, Std. dev. • Comparison • Purpose • Enforcement • Remediation • Source ID + ANALYTE OFINTERESr SELECT METHODS SET 000s - • GC/MS •GC • FieldGC • Methanol extraction I NO MINIMIZE: RESOURCES: MAXIMIZE: INFORMATION MAXIMIZE QUALITY FEASIBLE: V YES V • Personnel • Budget • Time • Politics I: ON•SITE_DATA _____ • Field analysis • Visual observations •Odors • Population esibdrty FIELD. IMPLEMENTATION DATA EVALUATION + I OBJECTIVE: MET NO 8 ------- preparation step is large then the best laboratory quality assurance prOgraJll will be inadequate (van Ee et aL, 1990). The greatest emphasis should be placed on the phase that contributes the largest component of error. For the analysis of soils for VOCs, the greatest sources of error are the sample collection and handling phases. The minimum confidence level (CL) required to make a decision from the data is defined by the OQOs. The minimum CL depends on the precision and accuracy in sampling and analysts and on the relative analyte concentration. Relative error may be reduced by increasing either the number or the mass of the samples to be analyzed. For instance, although 5-g aliquots collected in the field might exhibit unacceptable errors, 1 00-g samples will yield smaller errors and might therefore meet study or project requirements. Corn positing soil samples in methanol in the field also can reduce variance by attenuating short-range spatiai vanability. Field sampling personnel should coordinate with laboratory analysts to ensure that samples of a size appropnate to the anal yt icai method are collected. For example, lithe laboratory procedure for preparing aiiquots calls for removing a 5-9 aiiquot from a 1 25-mL wide-mouth jar, as per SW-846, Method 8240 (U.S EPA I 986b), then collecting a larger sample in the field will not reduce total measurement error, because addi- tional errors will be contributed from opening the container in the laboratory and from subsequent homogenization. Aiiquoting ofa5-g samplein thefield intoa40-rnL VOAvial that can be directly attached to the laboratory purge-and-trap unit signrficantly reduces loss of VOCs from the sample (U.S. EPA. 1991a). Significant losses of VOCs were observed when samples were homogenized as per Method 8240 speafica- tions. Smaller losses were observed for smaller aliquots (1 to 5g) placed in 40-mL VOA vials that had modified caps that allowed direct attachment to the purge-and-trap device. The procedure of collecting an aliquot in the field eliminates the need for sample preparation and eliminates subsequent VOC loss in the laboratory Field-screening procedures are gaining recognition as an effective means of locating sampling locations and obtaining real-time data. The benefits of soil field-screening procedures are’ (1) near real-time data to guide sampling activities, (2) concentration of Contract Laboratory Program (CLP) sample collection in crTtIcaJ areas, (3) reduced need fora second v sst to the site, and (4) reduced analytical load on the laboratory. Umitations of field-screening procedures are: (1) a peon knowledge of VOCs present at the site is needed to aco.irately identify the compounds, (2) methodologies and instruments are in their infancy and procedures for their use are not well documented arid (3) a more stringent level of quality assur- ance and quality control (QNQC) must be employed to ensure accurate and precise measurements. The potential benefits and limitations associated with soil-screening procedures must be carefully weighed and compared to the DOOs. Certain sampling and analytical methods have inherenUimrta- tions on the type of QNQC thai is applicable. For example, splitting sod samples in the field would not be appropnate for VOA due to excessive analyte loss. The higher the minimum CL needed to make a decision, the more ngorous the QAIOC protocols must be. As VOC concentrations in the soil sample approach the action or detection limit, the quantity and Ire- quency of QA/QC samples must be increased, or the number of samples must be increased, to ensure that the data quality obtained is appropriate to satisfy project objectives. One critical element in VOC analysis is the appropnale use of trip blanks, if a sample consists of a silty day loam, a hip blank of washed sand may not be realistic, for such a blank would not retain VOC cross contaminants in the same way as the sample. The trip blank soil matrix should have a sorptive capacity similar to the actual sample. In additIon, high- concentration and low-concentration samples should be shipped in separate coolers. DEViCE SELEC11OI4 CRITERIA The selection of a sampling device and sampling procedures requires the consideration of many factors induding the number of samples to be collected, available funds, soil charactensiics, site limitations, ability to sample the target domain, whether or not screening procedures are to be used, the size of sample needed, and the required precision and accuracy as given in the DOOs. The number of samples to be collected can greatly affect sam- pling costs and the time required to complete a site cflaractenza- lion. If many subsurface samples are needed, it may be possible to use sod-gas sampling coupled with on-site analysis as an integrated screening technique to reduce the area of interest and thus the number of samples needed. Such a sampling approach may be appli le for cases of near-surface centamlietion. Ultimately, the sampling, sample handling, containertong, and transpoit of the sod sample should minimize losses of volatiles and should avoid contamination of the sample. Soil sampling equipment should be readily decontaminated in the field if it is to be reused on the job site. Decontamination of sampling equ- ment may require the use of decontamination pads that have impervious liners, wash and rinse troughs, and careful harding of large equipment Whenever possible, a liner should be used inslde the sampling device to reduce potential cross coritamina- hon and carryover. Decontamination procedures take time, require extra equipment, and ultimately increase site character- ization costs. Ease and cost of decontamination are thus unpor- tent factors to be considered in device selection. Several soil-screening procedures are In use that indude he r4epaoe analysis of soils using organic vapor analyzers: w r (or NaCI-satiJrated water) extraction of soil, followed by static head pene analysis using an organic vapor analyzer (OVA) or gas thromatograph (GC); cclodmetrfc test 1db; methanol exti hon followed by he cpar* analysis or direct injection Into a GC; and sod-gas sampling (U.S. EPA. 1988). FIeld measurements may nat provide absolute values but often may be a superior means olobtairüng relative values. Thee. proosduree are gain- ing acceptenos. Sit. CharacWlatlcs The remoteness of a site and the physical setting may restrict access and, therefore, affect equipment sele on. Such f rs as vegetation, steep slopes, rugged or rocicy ten’aln, overhead power lines or other overhead restrictions, and l of roads can conb ute to esa problems. The presence of underground utilities, pipes, electrical lines, tanks and leech fields can also affect selection of sampling 9 ------- equipment. tf the location or absence of these hazards cannot be established, it is desirable to conduct a nonirmusive survey of the area and select a sampling approach that minimizes hazards. For example, hand toots and a baclthoe are more practicaJ under such arcumstances than a Large, hollow-stem auger. The selection of a Sampling device may be influenced by other contaminants of interest such as pesticides, metals, semivolatiie organic compounds, radionudides, and explo- sives. Where the site history indcaies that the matiix is other than soil, special consideration should be given to device se4e on. Concrete, reinforcement bars, scrap metaJ, and lum- ber will affect sampling device selection. Under some circum- stances, it may not be practical to collect deep soil samples The presence of ordnance, drums, concrete, voids, pyrophonc ma- tenals, and high-hazard radioactive materials may predude some sampling and may require development of alternate sampling designs, or even reconsideration of project objectives. SoIl Characte,laijcs The charactenst,cs of the soul matenai being sampled have a marked effect upon the selection of a sampling device. An investigator must evaluate soil characteristics, the type of VOC, and the depth at which a sample is to be collected before selection of a proper sampling device. Specific characteristics that must be considered are 1 Is the sod compacted, rocky, or rubble filled? If the answer is yes, then either hollow stem augers or pit sampling must be used. 2. Is the soil fine grained” If yes, use split spoons, Shelby tubes, liners, or hollow stem augers. 3. Arethereflowungsandsor ter . soil If yes, use samplers such as piston samplers that can retain these materials. SOIL-GAS MEASUREMENTS SoiL-gas measurements can serve a vanety of screening pur- poses in soil sampling and analysis programs, from initial site reconriaissa,i to remediaj monitoring efforts. Soil-gas mea- suremerits should be used for screening purposes only, and not for definItive determination of soil-bound VOCs. Field analysis is usually by hand-held detectors, portable GC or GC/MS, infrared detectors, ion mobility spe meters (IMS), industha] hygiene detector tubes, and, recerilty, fiber optic sensors At some sites, soul-gas sampling may be the only means of apouinng data on the presence or absence of VOCs in the soil. For example, when the size and density of rocks and cobbles at a site prevent insertion and withdrawal of the coring device and prevent sampling with shovels and U ’owels, unacceptable losses of VOCs would o ur. Soil-gas measurements, which can be made on site or with collected soil samples, can be used to identify volatile contaminants and to determine relative magnitudes of concentration. Smith et aJ. (1990) have shown a disparity in soil-gas VOC concentrations and the concentra- tion of VOCs found on the solid phase. Soil-gas measurements have several applications. These in- dude in situ soul-gas surveying, measurement of headspace conoarllrations above containerized soul samples, and scan- ning of soul contained in cores collected from drfferent depths. These applications are summarized in Table 6. Currently, no TABLE 6. APPliCATiONS OF SOIL-GAS MEASUREMENT TECHPIQUES IN SOIL SAMPliNG FOR VOCs Aon Uses Methods B ta4 ore Soul vapou’ Identify sowcos.and extent Aotve from soul surve nng of contanunation. Distinguish between soul and ground Water contamination. Detect VOCs Lmder asphalt, concrete. etc. probes into canisters, glass bofbs, gas sarr iUng bags. Passive sampling onto buried LsorTXne stt ates. Folewed by GC o other analysis. BENEFITS Rapid, inexpensive screemang of Large areas, avoid san ftng ulbcontaminated areas. UMfl ’AllONS: False positives and negatves, miss detedrig bcallzed si.idaoo sp ls, dsequdibnum between adsorbed arid vapor phase VOC con Soil headspa Screen large ntmters of soul Measure headspace above measurements samples containerized sod sample Containers range from plastic sanawudi bags to VOA vials Use GC, vapor detectors, IMS, etc BENEFITS Mou-e repmsen ve of adsorbed sold phase concant’aion UMITAT1ONS Losses of vapor phase component duiing san ing and sample transfer Screening Soil cores scanned to Locale Collect core soul cores depth where highest VOC levels are located (e.g , split spoon) and scan ion vapors new core surface tsing portable vapor monitor BENEFITS: Locate arid collect sod from hot spot in core for word case, UMrTAT)ONS False negatives and positives, environmental conditions can influence readings (a g , wind speed and thedion, terepereture, humidity) 10 ------- star ard protocols exist for soil-gas analysis; many investiga- tars have devised their own techniques, which have varying degrees of efficacy Independently, the American Society for Testing and Matenais (ASIM) and EPA EMSL-LV are preparing guidance documents for soil-gas measurement. These docu- ments should be available late in 1991 The required precision and accuracy of site characterization, as defin d in the DOOs, affect the selection of a sampling device. Where maximum precision and accuracy are required, sampling devices that collect art intact core should be used, particularly for more volatile VOCs in nonretentive matrices Augers and other devices that collect highly disturbed samples and expose the samples to the atmosphere can be used if lower precision and accuracy cart be tolerated. Collection of a larger number of samples to characienze a given area, however, ca_n compen- sate for a less precise sa_mpling approach The doser the expected contaminant level is to the action or detection limit, the more efficient the sampling device should be for obtaining an accurate measurement SOIL SAMPLING DEVICES Table 7 lists selection critena for different types of commercially available soil sampling devices based on soil type, moisture status, and power requirements The sampling device needed to achieve a certain sampling and analysis goal can be located in Table 7 arid the supplier of such a device ca_n be identified in Table 8. Table 8 is a partial list of commercially available soil sampling devices that are currently in use for sampling soils for VOC analysis. The list is by no means exhaustive and indusion (Conbriued co page 14) TABLE 7. CRITERIA FOR SELECTING SOIL SAMPUNG EQUIPUENTt Type of Sampler Obtains Core Samples Most Sut ls SoIl types Operation In Stony So I ls il le Sod Mo lsbzs Con t Ioos Refatlys Ses%ple S Labor J em.nds (#o(Per.oos) m 1 or Pow Operation A Mechanical San le Recovery 1 Hand-held Power augers No Coh/ h1ess Unfavorable trile,mediate 2 Solidstemfllgtnaugers No CoW h’ ss Favorable Wet to Large Pow 3 H w-steni augers Yes olVcohless Favtizilav Wet to y Lzge Pow 4 Budwt augers No Cohf h1esa Favorable Wetto&y Large Power 5 Baclthoes No ColVoehiess Favorable Wet to y Large Large 2. 2+ Power Power BSan ers 1 So w-ype augers No Coh Unfavorable lntorme Sma 2 Barrelaugers Srigle a. Po -hole augers No Coli Unfavorab le Wet b Dutch augers No Coh Unfavorable wet large Singe c Regular barrel augers No Coh Unfavorable lntacmed Large Smgle Mesu Sand augers No Cohiess Unfavorable frtteimed ta Large SlnØe Maiu e Mud augers No Coh Unfavorable Wet large Single 3 TL e-typO sarr ers Large Single a. Soil sainplet, Yes Coh Unfavorable W e tto y Sm l b Vethmeyer tt es Yes Coh Unfavorable lritarme Single c Shelby tubes Yes Coh Unfavorable Ir itermecIlate Large Single d RIng-Irted san tlers Yes Coliless Favorable Wet to nteiine ate Large 2+ Both e Co ussarr lers Yes Coh Unfavorab le etto thy Large 2i 2+ Beth Retonsan le rs Yes Coh Unfavorable Wet Large 9 Zerocomerrunationsarnplers Yes Coli Unfavorable Wettov terme Large Sm 2+ 2+ Both h SpUtspoonsarrçlers Yes Coh Unfavorable tiierme i 2+ 4 Bu samplers No Coh Favorable Wettocty Large Large Single Mwe i t wdftamUS EP& 1986.. AU fand-opera d neor of sai i . e nwwm saITlerL , be uKed by one poiscit Coh — 11 ------- TABLE & EXAMPLES OF COMMERCIALLY AVAiLABLE SOIL SAMPUNG DEViCES Le th thu) LD. t s) . AkwM1ll d D & P sge and Tr 3 r dy swT! e sods Uamsta nç Co. Sod San er 0.5 b a ee1 ç by 1.ow Lever 814 ICrVi Henry S’eet S ess eel Pwge and Tr m ho Aiexz* . VA 314 703 - 549 - 5999 frôar Dr Co. H evy Duty l. 18824 SØt e ows for easy P.O. Box 830 Spdt T A e Swrçler 1-1P2 4-1 /2 sarT e removal Suk ,PA Steel 717-586-2061 De. n Core Bairel 24860 Brass WI rsn Lm bsbed 1-7)8 te 6- 1 6 saiTile from cohess’e sods AMS COreSOdSarr IIeI- 21o12 S ess plastic Goodrit pesofsoiis He on Oregon Trail 1.1)2 3 aljwn, Ixuize AmencanFajs ,ID83211 Aboy , afless teflon Dual Pispose Sal 12,18 824 8ut yata, Teflon Ad s to AMS i s A ,*ii Recovary Pmbe 3/4 end I eea hwom& 1fte rnent Use 4 l3OAJby , Sod Recovery Auger 8 to 12 Pt ic &afless M dle to AMS exteasion 283 Teflon a lisnmum aiidoross-4 andes. Concord, Inc. Speedy Soil Sarr )ler 488 72 Autor ad sy m alows 007thAve N 3/16to3-1/2 rethevajo 24msod Far9o,P 58 102 Stauless saiTOeel l2sec. 701 - 0-1260 Zero Cor ai vnatjoi, Llrvt Han Hed San r CME 60 Butyi’ate Mayrtbe x ein Central Mine Equp. Co 2-1/2 to 5-3/8 ony sods. Adapts to CMS 6200 North Broedway Steel, s ess auger. St Lo as. MO 63147 800-325-8827 Searing Heed Coritriuous 60 ButyTate Versa e system. Adapts San leTi eSy ern 2-1)2 b o1 augers. Steel, ass Diedndi Dnlhng E p Heavy Duty Split 18 £24 Biass, plastic Full line of ssones P 0 Box 1670 T& Sa ier 2,2-1)2,3 amless, Teflon are avai fe. Lapoiie, IN 46350 Steel CoousSarr ler 60 Brass plastic 3,3-1)2 a irUess, Teflon (Continued) 12 ------- TABLE 8. (CONI1NUED) — — s) 1.0. (fficfles) Maufactures Sampllflg Dev1 Sampler Matertai’ Unera Feelures Geoprobe Systems Probe Drive 11-1/4 Remains n eteIyseaied 607 Barney St Soil Sampler 0.96 w1,ile pustied to depth in Saiina.KS Moy steel sod 913-825-1842 Giddings Maclime Co Conng Tubes 48 & 60 Butyrate A senes of optionaJ 8 in P 0 Drawer 2024 7/8 to 2-318 slots permi observation of Fort Collins, CO 80522 4130 M&ycl rome the sample. 303-485-5586 - JMC Environmentai msts 36848 PETGpL a C, Clements and Associates Sub-sod Probe Q 9 stainless peneüate the hardest ol soi . RR. lBox l8G Nickaiplated Newton, LA 50208 800-247-6630 Zero Contamination 12,18 & 24 PETG p1a c, Adepts top probe. Tubes 0.9 stainless Mobee Onhing Co. L mac Split 18 & 24 Brass. Ad Xs to Moble weC. e 3807M s on Ave. Ba!TeI Sampler 1-1i2 plastic lr ianapolis, IN 46227 800-42 4475 stinc. ZeroCitananatmon 12,18824 Handsarrqjei-goodb ’ 66 Aftmretht Drive Sampler 0.9 therical residee stiut e& Laiw Blufi, IL Chrome plated 800-323-1242 Thin Wall TiEs 30 Wil take ttd isbed sertçlee Sampler (Shelby) 2-1)2,3, 3-1t2 in ixiheens ee amid days. : ‘ S t Ti S ipler 24 Forced i by 1 —1/2to3 h * Icprseor wmQ . Vvtypecfaa. 1 I.. Ve8veyerSoi 48872 S ingT i Ee 3/4 u, dcr a*. Spragmk&Hen dtnc. S&H *8ai el 18824 Agenera çurp Scranton, PA 18501 Sw ip ler 2to3-1/2 s&1 da ced sq d 600-344-8506 ___ heunled. Not: e i eriam . U on in ioiid n be x eeusd ii wmu ins. 13 ------- in the list should not be construed as an endorsement f r their use. Commonly, soil samples are obtained from the near surface using shovels, scoops trowels, and spatules . These devices can be used to extract sod samples from trenches and pits excavated by beck hoes. A predeaned shovel or scoop can be used to expose fresh soil from the face of the test pit A thin- walled tibeor smaildameter. hand-held corer can be used to collect soil from the exposed face. Bulk samplers such as shovels and trowels cause considerable disturbance of the soil and expose the sample to the atmosphere, enhanong loss of VOCs. Seegnst and Jenssen (1990) have shown that sampling procedures that cause the least amount of disturbance provide the greatest VOC recoveries. Therefore, sampling devices that obtain urdstirbed soil samples using either hand-held or me- thanicai devices are recommended. Sampling devices that collect un bed saroples include spilt-spoon sai ers, rIng — continuous samplers, n Inabon san ers, sf4 Shelby ties These sampling devices can be used to collect surface sod samples or they can be used in con jun on wTth hollow-stem augers to collect enbsurface samples. The soil sampling devices discussed above are summarized in Table 9. Devices where the soil samples can be easily and qucidy removed and centajnenzed with the least amount of disturbance arid exposure to the atmosphere are highly recommended. U.S. EPA (19868) gives a more detailed discussion on the proper use of drill rigs and sampling devices. Liners are available for many of the devices listed in Table 9. Liners make sell removaJ from the coring device much easier and quicker. Liners reduce cross contamination between samples and the need for decontamination of the sampling device. The liner can run the entire length of the core or can be precut into sections of desired length. When sampling for VOCs, it is othcal to avoid interactions between the sample arid the liner and between the sample and the sampler. Such interactions may include either adsorption of VOCs from the sample or release of VOCs to the sample. Giliman and O’Hannesin (1990) studied the sorption of six monoaromatic hydrocarbons in ground water samples by seven matenais. The hydrocarbons included benzene, toluene, ethylbenzene, and 0-, m-, and p-xylene. The materials exam- ined were stainless steel, rigid PVC, flexible PVC, Pi P- Teflon, polyvinyl idene fluoride, fiberglass, and polyethylene. Stainless TABLE 9. $01 SAMPLERS FOR VOC ANALYStS mminded Not l ccnwnended Split oon wilners SoMd it ners Shaby tbe (thu wall tiles). Dc mg mud auger Hofow ern augers Air drilu ig auger V&wneyer King tiles Cable tool wliners Hand augers P ton sanplers’ Bw augers Zero nvnation Scoop mplers Probe -i*ive sanplers Excavalmg tools, e.g , thovels, baddioes steel sh ri,d riO s gnhtkarTt sorption slng an 8-week period. Au polymer rnatenais sorbed all compounds to some extent The order of sorption was as_follows: rigid PVC fiberglass .c polyvinylidene fluonde < t ’ I t- c polyethylene <1lex le PVC. Stainless steel or brass liners should be used since they exhibit the least adsorption of VOCs. Other materials such as PVC or acetate may be used, provided that contact time between the soil and the liner material is kept to a minimum. Stainless steel and brass liners have been sealed with plastic caps or paraffin before shipment to the laboratory for sectioning and analysis. VOC loss can result from permeation through the plastic or paraffin and volatilization through leaks in the seal. Acetate liners are available, but samples should not be held in these liners for any extended period, due to orpbon onto and permeation through the material. Altem vety, the soil can be extruded from the liner, and a portion can be placed into a wide- mouth glass jar. Smaller aiiquots can be taken from the center of the precut liner using subconng devices and the soil plug extruded into VOA vials. TRANSFER OF SOIL SAMPLES FROM DEViCE TO CONTAINER The sample transfer step is perhaps the most cnticai and least understood step in the sampling and analysis procedure. The key point in sample transfer, whether in the field or in the laboratory, isto mirumize disturbance and the amount of time the sample is exposed to the atmosphere. it is more important to transfer the sample rapidly to the container than to aoi .irate4y weigh the aiiquot which is transferred, or to spend considerable tine reducing headspace. Therefore, a combination of a device for obtaimng the appropriate mass of sample and placement of the aliquot into a container that can be directly connected to the analytical device in the laboratory is recommended. Severai designs are available for obtaining a S-g aliquot (or other size). Most subconng devices consist of a plunger/barrel design with an open end. The device shown in Figure 3 was constructed by Associated Design & Manufactunng Company (Alexandria, VA). Other designs include synnges with the tips removed, and cork borers (Table 8). The device is inserted Into the sample and an aJiquot is withdrawn. The aliquot, which is of a known volume and approximate weight, can then be extruded into a tared 40- mL VOA vIal. Routinely, the vial is then sealed with a Teflon-Imed septum cap. Teflon, however, may be permeable to VOCs. Auuminum-lined caps are available to reduce losses due to permeation. At the Laboratory, the vial must be opened and the contents of the vial must be transferred to a eparger tube. The transfer procedure will result in significant losses of VOCs from the head pace in the vial. The modified purge-and-trap cap shown in Figure 4 eliminates the loss of VOCs due to container opening and sample transfer. The soil is extruded from the subcorer into a tared 40-mL VOA vial and the modified cap is attached inthefleld. In thelaboratory, the vialisattacheddirecdy to a purge-and-trap device without ever being opened to the ambient air. Use of subcoring devices should produce analytical results of increased acouracy. In order to test this hypothesis, an expen- merit was conducted in which a bulk soil sample was spiked with 800 ig/lcg of different VOCs (Maskannec, 1990). Three aliquats were withdrawn by scooping, and three ahquats were withdrawn by using the sub-corer approach. The results are presented in Table 10. Although neither method produced quantitative recov- ery, the subcorer approach produced results that were generally 14 ------- Method sof % S n .q Compound Method Method” of iI C ethane 50 1 5 6 153 Oromomethane 31 536 4 67 Chloroethane 78 946 10 118 11-DKt oroethene 82 655 10 82 11 -Dichloroethajie 171 739 21 82 Chbrobrn 158 534 20 67 Carbon tmrathonde 125 658 16 82 12Dichiorop r opar 147 766 18 98 Inroethene 120 512 15 64 Benzene 170 536 21 80 11,2-TndIom e t haj, 78 477 10 60 Bromotorn 30 170 4 21 1,1 ,2,2-Trthomethnjie 46 271 6 34 Toluene 129 656 16 82 — Ethy eruene 57 298 7 37 68 332 8 42 Stymne 30 191 4 24 Noir meeEd c i sanple v a. mr method i e d sWm F se 3. S m eçè.d th 800 4cg ci e fl VOC five times higher than the standard approach, whereby the contents of a 1 25-mL wide-mouV, jar are poured into an alum - num tray and homogeruzed with a s iniess steel spatula A 5- g sample is then placed in the sparger tube (SW-846, Method 8240). Several compounds presented problems with both approaches: styrene potymenzes, blomoform purges poorly, and 1,1 ,2,2-tefr chloroethacie degrades qulctdy. lIT’ Stalnleee Steel Body O - Rlng 1116” Teflon Ball Receiving union from Purge .and-T ap Device 112” Stalnlesa St Body 0-Ring Hoia Cap 40 ml Vial Pu eNee e Fçure 4. ModIfied purge-and- 40-niL VOA conta nenzing samples m the field. Vl at thed dIrectly to a pwge-and-tr , system without exposure of sample to the a nosphece. Figure 3 Small-diameter hand-held subconng device made by Assocated Design & Manufactunng Company (Aiexandna, VA). TABLE 10. LABORATOHY COMPARISON OF STANDARD METHOD AND SUBCORER METHOD 15 ------- 1r1 IOth stidy (U.S. EPA. 1991a). a large quantity of well characterized soil was spiked with 33 VOCS and was hornog- enized. From the homOgem2 d mat0 , a 5-g aiiquot of soil was placed in a 40-mL VOA viai and sealed with a modified purge- and-trap cap (Figure 4). The remaining soil was placed in 125- mL wide-mouth jars. The samples were shipped via air camer and were analyzed by GC/MS with heated purge and trap. The 40-mL VOA vials were connected directly to a Tekmar purge- and-trap uni! without exposure to the atmosphere The wide- mouth jars were processed as per SW-846 Method 8240 spec- fications (U.S. EPA, 1986b). Table 11 compares the resutts of the GC/MS analyses using the two pretreatment techniques The modified method (40-mL VOA vial with a modified cap) yielded consistently higher VOC concentrations than the trath- lionel Method 8240 procedure (U.S. EPA, 1986b). The standard methods for VOC analysis, SW-846, Method 8240 and Test Method 624 (U.S. EPA, 1986b; U.S. EPA, 1982), call for the containerizing of soil samples in 40-mL VOA vials or 125- mL wide-mouth jars with minimal h sae 1 cp . As previously desa bed, wide-mouth jars may not be the most appropeate containers due to sample aliquoting requirements. Although wide-mouth jars may be equally as effective as 40-niL VOA vials in maintaining the VOC content of soil samples, the sample preparation procedure that is required with jar-held samples causes significant (>80%) loss of highly volatile VOCs (Siegrist and Jenrisen, 1990). However, if samples are collected in such containers, it is important to ensure sample integrity, preferably by using amber glass jars (for photosensitive compounds) with solid phenolic resin caps and foam-backed Teflon liners. Alum i- num-lined caps are not available for the wide-mouth jars. Soil shouldbewipedfromthethreadsof the jartoensureat ightsea l. The methanol-immersion procedure calls for the transfer of the sample into a glass jar conteining a known volume of chrornato- graphic-grade methanol (usually 100 niL) or in a 1:1 weight-to- volume ratio of soil to methanol. This has the effect of preserving the volatile components of the sample at the time the sample us placed in the container. Furthermore, surrogate compounds can be added at this lime in order to identify possible changes in the sample dunng transport and storage. The addition of methanol to the sample raises the detection limi from 5 to 10 rg(Kg to 100 to 500 Mg/kg, because of the attendant dilolion. However, the resulting data have been shown to be more representative of the original VOC content of the soil (Siegrist and Jennsen, 1990, Siegnst, 1990). In a companson of transfer techniques, Siegnst and Jennsen (1990) demonstrated thai minimum losses were obtained by using an undisturbed sample followed by immediate TABLE 11. COMPARiSON OF VOC CONCENTRATIONS 4 SPIKED SOIL ANALY D BY METHOD 8240 AI UO RED METHOD 8240 thod —Concun atlon g) ‘ 1 o . -d M ed VOC 8240t 8240tt Diftaiwice —Cancen on g) Moifted M ed VOC 8240t 8240ft Diffaisece t Metiod 8240 ç 1 -iit rdeuth r iT ig aL inç n9 vi bburaL y 1u9 tf Melhod 8240 usmg 4Oiri vai &ç nç l vi d ie , el ied b ulaLiy pqVrap arely — ll e e ii y pv r tian 0, rth Pvais 01 cwi e s 1I i y 7ealer tiaii 0, rth P - akie betwien 0.01 and 005 No iiw was 300 i 1c l i lorde 9 3 44 32 35 29” D8xornodiloromethane 1,l2.Tn oroeffisne 121 142 159 193 38 51 Chloroethane 6 36 30” ere-1,3-Did1oropropene 154 203 49 Meth 1ened ilonde 69 100 31” Broiwil c n 116 140 24 Caibon daiflde 32 82 50” Teuadioroethene 62 124 6Z ’ 1.1 - d omethene 12 35 zr 1,122-Tetiadloroidiane 137 162 25 1,1- dloroethane 34 83 49” Toliene 85 161 76’ 12.Didloroet lw ie 36 66 30” Chk robenzene 91 132 41” Ch lorotom, 56 96 40” Ethyberizene 85 135 50” 1,1 ,1-Tn omethane 26 90 54” Styvane 86 114 28’ Carbon Wonde 18 61 43” To Jx 1enes 57 85 28” Vu ’u 1 tate 18 26 8 12 -l dlomeV ’ iane 101 159 58” KETONES os-1,3-Dudiloropropene 136 189 53’ A tone 336 497 161’ Tnthbroethene 48 87 39” 2-8u icne 290 365 75 Benzene Bmmoc dloromethaiie 56 114 58 2-Hexanone 200 215 15 16 ------- immersion into methanol. The results for SIX VOCs are shown in Figure 5 At high VOC spike levels (mg/kg) the investigators found that headspace wrthnn the bottle Caused a decrease in the concontratlon of VOCs in the sample. At lower spike levels, however headspace did not seem to be a major contributor to VOC losses (Maskannec, 1990). In another study (U.S EPA. 1991a), soilcore and placed in a 40-mL VOA vial provided consistently higher TREATMENT A UNDISTURBED SOIL PLAS T1C BAG LOW HEADSPACE TREATMENT B UNDISTURBED SOIL GLASS JAR HIGI HEADSPACE TREATMENT C DIS11JRBED SOIL GLASS JAR LOW KEADSPACE S Concentration, ppm TREATMENT D UNDISTURBED SOIL GLASS JAR LOW HEADSPACE TREATMENT E UNDISTURBED SOIL GLASS JAR METhANOL TOLU9 E — cao oee Figure 5. VOC recovery as a function of sample treaUnent. 15 10 0 T AT TA T AT ITB T AT ITC T AT P T AT ITE — METHYLENE CHLORIDE — 1,2-OICItOROETh conceiihutlOfl, ppm 2.5 2 13 0 T _ATVB(T A T A1 (T THEA1 lT C T A1 lTD T A1 (T E 1,1,1 ,-T LOROEfl1ANE — T *tOROE 17 ------- VOC levels than a sample taken from the same core, placed in a 1 -mL wide-mouth jar, and later POUTed 0 (11, homogenized, and a 5-g aiiquot taken from the bulk material as per Method 8240 specifications. SOIL SAMPUNG SCENARIOS The foIIowrn rec riendations for soil sampling and sample handhng are presented for the three generai sampling sce- nanos described earlier 1. Open Test Pft or Trench Samples are often collected from exposed test pits or trenches where remedia on efftxts are in progress Srtes may also be encountered where large-diameter coring devices cannot be employed. In such instances, aude sampling devices, such as trowels, spoons, shovels, spades, scoops, hand ai.igers, or bucket wagers must be used to excavate the soil. The exposed f of an excavated test pit is soaped to uncover fresh material. Samples are collected from the soaped face by using a small-diameter, hand-held corer (Figure 3). If the nominal 5-g sample is to be collected, the appropriate volume (3 to 4 mL) is extruded into a tared 40-mL VOA vial and sealed with a mothfled purge-and-trap cap (Figure 4). The vial is thlied to 0° to 4°C and sent to the laboratory where the entire contents of the vial are purged without opening the vial (U.S. EPA 1991 b) Though this method minimizes losses of VOCs, the small sample size may exhibit greater short-range spatial variabdrty than larger samples. Alternatively, a small-diameter, hand-held sell corer (Figure 3) can be used to collect a larger volume of soil. The soil is extruded to fill a 40-niL VOA vial with a Tefion- ined septin, cap (minimal he 1sper ) , chilled, and sent to the laboratory The major weakness with this method is that VOCs are lost in the laboratory dunng sample homogenization, preparation of aJiquots from a subsample, and the transfer to the extraction or sparging devx . if large coarse fragments or highly compacted soils are encoun- tered, the use of a hand-held corer may not be possible In this ca_se a’ude sampling devices are used to rapidly collect and fill (minimal headepace) a 125- or 250-mL wide-mouth glass jar. The threads are wiped clean and the jar is sealed with a foam- backed Teflon-lined cap. The jar is chilled immediately to 0° to 4°C for shipment to the laboratory. Losses of VOCs are consid- erably greater with this method due to demuption of the matrix and losses in the laboratory during sample preparation. Metha- nol immersion may be more suitable for these matrices. 2. Surface Solla (.c 5 ft deep) The preferred soil sampling procedures reduce VOC losses by minimizing sample disturbance dunng collection and ti’aristerto a container. The collection of soil cores with direct extrusion into a container accomplishes this goal. A larger-diameter coring device (e.g., splrt-spoon sampler, Shelby tube, zero-contami- nation sampler) is used to collect an intact sample from the surface soil or from an augered hole. Many of these samplers can be used with liners, an insert that greatly reduces the time required to remove the soil and obtain a subsample. For subsamples collected from split spoons or extruded large- diameter cores, the section to be subsampled is scraped and laterally suboored, orthe extruded soil is cut or broken to expose fresh material atthe depth orzone of interest, then longitudinally suboored. For large-diameter cores that are collected in precut liners, the liner sections are separated with a stainless steel spatula, and a small-diameter hand-held corer is used to collect a subsample from the center of the liner section. The uppermost porton of the core should not be sampled, because it is more likely to be cross contaminated. The small diameter corer (Figure 3) is pushed into the soil, the outside of the corer is wiped clean, and the required core volume (typically about 3 to 4 mL or5 g) is extruded directiy into a tered 40-rnL glass VOA vial and sealed with a modified purge-and-trap cap (Figure 4) The vial threads and lip must be free of soil to ensure an airtight seal 3. Subsurface soIls 5 ft deep) The same sampling pnndples apply for the collection of deeper soil samples. Collection of soil cores with direct extrusion into a containergreattyreducesthelossofVOCs. Tube-type samplers such as split-spoon, Shelby tubes, and zero-contamination samplers are used inside a hollow-stem auger to obtain an intact sample from greater depths. The coring device is retrieved and a subsample is obtained in a similar manner as that descrIbed for surface soils. METhANOL IMMERSION PROCEDURE Soil collected by protocols outhned above can be placed in a tared wide-mouth glass jar containing pesticide-grade methanol (1:1 weight-to-volume ratio of soil to methanol). The ‘nmersion of relatively large soil samples into methanol has the advantage of extracting a much Larger sample that is probably less prone to short-range spatial varIability. This is of particular advantage with coarse-gramed soils, materials from wtiidi it is hard to obtain a 1 to 5-g subsample for analysis. Multiple small-diameter corers can be immersed in a single methanol-filled jar to produce a composite sample Compositing becomes pr caJ because VOCs are soluble in methanol, thus reducing losses. Appropriately collected corn- posrte samples can produce more representative data than a comparable number of individual samples. Short-range spatial varrabihty is greatly reduced. Another advantage is the ability to reanalyze samples. The main disadvantages of using methanol include the requirements for handling and stupping the media- nolandthedetectionhmhthatisraisedbyafactorof about 10 to 20. For the methanol-immersion procedure, jars filled with methanol and shipped to the laboratory are classified as a hazardous material, flammable liquid and must be labelled as per Department of Transportation specifications (49 CFR, 1982). If these disadvantages are unacceptable, then the modified purge-and-trap procedure may be applicable. 9ELD STORAGE Material containing VOCs should be kept away from the sample and the sample container. Hand lotion, labeling tape, adhesives, and ink from waterproof pens contain VOCs that are often anaiytes of interest in the sample. Samples and storage contain- ers should be kept away from vehicle and generator exhaust and other sources of VOCs. Any source of VOCs may cause contamination that may compromise the resulting data. 18 ------- Once samples are removed from the Sampling device and placed in the appropriate storage container, the containers should be placed in the dark at reduced temperatures (00 to 4°C). Excessively cold temperatures (< ‘10°C) should be avoided: studies have shown greater tosses of analytes due to reduced pressures in the container, sublimation of water, and concomitant release of water-soluble VOCs into the headspace Upop opening the container, the vacuum is quickly replaced with ambient air, thus purging out VOCs from the headspace (Maskannec et al , 1988). Extremely cold temperatures can also loosen the seal on the container cap. Caps should be retightened after 15 minutes at reduced temperatures. Samples should be kept in ice chests while in route to the shipment facility or laboratory At temperatures above freezing, bactenal action can have a significant impact on the observed soil VOC con- centration. Numerous preservation techniques are being evaluated at the University of Nevada Environmental Research Center in Las Vegas and at Oak Ridge National Laboratory. SHIPPING Given the short holding times required for VOC analysis under Method 8240 (10 days from sample collection to analysis), samples are usually shipped via air carner to the anaiyticai laboratory. Samples should be well packed and padded to prevent breakage Temperatures in cargo holds can increase to more than 50°C dunng transit, therefore, the need for adequate cold storage is critical. Styrofoam coolers are commercially available to accommodate 40-mL and 1 25-mLglass containers. Sufficient quantities of Blue tce or Freeze Ge1Th packs should be placed in the container to ensure that samples are cooled for the duration of the shipment. A maximum-minimum thermom- eter (non-mercury) should be shipped with the samples. ft sample containers are not adequately sealed, VOC losses can occur These losses may be exacerbated by the reduced atmosphenc pressures encountered in the cargo holds of air carriers. Figure 6 illustrates the changes in temperature and pressure in the cargo hold of various air carriers aircraft. Three major air camers have been monitored and have shown similar fluctuations in temperature and pressure (Lewis and Parolini, 1991). Lewis et aJ. (1990) noted decreases in VOC concentra- tions in soil samples that were shipped compared to samples that were analyzed in the field, If the container is of questionable or unknown integrity, it should either be evaluated pnor to use or a previously characterized container should be used. As discussed previously, samples that are immersed in metha- not have spectai shipping requirements. These samples must be shipped as Flammable Liquids’ under Department of Trans- portation (DOT) requirements, A seoondasy container is re- quired for shipment of any item dassthed as a flammable liquid. PRESERVATION Improvements ie opera1ionai factors such as sampling device efficiency, sample transfer, containerizing, shipping, storage, laboratory sample preparation, and analysis will reduce VOC losses from soils. Two pnncipai matrix-specific factors that can contnbute to the loss of VOC in soils are b4odegradation and volatilization. An effective preservation technique should act on these matrix-specific factors to reduce losses of VOCs. The required preservation technique for soil samples storage at 0° to 4°C in the dark. This technique retards biodegradation C) 0 It, I 15 14 13 12 11 I v FEDERAL EXPRESS ‘6 0 5 10 15 15 14 13 12 11 10 — - fljc X V .8 0 5 10 15 8 8 40 Elapsed Time (hr) Figure 6. Temperature and pressure fluctu ions recorded in the cargo hold of venous air carnors. Recording device was shipped from Las Vegas, NV, to Pearl River, NY, and returned. AIRBORNE 6 0 5 10 15 8 35 40 l x — — 19 ------- processes mediated by soil microorganIsms. Some micicorgan- isms, however, such as fungi, are biologicaily active even at 4°C. Woff at al. (1989) investigated several methods (i.e., chemical p4 irradiation) for sterilizing soil and concluded that mer snc chloride is one of the most effective preservatives that causes minimal changes to the chemical and physical proper- ties of the soil. Stuartet al.(1 990) utilized mercuric chlonde as an antimicrobial preservative to stabilize ground-water samples contaminated with gasoline Other researchers (U.S. EPA 1 991a) have used mercunc chloride to retard biodegradation of VOCs in soil samples. The soils were spiked with 150 j. i .g/1g of Target Compound Ust (TCL) VOCs and were preserved with 2.5 mg of mercunc chloride per 5g of soil. The results indicated that the amount of mercunc chionde needed to reduce biodegrada- tion was directly related to the soil’s organic carbon content. In addition, the levels of mercuric chloride added to samples did not interfere with sample handling or analysis. Currently, re- search is underway to quantitate the required mercuric chloride concentration as a fun on of soil organic content. The loss of VOCs through volatilization is reduced by optimizing sample handling procedures. When samples require laboratory pi ’etreatinent. severe losses of VOCs (up to 100%) have been observed. In order to minimize volatilization losses, several preservatives have been examined (U S EPA 1991 a), induding solid adsorbents, arihydrous salts, arid water/methanol extrac- ben mixtures The most efficient preservatives for reducing volatilization of VOCs from soils have been two solid adsorbents, Molecular Sieve - 5A’ (aluminum silicate desic- cant) and F1orasil’ ’ (magnesium silicate desiccant). The addi- tion of 0.2 mg per 5 g of soil greatly mci-eased the recovery of VOCs from spiked samples. The mechanism is believed to involve the displacement of water from adsorption sites on the soil particle and bmding of VOCs to these freed sites. Currently, research is in progress with soils obtained from actual contami- naled sites. LABORATORY PROCEDURES Sample Storage Most regulatory procedures specify storage of samples for VOA at 4°C in the dark. Sample coolers should be opened under chain-of-custody conditions, and the temperature inside the cooler should be verified and noted. Samples should be trans- ferred to controlled-temperature (4°C) refrigerators until analy- sis. In many cases, insufficient cooling is provided during transport In these cases, data quality may be compromised. Sample Preparation The two most commonly used methods that satisfy regulatory requirements torthe analysis of soil samplesfor VOCs are direct purge and trap and methanol extraction Each procedure has benefits and limitations with respect to sample preparation prior to VOC analysis of soils. The modified purge-and-trap procedure has the following char- actenstics Homogenization of contents of wide-mouth jar will cause significant VOC losses The collection of a 5-g aliquot in the field arid placement into a tamed vial sealed with a modified purge-and-trap cap is recommended. • Surrogate addition should be made to the soil in the field, if possible • May be more susceptible to short-range spatial variability. • Samples should be brought to ambient temperature before purging. • May be more suitable for low-level samples. The methanol-immersion procedure has the following charac- teristics. • The key is to minimize the time samples are exposed to the atmosphere pnor to immersion into methanol. • M inimumdetecbonlinirtscanberaisedbyafactorofiOto2O. • The best option for sample archival because VOCs are highly soluble in methanol. • Large-mass samples can be extracted in the field in a 11 ratio arid the methanol extract shipped to the laboratory for analysis. • Can collect composite samples. The analytical methods that can be used for the analysis of soils for VOCs are summarized in Table 12. An analytical method should be selected that is compatible with the recommended sample collection arid containerizing procedure discussed ear- her. CONCLUSIONS AND RECOMMENDATIONS Current research on sampling soils for VOC analyses answers many of the questions asked by RPMs and OSCs who conduct site characterization arid restoration. 1. There is no specific method or process that can be recom- mended for sampling soils for VOk A wide variety of sampling devices are currently used for collecting soil samples for VOk Sampling device selection is site-specific, and no single device can be recommended for use at all sites. Several different samplers, which cover a broad range of sampling conditions and circumstances, are rec- ommended for obtaining representative samples for VOC analysis (Table 7). Procedures may vary for different VOCs. Expenmerits have shown that a procedure that collects an undisturbed, intact sample with a device that allows direct transfer to a sample container (e.g., split-spoon, Shelby tube, or zero-contamination sampler) is supenor to a more disruptive procedure that uses a crude bulk sampler (e.g., shovel, trowel, scoop, or spade) for maintaining the mntegnty of VOCs in a soil sample. Large-diameter tube-type sam- pling devices are recommended for collection of near- surface samples The same types of devices can be used in conjunction with hollow-stem augers for collecting sub- surface samples 2. Transfer of the sample from the sampling device to the container isa critical step in the process. Losses of as much as 80% have been observed dun rig this step. The faster the soil can be removed from the sampling device and 20 ------- TABLE 12. METHODS FO P VOC ANALYSIS OF SOIL Method Sample $ ze Sample paradon SenSItMty OiaUly Extrac o&analySI3 (9) ProcedUre (jig /kg) ot . ctt , e / 8 5 Purge and trap 5-10 Li g on RCRA’ Sasap e transfer to /8010 pwgeandtrapis /8015 /8020 /8030 /8260 5380 / 8240 5-100 Methanol extraction 500-1000 RCRA Sensitivity loss but /8010 etransfer /8015 tacthtaled /8020 / 8030 /8260 5031 / 8240 5 Field purge 5-10 Seni- RCRA San e n only be /8010 quantitative analyzed /8015 fransferandshippng /8020 faa6t ed. / 8030 /8260 3810/8240 10 Heatto9 OcC 1000 Screecmg RCRA Canbepeslodrned / 8010 in Water bath fix pwge fe in the fiefd. /8015 andanaiyze orgajics /8020 hea&cpace /8030 /8260 3820 10 Hexadecsne 500-1000 Saeecwig RCRA FID responses vary extraction pnortoGC thtype&VOC. followedby orGC/MS GC/FID analysis 624 5 Purge and trap 5-10 Litigation CJ.P Sum to method 5030 /8240 m RCRA SW-848. US EP&1986b ‘US EPk 1982 21 ------- transferred into ai airtight sample container, the smaJler the VOC loss. Liners make the removal and subsamphng of soil from the collection device more efficient. 3 The best method for transfemng a sample from a large- diameter conng device (or exposed test pit) unto a sample container is by collecting the appropnate size aluquot (for laboratory analysis) with a small-diameter, hand-held corer and extruding the subsample unto a 40-mL VOA viaJ, then sealing the vial with a modified purge-and-trap cap. Alter- natively, contents of the large-diameter conng device can be sectioned and immersed in methanol 4 Small-diameter, hand-held corers can be used for col- lecting samples from a freshly exposed face of a trench or test pit, or for obtaining a subsample from a large-diameter ccnng device The use of a small-diameter, hand-held corer is recommended for obtaining subsamples from liner-held soil Collection of a sample of the appropnate size for a particular anaiytca] procedure us optimaL The required size of aliquot cart be extruded into a 40-mL VOA vial and sealed with a modified purge-and-trap cap. The possibility exists of composuting several small-diameter core samples by immersing them in a single jar containing methanol 5 Sample containers vary in terms of air-tlghthess Data are available to indicate that there is a dea’ease in pressure and an na-ease in temperature in the cargo holds of certain air caj’ners. This is the worst possible set of conditions for maintaining VOCs in containenzed soil samples. Intact seals on storage containers and adequate cooling is thus critical for maintaining VOCs in soil samples. Shipping and holding-time studies have shown that vials and jars may be equally suited for containing VOCs in soil samples, the laboratory pretreatment step needed to obtain an aliquot from a jar-held sample causes signrfi cant losses of VOCs Commercially available shipping packages with built-in cooling matenals (e g.. Freeze Gel Packs® or Blue Ice®) are available. Whenever possible, art integrated sampling approach should be employed to obtain the most represen- tative samples possible. Soul-gas surveying coupled with on-site soil sampling and analyses followed by the Re- source Conservation and Recovery Act (RCRA) or CLP laboratory analyses may provide valuable information on the partitioning of VOCs at a site - 6. The current preservation technique for soil samples is storage at 4°C in the da& Biological activity may continue at this temperature. The addition of mercunc chlonde to the soil may reduce biodegradation of VOCs. The amount of mercunc ctilonde to be added, however, isa function of the organic carbon content in the soil The most promising preservatives for reducing losses of VOCs through volatil- ization are solid adsorbents such as Molecular Sieve - 5A’ and RorasW ’ 22 ------- REFERENCES CfTED 49 CFR. 1982. Code of Federal Regulations, 49, Parts 100 to 177 October 1, 1982, pp 231 Acker, W L 1974 Basic Procedures for Soil Sampling and Core Drilling Acker Drill Co., Inc., Scranton, PA, 246 pp. An elh, J -D, G Milde, H Kemdorff, and R Schleyer. 1988 Waste deposit influences on groundwater quality as a tool for waste type and site selection for thai storage quality. rn P Baccini (Ed.) The Landfill Reactor and Final Storage, Swiss Workshop on Land Disposal of Solid Wastes, Gerzensee, March, 14-17, pp 399-415 Barcelona, M J 1989. Overview of the Sampling Process In Keith, L. H (Ed.), Pnnciples of Environmental Sampling, Amen- can Chemical Society, Washington, 0 C , pp. 3-23. Barth, D.S.B.J Mason,T H, Starks,andK.W.Brown. 1989. Soil Sampling Quality Assurance User’s Guide (2nd edition), EPA 600/8-89/046, U S EPA, EMSL-LV, Las Vegas, NV, March, 225 pp. Boucher, F. R. and G F. Lee. 1972 Adsorption of Lindane and Dieldnn pesticides on unconsolidated aquifer sands Env. Sci. Tech. 6.538-543. Bouwer, E. J. 1984 Biotransformation of organic micropollutants in the subsurface. Irn Petroleum Hydro ibons and Organic Chemicals in Ground Water. National Well Water Association, Dublin, OH. Cameron, A E 1963. Algae of southern Arizona - Part 1 Introduction to blue-green algae. Rev. AIg. N.S. 6(4)282-318. Chiou, C. T. and 1. D Shoup 1985. Sod sorption of organic vapors and effects of humidity on sorptive mechanism and capacity. Env. Sc. Tech. 19(12):1196-1200. Chiou, C. 1., 0 E. K Ie, and A. L Malcolm. 1988. Sorption of vapors of some organic liquids on soil humic acid arid its relation to partitioning of organic compounds in soil organic matter. Env. Sci. Tech. 22(3).298-303. Dean. J. D, P S. Huyakom, A. S. Donigan, Jr., K. A. Voos, A. W Schanz, Y J Meeks, and A. F Carsel. 1989. Risk of Unsaturated/Saturated Transport and Transformation of Chemical Concentrations (RUSTIC) - Volume I: Theory arid Code Venficaijon, EPN60 3-89/O48a, Environmentai Re- search Laboratory, U.S. EPA, Athens, GA, 203 pp. Davill, 0 A., A. B. Evans, W. A. Jury, T. H. Starks, B. Eldunk , and A Gholson. 1987. Soil Gas Screening for Detection and Map- ping of Volatile Organics. National Well Water Association, Dublin, OH, 27ap 0 . Farmer, W. J., M.-S Yang, J. Letey, and W. S. Spencer. 1980. Land Disposal of Hexachlorobenzene Wastes: Controlling Va- por Movement in Soils EPA-60012-8o -1 19. U.S. EPA. Environ- mental Research Laboratory, Cincinnati, OH, August 1980. Gillrnari, R W and S.F O’Harinesin. 1990. Sorption of aromatic hydrocarbons by matenais used in construction of ground-water sampling wells. In: 0. M. Nielsen and A I Johnson (eds), Ground Water and Vadose Zone Monitcnng. ASTM SW 1053, Arnencan Society of Testing MatenaJs, Philadelphia, PA, pp. 108-122. Hanisch, A C. and M A. McDevttt. 1984 Protocols for Sampling and Analysis of Surface Impoundments and Land Treatment] Disposal Sites for VOCs. Technical Note. EPA-EMB 68-02- 3850, Work Assignment ‘ii Jarnison, V W, P L Raymond, and J 0 Hudson. 1975 Biodegradation of high-octane gasoline. ln Proceedings of the Third International Biodegradation Symposium. Applied Sci- ence Publishers Ltd., London. - Johnson, A. E. 1976. Degradation of DOT by Fungi ResiDue Review 61:1-28. Jury, W A. 1984. A User’s Manual for the Environmental Fate Screening Model Programs SAM and BCM, Dept Soil and Environ. Sc.. Univ. of Cairfomia, Riverside, CA. Submitted to the California Depamnent of Health Services. Kanckhoff, S W., D. S. Brown, and T. A. Scott. 1979. Sorption of hydrophobic pollutants on natural sediments. Water Res. 13:241 -248. PQeopfer, R. 0. at aJ. 1985. Anaerobic degradation of bichioro- ethylene in soil. Env. So. Tech. 19277-284. Kobayashi, H. arid B. E. Rvttmai,. 1982. Microbial removal of hazardous organic compounds. Env. So. Tech. 16(3):170A- 183A. Lechevalier, H. A. and M. P. Lechevalier. 1976. A nomycetes found in sewage ti’eatrnent plants of the activated sludge type. In: Actinomycetes; The Boundry Microorganisms, Toppen Co., Ltd., Tokyo, Japan. Lewis, T. E., B. A. Deason, C. L Gerlech, and D. W. Bottrell 1990. Performance evaluation matenals for the analysis of volatile organic compounds in sod: A preliminary assessment. J. Env. Sd. Health A25(5)505-531. Lewis, 1. E. and J. A. Parohni. 1991. Temperature and pressure fluctuations in a r shipped samples: Implications on sample integrity. Analyst (in preparation). Lotse, E. G.. D. A. Greet, G. Chesters, G. B. Lee, and L W. Newland. 1968. Lindane adsorption by lake sediments. Env. Sc. Tech. 2:353-357. Maskannec N. P., 1990 Personal Communication. Maskannec, N. P., L H. Johnson, S. K. Holladay. 1988. Preanalytx J holding times. Paper presented at the Quality Assurance in Environmental Measurements Meeting, U.S. Army Toxic and Hazardous Materials Agency, Baltimore, MD. May 25-26, 1988. 23 ------- McCoy, 0 E. 1985 101. studies provide insight mb future of CERCLA. The Hazardous Waste Consultant 3/2.18-24 Plumb, R H, Jr. 1987. A practicai alternative to the RCRA organic indicator parameters. In. T. Bursztynsky (Ed ) Proceed- ings of Hazmacon 87, Santa Clara, CA, April 21-23, pp 135- 150 Plumb, R H., Jr., arid A. M Prtchford 1985. Volatile organic scans. Implications for ground water monrtonng Paper pre- sented at the National Water Well AssooationiAmencan Pe- troleum Institute Conference on Petroleum Hydrocarbons and Organic Chemicals in Ground Water, Houston, TX, November 13-15, 1985 Richardson, E. M and E. Epstein. 1971 Reterthon of three unsectiodes on different size soil partides suspended in water. Soil So Soc An, Proc 35:884-887. Roy, W. A arid P A Griffin. 1985 Mobility of organic solvents in water-saturated soil matenaJs, Environ. Gaol Wat. Sci. 7(4).241 -247 Sheri, T T arid G H Sewell 1982 Air pollution problems of uncontrolled hazardous waste sites In Proceedings of 1982 Superfund Conference. Hazardous Materials Control Research Institute, Washington, pp 76-80 Siegnst. R L 1990 Volatile organic compound measurements in contaminated soils The state of practice, methods develop- ment, and implications for remediation projects. J Haz. Mat. (in press) Siegnst, R. L and P 0 Jennsen. 1990. Evaluation of sampling method effects on volatile organic compound measurements in contaminated soils Env So Tech. 24 1387-1 392. Smith, J. A, C I Chiou, J A Karnmer, and D. E Kile 1990 Effect of soil moisture on the sorption of ti chloroethene vapor to vadose-zone soil at Picatinny Arsenal, New Jersey. Env. Sci. Tech. 24 676-683 Spencer, W F and M M CIiathl97O. Soil So Soc. Am Proc. 34:574 -579 Spencer,W F,M M Cliath,W kJury,andL-LThang.1988. Volatilization of organic chemicals from soil as related to their Henrys Law constants J Env. OuaJ 17(3) .504-509 Stuart, J. 0 V D. Roe, W M Nash, and G. A. Robbins. 1990. Manual headspace method to analyze for gasoline conlarnina- tion of ground water by capillary column gas chromatography Personal Communication U.S EPA 1982 Test Method 624 (Purgeables). Methods for Organic Chemical Analysis of Municipal and lndustnai Wastes, EPA-600/4-82-057, U.S EPA Environmental Support Labora- tory, Cincinnati, OH. U.S. EPA 1983. Characterization of Hazardous Waste Sites - A Methods Manual: Volume II - Available Sampling Methods. EPA-600/4-83-040. U.S. EPA. 1986a. Permit Guidance Manual on Unsaturated Zone Monitoring for Hazardous Waste Land Treatment Units, EPN53O/-SW-86-040, pp 11-62 U.S EPA 1 986b Test Methods for Evaluating Solid Waste (SW-84.6), Method 8240, Off Solid Waste and Emergency Response (3rd Edition) U.S. EPA. 1987. Data Quality Objectives for Remedial Re- sponse Activities: Development Process EPA’5401G-871003, Off. Solid Waste and Emergency Response, Washington, 0 C U.S EPA 1988. Field Screening Method Catalog Users Guide EPN54O/2-88/005, Sept 1988, Office of Emergency and Re- medial Response, Washington, D.C. U.S. EPA. 1991 a. Investigation of Sample and Sample Handling Techniques for the Measurement of Volatile Organic Com- pounds iii Soil. University of Nevada, Las Vegas, submitted to U.S. EPA, Environmental Monitoring Systems Laboratory, Las Vegas, NV (in preparation). U.S EPA. 1991b. ManualforSamplingSoilsforvola1 Organic Compounds. Environmen J Monitoring Systems Laboratory, Las Vegas, NV, 26 pp. (in preparation) van Ee, J. J., L J. Blume, and T. H. Starks. 1990. A Rationale for the Assessment of Errors in the Sampling of Soils, EPAJ600/ 4-90/013, Office of Research and Development. Environmental Monitoring Systems Laboratory, Las Vegas, NV, 57 pp. Verscf -iueren, K. 1983. Handbook of Environmental Data on Organic Chemicals, Van Nostrand Reinhold Company, New York, NY (2nd Edition). Vogel, T M and P L McCarty. 1985. Biotransfom,aton of tetrachloroethylene to trichloroethylene, did loroethylene, vinyl chloride, and carbon dioxide under methanogenic conditions AppI. Environ. Microbiol. 49:1080-1084. Voice, T. C. arid W. J Weber, Jr. 1983. Sorption of hydrophobic compounds by sediments, soils, and suspended solids - Theory and background. Wat. Res. 17(1 O):1 433-1441. Wolf, D. C.. T. H. Dee, H. D. Scott, and I L Lavy. 1989 Influence of sterilization methods on selected soil microbiologi- cal, physical, and chem J properties. J. Env QuaJ. 18:39-44 24 *u.& OOvtaMM 4y MINTING . ------- |