United States Environmental Protection Agency Environmental Monitoring and Support Laboratory Cincinnati OH 45268 Research and Development EPA-600/S4-84-054 July 1984 &ERA Project Summary EPA Method Study 29, Method 624 Purgeables The work which is described in the report was performed for the purpose of validating, through an interlaboratory study, proposed Method 624 for the analysis of the volatile organic priority pollutants. This method is based on purging and concentration of the various analytes on an adsorbent followed by thermal desorption onto a gas chromatographic column. A low resolution mass spectrometer serves as the measuring device. Participating laboratories were se- lected based upon technical evaluation of proposals and upon the analytical results of prestudy samples. The labora- tories were supplied with ampuls containing various concentrations of the pollutant compounds. These solu- tions were aliquoted into four different water types which were subsequently analyzed according to the appropriate methods. In addition to the sample concentrates, each laboratory was supplied with an industrial effluent which was known to contain various pollutants. The purpose of this sample was to ascertain the propensity of the method of produce false positives and false negatives. The data obtained from the interlab- oratory study were analyzed employing a series of computer programs known as the Interlaboratory Method Validation Study (IMVS) system which was de- signed to implement ASTM procedure O2777. The IMVS analyses included tests for the rejection of outliers (both laboratory and individual), estimation of mean recovery (accuracy), estimation of single-analyst and overall precision, and tests for the effects of water type on accuracy and precision. This report was submitted in partial fulfillment of contract number 68-03- 3102 by Radian Corporation under the sponsorship of the U.S. Environmental Protection Agency. The report covers a period from January, 1982 to June, 1983. This Project Summary was developed by EPA's Environmental Monitoring and Support Laboratory. Cincinnati, OH, to announce key findings of the research project that is fully documented In a separate report of the same title (see Project Report ordering information at back). Introduction The various analytical laboratories of the U.S. Environmental Protection Agency (USEPA) gather water quality data to provide information on water resources, to assist research activities, and to evaluate pollution abatement activities. The success of these pollution control activities depends upon the reliability of the data provided by the laboratories, particularly when legal action is involved. The Environmental Monitoring and Support Laboratory Cincinnati (EMSL- Cincinnati), of the USEPA develops analytical methods and conducts quality assurance programs for the water labora- tories. The quality assurance program of EMSL is designed to maximize the reliability and legal defensibility of all water quality information collected by USEPA laboratories. The responsibility for these activities of EMSL-Cincinnati is assigned to the Quality Assurance Branch (QAB). One of these activities is to conduct interlaboratory tests of the methods. This study reports the results of the validation effort on Method 624 for the volatile organic compounds. The interlaboratory study of USEPA Method 624 consisted of three distinct phases. Phase I involved the preparation and ampuling of concentrates of the ------- compounds. The prepared concentrations were then verified using GC methods. The second phase involved the selection of participating laboratories. Solicitations were made for both paid participants and volunteer participants. Selection of laboratories was based on experience, facilities, quality control procedures, and cost estimates received from laboratories. Final selection of fifteen laboratories was made after the successful analysis of a performance sample. No laboratories chose to participate in the study as volunteers. The third phase involved the conduct of the study. The prepared ampuls were distributed to each laboratory. Each laboratory supplied the required four water types into which the ampuls were spiked. In addition, a single water sample was supplied by Radian to evaluate the method's tendencies for false positives and false negatives. After analysis, results were reported on standard data sheets. Data were keypunched and validated by Radian. The final step in the study was to conduct an analysis of all data obtained using USEPA's IMVS computer programs. Procedure The design of the interlaboratory study of Method 624 was based on the technique described by W.J. Youden (1). According to this technique, samples are analyzed in pairs where the concentration of each analyte in the sample pairs is slightly different. The analyst is directed to perform a single analysis and report one value for each sample. The samples were prepared as concen- trates in sealed ampuls and shipped to the participating laboratories. Each laboratory was responsible for supplying laboratory pure water, finished drinking water, a surface water, and an industrial effluent water for use in the study (two laboratories, numbers 10 and 16 used water treatment plant effluents which may have had primarily municipal origins). The analyst was required to add an aliquot of each concentrate to a value of water from each of the four water types and subsequently to analyze the spiked water samples. Sample pairs for each method were prepared at three concentration levels; low, medium, and high, all of which were within the linear range of the mass spectrometer. In addition to the sample ampuls, an industrial effluent water selected by Radian was furnished to each participating laboratory for analysis. This sample was known to contain a number of the priority pollutants and was judged to be some- what difficult to analyze. The purpose of the industrial effluent sample was to determine the propensity of the method to produce false positives and false negatives. After all analyses were completed, the results were subjected to statistical analysis using USEPA's IMVS system to determine the precision and accuracy of Method 624. Test Design The following is a summary of the test design used based on Youden's nonrep- licate technique for samples. 1. Three Youden pairs of samples were analyzed for each analyte with the deviation from the mean of each pair being at least 5% but not more than 20%. The three pairs were spread over a usable and realistic range such that the lowest pair was somewhat above the minimum detection limit and the upper pair was within the linear range of the method. 2. The spiking samples were supplied as liquid concentrates in organic solvents sealed in glass ampuls. Sufficient sample was provided to allow withdrawal of the appropriate amount of solution to spike one water sample from each ampul. 3. Twenty-four volatile organic ampuls were provided to each of the 15 laboratories. 4. The concentrates were spiked into laboratory pure water, drinking water, a surface water, and an effluent waste water by the partici- pants prior to analysis. In addition, an industrial effluent sample was supplied to each .laboratory by Radian. This sample was analyzed without addition of analyte con- centrates. 5. Each of the 15 participating labora- tories was furnished with the following materials: Four Youden pair ampuls of each of three concentration levels for the volatile organics. (A total of 24 spiking sample ampuls.) Sufficient surrogate standard solution to analyze all samples and blanks. A1 liter sample of an industrial effluent to be analyzed without addition of spiking sample. Copies of Method 624. A questionnaire covering dif- ficulties encountered with the method and suggestions for m improvements. Data report forms to be com- pleted and returned to Radian. A set of instructions detailing the method for spiking the samples and the order in which samples were to be run. Results and Discussion Method 624 is acceptable for the analysis of purgeable priority pollutants. The accuracy of the method is judged very good while overall precision and single- analyst precision are considered accept- able. For most compounds, matrix does not significantly affect the analysis. Method 624 was characterized in terms of accuracy, overall precision, single- analyst precision, and the effect of water type on accuracy and precision through statistical analyses of 9,880 reported values. Estimates of accuracy and precision were made and expressed as regression equations, shown in Table 1 for each compound. The equations were based on the 8,446 data values remaining after eliminating 1,434 values (approxi- mately 15%) designated as outliers by the IMVS programs. The development and interpretation of these regression equa- tions are discussed in Section 5. To facilitate the interpretation of these equations. Table 2 was prepared. In Table 2, accuracy (percent recovery), overall precision (percent standard deviation), and single-analyst precision (percent standard deviation) were computed (using the regression equations) at a concentration of 100 fjg/L. Accuracy is obtained by comparing the mean recovery to the prepared values of the concentrations and computing the percent recovery. Overall, recoveries for the volatile organic compounds are very good for all of the water matrices with an average recovery of 100%. The mean recovery statistics (at 100 /ug/L) for the volatile organic compounds range from 68% fo'r bromomethane in the surface water matrix to 123% for cis-1,3- dichloropropene in the distilled water. One-half of the mean recoveries are between 94% and 105%, with one-fourth of the mean recoveries above and below these values. Recoveries for bromometh- ane are consistently low (ranging from 68% to 75%) for all water matrices. Mean recoveries for cis-1,3-dichloropropene and 1,2-dichloropropane are high with recoveries ranging from 116% to 123%, while the recovery of trans 1,3-dichloro- propene is uniformly low, averaging 83%. It is known that the isomers of 1,3- dichloropropene are relatively unstable ------- Tablt 1 . Regression Equations for Accuracy and Precision Water Type Benzene Bromodichloromethane Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy (10.8 - 480.01 SR = 0.26X- 1.74 S =0.25X- 1.33 X = 0.930 + 2.00 SR = 0.20X - 0.24 S =0.22X-0,75 X = 0.9SC + 1.40 SR = 0.15X-0.60 S =0.23X- 1.02 X = 0.94C + 1.88 SR = 0. 14X - 0.91 S = 0.22X-0.86 X = O.S9C + 1.60 (8.0 - 480.0) SR = 0.1 5X + 0.59 5 = 0.20X + 1. 13 X = 1.03C - 1.58 SR = 0.17X + 0.94 S = 0.20X + 3.76 X - 1.03C + 1.35 SR = 0. 18X + 0.43 S = 0.22X + 0.80 X = 1.00C - 1.02 SR = 0.23X-0.15 S = O.22X+1.01 X =0.940-0.93 Bromoform* (9.0 - 400.0) SR = 0.14X + 0.19 S = 0.20X + /. 18 X = 1.01C - 0.89 SR = 0.31 X + 1.36 S = 0.33X + 1.03 X =1. 13C - 1.07 SR = 0. 18X + 0.06 5 = 0.26X + 0.98 X =0.57C-0.67 SH = 0.28X - 0.02 5 = O.33 + 0.49 X = 0.95C - 1.65 Bromomethane (9. 1 - 6O7.O) SR = 0.27X - 0.50 S = 0.25X + 0.64 X =0.720-0.79 SR=0.29X -0.45 S = 0.34X + 0.57 X =0.690 - 1.14 SR=0.24X -0.15 S = 0.25X + 0.67 X =0.690-0.51 SR=0.37X -0.21 S = 0.41 X- 0.07 X =0.760-0.80 X = Mean Recovery C = True Value for the Concentration 'Revised regression equations and estimates of accuracy and precision are given in Table 2 Water Type Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy Carbon Tetrachloride* (9.0 - 400.0) SR = 0.11X^0.35 S = 0.14X + 0.17 X = 1.01C - 0.84 SR = 0.23X - 0.85 S = 0.24X - 0.87 X = 1.070 - 1.66 SR = 0. 16X + 0.98 S =0.19X^0.95 X = 1.010 - 0.22 SR = 0.20X - 0.29 S =0.20X^0.54 X =0.950-0.62 Chlorobenzane (13.5 - 600.0; Sfl = 0.;6X-0.09 S =0.26X-1.92 X =0.980 + 2.28 SR = 0.1 9X + 0.69 S = 0.22X - 0.30 X = 1.020 + 2.04 SR = 0.19X-0.81 S =0.29X-2.60 X =1.010 + 2.91 SR = 0.23X + 0. 13 S = 0.36X - 2.20 X =0.920 + 2.36 Chloroethane* (7.3 - 488.0) SR = 0.23X + 2.02 S =0.27X+1.95 X = 1.080 + 1.50 SR = 0.31 X- 0.71 S = 0.35X + 0.04 X =1.100 + 0.13 SR = 0.22X+ 1.63 S =0.28X+1.47 X = 1.090 + 1.83 SR = 0.32X + 0.25 S =0.38X-0.21 X =1. 120 + 0.44 Chloroform (4.5 - 300.0) SR=0.16X + 0.22 s =0. isx + o. re X =0.930 + 0.33 SR = 0.23X + 0.42 S = 0.31 X + 5.58 X =0.870 + 5.78 SR = O.22X - 0.30 S =O.23X-0.08 X =0.910 + 0.65 SR=0.14X + 0.33 S =0.18X + 0.65 X =0.940 + 0.37 X = Mean Recovery C = True Value for the Concentration "Revised regression equations and estimates of accuracy and precision are given in Table 2. Water Type Chloromethane* CIS-1,3,-Dichloropropene Dibromochloromethane Ethyl Benzene Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy (7.O - 469.0) SR = 0.41X+ 1.75 S =0.48X+1.21 X =0.940 + 2.37 SR = 0.43X + 0.09 S = 0.45X - 0.21 X =O.O9C + O.20 SR = 0.37X - 0.46 S =0.45X + 0.55 X =1.120 - 0.56 SR = 0.59X - 1.33 S =0.61X- 1.10 X = 1.020 - 0.38 (8.0 - 357.0) SR = 0.19X +0.44 S = 0.24X + 0.07 X = 1.240 - 0.55 SR = 0.21 X +0.38 S = 0.27X + 0.55 X = 1.210 - 0.47 SR = 0.26X - 0.90 S =0.32X-0.33 X =1.160 + 0.18 SR = 0.15X +0.33 S =0.25X + 0.01 X = 1.20C - 0.44 (8.1 - 360.0) SR = 0.17X-0.18 S =0.17X + 0.49 X =1.010-0.03 SR = 0.23X - 0.24 S = 0.26X + 0.88 X = 1.070 - 0.44 SR = 0.20X -0.39 S = 0.21 X - 0.18 X =1,010 + 0.10 SR = O.I8X-0.38 S = 0.26X - 0.87 X = 1.070 - 0.70 (15.O - 68O.O) SR = 0.14X + 1.00 S =O.26X- 1.72 X = 0.98C + 2.48 SR = 0.22X + 0.90 S =O.24X-0.77 X =0.990 + 2.97 SR = 0.15X + 0.38 S =0.22X-1.25 X = 1.010 + 3.88 SR=0.24X + 0.03 S =0.29X- 1.27 X =1.010 + 3.73 X = Mean Recovery C = True Value for the Concentration ^Revised regression equations and estimates of accuracy and precision are given in Table 2 ------- Table 1 (continued) Water Type Methylene Chloride'" Tetrach/oroethene Toluene Trans- 1 ,2-Dichloroethene* Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy (7.2 - 480.0) SR = 0.19X+0.76 S =0.30X+4.09 X =0.810 + 2.31 SR = 0.26X+ 5.78 S =0.36X+5.37 X =0.730 + 5.97 SR = 0.16X+9.45 S =0.25X+7.91 X = 0.800 + 8.57 SR = 0.30X + 3.54 S =0.44X+ 1.94 X =0.710+3.15 (9.0 - 400.0) SR = 0,13X -0.18 S = 0.16X-0.45 X = 1.060 + 0.60 SR = 0.23X + 0.04 S = 0.27X-0.64 X =0.980 + 0.71 SR = 0.18X -0.22 S = 0.25X - 1.16 X = 1.020 + 1.54 SR = 0.27X + 0.54 S = 0.31X - 0.15 X =0.870+1.62 (13.5 - 600.O) SR = 0.15X- 0.71 S =0.22X- 1.71 X = 0.980 + 2.03 SR = 0.18 + 0.71 S =0.24X-0.66 X =0.980 + 2.76 SR = 0.15X - 0.03 S =0.23X- 1.67 X = 1.000 + 2.25 SR = 0.22X - 0.93 S =026X- 1.07 X =0.920 + 2.63 (4.5 - 300.0) SR = 0.16X + 0.03 S =0.19X + 0.13 X =0.980 + 0.30 SR = 0.17X +0.20 S =0.17X + 0.52 X =1.050-0.17 SR = 0.16X + 0.10 S =0.16X + 0.37 X =0.980 + 0.26 SR =0.21 X -0.09 S =0.23X + 0.07 X =0.960 + 002 X = Mean Recovery C = True Value for the Concentration "Revised regression equations and estimates of accuracy and precision are given in Table 2 Water Type Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy Trans-1,3-Dichloropropene (9.4 -416.0) SR = 0.20X - 0.53 S = 0.26X - 0.09 X = 0.800 + 0.22 SR = 0, 13X + 0.94 S = 0.25X + 0.23 X = 0.830 - 0.58 SR = 0. 15X + 0.03 S =0.24X + 0.18 X =0.890 + 0.69 SR = 0.18X -O.37 S =0.22X-0.48 X = 0.820 - 0.08 Trichloroethene (5.4 - 360.0) SR = 0.1 3X + 0.36 S = 0.1 2X + 0.59 X = 1.040 + 2.27 SR = 0.23X - 0.34 S =026X-0.28 X = 1.030 + 1.65 SR = 0. 14X + 1.05 S = 0.1 9X-+ 0.94 X = 1.030 + 2.91 SR = 0.22X + 0.75 S = 0.33X - 0.03 X =0.990 + 1.76 Trichlorofluoromethane" (7.2 - 480.0) SR = 0.31X- 1.34 S = 0.36X - 0.48 X = 0.920 + 0.83 SR = 0.18X + 0.66 S =0.31X-015 X = 0.980 + 0.34 SR = 0.28X - 0.30 S = 0.31 X + 0.02 X =0.850 + 0.70 SR = 0.24X - 1.36 S =0.28X-0.56 X = 1.OOC + 0.25 1. 1 -Dich/oroethane" (10.8 - 480.0) SR = 0. 15X - 0.22 S =0. 15X + 0.53 X =0.980 + 1.09 SR =0.16X -0.21 S =0.14X + 0.82 X =1.010 + 0.11 SR= 0.1 IX +1.07 S =0.12X+1.06 X =0.990+1.13 SR = 0.23X - 0.27 S = 0.24X + 0.84 X = 1.040 + 0.39 X = Mean Recovery C = True Value for the Concentration "Revised regression equations and estimates of accuracy and precision are given in Table 2 Water Type Applicable Cone. Range Distilled Water Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision A ccuracy 1, 1 -Dichloroethene" (7.2 - 480.0) SR = 0.22X + 0.58 S = 0.37X + O.24 X = 1.010 + 1.43 SR = 0.16X+ 1.73 S =0.23X + 0.60 X = 0,940 + 2.07 SR = 0. 14X + 0.95 S - 0.21 X + 0.69 X =0.950 + 1.38 SR = 0.23X - 0.35 S = 0.23X + O.24 X =0.840+1.57 1, 1, 1 -Trichloroethane (9.0-4000) SR = 0. 12X - 0. 15 S = 0.21 X- 0.39 X =1.060 + 0.73 SR = 0.20X - 0.54 S =0.23X-0.22 X =1.110-0.53 SR = 0.23X - 0.27 S =0.28X-0.82 X =1.010 + 0.31 SR = 0. 18X - 0.81 S = 0.24X - 0.55 X =0.990 + 0.83 1, 1 ,2-Trichloroethane (10.8 - 480.0) SR = 0. 14X + O.02 S =0.18X + O.OO X =0.950+1.71 SR = 0,12X+ 1.44 S = 0.15X + 0.74 X = 1.020 + 1.80 SR = 0.1 6X -0.27 S = 0.21 X - 0.84 X = 1.040 + 1.55 SR = 0.18X + O.05 S = 0.23X - 1.06 X = 1 000 + 1.04 1. 1,2.2-Tetrachloroethane (15.O - 680.0) SR=0.16X + 0.69 S =0.20X + 0.41 X =0.930+1.76 SR = 0. 16X + 0.30 S =0.25X-0.83 X = 0.920 + 0.94 SR = 0 14X+ 1.08 S = 0.20X + 1.53 X = 0.990 + 1.45 SR = 0.37X - 1.22 S =030X + 0.91 X =0.870+2.09 X = Mean Recovery C = True Value for the Concentration "Revised regression equations and estimates of accuracy and precision are given in Table 2 4 ------- Tahiti, (Continued/ Water Type 1,2-Dichlorobenzene/1,4-D 1,2-Dichloroethane Applicable Cone. Range (16.O - 780.O) Distilled Water Single-Analyst Precision SR = 0.22X - 1.45 Overall Precision S -0.3OX-1.20 Accuracy X = 0.94C + 4.47 Tap Water Single-Analyst Precision SR = 0.36X - 2.57 Overall Precision S -0.38X-1.56 Accuracy X = 0.98C+4.65 Surface Water Single-Analyst Precision SR - 0.25X + 0.85 Overall Precision S = 0.30X + 1.48 Accuracy X = 0.97C + 6.92 Industrial Effluent Single-Analyst Precision SR = Q.25X + 2.55 Overall Precision S = 0.29X + 4.32 Accuracy X =0.95C + 5.14 X = Mean Recovery C = True Value for the Concentration Table 2. Revised Regression Equations Water Type Applicable Cone. Range lug/L) Distilled Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy (9.9 - 440.0) SR = 0.17X -0.32 S = 0.21X-0.38 X = 1.02C + 0.45 SR = 0.18X-0.21 S = 0.17X^0.14 X = 1.06C 0.45 SR = 0. 15X + 1.01 S = 0.18X+1.69 X =1. 01C + 0.97 SR = 0. 14X + O.96 S = 0.18X+1.44 X = 1.01C 0.28 lor Accuracy and Precision Bromoform (9.0 - 400) SR = 0. 12X + 0.36 S = 0.17X+1.38 X =1. 18C - 2.35 SR = 0.23X + 2.06 S = 0.33X+1.01 X =1.320-2.74 SR = 0.1 4X + 0.38 S = 0.24X + 1.08 X = 1.10C- 1.80 SR = 0.18X + 0,65 S = 0.25X + 1.02 X = 1.06C - 2.67 1 ,2-Dichloropropane (13.5 - 600.0> SR = 0. 14X - 0.85 S = 0.17X-0.41 X = 1. 18C + 2.00 SR=0.10X + 0.95 S = 0. 13X + 0.53 X =1.16C+1.70 SR = 0.13X-0.52 S = 0.17X-O.33 X =1.18C + 2.89 SR = 0. 13X + 0.77 S =0. 18X + 0.53 X = 1.22C - 0.25 Carbon Tetrachloride (9.0 - 400) SR - 0. 12X + 0.25 S =0.1 IX + 0.37 X =1.100+1.68 SR = 0. 18X - 0.53 S =0.20X-0.61 X =1.180-2.66 SR = 0. 15X + 7.07 S =0.18X + 0.98 X =1.070-0.73 SR = 0. 19X - 0.23 S =0. 19X + 0.59 X = 1.OOC - 1.07 1. 3-Dichlorobemene (7.2 - 480.0) SR=0.14X-0.48 S = 0.1 8X- 0.82 X = 1.06C + 1.68 SR = 0.22X + 3.41 S = 0.24X + 2.34 X = 1.02C + 3.80 SR=0.15X + 1.44 S = 0.16X^1.72 X =1.1 1C + 1.90 SR = 0. 15X + 2.01 S = 0.17X^1.83 X =1.030 + 1.79 Chloroethane (7.3 - 488) SR=0.14X+2.78 S =0.29X+1.75 X =1.180 + 0.81 SR = 0.29X - 0.52 S = 0.34X + 0. 13 X =1.170-0.37 SR=O.25X+ 1.37 S = 0.28X + 1.46 X =1.120+1.63 SR = 0.32X + 0.25 S = 0.4OX - 0.37 X = 1.240 - 0.41 X = Mean Recovery C = Prepared Concentration Water Type Chloromethane Methylene Chloride Trans-1,2-Dichloroethene Applicable Cone. Range fag/L) Distilled Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy (7.0 - 469) SR = 0.37X + 2.14 S =0.58X + 0.43 X = 1.030 + 1.81 SR = 0.38X + 0.40 S =0.55X-0.79 X =0.960-0.20 SR = 0.32X - 0.05 S =0.49X + 0.27 X = 1.230 - 1.31 (7.2 - 480) SR = 0.15X + 1.07 S =0.32X + 4.OO X =0.870+1.88 SR = 0.2OX + 4.96 S =0.38X + 5.19 X =0.780 + 5.66 SR=0.27X + 8.17 S =0.29X+7.48 X = 0.830 + 8.40 (4.5 - 3OO) SR = 0.14X + 0.09 S =0.19X + 0.17 X = 1. ISO + O.03 SR= 0.1 IX + 0.49 S =0.15X + 0.60 X =1.110-0.40 SR=0.17X S =O.15X + 0.4O X = 1.020 + 0.05 ------- Table 2, (continued) Water Type Chloromethane Methylene Chloride Trans-1,2-Dichloroethene Industrial Effluent Single-Analyst Precision Overall Precision Accuracy SR = 0.61X- 1.43 S =0.58X-0.95 X = /. J3C - 1.23 SR = 0.30X + 3.56 S = 0.42X + 2.06 X =0.800 + 2.50 SR = 0.26X - 0.29 S = 0.19X + 0.22 X = 1.02C - 0.23 X = Mean Recovery C = Prepared Concentration Table 2. (continued) Water Type Trichlorofluoromethane 1,1 -Dichloroethane 1.1 -Dichloroethene Applicable Cone. Range (fjg/L) Distilled Single-Analyst Precision Overall Precision Accuracy Tap Water Single-Analyst Precision Overall Precision Accuracy Surface Water Single-Analyst Precision Overall Precision Accuracy Industrial Effluent Single-Analyst Precision Overall Precision Accuracy (7.2 - 480) SR-0.33X- 1.48 S = 0.34X-0.39 X = 0.990 + 0.39 SR = 0.17X + O.80 S = 0.29X + O.04 X =1.050-0.19 SR = 0.33X -0.57 S = 0.31 X + 0.03 X =0.870 + 0.50 SR = 0.27X + 1.62 S =0.26X-0.43 X = 1.070 - 0.29 (10.8 - 480) SR = 0.13X - 0.05 S = 0.16X + 0.47 X = 1.05C + 0.36 SR = 0.14X - 0.08 S =0.14X + 0.52 X = 1.070 - 0.53 SR = 0.11X + 1.08 S = 0.12X+1.12 X =1.020 + 0.76 SR = 0.23X -0.27 S = 0.21 X+ 1.12 X =1.090-0.12 (7.2 - 480) SR=0.17X+ 1.06 S =0.43X-0.22 X = 1.120 + 0.61 SR = 0.12X + 2.08 S =0.24X + 0.53 X = 1.020 + 1.43 SR=0.16X + 0.87 S =0.24X + 0.51 X =1.010 + 0.91 SR = 0.24X - 0.39 S = 0.20X + 0.39 X = 0.930 + 0.94 X = Mean Recovery C = Prepared Concentration and may decompose to 1,2-dichloropro- pane. The overall standard deviation of the analytical results is an indication of the precision associated with the measure- ment generated by a group of laboratories. The percent relative standard deviation (RSD) at 100 /t/g/L for the volatile organic compounds range from 13% for trichloro- ethene, 1,1-dichloroethane, and 1,2- dichloropropane in the various water matrices to 60% for Chloromethane in the industrial effluent with a median value of 24%. Precision for Chloromethane is relatively poor for all water matrices with percent relative standard deviations ranging from 45% to 60%. One-half of the RSDs are between 20% and 29%. In 95% of the cases the RSDs are less than 44%. The percent relative standard deviation for a single analyst (RSD-SA) indicates the precision associated within a single laboratory. The RSD-SA for samples at 100 /ug/L ranges from 11 % for carbon tetrachloride (distilled water matrix) and 1,2-dichloropropane (tap water matrix) to 58% for Chloromethane in the industrial effluent with a median RSD-SA of 19%. Single-analyst precision for Chlorometh- ane is relatively poor with RSD-SAs ranging from 37% to 58%. One-half of the RSD-SAs at 100 /jg/L are between 15% and 23%. In 95% of the cases, the RSD- SAs are less than 36%. Three compounds used in this study, bromomethane, Chloromethane and chloroethane, are gases in pure form. Although there are no clear trends for accuracy in the gaseous species as opposed to less volatile compounds, it is possible that the low recovery observed for bromomethane and the poor precision for all three compounds may be due to inherent difficulties in handling gaseous and extremely volatile compounds during the various preparation and analytical proce- dures required in the method. Bromo- methane is also known to be unstable, which could also account for low recover- ies. The effect of water type was different for the various volatile organic compounds. For most compounds the water matrix does not have a great effect on either the accuracy or precision. Overall, recoveries for the volatile organic compounds averaged 100% in distilled water, 101% in tap water and surface water, and 97% in the industrial effluent matrix. Precision (RSD and RSD-SA) for the volatile organic compounds ranged from a median RSD of 21% and median RSD-SA of 16% for the distilled water to a median RSD of 25% and a median RSD-SA of 23% for the industrial effluent matrix. A trend toward higher recoveries (above 100%) forthe lowest concentration Youden pairs was observed for 10 compounds. One explanation could be sample contamination from the presence of these compounds in the laboratory. Methylene chloride displayed the most pronounced example with recoveries averaging 142%, 76% and 83% for the low, medium and high pairs respectively. Low-level contamination may be respon- sible for the 142% recovery of the low pair. Blank concentrations were also higher for methylene chloride than for many of the other compounds, indicating a greater likelihood of low-level sample contamination. This explanation is less clear for other compounds. For example, the trend is more pronounced for the chlorobenzenes than for benzene or chloroform, yet the latter compounds would be expected to be more ubiquitous in a laboratory environment. A review of the data remaining after the IMVS outlier screening indicated some ------- potential problems with the data for nine of the volatile organic compounds. For these compounds, results for ampul four were out-of-line (usually due to extremely low recoveries) with the remaining data. It is suspected that during production of ampul concentrate four, these volatile compounds were lost. The data for these medium level ampuls were eliminated, and the equations revised. Table 2 presents the revised equations. Four compounds in addition to those listed in Table 2 had questionable regression equations although the equa- tions were not revised. These were bromomethane, cis and trans 1,3- dichloropropene and 1,2-dichloropropane. Bromomethane exhibited poor recoveries which may have been due to its extreme volatility or to its reactivity. The dichloro- propenes are known to be unstable and to form dichloropropane upon decomposi- tion. Problems with these compounds were also encountered with USEPA Quality Control Samples and in the Interlaboratory Study for Method 601 Halogenated Purgeables by GC. This Project Summary was authored by staff of Radian Corporation, P. 0. Box 9948. Austin. JX 78766 Raymond Wesselman and Bob Graves are the EPA Project Officers (see below). The complete report, entitled "EPA Method Study 29, Method 624Purgeables," (Order No. PB 84-209 915; Cost: $20.50, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA22161 Telephone: 703-487-4650 The EPA Project Officers can be contacted at: Environmental Monitoring and Support Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 *USGPO: 1984-759-102-10632 ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 BULK RATE POSTAGE & FEES PAID EPA PERMIT No. G-35 Official Business Penalty for Private Use $300 ------- |