United States Environmental Protection Agency Environmental Monitoring Systems Laboratory Research Triangle Park NC 27711 Research and Development EPA-600/S4-83-008 May 1983 Project Summary Validation and Improvement of EPA Reference Method 25- Determination of Gaseous Nonmethane Organic Emissions as Carbon G. B. Howe, S. K. Gangwal, and R K. M. Jayanty EPA Reference Method 25 for mea- surement of total gaseous nonmethane organics as carbon in source emissions is being evaluated. Details are given of the evaluation of a commercial non- methane organic analyzer (Byron Model 401); design, construction, and testing of a sample collection and conditioning system; and two field tests of the method using dual identical sampling trains at a textile plant and a plywood veneer plant. Recommendations are made to improve and modify the method. This Project Summary was developed by EPA's Environmental Monitoring Systems Laboratory. Research Triangle Park, NC, to announce key findings of the research project that is fully doc- umented in a separate report of the same title (see Project Report ordering information at back). Introduction On October 5, 1979, under Section III of the amended Clean Air Act the U.S. Environmental Protection Agency (EPA) proposed standards limiting the emissions of volatile organic compounds (VOC) from new, modified, and reconstructed auto- mobile and light-duty surface-coating operations within assembly plants. The standards were based on EPA's determina- tion that such emissions contribute signif- icantly to air pollution by acting as pre- cursors for formation of ozone and other photochemical oxidants that impact ad- versely on health and welfare. The proposed standards include a method for measuring VOC emissions—EPA Ref- erence Method 25. Method 25 involves first withdrawing an emission sample from the source stack through a chilled conden- sate trap and into an evacuated sample tank The total gaseous nonmethane organics (NMO) are then determined by combining the results obtained from separate analyses of the condensate trap and sample tank fractions. The organic contents of the condensate trap are oxidized and quantita- tively collected in an evacuated vessel, and an aliquot of the resulting C02 is reduced to methane and measured by a flame ionization detector (FID). An aliquot of the gas collected in the sample tank is intro- duced into a gas chromatographic (GC) system to separate the nonmethane or- ganics from carbon monoxide (CO), carbon dioxide (C02), and methane (CH4). After this separation, the NMOs are oxidized to C02, reduced to CH4, and measured by an FID. In this way, variations in FID response to different compounds are eliminated, and all carbon is measured as methane. Application of the method revealed de- ficiencies that affect the accuracy and reproducibility of the results. As a result the present investigation was undertaken to systematically evaluate EPA Reference Method 25 and to recommend further modifications to improve the accuracy, precision, and reliability of the data collected In addition to a laboratory evaluation, two field tests of the method were conducted to determine the quality of the data that can be expected using Method 25. ------- Results and Discussion All necessary system components for constructing the sampling, condensate recovery, and conditioning systems were obtained and assembled. Byron Model 401 was chosen for NMO and C02 analysis because of its availability. Complete experimental details of sys- tem design and construction are available in the project report and are not repeated here. Laboratory evaluation consisted of separate evaluations of the sampling sys- tem, the NMO analyzer (Byron Model 401), and the condensate recovery and conditioning system. Two field tests were also conducted, one at a textile plant and the other at a veneer plywood plant In the laboratory evaluation, the sampling system yielded a recovery of 99.9 pe'rcent and a relative standard deviation of 0.08 percent for a dilute propane in air mixture sampled through a manifold. The accuracy of the dead volume data of the sampling train was a point of concern. The error is directly proportional to the dead volume of the sampling system from the probe tip to the sampling container and inversely pro- portional to the volume of the sampling container. However, the error can be determined experimentally and a correction applied, as was done during the second field test. The laboratory evaluation of the NMO analyzer showed excellent linearity for some organic compounds over concen- tration changes of as much as three orders of magnitude, but the responses to dif- ferent organics tested ranged over a factor of three, as shown in Table 1. Possible reasons for unequal responses are incom- plete oxidation (carbon laydown) on the catalyst modules, interaction of the air carrier gas with the organic on the GC columns, and severe tailing of compounds resulting in inaccurate integration. The Model 401 was linear for C02 measure- ment over 2.5 orders of magnitude. The implication of the above findings was that the NMO would yield more accurate results for organics that are trapped (and eventually converted to C02 prior to measurement) rather than for organics that are measured in the gaseous form from the sampling container. A modified NMO analyzer that would alleviate the problem of unequal carbon responses was designed and briefly described in the report This analyzer will be built and evaluated in future studies. Laboratory evaluation of the condensate recovery and conditioning system was carried out according to procedures speci- fied by EPA Method 25. The system passed the requirements of the method with a system C02 blank of 6.1 ppm. Table 1. Variation of the Byron Model 40 J Analyzer Response to Various Organic Compounds Compound Trichloroethylene Ethylene Methyl acetate Ethane Freon 1 13 Carbontetrachloride Isopropyl alcohol Decane Acetyl acetone Nonane Amyl acetate Tetrahydrofuran Tetrahydropyran Naphthalene Propane Toluene Hexane Methanol Propylene Benzene Concentration range, ppmC 10 - 50 - 600 - 150 - 60 - 60 - 40 40,000 1,300 1,400 30 20 300 260 400 455 350 500 60 25 9,000 128 330 100 30 600 Relative response" 603 473 215 608 652 859 365 864 428 825 787 436 451 721 470 805 429 272 463 647 *Area counts/ppmC catalyst oxidation efficiency of greater than 96 percent, and carbon recovery of greater than 90 percent for hexane and toluene subjected to trapping in both liquid (equivalent C02 < 10,000 ppm) and gaseous forms. Two field tests were conducted, one at a textile plant and the other at a veneer ply- wood plant. Identical dual sampling sys- tems were used to test the precision of the method under conditions of actual appli- cation. During each test, four sets of duplicate samples were collected. During the first field test, a 31-percent relative standard deviation of the pooled values was obtained. Several modifications were made to the sampling train and procedure for the second field test, which resulted in a pooled relative standard deviation of 8 percent The most significant modifications included the use of a needle valve-rotameter combination for maintaining constant flows during the test period (instead of just the needle valve as used in the first field test) and a modified leak-check procedure with a dead volume correction. In addition, two condensate traps in series, one at ice water temperature and the other at dry ice tem- perature, were used to circumvent the possible problem of moisture freeze-out in the dry ice trap. The best precision of results from dual sampling trains was ±1.45 percent for sample #4 from the second field test The carbon concentration in the polyester cloth drier exhaust ranged from 200 to 750 ppmC and that in the veneer drier exhaust ranged from 4,700 to 6,700 ppmC. Conclusions Laboratory evaluations of EPA Reference Method 25 revealed that recoveries of organic gases exceeded 90 percent when several standardized mixtures of these gases in air or nitrogen were analyzed. At first, poor precision was encountered in field testing. When dual trains were employed to sample a fabric finishing plant effluent stream, a 31 -percent relative standard deviation (RSD) was obtained for the pooled values. These results prompted modification of the sampling trains, im- provement of the condensible organics recovery procedure, and introduction of a sampling train dead volume correction. When these techniques were employed in the sampling of the effluent from a plywood veneer dryer using dual sampling trains, the pooled values yielded an 8-percent RSD. The Byron Model 401 analyzer showed excellent linearity for a variety of organic compounds, with the linear range extend- ing over three orders of magnitude in some instances. However, the response to samples of different compounds con- taining equal ppmC was found to be different. Incomplete oxidation of organic compounds to C02 within the analyzer and peak tailing appear to be among the many causes that may have contributed to the unequivalent response. This response variability is expected to result in inaccu- racies for measurement of the volatile por- tion of the sample that is not captured by the condensate trap and ends up in the sample tank. The variability of response. ------- however, should have no effect on the precision of results for samples collected via the dual sampling trains. Recommendations Many of the problems with the Byron Model 401 analyzer might be alleviated to some extent by constructing an improved NMO analyzer. 1. Using only one column in tho Byron Model 401 instead of the three may substantially reduce the peak tailing problems by reducing the dead volume. 2. A more efficient catalyst, such as the hopcalite oxidizing catalyst used in the liquid condensate recovery and condi- tioning system instead of the proprie- tary catalyst used in the Byron instru- ment may solve the problem of re- sponse variability associated with in- complete oxidation of organics. 3. Finally, an inert carrier gas such as helium or nitrogen used in the chromat- ographic portion of the analyzer instead of air may result in reduced interaction with the column material. A 60/80 'mesh Porapak-N (Waters Associates) column could be used to separate the NMO from C02 and other gases. The column would first elute C02 then the .N MOs would be backf lushed out of the column into an oxidation-reduction sys- tem and finally to an FID. This design would be much simpler than that origi- nally shown in Method 2 5 as described in the Federal Register. The use of a Heise bourdon tube pressure gauge in place of the U-tube mercury manometer is also recommended. Its advantages include ease of use, portability, and ruggedness and it is adequately ac- curate for making sample tank pressure measurements. The reuse of the sample tank as the intermediate collection vessel during con- densate recovery and conditioning would eliminate the necessity of measuring tank volumes and the associated error. This action is possible because the intermediate collection vessel volumn (Vv) and the sample tank volume (Vs) cancel out in the calculation of condensible organic con- centrations (Cc) if these volumes are equal for Method 25 equation derivations. Before the sample tank is reused, however, it should be flushed adequately with clean air to remove any remaining CC^. The Federal Register procedure speci- fies that during recovery of the condensate trap sample, the dry ice must be removed from the trap before switching the carrier gas flow through the trap. This method can create problems of clogged rotameters and sample loss because the trap gases begin to expand when the dry ice is removed. If the carrier gas is switched into this pressurized trap, water vapor or or- ganics could expand back into the rotam- eters. Therefore, the carrier gas should be passed through the condensate trap before the dry ice is removed. G. B. Howe, S. K. Gangwal. and R. K. M. Jayanty are with Research Triangle Institute, Research Triangle Park, NC 27709. Joseph E. Knoll is the EPA Project Officer (see below). The complete report, entitled "Validation and Improvement of EPA Reference Method 25—Determination of Gaseous Nonmethane Organic Emissions as Carbon," (Order No. PB 83-191 00 7; Cost: $11.50. subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Environmental Monitoring Systems Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Postage and Fees Paid Environmental Protection Agency EPA 335 Official Business Penalty for Private Use $300 PS 0000329 U S ENViR PROTECTION AGENCY REGION 5 LIBRARY 230 S DEARBORN STREET CHICAGO IL 60604 ------- |