RTI/2218/35-01 STATUS REPORT #4 STABILITY OF ORGANIC AUDIT MATERIALS AND RESULTS OF SOURCE TEST ANALYSIS AUDITS R. K. M. Jayanty W. F. Gutknecht C. E. Decker EPA Project Officers: EPA Contract No.: Joseph E. Knoll Darryl J. von Lehmden 68-02-3431 Prepared for U.S. Environmental Protection Agency Environmental Monitoring Systems Laboratory Quality Assurance Division Research Triangle Park, North Carolina 27711 June 1982 RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709 ------- TABLE OF CONTENTS SECTION NO. PAGE 1.0 INTRODUCTION 1 1.1 Objectives 1 1.2 Audit Materials Contained in the Repository ... 1 2.0 EXPERIMENTAL PROCEDURES 5 2.1 Instrumentation 5 2.2 Calibration 5 3.0 PERFORMANCE AUDITS 8 4.0 STABILITY STUDIES. 19 5.0 SUMMARY AND CONCLUSIONS 21 REFERENCES 22 ATTACHMENT 1 - STABILITY DATA AS OF MAY 1982 23 ATTACHMENT 2 - SAMPLE CALCULATIONS OF PERCENT CHANGE/MONTH . . 70 ------- SECTION 1.0 INTRODUCTION 1.1 OBJECTIVES The need for reliable standards for source emission measurement of hydrocarbons, halocarbons and sulfur compounds is well established. The Research Triangle Institute (RTI) under contract to the U.S. Environmental Protection Agency (USEPA) has responded to this need through development of an extensive repository of gaseous compounds. The main objectives of this ongoing project are (1) to provide gas mixtures to EPA, state/local agencies, or their contractors, as performance audits to assess the relative accuracy of source emission measurements in certain organic chemical manufacturing industries, (2) to corroborate the vendor's certified analysis of the gas mixtures by in-house analysis, (3) to determine the stability of the gas mixtures with time by in-house analysis, and (4) to explore the feasibility of new audit materials as requested by EPA. This report describes the present status of this project. Includ- ed in the report are (1) a description of the experimental procedures used for initial cylinder analyses and collection of stabilty data, (2) a description of the audit procedure, (3) presently available audit results, and (4) presently available stability data. Full details of the study with additional statistical analysis will be presented in the final report and also in a journal publication. 1.2 AUDIT MATERIALS CONTAINED IN THE REPOSITORY The RTI repository currently contains 40 different compounds for use in conducting performance audits during source testing. The compounds were selected based on the anticipated needs of the personnel of the Emissions Measurements Branch, Office of Air Quality Planning and Standards, USEPA. Table 1 lists the compounds, the concentration ------- ranges, the number of cylinders of each compound, and the cylinder construction material. In Table 1, the audit materials fall into two concentration ranges. The low concentration range between 5 and 20 parts per million (ppm) simulates' possible emission standard levels. The high concentration range between 50 and 700 ppm simulates expected source emission levels. The balance gas for all gas mixtures is nitrogen. In the case of some of the audit materials, a second compound, which serves as an internal standard, was added to the gas mixture. 2 ------- TABLE 1. AJDIT MATERIALS CLRRENTLY HELD IN THE REPOSITORY Lew Concentration Range High Concentration Range Canpomd No. of Concentration Cylinder No. of Concentration Cylinder Cylinders Range (ppni) Construction* Cylinders Range (ppn) Construction* Benzene 14 8 - 13 S 17 60 - 400 Al Ethylene 4 5 - 20 A1 4 300 - 700 Al 6 3000 - 20,000 Al Propylene 4 5 - 20 A1 4 300 - 700 Al Methane/Ethane - - 4 1000 - 6000(M), Al 200 - 700(E) Propare 4 5-20 A1 4 300 - 700 Al Toluene 2 5 - 20 S 2 300 - 700 S Hydrogen Sulfide 4 5 - 20 A1 2 300 - 700 Al MetohXylene 2 5 - 20 S 2 300 - 700 LS Methyl Acetate 2 5 - 20 S 2 300 - 700 S Chloroform 2 5 - 20 S 2 300 - 700 S Carbonyl Sulfide 2 5 - 20 S 2 100 - 300 S Methyl Mercaptan 4 3 - 10 Al - - Hexane 2 20 - 80 A1 2 1000 - 3000 LS 1,2-Dichloroethane 4 5 - 20 Al 4 100 - 600 Al Cy:lohexane - - 1 75 - 200 S Methyl Ethyl Ketone 1 30 - 80 S - - Methanol 1 30-80 Al - - 1,2-Dichloropropane 2 5 - 20 Al 2 300 - 700 LS Trichloroethylene 2 5-20 Al 2 100 - 600 Al ------- TABLE 1. AUDIT MATERIALS CLRRENTLY HELD IN THE REPOSITORY (Continued) Lew Concentration Range High Concentration Range Compound Ni. of Cyl inders Concentration Range (ppm) Cylinder* Construct ion to. of Cylinders Concentrat ion Range (ppn) Cylinder* Construct ion 1,1-Dichloro- ethylene 2 5 - 20 A1 2 100 - COO Al 1,2-Dibrcmo- ethylene 2 5 - 20 A1 2 100 - 600 A1 Perdiloro- ethylene 2 5 - 20 S 2 300 - 700 LS Vinyl Chloride 9 5-30 S - — 1,3-Butadiene 1 5 - 30 S - — Acrylonitrile 2 5 - 20 LS 2 300 - 700 LS Aniline 1 5 - 20 A1 - — Methyl Isobutyl Ketone 1 5 - 20 A1 1 50-100 Al Cyclohexanone 2 5 - 20 A1 - — Par a-d ichl orobenzene 2 5 - 20 S - — Ethyl anine 2 5 - 20 A1 - — Formaldehyde - — - - — Methylene Chloride 1 5 - 20 A1 - — Carbon Tetrachloride 1 5 - 2D A1 - — Freon 113 1 5-20 A1 - — Methyl Chloroform 1 5 -20 A1 - — Ethylene Oxide 1 5 -20 A1 - — Propylene Oxide 1 5 - 20 A1 1 75 - 200 Al Allyl Chloride 1 5 - 20 S 1 75 - 200 S Acrolein 1 5 - 20 A1 1 75 - 200 Al fUl Isll IW* UJO I/.OIC i X C _ Tfl *J l_vy Al - — Carbon Disulfide - — A1 - 75 - 200 Al *Cylinder construction: A1 - Aluninun, S - Steel, LS - Lew Pressure Steel ------- SECTION 2.0 EXPERIMENTAL PROCEDURES Analysis of the cylinder gases is required to corroborate the concentrations reported by the company which prepared the gas mixtures and also to measure concentration changes with time, that is, estimate stability of the standards. 2.1 INSTRUMENTATION Analyses are performed with (1) a Perkin-Elmer Model 3920B Gas Chromatograph with flame ionization and flame photometric detectors, and (2) a Tracor Model 560 Gas Chromatograph with a flame photometric detector. The Tracor instrument has been used principally for measurement of the sulfur-containing species. The gaseous samples are injected onto the columns by means of gas sampling valves constructed of Hastalloy C (high nickel content and low adsorptive properties). These valves are equipped with interchangeable sample loops to allow the injection of variable but known volumes of gas. To further facilitate the injection of varying sample sizes, a sample injection system (Figure 1) is employed. The operation of the system is based upon measurement of pressure differentials. Further details on the system are published in the open literature (1). The gas chromatographic parameters used in the measurement of individual compounds and problems that have arisen are listed in Attachment 1. 2.2 CALIBRATION Calibration of the gas chromatographs has involved measurement of known concentrations of gases in air or nitrogen. The source or method of preparation of calibration standards varies depending on the gas involved. National Bureau of Standards, Standard Reference Materials (NBS-SRM's) of methane and propane are used for the calibration of the GC for the measurement of methane and propane audit materials. These 5 ------- HELIUM O VALCO HASTALLOY-C _nl. VALVE \ T0 GC A/ ~ \ O SAMPLE GAS BOTTLE TEFLON VALVE C r 1 3 HEISE VACUUM - PRESSURE GAUGE NUPRO STAINLESS STEEL TOGGLE VALVES Q VACUUM PUMP Figure 1. Sample Injection system. ------- same gases are used to calibrate the chromatographic system for measurement of ethylene and propylene, assuming the FID response per carbon is constant from compound to compound. A second method for the quantitation of gaseous compounds involves the use of permeation tubes. Thus, the calibration gases for vinyl chloride and ethylene oxide have been generated using permeation tubes. The tube is placed in a temperature-controlled chamber and zero air is passed over the tube at a known flow rate. The resultant gaseous mixture is further diluted if necessary with additional zero air in a glass dilution bulb. The final mixture is collected in a gas sampling bulb or a Tedlar® bag and analyzed by GC-FID. The permeation rates of the tubes are determined periodically by weight loss. A third method for developing a standard is the "glass bulb" technique. A known volume of the compound, either gas or liquid, is injected into an evacuated glass bulb of known volume. The bulb is then returned to atmospheric pressure with a balance gas of choice. If a pure liquid is injected, total vaporization is assumed and the concentration is calculated by using the ideal gas law. Additional dilutions are also made, if necessary, by utilizing additional bulbs or by repeatedly pressurizing with a balance gas to a known pressure and then partially evacuating to a known pressure. With each of these approaches, multipoint calibration curves are prepared each time a sample is analyzed. Certain quality control procedures are followed, for example, equilibrating the permeation system and the glass bulbs with the sample gas before taking an aliquot for GC measurement. Also, an NBS standard cylinder of methane is used to verify the constancy of the detector response. Blank measurements are taken during the process of cylinder analysis and generally, blank measurements have shown no signal above the baseline. 7 ------- SECTION 3.0 PERFORMANCE AUDITS RTI supplies repository cylinders for audits upon request from EPA, state or local agency or a contractor. The contractor must be performing source emission tests at the request of EPA, state or local agency in order to qualify for the performance audit. When a request is received, the contents of the cylinders are analyzed, the tank pressures are measured and the cylinders are shipped by overland carrier. Tank regulators are also provided when requested. A letter is included with the cylinders which provides general instructions for performance of the audit. The audit concentrations and cylinder pressures are provided to the requesting agency audit coordinator. To date, 86 individual audits have been initiated, and 83 are complete. The audit results collected to date are presented in Table 2. Generally, the results of the audits show close agreement (+^ 10%) with the actual cylinder concentrations measured prior to shipment. 8 ------- TABLE 2. SUMWRY OF PERFORMANCE AUDIT RESULTS Audit RTI audit Client audit Status of No. CI ient Industry Audit material conc. (ppm) % bias (Avg.) audit 1 A Ethylene oxide product ion Ethylene in N2 Ethylene in N2 3,239 21,226 -22.5 -20.0 E 2 A Ethylene oxide production Methane/ethane in N2 Methane/ethane in N2 l,710Me/220Et 8,13CMe/597Et +9.00/-20.0 +9.00/-1.00 E 3 A Ethylene oxide production Methane/ethane in N2 Methane/ethane in N2 l,021Me/315Et 6,207Me/773Et +21.5/-4.50 +23.5/-4.50 E 4 A Acetone product ion Benzene in ^ Benzene in N2 79.0 374.0 -19.0 -11.0 E 5 A Maleic anhydride production Benzene in N2 Benzene in N2 138 300 -9.40 +4.70 E 6 A Ethylene oxide production Ethylene in N2 Ethylene in N2 5,442 18,918 -27.0 -33.0 E 7 B Maleic anhydride production Benzene in N2 Benzene in N2 80.0 355 +2.30 +27.5 E 8 C Maleic anhydride production Benzene in N2 Benzene in N2 101 387 +12.9 +14.5 E 9 D Ethyl benzene styrene manufacturer Benzene in N2 Benzene in N2 71.0 229 -2.80 -3.90 E 10 E Gasoline bulk terminal Benzene in N2 Benzene in N2 62.0 80.0 +3.80 +3.40 E 11 F Gasoline transfer Benzene in N2 terminal Benzene in N2 142 294 -3.50 +3.20 E 12 F Gasoline transfer Benzene in N2 terminal Benzene in N2 268 343 -11.8 -1.00 E ------- TABLE 2. SUfWRY OF PERFORMANCE AUDIT RESULTS (Continued) Audit No. Client Industry Audit material RTI Audit conc. (ppm) Client Audit % bias (Avg.) Status of audit 13 F Gasoline transfer Benzene in N2 terminal Benzene in N£ 129 318 I 1 O O T? E 14 F Gasoline transfer Benzene in N2 terminal 10.7 +2.60 E 15 C Nitrobenzene manufacturing Benzene in No Benzene in N^ 9.73 269 -4.60 -2.60 E 16 F Gasoline bulk terminal Benzene in N2 Benzene in N2 8.20 140 -2.30 -1.80 E 17a F Gasoline bulk terminal Benzene in N2 Benzene in N2 9.50 127 +10.4 -2.80 E 17b F Gasoline bulk terminal Benzene in N2 Benzene in N2 9.50 127 +12.5 -6.30 E 18 Coke a/en Hydrogen sulfide in Hydrogen sulfide in n2 n2 7.05 9.73 -24.8 -22.9 E 19 F Gasoline bulk terminal Benzene in N2 Benzene in N2 12.0 218 -0.80 +7.30 E 20 F Gasoline bulk terminal Benzene in N2 Benzene in N2 7.65 396 +16.3 +1.50 E 21 F Linear alkyl- benzene manu- facturing Benzene in N2 Benzene in N2 Benzene in N2 98 294 331 +5.70 +6.80 +4.50 E 22 F Gasoline bulk terminal Benzene in N2 Benzene in N2 9.85 81.0 -4.10 -6.80 E 23 F Gasoline bulk terminal Benzene in N2 Benzene in N2 10.2 61.0 +4.60 -9.50 E ------- TABLE 2. SUMQRY OF PERFORMANCE AUDIT RESULTS (Continued) Audit RTI audit Client audit Status of No. Client Industry Audit material cone, (ppm) % bias (Avg.) audit 24 H Industrial surface coating process To 1 iene in N2 Propylene in N2 Propane in N2 Methane/ethane in N2 14.8 474 20.3 l,640Me/195et -1.90 +0.20 -2.30 -13.5(as methane) E 25 C Acrylic acid and ester Production Propane in N2 Propane in N2 10.1 710 +8.60 +5.60 E 26 C Acrylic acid and ester Production Propane in N2 Propane in N2 5.10 607 +-17.6 -3.60 E 27 E Maleic anhydride Benzene in N2 Benzene in 10.2 218 NA m F 28A A Carbon adsorber Toluene in N2 Toluene in N2 8.55 405 -6.40 -1.00 E 288 A Carbon adsorber Toluene in N2 Toluene in N2 8.55 405 +4.10 NA E 28C A Carbon adsorber Toluene in N2 Toluene in N2 8.55 405 -8.80 m E 29 EPA, QAD Instrunent check-out Ethylene in N2 Ethylene in N2 Ethylene in N^ Ethylene in N2 Ethylene in n£ 4.75 19.6 312 3021 20456 +4.00 +3.10 -0.80 +5.30 -8.60 ------- TABLE 2. SUfWRY OF PERFORMANCE AUDIT RESULTS (Continued) Audit kll Audit Client Audit Static of No. Client Industry Audit material conc. (ppm) % bias (Avg.) audit 30 EPA, QAD Instrunent check-out Benzene in ^ Benzene in N2 Benzene in N2 Benzene in N2 8.20 78.0 133 348 -K).30 -0.90 -4.00 -0.90 E 31 EPA, QAD Instrunent check-out Toluene in Toluene in N£ 405 579 +3.20 +1.00 E 32 EPA, QflD Instrunent check-out Methyl acetate in N Methyl acetate in N Methyl acetate in N Methyl acetate in N 6.80 17.2 326 455 -2.60 +1.70 -1.50 -1.30 E 33 EPA, QAD Instrunent check-out Propylene in N2 Propylene in N2 Propylene in N2 Propylene in N2 4.90 19.7 300 685 -22.4 -7.80 +1.00 -1.80 E 34 EPA, QAD Instrunent check-out Propane in ^ Propane in N2 Propane in N2 14.6 303 439 -0.70 +7.60 +6.20 E 35a I Vegetable oil pi ant Hexane in N2 Hexaie in N2 82.2 1982 +8.10 +3.00 E 35b I Vegetable oil plant Hexane in N2 Hexane in N2 82.2 1982 -1.20 -1.30 E 36 A Carbon adsorber Toluene in No 8.20 -2.40 E ------- TABLE 2. SUMQRY OF PERFORMANCE AUDIT RESULTS (Continued) Audit RTI audit Client audit Status of No. Client Industry Audit material conc. (ppm) % bias (Avg.) audit 37 B Coke a/en Benzene in N2 Benzene in N2 12.1 105 ¦•0.80 +2.90 E 38 D Ethylbenzene/ styrene Benzene in N2 Benzene in N2 Benzene in N2 9.90 77.9 345 +5.70 +3.60 +1.50 E 39 B Coke a/ en Byproduct Benzene in N2 Benzene in N2 8.20 85.4 -2.60 -8.70 E 40 D Coke even Byproduct Benzene in N2 Benzene in N2 10.9 147 +20.0 +6.80 41 H Paint spray Benzene in N2 m-Xylene in N2 10.8 15.4 NA m F 42 H Tire manufacturing Cyclohexane in N2 93.4 -11.1 D 43 B Coke o/en Benzene in N2 Benzene in N2 7.54 225 +0.10 +0.40 D 44 D Ethyl benzene/ styrene Benzene in N2 Benzene in N2 Propane in N2 8.20 74.5 10.6 -3.40 -0.20 -3.00 D 45 F Industrial sir face coating Propane in Air Propane in Air 316 450 -3.20 -2.00 E 46 EPA, QAD Tire manufacturing Propane in Air Propane in Air 15 316 m m F ------- TABLE 2. SUWWRY OF PERFORMANCE AUDIT RESULTS (Continued) Audit No. Client Industry Audit material RTI audit conc. (ppm) Client audit % bias (Avg.)* Status of audit** 47 EPA, QAD Tire manufacturing Propane in air Propane in air 20.8 453 -18.4 +13.4 E 48 D Dimethyl terephthal ate production Meta-Xylene in air 487 -2.10 E 49 EPA, QAD Instrunent check-out Toluene in N2 Methanol in N2 61.5 55.2 m F 50 EPA, QAD Tire oven manufacturing Propane in air Propane in air Propane in air 4.90 613 718 -48.8 +16.9 +16.8 E 51 EPA, QAD Instrument check-out Propane in air Propane in air 20.8 316 +20.0 -9.20 E 52 D Styrene manufacturing Benzene in N2 Benzene in 1,3-Butadiene. in N2 106 358 20.9 -4.90 -3.70 +23.8 E 53 I Veg. oil manufacturing Cyclohexane in N£ 99.0 -3.50 E 54 M Research Chloroform in N2 Chloroform in N£ 16.5 531 NA NA F 55 J Ethylene in ^ 300 +1.40 E ------- No. Client TABLE 2. SUMWRY OF PERFORWCE AUDIT RESULTS (Continued) RTI audit Client audit Status of Industry Audit material conc. (ppm) % bias (Avg.)* audit** 56 57 58 59 EPA, QAD Reactivity of vent activated charcoal Instrunent check-out Coil coating Chloroform in No Hydrogen sulfide in N2 Propane in N2 Propane in Benzene in N2 Benzene in No 8.11 16.2 5.20 472 9.45 341 m m m -8.40 M m 60 M 61 EPA (State of Conn.) 62 0 63 64 M Audit not initiated Benzene in N2 Meta-xylene in N2 Hexaie in N2 Methyl mercaptan in N2 Benzene in N2 Methyl ethyl ketone in No B3 760 1986 4.44 13.4 44.5 W m m w m F F F F 65 Coke even Byproduct Recovery Benzene in N2 Benzene in N2 7.93 132 -2.90 +1.39 ------- Audit No. TABLE 2. SUMWRY OF PERFORMANCE AUDIT RESULTS (Continued) RTI audit Client Industry Audit material conc. Client audit Status of (ppni) %bias(Avg.)* audit** 66 67 68 69 70 71 72 73 EPA, Region II EPA, QAD EPA, Region I E EPA, QAD EPA, QAD Ribber manufactir ing Coke even Byproduct Recovery Instrunent Check-out Degreasing vent Instrument check-out Canbustion efficiency test Benzene in N2 Benzene in 1^ Benzene in Benzene in Hexane in N2 Hexane in N2 Propane in N2 Propane in Benzene in N2 Benzene in N2 Vinyl chloride in N2 Vinyl chloride in Propylene in N2 Propylene in Vinyl chloride in N2 Trichloroethylene in N2 Trichloroethylene in N2 Hexane in N2 Hydrogen sulfide in N2 Methyl mercaptan in N2 12.0 10.2 100 335 79.8 3076 9.97 314 8.29 75.7 5.74 28.3 328 725 7.5 14.9 566 3076 16.2 8.22 +14.2 0 ->6.40 +6.00 +1.80 -7.50 -3.20 -10.8 -2.20 -2.50 NA m -7.00 -8.30 m -0.40 -8.70 NA -7.50 -8.90 F E ------- TABLE 2. SUWARY OF PERFORMANCE AUDIT RESULTS (Continued) Audit RTI audit Client audit Status of No. Client Industry Audit material conc. (ppm) % bias (Avg.)* audit*"* 74 E Vinyl chloride manufacturing 1,2-Dichloroethane in N2 1,2-Dichloroethane in N2 9.30 462 46.00 +3.70 E 75 N Propane in air Propane in air 10.0 309 m m F 76 F Propane in air Propane in air 10.0 309 m m F 77 D Maleic anhydride Benzene in Np Benzene in 9.46 66.9 -6.60 -11.7 E 78 EPA, Region VII Benzene in N? Hexane in N2 120 30.2 m NA F 79 D Maleic anhydride Benzene in N2 Benzene in 9.46 128 -4.60 +12.5 E 80 F Plywood/veneer drying Propylene in N2 Propylene in N Toluene in N2 14.8 328 430 -4.70 +4.40 -0.80 E 81 P Pl,ywood/veneer drying Propylene in N2 Propylene in N2 Toluene in 20.3 479 487 +18.2 -22.5 +32.5 E 82 J Polypropylene manufacturing Propylene in N? Propane in N2 Propane in N2 9.63 19.70 296 -0.35 ¦+0.84 +0.45 E 83 I Coke even Hydrogen sulfide in N2 Hydrogen sulfide in N~ Carbonyl sulfide 428 647 101 +4.90 -16.54 +1.98 E ------- TABLE 2. SUNMARY OF PERFORMANCE AUDIT RESULTS (Continued) Audit No. Client Industry Audit material RTI audit conc. (ppm) Client audit % bias (Avg.)* Status of audit** M J Compliance testing Benzene in N£ Hexane in N? Toluene in Ng Methyl mercaptain in N2 7.45 72.6 15.0 5.40 +3.70 A 85 I Steel manufactir ing Hydrogen sulfide in N2 Carbonyl sulfide in N2 647 9.08 +8.50 +1.00 E 86 I Oil shale Hydrogen sulfide in N2 Carbonyl sulfide in N2 Methyl mercaptan in N2 437 117 8.42 A NA = Not analyzed 1m Client-Measured Concentration - RTI-Measured Concentration *Client%Bias = RTI-Measured Concentration **Status Codes: A = Cylinder shipped; audit results not >et received; B = Audit results received; C = Audit report sitmitted to EPA; D = Audit results received, audit report sifcmitted to EPA, cylinder not jet retirned by client; E = Audit complete; F = Audit completed without analysis of audit materials by client. 1977 - Audits 1-8 1978 - Audits 9-28 1979 - Audits 29-49 1980 - Audits 50-75 1981 - Audits 76-82 1982 - Audits 83-86 ------- SECTION 4.0 STABILITY STUDIES An ideal calibration standard or audit material should be stable over its total time of usage. Any change or instability should be less than the measurement error tolerated during its use. The stabilities of the compounds in the repository are being estimated through periodic analysis of the cylinder contents. Improved estimates of stabilities could be determined by increasing considerably the frequency of the analyses, however, this increase is beyond the scope and financial limits of this project. The primary objective of this project is to conduct performance audits during source testing to assess the relative accuracy of the analytical results. In this project, the gas mixtures in the repository are initially analyzed upon receipt from the specialty gas vendor to corroborate the vendor's analysis. If the RTI analysis differs from the vendor's value by more than 10 percent, the cylinder is given to a third party for analysis. The gas mixtures are again analyzed at 1 month, at 2 months, and at one year following the initial analysis. These subsequent analyses are made to determine the stability of the gas mixtures. In some cases, analyses are not peformed on the dates specified above; however, every attempt is made to acquire the data on this schedule. Cylinder concentrations are also determined prior to each performance audit, providing additional data for use in stability studies. The estimates of stability have been calculated in terms of percent change per month. These rates of change are calculated from a linear regression analysis of the concentration/time data. Stability estimates have been calculated only for cylinder contents analyzed three or more times. Two standard deviations of percentage change per month have been calculated only for cylinder contents analyzed four or more times. Attachment 1 shows the concentrations and time periods on which the calculations are based and the percent change per month for each cylinder. Sample calculations for a 100 ppm benzene cylinder (See 1H, page 26) are given in Attachment 2. 19 ------- As the number of analyses per cylinder increases, additional statistical analyses will be performed. These additional statistical analyses will include testing models other than the linear regression model. Such models are needed especially where the change is rapid at first but later becomes slow or nonexistent. The results will be presented in a final report and journal publication. Absolute accuracy of the cylinder analyses have not been determined due to lack of NBS-SRM's as standards for most gas mixtures. Absolute accuracy could be estimated by performing the analyses using several different analytical methods or by having a relatively large number of laboratories perform analyses on the same cylinders; both of these approaches are beyond the financial limits of this project. An examination of the data in Attachment 1 shows values for individual cylinder analyses usually vary by less than 10 percent for 4-8 analyses over 2-3 years. This variation indicates either a real change in cylinder contents (i.e., instability) and/or the precision of the measurement process. The possible sources of experimental error that could result in apparent changes in concentrations include (1) the variability of the analytical technique used for analysis, (2) stability of and/or accuracy of calibration standards, and (3) the ability to reproduce standards for which NBS SRM's do not exist. Each of the above sources of variability impacts to some extent on the resulting data presented in Attachment 1. In some cases, the cylinder contents were lost after only a few analyses. The possible reasons include leakage during storage or shipping, misuse, multiple use in audits or initial low pressure in the cylinder. 20 ------- SECTION 5.0 SUMMARY AND CONCLUSIONS Cylinder gases of hydrocarbons, halocarbons, and sulfur containing organic species have been used successfully as audit materials to assess the relative accuracy of gas chromatographic systems used to measure source emissions. Absolute accuracy has not been determined because of lack of standard reference materials; instead interlaboratory bias has been reported for the performance audits conducted during source testing. The inter 1aboratory bias determined has been generally within 10 percent for both low and high concentration gases (Table 2). Of the 40 gaseous compounds studied or currently under study, 26 have demonstrated sufficient stability in cylinders to be used further as audit materials. Four compounds (ethylamine, paradichlorobenzene, cyclohexanone, and aniline) are not recommended as audit materials for various reasons as discussed in Attachment 1. One gaseous compound (formaldehyde) was ordered but the speciality gas manufacturer indicated that cylinder gases of this compound could not be prepared. Another ten compounds (carbon tetrachloride, methylene chloride, methyl chloroform, Freon 113, ethylene oxide, propylene oxide, allyl chloride, acrolein, chlorobenzene, and carbondisulfide) have recently been added to the repository; the stability of these compounds remains to be studied. As stated previously, the percent change per month values reported in Attachment 1 are only estimates of stability. Detailed statistical analyses which would separate statistical deviations from true concentration changes with time are in progress and will be presented in a final report and a journal publication. 21 ------- REFERENCES S. K. Gangwal and 0. E. Wagoner, "Response Correlation of Low Molecular Weight Sulfur Compounds Using a Novel Flame Photometr Detector", J. Chrom. Sci., 17, 196-201 (1979). 22 ------- ATTACHMENT 1 Stabillty Study Data as of May 1982 1.0 BENZENE STABILITY STUDY 2.0 ETHYLENE STABILITY STUDY 3.0 PROPYLENE STABILITY STUDY 4.0 METHANE/ETHANE STABILITY STUDY 5.0 PROPANE STABILITY STUDY 6.0 TOLUENE STABILITY STUDY 7.0 HYDROGEN SULFIDE STABILITY STUDY 8.0 META-XYLENE STABILITY STUDY 9.0 METHYL ACETATE STABILITY STUDY 10.0 CHLOROFORM STABILITY STUDY 11.0 CARBONYL SULFIDE STABILITY STUDY 12.0 METHYL MERCAPTAN STABILITY STUDY 13.0 HEXANE STABILITY STUDY 14.0 1,2-DICHLOROETHANE STABILITY STUDY 15.0 CYCLOHEXANE STABILITY STUDY 16.0 METHYL ETHYL KETONE STABILITY STUDY 17.0 METHANOL STABILITY STUDY 18.0 1,2-DICHLOROPROPANE STABILITY STUDY 19.0 TRICHLOROETHYLENE STABILITY STUDY 20.0 1,1-DICHLOROETHYLENE STABILITY STUDY 21.0 1,2-DIBROMOETHYLENE STABILITY STUDY 22.0 PERCHLOROETHYLENE STABILITY STUDY 23 ------- 23.0 VINYL CHLORIDE STABILITY STUDY 24.0 1,3-BUTADIENE STABILITY STUDY 25.0 ACRYLONITRILE STABILITY STUDY 26.0 ANILINE STABILITY STUDY 27.0 METHYL ISOBUTYL KETONE STABILITY STUDY 28.0 CYCLOHEXANONE STABILITY STUDY 29.0 PARADICHLOROBENZENE STABILITY STUDY 30.0 ETHYLAMINE STABILITY STUDY 31.0 FORMALDEHYDE STABILITY STUDY 32.0 METHYLENE CHLORIDE 33.0 CARBON TETRACHLORIDE 34.0 FREON 113 35.0 METHYL CHLOROFORM 36.0 ETHYLENE OXIDE 37.0 PROPYLENE OXIDE 38.0 ALLYL CHLORIDE 39.0 ACROLEIN 40.0 CHLOROBENZENE 41.0 CARBONDISULFIDE 24 ------- 1.0 BENZENE STABILITY STUDY Cylinder No. 1A IB 1C ID IE IF 1G Cylinder Construction* A1 A1 A1 A1 S S S Manufacturer ppn 65.4 324 200 117 61 71 80 Concentration Date 7/27/77 7/27/77 7/27/77 7/27/77 2/10/78 2/10/78 2/10/78 PPn (79.0) (374) (241) (138) (62.0) (71.0) (80.0) Day 136 136 247 29 78 232 78 ppm (74.0) (337) (216) (144) (62.0) (73.0) (81.0) RTT Day 156 156 252 157 216 385 216 Concentration ppm (78.0) (350) (215) (134) (61.0) (75.0) (81.0) Day 167 167 381 252 385 586 385 ppm (80.0) (355) (218) (129) (65.0) (74.5) (84.0) Day 630 402 290 722 882 504 ppm (77.9) (331) (127) (66.9) (75.7) (85.4) Day ** 433 414 1337 1292 L292 Ppm (343) (127) (55.7) (65.7) (74.0) Day 969 1247 ppm (358) (132) Day 1274 ppm (348) 1491 (324) % Change/month -0.01 -0.10 -0.87 -0.12 -0.16 -0.14 -0.17 Two Std. Dev. 0.42 0.16 0.66 0.27 0.32 0.29 0.25 of % Change/month ~ ~ A1 = Aluninun; S = Steel; LS = Low Pressure Steel. ~~ Cylinder enpty. ANALYTICAL CONDITIONS: Flame ionization detector, 10% 0V-101 on Qramosorb WhP colum at 60 degrees Celsius. CALIBRATION: Reagent grade "Benzene" liquid is used as a standard. "Glass bulb" dilution technique is uilized for making the series of standards for calibration. 25 ------- 1.0 BENZENE STABILITY STUDY (Continued) Cylinder No. 1H Cylinder Construction*** S 11 S 1J S IK S 1L S 1M S IN S Manufacturer Concentration PPT! 100 139 232 265 296 326 344 Date ppn 2/8/78 (101) 2/9/78 (139) 2/9/78 (229) 2/9/78 (264) 2/9/78 (295) 2/9/78 (319) 2/9/78 (332) Day ppn 65 (102) 49 (139) 233 (237) 49 (261) 49 (292) 49 (316) 49 (327) RTI Concentration Day ppn 206 50 (98.0) (142) 386 (243) 50 (268) 51 (294) 51 (318) 54 (342) Day ppn 237 (101) 96 (139) 557 (225) 69 (254) 93 (298) 96 (323) 69 (335) Day ppn 434 (105) 127 (140) W (269) 205 (294) 433 (345) 809 (342) Day ppn 773 (106) 206 (138) 237 (302) 830 (335) ** Day ppn 831 (100) 505 (147) 809 (295) 1294 (320) Day ppn 1294 (92.0) 1293 (128) 12W (290) 1338 (128) % Change/month -0.12 -0.16 -0.02 -0.W -0.03 0.06 0.10 Two Std. Dev. of % Change/month 0.21 0.12 0.62 2.51 0.06 0.17 0.16 AAA A1 = Aliminim; S = Steel; LS = Low Pressure Steel. ** Empty 26 ------- 1.0 BENZENE STABILITY STUDY (Continued) Cylinder No. 10 Cylinder Construction*** S IP S 1Q S 1R S IS s IT S 1U S Manufacturer Concentration ppm 389 8.04 9.85 9.89 9.93 10.0 10.9 Date PP 2/9/78 (387) 4/21/78 (8.37) 4/21/78 (9.99) 4/21/78 (10.0) 4/21/78 (10.0) 4/21/78 (10.7) 4/21/78 (11.5) Day ppm 64 (369) 4 (8.33) 5 (9.88) 4 (10.1) 4 (10.1) 25 (10.2) 4 (10.7) RTI Concentration Day ppm 205 (396) 25 (8.20) 25 (10.1) 13 (9.73) 26 (9.80) 146 (9.20) 25 (10.8) Day ppm 809 (396) 26 (8.34) 332 (9.71) 332 (9.77) 56 (9.50) 362 (9.90) 332 (10.7) Day m 1294 (389) 56 (8.19) 1018 (9.46) 146 (3.90) 1222 (9.56) 434 (10.9) Day ppm 134 (7.81) 1270 (9.64) 628 (9.57) ** 759 (10.2) Day ppm 434 (8.21) 738 (9.45) 1222 (9.69) Day ppm 766 (7.93) 1222 (7.68) irk % Change/month 0.06 -0.16 -0.25 -0.10 -0.13 -0.15 -0.29 Two Std. Dev. of % Change/month 0.16 0.10 0.26 0.08 0.33 0.35 0.15 ** Cylinder anpty. AAA A1 = A1 uninun; S = Steel; LS = Low Pressure Steel. 27 ------- 1.0 BENZENE STABILITY SRJDY (Continued) Cylinder No. Cylinder Construction**1* IV ' s 1W S IX S 1Y S 1Z S 1AA S 1AB A1 Manufacturer Concentration m 12.2 8.09 11.0 11.2 8.09 9.14 270 Date ppm 4/25/78 (12.7) 5/19/78 (8.10) 5/4/78 (U.2) 5/4/78 (10.9) 5/4/78 (3.20) 5/4/78 (9.10) 7/27/77 (300) Day ppm 1 (12.5) 106 (7.70) 132 (10.2) 132 (9.90) 132 (7.04) 132 (7.80) 29 (319) RTI Concentration Day ppm 21 (12.3) 287 (8.10) ** 302 (10.7) 302 (7.70) 302 (8.50) 157 (312) Day ppm 109 (12.0) 488 (8.20) 393 (10.8) 473 (7.54) 1005 (8.17) Day ppm 358 (12.1) 7W (8.30) 1209 (8.42) Day ppm 755 (12.0) 1218 (11.7) 1194 (7.45) % Change/month -0.14 -0.09 -2.W 0.11 -0.27 -0.08 0.36 Two Std. Dev. of % Change/month 0.09 0.25 — 1.04 1.21 0.35 — Cylinder anpty. AAA A1 = Aluninun; S = Steel; LS = Low Pressure Steel. 28 ------- 2.0 ETHYLENE STABILITY STUDY Cylinder No. 2A 2B 2C 2D 2E 2F 2G Cylinder Construction*** f A1 A1 A1 A1 A1 A1 A1 Manufacturer ppn 2920 3000 4960 4970 19900 19900 4.95 Concentration Date 2/23/78 2/23/78 2/23/78 2/23/78 2/24/78 2/24/78 4/27/78 ppm (3066) (3127) (5214) (5202) (20438) (20622) (4.70) RTI Day 49 49 48 48 48 48 29 Concentration ppm (3115) (3177) (5341) (5284) (20780) (20822) (4.70) Day 198 198 .201 201 200 200 106 ppm (2883) (2942) (4662) (4913) (20150) (20320) (4.85) Day 809 809 809 809 808 808 741 ppm (3203) (3272) (5383) (5338) (18906) (18960) (4.62) Day 1180 ppn (5.12) % Change/month 0.18 0.19 0.14 0.10 -0.31 -0.32 +0.14 Two Std. Dev. 0.44 0.44 0.72 0.39 0.10 0.07 0.22 of Change/month AAA A1 = AT uninun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flame ionization detector, Durapak n-octane on Porasil C colinn at 30 degrees Celsius. CALIBRATION: NBS-SRM Propane is used for standard calibration. 29 ------- 2.0 ETHYLENE STABILITY STUDY (Continued) Cylinder No. 2H 21 2J 2< 2L 2M 2N Cylinder Construction*** A1 A1 . ,A1 A1 A1 A1 A1 Manufacturer ppn 10.0 15.0 19.9 300 448 603 701 Concentration Date 4/27/78 4/28/78 4/28/78 4/28/78 4/28/78 4/28/78 4/28/78 ppm (9.70) (14.4) (19.2) (306) (468) (629) (740) RTI Day 29 28 28 33 33 34 34 Concentration ppm (9.60) (14.4) (19.3) (319) (493) (646) (749) Day 106 m 104 105 104 104 104 ppm (9.90) (14.9) (20.3) (312) (473) (636) (737) Day 740 739 739 728 740 740 740 ppm (8.40) (18.0) (27.5) (300) (457) (606) (703) Day 1180 1179 1179 ppn (10.0) (14.4) (18.9) % Change/month -0.07 0.18 +0.29 -0.17 -0.19 -0.20 -0.23 Two Std. Dev. 0.44 0.66 1.14 0.22 0.27 0.14 0.08 of % Change/month AAA A1 = Aluninun; S = Steel; LS = Low Pressire Steel. 30 ------- 3.0 PROPYLENE STABILITY STUDY Cylinder No. 3A 3B 3C 3D 3E 3F 3G 3H Cylinder Construction*** A1 A1 A1 A1 A1 A1 A1 A1 Manufacturer ppn 4.94 9.91 14.8 20.0 298 446 585 683 Concentration Date 4/27/78 4/27/78 4/27/78 4/27/78 4/27/78 4/27/78 4/27/78 4/27/78 ppm (4.86) (9.83) (14.6) (19.8) (296) (442) (577) (672) Day 26 26 26 27 27 27 27 27 ppm (4.94) (9.85) (14.5) (19.0) (286) (428) (560) ?655) RTI Day 27 104 104 104 104 105 104 105 Concentration ppm (4.78) (10.3) (14.8) (20.0) (317) (474) (629) (729) Day 104 749 749 749 750 750 750 750 ppm (4.98) (9.76) (14.8) (20.3) (324) (479) (620) (721) Day 749 1250 820 820 ppm (4.93) (9.63) (328) (725) % Change/month 0.05 -0.08 0.06 0.16 0.38 0.33 0.27 0.25 Two Std. Dev. 0.17 0.12 0.09 0.25 0.29 0.47 0.56 0.32 of % Change/month AAA A1 = A1 iminun; S = Steel; LS = Low Pressure Steel. ANALYTIC/L CONDITIONS: Flare ionization detector, Durapek n-octane on Porasil C col urn at 30 degrees Celsius CALIBRATION: NBS-SRM Propane is used for standard calibration. ------- 4.0 NETHANE/ETHANE STABILITY STUDY Cylinder No. 4A 48 4C 40 Cylinder Construction*** A1 A1 A1 A1 Audit Material**** M E M E M E M E Manufacturer Concentration ppm 6000 714 8130 597 1000 295 1670 202 Date ppm 7/21/78 (6207) 7/21/78 (773) 7/21/78 (8130) 7/21/78 (654) 7/21/77 (1021) 7/21/77 (315) 7/21/77 (1710) 7/21/77 (220) RTI Concentration Day ppm Day m 264 (5982) 662 (6584) 163 (715) 264 (684) 35 (7551) 264 (7824) 35 (663) 163 (606) 264 (983) 1027 (1289) 163 (292) 264 (283) 35 (1563) 264 (1640) 29 (218) 157 (202) Day ppn 662 (703) 662 (8592) 264 (577) 1027 (284) 1027 (1953) 258 (195) Day PP 662 (598) 1027 (206) % Change/month 0.32 -0.34 0.43 -0.43 0.90 -0.20 0.59 -0.13 Two Std. Dev. of % Change/month — 0.61 0.53 0.57 — 0.36 0.41 0.39 *** A1 = Aluninun; S = Steel; IS = Low Pressire Steel. M = Methane; E = Ethane. ANALYTICAL CONDITIONS: Flame ionization detector, Durapak n-octane on Porasil C col urn at 30 degrees Celsius. CALIBRATION: N3S-SRM methane is used for standard calibration. 32 ------- 5.0 PROPANE STABILITY STUDY Cylinder No. 5A Cylinder Construction*** A1 58 A1 5C A1 5D A1 5E A1 5F A1 5G A1 5H A1 Manufacture- Concentration PPn 5.01 10.0 14.6 20.0 303 439 6C4 708 Date ppn 4/25/78 (4.90) 4/25/78 (9.70) 4/25/78 (14.3) 4/25/78 (19.5) 4/26/78 (304) 4/26/78 (441) 4/26/78 (615) 4/27/78 (730) Day ppn 24 (4.90) 24 (9.80) 25 (14.5) 25 (19.8) 24 (301) 24 (436) 27 (615) 26 (723) RTI Concentration Day PPn 108 (5.10) 108 (10.1) 108 (14.9) 108 (20.3) 107 (305) 107 (440) 107 (607) 106 (710) Day ppn 606 (4.89) 513 (10.6) 582 (15.0) 582 (20.8) 530 (316) 530 (450) 604 (613) 603 (718) Day ppn 729 (5.20) 752 (10.0) 736 (14.7) 736 (20.1) 581 (316) 581 (453) 735 (628) 734 (734) Day ppn 914 (10.0) 1252 (19.7) 735 (313) 728 (472) Day ppn 752 (314) Day ppn 913 (309) 1251 (296) % Change/month 0.11 0.09 0.10 0.01 0.01 0.25 0.08 0.3 Two Std. Dev. of % Change/month 0.26 0.22 0.16 0.14 0.12 0.12 0.10 0.13 A1 = A1 uninun; S = Steel, LS = Low Pressure Steel. ANALYTICAL GONDITIONS: Flane ionization detector, Dirapak n-octane on Porasil C col urn at 30 degrees Celsius. CALIBRATION: N3S-SRM Propane is used for standard calibration. ------- 6.0 TOLUENE STABILITY STUDY Cylinder No. Cylinder Construct ion***' 6A LS 6B LS 6C S 6D S Manufacture- Concentration ppn 408 606 16.2 9.11 Date ppn 12/6/78 (405) 12/6/78 (585) 10/3/78 (17.3) 10/3/78 (9.62) RTI Concentration Day ppn 3 (405) 3 (579) 48 (14.9) 64 (8.50) Day ppn 86 (394) 86 (577) 365 (15.0) • 66 (8.60) Day ppm 100 (393) 358 (615) 1046 (8.91) 160 (8.20) Day ppm iek 982 (491) ** Day ppn 985 (487) % Change/month -0.93 -0.50 -0.72 -2.6 7 Two Std. Dev. of % Change/month 0.07 0.29 0.56 1.97 Cylinder enpty. AAA AT = Alinrinim, S = Steel, LS = Low Pressure Steel. ANALVTIC/l CONDITIONS: Flare ionization detector, 10% 0V-101 Chrarosorb W-IP colunn at 60 degrees Celsius. CALIBRATION: Reagent g-ade "Toluene" liquid is used as a standard. "Glass bulb" technique is utilized for generation of series of standards for calibration. ANALTYICAL PROBLEMS: All analyses of Cylinder No. 6B before Day 982 and of Cylinder No. 6C before Day 1046 used glass calibration bulbs at roan tenperature rather than bulbs vJiich were heated to cbove toluene's boiling pint. As a result, toluene may have condensed on the walls of the roam-tenperature bulbs. This may explain viiy the earlier concentrations are greater than those of the most recent analyses. However, actual degradation may ha/e occurred in these cylinders. 34 ------- 7.0 HYDROGEN SULFIDE STABILITY STUDY Cylinder No. 7A 7B 7C 7D 7E 7F Cylinder Construction*** A1 A1 A1 A1 A1 A1 Manufacturer ppn 399 9.15 16.7 649 6.95 6.45 Concentration Date 10/1/78 7/7/78 10/1/78 10A/78 10/1/78 10/1/78 ppm (371) (9.73) (16.1) (641) (7.05) (4.94) R7I Day 38 87 38 38 87 38 Concentration ppm (424) (6.72) (16.5) (655) (5.75) (5.14) Day 111 124 111 111 124 111 ppm (414) (7.11) (15.7) (690) (5.62) (4.81) Day 1030 197 580 1030 197 580 ppm (437) (6.36) (16.2) (647)) (5.23) (4.35) Day 696 1030 696 1030 ppm (6.23) (17.5) (5.14) (3.71) Day 1116 1116 ppm (8.32) (5.38) % Change/month 0.28 -0.05 +0.22 -0.05 -0.43 -0.75 Two Std. Dev. 0.47 1.24 0.20 0.28 0.67 0.16 of % Change/month ~ Cylinder enpty. AAA A1 = Aluninun, S = Steel, LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flans ionization detector, 10% 0V-101 Chromosorb MHP col urn at 60 degrees Celsius. CALIBRATION: Reagent g-ade pire "Hydrogen sulfide" gas is used as a standard. Dilutions are made in a Tedlcr bag for generation of series of standards for calibration. ANALTYICAL PROBLEMS: All analyses of Cylinder No. 6B before Day 982 and of Cylinder No. 6C before Day 1046 used glass calibration bulbs at roan tonperature rather than bulbs v^hich were heated to above toluene's boiling pint. As a result, toluene may have condensed on the walls of the rocm-tenperature bulbs. This may explain nhy the earlier concentrations are greater than those of the most recent analyses. However, actual degradation may have occurred in these cylinders. 35 ------- 8.0 M-XYLENE STABILITY STUDY Cylinder No. Cylinder Construction*** 84 LS SB LS 8C S 8D S Manufacturer Concentration ppn 405 613 17.3 7.33 Date ppn 10/5/78 (480) 10/5/78 (720) 10/5/78 (16.6) 10/5/78 (6.20) Day ppm 63 (445) 63 (676) 63 (17.2) 63 (6.81) RTI Concentration Day ppm IS (425) 158 (656) 166 (20.8) 166 (6.82) Day m 412 (487) 606 (760)* 302 (16.4) 948 (4.36)* Day ppm 606 (507)* 1040 (534) 918 (14.1)* 1036 (5.66) Day ppm 1040 (364) 1036 (19.0) % Change/month 0.35** -1.65** 0.24 -0.41 Two Std. Dev. of % Change/month 0.74 0.77 0.84 0.63 ~ Questionable value. Not included in the calculation of % change/month. irk Calculated only through Day 412 because of change in analytical procedures as described under analytical problems. kkk A1 = Aluninun; S = Steel; LS = Low Pressire Steel. CALIBRATION: Reagent grade "M-Xylene" liquid is used. "Glass bulb" technique is used for genertion of series of standards for calibration. ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chrcmosorb WP colunn at 60, 120 or 140 degrees Celsius. ANALYTICAL PROBLEMS: All analyses before Day 948 used glass calibration bulbs at room tenperature rather then bulbs which were heated to ebo/e meta-xylene's boiling point. As a result, meta-xylene may have condensed on the walls of the rocm-temperature bulbs. This may explain tfiy the RTI concentration for cylinder Nos. 8A and 8B before day 1040 are greater than those for the analysis on Day 1040. 36 ------- 9.0 METHYL ACETATE STABILITY STUDY Cylinder No. 9A 96 9C 9D Cylinder Construction*** S S S S Manufacturer m 326 455 6.m 17.2 Concentration Date 10/13/78 10/13/78 10/13/78 10/13/78 m (271) (428) (5.29) (12.9) RTI Day 230 230 230 230 Concentration ppm (340) (437) (4.86) (12.5) Day 285 286 286 286 ppm (324) (442) (5.02) (11.8) Day 629 629 630 630 ppm (348) (479) (5.88) (12.5) % Change/month 1.15 0.60 0.68 -0.13 Two Std. Dev. 1.08 0.21 1.13 0.56 of % Change/month AAA A1 = Alinrinun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 10% OV-275 on Chromosorb VHP colimn at 50 degrees Celsius. CALIBRATION: Reagent grade "Methyl acetate" liquid is used as a standard. "Glass bulb" technique is utilized for generation of series of standards for calibration. 37 ------- 10.0 CHLOROFORM STABILITY STUDY Cylinder No. Cylinder Construction*** 10-A S 10-8 S 10-C S 10-0 s Manufacturer Concentration ppn 520 348 8.70 16.9 Date ppm 10/17/78 (529) 10/17/78 (345) 10/17/78 (8.08) 10/17/78 (17.6) RTT Concentration Day ppm 161 (515) 161 (351) 161 (7.39) 161 (16.5) Day ppm 256 (514) 256 (340) 256 (7.50) 256 (16.2) Day ppm 553 (531) 975 (325) 553 (8.11) 553 (16.5) % Change/month 0.06 -0.20 0.14 -0.29 Two Std. Dev. of % Change/month 0.31 0.12 0.88 0.51 *** A1 = Aluninim; S = Steel; LS = Low Pressire. ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chromosorb WP colimn at 50 or 100 degrees Celsius. CALIBRATION: Reagent grade "Chloroform" liquid is used as a standard. "Glass bulb" technique is utilized for generation of series of standards for calibration. 38 ------- 11.0 CARBONYL SULFIDE STABILITY STUDY Cylinder No. Cylinder Construction*** 11-A S 11-B S 11-C s 11-D S 11-E AL 11-F AL Manufacturer Concentration ppn 251 100 9.96 7.03 9.54 101 Date ppm 11/3/78 (276) 11/3/78 (109) 11/3/78 (9.10) 11/3/78 (6.81) 9/18/81 (12.9) 9/18/81 (111) RTI Concentration Day ppm 78 (281) 78 (111) 78 (8.66) 78 (6.48) Day ppm 185 (275) ~ 185 (95) ~ 185 (8.23) ~ 185 (6.41) ~ % Change/month -0.09 -2.13 -1.55 -0.92 ** ** Two Std. Dev. of % Change/month — — — — — — ~ Cylinder onpty. Calculations fbr % change/month are only done vlien three or rare analyses are available. Akk A1 = Aluninun; S = Steel; US = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare photcmetric detector, Carbopak B col urn at 50 degrees Celsius or Chrcmosil 330 colim at 60 degrees Celsius. CALIBRATION: Reagent grade pure "Carbonyl Sulfide" gas is used as a standard. Dilutions are made in Teflon® bag fbr generation of series of standards for calibration. ANALYTICAL PR08LEMS: Chly a Teflon colum and Teflon lines should be used. The air-to-hydrogen ratio is critical to the sensitivity of the FPD. 39 ------- 12.0 MTTHYL NERCAFTAN STABILITY STUDY Cylinder No. 12-A 12-B 12-C 12-D Cylinder Construction*** A1 A1 A1 A1 Manufacturer ppn 8.03 10.0 3.55 4.22 Concentration Date 1/24/79 1/24/79 1/24/79 1/24/79 ppn (5.66) (7.94) (3.65) (4.23) RTI Day 1M 104 104 104 Concentration ppn (5.60) (8.10) (3.50) (4.76) Day 139 139 139 139 ppm (5.65) (7.90) (3.56) (4.54) 985 985 985 ~ (5.40) (8.42) (3.64) % Change/month -0.14 0.18 0.05 2.04 Two Std. Dev. 0.04 0.10 0.17 of % Change/month A1 = A1 Lminim; S = Steel; LS = Low Pressure Steel. ~ Empty ANALYTICAL CONDITIONS: Flame photcmetric detector, Carbopak B col am at 50 degrees Celsius or Chrcmosil 330 col urn at 60 degrees Celsius. CALIBRATION: Reagent grade pure "Methyl mercaptan" gas is used a standard. Dilutions are made in a Teflorf® bag for generation of series of standards for calibration. ANALYTICAL PROBLEMS: Only a Teflon colimn and Teflon lines should be used. The air-to-hydrogen ratio is a critical variable. 40 ------- 13.0 HEXANE STABILITY STUDY Cylinder No. 13-A 13-3 13-C 13-D Cylinder Construction*** LS LS A1 A1 Manufacturer ppn 1975 2973 30.6 79.2 Concentration Date 2/6/79 2/6/79 2/6/79 2/6/79 ppn (2170) (3070) (30.8) (82.2) RTI Concentration Day 6 6 296 296 ppn (1982) (2855) (30.1) (81.0) Day 337 338 337 337 ppn (2069) (2946) (30.6) (81.3) Day 469 469 469 469 ppn (1986) (3076) (32.0) (79.8) Day 523 835 ppn (30.0) (80.2) Day 835 ppn (30.2) % Change/month -0.22 0.16 -0.05 -0.09 Two Std. Dev. 0.® 0.60 0.26 0.C8 of % Change/month AAA* A1 = Aluninun; S = Steel; l_S = Low Pressire Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chrorosorb kHP colunn at 60 or 100 degrees Celsius. CALIBRATION: Reagent grade "Hexane" liquid is used as a standard. "Glass bulb" technique is utilized for making series of standards for calibration. 41 ------- 14.0 1,2 DICHLCROETHANE (ETHYLENE DICHLCRIDE) STABILITY STUDY Cylinder No. 14A 14B 14C 14D 14E 14F 14G 14H Cylinder Construction*** A1 A1 A1 A1 A1 A1 A1 A1 Manufacture' ppn 14.4 9.64 100 526 6.92 12.5 97.9 439 Concentration Date 1/19/79 1/19/79 1/19/79 1/19/79 4/5/79 4/5/79 4/5/79 4/5/79 ppm (14.1) (9.20) (96.2) (498) (10.0) (15.2) (102) (463) RTI Day 58 58 58 58 30 30 30 30 Concentration ppm (15.2) (10.8) (103) (534) (9.42) (14.7) (105) (451) Day 155 155 155 155 69 69 69 69 ppm (14.9) (10.0) (98.2) (524) (9.30) (14.3) (99.0) (462) Day 811 811 501 501 586 811 425 589 ppm (14.2) (9.56) (87.3) (592) (9.14) (14.5) (87.3) (432) Day 835 835 920 920 811 835 m 697 ppm (13.5) (9.19) (102) (502) (9.70) (13.8) (101) (451) Day 835 m ppm (9.16) (453) % Change/month -0.22 -0.23 -0.01 +0.04 -0.09 -0.16 -0.18 -0.10 Two Std. Dev. 0.28 0.50 oo LO o 0.64 0.25 0.21 0.62 0.16 of % Change/month kk'k A1 = A1 uninun; S = Steel; IS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 10% OV-101 on Chrorosorb WP colum at Id) degrees Celsius. CALIBRATION: Reagent grade "1,2 Dichloroethane" liquid is used as a standard. "Glass bulb" technique is utilized for making series of standards for calibration. ------- 15.0 CYCLOHEXANE STABILITY STUDY Cylinder No. 15-A Cylinder Construction*** A1 Manufacturer ppn 99.1 Concentration Date 3/19/79 RTI PPn (106) Concentration Day 147 ppn (93.4) Day 394 ppn (99.0) 926 (102) % Change/month 0.02 Two Std. Dev. 0.55 of % Change/ronth *** A1 = Aliminun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flame ionization detector, 10% 0V-101 on Chromosorb WHP colimn at 100 degrees Celsius. CALIBRATION: Reagent grade "Cy:lohexane" liquid is used as a standard. "Glass bulb" technique is used for making series of standards for calibration. 43 ------- 16.0 METHYL ETHYL KETONE STABILITY STUDY Cylinder No. 16-A Cylinder Construction*** S Manufacturer ppn 43.7 Concentration Date 5/23/79 ppn (42.3) -RTI Day 28 Concentration ppm (40.0) Day 58 ppm (39.9) Day 380 Ppn (44.5) Day 653 ppn (38.7) % Change/month -0.08 Two Std. Dev. 0.67 of % Change/month AAA A1 = Aluninun; S = Steel; LS = Low Pressire Steel. ANALYTICAL CONDITIONS: Flare ionization detector, Chrarosorb 101 colimn at 180 degrees Celsius. CALIBRATION: Reagent grade "Methyl ethyl ketone" liquid is used as a standard. "Glass bulb" technique is utilized for making series of standards for calibration. 44 ------- 17.0 METHANOL STABILITY STUDY Cylinder No. 17-A Cylinder Construction*** A1 Manufacture- ppn 50.0 Concentration Date 5/17/79 ppn (58.8) RTI Day 21 Concentration ppm (52.3) Day 51 ppn (51.1) Day 196 PP (55.2) % Change/month -1.03 Two Std. Dev. 3.01 of % Change/month AAA A1 = Alimirum; S = Steel; US = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, Chrorosorb 101 colurm at 50 degrees Celsius or 0.2% Carbowax 1500 plus 0.1% SP-2100 on Carbowax C at 60 degrees Celsius. CALIBRATION: Reagent grade "Methanol" is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for calibration. 45 ------- 18.0 1,2-DICHLCROPROPANE (PROPYLENE DICHLCRIDE) STABILITY STUDY Cylinder No. Cylinder Construction*** ISA A1 188 A1 18C A1 18D A1 Manufacturer Concentration ppn 7.07 14.6 476 664 Date ppn 7/10/79 (6.06) 7/10/79 (15.6) 7 AO/79 (496) 7/10/79 (685) RTT Concentration Day Ppn 28 (5.52) 28 (16.4) 28 (455) 28 (621) Day ppn 48 (5.94) 48 (15.0) 48 (480) 48 (675) Day ppn 497 (6.03) 749 (16.3) 372 (497) 372 (685) Day ppn 749 (5.59) % Change/month -0.09 +0.16 +0.33 +0.30 Two Std. Dev. of % Change/month 0.43 0.43 0.89 1.05 kkk AT = A1 Lminun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flame ionization detector, 10% 0V-101 on Chrcmosorb WP colunn at 100 degrees Celsius. CALIBRATION: Reagent grade "1,2-Dichloropropane" liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for calibration. 46 ------- 19.0 TRICHLCROETHYLENE STABILITY STUDY Cylinder No. Cylinder Construction*** 19A A1 1SB A1 19C A1 19D A1 Manufacturer Concentration ppn 9.23 14.7 100 505 Date ppm 5/24/79 (9.58) 5/24/79 (14.3) 5/24/79 (102) 5/24/79 (506) RTI Concentration Day ppm 77 (10.2) 77 (15.1) 77 (103) 77 (503) Day PPn 92 (9.78) 92 (14.9) 92 (100) 92 (499) Day ppm 683 (9.03) 683 (13.6) 810 (105) 810 (522) 820 (8.91) 820 (13.5) 820 (94.6) 820 (490) % Change/month -0.37 -0.33 -0.08 0.02 Two Std. Dev. of % Change/month 0.23 0.21 0.31 0.19 ¦Hck A1 = Aliminun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chrcmosorb WP colimn at 100 degrees Celsius. CALIBRATION: Reagent grade "Trichloroethylene" liquid is used as a standard. "Glass bulb" technique is used for making series of standards for calibration. 47 ------- 20.0 1,1-DICHLCR0ETHYLENE (VINYLIDENE CHLORIDE) STABILITY STUDY Cylinder No. 2QA 2C8 20C 2CD Cylinder Construction*"** A1 A1 A1 A1 Manufacturer ppn 9.58 14.8 96.8 490 Concentration Date 6/1/79 6/1/79 6/1/79 6/1/79 ppn (10.3) (15.6) (101) (524) RTI Day 35 35 35 35 Concentration ppn (9.90) (15.1) (99.0) (510) Day 62 62 62 62 ppn (10.1) (15.5) (102) (505) Day 4M 404 817 404 ppn (11.5) (17.1) (94.0) (498) 818 818 (9.00) (14.2) % Change/month -0.25 -0.17 +0.40 -0.26 Two Std. Dev. 0.81 0.63 0.14 0.31 of % Change/month AAA A1 = Aluninun; S = Steel; LS = Low Pressire Steel. ANALYTICAL CONDITIONS: Flame ionization detector, 10% 0V-101 on Chrcmosorb WP colum at 100 degrees Celsius or 10% SP-2100 on Supelcoport colum at 100 degrees Celsius. CALIBRATION: Reagent grade "1,1-Dichloroethylene" pure liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for calibration. 48 ------- 21.0 1,2-DIBR0M0ETHYLENE (ETHYLENE DIBRCMIDE) STABILITY STUDY Cylinder No. 21A 21B 21C 21D Cylinder Construction*** LS LS LS LS Manufacturer ppn 10.0 14.9 99.9 301 Concentration Date 6/18/79 6/18/79 6/1/79 6/18/79 ppm (7.90) (12.2) (110) (265) RTI Day 61 61 61 61 Concentration ppn (7.80) (12.0) (107) (266) Day 89 89 89 89 ppm (7.40) (11-6) (105) (257) Day 722 772 787 643 ppm (7.72) (8.02) (99.2) (309) % Change/month -1.90* -1.53* -1.52* -0.83* Two Std. Dev. 0.35 0.10 0.17 0.29 of % Change/month * % change/month is calculated for only the first three analyses and are approximate because of change in calculation procedure as-described below in the analytical problens. A AA A1 = Aluninim; S = Steel; LS = Low Pressire Steel. ANALYTIC/^ CONDITIONS: Flcme ionization detector, 10% 0V-101 on Chrcmosorb W4P colunn at 100 degrees Celsius. CALIBRATION: Reagent grade "1,2-Dibrcmoethylene" pure liquid is used as a standard. "Glass bulb" dilution techniques is utilized for making series of standards for calibration. ANALYTICAL PROBLEMS: The gas mixtures and the cal ibration standards contain substantial anoints of both the cis and the trans iscmers of 1,2-Dibrcmoethylene. The first three sets of analyses are questionable because only one isaner was measured during the calibrations and cylinder analyses. 49 ------- 22.0 PERCHLOROETHYLENE STABILITY STUDY Cyl inder No. 22A 22B 22C 22D Cylinder Construction*** S S LS LS Manufacturer ppm 7.98 13.0 487 629 Concentration Date 7/6/79 7/6/79 7/6/79 7/6/79 ppn (8.40) (15.0) (419) (624) RTI Day 35 35 35 35 Concentration ppn (7.97) (14-9) (453) (642) Day 52 52 52 52 ppm (7.92) (14.7) (440) (619) Day 376 376 677 677 ppm (7.94) (14.5) (361) (542) 713 713 (387) (571) % Change/month -0.22 -0.23 -0.63 -0.51 Two Std. Dev. 0.61 0.17 0.37 0.22 of % Change/month AAA A1 = A1 rniiniri; S = Steel; LS = Low Pressire Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chrcmosorb WHP colimn at 50 or 100 degrees Celsius. CALIBRATION: Reagent grade "Perchloroethylene" liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for cal ibration. 50 ------- 23.0 VINYL CHLCRICE STABILITY STUDY Cylinder No. 23A 23B 23C 230 23E 23F 23G 23H 231 Cylinder Construction*** S S S S S S S S S Manufacture- ppn 5.S4 8.00 8.03 8.52 20.0 20.1 30.0 30.3 7.98 Concentration RTI Date 10/1/79 10/1/79 10A/79 10/1/79 10/1/79 10/1/79 10 A/79 10/1/79 10A/79 Concentration ppn (5.87) (7.71) (7.82) (7.85) (19.7) (20.1) (29.6) (29.8) (7.31) 18 18 18 18 18 18 18 18 18 (5.74) (7.50) (7.45) (7.61) (19.1) (19.3) (28.3) (28.7) (7.12) 700 700 700 700 700 700 700 700 (6.60) (8.44) (8.41) (20.7) (20.9) (30.4) (29.4) (8.39) % Change/month 0.59 ¦** 0.45 0.38 0.29 0.25 0.01 0.01 0.71 • Calculations for % change/month are only done nhen three or more analyses are avail able. ¦kirk A1 = Aluninun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flane ionization detector, 0.4% Carbowax 1500 on Carbopak C at 50 degrees Celsius. CALIBRATION: Vinyl chloride permeation tibe purchased from Metronics is used for calibration. Permeation tube is maintained at 30°C. 51 ------- 24.0 1,3 BUTADIENE STABILITY STUDY Cylinder No. 24A Cylinder Construction*** S Manufacturer ppm 22.6 Concentration RTI Date 3/21/80 Concentration ppm (20.9) Date 95 m (23.1) Day 480 ppn (24.0) % Change/month +0.73 kkk A1 = Aliminun; S = Steel; LS = Low Pressire Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 0.1% SP-1000 on Carbopak C colum at 90 degrees Celsius or 10% 0V-101 on Chrorrosorb WHP colum at 60 degrees Celsius. CALIBRATION: Reagent grade "1,3 Butadiene" liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for calibration. 52 ------- 25.0 ACRYLONITRILE STABILITY STUDY Cylinder No. 25A 25B 25C 25D Cylinder Construction*** LS LS LS LS Manufacturer PPn 20.1 348 11.7 638 Concentration Date 7/24/79 7/24/79 7/24/79 7/24/79 m (14.6) (411) (6.38) (678) RTI Day 185 185 185 185 Concentration ppm (12.7) (416) (3.35) (699) Day 349 349 349 349 ppm (13.2) (441) (2.87) (703) Day 841 841 841 841 ppm (9.96) (397) (4.05) (667) % Change/month -1.07 -0.14 -1.12 -0.10 Two Std. Dev. 0.46 0.47 3.49 0.26 of % Change/month *** A1 = Aluninim; S = Steel; LS= Low Pressure Steel ANALYTICAL CONDITIONS: Plane ionization detector, 4% Carbowax 2CM on Carbopak B at 50 or 150 degrees Celsius CALIBRATION: Acrylonitrile permeation tibe is used for GC-FID cali- bration. Permeation tube is maintained at 30°+_0.1°C. ANALYTICAL PROBLEMS: The large changes noted at the lew concentration levels are, at least in part, a result of difficulty in making precise measuranents at these levels. 53 ------- 26.0 ANILINE STABILITY STUDY Cylinder No. 26A 26B Cylinder Construction*** A1 A1 Manufacturer ppn 11.3 18.4 RTI See Analytical Problems Analysis •kick A1 = Aluninun; S = Steel; LS = Low Pressure Steel ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 on Chrcmosorb WHP colimn at 250 degrees Celsius. CALIBRATION: Reagent grade "Aniline" pure liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for calibration. ANALYTICAL PROBLEMS: Because aniline has an extronely high boiling point (186°C), special handling would be required to measure this canpoind. A completely heated systan for sampling in the vapor phase and for preparing standards would be required. Temperature-dependent conden- sation in the cylinder and the regulator causes the anouit of aniline which is delivered by the cylinder to vary. As a result, aniline is not considered to be practical as an audit material. 54 ------- 27.0 .METHYL ISCBUTYL KETCME STABILIPC STUDY Cylinder No. 27A 27C Cylinder Construction*** A1 A1 Manufacturer ppn 9.51 72.9 Concentration Date 12/18/80 7/8/81 ppm (10.2) (75.4) RTI Day 27 See Analytical Concentration ppm (10.6) Problems Day 83 ppm (9.53) Day 202 ppm (9.49) % Change/month -1.40 Two Std. Dev. 1.58 of % Change/month Calculations for % change/month are only done viien tlree or rrore analyses are available. AAA A1 = Aluninim; S = Steel; LS = Low Presstre Steel ANALYTIC/1 CONDITIONS: Flare ionization detector, 0.1% SP-1000 on Carbopak C col iron at 180° degrees Celsius. CALIBRATION: Reagent grade "Methyl isobutyl ketone" liquid is used as a standard. "Glass bulb" technique is utilized for making series of standards for calibration. ANALYTICAL PROBLEMS: Methyl isobutyl ketone at high concentra- trations is not practical as an audit material because pressirization of the cylinder above approximately 200 psi results in condensation of the analyte. 55 ------- 28.0 CYCLOHEXANONE STABILITY STUDY Cylinder No. 28A 2S8 Cylinder Construction*** A1 A1 Manufacturer ppn 10.1 19.0 Concentration RTI Date 12/11/8) 12/11/80 Analysis ppm (8.19) (25.5) Day 85 85 ppm (3.26) (17.1) % Change/month ** ** Calculations for % change/month are only done i«hen three or more analyses are available. AAA A1 = Aluninun; S = Steel; LS = Low Pressure Steel ANALYTICAL CONDITIONS: Flare ionization detector, 10% SP-1000 on Supelco- port colimn at 200 degrees Celsius. CALIBRATION: Reagent grade "Cyzlohexanone" liquid is used as a standard. "Glass bulb" technique is used for making series of standards for calibration. ANALYTICAL PROBLEMS: The analysis of cyrlohexanone gas is dependent on the temperatures of the cylinde- and the regulator and on the length of the sampling line between the regulator and the gas chronatograph. The concen- tration in the cylinder decreases with time. Therefore, cyzlchexanone is not practical as an audit material. 56 ------- 29.0 PARADICHLCRCBENZENE STABILITY STUDY Cylinder Mo. 2SA 29B Cylinder Construction*** S S Manufacturer ppm 15.6 38.1 Concentration RTI See Analytical Problems Analysis AAA AT = Aliminun; S = Steel; LS = Low Pressure Steel ANALYTICAL CONDITIONS: Flare ionization detector, 10% SP-1000 on Supelcoport col urn at 200 degrees Celsius. CALIBRATION: Reagent grade "Paradichlorobenzene" is used as a standard. "Glass bulb" technique is used for making series of standards for calibration. ANALYTICAL PROBLEMS: The stability study for this compound was terminated bacause of analytical difficulties and because the cylinder pressure was less than 200 psig. Paradichlorobenzene is a solid at room temperature with a melting pint of 54°C. Condensation in the cylinder, regulator and sampling lines was extreme. Paradichloro- benzene is not practical as an audit material. 57 ------- 30.0 ETHYLAMINE STABILITY STUDY Cylinder No. 30A 3G6 Cylinder Construction**11' S S Manufacturer ppn 10 20 Concentration RTI Analysis See Analytical Problems AAA A1 = aluninun; S = Steel; LS = Low Pressure Steel ANALYTICAL CONDITIONS: Flare ionization detector, 10% 0V-101 Chrcmosorb WHP colum at 250 degrees Celsius. CALIBRATION: Reagent grade "Ethylamine" liquid is used as a standard. "Glass bulb" technique is utilized for making series of standards for calibration. ANALYTICAL PROBLEMS: Because of vapor pressure considerations, the cylinders could not be fully pressurized. The pressure in the cylin- der is less than 200 psi. A canpletely heated systen for sampling in the vapor phase and for preparing standards would be required. Tanperature-dependent condensation in the cylinder and the regulator causes the aroint of aniline which is delivered by the cylinder to vary. As a result of these problems, ethylamine is not considered to be practical as an audit material. 58 ------- 31.0 FORMALDEHYDE STABILITY STUDY RTI Requested ppm 10 20 Concentration The speciality gas supplier indicated that they could not make gas mixtures containing formaldehyde. 59 ------- 32.0 METHYLENE CHLORIDE Cylinder No. 32A Cylinder Construction* A1 Manufacture' ppn 10.2 Concentration RTT Date 3/5/82 Concentration ppm (10.8) Day 31 ppm (10.8) Day 70 ppm (10.6) % Change/month 0.68 ~ A1 = Aliminun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 20 ft. x 1/8" SS colimn packed with 10% SP-1000 on 80/100 Supelcoport. 30 cm^/minute hfe carrie- gas. Colum tarp. = 100°C. Detector temp. = 175°C. CALIBRATION: Reagent grade "Methylene chloride" liquid is used as a standard. "Glass bulb" technique is utilize for making series of standards for calibration. 60 ------- 33.0 CARBON TETRACHLORIDE Cylinder Mo. 33A Cylinder Construction* AL Manufacturer ppn 11.3 Concentration RTI Date 3/4/82 Concentration Ppn (12.7) Day 74 ppm (11.7) ~ A1 = Aluninun; S = Steel; LS = Low Pressure Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 20 ft. x 1/8" SS col urn packed with 10% SP-1000 on 80/1000 Supelcoport. 30 cnr/minute He carrier gas. Colum tarp. = 1008C. Detector tenp. = 175°C. CALIBRATION: Reagent grade "Carbon tetrachloride" liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for cali- bration. 61 ------- 34.0 FREOM 113 Cylinder No. 34A Cylinde" Construction* A1 Manufacture- ppn 10.4 Concentration RTI Date 3/3/82 Concentration ppm (10.8) Day 34 ppm (10.1) Day 70 ppm (10.0) % Change/month -3.08 ~ A1 = Aluninun; S = Steel; LS = Low Pressire Steel. ANALYTICAL CONDITIONS: Flare ionization detector, 20 ft. x 1/8" SS col urn packed with 10% SP-1000 on 80/100 Supelcoport. 30 on^/minute Hs carrier gas. Colum taip. = 100°C. Detector tanp. = 175"C. CALIBRATION: Reagent grade "Freon 113" liquid is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for cali- bration. G2 ------- 35.0 METHYL CHLCRCFORM Cylinder No. 35A Cylinder Construction* A1 Manufacturer ppn 10.2 Concentration RTI Date 3/2/82 Concentration ppm (10.3) Day 70 ppm (11.8) ~ A1 = Aluninun; S = Steel; LS = Low Pressire Steel. ANALYTIC/L CONDITIONS: Flare ionization detector, 20 ft. x 1/8" SS colunn packed with 10% SP-1000 on 80/100 Supelcoport. 30 air/minute He carrier gas. Col urn terp. = 100°C. Detector teip. = 175°C. CALIBRATION: Reagent grade "Methyl chlorofonn" is used as a standard. "Glass bulb" dilution technique is utilized for making series of standards for cali- bration. 63 ------- 36.0 ETHYLENE OXIDE Cylinder No. 36A Cylinder Construction* A1 Manufacturer ppti 10.0 Concentration RTI Date 3/12/82 Concentration ppm (11-2) Day 73 ppn (9.60) Day 88 ppm (9.80) % Change/month -4.8 ~ A1 = Aluninun; S = Steel; US = Low Pressire Steel. ANALYTIC/L CONDITIONS: Flare ionization detector, 6 ft. x 1/8" SS colimn packed with 80/100 mesh Porapak Q5. 30 cnr/ minute He carrier gas. Col urn tamp. = 150°C. Detector = 175°C. CALIBRATION: Ethylene oxide permeation tibe purchased from Metronics is used for GC-FID calibration. Permeation tibe is maintained at 30°C. ANALYTICAL PROBLEMS: There appeared to be seme loss of ethylene oxide when a brass regulator was used on the cylinder. A stainless steel regulator was used for the analysis. 64 ------- 37.0 PROPYLENE 0XIC6 Cylinder No. 37A 378 Cylinder Construction* A1 A1 Requested ppn 10.0 100 Concentration RTI Date XX XX Concentration ppn ~ A1 = A1 iminun; S = Steel; LS = Low Pressure Steel. XX Cylinders recently ordered but not yet received fron the manufacturer. 65 ------- 38.0 ALLYL CHLCRIDE Cylinder No. 38A 388 Cylinder Construction* S S Requested ppn 10.0 100 Concentration RTT Date XX XX Concentration ppm ~ A1 = Aliminun; S = Steel; LS = Low Pressire Steel. XX Cylinders recently ordered but not^et received fran the manufacturer. 66 ------- 39.0 ACROLEIN Cylinder No. 39A 39B Cylinder Construction* A1 A1 Requested ppn 10.0 100 Concentration R7I Date XX XX Concentration ppm ~ AT = Aluninun; S = Steel; LS = Low Pressure Steel. XX Cylinders recently ordered but not .jet received frcm the manufacturer. 67 ------- 40.0 CHL0R0BENZENE Cylinder Mo. 4QA Cylinder Construction* A1 Requested ppn 10.0 Concentration R71 Date XX Concentration ppn ~ A1 = Aliminun; S = Steel; LS = Low Pressure Steel. XX Cylinders recently ordered but not^et received frcm the manufacturer. 63 ------- 41.0 CAR80NDISULFIDE Cylinder No. 41A Cylinder Construction* A1 Manufacturer ppn 100 Concentration RTI Date XX Concentration PPm ~ A1 = A1 uninun; S = Steel; LS = Low Pressire Steel. XX Cylinders recently ordered but not >et received fran the manufacturer. 69 ------- ATTACHMENT 2 SAMPLE CALCULATIONS OF % CHANGE/MONTH 70 ------- ATTACHMENT 2 Sample Calculations of % Change/Month Example: (Data shown for benzene cylinder.) Data: Date of Analysis Concentration, ppm 2/8/78 4/13/78 9/1/78 10/2/78 4/17/79 101 102 98 101 105 1) Linear Regression, X and Y Data Points, Slope X values correspond to the dates of analyses, with the first date being day 1 (2/8/78). The second X value is equal to the number of calendar days (i.e., 65) between the first analysis date and the second analysis date (4/13/78) and so on. The Y values are the concentrations (ppm) that were determined on the respective analysis dates. Perform the regression analysis and determine the slope of the line utilizing commercially-avai1 able hand calculators or by the following formula: X Y 1 101 102 98 101 105 65 206 237 434 m z(Xi-X)(Yi-Y) 0.0071 Standard deviation of slope = -I 1/2 £(xrx)2 0.0076 (N-2) z (X.-X)c 71 ------- m = slope of regression line X. = x-coordinate of an individual value l Y. = /-coordinate of an individual value l X" = average of all Y - average of all Yi N = number of analyses 2) Percent change/month % change/month = m (slope) x ^gg x days where b = y intercept = -^ (Y - bX) = 100.06 Therefore, % change/month = °-0071 x 100 x 30 = 0.21 100.06 3) Standard deviation of percent channe/month , , . Standard deviation of slope Standard deviation in units = X 100 X 30 = 0 2270 of % change/month Intercept 95% Confidence interval = 0.21 + 2 X 0.2270 = (-0.255, 0.664) for % change/month 72 ------- |