Document wqmMnNgejQRVDKOZ8vXO1OGv3

(conoco) Interoffice Communication To From Date Subject R. B. Martin D. S. Cox August 27, 1979 CHECKING THE CALIBRATION OF THE "REAL" MINIMONITOR AND REPRODUCIBILITY MEASUREMENT OF THE CARBON ANALYSIS A program was developed to determine if the calibration of the "REAL" badge, membrane would change; however, the work changed to more of a precision study for the carbon adsorption test. SUMMARY 1. A series of badges were exposed to known concentrations of VCM in nitrogen for known times and the calibration values measured. 2. Carbon was exposed to VCM (without using the badge) and the precision measured. Ultrasonic treatment improved the precision by almost a factor of 10. 3. Carbon was exposed to VCM and analysis done by the OSHA approved carbon disulfide method. The precision was only marginally better. CONCLUSIONS 1. The badge calibrations had changed indicating the membrane was partially blocked. Replicate measurements are needed to overcome the inherent var iability of the carbon analysis. 2. The head space method gives equivalent results to the carbon disulfide method. 3. Under normal analysis methods, the precision (1 standard deviation) of the carbon analysis is 19% relative (1 ppm TWA-8 hr.). There is some indication that improvement is possible using an ultra sonic treatment. EXPERIMENT I The badges were exposed to known concentrations of flowing vinyl chloride gas. The exposures were made in a special chamber so each badge had its own atmosphere unaffected by any of the other units. The analysis was done by F-42 head space analysis using a mixture of diniethlacetimide (DMA) and water (see Appendix for analysis details). The "Standard" used was the average value of 10 knowns. Table I shows the results of the calibration checks. Due to the poor relative standard deviation, a search was made for the source of the variability. ftHS ftp F 3 VU_ August 27, 1979 Checking the Calibration of the "REAL" Badge Page 2 EXPERIMENT II A systematic search for variability was made in the test. Six vials were filled with a 12 ppm VCM standard gas and run on the F-42 (Table 2). The errors due to the injection step were 2.7% relative. A solution of VCM in DMA was prepared and an equal amount of this solution put into five F-42 vials. The extraction solvent was added and the vials tested on the F-42 (Table 3). The precision was still +_2.7% relative and showed that the equilibrium of VCM in the level-space was constant. EXPERIMENT III The regular 90C method of analysis of the carbon standards (Table 4) showed a relative error of 19.3%. A test was made at a lower bath temperature (70C) to check for VCM degradation. None was evident. The relative error for standards in the 70C bath was not significantly better. EXPERIMENT IV Carbon was put into a flask and exposed to VCM vapors then the carbon was put into vials (equal weights) and run. The precision was much better (Table 5) indicating standard preparation contributes to the errors. EXPERIMENT V A test was carried out with propane to confirm that the errros were not due to the reactivity of VCM. Table 6 shows the results of this test. Propane behaved the same as VCM on carbon. EXPERIMENT VI . A set of standards were analyzed by the carbon disulfide method to confirm that the head space method was not the source of errors. Table 7 shows the results of that test. There was some improvement over the regular carbon analysis. EXPERIMENT VII A test was made using the ultrasonic bath to dislodge any gas bubbles from the carbon after extraction was complete. The results (Table 8) were favorable with a relative error of only 3.5%. More tests will be needed to confirm that the ultrasonic treatment results were typical and not a fluke. David S. Cox Mfg. Chemist 7 TABLE I CALIBRATION VALUES c "1 V 'S \ BADGE # Run #1 2 3 209 1.01 1.00 1.03 210 1.05 1.01 1.04 211 1.17 1.15 0.90 213 1.29 0.93 1.00 214 1.09 1.03 1.02 4 0.89 0.90 0.94 0.90 0.98 5 0.93 0.92 0.88 0.80 0.89 6 1.47 1.4S 0.87 1.47 1.66 7 1.25 1.30 1.23 1.58 1.60 8 1.41 1.28 1.09 1.18 9 1.96 1.79 1.86 - 2.44 10 1.43 1.10 - - 1.36 11 1.17 0.81 1.34 - 1.21 12 Average 1.23 1.23 1.42 1.18 1.78 1.20 _ 1.14 1.30 1.31 Rel. Std. Deviation + 10% 24% 29% 27% 32% K = CT m K = Calibration # T - Hours M = Microliters of VC on Carbon C = Concentration of the VC in roicroliters per liter (ppm) . 116 1.71 1.00 1.02 1.04 0.75 1.48 2.52 1.38 1.42 0.91 1.32 39% rk Vi4'*' whT v v/W V y 000043123 TABLE 2 ANALYSIS OF 12 PPM VC STANDARD GAS % Rel. Std. Deviation Area 2724 2771 2830 2774 2875 2662 2.7% TABLE 3 REPLICATE ANALYSIS OF VCM SOLUTION % Rel. Std. Deviation 2276 2368 2296 2208 2230 2.7' TABLE 4 ANALYSIS OF VC INJECTED INTO A VIAL CONTAINING CARBON (lOyttl VC) Water Bath 90 for 2 Hrs. fu 11 10.7 13.2 7.4 8.3 11.9 9.4 10.1 8.9 COMBINED % REL. STD . 8.1 14.4 10.8 8.2 9.3 9.9 7.9 10.7 9.9 10.8 DEV. 19 .3% 12.5 9.6 11.9 10.6 11.9 7.0 8.8 12.1 9.0 6.5 Bath 70 for : % Rel. Std Deviation 10.4 13.2 10.6 10.0 10.6 8.8 10.6 8.0 6.9 11.1 17.5% TABLE 5 ANALYSIS OF CARBON EXPOSED TO VC IN A GLASS FLASK (Equal Weights of Carbon) Area 90 for 1 Hr. Area @ 70 for 2 Hr. % Rel. Std. Deviation 1006 995 1028 1003 973 962 1031 866 894 966 S.6% 909 1021 925 917 1095 947 797 1006 1032 1023 8.8% TABLE 6 ANALYSIS OF PROPANE ON CARBON % Rel. Std. Deviation 2.18 2.20 3.30 3.89 4.14 2.58 4.34 3.32 26.3% TABLE 7 ANALYSIS OF CARBON WITH IQ^fll VC BY CARBON DISULFIDE METHOD 9.1 9.6 8.8 13.8 9.2 7.7 8.0 8.4 8.8 17.9 10.5 7.9 9.2 16.1 8.8 8.8 8.2 9.0 10% Rel. Std. Deviation TABLE 8 ANALYSIS OF CARBON BY F-42 AFTER ULTRASONIC TREATMENT 10 nl STD ON CARBON (DMA EXTRATANT) Rel. Std. Deviation AREA 3380 3598 3477 3509 3368 3494 3745 3426 3.5%