Document 44D9gaebDmJR5dz3ZzwV2M4p

/T*0 j it BUSINESS CONFIDENTIAL ^ "PROJECT REPORT -> * VINYL CHLORIDE MONOMER: llECt&VED a' USE OF MIRAN I PORTABLE INFRARED GAS ANALYZER J',u 13 1374 FOR CONTINUOUS AREA AIR ANALYSES 1L N. WHEF Cl*- Jp- author* G. F. Hurley supervisor. N. H. Ketcham date* June 4, 1974 projbct mo.. 9IQA20 pile no.. 19541 SUMMARY The Mlran I Portable Infrared Gas Analyzer can be used to continuously monitor for vinyl chloride monomer in air. With the instiument para meters chosen for the present work It was passible to determine concentrations in the 5 to 300 ppm range. While the manufacturer lists 0.7 ppm as the minimum detectable amount we feel that a more realistic value would be I to 2 ppm with the appropriate instrument settings. The mast serious problem in obtaining reliable results is caused by baseline drift. In field monitoring In the South Charleston Dispersion Vinyl Resins Unit it war found that daily adfustment of the baseline was necessary to maintain accuracy. A Miran II system, designed for such continuous monitoring, automatically and contin uously corrects for such drift. Polyethylene tubing was determined to be suitable for drawing sample for analysis from remote locations as distant as 100 feet. Variation in moisture content of the sampled air could produce minor err rs in the analysis. However, the error was small in 53 percent relative humidity air if calibration was done in 31 percent R.H. air. Because of the urgent need for monitoring, we relied on interference data supplied with the Miron which indicated passible interferences from 20 compounds "likely" to be present in processes which use vinyl chloride. After the on-site in vestigation, some addttonal compounds ware checked for passible interference. At the unit monitored vinyl acetate was used in some batches. We found It does in terfere at 10.7 p, so that, where both the chloride and the acetate were used, the vinyl chloride values will be high. In the range- under 30 ppm> the response Is equivalent; while over 30 ppm the response to the acetate is 50 to 60 percent of that due to the chloride. INTRODUCTION With the recently recognized need for reducing exposure to vinyl chloride, a need has developed for continuous location RESEARCH AND DEVELOPMENT DEPARTMENT CHEMICALS AND PLASTICS UNION CARBIDE CORPORATION SOUTH CHARLESTON, WEST VIRGINIA ucc 034630 BUSINESS CONFIDENTIAL -2 - 9I0A20 monitoring in addition to personal sampling. Since infrared absorption is one f th most promising meth ds for air analysis for vinyl chloride monomer, and w had a Milan I Portable Infrared Gas Analyzer it was decided to evaluate this instrument for continuous area monitoring. The Miran I Portable Infrared Gas Analyzer (Wilkes Scientific Corp.) was described recently in a report by J. E. Neff (1) who evaluated its use for methyl isocyanate ironitoring. Being a single beam dispersive instrument it can be set t any wavelength within the range 2.5 to 14.5 p or it can be used to scan the entire range. Thus, its operating principles are those of a typical infrared analyzer. Its uniqueness lies in the variable path gas cell provided which permits determination of chemicals in air in the parts per million range. In the list of compounds supplied by Wilkes which are amenable to IR analysis the minimum amount of vinyl chlorid detectable with the Miran I is given as 0.7 ppm* A pump, supplied, can be used to draw sample air into the gas cell In order to monitor continuously for any con taminant. This report gives the results of the laboratory investigation and discusses the use of the Miran I for continuous monitoring in the Dispersion Vinyl Resins Unit at the South Charleston plant. The Laboratory Investigation A. Calibration Curve For an initial plant application where the vinyl chloride concentration variation was not known it was decided to choose instrument parameters such that a range of about 20 to several hundred ppm of vinyl chloride could be determined. In the instrument, pictured in Figure 4, the black section on the left Is the gas cell. The samplepath length in the gas cell was set at 20 meters for the analysis and a 2.0 mm slit width was chosen. Standard samples were prepared in Saran bags by adding measured microliter quantities of pure vinyl chloride to known volumes of air. These standard samples were drawn into the evacuated infrared gas cell. A curve was prepared from the measured absorbance of vinyl chloride at 10.7 p which covered the range of 10 to 300 ppm (Figure 1). B. Remote Sampling The infrared light source is a heated nlchrome wire (82OP C). Far this reason the instrument is not intrinsically safe. Therefore, usage in a plant requires either some means to make it safe or operation in a non-hazardous area with ' ... sample tubing extended to remote areas where monitoring is desired. For an ex pedient solution we chose to use remote sampling for monitoring in the Dispersion Vinyl Resins Unit at South Charleston. A 100 ft. length of 0.5 inch polyethylene tubing was tested for use as a sampling ] )ne In a total recycle system as shown in Figure 2. To determine if there was any loss of vinyl chloride by absorption or reaction on the polyethylene ucc 034631 BUSINESS CONFIDENTIAL -3 - 9IGA20 after I ng contact the f II wing sequence was foil wed: 1. A nominal 50 ppm vinyl chloride sample made by adding a calculated measured volume of the pure gas to a clean 20 liter Goss Cylinder and pressurizing with dry air to 300 pslg was drawn into the evacuated Miran sample cell. A trans mission reading of 78 percent was obtained indieating 48 ppm of vinyl chloride. 2. The entire system shewn in Figure 2, pump, 100 ft. of tubing and cell was puiged several times with this standard mixture. To do this the cell was evacuated and then filled to 0 psig with the standard sample. This was recirculated through the system including the polyethylene tubing to purge out the air in the pump and tubing. After several minutes circulation was stopped, the cel! evacuated and the purge process was repeated. Readings obtained after successive purges were 78, 77*5, 78.0 per cent indicating steady state conditions. 3. Using the pump, a Neptune dyna-pump (Model 3) fitted with a teflon diaphragm,, the standard sample was recycled in the system at 4.61pm for 65 minutes. A 79 percent transmittance indicating 46 ppm vinyl chloride was obtained. This indicates that there was essentially no loss of vinyl chloride in the polyethylene tubing* Wall dbsorpHon for short terrh IcontaetWcs,' tested ds follows:. .. The polyethylene tubing and the Miron gas cell were evacuated. Then 10 ppm concentration of vinyl chloride in dry air was Introduced at the remote end of the tubing. The instrument showed 12 ppm vinyl chloride, which was the same as was found by sampling without the 100 feet of tubing. C. Effect of Moisture To determine the affect of moisture, the Instrument was set to 61 percent transmittance in dry air. Purging with room air, relative humidity (RH) 31 percent, changed the reading to 58 percent. With 53 percent relative humidity air the transmittance was 57 percent. The instrument was calibrated with room air of 31 percent relative humidity. We conclude that there is very little effect of moisture to 53 percent RH when calibration is made at 30 percent RH. D. Interference In the limited time available prior to the field menitortng, no attempt was made to determine possible interference at the 10.7 p wavelength chosen, (This is the most sensitive band for vinyl chloride). We relied on a Wilkes Report ucc 034632 BUSNIESS CONFIDENTIAL - 4- 9I0A20 (2) which discussed possible interferences. These are shown in tabl 2 for the predominant vinyl chloride bands. Subsequent to the monitoring in the Dispersion Vinyl Resins Unit, a limited laboratory investigation showed interferences of some compounds as dated with UCC vinyl chloride units. These are listed in table 3, together with absorption at 10.7 and 13.8 u. From this it can be seen that viiyl acetate would interfere at 10.7 p giving high vinyl chloride results, if present. The extent of the vinyl acetate interference at 10.7 p is shown in figure 5. For 200 ppm vinyl acetate, the calibration curve for vinyl chloride would show 105 ppm. At concentrations under 30 ppm the response is almost 1:1. Field Monitoring The instrument was transferred to the South Charleston plant where it was set up in the Dispersion Vinyl Resins Unit. The monitor was placed In th control room of Building 190 with 100 ft. of 0.5 In polyethylene tubing leading. to a portable remote sampling probe placed in the unit. A filter at the probe served to keep dust out of the system. The monitor system shown in Figure 3, includes a carbon trap for providing ambient air free of vinyl chloride for in strument zeroing. The septum is useful far injecting known volumes of vinyl chloride for calibration. Monitoring for vinyl chloride started on March 12 and was almost con tinuous until about May 3. Some of the Mlran results are given in Table I along with gas chromatographic analyses on similar samples. Initially, there were brief excursions to higher levels of vinyl chloride during periods when filters were dumped, reactors were being steamed or dumped, charging lines washed, etc. The levels reached varied, with some of the earlier higher values being over 350 ppm vinyl chloride. On one such excursion where a Miran value of 250 ppm was obtained the TLV Sniffer (Bacharach Instruments) showed 240 ppm. More recently values above 50 ppm have been infrequent. The only problem of consequence in the Miran I operation has been th instrument baseline drift. Daily adjustment hee been necessary to maintain a reasonable degree of accuracy. The maximum drift produced values where the Miran read 10 ppm while gas chromatograph samples gave only 0,5 to I ppm. DISCUSSIONS AND . The Mlrah I is a suitable instrument for monitoring for CONCLUSIONS vinyl chloride in production units If the following fac tors are considered: . I. Not being intrinsically safe, it must be placed in a non-hazardous area and sampling done remotely using inert tubing. UCC 034633 BUSINESS CONFIDENTIAL -5 - 9I0A20 2. An absorption band must be chosen which will be fre of interfering compounds. 3. Baseline drift must be corrected periodically - probably daily. Under satisfactory conditions/ it is sensitive to about I ppm and will perform reliably for continuous field monitoring. For more permanent installation!/ the Miran II system should be used. Here/ by virtue of having a double beam instrument, the affect of baseline drift is continually corrected. Also, the effect of one interfering element can be corrected by electronically cancelling its absorbance at the chosen wavelength. Results on the double beam instrument would be less effected by moisture on th salt window and dust an the mirror surfaces. RFERENCES 1. Evaluation of the Wilkes Miran Infrared Portcble Gas Analyzer for Determining Methyl Isocyanate and Other Compounds in Air, J. E. Neff, File No . 17992, Jan. 19, 1973, 2. Infrared Analysis of Vinyl Chloride at Concentrations Below 100 ppm. Appli cation Report No. 4, Wilkes Scientific Corp. Date of manuscript: 5-1-74 Date Typed: June 4, 1974 Attachments: 3 Tables, 5 Figures GH:ml ucc 034834 BUSINESS CONFIDENTIAL 91QA20 TABLE 1 VINYL CHLORIDE ANALYSES IN DISPERSION VINYL RESINS UNIT AT SOUTH CHARLESTON Date 3-13 3-13 3-13 3-13 3-22 3-13 3-14 3-14 3-14 3-14 3-14 3-19 3-19 3-19 3-19 3-19 3-19 4-4 4-5 4-5 Sampling Point Standard MIran cell exit It MIran probe in unit II II MIran cell exit II II II II Control room air Control room air II MIran Anal PP*" 90 0 0 0 12 0 12 8 10 8 10 10 10 8 6 8 6 5 7 8 Gas Chrom. Anal,, ppm 98 2 10 2 6 3 18.4 7.6 5.7 5.6 7.1 13 12 9 7 6 6 3.3 6 9.3 Remarks MIran zero drift effect No MIran zero adjustment for 3 days previous ucc 034635 BUSINESS CONFIDENTIAL 9I0A20 TABLE 2 POSSIBLE INTERFERENCES WITH VINYL CHLORIDE ANALYSIS Compound Acrylonitrile Allyl Chloride Chlorobromomethane Chloroform Ehtylene Ehtylene Dlchlorlde Freon - 11 Freon - 12 Freon - 13 Methacrylonitrtle Methyl Chloroform Methylene Chloride Methyl Methacrylate Perchloroethylene Styrene Tetrahydrofuran Trichloroethylene Toluene Vinylidine Chloride Vinyl idlne Fluoride KEY: W* Weak, Wilkes data (ref. 2) Vinyl Chlorld* Anolytlcal Bands 6.15 p -9.8 p 10.7 p 13.8 p M W S WW W W W S S M= Medium, W S S w s M S w s M M M W S S3 Strong W W W S W W S M S ucc 034636 busniess confidential TABLE 3 9I0A20 MEASURED INTERFERENCES WITH VINYL CHLORIDE ANALYSIS COMPOUNDS Vinyl Acetate Toluene Trlchloethylene Acetone Methyl Ethyl Ketone Isopropanol (Vinyl Chloride 20 meter path, 2 mm silt TRANSMISSION FOR< 1iw00 ppomr qt 10.7pi at 13.8 i 70% 95 32 92 80 71 1 99 57) FIGURE 1 91OA20 FIGURE 2 REMOTE SAMPLING TESTING 91QA20 100*11/2" Polyethylene Tubing ucc 034639 I FIGURE 3 PLANT MONITORING SYSTEM 91QA20 - '. -v1, 'r.- UV-$; V1-"' ; : T,;; 'i ;i- Mlran Gat Cell Vv/ ucc 034640 FIGURE 4 Ml RAN I PORTABLE INFRARED GAS ANALYZER 910A20 ucc 034641 Distribution: Mr. W. M. Barrows/ 511 Mr. E. M. Bartrug, Jr., 514 Mr, J. T. Bouchier, 515 Mr. T. W. Carmody, NYO-31 Mr. J. L. Davidson, 513 Mr. D, E. Deese, 515 Mr. M. E. Eisenhour, 515 Mr. D. A. El Iwood, 514 Mr. D. L. Engle/Mr. B. B. Cowser, 515 Mr. R. L. Frantz, 515 Mr. C. E. Fry, 514 Mr. F. J. Garcia-Sharp, 293 Mr. J. E. Giffin, 514 Mr. R. E. Graebert, 511 Mr. H. R. Guest, 511 Mr. R. J. Hanna, 511 Dr. E. Q. Hull, 511 Mr. S. W. Kasinski, 314 Mr, R. C* Lawrence, 511 Mr. R. G. Lilly, 514 Mr. S. V. Lucas, 380 Dr. W. R. Manning, 511 Mr. R. W, Martin, 511 Dr. W. P. Miller, 511 Mr. L. G. McMullen, 510 Mr. K. S. Morlock, 517 Mr. B. L. Murray, 514 Mr. W. D. Neal, 312 Mr. J. E. Neff, 511 Mr, D. R. Pauley, 511 Mr. R. E. Peele, 511 Mr, K. E. Ross, 515 Mr. A. R. Ryrholm/Mr. J. L. Carvajal, 514 Mr, J, J. Scharf, 514 Mr. J. C. Schonbeig, 519 Mr. F. J, Small, 511 Mr. E. Sod, 515 Dr. A. B. Steele, NYO-28 Dr, T. T. Szabo, 511 Mr. K. G. Townsend, 526 IMr. R, N, Wheeler, 514 Mr. Fred Williams, 512 Mr. F. A. Woods, 511 Mr. W. C. Young, 514 Dr. N. L. Zutty, NYO-32