Document Z42vXK15qv4y3Oo2o1qV80JrV

ENVIRONMENTAL VINYL CHLORIDE MONITORING Presented By: John Knowles George D. Clayton & Associates 25711 Southfield Southfield, Michigan At: AICHE Meeting Houston, Texas March, 1975 vvc 00204o5 ENVIRONMENTAL VINYL CHLORIDE MONITORING Early In 1974 it was reported to OSHA that several deaths due tc angiosarcoma, a rare liver cancer, may have been caused by occupational exposure to vinyl chloride (VC). As a result of this an emergency temporary standard was implemented lowering the permissible ceiling concentration to which industrial workers could be exposed from 500 to 50 ppm. Ongoing toxicological research demonstrated that liver cancer could be induced in mice at exposure levels as low as 50 ppm. A permanent standard was established limiting occupational exposure to one ppm averaged over any eight-hour period and a ceiling of five ppm averaged over not more than 15 minutes. In addition an "action level" of 0.5 ppm was set. The permanent standard requires that a 21 employee exposure levels be determined either by area or personnel sampling. If the exposures of work ers are maintained below the action level no routine monitoring is required, if between the action level and the permissible limits, monitoring must be at least quarterly and if above the permissible limits at least monthly. It is the intent of this discussion to review methods avail able t c c c n d u c t monitoring of e rr. o 1 c y e e exposures v: tr.tr. t r. e w c re place. In addition, procedures and techniques for monitoring emissions fr om industrial operations are presented and discussec. Obviously, the simplest way to conduct a survey is by using some sort of direct-reading instrument, that is, an instrument which is capable of sampling a volume of air, performing a quan titative analysis ar.c displaying the result. The advantages c: u s i n a sue n system include- V V C 0 0GO2.O2*-06 2 a) immediate estimations of the concentration of the con taminant, permitting on-site evaluations; b) provision of permanent four-hour records of contaminant concentrations; c) attachment of alarm system to instruments to warn work ers of build-up of hazardous situations; d) reduction of number of manual tests; e) reduction of number of laboratory analyses; f) provision of more convincing evidence for presentation at hearings and litigation proceedings; g) reduced cost of obtaining individual results. However, the majority of these systems measure some physical property of the contaminant being studied and consequently are prone to interferences. Before using results obtained with a direct-reading instrument supplementary data should be obtained relating to possible interferences. Calibration of any direct- reading instrument is an absolute necessity - the frequency of the calibration depends on the instrument in question and the degree of accuracy required. The main application of such instruments is in area moni toring, that is, in monitoring the concentration of VC in a given area at a given instant in time. Personnel monitoring imposes limitations on the sampling procedures and is dealt with later. There are many principles of operation which have been used in direct-reading instruments for VC. The ones discussed below are those with the highest applicability. vve 000020^07 1 C om bustion Conductivity If vinyl chloride is heated in air at 1000C it will decompose Co C0.} and HC1. 3 y bubbling the H C 1 into pure water a change in conductivity will be produced which can be measured. is very rapid, taking only about thirty seconds. The method Interferences are common since any substance which changes the conductivity of the sample solution will interfere. How ever, if this technique is used in a plant where interferences are known to be low or can be compensated for the method can pro vide a rapid means1' of analysis. In all instruments using conductance as the measured parameter temperature variation becomes a very important consideration and corrections should be made accordingly (2% per degree centigrade). Flame loni zation Detectors (F.I.D.) If an organic molecule is burned in a small flame, carbon ions are produced. By applying a potential across the flame anc meas uring the current produced the concentration of the organic mole cules can be measured. This is the principle or operation used in F.I.D. `s. As with the conductivity method above the analysis, while being extremely fast, is only useful for an atmosphere known to contain only / C since any organic ; o m pc un a will 5: ' In fra:;: Analysis Vinyl chloride absorbs I . R. radiation due to the movement in its chemical bonds, the amount of absorption oeing oepen cent method is not entire 1v specific as interfering substances may 4 commonly be present. As the spectra are additive the contribu tion by an unsuspected minor component can be overlooked, leading to incorrect data. Due to design limitations, with the best cur rently available instrumentation an accuracy of -+-107. is the best that can be obtained practically with this method. In general only about two minutes are required for analysis although additional time may be required to change samples. With a long path length sensitivities of 0.7 ppm can be expected. Gas Chromatography (G.C.) This procedure physically separates components of a mixture by passing them through a column containing an inert substance coated with an organic silicone oil. Each component will be absorbed by the stationary liquid and desorbed by the carrier gas stream in a unique manner. Over the length of the column che adsorption-desorption process is repeated many times resulting in a separation of the different components as they are eluted from the column. If the emerging gas stream is made part of the fuel supply for an F.I.D. the compounds can be measured by the response they produce. The signal produced is amplified and used to produce a peak on a strip chart recorder. The concentration of the compound is proportional to the peak area. Rugged, battery-operated, portable gas chromatographs have been refined to the point where they may now be considered prac tical for many field study applications. f These instruments come complete with gas sampling valve, rechargeable batteries, appropriate columns and self-contained oooozo^09 vve 5 supplies of gases which provide 8 Co 20 hours of operation de pendent on the flowrates and must be recharged using high pres sure gas regulators. The retention times of the compounds of analytical interest must be determined in the laboratory for a given type of column, as is true for the laboratory type chroma tographs , Although interferences can result from other compounds having similar retention times (these can often be removed by changing the column conditions) the primary advantage of the method in the present consideration is its ability to determine vinyl chloride in Che presence of many other compounds. The sensitivity of the method is excellent and extends down into the ppb range. For continuous monitoring the major disadvantage is the time involved in the analysis, which is approximately 30 minutes. The four above methods are the basis for the majority of direct-reading instruments used for vinyl chloride determinations. As stated earlier results obtained using these techniques are only meaningful when used in conjunction with results obtained using more specific sampling and analysis techniques. The method of choice for vinyl chloride analysis is gas . chromatography. This is relatively fast and provides the accuracy required by the OSHA standard. The standard VC concentration for the gas chromatograph can be prepared in two ways. 1) By injecting the 5.0 ml of pure vinyl chloride into 10 liters of air in a Tedlar bag. This gives a 500 ppm VC standard. (.1 i :: v yl i ; o 1 ; i n g 1 ml o f v i n y l r h l o r i <j o i l 0 1. 31 ( c. one. = 255.6 u g / m 1 ) . Further dilution pro pared as 11 e c u s s a r y , usual!;.- a L : L 0 scries to d e s i r u >_ lever level Five v,i 1 1 ilitc rs or 5 a 1 respectively of the above standard s are injected into a gas chromatograph equipped vith a column pa c /. e w i. t h 10::, f ;' 3 0 on 3 0 / 100 mesh c h r o~ o s o r b (other columns c an be with a column temperature of about 60. The limit of detection each i \ , .a i is o t the .jrJtr of 100 p i c o g r am s which gives a 'IcIoclL'.-, limit of around 10 ppb in air (5 ml injection volume) wh in- h i. :i .a; 1 I below the level specified in the 00 11 A stand a r d . 0 a m n 1 e ; f u r a n a 1 y a o s can be collected either by grab or iiit c grated campling methods, In grab sampling a gns collection dev Ice or known volume under measured conditions or temperature, and pressure Is used. The sample is collect a d over a p c rio d of s e v e r a l second:-: up to one or tVu minutes in a :s im um . A v i do varlGt y of c o 1 1 e c t i o n d a v i t c s is u sub to obtain grab s am p Luc. Tho s c i :: c 1 u c a v a c u u m flasks, vac u u m b c t L 1 u s , g a s - or 1 i u i. d - d 1 s p 1 a c e m e n t type collectors, glass b o i. r . ; s , springes a u plastic bags. Tine t e m d e r a t u r e and pressure at the time of s a n - nil n -- u s t be recorded to permit the retorting : f r. a 1 t i c a 1 u a T n instances where the atmosphere to be s a moled remains 0 i v constant with time, grab samples will represent the / :i v e r; n jr i ) . o >> c o n y a I; i. i Howc-vor, vhero the .Uaos fhe r i.c con c e n t: r: n t i o n vaii^s appreciably i n accordance with processing opn-.u ions i ii I oilier laciors, a s u f f i c i e n t number of samples ,nus t be taken to determine the average concentracion of the contami nant under each of the relevant processing or ambient conditions. Grab samples have the primary advantage over most other methods of sar.ipJ in:; in that che collection efficiency is 100 percent. Once the sample is collected a portion (usually 5 or- 10 ml) can be removed with a gas syringe and injected into the gns chrc- iiia tograph . f, y suitable choice of column conditions the vinyl chloride peak can be obtained free of interference effects and the concentrations of vinyl chloride calculated accurately. If other substances are known to be present in the atmosphere, such, infor mation should be submitted with the sample as they may cause interferences. For the same reason it is also a good idea if possible to submit a sample of the bulk materials used in the process for the purpose of identifying all atmospheric contam inants which may be collected along with the vinyl chloride. When sampling atmospheres which have variable concentrations of a contaminant (such as vinyl chloride), for evaluation of personnel exposures and time-weigh ted average concentration levels, the use of an integrated sampling method is required. In this method ambient air is drawn c-ver an adsorbing medium (or through an absorbing medium) at a fixed rate for a known period of time. The large volumes sampled by this method lead to increased sensitivity in the overall procedure, Analysis of the adsorbed vinyl chloride collected by this method produces an estimate o I: the average concentration of che VC i n the sampled air volume. 0 2) Hv dissolving 1 in l of vinyl chloride in 10 ml C$2 (cone. = 255.6 pi g / m 1 ) . Further dilution prepared as necessary, usually a 1:10 series Co desired lower level. Five milliliters or 5 p.1 respectively of the above standards arc injected into a gas chromatograph equipped with a column packed w 11 h 10 X Hi.', 3 0 on 8 0/ 100 mesh c h r ora osorb (other columns can be used) with a on inrun temperature of about 60. The limit of detection for each injeci j on is of the order of 100 picograms which gives a doLecLi.>n limit of around 10 ppb in air (5 ml injection volume) whit, h in we] I lie low the level specified in the 08 HA standard. Samples for analyses can be collected cither by grab or inte grated sampling methods. In grab sampling a gas collection device of known volume under measured conditions of temperature, and pressure is used. The sample is collected over a period of sev eral seconds up to one or two minutes maximum. A wide variety of collection devices is used to obtain grab samples. These include vacuum flasks, vacuum bottles, gas- or liquid-displacement type collectors, glass bottles, syringes and plastic bags. The temperature and pressure at the time of sam pling must be recorded to permit the reporting of analytical data under standard conditions, typically 2 5 0 C and 760 mm of mercury for in-plant samples. The grab sampling method is useful only for determining instan taneous vinyl chloride levels and Ls best used for determining VC levels in regions suspected of being out of Compliance. In instances where the atmosphere to be sampled remains reiative'y constant with time, grab samples will represent the VVC 000020411 8 B a s t c ,i 1 L y the Integra cod sampling method requires the follow ing components: 1) A sampling pump; 2) An adsorptive material capable of trapping ail of the VC passed over it during the sampling period. 3) A desorption mechanism to remove quantitatively all the VC. from the adsorbant at a later time to permit analysis. Any pumps used must be calibrated before use to ensure that the air flow measurexn.ent is accurate. In addition, if used for pe sonnel sampling, they should be light enough that they are comfort able to wear without interference with the employee's job. The adsorbing material most commonly used for organic vapor is activated charcoal. The capacity of a charcoal tube at any given flowrate is proportional to the surface area of the char coal and the affinity of the substance being sampled for the charcoal of a. given mesh size. Commercially produced tubes containing 100 and 20 milligrams of activated charcoal in the front and back-up sections;, respec tively, have been available for several years. Desorption of the organic vaoors adsorbed in these tubes by carbon disulfide rcll _ L owed as c h r cma t og r a o nl: termed t r: u 0 a sis o We have found that this method is ineffective in collecting VC at the normal flowrate of 1 1pm. The operational N1DSM pro cedure for VC presently in general use recommends a sampling rate of 50 ml/miuute for a volume of no more than 5.0 liters (giving a detection limit of less than 3 ppb in air). In order to utilize the 1 1pm sampling rate typically used to sample for organic vapors' with charcoal tubes, ww c have ^ 9 developed a larger sampling tube. This is an 8-mm (I.D.) glass > tube which contains two sections,of 900 mg each, of 20/50 mesh activated coconut charcoal. (Pittsburgh Activated Carbon). Each section has a breakthrough capacity of approximately 170 mg VC compared with 20 mg for the commercially available tube. Five and 50 ppm gas mixtures have been sampled at 1.0 liter/minute for up to two hours with no evidence of breakthrough. Water can cause an interference to the method by displacing VC from the charcoal. f If conditions of high humidity are encoun- tered sampling using the larger capacity tube becomes a neces sity. After sampling is complete both sections of the tube are de sorbed separately using carbon disulfide. If it is found that the back-up section of the tube contains significant amounts of VC there is a good chance that all the VC sampled has not been adsorbed onto the charcoal. If possible the desorption and the gas chromatographic analysis should be performed immediately upon completion of sampling. We have found that migration of the VC occurs within the tube leading to high VC values in the backup section. There is also some evi dence that polymerization may occur on the activated surface if the samples are left too long after sampling. Both the above phenomena can be avoided by immediately refrigerating the sam ples upon completion of the sampling. It has been shown that with these precautions a recovery of 95% of the sampled VC can be attained within 48 hours of collection. Ambient Air Surveillance The most logical extension of concern for exposure to VC beyond the workplace is in the ambient air quality downwind from - 10- an operation discharging VC to the atmosphere. The determination of the concentration of a specific pollu tant, such as VC, in the ambient air over a given area is a complex problem. Many factors such as local topographical dif ferences, atmospheric chemical reactions, fluctuating emission rates from various sources and variable meteorological conditions cause spatial and temporal influences in air quality. This space time dependence of air quality must be dealt with on a quantitative basis in order to acquire statistically reliable data of air quality and to meet ambient measurement objectives. -Specific guidelines for ambient VC sampling have been issued rccently by the U.S. EPA. Minimum site locations are specified and a sampling schedule is provided. These guidelines are intended for source monitoring only and are to be used in the "interim" until the state-of-the-art of VC sampling is improved. The EPA guidelines regarding ambient sampling in monitoring VC state that sampling sites should be selected downwind and in ^ the plume of atmospheric emissions from the plant. Samples should be collected only in areas where local residents or neighboring industries would be exposed. They specify that grab samples should be used to the maximum degree possible; however, the shorter the sampling period the larger the variation in concentration values becomes. This leads to poor precision in mean concentration for areas of high variance as the grab sampling data have little quan titative value and can be in no way related to long term exposure. The basic integrated sampling method dealt with above can be adapted for use in both ambient and stack sampling by enlarging the size of the adsorbent tube (two, 6.5 g charcoal beds). Sampling with these tubes for a minimum of one hour is at present ............................................................................................... i:................... .................. ... ...... .......... _........ . .ooo^5 11 the most accurate method of determining long term exposures. Optimum sampling rates for these tubes must be determined, gen erally a maximum rate of 1 1pm is recommended. Background ambi ent coucentrations are obtained by simultaneous sampling upwind of the plant. Although many of the sources of VC emissions to the atmos phere are fugitive in nature several major sources emit meas urable quantities of VC. Measurement of such emissions can be used to document m'aterial: losses, as an aid for designing and as a means of testing control equipment, as well as determining compliance with regulations. Samples are usually collected in triplicate for each stack condition. The sample time is normally 30 minutes but if the operation is cyclic in nature it should be extended to include at least one complete cycle. The sampling train normally consists of a probe, adsorption tube, flow meter and pump, however, if stack conditions are such that moisture or other condensate collects in the sampling hose in front of the charcoal tube, a dry trap should be included in the train. Any condensate collected in this trap should be treated as part of the sample and analyzed accordingly. vve 000020416