1/ 1 i !970 thing to a -<• ominal . n elecup provid,u ‘ these Industrial Hygiene Engineering and the Process - Environment System J. E. MUTCHLF.R* mbcr walls vapor or nc test K -f 3540 ic solvent, u of oner i twice aotti singly adsorption ■,( 0 liters 5( cd and nalysis. To volume it* . of pe:tu**tration 0.5,1, 1.3, to one sec• is measg ss tube li Parafilm. sj Environmental Research Laboratory, The Dow Chemical Company, Midland, Michigan ® A systems approach to industrial health would help integrate the concepts applied by engineers, industrial hycicnists. toxicologists, and physicians, and organize them into a strategy for health maintenance that is compatible and measurable along com­ mon lines. The advantages of a systems approach to industrial health are viewed in terms of (he application and reiincment of industrial hygiene standards, participation of the engineering professions, communication among all members of the occupational health team, and transition we arc enjoying with the increasing use of automation, process control and computer technology. Introduction I 'njeeted Oi with a vaa capped of all fourvions of •in. 100 if air were tl u : on the tuuixes of mnds were * .e tubes .• i assure t(s) by the in the de« ctly the o... sections aerials was ithqn tlisul- i TN HIS INTRODUCTION to Industrial ^■Hygiene’ Highlights, Professor Theodore Hatch1 reminds us that “it is the responsibility of industrial hygienists, toxicologists and phy­ sicians constantly to review the adequacy of the methods they employ to demonstrate health maintenance.” This is a chaiien.ee to all of us who have sensed the broadening horizons of our profession. For if we arc to apply the principles of industrial hygiene suc­ cessfully in this time of rampant technological change, we must keep abreast of new develop­ ments in the basic disciplines that, comprise our multidisciplinary profession. Effective management of occupational health programs requires—more than ever— coordination and teamwork among those spe­ cialties that function within this broad field. We need to know and understand how each discipline that deals with occupational health can complement and interact with other mem­ bers of the occupational health team. One measure, of the coropatibilitv of an interdisciplinary approach to occupational health is the emphasis we place on the con­ cepts and methods applied to industrial hy­ giene bv several specialties—by engineers, in­ dustrial hygienists, toxicologists and physicians •PfMtnt addreu: firnri* D. Clacton and .Auocia'ci. 25711 ScufJifield Road, Southfield, M'du«an Mi)?}, —concepts and methods which we have in common, and which can be reduced to quan­ tities measurable along common lines. One concept that relates to all the disci­ plines within the field of occupational health is that of an “acceptable exposure level” or an “industrial hygiene standard” for inhala­ tion of materials in the work place. Wc are familiar with threshold limit values (TLV4$). maximum acceptable concentrations (MAC’s';. and, in more recent years, the mul­ tiple guidelines of the Z-37 Committee of the America National Standards Institute (AN­ SI). Much attention has been given to the development of these standards, but it ap­ pears that there has been less concern for the validation and refinement of those guidelines, and clearly, even less concern for the applica­ tion of industrial hygiene standards in the work place. Specifically, two aspects of the application of standards are glossed over to our disad­ vantage. First is the fact that TLV’s and other hygienic standards arc not easy to use properl)’. Second, many of the engineers in industry who could he. are not using the con­ cept of an industrial hveicnc standard in their decision making. Therefore, we do not benefit as fully as we could from the assistance the engineering profession could give us in the overall control of industrial health hazards. U’hat can be done to improve this situation? 233 i 1~~ AP00000213 Mareh-AprH, 1970 534 EXPERIENCED .WORKERS, NCW employees'---. ENERGY INDUSTRIAL OPERATION __ WASTE “^EFFLUENTS utc in the total environmental burden of air, water and soil pollution. Third, there is an output irotu the system winch is o{ direct con­ cern to those of us interested in occupational health—the exposed workers and retirees, He have implied, then, that the work en- .SALABLE saw MATERIALS PRODUCTS Figure 1. An industrial operation. The Need for a Systems Approach Clearly, the application of knowledge from engineering and medical research, industrial toxicology, and field experience in environ­ mental control requires some communication system for the essential interplay to exist among the various parent disciplines within our field. If the communication is effective, that interaction can thrive. On the other hand, if each discipline inter­ ested in occupational health attacks a parti­ cular problem independently, we likely will not meet the total need. Thus, we should adopt an integrated approach in organizing ourselves to deal with our common concerns. In the integration of interdisciplinary inter­ est which focuses on a complex problem, one promising technique is ‘'systems analysis." Webster defines a system as a ‘‘regularly inter­ acting or interdependent group of items form­ ing a unified whole.1” In simplest terms, a system consists of a set of components which, when combined, produce a useful result. A “system” in terms of industrial health might be an organizational framework that defines the strategy for assuring the hralthfuiness of a worker population. To that end, the system might be composed of certain subsys­ tems which could be examined and treated separately. If we step back and look at an industrial operation and examine the various inputs and outputs from that industrial unit, wc observe an interesting fact: From a larger viewpoint an industrial operation takes three inputs— new employees, raw materials, and energy— and converts them into three distinguishable products (sec Figure 1'. First, there are the salable products which arc sold in the market­ place for profit. Second, there are waste effluents which take the form of solid, liquid, or gaseous materials—ami in turn may contrib- Specifically. every work situation has three ele­ ments: a process (perhaps a limited operation or function task!. a work environment, and a worker. Within an industrial unit the process, the work environment, and the worker arc not totally independent entities. Each is affected, or could be alTected, by the other two ele­ ments. In many cases, one cannot alter a process or even change a job procedure with­ out changing the quality of the work environ­ ment. A worker often has enough latitude in his work procedure to create differences in the degree of exposure he will encounter on his job. AU three of these items-~the process, titc work environment, and the worker—must receive attention in an occupational health program. It is at this point that some of the func­ tional differences within the occupational health team become important and ncrcssary. Clearly, the physicians, the nurses, and the supportive role of toxicology must deal with the worker and his health: the industrial hygienists must deal primarily with the work environment; and the logical extension of this approach would have the engineers dealing with the process—which, after all. is causing the environmental stress in the first place. The Need for Feedback Control of health hazards in the work en­ vironment can be neither comprehensive nor complete unless we systematically use the in­ formation that such considerations yield. In a self-rcgulatccj system the key element for con­ trol is "feedback.” This refers to a circularity of the flow among two or tnorc parts of the system structure. Such an arrangement assures that the output of the system is maintained at a desired level. Typically, this requires con­ stant monitoring of the system output, a com­ parison of thfc Output with standards, evalua­ tion of any discrepancies, and a flow of infor­ mation concerning the abnormal deviations Amt so th.u ncre-a Shllph' Likecritical must k mental feedli.-u statuki tinunlh and n; Fig..; environ the co// idating system.) er with CX|JO<(*[ hygicut carinj! . grates field im goal. This ijidusir: basis h\. live inc represet we ha\ coloeira exposed on his i sured jz: The . that ]>;i> the Pro oppose with th« If we portion ing wm control: The fV-oc Ficuri System, control applk.it Most er. with tin AP000002I4 American Industrial Hygiene Association Journal back to other elements in the system structure . so that the procedures may bo changed if necessary. Figure 2 shows an example of a simple closed-loop system with feediuirk. Likewise, in industrial he a ft h we have a critical need for feedback information. We must know the adequacy of specific environ­ mental control techniques. We also must have • feedback concerning our industrial hygiene standards. Useful standards should be con­ tinually verified or modified as our knowledge and experience grow.'1 Figure 3 shows a system both for assuring environmental control—including feedback on the control techniques employed—and for val­ idating industrial hygiene standards through systematic documentation of exjiosure togeth­ er with systematic medical surveillance of the exposed worker. This combination industrial hygiene and medical survey is a simple appli­ cation of a “systems analysis’’ because it inte­ grates several functional specialties in our field into a unit ihat moves toward the same goal. Tliis particular approach centers u]xm the Industrial hygiene standard that could be a basis but by no means the only basis for effec­ tive interaction among the parent disciplines ‘ represented in the overall system. Here then we have a strategy for harnessing the toxi­ cological feedback that is available fvotn the exposed worker in terms of the effect, if any, on his health at the levels of exposure mea­ sured in the work environment. The engineering function is represented in that part of the overall system that deals with the Process and the Work Environment—as opposed to that part of the system which deals with the Worker. If we separate out the Process-Environment portion of the overall system we notice a-strik­ ing similarity with the typical engineering control network. nf air, «• it an Cl illat. .sal •t*v u* •n« •m. ret* elf('ra(><>n .. d a ;>t css, :m- not w clc* am r a e viili'ii onin v in ■nee* in nter on pi :c», .Hist health tc •mcpuuonal ’COSSary.' ai the r; with idustrial h<’ work n ' this .ui.ilins causing v i cn(s i nor e tiic in•Id. In a I • conii larity t* of the H -usurcs i. ird at iiu*s con- Figure 4 depicts the Process-Environment System. Tltis is represented as a closed-loop control network that stresses the engineering application of industrial hygiene standards. Most engineers arc not primarily concerned with the requirements of industrial hygiene. /alua- - inforImntions t Fioure 2. A typical control system with feed­ back. Yet, effective communication on how engi­ neering techniques can be utilized is an essen­ tial need and deserves more attention in practice. The advantage we have in this respect is that engineers in genera] tend to deal with as large a system as they can handle from a com­ putational standpoint. Therefore, the indus­ trial hygiene standard could be used as an engineering standard in a larger system that includes the work environment. Specifically, the design of a process should be directly related to tire acceptable degree of environmental contamination by materials used or produced by the process. Environ­ mental problems, unfortunately, are often created bv poor engineering design. In other The Process-Environment System .1, a coms. 235 Ficcrc 3. An overall system in occupational health. l 1 AP000002t5 236 J Figure 4. The Process-Environment System. 1 words, environmental control must start on the drawing board. Second, the operation of the process can vary in a manner that will cause fluctuations in the level of contamination of the work en­ vironment. Again the acceptable exposure level serves as a guidepost toward which ali future chances and improvements in the op­ eration of the processes should be directed. Third, the industrial hygiene standard di­ rectly affects the measurejnent of contamina­ tion in a particular work environment—such measurements to be used for judging accept­ ability of control, following comparison with the standard that applies in that particular case. With airborne chemical X, for example, there may be several ways to measure and monitor airborne concentrations of that con­ taminant in the work room air. The level of the threshold limit will often govern what type of instrument is used, and thereby the accuracy of the resulting information. Fourth, we often think of a threshold limit value as the standard against which the esti­ mate of the exposure is judged, but it is not enough to take such judgment tar granted, for often the comparison of the measurement and the standard is made with inadequate infor­ mation. Especially in the use of multiple guidelines, such as those of the Z-37 Commit­ tee of ANSI, we reed a carefully planned approach—both to traihrr the proper infor­ mation and then to use it meaningfully. March-Aptil, 1970 Finally, the measured deviation from the industrial hygiene standard frequently points the way to the type of engineering control or improvement that results when the measured concentration exceeds the standard. In other words, how far must we journey to arrive at the desired degree of contaminant control? If the deviation from the TLV is near zero, there is a different set of alternatives available than if the measured exposure exceeds the TLV by a gross amount. Today's technological advances have brought us closer to the day when we can control the work environment just as we routinely control a manufacturing process. First we have seen the.development of con­ tinuous monitoring, which can be an effective diagnostic tool for the identification of pre­ dominant sources of contamination in the work environment, the proper selection of contaminant control methods, and the docu­ mentation of occupational exposures. The second development in this respect is the digital computer. When coupled to an en­ vironmental monitor the computer provides an effective means of reducing voluminous amounts of information into manageable form. These data, in turn, can be used to show both short-term and long-term trends of exposure. When correlated with day-to-day plant oper­ ation such information can result in even lower exposures.*-4 In addition, such data can be used to describe the exposures of workmen in what is in effect a continuing industrial hygiene survey. These two developments parallel the trend in the process industries where “on-stream process analyzers*’ and “computer applications*’ have mushroomed into prominence. Automatic regulation of product quality is commonplace—and the marriage of instru­ mentation and the computer has provided the engineering profession with some powerful new techniques. This is important to our field because our allies in the engineering profession arc in­ creasingly well-versed in such techniques, and to the extent that engineers become engaged in environmental control, we can expect to lake advantage of these and other technologi­ cal advances in controlling the quality of the AP00000216 I Aftil, mo \ k im the •rttly fioiiits t ‘irol or - i -ustiri'd h in mJjrr *• arrive at U i»mroi? n r aero, < available xc--d* (he ck t have we can m as we nt troccss. nt of conm effective or •>{ preoi in the Section of (’■■ docu• respect is ■-( to an on. r ’ovides ol. ninous -'able form, sfc-w both i' >osurc. tia'iit opert in even h• ta. can ’ > rkmcn industrial ( trend ’on-stream f applica. « nonce, q lity is of insiruovtrted the 1 werful ■oause our n re im <!e enliiii; 200 ppm Acceptable ntimlter ot penV concentrations ot 5-minute duration lor fbhr period dustrial hygiene standards that are available for carbon tetrachloride. To judge the ac­ ceptability of a particular work environment using these parameters, we would need a mea­ surement in the worker's breathing zone at least every 5 minutes for 8 hours. This mini­ mum of 96 samples would then have to be reduced to the various sampling statistics that could be compared with the standards. Relatively little attention Isas been given to this problem of characterizing the work cnviiujuncnt in a manner that will allow proper application of hygienic standards: it must re­ ceive more attention in the future. Table II shows one exposure summary that could be used to judge the intensity of exposure along several guidelines, both short-term and long­ term. This exposure summary ewers a 7-day pbriod and reduces over 100.000 measure­ ments to an informative, compact table. The information comes from a continuous environ­ mental monitor linked directly to a digital computer. Documentation and Feedback c The third advantage of the Process-Envi­ ronment concept is that adequate environ­ mental surveillance allows the use of feedback to provide a chcc-k on our environmental con­ trol efforts—to the extent that we correlate the environmental information with medical findings on the same worker population, we can use that feedback information to validate or rclinc our hygienic standards. As the problems of concern in the work room from environmental stresses become more subtle and complex, the need for sys­ tematic studies on the relationship between exposure and consequence increases, and we need more than evet to complement the re­ sults frotn laboratory studies with planned systematic findings on groups of people ex­ posed to the suspected agent under real conditions.3-®-* Summary * Progress in our field—occupational health —will continue to depend on a “team ap­ proach,!J As Professor Hatch1 pointed out, we must continually examine the methods we use to achieve results. Surely we must go beyond that and be willing to apply new techniques to our field whenever it can be demonstrated that such application will en­ able us to do a better job in conserving the health of the worker in the industrial environ­ ment. The development of our profession will depend on how well we integrate the common interests that each discipline brings to the field of occupational health. Our progress will also Table II Data Summary from a Continuous Environmental Monitor Department—chemical plant Material —rnethvl chiuridc Date —2/1J/69 to 2/23/69 Sampling Location (raw data) 4 3 Job Classification 5 6 A 61.3 79.3 26 A t 2 Mean concentration 14.3 3S.8 Standard deviation 13.9 29.3 9.3 19.4 9.1 22.4 21.0 19.3 100 ppm a.s 1.9 0.5 0.2 0.1 49.4 9.1 0-2 0.0 0.0 60.5 20.4 3.4 0.4 0.3 100.0 16 4 0.2 0.0 0.0 100.0 A 31.4 l.l 0.1 100.0 inn.o 14.5 0.3 0.0 39.6 13.5 1.4 0.1 0.0 933 -1.9 45.7 3(1.7 17.3 66,2 62. J 51.1 43.0 33.4 107.0 £4.6 fci.6 73.7 47.6 9I.G 82.2 61.0 34.9 45.C 164.7 145.0 3141.7 -*4.9 69.6 170.3 lSn.2 121.4 107.9 87.8 76.2 67.6 51.0 41.5 30.5 SoO ppm 42.6 Maximum concentration 30 mJnijfrj 4 Hours American Industrial Hygiene Association Journal lie measured by how well we find compatible, systematic methods of using the newest tools available to each basic discipline in our field. A systems approach to industrial, health would help us to take advantage of the tran­ sition we will enjoy with increasing use of automation, process control, and the digital computer. From a systems engineering point of view there is really little difference between controlling the process and controlling the work environment associated with it. A logi­ cal point of reference for the transition ahead would be the Process-Environment System. 239 References !. Ifttcit. T. F., in Cratlfv, L. V.: /riduttttaf ihikhi/in. Vo], I,, |i, !. Indi5tri.il )k?:cne Fuuniijtioe of Ammo. litr.. 1’j. 2. H'rbilrr'i WiruMi Sr:r CV.'* t'tti" hi'I'-’-n \, G. i C. Mrrri.1111 Ci'Iiiimiiv. >i>rmm. f'.. I).. K. 1). Silent, amt f. !,. Mutouw: Mt.mii.firn Liiiouiro to Vim I Clil.iria* Vir»ir; ilrciih Anjhir- and u('ntiiiHmi. Air S.uunltus. .inrr. Jai. /- mi i)71mat). 6- Hvroi, T. F.i Si«nifii;s«u DimernioM of tin1 Dosc-Rtsi.nnsr Rrljiionthiir. .4rri. Fill!iron, iitilth hj: 571 'I'-H.Sl. i. StOKiNcm. It. E.; Indutiriil Coniri)uj«!nn in TtireshoW Limit Yaiuot. Areh, Envirf. tjtallh Iv: 0!V (ItKij |, Reeeiied June 9, JW9 Gordon Research Conference The 1970 Gordon Research Conference on Toxicology and Safety Evaluations will be held at Kimball Union Academy. Meriden. Xcsv Hampshire. July 2731. This conference lias been of considerable interest to industrial hygienists for a number of years because of the close relauanshio of problem* encountered in industrial health work. This year the Chairman is J. Wesley Clayton. Jr. and Vice-Chairman is Edward I). Palmes. The program is as follows: July 27. (R. A. Scala, discussion leader): H. H. Cornish. '‘Serum isozvmc* and organ damage", j. L. Radnmski, "The metabolism of 1- and 2-naphthylamine as related to carcinogenesis.'1 iS. L. Fricss. discussion leader): J. M. Barnes. "Toxic substances and the nervous system.'1 July 28. (S. Epstein, discussion leader-: C. Jacobson, ‘"Test systems for the reproductive evaluation of mutagens''; D. J. Kilian. ‘‘The role of cvtocenics in evaluating the toxicity of chemical compounds.” 'H. Y. Mailing. Discussion leader): C. Valenti. “Effects of psychotropic drugs on mammalian chromosomes in vitro and in vivo observations.'' ' July 29. (C. H- Him-. discussion leader 1: R. E. Hodges. “The role of human studies in evaluating potentially hazardous agents'1: L. H. Schmidt. “What nonhuman primates have taught us relative to the ac:ion of chemicals on living systems.1' (Sidney Leskovitz, discussion leader) : B. Pernis, "Immunological problems in toxicology." July 30. (W. B. Ennis, discussion leader): D. O. King, “Rational Manage­ ment of the environment'*; K. C. Bartons. ‘‘Benefits of agricultural chemicals.’’ (J. P. Frawley, discussion leader): L. Cole, “The chemical threat to our en­ vironment.” July 31. (E. D. Palmes, discussion leader): T. Foin, “Technology, popula­ tion and the environment." Attendance at the Conference is by application only. Application forms and full information regarding requirements for attendance may be obtained by writing to Dr. Alexander M. Cruicksliank. Director Gordon Research Confer­ ence, Pastore Chemical Laboratory, Universitv of Rhode Island. Kingston. Rhode Island 02881 (Telephone: 1)01-783-101! U Applications for this Con­ ference should be submitted before June 1, 1970. Rented from AMERICAN* INDUSTRIAL HYCIKXE ASSOCIATION JOURNAL Volume 30, Xovrsnbcr-Dccctnber, 1069 Monitoring Exposures to Vinyl Chloride Vapor: Breath Analysis and Continuous Air Sampling EDWARD D. BAR.ETIA.* RICHARD D. STEWART, M.D.,* and JOHN E. MUTCHLERt Department of Environmental Sfedicl’-e Marquette School /if Medicine. Miluaukte. U'hconurt, end Environmental //- • '' -fion, Biochemical Re*farrh Laboratory, The Dote Chm.uti. Company, Midland, Michigan ^ An environmental survey was conducted to determine the tinw-veiphied avera^ exposure (TWA) of & group of chfniic.il plant workers to vinyl chloride (VC)1 vapor. This survey featured continuous multipoint atr sampling and analysis using an nfrared spectrophotometer. The inhalation exposure data were digiiircd and rccord:d on paper tape for subsequent computer analysis and derivation of daily TWA values for each worker. A breath sampling program was conducted concurrently with the environmental survey, and a series of breath decay curves relating postrxposurc breath concentration to vapor exposure were derived from the data. To validate the breath curves derived from on-the-job data. po«tevposure breath curves were also constructed from breath data obtained following experimental human txposures to carefully controlled concentrations of VC1 vapor. The close agreement between postexposure breath concentrations at the corresponding TN\ .Vi obtained by each 'of the methods suggests that cither continuous air monitoring or breath analysis is valid for cuitimine the worker’s individual daily exposure to NCI. and provides further evidence that breath analysis is a useful industrial hygiene technique for evaluating vapor exposure. bv the large number ami variety ot tasks per­ formed by todav's modem chemical plant worker. fnliowtner | Si « ams (o»c«M>etiateJI»TD X Of IM THC COKCCUMI'M IfCCCCCO TUT l-e-H FlGVRi. 2. Exposure profiles expressed as VCl vapor concentration versu« the exposure frequency distribution for the four job classifications. cap glass vial (Figure 4'. The overall length of the pipet was about 9 inches, so it could be conveniently and inconspicuously trans­ ported to and from work in a lunch bucket. The plastic caps were lined with six layer* of Saran film identical to that used for the construction of Saran air sampling bngv A 3/32-inch hole predrillcd through otto ol the caps provided an access for withdrawing samples. The Saran liners provided an ef­ fective gas barrier so that vnjror lo«cs were held to less than )0f< lor a boldine period of 3 days. When collecting a sample the subject was asked to remove the caps, place the pipet to his lips, and breathe normally in ihrough his nose and exhale through the piper three On-thc-Job Breath Sampling Three separate breath sampling programs were conducted concurrently with the envir­ onmental plant survey, designated by the boxed portions in Figure 3. Each worker collected three breath samples daily—the first on his arrival home from work, the second f> to 10 hours later, and a final sample be­ fore returning to work the following day The samples were collected in piputs constructed from short lengths of 20-mm soft glass tubing to which had been welded at each end the threaded portion ?f a 2-dram fF5-ml i screw- FtCURC 3. JVceklr mem Vtpor exposure concen­ trations measured durini: itie survey. Periods dmini' which breath sampling vs as conducted are represented bv the boxed-in area*. t AP00000222 540 Novembcr-Dcccmbi-r, 1909 G£.e .*>. 1: -___ '■■A . j Fici*t 4. Glass pipet (50 ml) used for collect* log breath samples. One cap lias a predrilled hole for gas sampling. Both raps have Snran liners which seal the pipet chamber. times. After expelling the fourth breath he quickly caps the tube, trapping a portion of alveolar air. The imjjortancc of writing the name, date, exact time of sampling, and the workshift most recently completed, on the label attached to each pipet. was stressed. Aliquots were drawn from the pipets with a 1-ml Hamilton gas-tight syringe and ana­ lyzed in an Aerograph A-600B gas chromato­ graph using X* carrier eas and a hydrogen flame detector. Separations were made with a 6-foot. Vi-inch 1.1). stainless-steel column packed with Carbowax 2QM alkaline on Chroinasorb W 60/80 mesh acid-washed. Exposure Chamber Operation Three experimental human exposures to VC1 were conducted at nominal vapor con* centrations of 50, 250. and 500 ppm. The exposure chamber was a room measuring 41 feet by 6 feet wide by 7.5 feet high. The room had a continuous positive air supply and exhaust system capable of maintaining a slight negative pressure within the cham­ ber. Continuous distribution of the cham­ ber air was achieved by iccirculaiing the air with a squirrel cage fan through a series of inlet and outlet duns spanning the length of the chamber. The YCl was metered into the duct carrying air exhausted by the squirrel cage fan and entered the room atmrwpltere via the recirculation system at a rate sufficient to maintain the desired atmospheric concen­ tration. The vapors were introduced from a pressurized storage cylinder through 6 feet of 54-inch hi), stainless-steel tubing into a rotameter prior to entering the circulating air duct. A heating tape wrapped around the stainless-steel tubing prevented condensation of the YCl and stabilized the flow of the vajror. The concentration of VC) in the chamber was constantly monitored with a Prrkin-EI- mcr infrared spectrophotometer equipped with a 10-mctcr path-length gas cell. A sam­ pling probe, consisting of 5/l6.mch I.D. Sa­ zan tubing, was centrally located during the exposure to represent the breathing zone of all subjects within the chamber. The proi/v was moved about prior to each exposure to detect imbalance of vapor concentration' within the chamber so that necessary cor­ rections in the recirculating system could be made. Air samples collected porinriicahv within the chamber throughout the c.Njxosurc day were analyzed by gas chromatogjaphv tor added assurance of analytical accuracy. Bo::: the infrared spectrophotometer and the e-is chromatograph were calibrated before each experiment and at intervals throughout :ivc exposure day. Each 7.5-hour exposure day included n 0.5-hour lunch period in an uncomaminavii area outside the exposure chamber. The TWA concentration was calculated on the basis of 7.5 hours of exposure. Clinical and jLa&orofory Procedures Each subject had been under careful medi­ cal surveillance by the medical department for a number of years, and each was given a complete medical examination a few davs prior to the VC1 exposures. Included v.v:c a complete urinalysis and 24-huur urine ;\c urobilinogen, complete blood count with '••(;• imentation rate, reticulocyte count. SGUT. SGPT. LDH. alkaline phosphatase, lll’X. creatinine, and bilirubin. Each subject received a tepeai pliysh.d examination l hour before entering the ex­ posure chamber. This examination inciud-m measurement of temperature, blood pres1'.::e. and pulse rate, a neurological examination, and collection of blood and breath samnies. A questionnaire noting the presence of any symptoms of illness (for example, hoadm h>!lection bag located outside tin* ihaniivr. Tidal volume and total expiratory rapacity AP000( .Ml American Industrial Hygiene Association Journal were measured in the morning and again late in die afternoon exjxvsuto periods, Subjective and neurological responses were measured before the subject entered the chamber. 15 minutes after entrance, and at 1-hour intervals thereafter. Flannagan Co­ ordination and Crawford Manual Dexterity Tests were conducted at imdmoining and again in the afternoon. Ktvath sampling be­ gan immediately after tlie subject left the exposure chamber. A 24-liour postuxposurc urine sample was collected and a blood sam­ ple was drawn the following morning for SGPT, LDH, alkaline phosphatase. BUN*, creatinine, and bilirubin determinations. Analysis of Breath Data The decay curves for the breath vinyl chlo­ ride concentrations were constructed bv step­ wise multiple regression using a digital com­ puter. An empirical relationship of the form Concentration ~ / (TWA, umf\ was select­ ed from a choice of several terms, each based on TWA and/or lime. The resulting regres­ sion equation best represents the ordered re­ lationship between breath vinyl chloride con­ centration, time-weighted average exposures, and postexposure time. Each breath decay curve has an associated. standard error of regression which can be used to compute the confidence band for any chosen level of significance. The 95 rJ- con­ fidence band for the mean of a group of ob­ servations was chosen in thh case to describe the statistical error associated with the breath data and the regression technique. Results Experinu ntal Breath Curves A total of 13 men participated in the throe experimental chamber exposures at nominal concentrations of 50, 250. and 500 ppm pro. ducing a total of 160 valid breath data points. Five of the six subjects exposed to 50 ppm were rc-exposcd at 500 ppm 2 days later. There was no measurable residual vinyl chloride detected on the breaths of the sub­ jects prior to the second exposure. Serial breath sampling was initiated immediately after the subjects left the exposure chamber and continued up to 20 hours following the exposures. Tabic I Experimental Human Exposure to Vinyl Chloride Chamber C»wcmujiict» (ppm s.u. X 59 2 261 8 493 T 4*1 i k»ngt Ippm] €*- 55 2815-243 MR-171 525-175 TWA* ippuu 40 248 459 ■Wl» XuiiiIh:i Ill SutljlXtv 6 4 4 i •T/oi<.wrie!ilrd average cenetniraliua based ee IS Im>«u» including a v.Miour lunch period in ju uucviiUntuuird area. ^Continuous exposure for S.5 hours. Table 1 shows the analyzed concentration to which the subjects were exposed. Calcu­ lations of the mean and standard deviation of exposure concentration arc based on chart readings from the infrared spectrophotometer taken at 5-minute intervals over the two 3.3hour exposure periods. The TWA is based on the total 7.5 liouvs which included a U.:>hour lunch period in an itn'contaimmuod area. The final breath decay curves intended for ttse as an index to YCI exposures teen? ad­ justed to TWA concentrations of 50. 250. Kicvkr. 5. Breath decay curves bated cm experi­ mental human exposures to 50, '250, and 5 512 DAILY VARIATION IN EXPOSURE TO VCL VAPOR < 1* TWA) VCi. 1C C'WiH'o*** »*v - . ^ - : ;i i' i *: " r Doyr 6f «M War'* a Fioure G. Variation in VC1 vapor exposure for three shifts. and 500 ppm (Fis:urc 5i. A IQO-ppm decay curve was interjjolatcd from the available data using regression analysis. These curves are presented with trie calculated 95<‘o con­ fidence bands for the mean of a group of ob­ servations. On-the-job Breath Curves Ten workmen participated in the on-thcjob breath sampling program, producing a total of 91 usable sets of data. Ten percent of the breath samples collected were discard­ ed because of pipet leakage or poor sampling techniques. Absolute breath levels ranged from about 20 ppm in one sample taken less than 1 hour after an 8-hour TWA of 250 ppm. to barely detectable levels : <0.05 ppm ' in samples taken after exposures at TW.-Vs below 50 with confidence, bands only slightly wider than those front controlled human experi­ ments. The close similarity between these curve* and those constructed from controlled exjtosure data am further illustrated in Fig­ ure 8. Human Responses From a subjective standpoint no significant untoward affects were noted at any of die exposure concentrations. The only complaints were those of two subjects who reported mild headache and some dryness of their eyes and nose during the 500-ppm exposure experi­ ments. Xo odor was detected by anyone entering the exposure chamber at 50 ppm. At 250 ppm all four subjects entering the chamber initially reported that they could detect a very slight odor of the chemical. Five of the seven subjects entering the expedite cham­ ber at 500 ppm were able to detect the odor of VCl. but after 5 minutes of exposure those five were unable to detect it even with forced inspiration. Three of the four subjects re­ entering the chamber after lunch were able PPm-c The cxtrcmclv broad variation in the TWA’s experienced by workmen during one of the periods in which breath sampling was being conducted is demonstrated for three shifts of men bearing the job classification “coagulator operator*’ (Figure fri. Minute, hourly, and daily fluctuations in the concen­ tration of a contaminant arc most descriptive­ ly revealed by continuous monitoring. 'This method of sampling quickly points out the fallacy of judging TWA and peak ex|>osme concentrations on the basis of spot sampling or periodic surveys of brief duration, The remarkable correlation between breath concentration and conysponditur TWA val­ ues made it possible to construct the series of breath decay curves shown in Figure 7. Figure •/. Breath decay curve* derived frc:i. breath data collected from workers fnllowinc onthc-job exposures lo VCl vapor lG-3nmr TWA'. AP00000225 jJttiriu'aM Industrial Hygiene Association Journal 543 lo doiect a faint odor of VCl. One subject could detect a faint odor on deep inspiration for approximately 15 minutes after entering the exposure chamber. The exposure had no noticeable effect on neurological responses, nor did it produce sig­ nificant changes in the results of mental, co­ ordination, or manual dexterity tests conduct­ ed during the exposure period. All clinical laboratory studies performed in the postexposure period were normal and not signifi­ cantly different from pre-exposure values. Discussion The object of the environmental health survey is lo identify the atmospheric contam­ inant, determine the exposure level, and re­ late this to the health hazard it pusems. If one is to judge hazard by ambient concen­ tration measurements, then those measure­ ments must accurately describe tin? exposure on a continuing individual basis. A carefully conducted survey combining continuous anal­ ysis of the work room atmosphere with a comprehensive job study will provide data valid for estimating time-weighted average ex|K»smc. On the other hand, breath decay curves constructed from bread) data collected dur­ ing continuous plant monitoring are in close agreement with those obtained from exposure chamber experiments with VCl. These curves should therefore be useful as a second meth­ od for assessing exposure lo VCl vapor. The choice of whether one or both meth­ ods should be used depends on prevailing circumstances and on the thoroughness de­ sired. For example, data useful in describ­ ing peak exposures are obtained from con­ tinuous monitoring. Concurrently, exposure trends and concentration gradients may help identify plant operational inefficiencies and equipment malfunctions by revealing specific sources of emission. Correcting these prob­ lems not only restores a healthful work en­ vironment but often results in bonus savings by reducing losses of raw material and prod­ uct. Continuous monitoring, however, is ex­ tremely costly both in time and in the equip­ ment required. The scope of data acquired is Ficunz 8. Comparison of breath decay curves derived from the on-ihe-job data ar.o the experi­ mental human exposure data, limited by the number of sampling probes, and these probes arc not always capable of accurately measuring the individual's daily exposure experiences, especially should these involve unusual incidences such as chemical spills or exposures outside the monitored area. Iireath analysis lias the advantage of in­ dividualizing: ertvlt worker’s intccvntcd daily cxjjosttrc. Breath decay curves, a< an index of exposure, offer a means of estimating the average daily individual exposure on the basis of a few breath samples taken serially in the postcxpostire period. Consequently breath analysts can bo used to diagnose as well as quantitate an exposure which has already occurred. It is a relatively inexpensive and simple method which can be put into opera­ tion without extensive and costly preliminary preparations. However, postexposure breath analysis does not provide information on the daily fluctua­ tions of exposure, and the peak exposure con­ centrations are not made evident by breath data. Finally, breath analysis is not applica- AP00000226 544 .Xtimnbtr’DiiCtnbir. 1969 blc to all chemicals. and breath decay curves References established for one chemical arc not useful as an index of exposure to any other chem­ ] Timt-ox. J. 11.. H. K. Hour, ami h. I. Sciikmuii Tlif Application of Computer Science to Indmtrinl Hi lic-nc, .1 mi. 1*4, H\t. Ae Vapor: KcLitihip n( Expired Air and Blo.ai Concentration^ to t'.xpnMire and Toncitv. .•krfi. Emir..*. Htilth 2; 5It>4J2 iMjv mill. ical. The decay curves presented here arc in­ tended to serve as an index of exposure to vinyl chloride vapor and are based on an exposure duration of 7.5 hours for the experi­ mental exposures, and 8 hours for the on-thejob study. The close agreement between the two sets of curves and the narrow confidence bands obtained in each case demonstrate the usefulness and accuracy of both methods for estimating TWA exposures and indicate the importance of breath anaivsis and the need for expanding its use in evaluating ex­ posures to other widely used volatile organic chemicals. r-'-'r. l 3. SmviRT, R. D., IE. It. Civv. D. S. Ekxv, C. 1,. Hmm. and J. it. 1‘trtnjoN: OS'-rr.atioru on the Cmi«enirati«ii of Triclilofpvtlivknc In Wood tnd Expired Air Itilkmini; L\|totur< ui 1 luntani. .inter. Itid. U\g. Auoe. 1.2,1 167-lTu (April 4. SnwntT. R. I).. and V. K. Rnxvr: Qiiinre An* d’Eiuiki >ur k l.l.l.Trkhlorotthane. Aten, .t/«Icdi<> Prefett. 29: IW-JW ria Annual Mcctine the American Confer­ ence of Govera&enlai Ieduiliul Hygienitts, ilauitan, Texts, May IMS, supervision to control exposures, full me the data to estimate the average conrrnu.ition to which men were exposed was jj. • feasible. To illustrate: The instrument r.. orded air concentrations at the rate of «■: even- six seconds: this is 432,000 times month. Because the first continuous air jnotiiu • was a success at day-to-day control, oti«-were installed in several plants during i ! 950’s. As each monitor was put into i:even more information on workmen's -v posurcs was -'going to waste'" simplv Ik-v.v there was too much of it. The advent • digital computers suggested a solution. ’I hue advantage of the fact that the Conttru.• lions Research Laboratory had a Biuhm.v 152 20 computer, the Environmental Rcmm;- • Laboratory in 1961 acquired a machine i« ■ translating air concentration data from recorder to punched paper tape. Contimi-air concentration data could then hr con bitted with computer analysis to yield a **>'•»• more complete description of inhakuii'n • ■ posurcs than had heretofore been yxvwihUThis paper recounts some of the prohk:: involved, from the installation of air mo­ toring equipment to the interpretation of t: data as summarized by the computer. A* t data are used, but only for illustrative poses, so plant, process, and even the nir <■; taminant arc immaterial. This is not a rci’"" jof an environmental survey; it is a ptutwi the application of computer science to it"- • trial hygiene. An am It- sensttn of signifu tango fro; mediately or no sig ;w-a ted ex; ■K'lcctivv; rial (si of -Itould bo major! va i ibration. must be c: ' relativelv i lion by tin an instrun. ;mce tesiiu operation. '•id of air i-nncentrati corded. At these speci have used '(rujncnts spcctronu-tablc for re Stnglc-pc plant is s: analyzer is the plant, i. dnuousiy tc Tubing use-, or son Preparing D There i>; fading vest point chart •“v.uix that n>o" the da •’dnimutn, t analog data '•n the slide ,;»l form (fo-. handled by ; Because d. •uinimizc pre includes a di 180 i AP00000228 .0 trican Industrial Hygiene Association Journal Obtaining Air Samples n automatic air monitoring system must te sensitive enough to detect concentrations tj significance to health. Sensitivity must t jc from concentrations that may be ini* r Jiatcly hazardous to those that have little or no significance even for prolonged, re­ plied exposures, The instrument should be j -Clive; it must respond only to the mateof interest. It must be stable; it should be unaffected by minor (or even • ijor) variations of temperature, humidity, •ration, and line \ultugc. The instrument must be capable of operating- unattended for relatively long periods of time. Daily atten>n by the operator, minor service weekly by instrument man. and occasional perform­ ance testing should be adequate to keep it in •eration. Finally, because a permanent rccd of air concentrations can be important, concentration data must be automatically re­ corded. Any continuous analyzer that satisfies • icsc specifications can be used. So far -we ave used only combustion-conductivity in­ struments and long-path gas cell infrared 'pectrometers; both types produce data suitble for computer analysis. Single-point monitoring in a production nlant is seldom economical. Usually the inalracr is positioned at a central location in lie plant, and air samples are brought con­ tinuously to it through probes of some sort.. Tubing used for these probes must not react with, or sorb, the air contaminant of interest. Preparing Date for Analysis There is no method presently available for reading results automatically from multiplepoint chart paper into a computer. This means that a device must be used to '‘digi­ tize" the data as they are obtained. At the minimum, the digitizer must translate the analog data (usually the position of a wiper on the slide wire of a potentiometer) to digi­ tal form (for example, numbers) that can be handled by ihe computer. Because data on the time of exposure can minimize programing difficulties, the digitizer includes a digital clock with an output of the 181 day number, hour, and minute. A switch selects a “tyjx* of data" digit (from 0 to 9) which is iiicorjiojatvd into the digitizer out­ put, An extra "type of data" (for example, front an infrared spectrometer or from a com­ bustion-conductivity analyzer i digit can allow the computer to reject data that obviously do not belong with that being processed. To minimize costs, punched paper tape is used to transfer data from the digitizer to the computer. For each datum the digitizer punches two "words” on the tape. The first “word” con­ tains the day number, hour, and minute, as well as the “type of data" identification num­ ber. The second "word" contains the probe number and the actual datum which, in this case, is a number from 0 to 999, propor­ tional to the concentration of the air con­ taminant. This information is obtained at a rate that may vary from two times a minute to once in 2.5 minutes. Because collecting each datum generated is not always neces­ sary, tiic digitizer can be programed to skip the collection of some data. In every case, once data from probe Xo. 1 are punched onto the tape, data from the other probes arc obtained in serial order, but the digitizer can ignore all but every second, fifth, or tenth sets of analyses. Programing th« Computer We decided that the minimum time in the data summary should be an cicht-hour shift. For this period the Dow Computations Re­ search Laboratory programed the computer to calculate the mean concentration at each location, the standard deviation of these data, the percentage of time that the concentration was above several preselected levels, and the appropriate time-weighted averages. Computers arc versatile. For instance, if one probe is located to sample air outside the building, the computer can be programed to rorrect automatically for deviations in the "background.” It can take into account peculiarities in the calibration curve of the instrument used for air analysts and can re­ cognize several kinds of errors in the data and use only “good” data for calculations. 1 I AP00000229 182 Alerch’Apnl, jfir.r, Computer Output The basic computer output consists of the mean concentration, standard deviation, num­ ber of analyses recorded, and percentage of time tile concentration was allow preselected levels, all at each location for each shift dur­ ing the survey. Shift or daily averages can also be obtained over any selected time inter­ val such as a week or a month. These data, further identified as to department, air con­ taminant, etc., are permanently stored on magnetic tape. The computer has been programed to re­ ject poor data such as that caused by a stick­ ing punch or by a faulty encoder. If poor data are being obtained, this fact is often signaled first by an unexplained decrease in (12.51 (26X3) + (3».3) (28.66) + (18.81 <93. too the number of analyses used by the computer to obtain the shift averages. This lias become the normal signal for nonroudne maintenance of the digitizing equipment. In addition io the basic output, the com­ puter calculates time-weighted averages of two kinds. The first kind is the “usual” timeweighted average concentration to which men are exposed. It is obtained by combining the air analysis data.with “job analysis” infor­ mation on the percentage of time spent by men in the vicinity of specific analyzer probes. Job analysis information is also combined with the data on the percentage of time dur­ ing which concentrations exceed the prese­ lected levels. This results in a second kind of average which is the “rfme-ueighted percen­ tage of time” exposures exceeded the prese­ lected levels. An example of how the time-weighted per­ centage of time is calculated is given helots'. Assume that the following information is true for an operator in a plant: Data Location Percentage of Time Spent Number at That Location 1 12.5 7 31.3 4 18.8 12 25.0 Unexposed 12.4 100.0 Furthermore, assume that, over the tinv period of interest, concentrations at tl»%. data locations behaved in this manner: .American Induct tin 1701-----160 i— i 1 rr of Tirnr Concentration Exceeds Selenrt: Values at Sjtccified Locations Concentration greater than 25 ppm 50 100 250 500 1 ISO — Data Location N'umlicr 4 7 12 1301— no — 26.63% 28.68% 93.44% 31.55' 16.39 22.13 20.08 23.77 6.55 10.65 17.62 11.47 1.63 2.04 16.39 3.68 0.00 0.00 0.00 O.on too — •90 — 80 — 70 - 60 - The time-weighted percentage of time sp. u: by operators in concentrations above 25 ppm will then be: 50 40 — ) + (25.0) (31.55) + (12.4) (0) ----------------- .---------------------- — 38.92 This means that on the average, during the rime prriod of interest, men in the opera­ tors classification encountered concentrationabove 25 ppm 38.920 of the time. Similar calculations show that those m*':. encountered concentrations above 50 pi*:. 28,24% of the time: above 100 ppm 10.9V) of the time: above 250 ppm 4.00% of :5 time; and above 500 ppm 0.00% of the tin.Data Analysis Automatic air monitoring reveals that it. an industrial situation the variation of <'< ■■■ contention with time can be large despite t: • use of a rather large time base. Figure 1 h -• plot of daily time-weighted average com.-*.-, trations to which operators were exposed each shift over one week. Each point on ti*’ graph is the mean of several hundred driesminations spaced equally over an eight-hum period. These are actual plant data. On the second shift there was little varia­ tion during the first four days, but over drlast three days of this period the shift avvj.vc varied by a factor of almost four. If ththreshold limit value (TLV1 for this mawrial were 50 ppm. an industrial hygionwt tak­ ing air samples during the first three day* m; the second shift would probably have ri< • clarcd that the hazard to health was h>" 4>l nonexistent. On the other hand, it he had TIME-WC. CONCE?. SO - 20 Li I t Fiovrf 30 to IV ppm. sampled on the In* second shift) he strong inclination Variation of c< can also be- strikh pnrison of the li second shifts duri Figure 2 places • text. Data for l'ig average concenu. were exposal. *11 diown in Figure 1 can be repeated f< l he information : based upon hunt! ! vidual air sample Average concet average concentr;. • 'ummarics provid : :‘gc conrcntratioi Mandard dwiatio 'itttwlartl tlrviado are two different yet know how tc cicmly. i AP00000230 ”1 | American Industrial Hygiene Association Journal 183 i0 lltt^ 4 ) i 4 t Fiqu&k I. DaSy time-weighted average exposure* for each shift from January $0 to February $. Time-weighted average for the week (all shifts) was 62.32 ppm. sampled on the last day of this period (on the second shift) lie could have experienced a strong inclination to “push the panic button." Variation of concentration between shifts can also be striking, as illustrated by a com­ parison of the third shift with the first or second shifts during the first three days. Figure 2- places Figure 1 in a broader con­ text. Data for Figure 2 are also time-weighted average concentrations to which operators acre exposed. The rather extreme variation shown in Figure 1 for_a time basis of one shift ran.be repeated for .‘Mime basts of one week. The information contained in Figure 2 is based upon hundreds of thousands of indi­ vidual air samples. Average concentrations and time-weighted average concentrations are not the only data sufnmaries provided by the computer; aver­ age concentrations are accompanied by the standard deviation of the data. Having the " standard deviation and using it quantitatively —.* -are two different things, Jwmvver. We do not yet know how to use such information effi­ ciently. On the other hand, the spread or varia­ bility of the data is indicated in a more mean­ ingful manner by the percentage of time the concentration exceeded certain levels. Figure 3 is a plot of the time-weighted percentage of time above these levels on log-probability paper. The interval during which data were gathered and the operational classification arc identical to those in Figure 2. This graph shows that the median concentration to which men on all shift* in this classification were exposed was 40 ppm and that they' were exposed to 500 ppm or higher Vjh of tlie time. The conventional plot (Figure 2) shows a maximum concentration' of about 135 ppm, but it is one of seven-day averages, whereas Figure 3 is a summation of instantaneous values. Both kinds of graph have advantages. A conventional plot illustrates better how ex­ posures vary with time, and any trends be­ come readily apparent. With this kind of graph, however, the only usable index of ex­ posure is the time-weighted average, a num­ ber of limited utility because it cannot reflect AP00000231