Document 1QVgwbw8n7RY9dbaM105oGzXj

FILE NAME: Union Carbide (UC) DATE: 1956 Sept DOC#: UC335 DOCUMENT DESCRIPTION: Journal Article - Symposium on Threshold Limits 'i. MELLON INSTITUTE LIBRARY SEP 2 7 1256 PITTSBURGH, PA. Symposium on Threshold Limits Present Trends in MAC's Introduction by WARREN A, COOK, Associate Professor Industriai Health and Hygiene, School of Public Health University of Michigan, Ann Arbor THE thinking today on the concept of threshold limits of injurious materials, designted by whatever term, has passed through an evolution to a gratifying- niaUrity. Initially the need to know how little of a substance might remain in an indus trial atmosphere without injury to those ex posed was insistently apparent. Such in formation began to be supplied through cor*'tation of concentrations of injurious sub stances with observed effect on the health of llie worker for such materials as granite and lead compounds.2 Results of short time animal experiments had been available ?*** before the present century from such tMMligators as K. B. Lehmann3 and later J0*# Yant and others at the U. S. Bureau f Mines,4 with the classic combined longPoaure animal experimentation and oc cupational study on benzene by Greenburg6 * the middle twenties. A* these threshold limit values became numerous and were more widely apit is perhaps not surprising that mis- J*Ptions and aberrant usages occurred. Persons erred in considering the levels A*?**precise than the facts justified; others y^JOcted them because the values could not Gtt" *ly fit all conditions of exposure and Wytiologica] response. And the detractors r* ^ f r a while more vociferous and seemed to be more sound than the pro- u lit8. With time, there has been a separation of _ S uable advantages of the threshold * t th e J o in t Session of th e A m ehican I n H ygiene A ssocia tio n -- A m erican Co nferen ce ^ t a l I n d u s t r ia l H y g ie n is t s a t t h e 1M>& I n C onference, B uffalo, A p ril 26, 1965. limits and the improper use of them. No where is the present concept of threshold limits more ably and competently expressed than in the papers and discussions of this Symposium. That the discussants might not have the very last word, it may be permissible for this prologue to assume in part the preroga tive of an epilogue! The following com mentaries appear pertinent. The last publication of the American Standards Association Z-37 Committee on Maximum Acceptable Concentrations of Toxic Dusts and Gasesfi is reported cor rectly as 1949 by Sachs who deplored the lack of revision of the 1941 Standard for benzene at the obsolete level of 100 parts per million. It is to be noted that this ASA Committee is now active in the revision and publication of MAC Standards for many of the materials on which information is available. The term "maximum acceptable concen trations" was adopted by the ASA Commit tee for a number of reasons. The word "ac ceptable" does not connote the legal control inherent in the words "allowable" and "lim it," but rather a simple acceptability with out further implication. Also, the expres sion, MAC, has been widely used over the years and is now a part of the language. In fact the Germans adopted a wording which gives the abbreviation, MAK, so that this term may be used in international conversa tion. Their words as used in the title of a paper by Oettei7 are "Maximale Arbeits platz-Konzentration." It is understood that this same expression is being used in the "MAK" tables of the German Association for Worker Protection. triai Hygiene Quarterly 297 geeks the maximum production of erials at minimum cost; and if our `tive system is to be retained and is to succeed, bureaus of industrial must assist in the attainment of by the simplest and most inexpenods possible. iwould contend that it is the duty -kyjrienist to tell the manufacturer ;run his business. Many industrial are so secretive and complicated oped over an extended period of at it is frequently impossible for an within a limited period of time to that production process to give all of the methods of control that 'enist would recommend. What is cooperation between state depart- manufacturers, not additional _r statute books that may be used judice of state departments as well facturers. is one other thought which is of consideration. Assuming that threshold limits of maximum allowable con centrations are established by law, the ad ministration of such codes may prove to be embarrassing to state departments charged with their administration and to industry at the instance of organized pressure groups of representatives of the public, labor unions, or possible competitors. Assuming the adoption of codes, it is a fact that oc casional violations may occur resulting from breakdown in machinery or equipment. Im mediately the manufacturer is placed in the position of having violated the law although such violation may be absolutely uninten tional and may be readily corrected. State Departments of Health and Labor do not need to be told by any organized group as to what they can or should do in the adminis tration of their functions. Therefore, a bet ter result will be obtained if use is made of the information developed by our scientists as to permissible concentrations not be cause some law requires this be done, but because it is simply good business to comply. cyclopedia of Instrumentation for Industrial Hygiene -STKIAL hygienists and workers in allied fields will be interested to know that heralded encyclopedia on industrial hygiene instrumentation is now generally ble. The book is divided into seven sections, each of which includes a compre ss review of instrumentation in the area being considered, technical papers on problems or types of instruments used, and descriptions of available instrumostly commercially available. The sections include the seven areas of ents: for measuring air contaminants in occupied spaces; for use in laborafor air pollution and meteorology; for air velocity and metering; for sound oration; for ionizing radiations; and for ultra violet, visible and infrared energy. Encyclopedia brings together in one place a fund of specific information on nents within the scope of the book. It was produced through the combined of personnel of the University of Michigan, manufacturers' research directors Cosigners, authorities in the several areas of instrumentation, and a staff of exI* Msigned by the U. S. Public Health Service to assemble and edit the contents w book. The comprehensive reviews include progress in the individual field of Fomentation, problems presented, limitations of present instrumentation, and needs .Bet by instruments now available. The descriptions of available instruments in the name of the instrument; name of the manufacturer; intended and special I *nd adaptations; operating principle; physical description; performance data, i as sampling rate, range, sensitivity; inteferences, limitations and safety hazards; operating instructions; calibration instructions; maintenance instructions; raphs, line drawings or wiring diagrams; bibliography. material in the book comprises some 1243 pages, 9 x 12 inches, with 1400 li ons, charts and diagrams. Copies may be obtained through the University of n Publication Distribution Service, Ann Arbor, Michigan. The book is priced 274 Septemh k, However, there can be no really valid ob known cancerigenicity of arsenic, cL jection to the xerm "Threshold Limits'' used and asbestos. Says Oettel 'In this by the committee of the American Confer ence of Governmental Industrial Hygienists or to "Hygienic Guides," the term adopted by the American Industrial Hygiene Asso ciation committee. This Symposium would not be complete without reference to the great danger. The lists should in remain logical." Waters' discussion of Brandt's cupies the final pages of this publi the Symposium, but it is belietn Waters would not wish his commen availability of the excellent AIHA Hygienic considered the last word on the sul Guides to which reference is made else threshold limits--rather that his < where in this Journal. tions are limited essentially to the One of the most perplexing problems in these values in legal codes, rigidly ] arriving at threshold limits is posed by the statute. As the last word in this Syni cancerigens. Stokinger's statement in this this introductory refers the reader! Symposium is the first publication on this phase of the subject. However, the use of the procedure suggested by Stokinger as ap concluding paragraphs of the pa Brandt for his discussion of the p h | of threshold limits. plied to nickel carbonyl has already drawn some fire. Quoting in part from a private communication from Oettel whose publica tions' s on "MAK's" include a well-consid ered presentation of the general subject: "We suggest that consideration be given to References 1. R u s s e l l , A . E ., Br it t e n , R . H .p T h o m f i Bloomfield, J . J . : T rad es II. E xposure P ublic H ealth B ull, T he H ealth of to Siliceous D ust N o. 187. 1929, W orker* (G ranite s 2. R u sse l l , A. E ., J o n e s. R. R.. Bloom fi Br it t e n , r , h .. T h o m pso n , L. R .; L ead P o l indicating the suspected cancerigens with an asterisk with reference to a footnote that the hazard exists and consequently especial ly effective control measures be instituted. No man can say today which concentration of the several industrial substances is actu S to ra g e B a tte ry P la n t. P ublic H ea lth Bull. S o . 3. L e h m a n n , K. B .: N u m e ro u s p a p e rs mi * 4 lis h e d in A r c h iv f r H y g ie n e f ro m th e 1880's. 4. Ya n t, W. P., an d colleagues: A cute G uinea Pigs to V apors of Some New Commi g an ic C om pounds. A series of p ap ers published H e a l th R e p o r ts d u r i n g th e L930's. < 5. Green burg, L. : Benzol P oisoning a* ^ ally required to be cancerigenic . . ." This statement follows his objection to the pres ent listing by the ACGIH of nickel car tria l H a z a rd . P ublic H ea lth R ep te., 41, 1367, . 1519; 1926. ; 6. A m e ric a n S ta n d a rd s A ssociation : Standard ceptable C oncentrations of Toxic D usts and G bonyl as 0.001 ppm on the basis of its alIeged cancerigenicitv. which incidentally Oettel's extensive experience with the sub stance causes him to doubt, while failing to include any acknowledgement of the 70 E. 4 5 th S tre e t. N e w Y o rk 17, N ew Y o rk . 7. O ettel, H . : D ie m ax im ale A rb eitsp latz-] t io n (=: M A K - W e r te ) s c h d lic h e r G a se . Di S ta u b e. Die B e ru fR g en o esen sch a ft, N o. 2, Fehl 8. O e t t e l , H . : G e w e rb lic h e V e r g if t u n g e n di D m p fe und S tau b arten . A rch iv f r Toxikolo. 1954. The Need for Threshold Limits MIRIAM SACHS, M.D. New Jersey State Department of Health, Bureau of Adult and Occupational Heaf Trenton, New Jersey . Probably one of the most basic needs in human endeavor is for man to have an orderliness set to his manner of thinking. Everyone has a need for discipline, for boundaries to be drawn, and for limits to be defined. It does not matter that the only use for some of the limits so defined may be to ignore them; they have still served their purpose as a point of aim for indi interpretation. If this seems cryptic, I ample concerning a common substan be cited. Almost everyone is familiar wit! threshold limit values for silica--foi with a free silicon dioxide content 50%, the limit is given as five million trial Hygiene Quarterly cubic foot of air i MPPCF), for dust f'erfL e.silica content below 5%, the lim- MPPFF, *-lUt MPPCF is the value ! 1 (i for dusts with 5% to 50% free **`*!>rrl`?jt s rather difficult to understand "!u!l| ngs could be so selective as to react ''* fibrosis if air containing more than 5 P('F of a 50% silica bearing dust was inbut vvould be expected to remain M i hv if a 40% silica content dust at the ^ I of 0 MPPCF was similarly inhaled. It ^""id be of doubtful accuracy for each in*"U ja| hygienist to interpolate on the : * j of this wide range and apparent dis " ncv. It has always seemed better prac- abandon the middle limits and when 'Pessary forward a report (concerning with a free-silica content of about *"' \ tbat "dust with such a dangerously *' b content of free silica should be main ned in the w'orking environment at less ian 5 MPPCF." For agencies who have not inducted any research or original investiJation along these lines, this might have n a precarious position under attack, fortunately, the trend is toward revision of he limit for silica-bearing dusts.1 An bsen+iai Tool Actually, threshold limits satisfy the need for a tool essential to the protection of the health of the worker. The quantity of tech nical information concerning the biological actions of old and new chemicals is so vast that no single practitioner of industrial hyfiene can be expected to be able to asimilate more than a very small fraction of this mass of material. A list of guide values prevents the chaos that would result if each investigator conducting plant surveys had n depend on his own comparatively retricted experiences and background. Two f industrial hygiene's sister specialities are admirable demonstrations of the benefits f having, and the disadvantages of not raving, an approximate measurement of hazard. In radiological health and radioogical safety the maximum permissible ex posure value, blessed by national and inter national committees, has made possible much of the rapid adoption of var."us radiological techniques, uses of radio-otopes, and forays into the field of nuclear energy. In air pollution investigation and control, 275 lack of standards for permissible levels of atmospheric contaminants has seriously hampered wise administration of air pollu tion control practices, and has made it al most impossible to weigh the effects of these contaminants on the public health. The need for threshold limit values and the manner of use is quite different for the staffs of governmental or official agencies and for persons employed by private indus try or working with research laboratories, universities or foundations. Official indus trial hygiene agencies must have a screen ing device to apply to literally thousands of industrial processes which are encountered in the day-to-day field work, from one year to the next. A screening device may very well be a single parameter but it should be used only for the purpose of suggesting more intensive and detailed study where the parameter is significantly exceeded. In multiphasic screening for chronic disease, the patient whose survey film is suspect for tuberculosis, cancer of the lung, or ab normal heart outline is referred to his private physician or to an appropriate clinic for more definitive diagnosis. In discussing any screening test proce dure, the question of false negatives versus false positives is certain to be reached. When the risk is permanent disability or death, the governmental industrial hygien ist is so conditioned that he favors the side of the false positive; he tends to be too strict rather than too lenient. For private industrial purposes, the threshold limit values satisfy the need for engineering bench marks. "Almost 100% confidence can be placed in the threshold limits where they are being used for evalu ating and controlling of an industrial en vironment. This is particularly true in the use of these limits as a base line for calcu lating control design specifications in that the usual engineering factors of safety are included in such designs. It should be re membered that under this connotation the base line for comparison need not be ex tremely accurate. Naturally, the more ac curate the figure is, the better economy that can be realized in engineering a control."2 One of the deans of industrial hygiene and toxicology, restated this point of view as, "We need a yardstick even if it's only a rub ber yardstick!" 276 Sepie Henry Smyth has given the most nearly complete description of the toxicological data required to satisfy all needs.3 (1) What uniform concentration is tolerable eight hours a day for a working lifetime? (2) What correction in the average must be made for brief peak concentrations? (3) What single brief exposure to a high concen tration is tolerable each day when there is no exposure the rest of the day? <4j What biological test upon the workman can meas ure his actual intake of the chemical at his job? (5) What are the earliest symptoms and objective signs of excessive exposure and how severe can they become before re moval from exposure fails to prevent perma nent injury? (6) What is the best treat ment for the effects of excessive single ex posure or excessive repeated exposure?" For a few industrial materials, all of these ques tions can be answered. The threshold limit values attempt to supply the answer only for question one. The introductory preface to the list advises caution in interpreting even this single answer. The introduction to the 1954 list has probably been well read. The preface to the 1948 list warrants repe tition: "While it will doubtless be very advantageous to have what might be called permanent or standard maximum allowable concentration values, it must be borne in mind that all our values at the present time With so clear a statement of the books, it has always been baffling that several persons speak \ terness and disdain about the thr its list. In all fairness, it mustf mitted that the acrimony arises the advice in the foregoing state frequently been disregarded and son latory agencies have not only alio have assisted the entrance of the limits into the sacred and rigid pre the law. This is an unfortunate restijj for a scientific instrument that requjj stant adjustment and calibration. A number of excellent papers ha written on maximum allowable co tions and threshold limits. In the of progress this panel discussion ah more than a review of bibliog sfl^hld be a reflection of the opinion ACGIH-AIHA joint membership ad expression of their perspective if threshold limits, their needs and th gestions for changes or additions, ingly, a simple one-page questionn the value of threshold limits was sen members of the American Industr giene Association and the Americ ference of Governmental Industr gienists (excluding those in foreig tries). A copy of the questionnaire on the next page. are fluid and subject to annual revision. They should not be adopted as fixed or legal Response to Inquiry values, but merely as guides to assist us in Q N E THOUSAND AND FIFTY q u e stic defining more or less safe working condi were sent out, of which 340 tions. . . It must be borne in mind that these turned by the February 10 deadline^ values are not indices of toxicity and are teen of these were discarded since th< not intended to approach that value. Ac simply courtesy returns with the qui cordingly, the comparative toxicity of these unanswered. compounds cannot be established on the Of the 326 completed forms, the e basis of their numerical maximum allow sions of "confidence" in the Threshold able concentration value. its were divided as: "People vary greatly in their response to Confidence Number of Replii n drugs and toxic substances. Therefore, it is 100% 53 ' a figment of the imagination to think that 80% 173 we can set down a precise limit below which 50% 37 : there is complete safety and immediately Less than 50% 9 : i above which there may be a high percentage Qualified 54 : 1 of cases of poisoning among those exposed. "With these facts in mind the Committee Total 326 ; has set values below which it is fair to ex pect reasonable protection and above which it is reasonable to expect that we can have occasional cases of poisoning,"4 XilC 5lUU|I u i CrU- ciiiowvoj rv*i sentially those who wrote they could noj ply a numerical rating to their reactiol the threshold limit values or who said' fg*t.inai Hygiene Quarterly ' gY State Department or Health '**' J 0F adult and Occupational Health V**ALL i7 West State Street Trenton 8, New J ersey The Need for Threshold Limits u oVv would you r a t e yo u r confidence in ACGIH Threshold Limits? check one) ___ 8 0 % ___ 5 0 % ____ LeS5 t h a n 5 0 % ---- Do you th in k we need a list of T h r e s h old Limits? Y es-- . N o_ What standards of performance would you prefer? When you use or in struct as to the use of ` T h r e s h o ld Limits, w hat qualifications or reservations (if any) do you advise? Other Comments , May I quote you by n a m e ? Yes__ No____ Nam e ________________________________ Title_____________________________ Affiliation____________________________ 1d. for example, 100% confidence in some the values and little or no confidence in her values, particularly values for some wer substances assigned because of emical or structural similarity to a ma rial with a well-documented value. At this int it should be emphasized that many, .my responses were careful to explain that eir confidence rating was for the "value" ; at their confidence in the Committee on reshold Limits was unqualified. The question, "Do you think we need .reshold Limits?" was answered Yes 323 No 3 Total 326 In the qualified answers, in the comments d in the suggestions for other standards performance, there were many pleas for /rouping of data and for additional in- rmation. These were expressed in many Ferent ways but lent themselves to a few ijor headings. There were 103 requests it the list indicate whether the value was injury" value or a "nuisance" or "com- :t" value; there were 73 requests for a 277 notation as to whether the value was based on animal experimentation or human experi ence; 75 persons desired a short exposure value, a one-hour tolerance or a value for "peaks" ; 32 wanted information on multiple exposures, synergism or additive effects; and 28 thought the references should be printed along with the assigned value. These suggestions may be summarized as follows : Requests for Number Comfort threshold (nuisance level) Chronic toxicity threshold 61 >/ 103 (injury level) 42 ^ Animal experimentation or human data 73 Short exposure threshold, one hour, peaks 75 Synergism, multiple or additive exposures 32 References on list 28 Several of these comments and requests were embodied in enclosed reprints, long letters and telephone calls. With all this ac tive and militant thinking, there should be no great fear that threshold limits will be applied blindly or that a numerical value will ever entirely replace thoughtful, scien tific appraisal of the many factors that must be considered concerning the health of th worker and the status of the workroom en vironment. Especially, the need for medi cal examination and a determination of the general physical condition of the worker seem well understood. What to be Expected of A C G IH List CANE pertinent question remains to be an swered. W'hat can logically be expected of a list, revised annually, compiled by a committee, each member of which is earning his living at a full-time job? One of the major values of the list is that it is revised annually, and new substances added as data are forthcoming. The excellent Z37 series prepared by the American Standards Asso ciation contains much more information than appears on the Threshold Limits list. However, of the toxic dusts and gases group there have appeared 15 leaflets, the most recent of which was entitled, "The Allow able Concentration of Methyl Chloride" which was published January 5, 1949. "The Allowable Concentration of Benzene" was 278 Sepie approved and published January 15, 1941 and states the permissible concentration as 100 parts per million parts of air by volume. So far as is generally known no revisions of this ASA figure have appeared although it is commonly agreed that the 1954 ACGIH value of 35 parts per million is a far safer working figure and one that is practical. If it would not unduly burden the Com mittee on Threshold Limits it would be ad vantageous to have a regrouping of the listed values according to comfort thresholds and chronic toxicity thresholds. And where they exist and are easily verified, it would be most helpful to have a special group of one-hour tolerances or peak short-time per missible exposures. The combined mem bership of today's audience might give con sideration to the publication of a monograph on threshold limits which could very well be a compilation of fundamentally sound material which has already been published. The monograph could include the papers al ready referred to in this article, with per haps several others that are equally well known.'v6'7's No list of values and no amount of re grouping and no increase in the number of background references can relieve the indus trial hygienist of the responsibility of ac quiring sufficient education and training to do his job properly and to think for himself. It seems desirable to add here another refer ence, "Education in Radiation Protection, 9 the Janeway lecture delivered by Lauriston S. Taylor. The words spoken by Mr. Taylor while directed particularly to the field of radiation exposure apply just as well to any industrial chemical exposure. He states: "The young radiologist should be educated in the background and significance of per missible dosage or exposures and should at all times retain a sense of proportion and a sense of humor relative thereto. For ex ample, the present permissible exposure for the whole body is 300 milliroentgens per week (measured in a ir). This figure was de termined from a limited amount of clinical and biological data to which was applied a series of well educated guesses, and yet we may find a person's weekly exposure re corded as, say 163 milliroentgens--a figure that few, if any, practical clinical instru ments can measure with an accuracy of bet ter than 10%. Again, if a film badge (accuracy 20%) shows a readi milliroentgens for one week, are alarmed? And are we to be compl a duplicate reading of 200 millir taken in the same place for the sa Actually both readings are the sam experimental limits of error and w be alarmed in both cases not prima cause the reading differ but because posure in each situation may be * sarily high. It is only rarely that permissible dosage cannot be obtain but little inconvenience or added ei "The aim should be to see how low. tical operating level can be achieu how high a level without transgresj 300 mr per week medicolegal uppe set by radiation safety experts. E' smallest amount of radiation has a feet on the living human body. Evei harm is normally undetectable byii dividual, genetic damage can rest! single minute exposures. We must perceptible harm or damage--we minimize the likelihood of undel damage even though we cannot elim entirely." The fundamental goal should be provement of industrial hygiene p Threshold Limits, however accura however well-documented, serve onl guide and a yardstick. No method erj been, or ever can be, devised that wilfl an exact advance prediction of bund ard. The closest we can come is l judgment. 1 References ! !. STOKINGER, H - E . : S t a n d a r d s f o r S a fe g u a i H eaith of th e In d u stria l W orker, Public H ealth 70 :1 , 1955. 2. C h u r c h , F . W . ( E sso R e se a rc h a n d E n C om pany, L inden, N .J J P erso n al C om m unicatJ 3. S myTH, H . F .: T o x ic o lo g ic al D a ta -- S o u re form ation and F u tu re N eeds. A lH A Q uarteri% 1954. 4. A m erican C onference of G overnm ental H y g ien ists, 1948. 5. Cook , W arren A .: M ax im u m A llow able tra tio n s of In d u stria l A tm ospheric C ontam ili dust-rial M edicine, 14:936-946, N o v em b er, 1945. 6. S il so n , J o h n E .: T h e S ig n ific a n c e o f M ai Jowable C o n cen tratio n s. M o n th ly R eview , 28:5, 1949. . 7. St o n e , R obert S . : T h e C o n c e p t o f a Maxi) m issible E x p o su re. R adiology, 58:639, 1952. 8. S c h r e n k . H . H .: I n te r p re ta tio n s o f ? L im i ts , A i H A QUARTERLY, 8 :5 5 , 1947. 9. T aylor, L a uriston S .: E d u c a tio n in P ro tectio n . A m er. J . o f R oentgenology, 73:193 Industrial Hygiene Quarterly 279 Prepared Discussion HENRY F. SMYTH, JR. Mellon Institute Pittsburgh R SACHS has established that the vast M majority of our colleagues recognize jr need for a list of threshold limits. She itlnguishes between the need of the official -encies for a screening device to apply to -k,,usands of industrial processes they enjnter in field work, and the need of the -ivate industrial hygienist for an engineer , bench mark for planning and evaluating vironmental control. To satisfy each of eie needs vapor concentrations must be tasured under operating conditions before e threshold limits values can be useful. There is a third group which uses thresh ,) limits and needs them because nothing iter suited to the need is available. Every e of our profession is a part-time member this group and some are full-time memrs. Reference is made to the arm-chair in strial hygienist. He may never measure a ncentr: tion and he may never observe an ijustria. operation. He uses threshold lim - to advise upon the selection of a chemical r a particular application, to select a suit,le application for a little known chemical d to bolster sales arguments for a chemiI or mixture of chemicals. When the industrial hygienist is asked to vise whether acetone, propyl acetate, or -thylene chloride would be the safest solnt in a particular application, it may be lieved that his answer is based on a proind knowledge of toxicology, but actually recalls the threshold limits, or refreshes - memory by consulting the list while the -stioner waits on the telephone. Then the -wer is given as if the list recorded eomrative toxicities, perhaps adding a bit of alitative information on fire hazard and e of injury from an excess. What better data are available for the -n-chair industrial hygienist than the an al list of threshold limits? He must have ne guidance if he is to do more than toss a n. If he is given a list based on uniformly ducted animal experiment, he is deprived the benefit of the years of carefully obved and. evaluated human experience which is recorded in some of the threshold limit values. Whether we want him to or not, he ( and that includes every one of us at times) will use threshold limits as estimates of comparative toxicity to meet his urgent requirement for finding some basis for opin ion. To what extent is he wrong? A threshold limit based on comfort, irritation or good engineering practices certainly is not com parable to one based on pathological effect. But the former limit is certainly lower than one based on injury. The error is on the side of safety if it is stated that material A has a threshold limit based on comfort higher than that of material B which is based on injury, therefore material A should be used in the process. No matter which material is chosen it will be employed by persons subjected to the same degree of industrial hygiene and medi cal scrutiny. By choosing the material with the higher threshold limit the probability of safe operation has been increased, even if the basis on which the choice was made is not numerically sound. There is one important short-coming in this argument which points out an im portant way in which threshold limits do not meet the needs of the arm-chair industrial hygienist. If the volatility of a material with a high threshold limit is greater than that of a material with a low limit, then in a particular process the less toxic material may be more hazardous, because a greater vapor concentration will be in the atmos phere. I maintain that the list of threshold lim its is our best available source for judg ments on comparative toxicity of vapors to humans. It can be improved for this pur pose, but there is nothing better for us to use. However, it refers to toxicity, not to hazard. It is the single-phased toxicity of K. B. Lehmann. A table of comparative hazards would go back to Lehmann's two-phased toxicity by combining vapor pressure (or evaporation ratej with toxicity. This table of comparative hazards is what the arm chair industrial hygienist really needs. It would go a long way toward satisfying all of his needs, particularly if it included in formation on nature of injury, on the penal ty for exceeding the limit and on the degree of hygienic urgency for observing the limit. 280 Septem Methods of Establishing Threshold Limits JAMES H. STERNER, M.D. Medical Director, Eastman Kodak Company Rochester, New York T hreshold limits are based upon in formation derived from many and di verse sources. For each evaluation, data developed by many different methods may need to be considered and weighed; and from this sometimes complicated, often con tradictory, and rarely adequate, complex of information, a significant value must be developed. It is customary to append, as reference material, the important published studies from which the conclusions were made. In these studies, particularly if the author felt that his contribution permitted such a judg ment, there is often a proposed threshold limit value. Not infrequently, of course, the values suggested by different investigators may disagree. The group which is charged with the responsibility for establishing threshold limits must consider, in addition to the character of the study, such qualities as accuracy, reliability, completeness, and purpose--and, a not insignificant factor, the reputation of the investigator. The subject, ``Methods of Establishing Threshold Limits," has a connotation be yond that of the procedures and techniques reported in the industrial hygiene literature. This has to do with the actual mechanism by which a group or a committee is designated to act as a body for establishing threshold limits, and the principles and practices which actually govern the operation of the committee. In the present evolutionary stage of industrial hygiene, the internal activities of this committee must weigh heavily in any consideration of "methods." Threshold Criteria TT would be worthwhile, if it were possible, to record all of the elements of a delibera tion in arriving at a standard. Often, there is considerable material, undocumented, which plays an important role in the evalua tion, in addition to the published studies. This may be limited and fragmentary data from industrial or governmental ini hygiene surveys; or it may be ini data from animal experiments; or r< cases of alleged intoxication from men's Compensation sources. A still less tangible factor, related experience, training, and critical ju, of the individuals performing the tion, is the ability to make a variet; trapolations. In one instance this volve the estimation of the probabli in man from data developed in one species of lower animals. The expe; toxicology data may be limited to subacute experiments; but even if cl long-term studies have been done, tin lation from a few years exposuri though a lifetime for the animal, to span of a working lifetime in man is cult step. In other cases, the extra] may be from experimental or clini developed with one chemical to the pi effect of an homologous compound or; terial with similar chemical structuri The behavior of a substance in othe: --as a therapeutic agent, an insectii even as a beverage (as in the case o; alcohol) may contribute important in: tion. The background of the indi making the judgment, with respect sonal experience ranging from animi perimentation through long-term c. observation of exposed workmen, witl spect to a practical and critical apprecii of the value and limitations of me( for making environmental measure; determines the ultimate value with all of these factors, concrete and absi are blended to form a valid, practical acceptable threshold limit. The more stantial the documented information, broader the sources of pertinent data, less the demand for these intangible f; to fill the gaps which are inherent in kind of procedure. The absolute test of a threshold limit lidiistnal Hygiene Quarterly 281 beini and probably will not be achieved ',ot. g measure of validity is strictly con- ^ ed in terms of a completely "safe and '^thfnl" environment for the occupational ^tim e of an individual. Practically, cri` a much less complete are accepted, altX -h the trend is to the constant improve- 1,0 t 0f our methods for evaluating long- rui and more subtle effects. These more re f te parameters of injury are seen in the ^reme in the study of radiation effects, lhere consideration is given to such factors \ shortening of the total life span, and of rnetic effects involving future generations. t,th the great majority of physical and nemical agents, we must be content with freehold limits predicated upon less ex- ..'nsive and less subtle end-points. If threshold limits, even with their pres et imperfections, are accepted as useful ntj desirable, they must continue to be abricated from information which is somemes inaccurate, frequently controversial, :!d always incomplete. In this discussion, , attempt will be made to define a pattern <acceptability for the various elements of idenee which may be considered in estab-hing threshold limits (this might be .iraphrased as "threshold criteria" for ireshold limits). This will vary with the arpose for which the limits are intended, -,echaracter of the group making the judg- -ent, and the need for such standards. At -is point, it might be suggested that the jrious bodies which are responsible for es- .blishing standards, attempt to define in -neral the purposes, criteria for accepta bly, and limitations of their function-- id perhaps, specifically Indicate the basis r their judgment in individual instances iere important factors other than the ap- uded reports played a significant part, itf publication of this information wouid t. of course, still ajl criticism, but it would viate much of the criticism which is .sed upon unfamiliarity with the manner which a decision is made. It is paradoxii. that the greater the need for a thresh- 1 limit (in terms of the numbers of inciduals actually being exposed, and the verity of exposure) the greater the justi- ition for accepting a tentative standard inadequate and incomplete information, is practice is defensible, of course, only this initial guiding limit is continuously and critically tested by a competent clinical study of the exposed people. Sources of Information ^J"HE TWO general sources from which sig nificant information is developed are the experimental laboratory where the exposure is deliberate, and the actual plant operation, where the exposure is incidental (sometimes accidental). The establishing of threshold limits depends increasingly upon a balance of information developed from both these areas. The more complete the laboratory in vestigation, the greater the security of the tentative standard for plant exposures. The inherent uncertainty of extrapolation from the experimental data necessarily places the final judgment upon the clinical evaluation of the exposed workmen. Industrial hygiene laboratory methods run a gamut from the simple, preliminary "screening" procedure using a few small animals to a relatively involved, carefully controlled, clinical experiment in which human subjects are deliberately exposed to a toxic agent. The studies with lower species should define a range from minimal or no effect through severe injury and lethality, and should indicate the various physiologi cal and pathological mechanisms of injury. The exposure levels for the human subjects usually attempt to define levels associated with "discomfort," "minimal," and "earliest reversible" effects. There is no formula at present by which it is possible to estimate the pattern or amount of laboratory experimentation which will be required of this component in estab lishing a threshold limit. In general, the more novel the physico-chemical properties of the agent, the more distantly related to other materials which have had industrial hygiene evaluation, the greater the amount and variety of toxicological procedures which must be employed. The direction and extent of further studies must be deter mined as the pattern of toxicological investi gation unfolds. The importance of the vari ous routes of absorption, the relationship of divided dose administration to the single ef fective dose, the relalive primary irritation and sensitization potencies, the behavior in relation to such factors as species of test animal, age, sex, concurrent disease, are only a few of the important items which 282 must be considered in deciding the importance of the role which a particular animal study may play. Seme of the experimental toxicological findings which suggest caution in evaluating a study may be noted. A lethal dose curve whose slope is gradual may overlap the curve of the physiological function upon which the threshold is to be based. A scattered configuration of delayed deaths suggests multiple effects or secondary pathology which may be difficult to evaluate. A substance which is a sensitizer or allergen, even though indicated solely by skin sensitization tests, may produce systemic or specific internal organ sensitization. Marked variance with respect to severe injury or lethality among several species of test animals, increases the difficulty of extrapolation. The failure to reproduce the disease pattern already identified in human subjects definiteiy limits the significance of animal studies. Less clearly identified effects, such as involvement of the central nervous system, or injury which is not easily reversible, such as aplastic anemia, signal caution. Experimental methods with deliberate exposure of human subjects to low levels of a toxic agent are finding increasing usefulness in establishing threshold limits. Initially, many of these tests were concerned with irritation or discomfort levels, and although no injurious effects were noted in animals exposed to higher concentrations, the human experiment values were frequently the determining factor in setting a threshold limit. A possible fallacy in this reasoning lies in the quite common experience of finding conditions in plant operations which are quite irritating or uncomfortable to a person first entering the environment but which are tolerated without complaint--or even, in some cases, with a positive expression of benefit--by the acclimated workmen. In such an instance it is difficult to decide which result is significant, the controlled short-term study using, usually, subjects whose experience with industrial conditions may be very limited, or the testimonial evidence of workmen, especially if unsupported by evidence which confirms the absence of injurious effects. Other applications of the controlled human-subject experiments may develop more important information. The studies of re- Septem tention of a toxin (with the lonf vestigations of lead and fluorine examples) must play an importai the establishment of threshold ] these substances. The obvious ad\ this technique lies in the accui which the relation can be estab contrasted with the more difflci trolled experience in the plant. Moi studies on a long-term basis are n< the cost in effort and in dollars im tinct limitations, This type of experimental appr value in determining the earliest reversible) changes in certain p functions, such as vascular inst measured by blood pressure ch metabolite excretion, as in the uri partition with benzol absorption jective, of course, is to recognize ble, functional change which, if i may lead to permanent injury. A difficulty may develop as the acui procedures increases, since the ch; quently are not specific for the may occur with many other fact affect the body--as for example, ar alcohol. On the other hand, specii tion of statistical methods to j posures may make these techniqu the most acute methods of signalii Value of Data on Worker Exposures 'J 'H E CONCEPT that a careful am hensive study of the exposed w the most significant factor in esta threshold limit, merits repetition tice, however, the number of subs which such complete studies hav< ported are few. A number of facto against such long-term, comprehi vestigations. The cost of an adeqi cal program, carried on over mi and with continuing negative ri quires the support of an unusual gent and understanding manager development of environmental ments so as to be effective for c with the clinical findings requiri degree of cooperation and plannin clinical and industrial hygiene And, finally, the job of organizin tensive data, of deciding that the i significant, particularly if they ari in the sense that no injury is fi strial Hygie-ne Quarterly 283 ,,ring for publication (since there is of P'TrTia] of Negative Data), demands of * J ,estigator courage to the point of be*Jfoolhardy- ^ if the various types of information which ^ be obtained from actual industrial exre3i the unsupported testimony of work- and supervisors, even though accom- *KRed by accurate measurements of en^ omental factors, is generally so un- *Vble as to merit little weight in establishr*1a threshold limit. Individuals who are 'ruble to tolerate the work conditions, or who have actually become ill and left -he job, ma- have keen eliminated so graduv that recognition of the cause and effect i^Utionship may not have developed in the remaining personnel. Furthermore, human Mture is such that under these circumunces the men remaining on the job are ,pt to dismiss the others who left as being weak to take it. As in the case of planning a program of ,*perimental laboratory methods for evalu ating a toxic agent, the in-plant clinical sur vey and environmental analysis must develop n relation to the specific hazard. To the (^jic elements in the medical examination may be added a variety of special test proce dures selected to detect the earliest changes m physiologic function. If little is known ,,bout the kind of toxic reaction which may develop, a "shot-gun" approach may be justified with the hope that one or more of the battery of tests will signal a harmful ef fect. As knowledge of the earliest reactions to specific agents increases, the selection of the most sensitive test procedures becomes more practical, and more reassuring. The recognition of an occupational dis ease is frequently much simpler than the proof that a particular exposure is free from any injurious effect. As exposures to :nxic materials are decreased from levels hieh can injure in a relatively short time, i point is reached where the signs and symp toms may develop only after a very long ex posure time, and the disease so mild as to hallenge the best diagnostic program. We an note again the studies with radiation, here minimal shortening of the life span *ith relatively low exposures can be demon strated in experimental animals but would ir impossible to detect, with our present 'rthniques, in the human subject. Clinical Observations 'J 'he broader the base of the clinical in vestigation which is associated with a finding of no injury or no significant injury at a particular exposure level, the greater the reliability of the conclusion, To the critical investigator, however, the job is never complete, never without some area in the study which could not have been strengthened or improved. The period of the study can extend into many years, and yet the end point can remain uncertain. As other techniques are added to the evaluation program--such as studies of morbidity and absenteeism, reasons for dis pensary visits, and analysis of cause of death--the problem of interpretation be comes increasingly complicated. An example may be cited in which two young women em ployees in the same small department developed leukemia within a few months of each other. This occurrence was readily accepted as a chance finding because both girls had clerical positions with no possible exposure to an industrial toxin. Had this oc curred following a common exposure, how ever brief or minor, to a new chemical with a long and unfamiliar name, it is quite like ly that medical testimony would have been developed, in a Workmen's Compensation hearing, attributing the disease to the ex posure. The effect of the incident might well have extended beyond the cost of compensa tion and into the area of threshold limits by the publication of a case report. A fine discretion is required in recognizing the first or isolated instances of injury to a new chemical and yet avoiding the inclusion of cases solely on legal or social motivation. In a study of the long-term effects of a solvent, records were kept of the causes for dispensary visits. A slightly higher in cidence of gastrointestinal complaints was found in the exposed group, and since this had been reported previously by others, it seemed significant. However, the incidence of respiratory complaints was as much lower for the exposed individuals. In an in terpretation, it would have been equally proper to assume that the solvent vapors "protected" the individuals against respira tory disease, and that this beneficial effect might offset the gastrointestinal difficulties. As might be suspected, when these relations were put to the test of statistical signifi-  284 Septemb, + cance, both could have been due easily to their needed contributions (which chance alone. from Dr. Sterner's intention) or de] A final word is in order concerning- the more experienced contributors, it shi methods for developing data by which the immediately pointed out that there* environment is described--the conditions of least two practical and very helpful exposure. As much variability can be, and of overcoming certain deficiencies ii is, encountered with this function as with in the basic data. One is the safety; the methods for laboratory experimentation the other, the periodic re-evaluation: and in-plant clinical examinations. The ac threshold limit values. The safety fac curacy and reliability of the analytical tech been built into most of the values niques, the relation of the time and site of threshold limits list.* The factors 2 sampling to the true exposure, the presence or even greater have been applied t of other agents which might modify the values. There are, of course, some ] single action of the toxin under study are exceptions, such as the present val some of the many factors which must be trichloroethylene, for which the th; critically examined in the light of their limit is the absolute ceiling, but gti usefulness for correlation with a given set such instances are rare. In gem of experimental or clinical findings. greater the uncertainty in the data1 It is obvious that there is no single cability to human industrial expostj method or pattern of methods which can larger the factor applied. This lowej satisfy the varied requirements for estab the limit value by an arbitrary safetij lishing threshold limits. It is equally obvious may at times provoke some controvaj that even though rigid standards were de cause now the value becomes one of d scribed which would satisfy a discriminat not fact. Be that as it may, the safetj ing jury of scientists, the available data incorporated in the air standards gi for establishing threshold limits for all but creased assurance of safety to man; a very few substances would fail to satisfy ful values. such limits. The real use for threshold lim The re-examination of the listed vi its, as a guide for industry in the control the committee provides annually fo) of exposures and as a measure for action by justment of all values, whatever thel governmental agencies, demands a con tity, upon submission to the commj tinuing improvement in the quality and new and experimentally supported fii quantity of the methods by which truly Repeated scrutiny and re-appraisal \ valid criteria may be achieved. sort can lead finally only to assignn safe values on which complete relia| Prepared Discussion be placed. The committee welcomes I information.** HERBERT E. STOKIN0ER, Ph.D. Chief, Toxicologicai Services Public Health Service Cincinnati, Ohio Dr. sterner has discussed the great diffi culties attendant on developing ade quate data for threshold limits, the many imponderables in their interpretation, and their unsatisfactory and necessarily always incomplete nature. All these considerations should certainly be thoughtfully considered and strongly stressed, because errors in judgment cannot be afforded. But lest these many considerations seem so formidable to many potential investigators as to prevent Need for more Data TN t h is connection another point i 1 in Dr. Sterner's discussion sho strengthened--namely, the need foi data substantiating the choice of si posure levels based on industrial expi Much useful information is undoubt the files of many plants. Indication was the fine evidence on six industri T he incorporated safety facto r in th e threat values, although added a t tim es because o f ui in th e v alue as related to h u m an exposure, o ft provides an ap p reciab le m arg in of safety. F or ti th e correctn ess o f th e term "threshold lim it" q u estio n ed . It m ig h t m o re p ro p e rly be replaced h y g i e n e s t a n d a r d . '* A llan Colem an, C hairm an Threshold Lim its tee, A C G IH , C o n n e c tic u t S t a te D e p a r tm e n t ol H a r tf o r d 1, C o n n e c tic u t. trial Hygiene Quarterly l*dvs 285 ceS derived from many years of plant **^rience that came to light last year at meetings from the presentation of o j-bert J- Weber.1 Others should have *! ilar material that should be brought to *h attention of the Threshold Limits Com ! e(tee. Like Weber's material, all of it need * t ,e novel or presented to show need for Ranging existing limits; equally valuable 're data confirming existing limits. As never before, interest in the value of the control ,f industrial environments is being shared >,v management generally. Greater numbers ',f industrial hygienists than ever before are t*ing engaged by industry. Is it too much to hope that meetings such as these will orient !he thinking of properly placed industrial hvgienists to secure much needed plant in formation to aid in the choice of safe lim- ts of human exposure? The type of information needed may be listed as follows: (1) Air concentrations should be determined for the substances under study through a complete cycle of plant operations and with reasonable regu larity in order to obtain a true picture of the range and fluctuations of exposure. (2) the data should have good accuracy. (3) The observations should be carried out over n reasonable period of time--a minimum of live years. (4) The air concentration data ihould be correlated with a good medical program. A pattern for such work is that of Dr. Stemer's 10-year study of workers exposure to butyl alcohol.2 There is a real need for more data based on industrial experience. The often-heard statement "the threshold limits are nothing but educated guesses" unquestionably re jects the wish at least that more data be Irmly based on industrial experience to sub stantiate the choice of limits. As a member if the Threshold Limits Committee, I was concerned over the statement and took the trouble to review each substance in the threshold limit list for 1955 as to the basis tor choice of the level. The results are shown :n Table I. It is possible that everyone would not arrive at precisely the same figures, but 1believe that their magnitude would not be much altered. Although the table shows that `he educated guesses account for a relative ly small number, it does confirm the often expressed feeling of the need for more !olidly based levels. Table I shows a number T able I. Basis for Choice of T hreshold Lim it Value Study Type No. of L istings Percent Total L is tinsra (223) A nim al Industry A nim al & M an " Educated A nim al & In d u stry Guess" M an 94 42 51 23 23 10 25 11 21 9 9 4 Source uncertain ' Based in p a rt on Cook, W , A., and from docum ented m aterial of m ittee A .C .G .I.H . 1953-1955, 3 Ind. M ed. T hreshold 14:936, L im its 1 1945, Com- of other interesting facts : (1) that most of the values have some sort of scientific basis; (2) that each level has been documented either by Warren A. Cook,3 or by the Committee on Threshold Limits; (3) that tb j values based on animal experiments account for the largest number, 42%; but (4) that values having some industrial basis account for 13 of the total. The values ascribed to the "man" cate gory arise from two sources--that of Nel son, et al,4 and those more recent publica tions of the Dow Chemical workers, Irish, Rowe, Spencer, Adams et al. The "Educated Guess" ^ few words should be said in defense of the "educated guess." A review of the values described as guesses indicates in the instances in which sound information has later become available that the "guess" was remarkably good. Two prominent examples only will suffice--hydrogen fluoride and uranium. A safe exposure level for hydro gen fluoride was set at 3 ppm on the very limited evidence supplied by a study in ani mals by Ronzani in 1909.5 Last year a reportfi culminating many years of study of fluoride exposure in the aluminum industry, involving thousands of air and urine analy sis for fluoride and studies of roentgenographic changes in bone, showed without question that air levels double the accepted limit gave rise to perceptible changes in bone in only a few individuals, and only after many years of exposure, thus vali dating the wisdom of this "educated guess." In the case of uranium, an engineering bench-mark had to be "guessed" at early in the days of the Manhattan Project. After a 286 Septembt review of the quite limited animal data on uranium then available, Dr. Stafford Warren suggested that the "safe" exposure level for uranium be the same as that for lead, 0.15 mg cu.m. After $500,000 and many years had been spent in research, the safe levels of exposure to uranium compounds were found to bracket this value very closely. Levels tor Cancerigens 'J 'here is still one group of substances for which some method should be devised for establishing safe air standards--the indus trial cancerigens. How shall we establish the limits for this type of substance? Thus far the question has been sidestepped com pletely. As a result, with one exception, nickel carbonyl, limits taking into consider ation potential cancerigenicity have not been assigned. Several industrial substances are known or suspected cancerigens; many more are suspect on the basis of animal ex periments. As a suggested method of ap proach, the following is offered: To the level judged safe for other types of systemic in jury add a safety factor for carcinogenicity. The magnitude of the safety factor is sug gested to be from 100 to 500. This provides at least a second power of 10, which, from the well-known dosage-response hypothesis, provides at least a fourfold longer interval before effects may be expected to occur, or conversely at least a response with !4 the intensity. This manner of approach has been used for nickel carbonyl. A tentatively safe level for systemic effects from repeated daily exposure has been set at 0.1 ppm; one- hundredth this level, or O.OO^ppm for nickel carbonyl on the basis that poisoning gives rise to a substant crease in the incidence of lung cance, realized that unfortunately the saf its for all industrial cancerigens can so readily resolved. This is especially; dye intermediates, such as benzidij naphthyl amines whose major route is not commonly via the lungs but t | the skin and gastrointestinal tract.! are laundry and protective equipment lems not solvable by air control. There are undoubtedly substand which the suggested procedure m* strictly apply, but imperfect as it _ the suggested method is felt to be a the right direction and serves better l exposures to industrial carcinogen considering the problem too difficult with at the present time. References 1. W eb Eft, H. J . : T h r e s h o ld L i m i ts , A Pane] sion. A I H A Quart., 16:38, 1955. 2. S t e r n e r , J . H.. C rou ch , H . C ., Bro c k * * C u s a c k . M .: A T e n - Y e a r S tu d y o f B u ty l Afc p o su re. A I H A Quart., 10:53, 1949. 3. Cook, W . A .: M a x im u m A llo w ab le C o n e of In d u strial A tm ospheric C o n tam in an ts. In 14:936, 1945. 4 . N e l s o n , K. W ,, E ge, J . F ., R o s s , M ., W o E ., SILVERMAN, L .: S e n s o ry R e s p o n s e T o C e ji trial S o lv e n ts . J . In d . H y g . T o x i c ., 25:282, 19 5. Ro nzan i. E .: U b er der E influss d er E in von reizenden G asen der In d u strien a u f die der O rganism us gegenber der in fek tiv en Kn A rc h . f. H yg ., 70:217, 1909. 6. Ir w in , D. A .: C lin ical F in d in g s W hich ticip ated A fter L ong-C ontinued E x p o su re to P resen ted before Sym posium on F luorine, L ab o rato ries, C in cin n ati, Ohio, M ay, 1954. Engineering and Chemical Application of Standi ALLEN D. BRANDT, Sc.D. Bethlehem Steel Company, Bethlehem, Pennsylvania The why and how of threshold limits for air contaminants have been discussed by the preceding speakers. The logical next and final consideration is the use or applica tion of these limits in preventing occupa tional diseases, in avoiding complaints from exposed employees, in maintaining employee efficiency, and in promoting good house keeping. To engineers engaged in industrie giene, threshold limits provide the refe line or bench mark upon which all cor tions and calculations for the control i contamination are based. It is charact of engineers and of others in the phj sciences to want to reduce to numbers! problems with which they deal, because i then can they be attacked in a precise : Industrial Hygiene Quarterly 287 ner. It has never been clear in my mind whether engineers enter this profession be cause of their love for and adeptness with numbers, or whether they acquire their de sire to work with numbers as a result of pursuing the engineering discipline. Nor is it important to the subject matter of this paper to know which is the cause and wdiich the effect. Suffice it to recognize at this time that this quality is omnipresent in engi neers, and if industrial health problems arising out of environmental conditions are to be prevented or controlled, the physiologi cal responses of the industrial employees will have to be translated into some physical yardstick readily understood by engineers whose function it is to regulate the environ ment in the best interests of the worker and of production. Threshold limits serve several very use ful purposes in the industrial health picture. Through careful studies of the air at exist ing operations or in existing plants, they permit deciding whether a health hazard is present or not, and if present, how7 severe it is. They are quite useful to the discern ing physician in determining whether the health impairment of an exposed worker is attributable to his employment, and in this respect may exert significant influence on the disposition of occupational disease claims. Where a health hazard from air con tamination has been found to exist, thresh old limits frequently are the guidepost indi cating which avenue of control to follow to achieve a satisfactory solution to the prob lem. Having ascertained the avenue to fol low, the threshold limits provide the only acientific basis as to how7far we must jour ney to arrive at the desired destination. Having alluded briefly to the principal roles played by threshold limits in industrial oygiene, let us now examine each one of 1 ^ * 0 more carefully and consider their ap plication in industry. Palliating Potential Health H azards I** This age of mechanization and high speed production, there are many places r 1lndustry where some of the materials behandled, consumed, or processed escapes the air. Whether this airborne con- **uiant constitutes a health hazard de ***** upon its concentration in the air breathed by exposed persons as related to its inherent harmfulness. The concentra tion of the material in question in the breathing zones of employees is capable of rather accurate measurement. If the harm fulness of the contaminant is known, also, in terms of a safe atmospheric concentra tion (the threshold limit), the presence or absence of a health hazard is apparent. In this way threshold limits are applied widely in industry to decide w7here better control of the environment is needed and where it is not needed. Assisting in the Diagnosis of Illness TVTot infrequently, there is a discourag- ^ ing duplicity of symptoms manifested by persons suffering from different illnesses, especially in the early stages. If the symp toms exhibited by an employee are compati ble with those that might be produced by the air contaminant to which he has been ex posed as well as w7ith several other stimuli of non-occupational origin, comparison of the concentration to wThich he actually was exposed w7ith an accepted threshold limit for this same material is of inestimable value in arriving at the most probable cause of the illness. If the exposure exceeds the thresh old limit and the symptoms are indicative of illness from such exposure, the occupa tional hazard may well be the cause. If, how ever, the atmospheric concentration of con taminant in the breathing zone of the worker has been w7ell below7 the threshold limit, it suggests to the examining physi cian that all other possible causes must be explored thoroughly before a diagnosis can be made. This application of threshold limits some times reaches out into the compensation courts. The employee's personal physician, having only the complainant's description of his job rather than factual data to guide him in the diagnosis, not infrequently will arrive at an erroneous conclusion. Contro versial medical opinions are introduced in evidence at the hearing, and the carefully measured atmospheric concentration of the material in question as compared to the threshold limit for this material is the only bit of uncontested scientific informa tion that the referee has available on which to decide the case. This is not to say that the controverted cases usually are decided on 288 September, this basis. Other factors, many of which are not scientific in nature, play a part in the adjudication of such compensation cases. Nevertheless, testimony of this kind has been helpful in arriving at the truth in some cases, and there is hope that an increasing amount of importance will be attached in the future by compensation courts to evi dence of this kind, as more and more experi ence is accumulated to support the threshold limits and as knowledge of this useful tool becomes more common in compensation circles. Control of the Hazard JT is in the control of atmospheric contami nants that threshold limits find their most important application in industry. The judicious use of threshold limits in selecting and designing control measures and equip ment will result in satisfactory working con ditions at a minimum in cost and interfer ence with production. Possibly the best way to add clarity and reality to this general ob servation is to implement it with conspicu ous examples. Given: a production type degreasing op eration in which the parts are cleaned mechanically by immersion in a tank of car bon tetrachloride. The tank is provided with a local exhaust system, but is so wide that considerable of the grossly contaminated air at the remote side is spilling into the work room air and nearby employees are exposed to vapor concentrations in the order of 250 ppm. Problem: how best to correct this condition. Obviously the condition can be corrected by rebuilding the exhaust system and in creasing its capacity, or by changing to a less toxic solvent. Imagine how insecure you would feel in facing this problem if there were no threshold limits to guide you. You would not have any idea how much re duction in vapor concentration is necessary to eliminate the health hazard. Even if a less toxic solvent were substituted, there would be no way of knowing whether it alone will solve the problem or whether improved process ventilation is needed, and if an in creased exhaust rate is needed, how much increase is required. To be on the safe side you probably would change to a safer sol vent and redesign the exhaust system with 100c/c capture as your goal. How much simpler it ifi with threalj limits to provide a scientific basis for g sideration. Obviously the exhaust sygfc could be redesigned to capture all the vaj rising from the tank surface and from f withdrawn parts. This would have the se ous disadvantage of increasing markedly | consumption rate of the solvent, carb tetrachloride. Since methyl chloroform h physical characteristics very similar to d bon tetrachloride, if it were used instead carbon tetrachloride the atmospheric o* centration of vapor would be about the san and since the threshold limit for meti chloroform is 500 ppm, the problem won be solved. However, this solution carri with it a substantial cost disadvantai methyl chloroform being more expend than carbon tetrachloride. Therefore, lj consider using trichloroethylene or chloroethylene in place of carbon tetz chloride. It too costs more than carbon tet| chloride, but because of its lower volatile the consumption rate will decrease if mu of the loss takes place via the air enteril the exhaust system. But what about solvi| the health hazard. The threshold limit fi trichloroethylene and for perehloroethylei is 200 ppm and the vapor concentration wi found to be in the order of 250 ppm wh< carbon tetrachloride was used. Because the lower vapor pressure of the propo substitutes, the amount of vapor gettii into the room air will be much less and tl exposure of the workers will be reduced 1 a value well below 200 ppm. Thus by eg ploying the yardstick of threshold limit two easy and certain solutions are found 1 this problem without difficulty. It may be well to emphasize at this poii that other important considerations mui not be overlooked. All factors must be born in mind whenever one is faced with an ai contaminant control problem. Sometime careful selection of a substitute will nc not only bring the health hazard unde control but also will increase the oj erating rate, or decrease the consumptio rate of a process ingredient thereby achiei ing considerable cost advantage as a bj product of the change. By way of illustrs tion, the slower evaporation rate of pel chloroethylene as compared with carbo: tetrachloride not only minimized the poten tial health hazard at a specific gravity test Industrial Hygiene Quarterly 289 ing operation but also resulted in an an nual saving of about $3000 in the cost of the liquid consumed. This, in spite of the fact that perchloroethvlene is much more ex pensive per unit of measure than is carbon tetrachloride. Criteria in Substitution 'J'HE application of threshold limits in the control of health hazards by substitution is such a useful tool that several other ex amples will be cited. Solvent degreasers and cleaners are used ao widely, and frequently so erratically, that it is difficult to be certain that adequate control precautions are being observed at U times, especially if the solvent in ques tion is relatively toxic. For this reason in dustry is engaged in a search for effective solvent cleaners which inherently are much less toxic than those in common use today. Because carbon tetrachloride has qualities which make it an exceptionally satisfactory olvent from the functional viewpoint, it is in constant demand by all operating people who are not fully aware of its danger. How ever, reputable jobbers and vendors of pro prietary solvents of all kinds have become increasingly conscious of the health con siderations attending the use of their pred a te and have been selecting their product WiWdients on the basis of toxicity ex|*6ssed in terms of threshold limits. It is "Wtening to note how commonly the supPuers of solvent degreasers and cleaners are ucing the valuable yardstick of threshold |*nuts to concoct ever safer products to sat**fy the complex degreasing and cleaning Hands of industry. it appears to us that we are on the doortep of a new era whjch will see great ad vances in solvent cleaning safety, thanks to hreshold limits. It would be remiss at this pot to mention that methyl chloroform Promises to be the carbon tetrachloride of `fmorrow. If properly inhibited to prevent ** corrosive action on certain metals, it PP^rs to be the perfect answer for those and degreasing jobs on which only 7*ri>on tetrachloride has been acceptable in `P ast. Threshold limits have played and continue t* " a y a key role in the control of the associated with abrasive blasting of " kinds. Originally, sand was used for all Jr- abrasive blasting. This created a serious health hazard on the part of the operators and nearby employees, and not infrequently it posed a serious wear problem on nearby machinery which had not been protected during the blasting. The problem was soon brought under control for most productiontype blasting by housing such operations in well ventilated rooms, chambers or booths. In addition, steel grit was substituted for sand on most work of this kind if the grit was recoverable and reusable. Because of the high cost of steel grit as compared with sand, it was not readily accepted for work in which most or all of the grit was lost. More recently, however, several types of grit have been placed on the market which are by-products of the metal smelting and refining industry. While not as economical as sand, they are much less expensive than steel grit. These substitute blasting grits came into use because the probable health hazard attending their use is much less than is the case with sand, in terms of the threshold limits of the ingredients and the amounts present. Selection of the particular by-product grit to be used on any blast cleaning job is made on the basis of avail ability and the probable health hazard as ex pressed by the threshold limits of the in gredients and the amounts present. But the application of threshold limits on abrasive blast cleaning jobs does not end here. Since most blast cleaning jobs are not done mechanically in well ventilated booths and require respiratory protection for the operator, threshold limits enter the con sideration as to the proper respirator to be worn. Supplied-air respirators or helmets require a supply of air of respirable quality under pressure. Consequently for isolated jobs where a separate blower or semi-com pressor would be required to furnish air to the respirator, conventional dust respirators of the mechanical filter type may provide adequate respiratory protection depending upon the harmfulness of the dust produced as determined by the threshold limits of the grit ingredient and those of the surface be ing cleaned. In reverse order, of course, sand may be used with safety if the blaster can conveniently wear a supplied-air helmet and a source of respirable air for the res pirator is readily available. It must not be inferred from the fore- 290 Septembt going that threshold limits per se of a process material or the ingredients thereof determine the extent of a health hazard. The atmospheric concentration of the contami nant in the air to which the employee is ex posed is equally important. But when con sidering a type, or class, of operations in which the employee's exposure does not vary tremendously, as a rule, from one operation to another, substantial changes in the threshold limits of the process material in gredients overshadow the smaller variations that take place in atmospheric concentration of the contaminant. This is not to say that all abrasive blasting operations produce the same concentration of dust in the ambient air, but rather that, all other things being equal, a change in the nature of the blasting material will not affect the dust concen tration nearby as profoundly as it will the applicable threshold limit, if the sub stitute material is well chosen. For ex ample, it is extremely unlikely that steel grit or even a substitute grit pro duced as a by-product in the glass or in the smelting industry and containing less than 5 cr free silica would create as severe a hazard as does sand in blast cleaning of unpainted surfaces. Obviously, if painted surfaces are blast-cleaned, a new factor en ters the picture which may well overshadow the composition of the grit. However, here again threshold limits would be consulted in conjunction with the dust concentration to determine (1) the severity of the health hazard, (2) whether there is any advantage in using a more expensive substitute grit, and i3) what is required in the way of personal protection. Threshold limits are especially useful in designing for the control of air contami nants in all cases where the sources are of such nature, location and distribution as to permit effective control by general ventila tion. In the case of solvents or other liquids being used in a room, the required ventila tion rate can be calculated very easily if the consumption rate of the liquids is known. The same is true for particulate matter if the rate of dust generation is known, or if information is available as to the dust con centration and ventilation rate prevailing at a comparable operation elsewhere. By em ploying the threshold limit for the dust in question, the ventilation for the operation under consideration can be calculated* ly, rather than guessing what it she or following some rule of thumb, existing shops or rooms where air con nants are present in objectionable tration, the threshold limit of the co nants in conjunction with the current! eral ventilation rate (which can be ured) permits calculating accurately is needed in the way of fans to eliminaft objectionable or potentially harmful tions. Threshold limits remove much conjecture otherwise involved in pro of general ventilation for purposes of i fume and gas control. Classification of Material for Control Pu JgVEN THOUGH it is desirable to con each problem separately, not infreq ly it is expeditious to set down rules go ing the control of hazards created by \ operations that differ considerably in In such instances the toxicity of the pi materials and the by-products in ter their threshold limits may be grouped! used in conjunction with similar group of the other variable factors influencing hazard, to arrive at a set of relative rules to cover the whole gamut of oper in question. For example, let us co open-surface tank operations of all whether they be pickling, degreasing, i cleaning, painting, plating or other ment. By dividing all the materials inv in these operations into three groups i ing to their threshold limits and into i groups on the basis of the other gov^ factors, it was possible to classify all ] ble types of open-surface tank oper into four classes as regards the ventill rate required at the tank for adequate i trol of the contaminants released.1^ Wii out threshold limits to serve as a guide! would have been well nigh impossible^ develop these few rules governing s o ; different materials and operations. Similar codification is under consid tion by American Standards Asso Committee Z-9--Exhaust Systen other groups of processes and operatl having like industrial hygiene aspects,-I for example, bulk materials handling, face coating operations, mechanical cutttai and abrading operations, and abrasive blast* ing operations. Without threshold limita Industrial Hygiene Quarterly 291 ucb codification would be beset with un ending research, consultation and guesswork ot inch magnitude as to make consummation dubious and as to place the final result under Mrious suspicion. Experience THE&e is much merit in the proverb "the * tu te of the pudding is in the eating." Therefore, past experience may well be ex amined in the use of industrial hygiene xtandards to prevent occupational illnesses. In this respect, the literature is most dis appointing, not because the experience has been unfavorable but rather because the literature contains few reports of first hand iM fltigations into this relationship. The 1M which probably best demonstrates the 1W. bf threshold limits and the results dlleved may be found in the report by Mc Connell, et al., on the occupational disease sperience in the government-owned ord- Bce plants during World War II.3 Since It wee my privilege to take part in this pro- **> from its inception, I can speak of it Ith some feeling. When the United States first began am- nnition production at an increased rate 1940, very little information was mrailable on the toxicity of TNT of such na- that would permit setting a threshold TNT consumption in the U. S. in ^ " Period between 1918 and 1940 was esmillally nil and threshold limits were not so common as they are now. That is toxic was not questioned, for ac to the best information available 17,000 persons in ammunition plants poisoned, 475 of them fatally, during *"*1 War I. It was this sad experience rifle d with the related problems of absen- rr**1 inefficiency and manpower shortage the War Department to undertake with the U.S. Public Health Service afcgressive industrial _5munition plants. health program in Inasmuch as the n, sdereicoiudsedpaottetnhteiaoluptsreotblteomcownjausreTuNpTa, allowable concentration for this see, the standard to be based on the formation available as to its toxicity, Srative physiology, and the concentra- tt Was known to be achievable by good ng practice. This "educated guess" Out to be 1.5 mg m3, a value still in common use. Reference to the report by McConnell, et al., shows how closely the in cidence of illness from TNT paralleled the exposure curve. Owing to the tremendous scarcity of the equipment needed for engi neering control of the fumes and dust, such as motors, fans, and sheet metal for piping and hoods, the weighted-average level of 1.5 mg m3 was not reached until the latter part of 1944. Howrever, the fatalities from TNT were held to 22 for the entire period of the war1and the total number of lost time cases was only 56 as of May 1, 1945, the end of the report period covered in reference 3. The incidence of early symptoms of TNT poisoning was quite high, and there is little doubt that the number of lost-time cases as well as fatalities would have been much higher if it had not been for the excellent periodic medical examination program then in effect in the ammunition plants, especial ly in the loading plants where the exposures were the worst. It is interesting to note that the case rate of early systemic effects was close to 500 per 1000 man years of exposure for the first six months of 1943 when the weighted average exposure to TNT was about 2.7 mg m3, and about 15 of this rate for the last six months of 1944 when the ex posure had been reduced to about 1.4 mg/m3. The control program in the ordnance plants during World War II would have been inestimably more difficult, and there is little doubt in my mind that the final report would portray a different story, a much sadder one, from that told by McConnell, et al., if there had been no standard for TNT. In the face of the dire shortages of equip ment needed directly by the Armed Forces, and other equipment needed to produce the weapons of war, authorization to divert some of this material for dust controt pur poses would have been out of the question in the absence of a threshold limit to indi cate how badly and wThere such equipment was needed. As in many other factors af fecting our daily life, standards are the very foundation of industrial hygiene. Without them the engineer in this field is placed in the same position as the boy who is sent to the hardware store by his father to buy a "big" bolt. The boy would have as good a chance of getting a bolt of the right size as an engineer would have of preventing the vast majority of occupational illnesses and Septembi 292 complaints if there were no threshold limits. In conclusion a few remarks are in order dealing with the philosophy of and justifica tion for threshold limits, and are added here because they serve as a reply to the criti cisms often leveled at industrial hygiene standards. It is not unusual to hear that these limits are useless, if not even dangerous, because in the event that the atmospheric concen tration is measured inaccurately, reliance on the standards creates a false sense of se curity on the one hand, or needless alarm on the other, depending upon whether the measured result is in error downward or upward. It requires little mental gymnastics to realize that this is equivalent to con cluding that there is no sense in asking your grocer for a dozen oranges simply because he may make a mistake in counting them. By definition, threshold limits are those concen trations of contaminants which in the light of current knowledge will not cause harm to persons exposed continuously day in and day out during the normal working hours. Reference is to the concentration actually existing, not to some value that the investi gator might conclude exists. Concentrations can be measured accurately, but to do so re quires skill, understanding, patience and energy. It is rather discouraging to note how frequently conclusions are based on wholly inadequate data. It is not surprising either that the concentrations frequently "measured" have little relation to the true weighted average concentration. . The philosophy of threshold limits is that each one represents a concentration of the substance in question that will have no demonstrable adverse effect on the health of exposed persons. Not all persons react the same to stimuli of this nature, but rather any given population follows approximate ly an average random distribution pattern as regards response to atmospheric contami nants. That is to say, there is an occasional person at one end of the curve who is un affected by relatively overwhelming concen trations and there is an occasional person at the other end who is affected by extreme ly low concentrations. To prevent any ad verse effect upon every last man may require such low concentrations as to be impractic able. The threshold limit is merely a con centration that intersects the random dis- tribution curve at a very low pointlow this point is for any substance canstated. That it is not the same for all terials is obvious, and that it serves tovent harm to all but a relatively f,, obvious, also. Serious or irreparable da to the occasional person who falls to the' of the threshold limit on the random ' tribution curve can be avoided by an priate medical examination program. References 1. A m e r ic a n V entilation and S ta n d a rd s O peration A ssociation: S afety Cod, of O pen S urface Tanka, 1 ^ ^ N e w Y1 Uo rl ka , i1v9e5i1*. 2 Brandt, A, D. Exh& ust System s. E le tti. E n g in e erin g H andbook, C h a p te r 27, R einhold Pu C o rp o ratio n , N ew Y ork, 1955. 3. M c C o n n e l l , W , J . , F l i n n , R . H ., a n d Bl-- D .: O ccu p atio n al D iseases in G overnm ent-O w ned n a n c e E x p lo siv e s P la n ts . O ccu p . M ed. 1 :551-618, 1946. ,,,, ,, 4. M cCo n n e l l , W . J ,, a n d F l in n . R . H .: S o f T w e n ty -T w o T r in itr o to lu e n e F a ta litie s in Woi II. J . In d u et. H y g . & T o xic.. 28:76-8$, M ay, 1948. Prepared Discussion ARTHUR C. STERN, Chief Air Pollution Community Program Robert A. Taft Sanitary Engineering U.S. Public Health Service Cincinnati, Ohio IN AN ENDEAVORto safeguard the heal! the worker and the public from expto harmful substances, two basically c ent types of standards have evolyed.fi type--the threshold limit, or maximun lowable concentration--is the perform* type of standard. The other, about w little has been said by the previous speake is the engineering type of standard. v To make the distinction between the t f types of standards more concrete, take r specific example of a stave-type tumbO mill (Fig. 1) in a ferrous foundry. Int are placed castings with sand both cling to their outer surfaces and in their intet cavities in the form of cores. The expr purpose of placing these castings into mill is to clean them of this sand. The U is, therefore, by its very nature, a devf designed and operated to cause sand leave the castings and enter the ambient While it is true that most of this sand * St. 293 Fig. I. Sfave mill. e floor under the mill, it is equally , a tremendous number of free silica will become air-borne by the rom the viewpoint of the performdard, it is necessary to keep the `re of the cleaning room below the limit for free silica by whatever owner of the foundry may choose Standards VIEWPOINT of engineering stand ' tile reasoning is somewhat as fol I widely recognized in the foundry that stave mills are bad dust projuid require enclosure. Industry, by trial and error, has a satisfactory type of ventilated (Fig. 2) which not only keeps the Of the workroom, but also keeps the the mill free of much of the sand otherwise accumulate there and be carted away. type of enclosure, having been by most of the industry, should become the standard of the entire .^engineering standard should therePjritten so specifying this enclosure P^ect to structure and ventilation g foundryman building an enclosure p these specifications will achieve `1of his stave mill (Fig. 3). *ty of Standards i ***ORE some of the arguments against each of these approaches, at the same objective--the safethe worker. The principal ar gument in favor of the performance stand ard is that it is noboby's business except the owner's as to how he achieves control of the hazard, just so long as he does, in fact, suc ceed in so doing. The corollary argument against the use of the engineering standard is that there is no positive guarantee that once the owner has invested his money in an enclosure conforming to the standard speci fication, the exposure of the workers in the cleaning room will be below threshold limit. The same logic can be used to plead the case for the engineering standard. The dust load in the cleaning room air comes from many sources in addition to the stave mills (Fig. 4). If some of these sources remain uncontrolled after the stave mill has been enclosed and ventilated according to stand ard specification, it is entirely possible, even probable, that dust concentration in the room will remain above threshold despite entirely satisfactory dust control by the stave mill enclosure. Engineering standards thus have the inherent capacity to allow a step by step orderly approach toward the ultimate goal of achieving below-threshold concentrations in all parts of the plant at all times. New Construction Assume that a new foundry is being de^ signed. To apply performance standards, Z9i September, the foundry must first be built, placed in cleaning room equipment and applying operation, and then subjected to air sam to the best of his ability. In this situa pling and analysis. No reputable and well- argument arises, not as to the need for informed engineer would design this new engineering standards, but rather plant without seeking the engineering their accuracy. Should the opening for standards for the control of dust from the be 3V2" or should it be 4"? If the stani says 900 cfm, and the signer is firmly convi that 450 cfm are suffic he should be prepare< back up his independem judgment by later pj that concentrations are low threshold in the ( pleted workroom when lower value of cfm is ployed. But what if the desij reluctantly makes the pacity 900 cfm to conf( TUM BLER Duct velocity : 5COOFPM Duct velocity* 3500FPM minimum. to the standard? Is tiff equal necessity for Entry toss - 2 0 to 4.0 VP (depends on design) Entry loss tones with fotte-ofFO 2 5 0 5 0 VP Squore miti side diem r Up to 24 ncf. 25to30 " 3/ to 36 * 3? to 42 * EX H A U ST VOLUMES Round mffl f.D tranches Upto 24 included 24 - 30 30 - 36 36 - 42 42 - 46 CFM Trunnion Stare* 430 800 680 900 990 900 /330 1330 1750 1750 sponsors of the standard guarantee compliance v the threshold limit? 1 difficulty of so doing in possible presence of traneous sources of the ( taminant that have alrei been mentioned almost variably makes it impoi ble for the sponsor to 43/0 43 * 49/0 54 m 43 - 54 54 - 60 2200 2200 2730 2730 dertake such guarantee^ 55/06O* 60 - 66 3300 3300 Double Jeopardy 6! to 66" 67 to 72 " 66 - 72 3920 3920 4600 4600 ^There the engmeei standard is a tool in For lengths Over 7 0 " , increase CFM Uroportionately hands of an enforcem Fig. 3. agency, the user of 31 Tumbling mills. standards frequently fi that he is placed in dou jeopardy. He has to compj with engineering standard with which he may di< agree, and then, having ployed all such standards! every operation, may be not in compliance the performance standai The wise enforceme* agency will let it be knoMf that it will not permit sue double jeopardy to occui Fig. 4. and that any factory own Floor plan showing layout of equipment including stave mills. who adheres throughout t iai Hygiene Quarterly engineering standards known to the ill not be asked to re-do the job. then, is the plight of the working"ht in the middle? Is he likely to pawn, forced to work in an un atmosphere because the sponsorcy is unwilling to admit a mistake effectiveness of its engineering ? We are indeed fortunate that ty of our engineering standards is that such cases occur infrequently, (hey do occur, they are almost always e cases with respect to the threshand are well within the factor of .lit into the threshold limit itself. ndards -TIONSHIP between performance aeering standards for in-plant i has its counterpart in that be- ospheric and emission standards d of air pollution. A major difthat, whereas there are a large f l accepted threshold limits for in"phere, there are almost no stand^only accepted for the outside -e. iasion standard is very much in position as the engineering stand , In-plant controls in that adherence ae specific stack does not guarantee of the air of the surrounding Thousands of factories have t engineering standards to the 'proven that when every source of ?t release is controlled as specified licable engineering standard, the been to maintain the factory `c contaminant concentration bethreshold limit. Such beautiful tions, which are easy when the re in question is confined by four a roof, become almost impossible the open air over cities and towns. it is reasonable to believe that in a small confined space will -tually work in the largest of con- that having the inevitable in ceiling as a roof and four topo'rI or meteorological side walls. lor the Open Air OK once again at the principal - t for performance standard in t situation, i.e., that it is no 295 body's business but the plant owner's as to how he achieves control of the hazard--so long as he does so. Is this doctrine equally applicable to the air over a city? It hardly seems so. No individual owner or group of owners could or would be willing to take the responsibility for the maintenance of a specified level of contamination of the air of the community in which they operate. They would quickly recognize the many factors over which they do not have the same measure of control that exists inside the factory building. For them, the emis sion standard is a boon. Here the argument is not with the con cept of the emission standard, but with its numerical value for a particular installa tion. One of the most interesting develop ments of the past decade along these lines has been the fact that public utility power plant designers have consistently specified equipment for fly ash control to meet emis sion standards much more stringent than public regulatory bodies have required. In dustrial foresight and conscience is in this, and a number of other industrial hygiene and air pollution control fronts, well ahead of public regulatory demands. Conclusion TN both these areas, this is the hope of the future. The dictum of a Massachusetts court of many years ago that the best plants in an industry may reasonably be used to set the standard for the entire industry provides the real key to future progress. America is blessed with real industrial leadership and statesmanship. Neither in dustry's leaders and statesmen nor the pub lic will tolerate the acceptance of stand ards which are below the best that industry can, by example, provide. Summary ' J 'here are two general types of standards --performance type standards and engi neering standards. The former specify threshold limits. The latter specify the per formance of a particular piece of equip ment. Arguments are given pro and con as to the relative merits of these two ap proaches to the same objective--the safe guarding of the worker. The engineering standard approach is particularly valuable in new construction where no equipment September, ; 296 exists at which to measure atmospheric con centration before designing safeguards. The similarity of engineering standards for the in-plant situation to emission standards for atmospheric pollution control is noted. Prepared Discussion THEODORE C. WATERS, L L.B. Baltimore, Maryland This comment and discussion of the ex cellent paper presented by Dr. Brandt will be directed to the legal aspects of the application of threshold limits in the form of maximum allowable concentrations for the control of air contamination. May I identify myself as one who has served as counsel for industrial organiza tions; and while my views may be prejudiced on their behalf, I feel that the industrial point of view is important in our discussion of this matter. _ . It is needless to say that today industry is prepared and wishes to accept full responsi bility for the control of potential hazards incident to employment. Simply stated, it. is good business both from the financial standpoint and also in the promotion of industrial relationships for employers to concern themselves wfith the protection of employees from all types of industrial in juries, be they accidental or resulting from occupational disease. The ultimate cost of compensation for such injuries may well ex ceed the cost resulting from installation of effective methods for engineering and medi cal control, aside from the fact that the em ployer's relation to his employees is pro moted when employees know that the em ployer is doing everything practical to effect protection. Question arises to the propriety and ad visability of the adoption of codes of in dustrial hygiene defining maximum allow able concentrations of toxic materials. The word "code" is defined as follows. A body of law established by the legislative authority of the state, and designed to regu late completely, so far as a statute may, the subject to which it relates." The adoption of codes having the full force and effect of law is unnecessary, inadvisable, and may well lead to unfortunate problems of admi tion to which brief reference may be Under the constitutions of the va states and the powers granted to state partments of Health and state Depart of Labor, there exists adequate auth for such departments to effectively cob any hazardous condition injurious 1 health. As a practical matter, there ii legal need for the adoption of codes pre; ing maximum allowable concentrations, fair to state that state Departmental Health and state Departments of La| administering divisions of industrial ( giene have enjoyed the confidence of public, including both management labor, and our several state depart! have been and are doing excellent job , the administration of their affairs. Thf fore, the use of maximum allowable eonej trations should be in the nature of sq ing as a guide to industry with parties reference to engineering standards. Indi try has and will continue to seek 1 advice of proper state departments to ci trol hazardous conditions effectively, g the desired objectives will be obtaii through education and dissemination of formation through the cooperation betwi state divisions of industrial hygiene i industry rather than the attempted enfoi ment of a given code as a matter of law. Dr. Brandt and Mr. Stern have discus in detail the practical uses of threshold 1 its with the purpose of eliminating or < trolling occupational hazards. With all the information presently available science, it is seriously to be questio whether the present schedule of reo mended maximum allowances of concen tions are scientifically correct. True, t represent the best thinking of those sc tists who have studied the problem; h ever, it is a fact that from day to day, ini trial processes are changing, new chem compounds are coming into commercial and tomorrow some new problem may presented to industry and to public ministrative agencies entirely different f any that may have been presented bel Industry welcomes technical mfoyma with respect to the existence of occupati hazards. They will continue to welc technical advice as to practical method control. It must be remembered that