of the Act (29 U.S.C.

polymers made from AN except for out regard to the use of respiratory 657) requires employers to notify materials exempted under paragraph protection. Section 6(b)(7) of the Act promptly all employees exposed above (a)(2). (29 U.S.C. 655) mandates that any the PEL. The final standard imple- This proposed definition of "PAN" standard promulgated under section ments section 8(c)(3) by requiring an has been deleted from the final stand- 6(b) shall, where appropriate, provide employer to notify each employee in ard for several reasons. The standard for monitoring or measuring employee writing of the employee's measurewas and is intended to protect employexposure at such locations and interment within 5 working days after the FEDERAL REGISTER, VOL. 43, NO. 192-TUESDAY, OCTOBER 3, 1978 45796 RULES AND REGULATIONS receipt of the results of any required to the ETS, OSHA was most conand precision requirements set forth measurements. cerned about this situation, particulartherein. The testing that OSHA has OSHA recognizes that the standard ly in light of the industry recommenperformed, supplemented by the is intended to set only minimum monidations at that time that the standard NIOSH data, serves to demonstrate toring requirements for employers. It be set as low as 2 ppm. Therefore, testthat there is at least one method availis reasonable, then, that the standard ing was performed at the OSHA Anaable which meets those requirements. should only require the notification to lytical Laboratory in Salt Lake City, We are fully aware that there are employees when monitoring required Utah, to determine whether NIOSH many other methods in use throughby the standard is carried out. Addimethod S-156, which was the method out industry to sample for AN, as well tional actions by the employer in monused under the existing 20 ppm standas methods which are in varying itoring exposures, while certainly enard, could also be used in sampling stages of development. Such methods couraged by OSHA, are not required lower concentrations (see appendix D). include the use of porous polymer by the standard, and direct notifica- The test results clearly showed that tubes instead of charcoal tubes for coltion need not be given to employees. S-156 could be used for sampling and lecting the sample, gas bags, badge- The standard requires that measureanalysis below 1 ppm with sufficient type dosimeters, and the use of portaments be made by monitoring which is accuracy (Exs. 18; 18A). Testing was ble and fixed-point sampling devices representative of each employee's experformed in the presence of high huand systems, Considerable data has posure to AN over and 8-hour period midity (85 percent) and other monbeen submitted to the record by many without regard to respiratory protecomers, such as styrene and butadiene, companies concerning their experience tion. Exposure measurements for each to determine whether these elements with these methods. By demonstrating individual employee would, certainly, had any detrimental effect on the the availability and effectiveness of S- be an indication of that employee's exsampling and analysis. The effects of 156 as modified, we do not intend to posure. However, this may be unnecesthese factors was negligible. The Dishort-circuit any of the efforts by sarily burdensome in some instances. rector of the OSHA Laboratory, Floyd these companies to develop their own Monitoring which can be shown to be Madsen, appeared at the rulemaking monitoring procedures, some of which representative of an employee's expohearing and was questioned at length may prove to be even more accurate. sure would provide the necessary inby industry representatives concerning Industry representatives have formation and in many instances his testing procedures and results. In argued that it is essential for a method would require fewer samples. This is addition, representatives of SPI were to be "field validated" in order to be particularly true in manufacturing of permitted to visit the Laboratory after used for compliance with the AN AN and other 24-hour continuous the hearing. in order to talk with Mr. standard (Exs. 75; 83). Yet it should be processes. Madsen further concerning sampling noted that such representatives have The proposal required monitoring and analysis, in response to a direct realso stated that the range of field conmonthly for employees exposed above quest at the hearing by both SPI and ditions that may be encountered in the permissible exposure limit and Vistron. Although such a visit did, the AN industry is so wide and variquarterly for employees exposed beindeed, go beyond what is required of able that it would be necessary to tween the action level and the PEL. As the Agency during informal rulemak- "field validate" a method in each indinoted above, many employers have ining. OSHA determined that it was, in vidual workplace in order to show that dicated that they perform monitoring this limited circumstance, an appropriit was, indeed, a workable method on a more frequent basis. It is evident ate gesture toward developing a comacross the board (Tr. 933). It is evident that the higher the frequency of meaplete record on the sampling issue. that no single analytical method will surement, the higher the accuracy of Laboratory testing has also been work in every imaginable workplace the employee's exposure profile. The performed by NIOSH at their laborasituation, either for AN or for most companies involved with the manufactories in Cincinnati, also involving any other organic vapor. In addition, turing and polymerizing of AN indicatmethod S-156. Data submitted to the it is also clear that whatever method ed little, if any, disagreement with the record by Dr. Judd Posner of NIOSH an employer may choose to use to monitoring frequency in the proposal. also demonstrated that AN could be monitor for AN, he must validate it OSHA has carried forward these prosampled and analysed below 1 ppm for use in his particular plant. The visions to the final standard. We bewith accuracy (Exs. 43; 66). Since this need for such individual validation was lieve that in the industries affected by data was submitted during the course espoused by industry representatives this standard, monthly and quarterly of the hearing, Dr. Posner invited in- (Tr. 933). monitoring will reflect employee expoterested persons at the hearing to The method of "field validation" sure with sufficient accuracy to assure submit questions to him in writing, to proposed by Dr. Ward of Monsanto that suitable precautions will be taken which he would reply during the postpresents an interesting approach as needed to protect employees from hearing comment period. It is imporwhich shows promise (Ex. 136). Howexposure to AN. tant to note that no such questions ever, there is nothing to indicate that Industry representatives have conwere received. A complete write-up of a method is or has been in use in the centrated much of their criticism on the analytical method and sampling past to "field validate" sampling methwhat they have contended to be the technique used by NIOSH is available ods for AN or other organic vapors. In lack of a method of sampling and analin the record (see appendix D). In esaddition, Dr. Posner of NIOSH has ysis for AN in the range of 1 ppm and sence, it sets forth the same method submitted an analysis of Dr. Ward's below which will meet the accuracy utilized by the OSHA Laboratory, validation method which points up and precision requirements set forth with the major exceptions being the that it, too, is not without pitfalls in in the proposed standard. This issue use of a different desorbing solvent its implementation (Ex. 140). There is was brought to the surface by the and the use of flame ionization withno single accepted method of "field original NIOSH recommendations in out the nitrogen phosphorus detector. validating" a sampling method to September 1977 (Ex. 65). In those rec- OSHA does not require that any one obtain data which will support its use ommendations, NIOSH stated that specific method of sampling and analunder all workplace conditions. OSHA should issue a 4 ppm standard, ysis be used for compliance with the A detailed exposition of the range of based upon what they characterized as monitoring provisions of the standard. available methods of sampling and the limitations of existing sampling This provision is set forth in performanalysis for AN was contained in the and analytical methods for AN. As we ance language, and permits the use of written comments submitted by Monpointed out at length in the preamble any method which meets the accuracy santo (Ex. 11:(27)). Table A contains a FEDERAL REGISTER, VOL. 43, NO. 192-TUESDAY, OCTOBER 3, 1978 RULES AND REGULATIONS 45797 listing of five methods which were recovery (with methanol or carbon dieither validated by Monsanto (field sulfide) falls below 75 percent by deand/or laboratory validation) or, in veloping a "working curve," in lieu of the case of the gasbags, showed a dethe NIOSH protocol. They emphatection limit below 1 ppm. The field sized that appropriate validation techvalidation was, presumedly, carried niques are necessary for the use of all out under the protocol submitted by monitoring procedures in the work- Ward (Ex. 136). Although Monsanto place setting, whether charcoal tubes reported difficulty in using NIOSH or other means of sampling are used. method S-156, it is clear that they They have chosen to rely upon alterhave been able to develop or apply native systems, such as porous polyother methods of sampling and analymer adsorption/thermal desorption, sis to their many and varied opersince they consider them to be less ations. subject to the types of variables associ- Monsanto, in their written comated with charcoal-based systems. ments, discussed the various problems As was noted by Madsen in his testithat they had experienced with mony at the hearing, Monsanto con- NIOSH method S-156 (Ex. 11:(27)). curs that the use of a nitrogen-specific They noted that the only charcoal lot gas chromatographic (GC) detector which had shown satisfactory results has the potential of enhancing the dewas lot 107. This point was noted by tection limit of all of the methods Madsen in his testimony (Tr. 132). In listed in the table by a factor of 10 dealing with the problems of sample (Ex. 11:(27)). It is also important to "breakthrough" reported by various note that other methods which incompanies, including Borg-Warner volve gas chromatographic analysis, (Ex. 83), Monsanto noted that some have this same potential. companies are using two larger char- Monsanto has also reported using coal tubes, containing up to 1,000 mg., both the Century OVA 128 and Cenin series, to replace the 150-mg. tubes tury flare tubes at levels well below 1 generally used. They have concluded, ppm, noting only that "the methods however, that there is no evidence of are claimed to be selective." Century any effect of humidity or temperature Systems testified independently that at low AN burden. their instrumentation was capable of Monsanto noted that charcoal tube measuring AN at concentrations well methods can be used for AN at concenbelow 1 ppm as an 8-hour TWA (Tr. trations below 2 ppm even where AN 873-874). FEDERAL REGISTER, VOL. 43, NO. 192-TUESDAY, OCTOBER 3, 1978 45798 RULES AND REGULATIONS TABLE B: ADDITIONAL METHODS OF MONITORING AND ANALYSIS FOR AN BELOW 1 PPM (Adapted from Monsanto comments (Exhibit (27))) Method Adsorption/Desorption/ Validation: Analysis concentration, detection limits 1. Monsanto Poropak-N field validated down to PoropaK-N (porous Thermal GC 0.1 ppm; detection limit polymer) approaching 0.02 ppm for a 3-liter air sample. 2. Modified 1 to 2% field validated at NIOSH Charcoal Acetone in/GC levels above 0.08 ppm P&CAM 127 CS₂ within ± 2 std. devs. for a TWA6. 3. NIOSH laboratory and field P&CAM 127 Charcoal/CS2/GC validated at and above 0.2 ppm for a TWAS; Reported loss of AN when desorbed sample is stored over time. 4. Passive Dosimetry Abcor R Acetone / GC laboratory validated 0.94 Gasbadge/in CS₂/ ppm; field tested at 0.8-3.8 ppm by reference to method no. 2 above. 5. Gas sample bags Gas bag/GC Detection limit of ca a 2.5 CC. gas sample. Reported memory effects at low AN levels. FEDERAL REGISTER, VOL. 43, NO. 192-TUESDAY, OCTOBER 3, 1978 RULES AND REGULATIONS 45799 Monsanto was confident that "the The monitoring defined by the is certainly one approach toward decharcoal tube or passive dosimeter OSHA Laboratory involves the use of termining representativeness of expomethods could be practiced by downcharcoal tubes and personal sampling sure measurements. There are other stream processors with some guidance pumps, with subsequent analysis by approaches which would be suitable, from trained professional" (Ex. gas chromatograph with nitrogen as well. The employer has the burden 11:(27)). The record reflects that the phosphorus detector. Laboratory valiof showing, by whatever means at his employers covered by this standard dation data indicates that this method disposal, that his measurements are possess considerably more technical can be used with sufficient accuracy representative of each employee's excapability in this area than downand precision down to the 0.5-ppm posure to AN. stream employers would be expected level and below. Floyd Madsen was The rulemaking record, along with to have. The broadening of the exempquestioned extensively at the hearing experience under the ETS, indicate tion provisions will limit the scope of as to his findings by industry reprethat most affected employers have the standard to those companies best sentatives (Tr. 117-159). In addition, successfully implemented a monitorable to provide an effective monitoring several industry personnel visited ing method which gives satisfactory program which will produce meaning- Madsen at the Laboratory during the ful results. results in their operations, at concenposthearing comment period (Ex. 142). trations in the 1- to 2-ppm range. How- Companies such as Goodyear, Dow Neither the questioning at the hearing Badische, and Dow Chemical have ever, OSHA cannot ignore the difficulnor the laboratory visit have raised submitted details of the sampling and ties that various companies have reany significant questions as to Madanalytical methods that they have ported in demonstrating the required sen's findings. found valid for use in their respective accuracy and precision specified in The primary differences between operation (Exs. 138; 11:(17); 11:(25)). both the ETS and proposal at a (95 the NIOSH and OSHA variations on Still others have installed continuous percent confidence level, ±25 percent S-156 are the method of desorption monitoring systems which provide at or above 2 ppm, and ±35 percent (NIOSH does not use ultrasonic agitaarea measurements of airborne AN below 2 ppm) (Exs. 83; 159). We have tion of the sample), the desorbing solconcentrations (Exs. 11:(10); 11:(8)). determined that these criteria should vent (NIOSH use 1 percent acetone in As the record makes clear, there are carbon disulfide, whereas OSHA uses be made more flexible, in order to a number of adequate methods availamethanol), and the analytical equipassist employers who have encounble for sampling and analysis of AN ment (NIOSH uses flame ionization, tered unique sampling problems. concentrations at and below 1 ppm. whereas OSHA adds a nitrogen-phos- Therefore, the final standard permits OSHA recognizes that one method phorus detector). the use of methods with accuracy and may not work under all workplace con- Method S-156 and variations thereprecision at a 95-percent confidence ditions due to such variables as temon have been adopted and accepted for level, of ±35 percent at or above 2 perature, humidity, and other unique general use in field monitoring at the ppm, and ±50 percent below 1 and 2 factors. Regardless of the method se- 1- to 2-ppm level by several major comppm. The difference (35 percent VS. 50 lected, an employer must validate the panies that manufacture and/or use percent) reflects a recognition that acmethod that he intends to use under AN in their facilities. Workplace concuracy of monitoring decreases below the conditions present in the particuditions and other factors vary widely 2 ppm. lar workplace to assure himself that it among these companies (cf. Ex. 2:(22); The accuracy and precision requirewill be sufficiently accurate. It is for 11:(14); 138; Tr. 1196-1199). ments are intended to provide the emthis reason that the monitoring accu- The standard requires that the meaployer with a degree of confidence in racy and precision required by the surements be made by monitoring his sampling results. As noted earlier, standard are stated in performance which is representative of each emmonitoring is not carried out for its terms. The OSHA and NIOSH methployee's exposure to AN over an 8- own sake, but for the purposes of deods of monitoring employee exposures hour period without regard to the use termining what is required for compliare included as an appendix to the of respiratory protection. Exposure ance in a given operation. OSHA anfinal standard as a means of assisting measurements for each individual emticipates that the wide range of availathe employer in the selection of an apployee would, of course, be an indicable methods, combined with more propriate method for his workplace. tion of that employee's exposure. flexible criteria for the method em- It was also contended that because However, this may be unnecessarily ployed, will enable employers to perneither the OSHA nor the NIOSH burdensome in some instances. Moniform their required monitoring withvariations of method S-156 had been toring which is truly representative of out serious difficulties. "field validated" under all workplace an employee's exposure would provide The employee exposure measureinsults and exposure conditions, they the necessary information and in ments are to be made without regard could not serve as reference methods. many instances would involve fewer to the use of respiratory protection. This assertion totally misconstrues the samples. One commentator took issue with this true purpose of a so-called "reference In their testimony, the NIOSH panel provision, contending that the sammethod." The applicability of any expressed concern that OSHA did not pling does not reflect the actual expo- "reference method" was described specify a means of determining if an sure of the employee who is being promost accurately at the hearing by Dr. exposure measurement was "repretected by a respirator (Ex. 149). Ward: sentative" of an employee's exposure. Whereas this statement may be true, There is always a need for a reference They believed that this presented a it overlooks the fact that exposure method. However, the reference method potential for actually reducing the monitoring is not a single-purpose accannot serve under all applications. It can protection afforded to workers extivity. It is necessary to know employserve under laboratory conditions and it has posed to AN. OSHA believes that the ee exposure levels without the use of merit in that respect (Tr. 791). standard should require the desired respiratory protection in order to OSHA and NIOSH have both perperformance, i.e., a showing of repreevaluate the effectiveness of the reformed laboratory testing of their sentativeness, rather than specifying quired engineering and work practice variations of NIOSH method S-156 one method for determining it. We controls, and to determine whether and have found the method to be a recognize, however, that NIOSH has additional controls must be instituted. suitable reference method for use in published a document outlining a sam- In addition, monitoring is necessary to the range of concentrations covered by pling strategy which is directed at this determine which, if any, respirator the standard (Exs. 18; 18A; 43; 66). problem (Ex. 14:(157)). That strategy must be used by the employee. PEDERAL REGISTER, VOL. 43, NO. 192-TUESDAY, OCTOBER 3, 1978 45800 RULES AND REGULATIONS As noted by numerous participants those operations where exposures The alarm system required by parain this rulemaking proceeding, expowould be well below the permissible graph (i) of the standard will detect sure conditions vary throughout the exposure limit. AN concentrations which exceed the day in many operations. At least one The standard requires that whenceiling limit. When an alarm is soundsample is to be taken during each shift ever there has been a production, ed, the employer will have to take in order to assure that exposure meaprocess, or control change which may action to locate the source of the resurements represent exposures of emresult in new or additional exposures lease which is causing the high levels ployees on all shifts. Employees workto AN, or whenever the employer has of AN, and will then have to deconing in the same area doing different reason to suspect an increase in emtaminate the leak or spill area. In this jobs may have different exposures. ployee exposure, the employer shall regard, the weekly survey, as pro- Therefore, the standard requires samrepeat the required monitoring and posed, would be redundant. pling for each job classification, as measurements for those employees af- OSHA has determined, however, well. The proposed standard did not fected by such change or increase. that some form of leak detection procontain detailed procedures in this The redetermination procedure is gram should be required by the standregard. However, in order to assure acnecessary to assure that the most ard. The details of such a program, curate representative monitoring of recent monitoring results accurately however, should be left to the employemployee exposures, OSHA has deterrepresent the existing exposure condier, who is already under the obligation mined that the variability of expotions. This is necessary so that the emto control employee exposures to sures warrants such procedures. ployer may take appropriate actions within the permissible exposure limits. The standard requires that the inisuch as instituting additional engi- The leak detection program in the tial monitoring be conducted, and the neering controls and providing approfinal standard has been placed in the results thereof obtained, as of the efpriate respiratory protection. housekeeping section of the standard, fective date of the standard. OSHA be- The final standard also provides paragraph (k). lieves that this provides more than that employers may discontinue peri- Regulated areas and notification: enough time for employers to comodic monitoring for those employees Paragraphs (d) and (f). The final plete the sampling and obtain the refor whom two consecutive measurestandard requires the employer to essults. Moreover, the standard permits ments, taken at least 7 days apart, tablish regulated areas wherever AN employers who have monitored under show exposures to be below the action exposures exceed the PEL, and to the ETS, or since January 17, 1978, to level. Where employee exposure mealimit access to these areas to authorutilize these measurements for pursurements fall below the permissible ized persons. This requirement will poses of complying with the initial exposure limits but are at or above the help to limit exposure to AN by premonitoring requirements of the final action level, the employer may alter venting persons who are not authorstandard. Employers who have already the monitoring schedule for those emized from entering areas with AN exmonitored their employees' exposures ployees from monthly to quarterly posures above 2 ppm. It will also assist within this period will, therefore, not after two consecutive measurements in designating areas where precautionhave to conduct initial monitoring taken at least 7 days apart indicate ary signs are to be posted, while also unless, because of a process, control, the reduction in levels. designating areas where employees or personnel change, they are required