SECRETARY OF LABOR,
Complainant,

v.

KASTALON, INC.
Respondent.

OSHRC Docket No. 79-3561

SECRETARY OF LABOR,
Complainant,

v.

CONAP, INC.,
Respondent.

OSHRC Docket No. 79-5543


DECISION

Before: BUCKLEY, Chairman, RADER and WALL, Commissioners.
BY THE COMMISSION:


These consolidated cases are before the Occupational Safety and Health Review Commission under 29 U.S.C. § 661(j), section 12(j) of the Occupational Safety and Health Act of 1970, 29 U.S.C. §§ 651-678 ("the Act"). The Commission is an adjudicatory agency, independent of the Department of Labor and the Occupational Safety and Health Administration. It was established to resolve disputes arising out of enforcement actions brought by the Secretary of Labor under the Act and has no regulatory functions. See section 10(c) of the Act, 29 U.S.C. § 659(c).

Kastalon, Inc., and Conap, Inc., manufacture polyurethane products. Among the chemicals they use in the manufacturing process is 4,4' -Methylene bis (2-chloroaniline), which is commonly called "MOCA."[[1]] MOCA has been shown to cause cancer in laboratory animals. The Secretary of Labor contends it must therefore be regarded as a cancer-causing agent ("carcinogen") in humans and that exposure to it must be strictly controlled. The Secretary alleges that Kastalon and Conap violated the Act by failing to take adequate measures to protect their employees from exposure to MOCA. In particular, the Secretary alleges that the employers violated section 5(a)(1) of the Act, the so-called "general duty clause."[[2]] The Secretary also alleges that Conap violated standards governing the provision and use of personal protective equipment. For the reasons that follow, we conclude that the Secretary did not prove that the employers violated the Act as alleged.
I
The Act gives the Secretary two tools by which he can seek elimination of workplace hazards. First, the Secretary can promulgate occupational safety and health standards and issue citations to employers who fail to comply with those standards. 29 U.S.C. § 655 (authority to promulgate standards); 29 U.S.C. § 654(a)(2) (employers are required to comply with standards). Second, where no occupational safety or health standard is applicable, the Secretary can attempt to obtain correction of a workplace hazard through the general duty clause, which requires that employers provide workplaces free of recognized hazards that are causing or are likely to cause death or serious physical harm. The Act's legislative history shows that Congress intended that the Secretary would primarily rely on specific standards, rather than the broad mandate of the general duty clause, to seek the correction of workplace hazards. See S. Rept. No. 91-1282, 91st Cong., 2d Sess. (1970) at 9-10, reprinted in Senate Committee on Labor and Public Welfare, 92 Cong., 1st Sess., Legislative History of the Occupational Safety and Health Act of 1970, at 149-50 (Comm. Print, 1971); remarks of Rep. Steiger on conference bill, Id. at 1217 ("general duty requirement should not be used set ad hoc standards."); Brisk Waterproofing Co., 73 OSAHRC 30/E1, 1 BNA OSHC 1263, 1264, 1973-74 CCH OSHD ¶ 16,345, p. 21,261 (No. 1046, 1973). "Reliance upon the general duty clause was discouraged because to do so would provide little advance warning of what specifically is required in order that employers could maintain a safe and healthful workplace." Id. Thus, the Commission has consistently held that the Secretary cannot rely on the general duty clause when a specific standard applies to the facts. Id.; Daniel International, Inc., 82 OSAHRC 23/D3, 10 BNA OSHC 1556, 1982 CCH OSHD ¶ 26,033 (No. 78-4279, 1982).

No occupational safety or health standard governs MOCA exposure. In 1974, the Secretary did issue a detailed standard regulating workplace use of MOCA, but that standard was ruled invalid because the Secretary had not conformed to the Act's requirements in promulgating it. Synthetic Organic Chemical Manufacturers Association v. Brennan, 503 F.2d 1155 (3d Cir. 1974), cert. denied, 420 U.S. 973, reh'g denied, 423 U.S. 886 (1975).[[3]] In early 1975,the Secretary again proposed the same standard and held a hearing on the proposal. The polyurethane industry, through the Polyurethane Manufacturers Association, a trade association that includes Kastalon and Conap among its members, participated in the hearing and urged the Secretary to adopt a more reasonable standard governing MOCA. However, the Secretary took no action on the proposal. Beginning in 1976, the Secretary began to issue citations alleging that employers were in violation of the general duty clause for not taking adequate precautions against MOCA exposure.[[4]] The alleged violations in Kastalon's and Conap's workplaces occurred in late 1978 and early 1979.

The administrative law judges, James H. Alesia in Kastalon and Mary E. Cerbone in Conap, vacated the section 5(a)(1) citations primarily because they concluded that the Secretary was using the general duty clause to circumvent both the court's decision holding the original MOCA standard invalid and the statutory requirements for promulgating standards. The judges noted that many of the abatement requirements the Secretary was seeking to have Kastalon and Conap implement were taken directly from the invalid MOCA standard,[[5]] and they viewed the citations as an attempt by the Secretary to enforce the invalid standard through the general duty clause.

We are troubled, as were the judges, by the Secretary's apparent attempt to enforce an invalidated standard through citations under the general duty clause.  As noted above, Congress contemplated that the Secretary would place primary reliance on standards to achieve safe and healthful workplaces and envisioned that the general duty clause would be used sparingly. Particularly in a situation like this, where a standard has been proposed and rulemaking proceedings have been conducted, the Secretary's failure to complete the rule-making, coupled with his issuance of citations under the general duty clause, do not promote the goals of "fairness and mature consideration of rules of general application" that the Act's rulemaking provisions were designed to foster.  See NLRB v. Wyman-Gordon Co., 394 U.S. 759, 764 (1969).

On the other hand, the Secretary contends that he did not issue a MOCA standard because he was simultaneously conducting a number of other rulemaking proceedings that had higher priority. The Secretary also maintains that his failure to promulgate a standard should not deny employees any protection that Congress intended them to have under the general duty clause. We recognize that the Secretary has discretion to take into account other rulemaking proceedings in which he is engaged in setting a timetable for issuing a particular standard. Public Citizen Health Research Group v. Auchter, 702 F.2d 1150, 1157-58 (D.C. Cir. 1983). However, we question whether this discretion coupled with the Secretary's aim of protecting employee health justifies the use of the general duty clause in effect to side step the procedural requirements of the Act and impose the requirements of the invalidated MOCA standard on employers. We need not, however, decide whether the Secretary exceeded the limits of his discretion here. Assuming arguendo that the general duty clause citations were properly issued, we conclude that the Secretary failed to prove that Kastalon and Conap violated section 5(a)(1), and we vacate the citations on that basis.

II

Kastalon and Conap both use MOCA as a raw material in the manufacture of polyurethane products. Both companies purchase MOCA in the form of dry pellets about the size of kernels of corn. To manufacture polyurethane, MOCA is first melted and then mixed in proper proportions with another chemical, referred to as a prepolymer, to form liquid polyurethane. The liquid polyurethane is then molded into the shape of the part being fabricated. Employee contact with MOCA can occur when the chemical is removed from its shipping container, when it is heated in order to melt it, and when the molten MOCA is mixed with the prepolymer. Spills of MOCA can also lead to employee contact.

By the time of the alleged violations, Conap had eliminated human contact with MOCA in most of its MOCA-handling operations. Most of the polyurethane products Conap fabricated were made in the following way. A closed drum of MOCA was placed in an enclosed chamber, with access to the MOCA provided through rubber gloves in one side of the the chamber.  A worker would place his hands in the gloves, remove the top of the MOCA barrel, and insert a vacuum tube. The dry MOCA would be sucked through the vacuum tube into a MOCA melter, from which it would be transferred without human contact to an automatic process control ("APC") machine. The melted MOCA would be automatically mixed with the prepolymer in the APC machine, which would then discharge the liquid polyurethane into molds. Conap also used other processes in which the possibility of employee contact with MOCA was present, but it was in the process of automating some of these processes. The company also provided personal protective equipment to minimize employee contact with MOCA.

Kastalon used much less MOCA than did Conap, and its MOCA-handling was not automated, as was Conap's. However, after the MOCA was transferred from its original shipping container, all handling of MOCA was done under a ventilated laboratory hood. Kastalon also provided personal protective equipment to minimize employee contact with the chemical and rotated the employees who handled MOCA, thus limiting the amount of MOCA to which any individual employee was exposed.

Kastalon's and Conap's polyurethane manufacturing operations resulted in employee exposure to MOCA.  MOCA is readily absorbed through the skin, and the handling of MOCA in Kastalon's and Conap's plants resulted in some degree of employee skin contact with the chemical.  Because it has an extremely low vapor pressure, MOCA does not present a hazard of airborne contamination except to the extent that MOCA dust is generated and becomes airborne. The record indicates that certain operations performed in Kastalon's and Conap's plants did generate visible quantities of airborne MOCA dust for brief periods of time. The Secretary's industrial hygienist measured the concentration of airborne MOCA to which two Kastalon employees were exposed. These measurements showed 8-hour time weighted average concentrations of 9.9 and 12 micrograms per cubic meter. Wipe samples of surfaces in Kastalon's plant showed that MOCA was present on the surfaces in amounts ranging from 87 to 1830 micrograms per 100 square centimeters of surface area. In Conap's plant, airborne samples, and wipe samples from employees' skin taken during the inspection yielded no detectable amounts. However, measurements of the concentration of MOCA in the urine of certain of Conap's employees showed that the employees were absorbing measurable amounts of MOCA. These measurements showed concentrations of MOCA ranging from 100 to 620 micrograms per liter of urine.

To prove that an employer violated section 5(a)(1), the Secretary must show (1) that a condition or activity in the employer's workplace presented a hazard to employees, (2) that the cited employer or the employer's industry recognized that hazard, (3) that the hazard was likely to cause death or serious physical harm, and (4) that feasible means existed to eliminate or materially reduce the hazard. Pelron Corp., 86 OSAHRC ____, 12 BNA OSHC 1833, 1986 CCH OSHD ¶ 27,605 (No. 82-388, 1986); Davey Tree Expert Co., 84 OSAHRC 11/D11, 11 BNA OSHC 1898, 1984 CCH OSHD ¶ 26,852 (No. 77-2350, 1984).

The threshold question in this case is whether the Secretary proved that Kastalon's and Conap's employees were exposed to a hazard. The Secretary argues that MOCA must be regarded as a human carcinogen based on experiments in which cancer was induced in laboratory animals, and that any amount of exposure to a carcinogen is hazardous. Although the record does not support a finding that MOCA is a known human carcinogen, it does establish that MOCA may be carcinogenic in man and therefore it should be considered a suspect human carcinogen. Expert witnesses for both sides testified to that effect. The mere fact that a chemical is a suspect human carcinogen, however, does not establish that a hazard exists within the general duty clause of the Act. In order to establish the existence of a hazard, the Secretary cannot simply show that a suspect human carcinogen is present in the workplace or that employees are exposed to it in some detectable amounts. The Secretary must prove that the amount to which employees are exposed presents a significant risk of harm.

Our conclusion that the evidence must establish that a significant risk of harm exists at Kastalon's and Conap's workplaces rests on the Supreme Court decision in Industrial Union Department, AFL-CIO v. American Petroleum Institute, 448 U.S. 607 (1980) ("the Benzene Case"). In the Benzene Case, the Supreme Court concluded that an OSHA standard governing exposure to benzene was invalid. Benzene is a chemical that has long been known to cause nonmalignant diseases in sufficiently large exposures. In order to protect against these diseases, the Secretary adopted a standard in 1971 limiting exposure over an 8-hour time period to an average of 10 parts per million ("ppm").[[6]] 448 U.S. at 617.

In the 1970's, evidence began to accumulate that benzene can cause leukemia as well as the nonmalignant diseases it had previously been known to cause.   The Secretary thereupon initiated notice-and-comment rulemaking proceedings aimed at lowering the permissible exposure limit. He proposed, and ultimately adopted, a standard limiting employee exposure to an 8-hour average of 1 ppm. In reducing the permissible exposure by a factor of ten, the Secretary proceeded on the assumption that no safe limit for exposure to a carcinogen could be established. He believed that the Act required him to establish as a permissible exposure limit the lowest limit that industry could feasibly achieve in the absence of evidence establishing that a higher limit was safe. Concluding that 1 ppm was a feasible limit, the Secretary established that as the permissible exposure limit. 448 U.S. at 623-26.

The Supreme Court concluded that the new standard was invalid because the Secretary had not justified reducing the permissible exposure limit from 10 ppm to 1 ppm. The Court primarily criticized the Secretary's reasoning that the permissible exposure limit had to be set at the lowest feasible level if a safe level of exposure could not be established. The Court noted that Congress did not intend the Act to create risk-free workplaces and it therefore concluded that a precondition to the adoption of a standard under section 6(b) of the Act was a finding that workers were exposed to a significant risk of harm under the conditions currently existing in the workplace. 448 U.S. at 642. In issuing the new benzene standard, the Secretary had not made a finding that benzene posed a significant risk at the preexisting permissible exposure limit of 10 ppm. Instead, the Secretary had found that the evidence presented in the rule-making record did not establish any level as safe, and therefore he presumed that exposure in any amount was hazardous. This finding, under the Court's reasoning, was insufficient to permit a new standard to be promulgated. Id. at 659.

The Court's observation that Congress did not intend for the Act to create risk-free workplaces applies equally to enforcement of the general duty clause as well as to the Secretary's authority to issue standards. Congress intended that the general duty clause would provide protection when no standard had yet been adopted, not provide protection that would go beyond what standards could permissibly provide. Indeed, the general duty clause is more restrictive than standards, for it applies only to hazards that are recognized by an employer or the employer's industry and that are causing or are likely to cause death or serious physical harm. Thus, in order to prove the existence of a hazard within the meaning of the general duty clause, the Secretary cannot merely show that there may be some degree of risk to employees. He must show, at a minimum, that employees are exposed to a significant risk of harm. Cf. Anoplate Corp., 86 OSAHRC, _____,12 BNA OSHC 1678, 1681, 1986 CCH OSHD ¶ 27,519, p. 35,680 (No. 80-4109, 1986) (Secretary must show significant risk when standard requires proof of a hazard as an element of the alleged violation).[[7]]

The Secretary contends that the concept of "significant risk" is already subsumed in the general duty clause in the requirement that a hazard be "recognized." He argues that the recognition requirement "encompasses notions of what workplace conditions reasonably prudent persons believe should be guarded against," and thus assures that conditions that do not present a significant risk will not be subject to section 5(a)(1). Implicit in this argument is the suggestion that any precautions employers take to protect the health or safety of their employees must be directed at a significant risk of harm or the employers would not take them. We do not believe this is a valid premise. Employers may decide to limit the exposure of their employees to chemicals of unknown toxicity or to take other precautions against injury or illness out of an abundance of caution rather than a belief that the absence of such precautions would expose employees to a significant risk of harm. The mere fact that employers take certain precautions does not prove that those precautions are required by an occupational safety or health standard or by the general duty clause. The Commission and courts have consistently held that it would be inappropriate to base a finding that a hazard exists solely on evidence that an employer has taken certain precautions because to do so would dissuade employers from taking voluntary protective measures beyond those the law requires. General Motors Corp., GM Parts Division, 84 OSAHRC 23/A2, 11 BNA OSHC 2062, 2066, 1984 CCH OSHD ¶ 26,961, p. 34,611 (No. 78-1443, 1984). aff'd, 764 F.2d 32 (1st Cir. 1985); Cotter & Co. v. OSHRC, 598 F.2d 911, 915 (5th Cir. 1979); Diebold, Inc. v. Marshall, 585 F.2d 1327, 1338 (6th Cir. 1978); Cape & Vineyard Division v. OSHRC, 512 F.2d 1148, 1154 (1st Cir. 1975). Accordingly, we reject the Secretary's argument that the "recognition" element of section 5(a)(1) eliminates the need for him to show that employees are exposed to a significant risk of harm in order to show that a hazard exists within the meaning of the general duty clause. We turn to whether the Secretary met his burden of proof in this case.

III

MOCA is a member of a family of chemicals referred to as aromatic amines. Several members of this group are known to cause cancer in humans. To date, there is no evidence directly linking MOCA exposure to cancer in humans.[[8]] However, there is typically a long latency period, often 30 to 40 years, between human exposure to a carcinogen and the onset of cancer. Therefore, the absence of direct evidence that a chemical is a human carcinogen, particularly as when here the chemical has not been in use for a long period of time, is inconclusive evidence that the chemical is not a human carcinogen.

Because of the long latency period between human exposure to a carcinogen and the onset of cancer, humans may be exposed to a chemical for many years before it can be definitively established that the chemical is carcinogenic. See, e.g., Society of the Plastics Industry v. OSHA, 509 F.2d 1301, 1305-06 (2d Cir. 1975) (vinyl chloride, which began to be used in 1939, was found to be a carcinogen in the early 1970's). In order to determine whether a chemical may be a human carcinogen before cases of cancer occur in humans, scientists perform experiments in which they determine whether laboratory animals exposed to the chemical contract cancer. Their supposition is that chemicals that induce cancer in small mammals are also likely to be human carcinogens.

MOCA has been the subject of several laboratory experiments in which the chemical induced cancer in three species of laboratory mammals: rats, mice, and beagles. The beagle study. performed by du Pont, demonstrated that beagles who ingested MOCA contracted bladder cancer.[[9]] Expert witnesses for both sides, Dr. John D. Scribner for the Secretary and Dr. George T. Bryan for the employers, agreed that the induction of bladder cancer in dogs strongly indicated that MOCA would also cause bladder cancer in humans. Dr. Scribner noted that other aromatic amines -- benzidine, beta napthalamine, and 4 amilnobiphenyl -- were known to cause bladder cancer in both dogs and humans. Dr. Bryan, a qualified expert in carcinogenicity, had been engaged by the Polyurethane Manufacturers Association ("PMA") to evaluate the available data regarding MOCA, to advise its members on how to deal with MOCA, and to assist it in the rule-making proceedings initiated by the Secretary. He believed that MOCA should be strictly regulated based on the studies showing that it induced cancer in laboratory animals. He stated that "no level of exposure to a carcinogenic substance, however low it may be, can be scientifically established . . . to be a 'safe level' for humans."

The Secretary contends that these statements by Dr. Bryan constitute evidence that an expert familiar with the polyurethane industry recognized that MOCA is hazardous in any detectable exposure. However whether Dr. Bryan's statements show that the polyurethane industry "recognized" a hazard is not our initial inquiry. The first question we must address is whether the evidence shows any hazard at all; only then does the question of whether the hazard is "recognized" arise. Dr. Bryan's statement that no level of exposure to a carcinogen can be scientifically established as "safe" does not prove that Kastalon's and Conap's employees were exposed to a significant risk of harm under the conditions present in their workplaces.[[10]] The central point of the Benzene Case is that a significant risk at any level of exposure cannot be presumed simply because a "safe" level of exposure cannot be established. 448 U.S. at 652-53. The Court rejected the Secretary's policy that, "in the absence of definitive proof of a safe level, it must be assumed that any level above zero presents some increased risk of cancer." Id. at 635-36 (emphasis by the Court, footnotes omitted). The Secretary cannot impose this discredited policy on an industry under the general duty clause simply by finding an expert familiar with the industry to testify that a safe level cannot be established for a carcinogen used by the industry. The absence of a "safe" level does not mean that there is a "significant risk" of harm at any particular level. As the Court made plain, risk-free workplaces are not required by the Act.

The Secretary also argues that industry recognition of a hazard should be predicated on advice given by Dr. Bryan to the PMA that its members should "insure" that employees are not exposed to detectable quantities of MOCA. This is simply the "no safe level" argument in another guise. Moreover, we do not agree with the Secretary's interpretation of the contents of the letter Dr. Bryan wrote to the PMA. The relevant portion of the letter states:

Tumors of the bladder frequently are produced in animals fed chemicals of the class called aromatic amines. Representative chemicals of this class are 2-naphthylamine, 4-aminobiphenyl, and benzidine. Each of these chemicals are recognized as human urinary bladder carcinogens, and each has caused urinary tract tumors in dogs when fed. These tumors are qualitatively similar to those produced by MOCA in dogs. Finally, MOCA is a member of the aromatic amine class of chemicals.

I believe that these data will have a substantial impact on the MOCA regulation proceeding. They demonstrate that MOCA possesses the same carcinogenic potential as other recognized human carcinogens. The inference that most will draw is that individuals working with MOCA are at definite risk of developing cancer. This inference might be blunted somewhat by a demonstration that under conditions of working with MOCA potentially exposed persons do not incorporate MOCA into their bodies, or excrete it or any of its metabolites into body fluids in detectable quantities. Additional evidence desirable in support of the view that MOCA does not represent an actual hazard would be a demonstration that no detectable quantities of MOCA are present in the work environment under conditions of usual usage. If these are not possible or feasible, then MOCA must be dealt with as a known animal carcinogen and as a potential human carcinogen.

This letter reiterates Dr. Bryan's consistently expressed view that MOCA should be regarded as a suspect human carcinogen based on its carcinogenic activity in animals. Although it says that a showing that employees are not exposed to detectable quantities of MOCA would negate any inference that a hazard exists, Dr. Bryan did not state that a hazard exists whenever there are detectable levels of MOCA. Also, contrary to the Secretary's argument, the letter does not advise employers that employees should not be exposed to any detectable quantity of MOCA.

The Secretary also argues that industry recognition of a hazard was shown because industry generally recognized the need to take precautions against MOCA exposure. He particularly relies on recommended precautions for MOCA handling published by du Pont, which was then the chemical's primary manufacturer. Du Pont included with each barrel of MOCA it sold a booklet entitled "Safe Handling of MOCA Urethane Curative." This booklet recommended that MOCA users take certain precautions the Secretary alleges Kastalon and Conap failed to take. The Secretary also points out that the polyurethane industry, through the Polyurethane Manufacturers Association, was urging both the federal government and the State of California to adopt MOCA standards and was informing its members, through semiannual meetings and in publications, of the need to control MOCA exposures. These PMA activities, the Secretary argues, show that the polyurethane manufacturing industry recognized the need for controlling exposure to MOCA. The Secretary further argues that hazard recognition is shown by evidence that individual companies in the polyurethane industry took certain precautions when handling MOCA.

The Secretary's argument demonstrates that the evidence of MOCA's carcinogenicity is sufficiently compelling that reasonable people recognize it is necessary to handle the chemical carefully. That evidence does not, however, show that the levels of MOCA to which Kastalon's and Conap's employees were exposed presented a significant risk of harm. An industry's general recognition that a chemical may be carcinogenic in man and that exposure to it should be controlled does not eliminate the need for the Secretary to prove that a particular employer's employees are exposed to dangerous levels in order to prove a violation of the general duty clause.[[11]]

IV

In order to establish a significant risk in a section 5(a)(1) case involving a carcinogen, the Secretary must show the probability that employees will contract cancer under the conditions present in the workplace. This is consistent with the Supreme Court decision in the Benzene Case. There the Court ruled that the Secretary must establish a significant risk exists through a body of reputable scientific thought. It noted that animal studies, epidemiological evidence and worker mortality rates could be used to establish the existence of a significant risk. 448 U.S. at 656 & n. 64.

The only evidence in the record which tends to show the probability of employees contracting cancer from exposure to MOCA is a quantitative risk assessment prepared by Dr. John D. Scribner, an expert in chemical carcinogenesis. A quantitative risk assessment is an estimate of the probability that humans will contract a disease. It is often based on evidence gathered from animal tests.[[12]]

Dr. Scribner's quantitative risk assessment relied heavily on evidence gathered in the du Pont study that showed MOCA induced cancerous tumors in beagles. In that study, six dogs were fed MOCA in their daily diet in an amount equal to 14 milligrams per kilogram of body weight. One dog died early in the study. The remaining five contracted lower urinary tract cancers--four bladder tumors and one tumor of the ureter--after 8 to 9 years. Of a group of six control dogs, none contracted cancer within that same time span. Dr. Scribner calculated that each exposed dog had received a total lifetime dose of 14 grams of MOCA per kilogram of body weight, and he observed that this dose produced a 100% chance of contracting cancer, as all the exposed dogs had done so.

Dr. Scribner then attempted to correlate this information with evidence of human exposures. He noted that a MOCA rat study at the National Institute for Occupational Safety and Health showed that the rats excreted in their urine about 0.03% of the MOCA they consumed, and that this was approximately the same as the percentage of benzidine that a human excreted when benzidine was applied to his skin. Dr. Scribner therefore estimated that the total amount of MOCA to which a person was exposed in a day would be about 3000 times the amount the person excreted in his urine that day. Assuming that a worker excreted 50 micrograms of MOCA per day in his urine and worked 250 days per year, Dr. Scribner calculated that the worker would receive a total dose of 14 grams per kilogram of body weight after 39 years. As this was the total dose at which the beagle dogs exhibited a 100% incidence of tumors, Dr. Scribner postulated that the worker would be at a 100% risk of contracting cancer after 39 years. To buttress his estimate, Dr. Scribner noted that MOCA and benzidine were about equally potent at producing cancer in dogs, and that a study had shown that 50% of workers had contracted bladder cancer after a total exposure to benzidine of 5 grams per kilogram of body weight. Assuming that MOCA and benzidine proved to be equally potent in humans, as they were in dogs, the human experience with benzidine demonstrated that workers exposed to similar amounts of MOCA as benzidine would experience a significant risk of contracting cancer.

The employers' expert, Dr. Bryan, criticized the central assumption underlying Dr. Scribner's risk assessment; that the quantitative response of humans and dogs to MOCA would be equal. Dr. Bryan believed that quantitative estimates of the risk to humans from animal responses to a carcinogen could not be made with any reasonable degree of scientific certainty. He noted that it was difficult to obtain the same quantitative response even in different experiments with animals of the same species due to the many variables that enter into such experiments. It was even more difficult to extrapolate from laboratory animals to humans. Dr. Bryan stated, for example, that the animals used in cancer experiments are genetically similar (homogeneous) while a typical group of exposed workers would be genetically mixed (heterogeneous). Because an animal's or human's genetic structure greatly affects its response to a carcinogen, one cannot assume that a heterogeneous group of humans will respond quantitatively in the same manner as a homogeneous group of laboratory animals.

Dr. Bryan was also critical of Dr. Scribner's assumptions that the carcinogenic potential of MOCA in humans could be equated to that of benzidine. Dr. Bryan noted that humans are typically exposed to a variety of substances while laboratory tests on animals attempt to isolate the effect only one chemical produces in the animal. He postulated that the workers who had contracted bladder cancer after exposure to benzidine had also been exposed to a number of other chemicals and that the high incidence of bladder cancer should be attributed to the mixture of exposures and not solely to benzidine exposure. Furthermore, if MOCA was as potent a carcinogen as benzidine, Dr. Bryan was of the opinion that human cases of bladder cancer would already have begun to appear among MOCA workers. The absence of any such cancers, as shown by the study of du Pont workers in which no adverse health effects attributable to MOCA were found, indicated that Dr. Scribner had overestimated the risk of bladder cancer in humans from MOCA.

Dr. Bryan's reasons for believing that extrapolations from animal experiments to estimate human risks were not valid within a reasonable degree of scientific certainty largely correspond to the reasons expressed by the Secretary for rejecting quantitative risk assessment in a rulemaking directed at developing a general policy for regulating carcinogens. That policy, which the Secretary issued on January 22, 1980, examined the considerations relevant to the regulation of carcinogens. "Identification, Classification and Regulation of Potential Occupational Carcinogens," 45 Fed. Reg. 5002 (1980). One of the principal issues considered in the rulemaking was whether a substance's carcinogenic potency should be considered in establishing a permissible exposure level, i.e., whether quantitative risk assessment could reliably estimate the risk presented to workers by a carcinogen. After receiving numerous expert opinions on both sides of the question, the Secretary rejected such an approach, saying:

Extrapolation from animal data to predict risks in humans introduces many additional uncertainties. These include selection of appropriate scaling factors for size, lifespan, and metabolic rate; differences in routes of exposure, duration and schedule of exposure, absorption, metabolism, and pharmacokinetics; differences in intrinsic susceptibility and repair capabilities; intra-population variation in susceptibility; and exposure to other carcinogens and intrinsic and extrinsic modifying factors. At least theoretically, these factors can affect the relative response of humans and animals by many orders of magnitude.[[13]]

Id. at 5200. Thus, the Secretary has concluded that quantitative extrapolations from animal data are too speculative to provide useful estimations of human risk.

The number of assumptions made by Dr. Scribner to extrapolate human risk from the animal studies on which he relied renders Dr. Scribner's conclusions too speculative for purposes of section 5(a)(1). Moreover, some of the studies relied upon in making the quantitative risk assessment involved exposure to benzidine rather than MOCA. Dr. Bryan's testimony demonstrates that conclusions as to the carcinogenicity of one chemical cannot readily be drawn from experience with a different chemical. Dr. Scribner's risk assessment was based on the assumption that the exposed workers would excrete 50 micrograms of MOCA a day for 250 days per year. However, this assumption is based on the urine tests of only a few Conap employees on a single day. As to Kastalon, the record is silent on the amount of MOCA in the employees' urine. We are therefore unable to correlate Dr. Scribner's risk assessment to the levels of MOCA to which the workers were exposed. Nor did Dr. Scribner take into consideration the information obtained from the du Pont study which showed no increase in the incidence of bladder cancer to du Pont workers exposed to MOCA. We therefore conclude that the quantitative risk assessment presented by the Secretary is too speculative to establish that Kastalon's and Conap's employees were exposed to a significant risk of harm within the meaning of section 5(a)(1) of the Act. Accordingly, we vacate the citations issued to Kastalon and Conap alleging violations of the general duty clause.

Because we conclude that the Secretary failed to prove that Kastalon's and Conap's employees were exposed to a hazard, we need not and do not decide whether the Secretary proved the remaining elements of a general duty clause violation.

V

We now turn to the Secretary's allegations that Conap violated standards governing the provision and use of personal protective equipment.

As noted above, the primary means by which MOCA can enter the body is through skin absorption. MOCA is only present as an air contaminant if handling of the chemical generates airborne dust. In the personal protective equipment citations issued to Conap, the Secretary alleges that the company did not provide and/or require employees to use equipment that would minimize skin contact and inhalation of MOCA. The standards cited by the Secretary, 29 C.F.R. §§ 1910.132(a) and 1910.134(a)(2), provide as follows:

§ 1910.132 General requirements.

(a) Application. Protective equipment, including personal protective equipment for eyes, face, head, and extremities, protective clothing, respiratory devices, and protective shields and barriers, shall be provided, used, and maintained in a sanitary and reliable condition wherever it is necessary by reason of hazards of processes or environment, chemical hazards, radiological hazards, or mechanical irritants encountered in a manner capable of causing injury or impairment in the function of any part of the body through absorption, inhalation or physical contact.

§ 1910.134 Respiratory protection.

(a) Permissible practice.

* * *
(2) Respirators shall be provided by the employer when such equipment is necessary to protect the health of the employee. The employer shall provide the respirators which are applicable and suitable for the purpose intended. The employer shall be responsible for the establishment and maintenance of a respiratory protective program which shall include the requirements outlined in paragraph (b) of this section.

The Secretary alleges that Conap violated these standards because employees in various areas of the plant were not using impervious gloves, impervious suits, shoe covers, goggles, head coverings, and appropriate respirators. One of the citation items also alleges that Conap violated section 1910.132(a) because MOCA was present on surfaces of a piece of equipment. In the Secretary's view, that means the equipment was not maintained in a sanitary or reliable condition, as the standard requires.

We vacate these citation items for the same reason we vacate the general duty clause citations: the Secretary failed to prove that Conap's employees were exposed to a hazard. Section 1910.132(a) requires that personal protective equipment be provided and used "wherever it is necessary by reason of hazards of processes or environment," and section 1910.134(a)(2) mandates that employers provide respirators "when such equipment is necessary to protect the health of the employee." Thus, like the general duty clause, the Secretary must show that a significant risk to employees is present as a necessary element of proof that an employer violated either standard. See Donovan v. General Motors Corp., 764 F.2d 32 (1st Cir. 1985) (significant risk test for proof of a hazard applies under section 1910.132(a)); Modern Drop Forge Co. v. Secretary of Labor, 683 F.2d 1105, 1114-15 (7th Cir. 1982) (significant risk test applies where standard does not presume existence of a hazard); Anoplate Corp., 12 BNA OSHC at 1681, 1986 CCH OSHD at p. 35,680 (where standard requires proof of a hazard, Secretary must show a significant risk to employees exists). As the Secretary failed to prove that MOCA presented a significant risk to employees at the exposure levels present in Conap's plant, it follows that he failed to prove that Conap violated the personal protective equipment and respirator standards.

Although we vacate these items due to lack of proof of a hazard, we also note that there are other deficiencies in the Secretary's evidence as to certain items. In item 8 of the citation, the Secretary alleged that one employee wore an unapproved respirator and another wore no respirator at all. However, none of the measurements made by the Secretary showed any detectable levels of airborne MOCA, so even assuming that exposure to some amount of MOCA is hazardous, the Secretary failed to prove the need for respirators.[[14]]

Conap also provided various items of protective equipment to minimize skin contact with MOCA, including polyethylene gloves, full-body impervious paper uniforms, and paper hats. Use of this equipment was left to the discretion of employees. The Secretary alleged in items 7(b), (d), and (e) of the citation that Conap violated section 1910.132(a) by not requiring that the equipment be used in various areas of the plant. The Secretary also alleged in items 7(a) and (c) that Conap violated section 1910.132(a) because employees were not using goggles or other eye protection and were not wearing shoe covers to prevent spilled MOCA from being tracked throughout the plant.

Some of these allegations involve employees who were not exposed to MOCA. For example, in subitems 7(b)(3), (c)(2), and (e)(2), the Secretary alleged that employees in the room containing the Automatic Process Control machines ("the APC room") were not wearing impervious gloves, shoe covers, and impervious clothing. However, employees did not handle MOCA in the APC room. The MOCA was mixed with the prepolymer in the APC machine before the mixture was discharged from the machines. Employees could contact the mixture after it came out of the machines, but there is no evidence that this would be harmful. To the contrary, Robert Turkington, an industrial hygienist who testified for the Secretary, stated that employees could not absorb MOCA by contacting the mixture of MOCA and prepolymer. Thus, the Secretary failed to show that there was any possibility of employee absorption of MOCA in the APC room.

The Secretary also alleged in item 7(b)(1) a violation of section 1910.132(a) because employees in the belts room did not wear gloves all the time.   Employees weighed and melted MOCA in the belt room, but they also performed other operations there.[[15]] Although the record shows that employees in the room did not wear gloves all the time, that does not prove that employees failed to wear gloves when they handled MOCA.

The Secretary alleged in item 7(a) that employees did not wear eye protection in two areas where MOCA was handled. However, there is no evidence that MOCA presented any special danger to the eye. The precautions recommended by du Pont, on which the Secretary places so much emphasis, do not mention the need for eye protection. Thus, the Secretary failed to show any hazard necessitating eye protection.

The Secretary alleged in item 7(c) that Conap violated section 1910.132(a) because employees did not wear shoe covers in the plant areas where MOCA was used. According to the Secretary, employees should have worn shoe covers in rooms where MOCA was used and should have discarded them before leaving the room, thus preventing MOCA from being tracked into areas of the plant where MOCA was not used. The Secretary did not show, however, that such a practice would significantly reduce employee exposure to MOCA. There is no evidence that MOCA was in fact being tracked throughout Conap's plant. Also, we assume that all employees wore shoes, so even if MOCA was present on the floors, there would appear to be little, if any chance, it could be absorbed through the skin of employees. Even under the Secretary's argument that any measurable exposure to MOCA presents a hazard, the Secretary failed to prove the need for shoe covers.

The Secretary alleged in item 6 that Conap failed to maintain a glove box in a sanitary and reliable condition. As described earlier, the glove box is one of the devices Conap used to minimize employee contact with MOCA. It is an enclosed chamber into which a closed container of MOCA is placed and manipulated through rubber gloves installed on the side of the chamber. Wipe samples taken during the inspection of Conap's plant revealed 3.6 micrograms of MOCA inside one glove, 9.9 micrograms inside the other. and 460 micrograms on top of the glove box. The Secretary argues that these measurements show that the glove box was not maintained in a sanitary condition. We do not agree. The drums of MOCA used by Conap each contained 132 pounds of the chemical. Translated into micrograms, each drum contained approximately 60 billion micrograms.[[16]] To find a violation based on the presence of a few micrograms, or even a few hundred micrograms, would impose on employers the duty to eliminate all traces of a chemical from all surfaces employees may contact. Such a requirement is simply unrealistic. The purpose of the glove box was to minimize employee contact with MOCA, and the fact it did not eliminate such contact entirely does not show it was not maintained in a sanitary condition.

VI

The evidence in these cases strongly suggests that MOCA is a human carcinogen. Certainly it would be preferable not to be exposed to a carcinogen than to be exposed to even a small amount. However, many substances to which people are exposed daily are known carcinogens, and the Benzene Case recognizes that human activity cannot be guided by the sole objective of totally removing exposure to carcinogens. Instead, a reasoned judgment must be made as to whether the amounts to which people are exposed present a significant risk of harm. In these cases, the Secretary alleges that Kastalon and Conap violated the general duty clause and certain general standards because their employees were exposed to MOCA. The Secretary has not, however, proven by a preponderance of the evidence of record that the amounts of MOCA to which Kastalon's and Conap's employees were exposed were hazardous. We are therefore compelled to conclude that the Secretary has failed to sustain his burden of proof and to vacate the citations.

From a broader perspective, to find violations here would be beyond the limited adjudicatory role of the Commission. We would effectively be establishing a permissible exposure limit through the adjudicatory process without regulated industries having the opportunity for input under the Act's notice and comment rulemaking procedures. While this may be appropriate where appropriate exposure limits have been firmly and scientifically established, see, e.g., American Smelting and Refining Co. v. OSHRC, 501 F.2d 504 (8th Cir. 1974), it is not desirable where it requires "findings . . . on the frontiers of scientific knowledge."

Benzene Case, 448 U.S. at 656. Such findings are uniquely suited to rulemaking, in which the Secretary can "make a policy judgment as to whether a specific risk of health impairment is significant in terms of the policy objectives of the statute." Id. at 663 (Burger, C.J., concurring). Moreover, in a rulemaking proceeding, all interested persons have the opportunity to be heard, and a rule gives employers notice of what they must do to provide safe and healthful workplaces. If the Secretary believes that MOCA presents a significant risk in the workplace, he retains the authority under section 6(b) of the Act to complete the rulemaking he has already started.

Finally, we note that Kastalon contends that the costs it would incur if the Secretary's abatement order is enforced would render it uncompetitive within its industry and thus put it out of business. Although it is not necessary to reach the merits of this contention, we mention it here because it reinforces the importance of proceeding through rulemaking instead of the general duty clause where a decision can be expected to have broad impact within an industry or industries. As the Court of Appeals for the District of Columbia observed:

[I]f adoption of the precaution would clearly threaten the economic viability of the employer, the Secretary should propose the precaution by way of promulgated regulations, subject to advance industry comment, rather than through adventurous enforcement of the general duty clause.

National Realty & Construction Co. v. OSHRC, 489 F.2d 1257, 1266 n.37 (D.C. Cir. 1973).

The judges' decisions are affirmed and the citations are vacated. FOR THE COMMISSION Executive Secretary

DATED: JUL 23 1985


The Administrative Law Judge decision in this matter is unavailable in this format. To obtain a copy of this document, please request one from our Public Information Office by e-mail ( lwhitsett@oshrc.gov ), telephone (202-606-5398), fax (202-606-5050), or TTY (202-606-5386).


 

 


FOOTNOTES:

[[1]] "MOCA" is a registered trade name of E.I. du Pont de Nemours & Co., Inc., which at one time was the chemical's primary manufacturer. Du Pont no longer manufactures MOCA.

[[2]] Section 5(a)(1), 29 U.S.C. § 654(a)(1), provides:
Each employer shall furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.

[[3]] Section 6(b) of the Act, 29 U.S.C. § 655(b), provides that the Secretary may promulgate occupational safety and health standards following rulemaking proceedings in which interested persons are given notice of the proposed rule and an opportunity to be heard. These notice-and-comment rulemaking provisions are similar to those in 5 U.S.C. § 553, section 4 of the Administrative Procedure Act, 5 U.S.C. § 551 et seq. ("the APA"), but impose additional requirements not found in the APA. Among those additional requirements are provisions governing the use of advisory committees. The Secretary may convene such a committee to advise him regarding a standard under consideration. Where he does so, he must publish a proposed standard within 60 days of receiving the committee's recommendations. Interested persons are given 30 days to submit comments and request a hearing. In the case of the MOCA standard, the Secretary published the proposed standard before receiving the advisory committee's recommendations, and he held a hearing on the standard less than 30 days after receiving the recommendations. The court held that this procedure did not conform to the statutory requirements because it did not give interested persons an adequate opportunity to review the committee's recommendations before submitting comments on the standard and participating in the hearing.

[[4]] The record indicates that two companies in addition to Kastalon and Conap were cited. A 1976 citation issued to Roadway Tire & Rubber Company was resolved in a settlement agreement. A citation issued in 1979 to Nazar Rubber Company was vacated by an administrative law judge, and the Secretary did not seek review of that decision. Nazar Rubber Co., 81 OSAHRC 52/D13, 9 BNA OSHC 1960, 1981 CCH OSHC ¶ 25,437 (No. 79-4005, 1981).

[[5]] This was particularly true in Kastalon, where the abatement requirements listed by the Secretary were largely taken verbatim from the invalidated MOCA standard.

[[6]] The Secretary adopted this standard under section 6(a) of the Act, 29 U.S.C. § 655(a), which required him to summarily promulgate national consensus standards and established federal standards as occupational safety and health standards within two years of the Act's effective date without notice-and-comment rulemaking proceedings. The OSHA standard was originally adopted as a standard by the American National Standards Institute in 1969.

[[7]] We note that the United States Court of Appeals for the Fifth Circuit has held that the significant risk test should not be applied in enforcing the general duty clause. Kelly Springfield Tire Co. v. Donovan, 729 F.2d 317, 323-24 (5th Cir. 1984). We respectfully disagree with that court's conclusion. The Fifth Circuit reasoned that the Benzene Case dealt with the promulgation of standards and was therefore of limited relevance in the different context of the general duty clause. We believe, however, that the Supreme Court's conclusion that the Act was not intended to create risk-free workplaces applies equally to the enforcement of the general duty clause as to the promulgation of standards. As we have stated, Congress did not intend for the general duty clause to provide broader protection than could be achieved through the promulgation of standards. Thus, a limitation on the Secretary's authority to issue standards necessarily also limits the scope of the general duty clause.

[[8]] The only record evidence regarding a specific study of human exposure to MOCA involves a study by du Pont. Du Pont began development work with MOCA in 1954 and commenced full-scale commercial production in 1962. Between 1967 and 1971, Du Pont studied the medical records of all its employees who had been involved with MOCA and found no adverse health effects attributable to MOCA.

[[9]] In the du Pont study, six dogs were fed MOCA in their daily diet. One died early in the experiment from unrelated causes, and the remaining five all ultimately contracted cancer. None of the six dogs in the control group contracted cancer, indicating that MOCA had induced the cancers in the exposed group. See note 12 infra.

[[10]] The Secretary cites National Realty & Construction Co. v. OSHRC, 489 F.2d 1257, 1265 n. 32 (D.C. Cir. 1973), for the proposition that a hazard is recognized within the meaning of the general duty clause if "safety experts who are familiar with the circumstances of the industry or activity in question" would have recognized the existence of a hazard. The court made this statement in the context of holding that the general duty imposed on employers by section 5(a)(1) was an achievable one, and that the test for recognition of the hazard was therefore an objective one. The court stated, in full, "The standard would be the common knowledge of safety experts who are familiar with the circumstances of the industry or activity in question." Id.

In this case, it is very doubtful whether Dr. Bryan's opinions can be said to represent the "common knowledge" of the polyurethane industry. Dr. Bryan did not work in the industry, but was retained by the Polyurethane Manufacturers Association to provide it with expertise the industry did not possess. In general, we think it would be unrealistic to hold an industry to the standard of knowledge of an outside expert it engages to provide it with specialized expertise. However, because we conclude that the evidence does not establish the existence of a hazard, we need not decide whether Dr. Bryan's recognition of a hazard would be sufficient to find that the polyurethane industry recognized the hazard.

[[11]] We also note that the general knowledge that a chemical is a suspect human carcinogen and should be controlled does not equate with recognition that exposure to any amount of the chemical is hazardous within the meaning of the general duty clause. We, therefore, do not agree with the Secretary that the polyurethane manufacturing industry recognized the existence of a hazard at any detectable level of MOCA. Additionally, we note that Kastalon and Conap took considerable steps to reduce the exposure of their employees to MOCA. No employees had been known to suffer ill effects from working with MOCA. There is no evidence that Kastalon and Conap believed the measures they took were inadequate or that they recognized that a hazard existed at their worksites.

[[12]] In a typical animal experiment, a certain number of laboratory animals are exposed to a measured quantity of a chemical while a control group of similar animals is maintained in a similar fashion but is not exposed to the chemical. If a larger number of exposed animals than control animals contract a disease and the disparity is statistically significant, it can be concluded that the chemical causes the disease. The statistical probability that an animal will contract the disease at the exposure level used in the experiment can also be calculated. If the experiment is conducted with different groups of animals exposed at several different levels, the degree to which the probability of contracting the disease varies with the exposure level -- the "dose-response curve" -- can be obtained. In order to extrapolate this quantitative animal data to determine the risk the chemical presents to humans, a number of assumptions must be made. First some assumption must be made about the relative sensitivity to the chemical exhibited by humans and the experimental animals. Assumptions must also be made about differences in conditions under which humans and laboratory animals are exposed. For example, the experiment may involve animal ingestion of the substance, but human exposure may typically be through inhalation. In such a situation, an assumption must be made about the relative risk posed by the different exposure routes. Also, animal experiments often involve exposure levels that are much higher than humans typically encounter in the workplace. In order to estimate the risk to humans from the data gathered in such experiments, assumptions must be made about the relative risk at high and low dosages. One possible assumption, for example, is that the risk is directly proportional to the dosage--in technical terms, that the "dose-response curve" is a straight line. Experience shows, however, that it is extremely difficult to predict the risk to humans at low dosages from the results of information gathered at high dosages, for substances that are toxic at high dosages are often relatively less harmful or even completely safe at the typical dosages workers receive.

[[13]] The Secretary noted that estimates of the probability that humans could develop cancer from saccharin ingestion calculated by different scientists assuming different models for cancer induction varied by seven orders of magnitude, i.e., by a factor of ten million. (One order of magnitude represents a tenfold difference in the estimated risk; two orders of magnitude represents a hundredfold difference). 45 Fed. Reg. at 5196-5200.

[[14]] The Secretary also alleged in item 9 that Conap did not establish and maintain a respiratory protection program because it did not train employees in the use of respirators and did not have written respirator instructions. However, absent evidence of a hazard requiring the use of respirators, an employer cannot be found in violation for failing to implement a respirator program. Gulf Oil Corp., 83 OSAHRC 21/A2, 11 BNA OSHC 1476, 1480, 1983 CCH OSHD ¶ 26,529, p. 33,819 (No. 76-5014, 1983).

[[15]] Much of the Secretary's evidence pertaining to the belt room was directed at an operation in which the liquid mixture of MOCA and prepolymer was poured over a mandrel on a lathe to form polyurethane belts. The evidence establishes that this was a "very sloppy" process and resulted in considerable spillage. However, as noted above, the evidence shows that there was no hazard of MOCA contact once the MOCA was mixed with the prepolymer. Thus, the evidence of the sloppiness of the belt-forming operation is irrelevant to whether there was a hazard of MOCA exposure in the belt room.

[[16]] One pound equals 454 grams. Each gram contains one million micrograms. Thus, the number of micrograms in a drum is 132 multiplied by 454 multiplied by one million.