Congress established the Occupational Safety and Health Review Commission (OSHRC) to keep OSHA within the bounds of the law. Creation of the commission was a key compromise that made possible the passage of the Occupational Safety and Health Act in 1970. But for the compromise to work, and for employers to be protected against the inevitable excesses of an enforcement-minded OSHA, the commission must do its job with a critical eye.
Unfortunately, OSHRC has not been sufficiently critical with respect to a key provision of OSHA's Lockout Standard: the Minor Servicing Exception. The OSHRC errors in interpreting the exception and in determining whether it is unconstitutionally vague has caused the exception and, with it, the Lockout Standard as a whole, to be applied irrationally and unpredictably. The judges on the commission even have applied it in a way that OSHA now has admitted is wrong. The result has been widespread confusion on the part of employers, OSHA inspectors and the OSHRC judges.
The Lockout Standard
OSHA's Lockout Standard, 29 C.F.R. § 1910.147, generally forbids an employee from servicing or maintaining a machine unless machine power not only is turned off, but a lock is placed on its switch to keep another employee from unexpectedly turning the power back on.
During the drafting of the standard, OSHA realized that it could not require lockout for minor servicing operations, such as tool changes. Such an exception already was provided by the national consensus standard that OSHA used as the “primary basis” for the standard's proposed version: ANSI Z244.1-1982, “American National Standard for Personnel Protection – Lockout/Tagout of Energy Sources – Minimum Safety Requirements.” Without such an exception, lockout would be required so often, and even when it is not needed to address actual hazards, as to make the standard infeasible. That is why, for example, OSHA stated in the standard's preamble that employers need “flexibility” when an operator needs to “replace a worn carbide cutting tool insert in a tool holder.”
The final version of the Lockout Standard therefore provided the following exception to paragraph (a)(2)(ii): “Minor tool changes and adjustments, and other minor servicing activities, which take place during normal production operations, are not covered by this standard if they are routine, repetitive, and integral to the use of the equipment for production, provided that the work is performed using alternative measures which provide effective protection (See Subpart O of this Part).”
Problems Right Off the Bat
The wording of the Minor Servicing Exception raises numerous questions:
- What is “minor?”
- Which operations are “routine?”
- Which are “repetitive?”
- Is there a practical difference between “routine” and “repetitive/” If so, what is it?
- How often is often enough to be “repetitive?” Hourly? Once a shift? Daily? Weekly? Monthly?
- What does “integral to the use of the equipment for production” mean? Is lubrication “integral?” Is unjamming “integral?”
Nothing in the words of the Lockout Standard even hints at how to answer these questions.
Worse yet, the exception states that it applies only “during normal production operations.” Does turning off the machine to replace a tool mean that the machine is no longer in a “normal production operation?” Although that would make no sense, the sad fact is that a hypothetical clever lawyer for OSHA might well argue to a judge that “yes” is the legally correct answer.
First, OSHA's clever lawyer would start with the words of the exception itself. He would say that “once the machine is turned off, Your Honor, the machine is no longer in ‘operation.' And even if it were in ‘operation,' it is not in ‘production,' and, even if it were in ‘production,' it is certainly not in ‘normal' production.'” Sounds good, right?
Second, OSHA's clever lawyer would point to the Lockout Standard's definition of “normal production operations” as “the utilization of a machine or equipment to perform its intended production function.” He would observe that while the tool is being changed, the machine is not being “utilized” and certainly not to perform its “intended production” function. He also would point to OSHA's compliance directive on the Lockout Standard, CPL 02-00-147 (2008), which states that “activities requiring … shutoff and disassembly, such as changing a machine tool or cutting blade, usually take place outside of the normal production process ….”
Third, OSHA's clever lawyer would argue that the judge's hands are tied by the decision of the review commissioners in Westvaco Corp., 16 OSHC 1374, 1380 (OSHRC 1993).
There, the commission held that adjustments of shaft heads to accommodate customer specifications in anticipation of the next production run is “setting up” rather than normal production operations because it occurs before such operations commence, i.e., when the machine is not running. It did not seem to have occurred to the review commissioners that there is such a thing as minor set up, or that its implicit emphasis on the machine being off would effectively swallow up the exception. Nevertheless, Westvaco is commission-precedent-binding on the judge.
Fourth, OSHA's clever lawyer would then remind the judge that OSHA does not have to be right to win. The U.S. Supreme Court held in 1991 that OSHA prevails if its interpretation is merely “reasonable,” even if a judge thinks that it is wrong.
But would OSHA's superficially plausible interpretation be reasonable? No, it would not, although it would require an independent-minded commission to say so. OSHA's position means that the only way to meet the exception would be, for example, to replace a tool bit while a chuck is still turning, which is dangerous madness. Applying the view of OSHA's clever lawyer would mean that the exception would swallow itself up and could never practically apply.
The Nightmare Realized?
Unfortunately, arguments by OSHA such as those sketched above have been taken seriously and even adopted. For example, in Matsu Alabama Inc., No. 13-1713 (final order Nov. 2, 2015), employees sometimes would do minor things to dies, such as changing the date to be stamped onto manufactured parts. Although that is a classic case of the “minor … adjustment” mentioned in the Minor Servicing Exception, a commission judge nevertheless rejected the employer's argument that the exception applied.
He did not hold that the employer failed to show that the task was “minor,” “routine,” “repetitive” or “integral to the use of the equipment for production.” Instead the judge reasoned that, “Clearly, [the machines] could not have been [used “during normal production operations”] since the machines were shut down and were not being utilized to perform their intended production function.” (An appeal is pending but it does not raise this issue.)
To be fair, the judge in the Matsu Alabama case apparently thought that his hands were tied by the review commission precedent in Westvaco, discussed above. That's the decision that overlooked that there is such a thing as minor set up and did not realize that the commission's emphasis on the machine being off effectively would swallow up the exception.
To its credit, however, OSHA recently repudiated the reasoning of the Matsu Alabama decision and the implication of the Westvaco decision. In a settlement agreement dated June 8, 2016 (which this writer negotiated), OSHA agreed in writing that “merely shutting off a machine (for example, to change a tool bit or blade) does not make the minor servicing exception inapplicable.” Spread the word!
It Gets Worse
Many other uncertainties about the exception plague its application in the real world. What about quickly re-programming a computer-driven machine so that it can make a different part for a different customer or product line? Isn't that a “minor … adjustment?” Would interrupting production to do that for 30 seconds take one out of “normal production operations.” What about five minutes of programming time? And what does the amount of down time have to do with safety anyway? Who cares what the delay is? The exception is so badly written that the gap between its words and real-world rationality creates an unbridgeable compliance nightmare.
And what does the exception mean by “alternative measures” that provide “effective” protection? Some OSHA inspectors think that because the parenthetical statement “(See Subpart O of this Part)” immediately follows the phrase “effective protection,” alternative measures are not “effective” unless they meet Subpart O's machine guarding requirements. That is not true.
The preamble to the final version of the Lockout Standard states that the requirement for effective alternative protection will only “generally” involve compliance with those machine guarding requirements. Moreover, the exception used the equivocal statement that the employer should “see” Subpart O, not “comply” with Subpart O. It seems obvious that OSHA must have written the exception and its preamble this way because minor servicing often requires an employee to reach around or move a barrier guard out of the way. For example, one cannot replace a drill bit in a chuck of a drill press with a barrier guard in place.
This leaves the employers to find guidance from only the word “effective,” which is so subjective that it could mean anything. Suppose the operator is the only employee near the machine while the drill bit or die block is being replaced? Would his immediate watchful presence be “effective?” Wouldn't it prevent an “unexpected” restart? Would turning off the power and hanging a “Do Not Start” tag on the switch be “effective?” What about die blocks between rams of mechanical power presses? Suppose they have “Do Not Remove” tags on them? What about relying on a machine interlock? A light curtain? The standard leaves employers at sea on all these common questions.
And this raises another issue: Interlocks and light curtains often rely on control circuitry. OSHA inspectors are taught that to be “effective,” the control circuitry of interlocks and light curtains must be “dual channel.” That redundancy requirement appears nowhere in the Lockout Standard or in Subpart O's machine guarding standards. Not even OSHA's most sophisticated machine guarding standard, §1910.217 on mechanical power presses, requires it. Moreover, although OSHA's drafting model for its Lockout Standard, ANSI Z244.1-1982, mentioned “specially designed control circuits” in ANSI's version of the minor servicing exception, OSHA omitted that language from the Lockout Standard. By what right does OSHA read it back in?
As discussed above, crucial terms in the Minor Servicing Exception are quite amorphous, including “minor” and “effective.” Are they so amorphous as to violate the constitutional ban on vagueness?
In Dayton Tire, 23 BNA OSHC 1247, 1250 (OSHRC 2010), the employer made just that argument. Although the review commission rejected it, nothing in its opinion came to grips with the real-world problem rightly raised there; that the Lockout Standard gives employers no notice of what is “minor” or “effective.”
The commission started its discussion with an irrelevancy – that the words “minor” and “effective” are “necessarily broad enough to cover the myriad servicing activities to which” they might apply. Yes, that is true, but it is irrelevant if those words fail to give fair notice of what they mean.
The commission then wrote that “a reasonable employer in the tire manufacturing industry could determine what constitutes ‘effective' alternative protection, given that the [Standard's] stated purpose … is to prevent injury” from unexpected startup. So, we are told, “effective” is not vague because it is unequivocal: It commands one to prevent injury, period.
That sounds good until you think about it. The commission's reasoning boils down to an assertion that the exception is not vague because it requires perfection. But that ignores holdings by the Supreme Court that OSH Act standards may not require protection at all costs and regardless of risk. Alternative measures must be permissible even if they are less than perfect, and it is OSHA's job to give employers fair notice of what those permissible alternatives are or how to determine what they are.
The last piece of reasoning used by the commission in rejecting the employer's vagueness argument was that to state the “alternative” measures it requires, the exception cross-references the machine guarding standards in Subpart O. That would be a logical point if the exception expressly required that alternative measures “comply with” Subpart O. But as discussed above, that is not what the exception says. It instead used the equivocal statement that the employer should “see” Subpart O.
The commission also seemed unaware that the standard's preamble said that alternative measures would only “generally” be Subpart O-compliant guarding. The commission seemed to lack a real-world appreciation of why OSHA likely equivocated on this point; that minor tool changes often require employees to reach around or move barrier guards.
The Occupational Safety and Health Review Commission has a difficult job to do, and it usually does it as well as can be reasonably expected. In the case of the Lockout Standard and its Minor Servicing Exception, however, the commission should re-examine its past pronouncements with a critical eye. EHS
Art Sapper is senior counsel to the Workplace Safety and Health Group of Ogletree, Deakins, Nash, Smoak & Stewart, P.C. Sapper is the former deputy general counsel of the Occupational Safety and Health Review Commission and a former professor of OSHA law at Georgetown University Law Center. He can be reached at [email protected] and at 202-263-0270.