Company safety data sheets on new chemicals frequently lack the worker protections EPA claims they include

Richard Denison, Ph.D., is a Lead Senior Scientist.

Readers of this blog know how concerned EDF is over the Trump EPA’s approval of many dozens of new chemicals based on its mere “expectation” that workers across supply chains will always employ personal protective equipment (PPE) just because it is recommended in the manufacturer’s non-binding safety data sheet (SDS).[pullquote]How much farther under the bus will the Trump EPA throw American workers?[/pullquote]

The typical course has been for EPA to identify risks to workers from a new chemical it is reviewing under the Toxic Substances Control Act (TSCA), but then – instead of issuing an order imposing binding conditions on the chemical’s entry onto the market, as TSCA requires – to find that the chemical is “not likely to present an unreasonable risk” and impose no conditions whatsoever on its manufacturer.  This sleight of hand is pulled off by EPA stating that it:

expects employers will require and workers will use appropriate personal protective equipment (PPE) … consistent with the Safety Data Sheet prepared by the new chemical submitter, in a manner adequate to protect them.

We have detailed earlier the myriad ways in which this approach strays from the law, is bad policy and won’t protect workers.  But here’s yet another gaping problem:  When we are able to look at the actual SDSs – that is, when EPA has made them available and when they are not totally redacted – we are frequently finding that the specific PPE that EPA claims to be specified in the SDSs – and that EPA asserts is sufficient to protect all workers handling the chemical – is not in the SDSs.  

EDF recently examined the SDSs for each of five new chemicals where EPA has declared them “not likely to present an unreasonable risk” and included the language I cited above.  EPA has also included the five in a proposed Significant New Use Rule (SNUR) that would require companies to notify EPA if they intend to use chemical in a particular manner that EPA has defined as a “significant new use.”  On Monday, EDF filed extensive critical comments on those proposed SNURs.

The reason we are focusing here on these chemicals is because, by law, EPA had to establish a rulemaking docket for the SNUR and place in that docket certain supporting documents pertaining to each new chemical.  Among those documents is (supposed to be) the chemical’s corresponding SDS.

Unfortunately, for two of the five chemicals (identified as P-18-0073 and P-19-0010, because the companies claimed their actual identities to be confidential), EPA failed to provide a copy of the SDS in its docket even though it is part of the documentation the company was required to submit to EPA.  For another of the five (P-17-0239), the copy of the SDS EPA included in the docket is totally redacted – even though much if not all of its content comprises health and safety information not eligible for confidential business information (CBI) protection under TSCA and, for the remainder, there is no evidence EPA has reviewed and approved any CBI claims the company asserted for the SDS.

That leaves us with the SDSs for the remaining two cases (P-18-0048 and P-18-0122), which are unredacted.  Now we can compare what they specify by way of PPE to the specific PPE that EPA relied on in determining these chemicals are “not likely to present an unreasonable risk.”

P-18-0048:  Here is what the “not likely” determination document for P-18-0048 states:

Risks to workers:  Reproductive toxicity via dermal exposure; corrosion to all tissues via dermal and inhalation exposures.

PPE EPA relies on:  EPA identifies as “appropriate PPE” the use of “impervious gloves and a respirator.” EPA goes on to state:

EPA expects that employers will require and workers will use appropriate personal protective equipment, including dermal and respiratory protection with an Assigned Protection Factor [APF] of 50, consistent with the Safety Data Sheet submitted with the PMN [premanufacture notice], in a manner adequate to protect them. (p. 6, emphasis added)

The associated SDS does recommend wearing “protective gloves,” “suitable protective equipment,” and “appropriate chemical resistant gloves.”  Its only reference to respiratory protection, however, is this:

[I]n the case of insufficient ventilation, wear suitable respiratory equipment.

Nowhere does the SDS specify use of a respirator with an APF of 50.  The SDS is clearly not consistent with EPA’s own description of it.

P-18-0122:  Here is what the “not likely” determination document for P-18-0122 states:

Risks to workers:  Lung toxicity via inhalation; irritation to skin, eyes, lung and GI tract.

PPE EPA relies on:

Risks will be mitigated if exposures are controlled by the use of appropriate PPE, including respiratory protection with an APF of 10.  Risks could not be quantified for irritation hazards, but appropriate PPE, including impervious gloves and protective eye wear, would mitigate concerns.  EPA expects that employers will require and workers will use appropriate personal protective equipment (i.e., impervious gloves, protective eye wear, and a respirator), consistent with the Safety Data Sheet prepared by the PMN submitter, in a manner adequate to protect them.  (pp. 5-6, emphases added)

While the corresponding SDS does recommend certain types of gloves and safety glasses, it specifically states:

Other protective equipment is not generally required under normal working conditions.

The only mention of use of a respirator anywhere in the SDS is where an OSHA regulatory workplace standard is exceeded – which is clearly not the case here, as no such standards exist for the new chemical.  Nowhere does the SDS specify use of a respirator with an APF of 10.  Here again, the SDS is clearly not consistent with EPA’s own description of it.

Other recent cases found

This finding spurred us to look further at other “not likely” determinations and the corresponding SDSs.  This is slower-going, because there is no electronically accessible docket.  That’s not only because EPA has not proposed a SNUR for other new chemicals to which it recently gave the green light; it’s also because EPA has failed to comply with its own regulations requiring it to provide electronic access to all new chemical submissions it receives.

As we have described elsewhere, EDF has had no choice but to request the “public files” for these chemicals through EPA’s Docket Center, which can take several weeks (they come by snail mail on a CD-ROM).

I looked at a number of recent new chemicals EPA has green-lighted for which we have received public files.  In one case no SDS was provided in the public file, while in two others the SDS was there but again totally redacted.  In some of the remaining cases the SDS recommended PPE that matched that EPA described in its “not likely” document, or at least came close.

But in other cases, there was not a match.  Here are two examples:

P-19-0021/22:  Here is what the “not likely” determination document for P-19-0021 and P-19-0022 states:

Risks to workers:  Lung overload via inhalation.

PPE EPA relies on:

Risks will be mitigated if exposures are controlled by the use of appropriate PPE, including a respirator with APF of 50.  EPA expects that workers will use appropriate PPE consistent with the SDS prepared by the PMN submitter, in a manner adequate to protect them. (p. 5, emphases added)

The associated SDS makes only this reference to respiratory protection:

Respiratory protection Mist respirator, include single use respirator

Nowhere does the SDS specify use of a respirator with an APF of 50.  The SDS is clearly not consistent with EPA’s own description of it.

P-18-0212:  Here is what the “not likely” determination document for P-18-0212 states:

Risks to workers:  Systemic effects via inhalation exposure; portal of entry/contact effects to the eyes, lungs and skin following ocular, inhalation, and dermal exposures

PPE EPA relies on:

The risks and hazards identified will be mitigated if exposures are controlled by the use of appropriate PPE, including impervious gloves, respirators with an APF of at least 10, and eye protection.  EPA expects that workers will use appropriate personal protective equipment (i.e., impervious gloves, respirator with an APF of at least 10, and eye protection), consistent with the Safety Data Sheet submitted with the PMN, in a manner adequate to protect them. (p. 5, emphases added)

The associated SDS makes this reference to respiratory protection:

Respiratory Protection:  For operations where inhalation exposure can occur use an approved respirator.  Recommendations are listed below.  Other protective respiratory equipment may be used based on user’s own risk assessment.  Recommended respirators include those certified by NIOSH.

Recommended:  Full Face Mask with a combination particulate/organic vapor cartridge.

Nowhere does the SDS specify use of a respirator with an APF of 10.  The SDS is clearly not consistent with EPA’s own description of it.

Conclusion

In each of these cases, EPA identified a particular type of respirator as necessary for its finding that the chemical is not likely to present an unreasonable risk, and in each case, EPA asserted that the corresponding SDS specified that type of equipment.  But in fact, in each case, the SDS does not specify that type of respirator.  EPA’s decisions run counter to the actual evidence before the agency, and EPA has actually mischaracterized that evidence.  That amounts to arbitrary decision-making.  Practically speaking, this mismatch means that workers could follow the SDS to a T and be using a respirator that is not sufficient to protect them against the chemical’s identified risks.

As we have noted before, EPA’s reliance on SDS-recommended PPE flouts the law and falls vastly short of what TSCA requires EPA to do to protect workers.  Amended TSCA requires EPA to issue binding orders to mitigate identified risks posed to workers by new chemicals, which it has identified in each of the cases we cite above.  EPA’s mere “expectation” that PPE will universally be available, used and effective is wholly insufficient to address the identified risks.  The recommendations in an SDS are not binding on employers, neither on manufacturers nor on other companies downstream in supply chains.  Failure to always use PPE or for it always to be effective is clearly reasonably foreseeable, and EPA is required to mitigate risks from “reasonably foreseen conditions of use” of a new chemical.  PPE is the option of last resort under the longstanding Industrial Hygiene Hierarchy of Controls adopted by OSHA and embraced by the industrial hygiene community.  Reliance on expected use of PPE shifts the burden of protection off of EPA and employers and onto the backs of workers.

Now, we find that even the PPE EPA identifies as necessary to be in an SDS in order to determine that a new chemical is not likely to present an unreasonable risk is frequently absent from the SDS.  Even under its own flawed theories, EPA is utterly failing to protect workers from the risks of these chemicals.

How much farther under the bus will the Trump EPA throw American workers?