The Trump EPA is throwing workers facing risks from new TSCA chemicals under the bus

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

[For more on this topic, see our recent 3-part myth-busting series:
Part 1          Part 2         Part 3]

We have blogged before (see here and here) about the steps initiated in mid-2018 by the Trump EPA to weaken new chemical reviews under the Toxic Substances Control Act (TSCA) – rendering them even less health-protective than under TSCA prior to the 2016 reforms enacted in the Lautenberg Act.

As these debilitating policy changes – still never publicly described or released, and apparently still not written down even for use within EPA – have taken hold, we have seen dozens of flawed new chemical decisions emerge.  We blogged extensively about the first such decision made under the new regimen in late July 2018.  Since then, about 60 more final determinations reflecting the new policies have been posted on EPA’s website.  These decisions pertain mostly to premanufacture notifications (PMNs), along with a few for significant new use notices (SNUNs).  At least 80% of these chemicals were cleared to enter commerce without being subject to any conditions whatsoever.  EPA accomplished this by issuing a final determination that each cleared chemical, or significant new use of a chemical, is “not likely to present an unreasonable risk.”  For these determinations, EPA is required under TSCA to post a statement of its finding, which it does in another table on its website.

We have been closely examining these “not likely” determination documents.  Some deeply disturbing patterns are emerging.  This post will describe one of them.[pullquote]A new addition to the long and growing list of illegal actions EPA has taken to render the new chemicals program weaker than under the old TSCA.[/pullquote]

Most striking is that for a significant majority of these chemicals, EPA either identified significant risks to workers or indicated it had insufficient information to determine the level of risk to workers.  Under the 2016 reforms to TSCA, either finding – that there are or may be risks or that there is insufficient information to determine the level of risk – requires EPA to issue an order specifying conditions sufficient to eliminate the risk.  Yet EPA did no such thing; instead, it cleared the chemicals for unfettered market access.  

Hazards posed to workers

It is not as if these chemicals are inherently benign; indeed, EPA identified significant hazards for many of them.  In the summary of EPA’s “not likely” determination document for PMN P-18-0221, the agency states that it identified the “potential for the following human health hazards: skin and lung sensitization, mutagenicity, carcinogenicity, and developmental, reproductive, liver, and kidney toxicity.”  Yet the very next sentence states:  “EPA concludes that the new chemical substance is not likely to present an unreasonable risk under the conditions of use.”

Here’s another example, the “not likely” determination document for PMN P-18-0324.  EPA identified “potential for the following human health hazards: lung toxicity (waterproofing), irritation, neurotoxicity and developmental toxicity.”  Yet, again, “EPA concludes that the new chemical substance is not likely to present an unreasonable risk under the conditions of use.”

Risks posed to workers

In both of these cases, as in many others, EPA also identified risks to workers from these same hazards, with predicted exposures well in excess of “safe” levels.

So how does EPA make these risks to workers magically disappear – without imposing a single condition on such chemicals?  By simply “expecting” them out of existence.  All of these “not likely” determination documents include a sentence like this (emphases added):

EPA expects that workers will use appropriate personal protective equipment (i.e., impervious gloves), consistent with the Safety Data Sheet prepared by the PMN submitter, in a manner adequate to protect them.

There are so many things wrong with this.  Here are some:

Reliance on nonbinding documents:  EPA provides absolutely no evidence to support its “expectation.”  It simply assumes that all workers coming into contact with the chemical will always “use appropriate personal protective equipment” (PPE) because the company submitting the PMN for the chemical has prepared a safety data sheet (SDS) recommending its use.  But SDSs impose no binding requirements either on employers or their employees to do so.  The mere presence of language in an SDS is completely insufficient to conclude that PPE is actually utilized or is sufficiently effective and protective.  While SDSs are required to be provided as a hazard communication tool, the only legal requirement under OSHA is that the employer provide the SDS to employees and train them on how to access and understand them.  For example, the 2012 OSHA Hazard Communications Standard explains (emphases added):

While the current HCS [Hazard Communication Standard] and this final standard require the provision of information on recommended control measures, including respiratory protection, personal protective equipment, and engineering controls, there is no requirement for employers to implement the recommended controls.  An employer should use all available information when designing an appropriate protective program, but a recommendation on a safety data sheet by itself would not trigger the need to implement new controls.

Any legal requirement under OSHA that SDS recommendations be followed would come through a separate requirement such as where there is an OSHA exposure limit for the substance.  For new chemicals, for which OSHA obviously has no such standards, there is no OSHA obligation beyond OSHA’s rarely used general duty clause for employers to provide any protection for workers.

Even companies’ own SDSs acknowledge this reality, including when it comes to an SDS that accompanies a chemical as it moves downstream to other processors and users.  For example, the SDS for PMN P-18-0070 states that the manufacturer:

assume[s] no responsibility regarding the suitability of this information for the user’s intended purposes or for the consequences of this use. Individuals should make a determination as to the suitability of the information for the particular purpose(s).

Despite these limitations of SDSs, EPA simply assumes, with no actual evidence, that there will be 100% use and efficacy of the PPE specified in a SDS by all workers throughout a chemical’s supply chain.

It’s also worth noting EPA has erected major barriers to the public gaining access to the SDSs companies submit with their PMNs.  We have blogged extensively about EPA’s failure to provide ready access to PMN public files, which include SDSs.  Moreover, even once we gain such access, SDSs are often heavily or wholly redacted based on illegal claims – apparently unreviewed by EPA and hence allowed to stand – that they constitute confidential marketing and sales information protected from disclosure.

Reliance on ineffective nonbinding documents:  There is significant evidence that SDSs are frequently not understood or followed.  For example, one recent systematic search and review of the literature identified serious problems with the use of SDSs even as hazard communication tools: they are often inaccurate, incomplete, and too technical for workers to understand.  The 2012 OSHA Hazard Communications Standard corroborates these findings.  For example, the Standard reports that “several studies show that employees do not understand approximately one-third of the safety and health information listed on SDSs prepared in accordance with the current standard” and that “[s]tudies also report that roughly 40% of persons reviewing SDSs found them difficult to understand.”  Also, see OSHA’s Inspection Procedures for the Hazard Communication Standard for more on the limitations of SDSs.

Reliance on ineffective methods specified in ineffective nonbinding documents:  Reliance on PPE has major practical limitations and, at best, exhibits mixed effectiveness in the real world.  For example, OSHA concluded that respirators are the “least satisfactory approach to exposure control.”  The agency provides the following explanation:

[T]o be effective, respirators must be individually selected, fitted and periodically refitted, conscientiously and properly worn, regularly maintained, and replaced as necessary.  The absence of any one of these conditions can reduce or eliminate the protection the respirator provides.

Respirator effectiveness ultimately relies on the practices of individual workers who must wear them. … Furthermore, respirators can impose substantial physiological burdens on workers, including the burden imposed by the weight of the respirator; increased breathing resistance during operation; limitations on auditory, visual, and olfactory sensations; and isolation from the workplace environment.

OSHA therefore continues to consider the use of respirators to be the least satisfactory approach to exposure control … .

Yet EPA relies solely on PPE identified in the SDSs as the basis for asserting that unreasonable risks to workers will be eliminated.

Failure to rely on the best available science and policy:  EPA’s reflexive impulse to resort to reliance on PPE fails to reflect the best available science and policy. The best available science and policy have long been memorialized through OSHA’s Industrial Hygiene Hierarchy of Controls (HOC), which prioritizes measures to eliminate or reduce the presence of a hazard in occupational settings (e.g., substitution/use of less toxic chemicals and institution of engineering controls) over measures like PPE that shift burdens onto the workers themselves.  The HOC, not PPE, exemplifies the best available science and policy for creating safe, healthful workplace environments.

In 2016, EPA proposed updates to its SNUR regulations “to align these regulations with revisions to the Occupational Safety and Health Administration’s (OSHA) Hazard Communications Standard (HCS).”  EDF filed comments supporting the changes, as did OSHA itself, most notably lauding EPA’s proposal to incorporate the HOC into its regulation.

Unfortunately, EPA has not finalized the proposed modifications to its SNUR regulations.  While EPA had been incorporating language reflecting the HOC into consent orders it was issuing for new chemicals, it now appears EPA is abandoning reliance on the HOC altogether, both by failing to issue orders and basing its “not likely” determinations on assumed use of PPE.

Another industry wish granted

If all this has a familiar ring to it, it should.  Under this Administration, the chemical industry has pressured EPA to weaken or abandon altogether workplace restrictions on new chemicals under TSCA.  It seems they’re once again getting their way.

The 2016 amendments to TSCA enhanced EPA’s obligations to protect workers, obligations that date back to the original law passed in 1976.  TSCA now explicitly defines workers to be a “potentially exposed or susceptible subpopulation.”  And it specifically requires EPA to identify, assess and mitigate unreasonable risks to workers.

EPA appears at least in limited ways to be identifying risks to workers posed by new chemicals.  But when it comes to actually protecting them from such risks – which would require imposing binding requirements on companies – that’s when workers get thrown under the bus.  If EPA’s toxics office were serious about environmental justice as a priority under TSCA, as it has recently asserted it is with respect to workers exposed to pesticides, it would not so casually cast aside risks to workers from new chemicals.

Add this to the long and growing list of illegal actions EPA has taken to render the new chemicals program weaker than under the old TSCA.

For more on this topic, see our recent 3-part myth-busting series:
Part 1          Part 2         Part 3