Richard Denison, Ph.D., is a Senior Scientist.
One thing I’ve learned in observing EPA try to operate under the Toxic Substances Control Act (TSCA) over the years is that – faced with limited authority and significant evidentiary and resource burdens – the Agency often has to resort to a workaround to get something it needs to do done.
Can’t ban a nasty chemical? Wait until it’s voluntarily withdrawn and then pounce on it with a Significant New Use Rule (SNUR) to try to wedge the door closed. Witness PFOS and penta and octaBDE. (Under TSCA, without a SNUR in place on a chemical, a new producer or importer could start up without even letting EPA know; where EPA has issued a SNUR for a chemical, advance notification is required and EPA least has a chance to weigh in before production or import proceeds.)
Can’t require an up-front minimum data set for new chemicals? Recommend to companies that for certain chemicals they submit such a data set along with the pre-manufacturing notice (PMN) they’re required to file, or risk having EPA extend the review of their new chemical or negotiate with them to do the testing. EPA has made such “recommendations” for those relatively few new chemicals where the company “anticipates” at the outset producing it in large amounts in the first three years or where significant release or exposure is projected.
The latest such workaround? EPA’s simultaneous issuance of a proposed test rule and a proposed SNUR for a batch of high production volume (HPV) “orphan” chemicals that no company agreed to sponsor under the Agency’s voluntary HPV Challenge Program.
Necessity is the mother of invention, they say, and these creative new proposals are a case in point. But, my oh my, there’s gotta be a better way….
EPA bears a heavy cross to require testing
Now, to justify requiring testing under TSCA, EPA has a number of substantial burdens it must meet. First, EPA must first make one of the following statutory findings:
1. The substance “may present an unreasonable risk of injury to health or the environment” and the probability of exposure to it is more than just theoretical.
2. The substance is produced in “substantial quantities” and either enters the environment in “substantial quantities” or there is “substantial or significant” human exposure.
In practice, EPA is rarely able to make the first, “risk-based finding” because it lacks the very data that testing would provide. In virtually all cases where it has issued test rules, EPA has relied on making the second “exposure-based finding.”
Here are the criteria EPA uses to define “substantial” production and environmental release:
- substantial production/importation is one million or more pounds per year; and
- substantial release is one million or more pounds per year, or 10% of the total amount produced or imported.
EPA defines “substantial or significant” human exposure as that affecting 1,000 workers, 10,000 consumers, or 100,000 members of the general population. Here again, however, EPA has to have sufficient information to document that these thresholds are met.
Finally, there are two other, non-trivial findings EPA must make in order to issue a test rule. EPA must show that:
3. existing data are inadequate for risk assessment,
4. testing is needed to develop the data.
Issuing a SNUR is not as difficult as is a test rule; while EPA has a list of factors it must consider, it doesn’t have specific findings it must make. However, there are two significant limitations:
- A SNUR cannot reach any activities associated with a chemical that are already underway, because by definition they are not "new."
- A SNUR does not regulate a chemical's production or use; it only requires notification of EPA and provides an opportunity for an EPA review. To regulate, EPA would have to negotiate a Section 5(e) consent orderwith the notifier or demonstrate "unreasonable risk" to justify a restriction under Section 6 (a task that has proven virtually impossible in practice).
So how has EPA met these burdens in the present case?
All 45 of the chemicals that are the subject of the latest proposed rules are HPVs, i.e., produced at or above one million pounds per year. So they meet the substantial production threshold. But, as just noted, that’s not enough for EPA to be able to justify testing.
The 45 HPV chemicals that are the subject of the latest proposed rules fall into two classes:
- For 23 of them, EPA is proposing to require testing because it indicates it has sufficient information to meet the burden of showing “substantial or significant” human exposure. In the current context, EPA has based these findings on evidence that: (1) the chemicals are present in consumer products used by large numbers of the general population, and (2) more than 1,000 workers are reasonably expected to be exposed within one or more corporate entities producing the chemical.
- For the other 22, EPA is proposing a SNUR, because it lacks such information; the SNUR would require that EPA be notified before a company begins producing, importing or processing one of these chemicals for use in a consumer product or in a manner that would potentially expose more than 1,000 workers within a corporate entity.
Now, here’s the creative part:
- If EPA receives reliable evidence in comments it receives on the proposed rules indicating that a chemical in the first group is not produced or used in the manner it believes to be the case that justifies requiring testing, it will instead issue a SNUR for that chemical that would require notification should such production or use be pursued by a company.
- If EPA receives reliable evidence in comments it receives on the proposed rules indicating that a chemical in the second group is currently being produced or used in the manner that would trigger the SNUR – and hence that it should not be included in the SNUR because such production or use is in fact not new – EPA intends to issue a test rule for that chemical because such production and use would now meet the additional triggers needed to require testing.
Lest you’re thinking this paired-rule approach proposed by EPA is too clever by half, it’s good to remember that, way back in 1998 when it launched the HPV Challenge Program, EPA said its expectation wasthat all HPV chemicals would have, as a condition for being on the market, a minimum set of screening-level hazard data:
“EPA expects that, over time, the testing of new HPV chemicals will become routine, and companies may wish to test new HPV chemicals as they appear.”
Industry agreed, after considerable pressure was applied, and the American Chemistry Council (ACC) unilaterally launched in 2005 its short-lived voluntary Extended HPV Program that was intended to spur testing of the many hundreds of chemicals that have reached HPV production levels after the HPV Challenge was launched and hence were not included in it. But the Extended HPV Program was flawed from the outset, never attracting more than a small number of volunteer companies and delivering little new data.
In writing this post, I had planned to link to ACC’s webpages for the Extended HPV Program, but alas, they’ve disappeared from ACC’s website, and I could only find a single mention of it anywhere on the site, buried on this brief page on the HPV Challenge. That speaks volumes (pun intended).
At any rate, given the industry’s acknowledgment that at least all HPV chemicals ought to have a base set of hazard information publicly available, one would hope they will support EPA’s modest workaround. Even EPA’s proposed approach falls well short of ensuring all HPV’s have such data. That’s because under TSCA, the fact that a chemical is produced in volumes of one million pounds or more is not sufficient to trigger testing.
What does all this mean for TSCA reform?
The machinations EPA must go through to try to get basic safety data even for those chemicals produced in the highest volumes are pretty sad, in my view. What should be done?
First, a basic set of safety data ought to be mandated, not only for HPV chemicals, but for all chemicals in commerce. That straightforward notion is embodied in the preamble to TSCA itself and bears repeating:
“It is the policy of the United States that … adequate data should be developed with respect to the effect of chemical substances and mixtures on health and the environment and that the development of such data should be the responsibility of those who manufacture and those who process such chemical substances and mixtures.”
Second, EPA shouldn’t have to go through a case-by-case rulemaking, as required to issue a SNUR, just to be able to learn about and keep up with significant changes to the production and use of a chemical. Here again, there ought to be an ongoing, universal requirement that companies notify EPA when such changes occur. And EPA should have authority to look at, or re-look at, a chemical when that information warrants.
Those modest asks, it seems to me, are common- sense underpinnings of a sound chemicals management system in which the public can have confidence.
The Wikipedia entry for “workaround” notes: “Workarounds can also be a useful source of ideas for improvement of products or services.”
In this context, “improvement of products and services” ought to serve as code for legislative reform of TSCA.