Richard Denison, Ph.D., is a Lead Senior Scientist.
[Corrections added below in bracketed italics on 1-17-20]
Last month EPA finally released its long-awaited update to its controversial 2017 New Chemicals Decision-Making Framework, which describes how EPA is conducting risk reviews of new chemicals under the Toxic Substances Control Act (TSCA). While we are still reviewing it and will be filing comments, it is clear the new document suffers from many of the same problems as the prior version, as well as raising additional concerns.
A core problem of both documents is EPA’s illegal bifurcation of its treatment of a new chemical’s “intended” conditions of use – those proposed by the company submitting a premanufacture notification (PMN) to EPA – from the chemical’s “reasonably foreseen” conditions of use. EPA does so despite TSCA’s clear instruction that EPA address potential risks from both categories in an integrated manner and at the same time. EPA’s frameworks instead relegate any consideration of “reasonably foreseen” conditions of use to a separate, later process undertaken upon receipt of a separate notification submitted to EPA in response to a Significant New Use Rule (SNUR) – assuming EPA has actually promulgated a final SNUR for the chemical in question.
EPA has now used this bifurcated approach to greenlight hundreds of new chemicals for market entry – finding that they are “not likely to present an unreasonable risk” based on a review only of the chemicals’ intended conditions of use. EDF has blogged in detail about the inadequacies of EPA’s “SNUR-only approach” and the disturbing extent of its application.
EPA has yet to provide any legal justification for its approach – how it believes the approach comports with TSCA – let alone demonstrate how it provides for protection of health and the environment despite deviating from what Congress intended EPA do under the law.
[pullquote]How on earth can EPA assert with a straight face that it is promulgating SNURs that precede its “not likely” determinations?[/pullquote]While we will have much more to say on the new framework, in the remainder of this post I want to focus on EPA’s characterization of its reliance on the SNUR-only approach. EPA now says it has expanded that approach to two different scenarios: One is “SNURs that Precede “Not Likely” Determinations” and the other is “SNURs that Follow “Not Likely” Determinations.”
We have examined the accuracy of EPA’s claim that the first type of SNUR precedes EPA’s “Not Likely” determination for a given new chemical. The timing is critical here: If there is a significant lag between EPA’s “Not Likely” determination and the issuance of a SNUR, all kinds of problems arise, which we have discussed previously. To name two:
- If a company engages in what EPA plans to deem a “significant new use” during the gap between the determination and at least proposal of a SNUR, then EPA cannot subject that use to the notification requirements of the SNUR because the use is “ongoing” and no longer “new.” That includes a new use engaged in by the company that got a green light for its chemical based on EPA’s review of only its intended conditions of use.
- Such a company that wants to have the ability to engage in uses beyond those it said it initially intended would have serious incentives to seek to avoid having EPA issue the SNUR. Because SNURs are done through rulemaking, the company can urge EPA to block or modify the SNUR through the rulemaking process. It can also apply pressure on EPA not to pursue a SNUR at all.
So is EPA being accurate when it claims that a large number of its SNURs precede EPA’s “Not Likely” determination for those same new chemicals?
The answer is a resounding no: Not a single final SNUR has been issued for any of the hundreds of new chemicals receiving “not likely” determinations prior to the dates of those determinations.
Being charitable, we thought maybe EPA meant that it proposes the SNURs prior to making the “not likely” determinations. So we compared the dates of the determinations with the dates on which the corresponding SNURs were proposed. Here’s what we found:
Since the 2016 reforms to TSCA, EPA has proposed SNURs for 68 new chemicals receiving “not likely” determinations where EPA’s determination indicates it was based on issuance of the SNUR. Of those 68 chemicals:
- For 59 of them (87%), the date of the SNUR proposal followed the date of the determination, lagging behind by anywhere from 7 to 55 days [CORRECTION: “4 to 55 days”]. (Actually the SNURs for the six chemicals with a 55-day lag have [CORRECTION: “SNUR for the chemical with a 55-day lag has”] yet to be proposed, so that gap gets bigger with each passing day.)
- Only for 9 of them (13%), did the date of SNUR proposal precede the date of the determination.
- For only 13 of these 68 chemicals has EPA finalized a SNUR – and those came a whopping 178 days after their “not likely” determinations were made. For the remaining 55 chemicals, EPA has not finalized a SNUR even now, much less before making the “not likely” determination.
How on earth EPA can assert with a straight face that it is promulgating SNURs that precede its “not likely” determinations is beyond me.
In the past few months, EPA has started issuing the second kind of SNURs: ones that clearly are intended to follow the corresponding “not likely” determinations. To date, at least 84 new chemicals have been subject to this approach. Here, EPA’s determination makes no mention of a forthcoming SNUR, but later, often many months later, EPA proposes one that states EPA identified concerns “if the chemicals were not used following the limitations identified by the submitters in the notices.” In these cases, EPA never identifies any of these uses as “reasonably foreseen”, effectively eliminating a whole category of uses Congress told it to evaluate. EPA has never explained or legally justified this approach, which suffers from the same fundamental flaws as its other SNUR-only approach. And to the extent this approach results in an even larger lag between the dates of the “not likely” determinations and the SNURs, it exacerbates the problems I identified earlier.
Meanwhile industry continues to complain loudly about EPA issuing any SNURs at all – most recently at EPA’s December 10, 2019 public meeting previewing the new framework. This is highly ironic, given that it was the industry that demanded and got EPA to pursue its SNUR-only approach in the first place.
Industry seems to want to have it both ways: Have EPA review their PMNs solely based on their asserted intended conditions of use and exclude any reasonably foreseen conditions of use or find there are none. Get their “not likely” determinations on that basis. And then have EPA be done with it and never issue a SNUR that places any requirements whatsoever on the company’s ability to depart however it wants from these asserted intended conditions of use.
What a mess EPA has gotten itself into, starting the minute it chose to depart from the clear approach laid out in TSCA.