The U.S. Environmental Protection Agency (EPA) released on December 20, 2019, an updated version of the “Working Approach” document that builds upon EPA’s November 2017 “New Chemicals Decision-Making Framework: Working Approach to Making Determinations under Section 5 of TSCA.” EPA states that the updated document, “TSCA New Chemical Determinations: A Working Approach for Making Determinations under TSCA Section 5,” explains its approach for making one of the five affirmative determinations on new chemical notices under the Toxic Substances Control Act (TSCA):
- The chemical or significant new use presents an unreasonable risk of injury to health or the environment;
- Available information is insufficient to allow EPA to make a reasoned evaluation of the health and environmental effects associated with the chemical or significant new use;
- In the absence of sufficient information, the chemical or significant new use may present an unreasonable risk of injury to health or the environment;
- The chemical is or will be produced in substantial quantities and either enters or may enter the environment in substantial quantities or there is or may be significant or substantial exposure to the chemical; or
- The chemical or significant new use is not likely to present an unreasonable risk of injury to health or the environment.
Upon publication in the Federal Register, EPA will accept comment on the updated document for 45 days in docket EPA-HQ-OPPT-2019-0684 at https://www.regulations.gov.
EPA notes that the updated document reflects feedback from a 2017 public meeting and comment period and EPA’s additional experience implementing the 2016 amendments to TSCA Section 5, and includes:
- Additional clarification and detail throughout;
- General guiding principles and concepts for making determinations;
- Decision-making logic and key questions that EPA must address; and
- Example applications of the Working Approach to reach each of the affirmative determinations under TSCA Section 5(a)(3).
It has been more than two years since EPA first published its new chemicals decision framework, and TSCA stakeholders have been eager to see the updated version. The significant time lag reflects, in part, the tremendous workload EPA is facing under new TSCA, for both new and existing chemicals. The delay also reflects EPA’s significant effort to provide a strong scientific and legal basis for its approach.
Bergeson & Campbell, P.C. (B&C®) can report that, in its view, the document accurately reflects EPA’s current approach to and practices regarding reviewing new chemicals.
EPA has reiterated that its decisions are and will continue to be based on risk, without consideration of cost or other non-risk factors, rather than only hazard. This approach is required by the statutory language. B&C and other TSCA practitioners had commented to EPA in 2017 that EPA’s early interpretation of “reasonably foreseeable” was essentially a hazard-based standard. In the early months after enactment, if EPA identified a hazard for the new chemical other than one considered “low,” EPA imposed a regulatory restriction to prevent any possible exceedance of a concern level.
In response to stakeholder requests, EPA includes in the revised Framework document a detailed elucidation of how EPA determines conditions of use as intended, known, and reasonably foreseeable. This is a welcome clarification. It gives submitters much more clarity as to EPA’s thinking, and it provides EPA a system to justify and document its views. Having such a structure provides a more predictable and defensible basis for EPA to make its determinations.
EPA also elucidates its views on how it interprets “sufficient information.” Although EPA does not so state in this section of the document, it is clear to B&C that EPA is following its mandate to minimize vertebrate testing in continuing to rely upon analogs and models as surrogates for test data on the substance itself. EPA does state in the section of the document related to testing that required testing will be “structured to reduce and replace vertebrate animal testing to the extent practicable and scientifically justified.” EPA also confirms that it will continue to take a tiered approach, following the standard practice for risk assessments. Some stakeholders have been critical that EPA has moved from requiring testing in every order it issued in 2016-2017 to rarely requiring testing in 2019. The underlying assumption is that EPA’s early implementation met all the statutory requirements and the current practice does not. B&C views EPA’s current practice of including “potentially useful” information in orders or significant new use rules (SNUR) as an appropriate balance between EPA’s requirement to make a determination on the unreasonable risk for permissible conditions of use and EPA’s likely need for additional information to inform a determination for the conditions of use prohibited by the order or SNUR. For example, if an order and SNUR prohibit releases of a substance to surface water because EPA predicts high concern for aquatic toxicity, EPA does not need test data on the substance to confirm that the substance is highly toxic to aquatic species that necessitate a prohibition of releases to water. EPA would only need the test data to determine if water releases might not present unreasonable risk if an entity seeks permission to release to water in the future.
EPA also explains how it interprets “unreasonable risk of injury to health or the environment.” EPA states that it makes this determination on a case-by-case basis and that it considers the severity and/or reversibility of effects of the substance or its degradants, the nature of exposures, including the duration and magnitude of potential exposures, and the characteristics of the exposed population. B&C finds this explanation helpful, but inclusion of an example or two would strengthen it.
In this version of the Framework, EPA explains its approach to worker protection. EPA notes that in reviewing a premanufacture notice (PMN), EPA considers worker exposures, including considerations of engineering controls, and calculates potential worker exposures absent personal protective equipment (PPE). If EPA finds potential unreasonable risk without PPE, EPA then evaluates whether PPE would sufficiently mitigate exposures so as to pose no unreasonable risk. EPA then states that it relies upon the Occupational Safety and Health Administration (OSHA) authority that compels employers to inform employees of the hazards, train those employees in proper use of PPE, and require that employees take the steps necessary to protect themselves. As B&C has demonstrated on behalf of the New Chemicals Coalition and provided to EPA in comments and in meetings, OSHA violations for gloves, goggles, or general dermal protection were quite rare (less than one percent of violations) among the 12 million violations that OSHA has issued. B&C argued, and EPA agreed, that this low rate of violations meets the statutory criteria of “not likely to present unreasonable risk under the [reasonably foreseeable] conditions of use.” It is not that glove or goggle non-use never happens, but the statutory language does not require that level of certainty and does not permit EPA to take regulatory action if there is only a very limited possibility of unreasonable risk. Critics of this approach insist that EPA is putting workers at risk because some workers are not properly informed, trained, or equipped, in violation of OSHA requirements. Those critics do not explain how a duplicative TSCA requirement would remedy that situation. EPA states that it evaluates whether standard hazard communication and PPE are reasonably foreseeable to provide the necessary protection for workers; if so, EPA relies upon the OSHA requirements; if not, EPA uses its authority under TSCA to impose exposure limits (that would presumably be met with engineering controls), hazard communication, and use of PPE. In B&C’s experience, EPA follows this approach. If use of impervious gloves, goggles, and common respiratory protection are sufficient to protect workers from unreasonable risk, EPA relies upon the OSHA authority. On the other hand, if EPA finds that the substance is unusually hazardous, or that non-standard measures are required to protect workers, EPA is using its authority under TSCA and imposing limits (such as a New Chemical Exposure Limit) with an order and corresponding SNUR.
Probably the most controversial aspect of EPA’s Framework is EPA’s use of SNURs without corresponding orders. Several groups have commented repeatedly that EPA’s use of so-called “non-order SNURs” violates the plain text of the statute. While B&C agrees that the statute does not specifically anticipate the non-order SNUR construct, B&C disagrees that the statute disallows this construction. B&C views EPA’s use of a SNUR to prohibit potentially problematic conditions of use as an effective way to limit what is “reasonably foreseeable,” allowing EPA and the submitter to proceed with the SNUR protections in place. Those who are critical of this approach state that SNURs are not sufficiently protective to satisfy the statutory requirements in a way that only orders can. A violation of either an order or a SNUR is still a Section 5 violation. In B&C’s view, SNURs are arguably more protective than orders. An order only applies to the signatory of the order (the submitter), while a SNUR applies to all potential manufacturers, processors, or users.
Once the submitter files a Notice of Commencement (NOC) for a substance that is the subject of an order and the substance is placed on the TSCA Inventory, there is no protection against another entity undertaking the conditions of use that are prohibited in the order until EPA publishes the SNUR corresponding to the order. For example, if an order prohibits domestic manufacture, once the original submitter commences import, if another company finds the substance listed on the TSCA Inventory, that company could immediately manufacture the substance in the United States without notifying EPA. By doing so, the second company could prevent EPA from promulgating the “import only” SNUR that corresponds to the order because the domestic manufacture activity would have occurred prior to EPA proposing the SNUR -- making domestic manufacture an ongoing use, not a significant new use. EPA could close this gap by proposing the SNUR contemporaneously with issuing the order, but that is essentially what EPA is doing with its approach of the SNUR preceding the determination.
If the problematic condition of use is not intended, but is reasonably foreseeable, by proposing the SNUR contemporaneously with making the determination, EPA ensures that if the condition of use commences prior to promulgation of the SNUR, the condition of use was not ongoing at the time of the SNUR’s proposal, and that condition of use will be prohibited when the SNUR takes effect.
Some have commented that SNURs are not as protective as orders because SNURs only require “notification” to EPA, whereas orders prohibit activities specified in the order. This view fundamentally misconstrues the effect of a SNUR and the permanence of an order. A party to the order can petition EPA to modify the order. When a submitter develops data (as required by an order or for another reason) and the data inform EPA’s assessment of risk of the substance, it is common for the submitter to request that EPA revise the order to remove the warnings or restrictions associated with that hazard. If an order prohibits release to water and later the manufacturer performs aquatic toxicity testing that demonstrates to EPA’s satisfaction that releases to water do not pose unreasonable risk (e.g., the substance is demonstrated to be low hazard), the submitter can petition EPA to remove the prohibition in the order against water releases. EPA would review the information and the conditions of use, decide whether releases to water would, might, or would not present unreasonable risk, and modify the order as appropriate.
If instead of an order, there is a SNUR prohibiting releases to water, a company seeking permission to release to water (i.e., to modify or lift the SNUR prohibition) would perform the testing to EPA’s satisfaction and submit the data with the anticipated conditions of use in a Significant New Use Notice (SNUN). EPA would then review the data and the intended, known, and reasonably foreseeable conditions of use and make a determination, just as it would if the substance were new or if EPA were considering modifying an order. The submitter could not undertake the Significant New Use (SNU) unless and until EPA makes its determination and takes the necessary action. In either case (order or SNUR), the activity is prohibited until EPA reviews the information, makes its determination, and takes any required action -- just as if the substance were a new chemical. It is decidedly not the case that submission of a SNUN is all that is required to undertake a SNU.
B&C applauds EPA’s significant effort in being more transparent with its policies and practices. It benefits all TSCA stakeholders to understand more clearly EPA’s approach. In B&C’s experience, the policies and procedures stated in this updated Framework document accurately represent EPA’s current practices. While it is possible that one or more stakeholders may bring a suit alleging that EPA is not permitted to use SNURs to limit reasonably foreseeable conditions of use without first issuing a corresponding order, B&C does not agree that doing so is less protective than first issuing an order. It is also possible that this Framework will invite litigation related to worker protection. Time will tell.
EPA has posted a document summarizing public comments received on the 2017 document and its responses. EPA discussed the updated document at a public meeting on December 10, 2019, and posted its presentation online.