On May 8, 2017, the U.S. Environmental Protection Agency (EPA) issued in the Federal Register a notification stating that a public meeting will be held to elicit public input on approaches for assigning and applying Unique Identifiers (UID) that are now required whenever it approves a Confidential Business Information (CBI) claim for the specific chemical identity of a chemical substance. 82 Fed. Reg. 21386. The meeting will be held on May 24, 2017, from 1:00 p.m. to 4:00 p.m. (EDT) in Washington, D.C. Registration is available online and requested by May 22, 2017. The meeting can be accessed remotely. EPA will also be soliciting comments via submittal to Docket ID number EPA-HQ-OPPT-2017-0144 on www.regulations.gov. Comments are due by July 7, 2017. The notice is thoughtful, well written, and timely, as discussed below.
TSCA Section 14 Requirement to Assign a UID
Section 14(g)(4)(A) of the Toxic Substances Control Act (TSCA), as amended by the Frank R. Lautenberg Chemical Safety for the 21st Century Act, directs EPA to “(i) develop a system to assign a unique identifier to each specific chemical identity for which the Administrator approves a request for protection from disclosure, which shall not be either the specific chemical identity or a structurally descriptive generic term; and (ii) apply that identifier consistently to all information relevant to the applicable chemical substance.” EPA must also, under amended TSCA, annually publish and update a list of chemical substances, referred to by their UIDs; ensure that any nonconfidential information received identifies the chemical substance using the UID while the specific chemical identity of the chemical substance is protected from disclosure; and for each claim of protection, clearly link the specific chemical identity to the UID.
Assigning the UID
The notice describes what EPA is considering in terms of assigning the UID, specifically “using a numeric identifier, which will incorporate the year the claim was approved.” EPA states that including this date will “facilitate tracking of the expiration of the CBI claims for specific chemical identity made in that document,” pursuant to TSCA Section 14(f)(3). EPA rejected using accession numbers given that language in TSCA Section 8(b)(7)(B) “suggests that accession numbers were intended to be distinct from the unique identifier.” EPA notes that expanding the use of accession numbers beyond their current use and purpose could be confusing. Accession numbers are currently assigned following a notice of commencement of commercial manufacture or import under Section 5, and identify chemicals that are or were formerly on the confidential portion of the TSCA Inventory. The UID would be an outcome of a CBI determination under Section 14, however, and at this stage, very few chemicals with accession numbers have been subject to this new CBI review requirement.
Application of the UID
TSCA Section 14(g)(4)(A)(ii) requires that EPA apply the UID to “all information relevant to the applicable chemical substance.” Section 14(g)(4)(C) also instructs EPA to ensure that any nonconfidential information received while the specific chemical identity of the substance is protected from disclosure, identifies the chemical substance using the UID. In addition, under Section 14(g)(4)(D), after the underlying CBI claim expires, is denied by EPA, or is withdrawn, EPA must clearly link the specific chemical identity to the UID in public documents that previously used the UID to “protect the specific chemical identity in information that the Administrator has made public.”
In the notice, EPA also draws special attention to its concerns regarding the use of the UID. According to EPA, the requirements to apply the UID to other, non-confidential information, and to use the UID in a manner to protect the specific chemical identity in public information while the specific chemical identity is protected from disclosure, “do not appear to be completely reconciled.” EPA states that it identified several situations where applying the same UID to every instance where information pertaining to the same chemical substance is reported under TSCA could cause CBI, including specific chemical identity, to be revealed.
EPA notes that if documents concerning the same substance submitted by different companies at different times and for different purposes were to always be assigned the same UID, then each company could learn that the substance was marketed in the U.S. by another company, and possibly learn of new uses and other information concerning the substance. EPA provides the following example:
Example 1: Company A files a Premanufacture Notice (PMN) and later commences import of Chemical X, for which its CBI claim for chemical identity is approved by EPA, resulting in Chemical X being placed on the confidential portion of the TSCA Inventory. Company B subsequently files a notice of substantial risk under TSCA section 8(e) on the same substance, which it is utilizing for research and development, also claiming chemical identity as CBI. EPA approves this claim and assigns the same unique identifier.
In the example above, Company B can determine the confidential information that Chemical X is in U.S. commerce, without having to submit and meet the terms of a bona fide request. Both Company A and Company B can learn that another company has an active interest in Chemical X. Each company can also determine any non-CBI information about the other company, uses, or other information that might be in the other company’s submission.
EPA notes that if the specific chemical identity is not uniformly claimed as CBI in all such submissions, applying the UID to a submission with a non-confidential chemical identity “effectively destroys the CBI claim for chemical identity in all other documents that use the same unique identifier.” E.g., according to EPA, if Company B chose not to claim chemical identity as CBI in its Section 8(e) filing regarding a research and development use, and EPA applied the UID to the Section 8(e) submission, “this submission could be readily linked to Company A’s submission, and the confidential chemical identity in Company A’s submission would be revealed to the public, along with the fact that Chemical X is in commerce in the United States.”
According to EPA, there are additional circumstances where the action or inaction of one company could cause the CBI of another company to be revealed:
Example 2: EPA receives and approves Company A’s CBI claim for chemical identity in a Notice of Commencement (NOC). The substance is placed on the confidential portion of the Inventory and a unique identifier is assigned. Subsequently, Company B files a bona fide notice concerning the same substance. Company B does not claim chemical identity as CBI. In accordance with section 14(g)(4)(C), EPA applies the unique identifier to the public version of the bona fide submission.
EPA states that applying the UID to the bona fide submission effectively discloses the identity of Company A’s chemical, and reveals that the substance is in U.S. commerce. Because it is no longer secret that the substance is in U.S. commerce, the substance would be removed from the confidential portion of the Inventory, and all information concerning uses, company identity, and other information that was not claimed as CBI in the underlying PMN, NOC, and the bona fide notice could be linked together, potentially further disclosing information about the chemical that the other company may have claimed as confidential.
EPA considered two alternative approaches to applying the UID to other submissions for the same chemical substance to provide meaningful information to the public without compromising trade secrets. According to EPA, these approaches are intended to give the “greatest possible effect” to the language of Section 14(g)(4) while maintaining the “EPA-approved confidentiality” of certain chemical identities. EPA requests comment on applying the UID to all submissions containing a particular chemical substance and on these alternative approaches, as well as suggestions for other approaches.
In the first alternative, Section 14(g)(4)(C) may be read as instructing EPA to ensure that any non-confidential information received by EPA concerning a confidential chemical substance should identify the substance using only the UID, so long as the confidential identity remains protected from disclosure. In this way, according to EPA, the public (including other companies) could identify the various submissions concerning a particular chemical, but could not identify the specific chemical. EPA notes that the need to treat information not claimed as confidential as such could be viewed as inconsistent with policy to limit CBI protection. It also presents implementation challenges, including requiring EPA to screen carefully incoming, non-CBI submissions against its list of confidential chemical names, and to treat as CBI information for which no such claim was made. Further, EPA states, the facts of a specific case could affect whether the original, unaltered, non-CBI submissions could be prevented from release pursuant to a Freedom of Information Act request.
The second alternative that EPA considered is to apply UIDs, once assigned, to other submissions concerning that chemical substance, but only those that are submitted by the same person/company. Submissions concerning the same substance, but made by a different company, would be assigned a different UID. EPA would not apply the UID to or associate it with non-confidential information if the effect would be to reveal the identity of an approved confidential chemical that is otherwise protected from disclosure under Section 14. According to EPA, the public would be able to link some submissions on the same chemical, but not necessarily all submissions. EPA states that this option “at least partly fulfills the intent to link information concerning the same substance, while maintaining the approved confidentiality claims of each submitter, until such claims are withdrawn, expire, or are subsequently denied by EPA.”
As noted, the notice is thoughtful and well written. EPA is to be commended for raising significant issues, especially the potential conflict that could arise if a single UID is used for a substance. We offer a few suggestions below, based on our extensive understanding of TSCA and the complexity of the issues this topic involves.
First, we suggest that EPA not incorporate the year that the claim was approved in the identifier. While it is important for EPA to be able to track claim expiration dates, the year the claim was approved will be a poor indicator of the status of the claim at any particular time. Claims may be renewed in perpetuity, so a reviewer would always need to check the current expiration date to determine when a claim is due to sunset. Conversely, a claim could be withdrawn shortly after the UID is assigned, such as when a valid CBI claim for chemical identity in a PMN is withdrawn in a NOC. The expiration date should be in the metadata tied to the UID, not part of the UID itself.
Second, we agree that language in TSCA Section 8(b)(7)(B) implies that the UID and accession number are separate identifiers for a substance. Also, UIDs must be assigned to a substance when EPA reviews and finds the CBI claim valid (or as discussed in the next item, at the time that the substance is submitted) and many valid claims will exist in submissions that do not relate to NOCs. In fact, a minority of substances with valid CBI chemical identities are subjects of NOCs (historically less than half of PMNs proceed to NOCs; other submission types, such as LVEs and Section 8(e) submissions are never listed on the Inventory). On the other hand, going forward, EPA may elect to use a UID to represent substances listed on the confidential portion of the Inventory, rather than assigning an accession number to the Inventory listing of a substance that also has a UID; we do not find any prohibition to using a UID in place of a traditional accession number.
As a practical matter, EPA may find that it is more expeditious to assign a UID upon submission of a CBI chemical identity rather than upon determination of a valid CBI claim. The statute requires assignment of a UID for approved CBI claims for chemical identity, but does not prevent assigning UIDs in advance of approval. In this way, all information submitted related to the specific substance from the specific claimant can immediately be tagged with the UID and therefore tracked together. If a UID is assigned and the CBI claim is determined to be not valid, the UID can be associated with the specific non-CBI identity in the same fashion that would apply upon expiration or withdrawal of a previously valid CBI claim.
Third, EPA is to be commended for recognizing the conflict that arises if a single UID is assigned a substance that may be submitted by multiple entities. Section 14(g)(4)(A) can be read to mean that the UID applies to “each specific chemical identity for which [EPA] approves a request for protection.” The language does not state “each specific chemical substance.” If 14(g)(4)(A) is read to mean that valid CBI chemical identity includes both the specific name of the substance and the validity of the CBI claim which is unique to the claimant, EPA should assign separate UIDs to each combination of substance and claimant as proposed in the second alternative described in the notice. This avoids inadvertent disclosure of valid CBI by linking a submission of non-CBI information for a substance with a submission that has a valid CBI claim. Assigning a UID to a particular submitter also facilitates EPA’s obligation to track, review, and sunset CBI claims by that submitter.