EPA simply must do better on transparency and chemical data access

Richard Denison, Ph.D., is a Lead Senior Scientist.  Lindsay McCormick is a Research Analyst.

[This is Part Two of a two-part series.  Here is a link to Part One.]

Under this Administration, EPA has taken some significant steps toward reversing decades of passivity and secretive practices that evolved under the Toxic Substances Control Act (TSCA) when it came to transparency in decision-making and providing access to chemical information it obtains or develops.

Several initiatives undertaken through what EPA originally termed its enhanced chemicals management program have developed and laid out clearer policies and procedures in areas such as:  chemical prioritization (leading to its Work Plan Chemicals Program); enforcing limits on and reviewing confidential business information (CBI) claims asserted by industry (leading to its declassification of hundreds of previously hidden chemical identities and health and safety studies that companies had illegitimately claimed confidential or no longer merited protection from disclosure, but that EPA had not bothered to review or challenge before now); and EPA’s regulatory efforts to reduce risks from exposures to toxic chemicals (leading to its Action Plans on high-concern chemicals and proposed follow-up activities for work plan chemicals where assessments – the first completed in decades – have identified significant risks).

EPA has also developed new databases and tools to provide greater access to chemical information in its possession and regulatory decisions and supporting documents it develops; these include the Chemical Data Access Tool and ChemView.

All of these efforts are still very much works in progress but hold significant potential to improve transparency, information access and risk reduction.

But sometimes the Agency does something that makes clear just how far it still has to go in these respects.  Unfortunately, a case in point is its recent effort toward assessing risks of a cluster of flame retardant chemicals, the brominated phthalates, some of which are in wide use and are showing up in everything from house dust to dolphins.  

We just blogged about the comments we submitted last week on EPA’s “Problem Formulation and Data Needs Assessment” for the brominated phthalates cluster [hereafter “Data Needs Assessment”].

Our comments also addressed a serious lack of transparency in the EPA documents, which is the focus of this blog post.  Specifically, our review turned up:

• unchallenged unlawful confidentiality claims for health and safety information;
• inappropriate withholding of non-confidential identifying information for cluster members;
• public access only to summaries of industry studies, the quality and completeness of which is impossible for the public to judge and may not have been examined by EPA.

The result of these deficiencies is that the public lacks access to all or even most of the information on which EPA is relying in deciding not to conduct risk assessments of these chemicals at this time.

Below we summarize our comments on each of these concerns.

Unchallenged unlawful confidentiality claims for health and safety information

EPA repeatedly cites, as a basis for statements in its documents, an August 27, 2012 letter from Chemtura Corporation to EPA that lists health and safety studies the company was transmitting to EPA on one member of the cluster, TBB (CAS# 183658-27-7).  The letter notes that it enclosed a “CD containing CONFIDENTIAL studies” on TBB (emphasis in original).

EPA’s Data Needs Assessment includes other references to “confidential” health and safety studies or data, information from which it excluded from its documents.  Yet TSCA expressly precludes health and safety studies and their underlying data from being protected from public disclosure as confidential business information (CBI), with two narrow exceptions for information that would reveal process or mixture portionality (see section 14(b)).

Despite this, EPA never challenged the company’s blanket confidentiality claim covering these studies and as a result has not made the studies available or provided any means of accessing them in the present documents.  Until we raised the issue with EPA staff, it appears Chemtura’s illegitimate claim had gone unnoticed.

EPA should immediately reject the confidentiality claims asserted by Chemtura, which should never have been allowed in the first place, and provide public access to these health and safety studies.  We are hopeful that our raising these concerns with EPA staff will lead to prompt rectification of this situation, and that procedures are put in place or enforced to ensure this does not happen again.

Inappropriate withholding of non-confidential identifying information for cluster members

When EPA released its Problem Formulation and Data Needs Assessment last August, it listed two members of the brominated phthalates cluster only as “Confidential A” and “Confidential B.”  No further identifying information was provided in the cluster documents.  This was wholly unacceptable.

TSCA and its implementing regulations require that, where the specific identity of a chemical is deemed confidential business information (CBI), the CBI claimant is to provide a structurally descriptive generic name and EPA is to use that name to publicly identify the chemical (see TSCA Section 5(d)(2); and 40 CFR §§ 720.85(b)(5), 720.90(c) and(d)).  Once manufacture of such a chemical commences, the generic name and a tracking number assigned by EPA are to be placed on the non-confidential portion of the TSCA Inventory.

Yet EPA did not even provide the generic names and tracking numbers – which by definition are NOT confidential – for these two cluster members.  Importantly, those generic identifiers are already public, including on the TSCA Inventory, but their exclusion from the Data Needs Assessment made it impossible for the public to find even the limited information available on them because EPA replaced them with “Confidential A” and “Confidential B.”

EDF raised our strong concerns about this approach directly to EPA staff.  We were told that EPA had unilaterally decided to withhold the generic names and accession numbers for these two chemicals, based on an argument that associating even the generic names of these two chemicals with this cluster might somehow reveal their confidential identities.  That argument appears far-fetched – how would generic identifiers allow someone to glean the specific identities? – and it flies in the face of the very purpose of a generic name – which is required to be used publicly to refer to a chemical for which the specific identity is confidential and cannot be disclosed.

In response to our pressing on this, EPA contacted the company(ies) and was told, we understand, that they did not object to the generic names being made available.  After several additional inquiries from us, EPA disclosed the generic names and Premanufacture Notice (PMN) case numbers for these two chemicals, through a memo it released on December 7^th, 2015.  EDF was then able to use these identifiers to locate Federal Register notices that announced receipt of their respective PMNs and of their subsequent notices of commencement of manufacture (NOCs).  The limited information we were able to find is below.

While inconspicuous in EPA’s documents, the Agency issued a TSCA Section 5(e) consent order in 2009 on Confidential A as a result of EPA’s review of the chemical prior to its first manufacture.  Once EPA made public the PMN case number for Confidential A, as described above, we were able to obtain a heavily redacted version of this consent order.  The consent order indicates serious concerns, based on data from chemicals structurally similar to Confidential A, regarding liver and kidney toxicity; potential persistent, bioaccumulative and toxic (PBT) characteristics; and potential carcinogenicity due to formation of byproducts during combustion of consumer products in municipal incinerators.

In the consent order, the Agency concluded that “the information available to the Agency is insufficient to permit a reasoned evaluation of the human health and environmental effects of the PMN substance,” and imposed testing requirements that would only be triggered if production reached a specified (redacted) volume.  EPA’s Data Needs Assessment indicates the triggering volume for testing has not been reached.

Confidential B, which is structurally related to Confidential A and other cluster chemicals, was apparently not subject to a consent order or any testing.  It is now a high production volume (HPV) chemical, produced at more than 1 million pounds annually.

Information EDF found on Confidential A and Confidential B:

Links to the respective receipt notices for the PMNs and NOCs, and to the Consent Order, are as follows:

• Confidential A/P-04-0404: PMN Receipt, Notice of Commencement receipt, Consent Order
• Confidential B/P-96-0965: PMN Receipt, Notice of Commencement receipt

The data needs for Confidential A and Confidential B are completely obscured in EPA’s documents – an omission we can only presume is related to EPA’s inappropriate withholding of even generic identifying information on these chemicals.  EPA states only:  “The nature of Confidential A and Confidential B cannot be disclosed.  Data Gaps and Data Needs should be considered for both reactive and additive uses.”  We fail to understand how identifying data needs for these chemicals would constitute disclosure of legitimate CBI.

Had we not pressed EPA to release generic identifying information on these chemicals, the public would not be able to discern that:

• they were new chemicals that went through the PMN process;
• they are now listed on the TSCA Inventory and are in commerce;
• for one of the chemicals (Confidential A), it is subject to a consent order EPA negotiated based on significant toxicity concerns, which includes yet-to-be-triggered testing requirements;
• for the other of the two (Confidential B), it apparently had no conditions placed by EPA on its commercialization, because no associated consent order or significant new use rule appears to have been issued.

In sum, we are very concerned that EPA unilaterally withheld non-confidential information about these two chemicals from the public, based on an argument that is wholly inconsistent with TSCA and its implementing regulations, which expressly provides for the public disclosure of generic identifiers where specific chemical identity is claimed to be CBI.  EPA should publicly describe steps it will take to ensure withholding of such information from the public does not occur again.

Public access only to summaries of industry studies, the quality and completeness of which is impossible for the public to judge and may not have been examined by EPA.

When we made inquiries about the “confidential” or missing studies referred to in Chemtura’s 2012 letter, EPA indicated that “robust summaries” of all of them were provided in a 2004 risk assessment conducted by the Australian government on a mixture containing some of the chemicals.

While such summaries of many of the referenced studies are indeed in an attachment to the Australian assessment, there are not summaries for at least four.  After considerable further investigation, we were able to discern that these four studies were mandated to be conducted by consent orders EPA negotiated with Chemtura, three of them several years after publication of the Australian report.  No robust study summaries appear to be publicly available anywhere for these studies.

Even for those studies for which summaries are available through the Australian report, it is entirely unclear whether those summaries, which likely were prepared by Chemtura, were reviewed against the full studies for quality and completeness by the Australian government agency or by EPA.  It appears most likely that the summaries were prepared by the manufacturer and simply passed along by the Australian government agency via their attachment to its assessment, and now by EPA via its reference to the Australian report.  It is not clear the extent to which EPA itself has access to the full studies, as that is not indicated anywhere in EPA’s Data Needs Assessment.

This is an unacceptable situation:  It violates the clear language of TSCA, which precludes health and safety studies and their underlying data from being protected from disclosure.  It obscures the extent to which EPA is relying on the manufacturer’s conclusions as to the findings of the studies.  And it denies the public any ability to independently assess the conclusions EPA draws as to the adequacy of available data or the data needs EPA has identified.

EPA needs to act promptly to rectify this situation:

• It should promptly reject confidentiality claims asserted by Chemtura, and provide public access to these health and safety studies.
• If EPA intends to utilize robust summaries as a primary means of providing initial public access to health and safety studies, it must, at a minimum:
  • itself prepare the summaries, or carefully review summaries prepared by another entity (whether Chemtura or another government) against the full studies and vouch for their accuracy and completeness;
  • make clear that such studies are not eligible for CBI protection and make the full studies publicly available upon request; and
  • if EPA does not have access to a full study it cites, EPA must:
    • make that abundantly clear;
    • promptly seek access to the full study, using all available authorities; and
    • clearly identify the preparer of any summary of the study to which EPA refers, and in the case of summaries prepared by a company making or using the chemical, take into account the potential for bias in any decision EPA makes that relies to any extent on such a summary.
  • It must put in place procedures and enforce practices internally to ensure this situation does not arise in future Work Plan chemical reviews.

Conclusion

We understand that the Work Plan Chemical Program is relatively new and that the current Problem Formulation and Data Needs Assessment is the first to be published through this program.  However, we are concerned that the early products of this program will be precedent-setting to some extent.   We urge EPA to strive to improve transparency, clarity, and data availability regarding these chemicals as they move through review under the program, and also to recognize the need for improvement in conducting reviews of other Work Plan chemicals in the future.