EPA Whistleblowers Provide New Evidence of Ongoing Failure to Assess Dangerous Chemicals

Managers in the EPA’s New Chemicals Division have refused to assess the risk of cancer and other harms from chemicals deemed to be "corrosive."

Photo illustration: The Intercept/Getty Images

A group of whistleblowers has provided evidence that the Environmental Protection Agency has not adequately assessed the health risks posed by several new chemicals on the grounds that they are corrosive. Managers in the New Chemicals Division have repeatedly and incorrectly used the idea that a chemical may cause irritation to the eyes, skin, and respiratory tract as an excuse to avoid assessing the risk of other harms it may cause. Those harms include cancer, miscarriage, and neurotoxicity, according to the whistleblowers, who work as health assessors in the division. In some cases described in a complaint that the whistleblowers shared with The Intercept and will soon submit to the EPA inspector general, the risks were calculated, found to be significant, and later deleted from official documents.

The theory behind the EPA’s decision not to calculate the risk of repeated exposure to certain corrosive chemicals — or to remove information about those risks — is that after the unpleasantness of the first exposure, people will avoid contact with the chemical in the future. But according to the group of health assessors who have been providing The Intercept with insider accounts of corruption in the EPA’s chemical assessment process over the past year, this logic is flawed for many reasons. Perhaps the most significant problem is that people may not actually experience or notice any effect from an initial exposure — either because the chemical has been incorrectly deemed corrosive or because it is corrosive only at concentrations higher than the levels to which people are exposed. Neither circumstance has any bearing on whether the chemical presents other risks. Workers may also be forced to have repeated contact with chemicals to stay employed.

“They’re trying to say that if a chemical is corrosive, people will just avoid it, which is nonsense,” said William Irwin, an EPA toxicologist who is among the small group of scientists that has been calling attention to flaws in the agency’s assessment of new chemicals. “That’s not the way things work. People have to do their jobs.”

In some cases, according to the complaint, high-level EPA staff members have argued against the calculation or mention of chemicals’ systemic risks based on the idea that workers will be protected from their dangers because they will wear personal protective equipment. But according to Sarah Gallagher, one of the whistleblowers and a human health assessor in the EPA’s Office of Pollution Prevention and Toxics, that’s not a safe assumption.

“PPE gives some protection, but it is often not complete, and some chemicals can sneak through gloves. In order to get the best protection, you need to ensure that the chemical does not readily break down the gloves so that they lose protectiveness,” said Gallagher. “There are tests for that, but my understanding is that the New Chemicals Division is not ensuring the right gloves are used for new chemicals.”

In an emailed response to questions from The Intercept, Timothy Carroll, the EPA’s deputy press secretary, confirmed that, for some corrosive chemicals, assessments do not include health effects beyond the corrosivity. “When occupational exposures from a chemical EPA finds to be highly corrosive are expected to occur, meaning it would immediately burn the skin/lungs, chronic effects that could occur from prolonged exposure to that chemical would not be expected to occur and therefore the risk assessment would quantify the corrosivity risk alone.” But in other cases, the EPA does consider other health risks, according to Carroll. “In cases where a chemical is not expected to be sufficiently corrosive to support an assumption that prolonged exposures would not occur or there are non-occupational uses that are described in the [pre-manufacture notice] or reasonably foreseeable that could result in prolonged or repeated exposures, EPA would characterize risks beyond corrosivity.”

Carroll’s email also said that the agency had recently changed its approach to addressing personal protective equipment, such as respirators and gloves, in chemical assessments. “In order to ensure that neither prolonged nor repeated exposures to the chemical would occur when the chemical begins to be manufactured, EPA would require the use of PPE or other measures during the risk management phase for that substance,” said Carroll. “This is a shift away from the last Administration’s policy of assuming that workers always have access to and correctly use PPE. Under the Biden Administration, in conducting initial risk assessments for new chemicals, EPA does not assume workers are utilizing PPE when calculating risks.” When the EPA does require PPE, it issues a special order that limits that production and use of the chemical, according to Carroll.

A Single Exposure

Another hole in what the whistleblowers have called the “burned-finger hypothesis” is that, for some chemicals, a single exposure can cause irreversible effects. So even if someone’s eyes or skin are irritated or burned on a first encounter, other serious harms could already have occurred. Consider a new chemical whose assessment was finalized in March: The company that submitted the chemical for review did not provide the agency with studies of its reproductive health effects. So, as is often the case, the health assessors had to resort to studies of structurally similar chemicals, or analogues, which often have similar health effects, to glean any information about the risks it might pose.

Studies of one analogous chemical showed that a single exposure could cause mice to “resorb” — or essentially miscarry — their fetuses. In fact, just hours after a single exposure to the chemical, more than 63 percent of pregnant mice experienced fetal death. Yet the risk assessment of the new chemical, which was finalized on March 7, did not mention the study or the possibility that the new chemical might cause miscarriage in humans. Instead, it noted that “the corrosiveness of the new chemical substance” limited repeated exposures — and found the chemical “not likely” to present an unreasonable risk to health.

While the assessment did not calculate or note the miscarriage risk on the grounds that the chemical’s corrosivity would cause people to avoid coming into contact with it, the whistleblowers also determined that workers exposed to it would probably not experience any skin damage — and thus were unlikely to realize that it was corrosive.

Like the vast majority of the other chemicals being reviewed by the EPA’s new chemicals program and described in this series, the information regarding that chemical was submitted to the agency as confidential business information. Although the health assessors know the product’s exact structure and name, they could suffer severe penalties if they made that information public.

Cancer Risk

The health effects that are omitted from assessment on the grounds that chemicals are corrosive include cancer. This was the case for a compound that the New Chemicals Division was reviewing in March 2021. There were clear reasons to suspect that the chemical, which is to be used to produce other chemicals, was a carcinogen. And information submitted by the company made it clear that workers and people living near the factories where it is used could be exposed. But the assessment, which has not yet been finalized, does not include the calculations of any systemic health risks, in part because the chemical was known to irritate the eyes and skin. Instead, the assessment notes that the “corrosivity of the new chemical substance may be protective of any potential systemic effects.”


EPA Withheld Reports of Substantial Risk Posed by 1,240 Chemicals

This chemical was also submitted to the EPA without any heath data, so the health assessors had to resort to analogues to glean information about the risks it might pose. When they did, they found that the structure alone raised serious concerns. Similar chemicals cause cancer, liver effects, and neurotoxicity. The whistleblowers then used a close analogue — a chemical called benzyl chloride that the EPA had deemed a “probable human carcinogen” in 2008 — to gauge the likelihood that the new compound might cause cancer. The results were alarming: One out of every 118 people exposed would be expected to develop cancer. In comparison, the EPA usually considers one cancer in every 100,000 people exposed the upper limit of acceptability.

But rather than using benzyl chloride to predict the carcinogenicity of the new chemical, without explanation the assessment relied on another analogue to gauge the new chemical’s cancer risk, as the whistleblowers point out in their complaint. That chemical had not been subject to a repeated-dose carcinogenicity study, which the agency requires for assessing the likelihood that a chemical will cause cancer.

If the EPA did not want to base its assessment of the new chemical on benzyl chloride, the analogue that had already been found to be a probable carcinogen, it could have asked the company to perform its own cancer studies on the new chemical. Or it could have searched for another appropriate analogue that had been assessed for its cancer-causing potential. Or, if it didn’t obtain adequate evidence of its safety, the agency could have prohibited the company from using the chemical. Instead, the latest version of the assessment deemed this chemical, too, “not likely to present an unreasonable health risk.”

Fundamentally Inaccurate

In March 2020, Gallagher, the human health assessor, found that another chemical presented risks to workers. Experiments showed that one analogous chemical caused rats to have tremors and behave aggressively. Another analogue caused reproductive effects in male rats and mice. Information about both hazards were included in a version of the assessment that was finalized on April 8, 2020.

But a month later, a manager in the New Chemicals Division created a new assessment. In this version, the information about the hazards had been deleted. Instead, in a section of the document headed “workers,” the document explained: “Risks were not evaluated for workers via repeated dermal exposures because dermal exposures are not considered likely due to the corrosivity of the new chemical substance.”

“It’s not just that we did the calculations. We did the calculations and found risks, and then they got rid of them.”

According to the whistleblowers, this statement is false. “It is intentionally misleading for EPA to put into a report that we did not calculate risk when we did,” said Martin Phillips, a chemist and human health assessor who works in the EPA’s Office of Pollution Prevention and Toxics. “It’s lying about what we did. It’s not just that we did the calculations. We did the calculations and found risks, and then they got rid of them and said that we didn’t calculate them. It’s fundamentally inaccurate.”

According to Public Employees for Environmental Responsibility, the organization that’s representing the whistleblowers, the statements may be a violation of the law. “I hope that the inspector general evaluates whether these false statement are violations of the criminal statute,” said Kyla Bennett, director of science policy for PEER. “EPA is not allowed to make knowingly materially false statements.”

To further complicate the assessment, a quick check of the pH of the chemical done by Gallagher revealed that it was neither acidic nor basic enough to cause skin damage. In other words, the chemical wasn’t corrosive after all. Gallagher repeatedly raised the issue with her colleagues after she made the discovery, but the assessment was not corrected. It was finalized on May 29, 2020.

In the emailed response to questions from The Intercept, the EPA’s Carroll wrote, “EPA is committed to ensuring the highest level of scientific integrity across the agency and takes seriously all allegations of violations of scientific integrity. Additionally, EPA is committed to fostering a healthy work environment that promotes respect between all levels of staff, supports work-life balance, provides for an open exchange of differing scientific and policy views, and achieves our mission of protecting human health and the environment. Where scientists identify a difference in scientific opinion, EPA has a transparent process that allows for expression, elevation, and resolution.”

The email went on to say, “The agency will fully cooperate with any and all future investigation by the Office of Inspector General.” The EPA inspector general is currently investigating numerous complaints previously filed by the whistleblowers.

An Ongoing Problem

For at least two years, the whistleblowers have repeatedly argued against the use of corrosivity to dismiss other health hazards — a strategy they say is in keeping with other EPA efforts to make dangerous chemicals seems safer than they are. Since The Intercept began reporting on their complaints more than a year ago, the EPA has taken several important steps to improve the regulation of new chemicals.

But according to the whistleblowers, the dismissal of serious health concerns with the mention of corrosivity continues. Just two weeks ago, Kyoungju Choi, a toxicologist in the New Chemicals Division, was asked to assess a compound. She noted that an analogue had developmental and reproductive effects on rats. But per the instructions of a senior staff member in her division, she was offered the option of dismissing these hazards because the chemical is corrosive.

“Then there would have been no other hazards,” said Choi. Although she felt pressure to dismiss the health concerns, Choi opted instead to lay them out in the document. While the assessment is still in draft form, she is hopeful that her warning will survive the EPA’s fraught assessment process and go on to protect workers and their children from harm.

Join The Conversation