Chemicals on the Tracks

The Norfolk Southern Train Derailment: A Disaster but Not a Surprise
By Laura Rabinow

The Norfolk Southern train derailment on February 3, 2023, near East Palestine, Ohio, and its impacts on local residents surprised and concerned many people across the country. But, as a disaster, and as a toxics exposure case, this crisis is unfolding in a largely predictable and patterned way.

The freight train had 150 rail cars carrying a wide variety of cargo. Twenty of those cars contained hazardous chemicals, 11 of which were derailed. Five of the eleven cars were carrying vinyl chloride, with a number of other cars carrying different toxic chemicals.

Following the derailment, the state ordered a period of mandatory evacuation of nearby residences during which regulators conducted a controlled release or burning of the vinyl chloride in an effort to avoid an explosion. Testing by federal regulators later determined that “since the fire went out on February 8, EPA air monitoring has not detected any levels of health concern in the community that are attributed to the train derailment,” and based on that data, Ohio lifted its evacuation order. As residents moved back into their homes, however, they continued to report concerning health and environmental impacts like strong odors, rashes, and headaches. More recently, residents have been newly diagnosed with illnesses such as bronchitis that may be related to the incident.

What is Vinyl Chloride?

Vinyl chloride is primarily used to produce polyvinyl chloride (PVC), a ubiquitous plastic. PVC was first produced in the 1930s and began to be used more broadly in consumer products in the 1960s, including water and sewer pipes, wiring and cables, shrink wrap, car upholstery, windows, and medical equipment. There were earlier indications of negative health impacts on workers at plants that manufactured PVC, but manufacturers worked to delay the public discovery of these impacts and subsequent regulation of vinyl chloride until the 1970s. Health impacts related to acute exposure to vinyl chloride in humans include dizziness, drowsiness, and headaches, while in animals it has effects on the central nervous system, kidneys, and lungs. More chronic impacts to humans include liver damage, central nervous system effects, associations with increased incidence of birth defects for exposed pregnant women, testicular damage and decreased male fertility, and liver cancer.

The dissonance between what residents continue to experience and what testing efforts have—and perhaps more importantly haven’t—shown, may be surprising and concerning to many. But, it is a common example of how our chemical regulatory systems function—and don’t function—in these kinds of cases.

Disaster studies scholars have long observed how events like the train derailment near East Palestine aren’t atypical, or momentary. They have pointed out how these events reflect “long disasters” whose roots and impacts extend over time and “normal accidents” that occur as an inherent result of complex systems and that “grow with the complexity of the system, including the complexity added by the safety features.” Jennifer Homendy, the chair of the National Transportation Safety Board, recently stated that while there was no evidence that the train’s crew “did anything wrong” in this particular instance:

… the derailment was 100% preventable…. We call things accidents—there is no accident. Every single event that we investigate is preventable.

Homendy’s statement reflects a key point: no wrong action needs to be taken by individuals immediately preceding the event for this kind of disaster to happen.

The failings that contributed to this derailment and many other related disasters—such as the chemical contamination in and around Parkersberg, West Virginia, by perfluoroalkyl substances, the BP oil spill in the Gulf of Mexico, and the derailment of a train in Paulsboro, New Jersey, also carrying vinyl chloride—are built into the structure of the country’s broader regulatory and industrial systems and their normal functioning. While the derailment and the experiences of residents in and around East Palestine may be shocking, they serve to highlight a number of existing and well-documented limitations of our regulatory systems.

The one-by-one approach to establishing standards for chemicals contaminants in the United States limits what we know about the potential impacts of a disaster. When contaminants interact there can be unknown or unstudied cumulative and synergistic effects.

The US’s chemical-by-chemical approach to environmental health and safety regulation has meant that our regulatory systems at both the state and federal levels are not structured to address potential cumulative and synergistic effects that toxic or hazardous chemicals and other contaminants may have. Cumulative effects generally refer to those that work additively or multiplicatively together, while synergistic effects refer to combined effects that are greater than the combination of individual effects. Research has at times shown significant impacts when hazardous substances interact with each other or with complex and dynamic ecosystems. Greater attention has been called to these cumulative and synergistic effects in environmental health and justice research that has found evidence of uneven distribution of harms and the layered or cumulative effects of exposures that are placed disproportionately on Black, Brown, and poor communities.

More recent regulatory actions by the EPA have begun to acknowledge and address this uneven distribution and cumulative burden. Advancements include updates to the agency’s environmental justice mapping tool EJScreen and National Research Programs, as well as by HHS, CDC, and others. But legal scholars and environmental regulators have noted ways in which current federal environmental law still “provide[s] no framework for addressing cumulative risk.” The inability of our environmental regulatory systems to effectively account for these types of cumulative and interactive risks and their unequal distribution can lead to dangerous ignorance about disasters. The recent Ohio train derailment reflects how regulatory policies and practices, and the ways in which they are co-shaped by industry, produce not only knowledge but ignorance, including gaps in and insufficient research related to cumulative impacts, and unseen science that is kept secret by industry or claimed as confidential business information.

Environmental standards do not only reflect health impacts, but technological capacity and decisions about ensuring production.

In the mid-1980s, the country’s approach to environmental regulatory policy began shifting from simply protecting the environment from pollution and toward more risk assessment, attempting to answer questions such as “how hazardous is this pollution exactly?” One challenge in moving toward this risk assessment model, however, is the technological constraints in the ability to test, monitor, and mitigate potential pollutants at a given level and to do so at scale. This technological capacity informs and limits what standards it is possible to enforce, short of compelling those regulated to develop or fund necessary technologies. These constraints can also result in a gap between the health-based level that is determined to be protective of the public and the level that is determined to be enforceable through regulatory standards.

When establishing standards, regulators also make decisions that weigh the “chances of harm against the need to ensure the flow of commerce.” In the case of the Norfolk Southern train derailment in East Palestine, the EPA has historically recognized this gap with respect to vinyl chloride and has made the decision not to prohibit its production. As the notice in the Federal Register by the EPA for the initial regulation of vinyl chloride in 1975 states:

The standard was designed to minimize the health risks associated with VC [vinyl chloride] by requiring reasonable control measures. As stated in the preamble to the proposed standard (40 Fr 59532, December 24, 1975) there is no known threshold level of effects for VC. Therefore, the only approach that would eliminate health risks associated with VC would ban its production and use. This approach was not selected.

The California Air Resources Board has similarly noted with respect to the state’s ambient air quality standard for vinyl chloride, set in 1978, “the level of the standard, 0.01 ppm as a 24-hour average was chosen because it was the lowest level that could be reliably measured at the time the standard was promulgated, and could thus function as an enforceable exposure limit.” Despite the recognition by the state that, “as a carcinogen, no level of exposure to vinyl chloride is considered as being completely safe and without risk,” this standard continues to be in place—though recent regulatory efforts to shift that have been noted.

The regulatory systems for chemical products are largely permissive ones that have placed the burden to prove harm from environmental toxics exposures on impacted communities and regulators after the fact.

Demonstrations of the degree to which an impact can be causally attributed to a particular contaminant through a specific exposure pathway are especially difficult from a legal standpoint. Affected parties trying to address impacts often face a lack of baseline environmental health data for communities, and toxicological data or studies for many chemicals and their interactions may not be available. Since 2016, there have been significant changes to the Toxic Substances Control Act of 1976 (TSCA)—the nation’s landmark chemical regulatory statute—that address the degree to which our regulatory system is a permissive one. But only as of late are we beginning to grapple with how to effectively implement these changes: evaluating earlier exemptions given for chemicals predating TSCA, the claiming of confidential business information, significant new use rules, and pre-manufacture notice requirements or low volume exemptions—provisions that have allowed chemicals to come onto or remain on the market with less oversight or knowledge about their potential harms.

Crises like these reflect that chemical contaminants and the risks posed by them could be better understood as complex sets of technical, environmental, regulatory, and social systems. Considering the risks in each of these systems or their components independently fails to recognize the way interconnected systems impact each other or adequately assess their combined risks.

A policy response to this crisis based on an a framework of complex and interrelated technical, environmental, regulatory, and social systems would include the collective consideration of the regulation of wheel bearings on rail cars; the detectors used to alert trains of a potential derailment, the thresholds at which they are set, and the authority of railroads to set them; an evaluation of cost-cutting measures that have been put into place that may make a train heavier and put more wear and tear on those mechanisms or alter the timing and placement needed for detectors to be effective. It would also include the consideration of safety requirements related to the staffing levels of the train’s crew and, for train car tanks carrying hazardous materials, whether or not and at what threshold a train as a whole is labeled highly hazardous. Related regulatory requirements and responsibility for how and where the waste produced from those hazardous materials is disposed of would also be considered.

In response to events like the Norfolk Southern derailment, efforts by regulators and industry to make a disaster manageable can obscure potential harms.

Regulators and responsible parties responding to toxics disasters work to monitor and test for contaminants. In doing so, they define the environment, its components, its contaminants, and by broader application the disaster as a whole. Working through the limitations of existing regulatory standards and technical capacities discussed above, efforts to test and monitor are designed to render complex socio-technical systems and ecosystems governable by making them discreet and measurable. This can be beneficial from a crisis response perspective, because it works to define the bounds of the disaster and make its impacts knowable and manageable. However, the establishment of boundaries can also place the ability to manage crises and maintain authority in tension with the recognition of further harms, unknown information, and the limits of regulatory knowledge.

The dissonance between what residents continue to experience and what testing efforts have—and perhaps more importantly haven’t—shown, may be surprising and concerning to many. But, it is a common example of how our chemical regulatory systems function in these kinds of cases.

Lingering Questions

These facets of the country’s chemical regulatory systems all work to make environmental health disasters like the train derailment in East Palestine, Ohio, ‘normal’ events that we can expect communities will continue to experience. As of now, authorities have concluded that levels of contaminants there do not exceed current standards. But, as more is learned about the exposures in Ohio and the experiences of health impacts by nearby residents, these facets of our regulatory system point towards some important and more immediate questions.

  • What are all of the layered and interconnected social, environmental, and technical systems that have contributed to this (and other) disasters?
  • Many news accounts have state that authorities found no levels of contaminants that exceeded standards, but there is limited information on these measurements. To better understand this declaration, authorities must answer a number of questions:
    • What specific standards or guidance values are being referred to for each contaminant tested across different environmental media (air, soil, drinking water, non-potable water) and different parts of the contaminants’ ‘life-cycle’ (manufacture, transport, waste)?
    • How do each of these standards or guidance values differ from underlying health-based values?
    • What processes for measuring contaminants are being used, how specific to each contaminant are they, and what are the related technological limitations, particularly with respect to health-based values? What are the longer histories of these standards, how they came to be, and who has shaped them?
  • What don’t we know about these contaminants and potential harms? Specifically, to what extent do we understand the potential cumulative or synergistic interactions of the different contaminants present in a dynamic ecosystem?
  • What baseline local data is or isn’t available to compare contaminant levels before and after the derailment to allow for the testing of environmental health impacts?
  • And what resources are being made available to residents to assess and address their personal and environmental health impacts?

These are more immediate and localized questions that need to be clarified. As policymakers and industry put forth proposals for regulatory reform, the existing limitations of what we can know must be considered. These reforms must address the interconnected and complex systems that continue to result in such disasters.


Laura Rabinow is deputy director of research at the Rockefeller Institute