After a Norfolk Southern train carrying industrial chemicals derailed in East Palestine, Ohio on Feb. 3, 2023, more than 115,000 gallons of vinyl chloride leaked into the air, water, and soil. A controlled venting and burning followed to prevent the potential for an uncontrollable explosion.
An East Palestine resident who lives near the chemical-contaminated creek was told by Norfolk Southern Railroad, “Based on air testing, it’s safe to move back home.” Still concerned, she insisted on water and soil testing, too. Based on those results, a toxicologist said her house was unsafe to live in. She’s outraged and doesn’t trust Norfolk Southern.1
Many others continued to question the air, water, and soil testing and whether it was safe. Answers are often technical, vague, or unsatisfying, making it more difficult than it should be to understand. Let’s sort through what monitoring and testing are, how they differ, and how the data can inform decision-making.
Monitoring versus sampling: what are the differences?
Monitoring is performed using meters and can provide useful data in real time. Sampling refers to the process of collecting samples (e.g., air, water, soil, or other) in containers or on sampling media, and sending them to a lab for various analyses. Monitoring and sampling are complementary and often done in conjunction for a fuller picture of various environmental conditions as they change. Both are typically used across a wide variety of situations—in routine conditions and, like in this case, emergency scenarios.
Multi-gas meters with built-in photoionization detectors (PIDs) are the main types used for air monitoring (e.g., MultiRAE and AreaRAE). They detect oxygen levels (normal air is 20.9 percent O2), combustible gases, (i.e., the lower explosive limit or LEL), and a few specific toxic gases (e.g., carbon monoxide (CO) and hydrogen sulfide (H2S)). These are perhaps the most common meters used in field investigations, especially in chemical spills or suspected gas leaks. Being able to quickly know levels of O2, the LEL, CO, and H2S as they change in real time is a definite boon. Add in using a wide spectrum analyzer like a PID for volatile organic compounds (VOCs) and one can learn a lot about the air where they are screening. Some will data log and can be used to record readings over time, but even with all these advantages, they are limited and don’t tell the whole picture.
Monitoring and sampling are complementary and often done in conjunction for a fuller picture of various environmental conditions as they change.
One issue is the need to both identify and quantify other chemicals. For the Ohio train derailment, the EPA added a detector for hydrogen cyanide and used another gas meter (UltraRAE) configured to detect benzene. Another meter used to collect data from the site is a tape-based gas detector (Honeywell’s SPM flex) configured to detect phosgene and mineral acids (e.g., hydrogen chloride).
Another issue involves trying to detect VOCs at levels in parts per million (ppm) or even parts per billion (ppb). PIDs are one tool but aren’t qualitative (i.e., they don’t define what each detected chemical actually is). For that, field personnel sometimes use portable analyzers.
In East Palestine, field personnel used a PID to monitor for VOCs. It detects VOCs with ionization potentials (IPs) less than its ultraviolet lamp’s IP (e.g., a PID with a 10.2 or higher UV bulb will detect vinyl chloride—IP is 9.99 eV). But a PID can’t tell the user which VOCs make up that total number in ppm. Vinyl acetate (IP of 9.19 eV) and butyl acrylate (IP of 9.6 eV) are also being screened with PIDs. These are very commonly used in a wide variety of applications including general indoor air quality, leak detection, emergencies and spills, or other odor-chasing events.
The EPA has also used a dust meter to determine particulate matter under 2.5 microns (PM2.5) and under 10 microns (PM10) from smoke. Particulate matter poses a health threat and can become serious during burning (e.g., wildfires). Add in chemicals being burned and the risk and complexities of assessment both increase. Particulate matter at those diameters and below is a good minimum for assessing PM quantitatively. Determining the smoke’s chemical make-up is better done with air sampling and lab analyses.
Monitoring was done by the US EPA, Ohio EPA, Ohio’s 52nd Civil Support Team (of the National Guard), and Norfolk Southern and/or their contractors. The latest rounds of air monitoring (done on Feb. 10 and Feb. 16-17, 2023) were normal (O2), non-detected (LEL and toxic gases), and below limits of concern (PM2.5 and PM10).2 PM2.5 and PM10 monitoring may have stopped due to lack of smoke, without which there is little added risk and so monitoring isn’t justified.
Sampling and analysis
Air, water, and soil samples are continuing to be collected and analyzed by labs. The EPA is collecting air samples using Summa canisters, which collect air over time in a specific area. The air is analyzed for a large suite of VOCs, such as those in analytical methods (e. g., EPA’s TO-15). These results will supplement the air monitoring, including potentially low levels of specific chemicals.
Field personnel and testing labs will continue to play key roles in any future hazardous materials incidents or chemical release events.
Water samples are typically collected from various sources, such as surface waters (lakes, ponds, creeks, rivers, etc.), public water supplies, private wells, homes, and newly drilled environmental wells into local aquifers. The EPA detected vinyl chloride in samples from Sulphur Run, Leslie Run, Bull Creek, North Fork Little Beaver Creek, Little Beaver Creek, and the Ohio River. Ohio EPA posted their water sample analyses3 and summarized it as very low levels for earlier results and undetected for more recent results for treated drinking water (e.g., at or near the method detection limit or MDL, which is effectively zero), especially as it dilutes moving downstream in flowing creeks and rivers. Concerns will remain for impacts to private wells and public water supplies over time. Environmental monitoring wells are being or will be drilled by Norfolk Southern contractors. Those results will also provide further clarity of groundwater contamination or purity.
Soil samples are collected at or near the contamination and release site where it was vented and burned. The EPA criticized Norfolk Southern for not removing more contaminated soils. As a result, Norfolk Southern has committed to removing the impacted soil under the train tracks, including the added work of first removing the rails and tracks and replacing them after the impacted soil is removed and clean fill replaces it. In the future, there may be added emphasis for ongoing soil sampling, results, and actions.4
Limitations and variables
Monitoring and testing are limited by several variables. First, readings are often called snapshots in time—each is a single data point. This makes it hard to generalize or extrapolate, especially about risks to the public. Second, authorities focus on legal limits set by agencies, scientific bodies, and others. They often struggle to answer concerns in layperson’s language. Plus, testing chemicals is complicated when burned—they’re a complex mix including byproducts of combustion. This can require both quantitative monitoring (i.e., PM10 and PM2.5) combined with qualitative analyses for VOCs.
Some people are sensitive to chemicals at lower levels. They can produce adverse effects, like eye irritation, headaches, nausea, or odors, even at levels below the regulatory limits. If you smell it, it smells bad, or you’re experiencing headaches, nausea, etc., these are all signs of negative effects. It’s a struggle to reconcile experiencing these effects with being told that results are below levels of concern set by the government.
Over time, a plume may move within an aquifer. Vapors can migrate through soil into homes and elsewhere. This subsequent contamination, even years later, can cause significant problems for people in affected homes or buildings. When you’re used to clean air, any contaminated air, off-putting smells, and health concerns are problems that need addressing.
What happens next?
Spill response, clean up, and monitoring will continue until the EPA is satisfied. This is likely to include both results that are at least under levels of concern (if not limits of detection) plus a lessening or lack of pressure by officials, locals, and others. Not all residents and members of the public will be satisfied with the response, especially those whose trust was lost. There will likely be ongoing research studies of longer-term effects on people, animals, vegetation, and levels in air, water, and soil. People will be concerned by any illnesses or deaths that appear out of the ordinary, or if a few are similar. Despite the aftermath and contamination of this train derailment, hazardous chemicals will continue to be moved by rail through cities and towns across the country. Field personnel and testing labs will continue to play key roles in any future hazardous materials incidents or chemical release events. Regardless of their affiliation or role, industrial hygienists, field geologists, hazmat technicians, lab analysts, and others will be called upon to monitor and test the air, water, and soil to ensure it’s safe.