Risk Communication Considerations for the West Virginia Elk River Chemical Spill

During a crisis, there are certain evidence-based principles and a basic risk management framework by which stakeholders should be guided.

Every day, we each analyze risk and perform cost/benefit analyses for even mundane tasks. For many people, this is particularly true when it comes to our health, from eating fatty foods to avoiding toxins. Risk assessment is also crucial on a larger scale when it comes to health, in situations where it is imperative to determine a localized or widespread threat to human health, provide information about how to best manage risks, and communicate with the public about the proper response to threats. When the threat is environmental, communication with the public becomes even more important, as environmental threats often affect large numbers of people. One such threat was the chemical spill of 4-methylcyclohexane methanol at Elk River in West Virginia in 2014, which left approximately 200,000 without water for over four days.

The problem was first discovered by residents in the area, who smelled the chemicals in the air. From there, accounts differ. The West Virginia American Water Company, who ran the intake, treatment, and distribution center where the spill occurred, said that an employee discovered the leak and cleanup began immediately, while the state Department of Environmental Protection (DEP) has said that the spill had clearly been happening for several days and had not been cleaned up when they came to investigate.

One reason for these conflicting accounts of how the risk was identified is that, regardless of whether the water company or DEP first saw the spill at the site, the residents knew about the spill before it was investigated. According to Eve Wittenberg, PhD, Senior Research Scientist in the Center for Health Decision Science and a professor at the Harvard T.H. Chan School of Public Health, “it complicates things when the public finds out about an incident first because there’s no opportunity for a thoughtful communication strategy to be developed. Stakeholders are more likely to speak off-the-cuff, and in this case, it exacerbated the lack of coordination among the agencies and companies involved.”

During a crisis such as this, there are certain evidence-based principles and a basic risk management framework that stakeholders should be guided by, although experts acknowledge that it is unlikely anyone will follow these principles exactly in a real-life situation. These principles include having the message come from a trusted source, ensuring the message is clear, and dealing with what people are actually experiencing in the moment. In the West Virginia case, these principles would entail messaging from an organization such as the Centers for Disease Control and providing clear instructions on whether or not the water is safe to drink, which is what residents say they were most concerned about.

Risk assessment is crucial to determining threats to human health.

However, this premise was not evident during the first press conference, which took place the day after the spill was discovered. During this press conference, Jeff McIntyre, the president of West Virginia American Water told the media and public that he “can’t tell you that the water is unsafe, but (I) also can’t tell you that the water is safe.” In addition to not providing a clear answer on the safety of the water in the area, this statement came after the water use ban, and therefore further muddled the message that the public was receiving.

“It can be very hard for agencies to assess what people are actually experiencing in the moment,” said Wittenberg. “Many times, people just want to be calmed, or they want to know what to do. Agencies need to be clear and tell the public what they should be doing, but in a way that’s consistent with what’s happening. In many cases, the people know what the right thing is to do and it’s important to acknowledge that.”

Furthermore, said Wittenberg, there is little cost-benefit analysis from those in charge during crisis situations. Those analyses are in the background during a particular situation, but they largely come into play further back in the risk management process, such as when setting guidelines for acceptable chemical exposures. However, in a case such as this, there is the underlying question of whether the set level which came out of the quantitative risk analysis is actually a reasonable level for real life – the Environmental Protection Agency might consider a certain chemical level safe for water, but would you let your children drink it? The cost-benefit analysis done by government or corporate stakeholders may come into conflict with the risk evaluation done by the public during a crisis situation, as it did in West Virginia.

After the West Virginia American Water Company unsuccessfully tried to mitigate the risk by filtering the water for about four hours, the DEP enacted a ban on using the water in the area of the spill, which included nine counties in the Charleston area. The approximately 200,000 affected residents were advised not to drink, bathe in, cook with, or wash with the water. Over the next three days, 169 people sought treatment for symptoms including nausea and rashes, while 700 contacted the state’s poison control center. In the end, 14 people were hospitalized.

Agencies need to be clear and tell the public what they should be doing.

After evaluating the risk to the public, both West Virginia and the federal government, which sent the Federal Emergency Management Agency (FEMA) to distribute potable water, declared a state of emergency. After five days, the state began gradually lifting the water use ban.

The long-term impact of the spill is difficult to assess not just in terms of environmental risk, but also in terms of health risk and psychological risk. Experts told the media in the days and weeks following the spill that there has been little research on the toxicology of 4-methylcyclohexane methanol despite its use in approximately 25 percent of coal processing plants in West Virginia. According to analysis by the independent Environmental Working Group, the safety level used to lift the drinking water restriction was based on an unpublished animal study and is thus a very crude indicator for human health.

In addition, the constantly changing message and lack of coordination between stakeholders can start to erode the public’s trust in credible sources, said Wittenberg. When this happens, future messages lose credibility and people start to rely on their own decision making. While this is fine in many cases, the public may not have all the tools and information to assess risk as well as they need to, which can cause problems in risk control and management. However, said Wittenberg, the important thing to remember is that, on both sides of the equation, people are really trying to do their best.


Dr. Eve Whittenberg teaches in Environmental Health Risk: Analysis and Applications at Harvard T.H. Chan School of Public Health.