A disaster waiting to happen: The Chemical Safety Board Investigation Report on 2008 Bayer CropScience fatal pesticide chemical explosion

by Elizabeth Grossman

On August 28, 2008 at 10:53 p.m., a massive explosion and fire, caused by a runaway chemical reaction, ripped through the Bayer CropScience pesticide plant in Institute, West Virginia. It killed two workers and injured eight employees, two contractors, and six fire-fighters, all of whom were treated for possible toxic chemical exposure. On January 20, 2011, the Chemical Safety and Hazard Investigation Board (CSB) released its investigation report. Each of its key findings points to a disaster waiting to happen.

The incident occurred at the facility's Methomyl-Larvin unit, where the pesticide Larvin is produced in a process that includes chemical reactions involving a number of highly hazardous, flammable, and explosive chemicals. Among them: methyl isocyanate (MIC) (the chemical released in the 1984 Union Carbide Bhopal disaster), phosgene (so deadly it was used as a nerve gas chemical weapon during World War I and was responsible for the death in 2010 of a worker at a DuPont plant in West Virginia), trichloroethane, ammonia, and chlorine. It happened when the unit that uses MIC in a series of complex reactions to produce methomyl (the base chemical of the pesticide called Larvin) was being restarted after being out of operation so that a new, computerized control system and other equipment could be installed. In its investigation the CSB discovered that realignment of personnel, employee cutbacks, extended work hours, lack of training, and malfunctioning equipment - and company failure to correct chronic safety problems - all contributed to the disaster.

Appropriate procedures not followed
The CSB found that company "pressure to resume production" resulted in the unit being restarted to soon, before the required safety review was conducted and before proper equipment installation and calibration were complete. The CSB also found that critical valves were improperly adjusted. The combination led to an uncontrolled chemical reaction that created extreme heat and pressure inside a 4,500 gallon vessel that ultimately exploded, releasing a highly flammable solvent that ignited in an "immense fireball" and burned intensely for over four hours.

In the days leading up to the incident, the CSB learned, the unit's only technical advisor had worked as many as 15 to 17 hours a day. The unit's engineer at the time had less than a year of experience and told CSB investigators he knew little about the unit's equipment and chemistry. Plant workers were unfamiliar with the new computerized equipment, which they found time-consuming and confusing for critical tasks, particularly during start-up. The unit's extremely complex standard operating procedure, whose manual was hundreds of pages long, also added to the problems.

The recent organizational changes at the plant "directly contributed to the incident causes," reports the CSB. Management did not directly advise or oversee the restart operation, the CSB wrote, and "was so far removed from the process operation that they were unaware" that operators routinely by-passed critical safety measures that led directly to the explosion.

"The deaths of the workers as a result of this accident were all the more tragic because it could have been prevented had Bayer CropScience provided adequate training, and required a comprehensive pre-startup equipment checkout and strict conformance with appropriate startup procedures," said CSB Chairperson Dr. Rafael Moure-Eraso during the press conference presenting the report.

Keeping the community in the dark
The workers affected by the blast included firefighters who couldn't get accurate information about what they might be exposed to. CSB discovered that the gas monitoring system at the unit was not working. Plant workers and Bayer emergency personnel assumed it was, and therefore said no MIC had been released. Bayer's fenceline air monitors were also found inadequate. Some monitors had in fact been turned off to prevent false-positive alarms. In addition, there was initially no direct communication between Bayer emergency personnel and community first responders, which left local response agencies without immediate reliable information during the event.

"The Bayer fire brigade was at the scene in minutes, but Bayer management withheld information from the county emergency response agencies that were desperate for information about what happened, what chemicals were possibly involved," said John Bresland, CSB chairman at the time of the incident. "The Bayer incident commander, inside the plant, recommended a shelter in place; but this was never communicated to 911 operators. After an hour of being refused critical information, local authorities ordered a shelter-in-place, as a precaution."

It's fortunate that no massive release of methyl isocyanate occurred during the disaster. But Bayer initially withheld key chemical information from being released publicly, claiming it to be Sensitive Security Information (SSI) - a claim subsequently overruled. "[W]e concede that our pursuit of SSI coverage was motivated, in part, by a desire to prevent that public debate [concerning the use of MIC] from occurring in the first place," said William Buckner, the president of Bayer CropScience in his April 2009 testimony to Congress.

The community suffered some property damages from the explosion but no apparent MIC effects. A lack of information about potential chemical exposures caused concern among some local firefighters who complained of symptoms consistent with toxic chemical exposure on the day after the fire. And the community as a whole remains in the dark about what they may have been exposed to since no data on chemical releases was - or apparently ever will be - made available. "No reliable data or analytical methods were available to determine what chemicals were released, or predict any exposure concentrations," wrote the CSB.

A history of problems
The August 2008 explosion was not the first disaster at the Institute plant. In 1993, when the facility was owned and operated by Rhone-Poulenc, an explosion in the same methomyl unit killed one worker and injured two. Similar improper control mechanisms led to both explosions, and the problems were never fully corrected. In some cases Bayer said corrections had been made, but the fixes had not been fully implemented.

Over the years, OSHA inspections of the facility found numerous repeat violations. "At Bayer, longstanding operating procedure deficiencies played a significant role in the accident," says the CSB report. Just three weeks before the 2008 explosion, an internal Bayer review found four dozen hazards that were supposed to be corrected but had not been.

The CSB investigation found a lack of oversight not only from Bayer, but also from OSHA and the EPA. While both agencies had conducted "process safety related audits and inspections" at the Bayer facility before the August 2008 incident, "the inspections did not detect or correct all the serious, longstanding process safety problems that were revealed by investigations conducted after the incident," writes the CSB. Further, while OSHA cited Bayer for "deficient process hazard analyses in 2005," the agency never verified that Bayer had fully corrected these deficiencies - failures that ultimately contributed to the August 2008 explosion.

Planned and recommended changes
In 2010 Bayer announced it would stop storing MIC above ground at the plant and on January 11, 2011, it announced it would end production of the pesticides that require production, storage and use of MIC and phosgene at the Institute facility. That change will happen by the end of 2012 and will end all methyl isocyanate production by Bayer. It will continue to produce Larvin at the Institute plant but do so using methomyl "sourced externally."

"Bayer's decision to end pesticide production using MIC was, I understand, done for its own business reasons. But for whatever reasons, the eventual elimination of this chemical will enhance safety in the Kanawha Valley, for workers and residents alike, and is a positive development in my view," said Dr. Moure-Eraso.

The CSB report recommendations include numerous internal plant operational changes and changes to emergency communications, as well as a recommendation that the West Virginia Department of Health and Human Resources establish a "Hazardous Chemical Release Prevention Program" that would have the authority to inspect and regulate chemical plants, and make public its ongoing findings. Following its January 20 public hearing in Institute, and a public comment period, the CSB board will vote on the report conclusions and recommendations.

While the January 20 press conference was being held, the CSB website list of U.S. chemical accidents continued to lengthen. With luck, lessons learned from the Institute disaster can prompt systemic changes that will stop this daily litany of chemical releases, fires, and explosions. But among the key elements to be determined is how the kind of information Bayer initially claimed as Sensitive Security Information will be handled in the future.

Elizabeth Grossman is the author of Chasing Molecules: Poisonous Products, Human Health, and the Promise of Green Chemistry, High Tech Trash: Digital Devices, Hidden Toxics, and Human Health, and other books. Her work has appeared in a variety of publications including Scientific American, Salon, The Washington Post, The Nation, Mother Jones, Grist, and the Huffington Post. Chasing Molecules was chosen by Booklist as one of the Top 10 Science & Technology Books of 2009 and won a 2010 Gold Nautilus Award for investigative journalism.