Experts question decision to vent and burn toxic vinyl chloride from derailed tankers in East Palestine, Ohio

Experts called by the US National Transportation Safety Board, or NTSB, to testify about the February train derailment in East Palestine, Ohio, said at a hearing Thursday that tank cars containing vinyl chloride were not undergoing a chemical reaction that would have caused them to explode – calling into question the decision of first responders to vent and burn the hazardous chemical inside.

In the aftermath of the derailment, residents of East Palestine have demanded an explanation for the decision to vent and burn toxic vinyl chloride from five derailed tanker cars, which sent a black cloud of smoke drifting ominously over the town and southeast into nearby Pennsylvania on Monday February 6, three days after the train initially derailed.

At Thursday’s investigative field hearing convened by the NTSB, representatives from train operator Norfolk Southern and emergency response contractors from two companies it hired to help after the crash testified that they were worried the railcars were in danger of exploding based on a number of factors:

• Temperature readings taken with thermal imaging cameras aimed at the outside of the tanker cars indicated that they were high and fluctuating.
• Pressure relief valves on several cars, which had been periodically discharging liquified gas, seemed to stop working. Contractors worried that the valves had been “gummed up” by the formation of solid polyvinyl chloride, an indication that a dangerous explosive reaction was underway.
• Chemical safety data sheets for vinyl chloride indicated that the chemical could begin a dangerous explosive reaction if exposed to significant heat alone.

Keith Drabick, chief of the East Palestine Fire Department, who was leading the response at the derailment scene, testified that Norfolk Southern and its consultants had weighed the uncertainties on the ground and concluded that a “vent and burn” was the only viable solution. That involved placing small charges on the railcars and blowing holes in the tanks, letting the liquified vinyl chloride drain into pits dug into the ground, where it was burned.

He said that ultimately, the final call was his.

“The final yes was given by me, based on the consensus by everybody in the unified command,” he testified Thursday.

Drabick said that in a meeting with Gov. Mike DeWine and members of Norfolk Southern on Sunday night – two days after the February 3 derailment – he was told to make the call quickly.

“When we got taken down to that room, we were told we had 13 minutes to make that decision” because weather and other conditions that might come into play as daylight faded, Drabick said, at the hearing.

In written testimony released on Thursday, Drabick told investigators “I kind of got blindsided, which probably is the biggest thing that bothers me out of all of this.”

Norfolk Southern declined to comment to CNN for this story and pointed to its experts’ testimony in the hearing.

Evidence tells a different story

Evidence presented at the hearing on Thursday, including photographs, laboratory tests, and temperature readings from the cars, suggests that the railroad and its experts misinterpreted the signs that a reaction was occurring.

NTSB investigator Paul Stancil showed a graph of the temperature readings taken from one of the vinyl chloride tankers investigators were most worried about. The temperatures taken on Sunday night, February 5, appeared to be going down, rather than up.

“Would you agree, that with the exception of the one spike in temperature during the middle of the night, the temperature was on a downward or decreasing trend?” Stancil asked Drew McCarty of Specialized Professional Services Inc., one of the contractors Norfolk Southern hired to assist with the response.

“Why was this tank car still a concern when the temperature of the tank car was decreasing?” Stancil asked.

McCarty said he had not put a lot of stock in the temperature readings because he didn’t have confidence that they were accurate. He said the workers were using thermal imaging cameras to read the temperature on the outside of the tankers, and because of the damage, they weren’t sure they were always able to hit the metal shell of the tankers.

“We, the contractor team and Norfolk Southern, didn’t put a lot of merit in those readings on those cars because we thought we might be contacting with the thermometer, the thermal blanket or insulation, and may or may not have even been touching the tank car,” he said.

McCarty said a greater concern had been the intense heat the cars were exposed to as fires blazed around them for hours.

What’s more, he said, the pressure relief valves on the cars, which are designed to vent liquified gas when pressure builds, had been working and then suddenly seemed to stop.

“Something either gummed it up or mechanically messed it up, and either way, it was incredibly dangerous,” McCarty said.

He said the consensus of the railroad and its experts at the time was that a polymerization reaction was happening inside the cars. When liquified vinyl chloride polymerizes or quickly turns to solid polyvinyl chloride, the familiar white PVC that many plumbing pipes are made of, it generates immense heat. Experts feared that reaction could cause the cars to explode and send metal shrapnel flying for up to a mile.

They feared that polyvinyl chloride forming inside the cars had gummed the pressure relief valves shut.

Evidence presented on Thursday showed that the pressure relief valves became blocked from the outside, not the inside, as first responders had assumed.

On Thursday, however, NTSB investigators said that some of the valves had sustained damage and become clogged but that it was because their aluminum housing melted in the fire.

What’s more, lab tests on swabs of the cars taken after the vent and burn conducted by OxyVinyls, a subsidiary of Occidental Chemical that made the vinyl chloride, found no evidence that polyvinyl chloride had formed.

Paul Thomas, vice president of health, environment, safety and security at OxyVinyls, said at the hearing that the company convened a team of safety experts when it was told about the derailment, and the team was interacting with Norfolk Southern and its contractors by phone until Sunday, two days after the crash.

In two conference calls, Thomas said, they had tried to convey that based on the evidence, they didn’t think an explosive polymerization reaction was occurring.

“The reaction would have an obvious temperature rise that would continue throughout the duration of the reaction. For this reason, we emphasized to Norfolk Southern and its contractors the importance of monitoring the temperature of the rail cars,” Thomas said in his opening statement.

Based on those temperature readings, he testified, “we were trying to communicate we didn’t believe polymerization was going on.”

But Thomas said OxyVinyls was not present in the meetings where the decision was made to vent and burn the cars. Instead, its experts were told to communicate indirectly through a contractor who would then relay the information to Norfolk Southern.

Robert Wood, who was the incident commander for Norfolk Southern, testified that the company also relied on information in OxyVinyl’s material safety data sheet for vinyl chloride to guide the decision to vent and burn the cars. That sheet cautions anyone working with vinyl chloride to avoid extreme heat, which could cause explosive polymerization.

Independent expert William Carroll of the University of Indiana testified on Thursday that he was puzzled by term “explosive polymerization.” “That made no sense to me,” he said, based on his years of experience with vinyl chloride.

He said he traced the term back to two 1970 scientific papers that are no longer in print. References to those articles say that vinyl chloride will polymerize in the presence of oxygen, he said.

However, OxyVinyls testified that before shipping the vinyl chloride, it had stabilized the chemical by removing oxygen from the tanks so almost none was present.

Without oxygen or another initiator, Carroll said, it wouldn’t have been possible for the chemical to polymerize.

He noted “vinyl chloride doesn’t polymerize only on the action of heat. It does not spontaneously polymerize.”

Drabick also said that even if he had been aware that OxyVinyls did not think the vinyl chloride was undergoing an explosive reaction, it might not have changed the outcome of the vent and burn. There were simply no options left for removing vinyl chloride from the damaged tankers, he said.

McCarty says first responders considered attempting a “hot tap,” in which they cut holes in the tankers with a welding gun and pump the vinyl chloride into another container.

However, the team decided that was too risky because they didn’t know the level of liquid vinyl chloride still left in the cars. Without liquid in the tanks at a certain level, a hot torch cutting into the cars could have caused an explosion.

“We made an absolute safety decision for the good of this community and for our own people,” McCarty said.

The NTSB hearing continues on Friday.

For more CNN news and newsletters create an account at CNN.com