THE US Chemical Safety Board (CSB) has found that a lack of hazard identification processes contributed to the tank explosions that killed three people at the Midland Resource Recovery (MRR) facility in West Virginia in 2017.
One of the services performed by MRR is the decommissioning of the tanks used to add the odorants known as mercaptans to natural gas. The MRR process added sodium hypochlorite to the odouriser vessels to remove the odour from the steel so that it could be scrapped. The vessels were sealed for a time, creating what MRR referred to as process water. On 24 May 2017, when the process water was being drained from a tank, an explosion occurred which killed two workers, one of whom was the founder and President of MRR. Another worker was seriously injured. A second explosion occurred on 20 June which killed a contractor who had been hired by MRR to perform investigation and mitigation work.
After the initial explosion, CSB began an investigation which found that it was plausible that methanol was also present in the MRR tanks as it was sometimes used to wash tanks. The CSB also found research which suggested that the methanol mixed with sodium hypochlorite can form methyl hypochlorite, a highly explosive chemical. The final report states that the MRR process had created the possibility that each chemically treated odouriser was essentially a bomb.
MRR’s legal counsel had hired contractor Specialized Professional Services (SPSI) to drain the remaining tanks. The CSB was not allowed to communicate its findings to SPSI and an MRR attorney said SPSI had a different theory as to the cause of the original explosion. The CSB investigators on site took shelter behind a shipping container. When SPSI attempted to drain a tank on 20 June, an explosion occurred which killed the SPSI field supervisor.
The CSB report determined that there was no way to know exactly what chemicals were present in the odourisers at the time of the explosions, but concluded that dangerous reactions occurred when the tanks were unsealed. The presence of methanol is one possible explanation. The CSB found that MRR did not have a formal hazard identification process to analyse the chemicals inside the odouriser vessels. MRR allowed the sodium hypochlorite to sit in the MP odourisers for over a month, potentially allowing unstable chemicals to form. It had no established limits on how long the sodium hypochlorite should be left for and also did not accurately control the dilution of sodium hypochlorite. There were also no safeguards to prevent uncontrolled chemical reactions.
CSB Interim Executive Authority Kristen Kulinowski said: “MRR did not have, and federal regulations did not require, a comprehensive safety management system to identify and control hazards from reactive chemicals. As a result, two serious explosions occurred.”
The CSB has issued key lessons for companies that deal with reactive chemistry. These include the need for a robust safety management system to prevent reactive chemical incidents and the need for a thorough and complete understanding of the reactive chemistry.
The CSB has been calling on the Occupational Safety and Health Administration (OSHA) and the Environmental Protection Agency (EPA) since 2002 to cover reactive chemicals and hazards in their regulations, but neither agency has acted on the CSB’s recommendations. “The CSB has long been concerned about the persisting gaps in federal safety regulations for reactive chemical hazards, and tragic incidents like the two explosions at MRR continue to occur,” said Kulinowski. “It is past time for OSHA and the EPA to adopt our recommendations to update their regulations to cover catastrophic reactive hazards that have the potential to seriously affect workers and the public.”
MRR still decommissions vessels but no longer uses sodium hypochlorite.
Catch up on the latest news, views and jobs from The Chemical Engineer. Below are the four latest issues. View a wider selection of the archive from within the Magazine section of this site.