The CSB's Clairton final report: an undocumented step, a brittle valve, and buildings on top of the hazard

On 10 August 2026 the U.S. Chemical Safety Board released its final investigation report on the 11 August 2025 explosion at the U.S. Steel Clairton Coke Works in Clairton, Pennsylvania. Two workers were killed, 11 were injured — five seriously — and the CSB estimates $52.5 million in property damage. The full report is worth reading closely, because the mechanism is assembled from ordinary, familiar pieces rather than an exotic failure.

What happened

Clairton, built in 1901, is the largest coke manufacturing facility in the Western Hemisphere, operating six coke batteries and 455 ovens. Coke oven gas — a flammable, toxic mixture of hydrogen, methane, nitrogen and carbon monoxide left after tar, ammonia, light oil and sulfur are removed — is driven off during production and piped as fuel to heat the ovens.

On the day of the incident, a supervisor decided to “exercise” the Battery 13 isolation valve — a cast-iron double-disc gate valve — to confirm it operated correctly. Exercising meant closing the valve and reopening it through its full range. Workers had long had trouble fully closing these valves because coke-oven residue accumulated in the valve seat, and over at least the prior three years they had adopted a practice of injecting steam or pressurised water into the cleanout port to wash the residue out. On 11 August, contractors from MPW Industrial Services used a pump truck to apply pressurised water while the double gates were closed or nearly closed. That created a sealed cavity between the two gates; the water filled it, pressure climbed past what the valve body could contain, and the valve — cast iron, originally manufactured in 1953 and refurbished in 2013 — failed catastrophically. Toxic flammable coke oven gas escaped and ignited within minutes.

The explosion tore through the “transfer area” between Batteries 13 and 14, where several routinely occupied buildings — control rooms and a break room — stood less than 20 feet directly above the coke-oven-gas piping. None were designed or built to withstand an explosion; all were catastrophically damaged. Both fatally injured workers were in or near control rooms directly above the piping.

The three findings, read as a study checklist

The CSB identified three key safety issues, and each maps onto a decision a study team actually makes.

Procedures and hazard analysis. The water-washing step was real, routine, and entirely undocumented — U.S. Steel’s exercising procedure never mentioned water, and there was no formal procedure for washing valves. A step that lives only in informal practice never reaches hazard review, so the hydraulic-lock scenario that killed two people was never on anyone’s worksheet. The CSB recommended U.S. Steel develop a written procedure for washing valve seats with pressurised water to minimise the hazards, and build a comprehensive process safety management system for all coke-oven-gas processes at the site.

Facility siting. Occupied buildings sat directly above a major-accident source. The CSB’s Investigator in Charge put it plainly: had the buildings been located away from the coke batteries, the incident “could have been far less severe.” The board recommended a siting evaluation for all occupied and potentially occupied buildings at Clairton, with identified hazards mitigated — the same occupied-building-siting question API 752/753-style reviews are built around.

Process safety management systems. The board recommended that Nippon Steel North America develop a corporate process-safety governance program, implement PSM across its facilities, and audit regularly and correct deficiencies promptly; and that MPW develop written policies for cleaning piping systems containing flammable or toxic gas and train workers on them.

Why it matters for study teams

Three threads stand out for anyone facilitating or reviewing studies. First, non-routine and maintenance operations are where informal practice accumulates — the value of a study depends on whether those steps are captured honestly rather than idealised. Second, materials of construction matter: cast iron is widely warned against in flammable service, and a brittle body under an unanticipated hydraulic load has no ductility to fall back on. Third, facility siting is a hazard control, not a real-estate afterthought; occupied buildings over a release source convert a loss of containment into a fatality count.

As CSB Board Member Sylvia Johnson noted, this was “the result of workers routinely performing a task incorrectly over a period of years until it ultimately led to a catastrophic explosion.” The corrective is unglamorous: write the procedure for every task that can cause harm, review it for hazards, and train to it.