A pH loop, a pooling pipe, and a scrubber switched off: the CSB's Woodland Pulp update
On 14 July 2026 the U.S. Chemical Safety Board released a factual investigation update on the hydrogen-sulfide release that killed two employees and exposed ten others at the Woodland Pulp mill in Baileyville, Maine, on 27 January 2026. The investigation is still open — the CSB has not yet issued conclusions or recommendations — but the update lays out the sequence of events in enough detail to be worth reading closely, because the mechanism is built entirely out of deviations a process hazard analysis is designed to surface.
What happened
The account in the CSB’s update and the full update document runs as follows. Woodland Pulp chemically processes wood chips into pulp in an area called the Bleach Plant, housed within a building called the Kraft Mill. Liquid waste from the Bleach Plant drains into an acid sewer pipe more than 1,000 feet long, which flows by gravity to the site’s wastewater treatment plant. One section of that pipe, an “upward transition,” runs uphill, so liquid tends to accumulate there.
The discharges are often alkaline and can carry sulfur compounds. To control pH, the mill sometimes adds sulfuric acid to the acid sewer — but sulfuric acid combined with high-pH, sulfur-bearing fluids reacts to produce hydrogen sulfide. Under normal conditions any small amount of H₂S generated in the pipe is pulled to a scrubber in the Bleach Plant and removed before it can reach the atmosphere.
On 26 January, managers decided to shut most of the mill down because a sharp rise in natural-gas prices had pushed operating costs up. Overnight into 27 January, operators drained Bleach Plant equipment, sending high-pH sulfur-bearing fluid into the acid sewer. A pH probe near the wastewater plant detected the high pH and automatically called for more sulfuric acid beginning around 4:00 a.m. But with reduced flow during the shutdown, liquid was pooling at the upward transition and taking far longer to reach the downstream probe. The probe kept reading high pH and kept calling for acid — which kept reacting with the sulfur compounds and generating H₂S in the pipe.
At about 11:40 a.m. the Bleach Plant scrubber fan was switched off as part of the shutdown. Without the fan pulling gas to the scrubber, the accumulated H₂S migrated through connected piping into two process vessels elsewhere in the building and escaped through openings into the Kraft Mill. Two employees working on an unrelated equipment-drawing project on the second floor collapsed. The CSB notes there was no system to account for who was in the building or where, no building ventilation, and no fixed H₂S detectors or alarms — nor were personal monitors provided or required. The two employees were not found until hours after the release had dissipated.
Why it reads like a HAZOP node
Strip the site-specific detail away and the sequence is a stack of guideword deviations that each defeated an assumption the control system was built on.
The pH-control loop assumed representative flow past the probe. During the transient, less flow and more level at the upward transition broke that assumption: the sensor was measuring stagnant liquid far upstream of where it thought it was, so a control action intended to neutralize pH instead drove a runaway generation of a toxic gas. That is exactly the kind of “the instrument is telling the truth about the wrong thing” failure a study team is meant to probe when it asks what happens to a measurement-and-control loop under startup, shutdown, and low-throughput conditions.
The scrubber was a credited safeguard — but credited for normal operation. Taking its fan offline as a routine shutdown step is a loss-of-containment-mitigation event occurring precisely in the operating state where the hazard was being generated. Transient operations, shutdowns, and startups are where safeguards are most often bypassed, de-energized, or sequenced off, and they are the nodes most easily skimmed in a study focused on the steady-state process.
The remaining layers the CSB flags — detection, ventilation, and personnel accounting — are the mitigation and emergency-response barriers that a layer-of-protection view exists to force onto the table for a known H₂S hazard. The CSB’s update records that the company was aware H₂S could form in the acid sewer piping; the gap it points to is between recognizing a hazard and having systems in place to monitor and mitigate it.
The trigger worth carrying
The sequence began with a business decision — shut the mill because gas got expensive — not an equipment failure. That is a familiar shape: the initiating condition for a process-safety event is frequently a management or economic decision that quietly moves the plant into a state the original design and analysis did not fully cover. For anyone facilitating or revalidating a study, the practical takeaway from this update is narrow and concrete: node the transient states as carefully as the steady state, and ask specifically which controls and safeguards change behavior — or get switched off — when the plant is being brought down or back up.
The CSB has said its investigation continues, with work still to do on detection and alarms, access control, and the site’s broader process-safety practices, and that complete findings and any recommendations will come in its final report. This piece reflects only the factual update issued on 14 July 2026.