Ten months before it failed, the inspection report said the tank was not fit for continued service.
What happened
On 26 May 2026, at about 7:09 am, a 1.2-million-gallon atmospheric storage tank failed at the Nippon Dynawave Packaging pulp and paper mill in Longview, Washington. It held white liquor, a hot caustic solution of sodium hydroxide and sodium sulphide at about 200 °F (about 93 °C). About 900,000 gallons were released into the courtyard where the tank stood and into the surrounding buildings. Eleven employees were killed and others were seriously injured.
The US Chemical Safety Board (CSB) published an investigation update in August 2026. It is an interim document: the CSB is still determining the failure mode and the cause.
What the CSB has recorded so far
The tank was built in 1987 from carbon steel to API Standard 650, with five shell courses. Its shell was repaired in 1999, 2002, 2006, 2011 and 2017. It had no secondary containment.
In July 2025 an external visual inspection and external ultrasonic thickness testing found significant thinning below the minimum safe thickness in the lower regions of the tank. The inspection contractor reported the tank "not fit for continued service" unless repaired, noted a high likelihood and consequence of failure, and recommended internal inspection and repair of shell courses 1, 2 and 3.
Further external thickness inspections in October 2025 and February 2026 found large portions of the shell still below the calculated minimum safe thickness.
The CSB records that the tank was not internally inspected, the shell was not repaired, and the tank was neither taken out of service nor derated before the failure. On the morning of the failure a process upset stopped liquor leaving the tank while liquor kept flowing in; the tank was about 90% full roughly 45 minutes before it failed.
What it means for asset owners
The CSB's final report will set out the failure mode and the causes, and nothing here anticipates it. The update does already show a pattern that is not unique to one site: the thickness was measured, the result was below minimum, and the warning was written down, but the decision did not follow.
A wall-thickness reading below the calculated minimum is not a maintenance item to be scheduled when convenient. It is a fitness-for-service question with only a few answers: repair, derate (for example by limiting fill height), or remove from service, each with a named owner and a date.
What an engineer should check
- Which of your tanks and vessels have thickness readings below the calculated minimum today?
- For each one, is there a recorded decision (repair, derate or remove), who made it, and when does it expire?
- Is the thickness survey only external? What would an internal inspection show that an external survey cannot?
Source. US Chemical Safety Board, Investigation Update: G Tank failure at Nippon Dynawave Packaging, Longview, Washington, August 2026. Read the update. This article summarises the source in our own words; see the original for full details. Figures are as stated by the CSB. It is general technical information, not advice on a specific asset.
Related reading
- How a corrosion failure investigation works
- Materials selection for corrosive environments
- MTIS corrosion and integrity services
- Follow MTIS on LinkedIn for new failure reviews and industry briefings.
A tank or vessel below minimum thickness?
Send us the inspection data. We will tell you what it supports and what it does not. Request a project quote.