An explosion that killed one and injured another contract welder at a DuPont plant November 9, 2010, was the result of flammable vinyl fluoride inside a large process tank, a hazard which had been overlooked by DuPont engineers, said officials at the Chemical Safety Board (CSB).
The CSB found sparks or heat from the welding, which took place on top of the tank, most likely ignited the vapor. The CSB said a primary cause of the blast was the failure of the company to require the interior of storage tanks – where hot work is occurring – to undergo monitoring for flammable vapor. A proposed recommendation urges DuPont to require monitoring the inside of storage before performing any hot work, defined as welding, cutting, grinding, or other spark-producing activities.
“I find it tragic that we continue to see lives lost from hot work accidents, which occur all too frequently despite long-known procedures that can prevent them,” said CSB Chairperson Rafael Moure-Eraso. “Facility managers have an obligation to assure the absence of a flammable atmosphere in areas where hot work is to take place. Explosion hazards can be eliminated by testing inside tanks as well as in the areas around them.”
The accident occurred at the DuPont chemical plant in Tonawanda, a suburb of Buffalo, which employs 600 workers. The facility produces polymers and surface materials for countertops, sold under the trade names Tedlar and Corian. The process for making Tedlar involves transferring polyvinyl fluoride (PVF) slurry from a reactor through a flash tank and then into storage tanks. An overflow line also interconnected the tanks. The CSB found the company erroneously determined any vinyl fluoride vapor that might enter the tanks would remain below flammable limits.
Days before the incident DuPont shut down the process for tank maintenance due to corrosion on tank agitator supports. They locked out the fill lines for safety. Tanks 2 and 3 ended up repaired and the process restarted, but they delayed work on tank 1 because the necessary parts were not available. Finally, a contract welder and foreman went in to repair the agitator support atop tank 1. Although tank 1 remained locked out from the main process, the overflow line remained open which connected tank 1 to tanks 2 and 3. The CSB determined flammable vinyl fluoride flowed through the overflow line into tank 1 and accumulated to explosive concentrations. Investigators found while a facility hot work permit went out for the task, DuPont personnel who signed it were not sufficiently knowledgeable about the Tedlar chemical process.
Although DuPont personnel monitored the atmosphere above the tank prior to authorizing hot work, no monitoring occurred inside the tank to see if any flammable vapor existed there. The CSB investigation found the hot work ignited the vapor as a result of the increased temperature of the metal tank, sparks falling into the tank, or vapor wafting from the tank into the hot work area.
The explosion blew most of the top off the tank, leaving it and the agitator assembly hanging over the edge. The welder died instantly from blunt force trauma, and the foreman received first-degree burns and minor injuries.
“Our investigation found that DuPont’s process hazard analysis incorrectly assumed that vinyl fluoride in the Tedlar process could not reach flammable levels in the slurry tanks,” said CSB Team Lead Johnnie Banks said. “And, critically, DuPont personnel did not properly isolate and lock out tank 1 from tanks 2 and 3 prior to authorizing the hot work. The flammable vapor was able to pass through the overflow line into the tank the welder was working on, unknown to him or to the operators who signed off the hot work permit.”
The CSB also determined DuPont should have included the three storage tanks as part of the Tedlar process covered by OSHA Process Safety Management rules. Yet on the day of the accident, a compressor failure led to higher concentrations of vinyl fluoride vapor in the polyvinyl fluoride slurry. Furthermore, a U-shaped seal loop on the flash tank overflow line had a “fishmouth” split in the pipe that could emit vinyl fluoride vapor. Engineers concluded further operation with the broken seal loop presented no hazards, but the CSB determined the pipe split provided a potential pathway for flammable VF gas to enter the tanks.