Issue Date: October 10, 2011
Investigating A Fatal Phosgene Leak
DuPont manufacturing practices were blasted by the Chemical Safety & Hazard Investigation Board (CSB) in a new final report that examined a fatal accident at the corporation’s Belle, W.Va., chemical plant.
The report looked at three accidents that occurred within 33 hours, one of which involved phosgene and killed a worker. The investigation found company practices that “surprised and alarmed” CSB, says board Chairman Rafael Moure-Eraso.
“That these preventable accidents happened at a company with DuPont’s reputation for safety should indicate the need for every chemical plant to redouble its efforts to analyze potential hazards and take steps to prevent tragedy,” Moure-Eraso says.
The series of accidents occurred in late January 2010 at the DuPont chemical manufacturing plant. Methyl chloride and oleum were also released, but most attention has focused on phosgene because it was the most hazardous chemical released and led to the worker’s death.
The DuPont plant rests on 700 acres along the Kanawha River, 8 miles east of Charleston, W.Va. The plant produces a variety of chemical products.
On Jan. 22, 2010, a plant alarm sounded and workers discovered that methyl chloride, a toxic and extremely flammable gas, had been leaking to the atmosphere unnoticed for five days and that some 2,000 lb had been released. Earlier alarms had gone off but were shrugged off because of the plant’s history of false alarms, says CSB lead team investigator Johnnie Banks.
The next morning, plant workers discovered that oleum, a concentrated form of sulfuric acid, had leaked through a corroded pipe, mixed with steam, and produced a fuming cloud of sulfur trioxide. Later that day, the phosgene release occurred, and three workers were exposed, one of whom died a day later.
Phosgene, which has a history as a chemical weapon in addition to being an industrial intermediate, attacks proteins in the lungs’ alveoli, disrupting blood-oxygen exchange and resulting in suffocation.
Liquid when under pressure or below 8 °C, phosgene was stored at the plant in 1-ton tanks in a shed area open to the atmosphere. The chemical was used in the manufacture of five products. A transfer hose moved the chemical in liquid form to various plant production units, according to CSB’s report. However, the unit where the phosgene leaked was in intermittent service at the time of the accident. The temperature of the phosgene in the hose increased, building up pressure that led to the leak.
The board says the phosgene hose, which was made from Teflon-lined stainless steel, was susceptible to corrosion by phosgene.
As far back as 1987, DuPont realized the hazard in using braided stainless steel hoses lined with Teflon, says Banks, pointing to internal company documents obtained by CSB. At that time, a DuPont expert recommended changing to hoses lined with Monel, a nickel alloy used in corrosive applications, Banks adds. The expert noted that Monel is more expensive than the stainless steel Teflon hose but will last at least three months, according to DuPont documents in the CSB report. The company decided not to make the change to Monel, however.
Currently, DuPont practices require the hoses to be replaced monthly. But the hose that burst in front of the workers had been in service for seven months, according to CSB’s report. The company had established a computerized warning system that was supposed to indicate when the hoses should be changed, but the automated system had been modified and no longer issued a warning, according to CSB.
Another safer measure not taken by DuPont cited in the report involved the shed where phosgene was stored at the plant. In the 1980s, according to CSB, DuPont considered enclosing the storage area and venting it through scrubbers to block phosgene from entering the atmosphere.
DuPont’s analysis concluded that the change would be safer for the public and workers but more expensive. The company did not make the change.
The report quotes from an internal document written by a DuPont employee in 1988: “It may be that in the present circumstances the business can afford $2 million for an enclosure; however, in the long run can we afford to take such action which has such a small impact on safety and yet sets a precedent for all highly toxic material activities?”
In a process safety analysis conducted in 2000, DuPont again recognized the need to enclose phosgene, the report says. But management put the work off by granting four extensions between 2004 and 2009. The board concluded that an enclosure, scrubber system, and routine requirement for protective breathing equipment before personnel enter the enclosure would have prevented exposure. None of these measures were in place at the time of the 2010 accident.
“The company just kicked the can down the road for 20 years,” Banks says.
However, since the accident, DuPont has dropped use of phosgene at the plant, notes Nathan Pepper, a company spokesman. The oleum-related process that led to an accident in the sulfuric acid recovery unit was also shut down. Both closures were due to changes in business plans for the company. In the case of phosgene, the phaseout was driven by a decision to drop manufacturing of crop chemicals, Pepper says.
Phosgene is used at a “handful” of other DuPont sites globally, according to Pepper. He tells C&EN that he is unsure how it is handled at the other facilities. But a DuPont plant in Mobile, Ala., was held up as a model by CSB in its report.
The Mobile facility uses the same 1-ton phosgene cylinders and Teflon-lined, stainless steel braided hoses similar to those used at the Belle plant, but Mobile’s hoses are shorter and can handle a greater maximum operating temperature and pressure, according to CSB.
And Mobile’s cylinders are housed in an enclosed room that vents to an emergency scrubber that creates a slight negative pressure and scrubs a discharge before venting it to the atmosphere. It has the capacity to capture vapors from an entire cylinder. Also, the facility has audible alarms inside the enclosure and a flashing light outside to warn employees of dangers. The emergency scrubber is automatically triggered by leak detection sensors.
In its report on the Belle accidents, along with the phosgene recommendations, the board urges DuPont to create a system to encourage workers to anonymously report accident near misses and to improve a host of deficits relating to equipment maintenance and repair, alarms, internal communications, and other areas.
Pepper says the company is examining the report and had already made many of the changes after its internal accident review and report. Among them, Pepper says, DuPont has conducted a comprehensive and intensive operations safety review, including more audits, more opportunities for employees to participate in a hazard review program, and a new alarm management system.
But Daniel M. Horowitz, CSB managing director, notes: “If this happened at DuPont, which is perceived as the industry’s safety leader, it could happen anywhere. It shows the kind of trade-offs companies sometimes make.”
Horowitz particularly focused on the need to modernize the national standard for compressed gas. “These industry standards are 40 years out-of-date,” he says.
The board recommends that the Compressed Gas Association and the Occupational Safety & Health Administration (OSHA) adopt the National Fire Protection Association code for phosgene and other highly toxic gases and urges that the American Chemistry Council, a chemical industry trade association, revise its phosgene safe practices guidelines. The changes should advise chemical companies not to use hoses made with permeable materials and to immediately report and investigate near misses involving phosgene.
ACC, in a letter from Phosgene Panel Manager Sahar Osman-Sypher, says the panel will comprehensively reevaluate the manual and consider CSB’s recommendations. No timeline for this was given.
The report also notes that OSHA had not conducted a planned inspection at the Belle facility for 18 years, although it did conduct compliance-specific investigations in 1995 and 2004.
The report took the board more than a year-and-a-half to complete. The board’s goal has been to complete reports within one year, but that goal has proven to be a challenge for CSB with its small staff and a full docket.
With the DuPont final report, CSB still has 17 unfinished investigations, a large number for an organization with 38 staff. Before the release of the DuPont report, the board had not issued a final report in more than eight months.
When the board decided to conduct the DuPont investigation, board members split over whether CSB had the resources to take on the investigation. It had 17 investigations ongoing at the time, including two in the Charleston area—one involving phosgene and another methyl isocyanate at, respectively, DuPont and Bayer (C&EN, Feb. 15, 2010, page 40).
The board’s resources are being further taxed: In mid-September, two of its five board members announced they were leaving because their terms had expired. The board members are presidential appointees, so the seats will remain open until President Barack Obama nominates and the Senate confirms new members.
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society