Amine absorbers: chemical time bombs?
On July 23, 1984, workers at an oil refinery near Chicago noticed a leak in a pressure vessel used to remove hydrogen sulfide from propane and butane through an "amine absorption' process. A catastrophic explosion split the 61.5-foot-tall vessel 15 minutes later, rocketing the top 46-foot section more than a half-mile away and initiating a fire. The accident left 17 dead and 17 more injured, and caused more than $100 million worth of damage. A recently completed autopsy of the vessel's remains by the National Bureau of Standards (NBS) in Boulder, Colo., has identified the chemical events that apparently initiated the 1984 explosion--and that may wreak similar havoc at thousands of similar chemical-process vessels elsewhere.
"We learned that hydrogen was responsible,' says Harry I. McHenry, the lead NBS investigator--but in ways that were not apparent, either at the start of the NBS investigation or earlier, when a team of private engineering consultants probed the accident's cause. The NBS team found that shallow cracks had developed where repair welds had been made during the vessel's 14 years in service. Hard, brittle steel is susceptible to hydrogen-stress cracking when it's simultaneously subjected to hydrogen--here supplied by the hydrogen sulfide in the vessel--and high internal pressures. Unlike the vessel's original factory-welded seams, McHenry says, affected repair welds had not been tempered (heat treated upon finishing) to make them soft and ductile.
Once the initial cracks formed in the weld-embrittled steel, another process took over--hydrogen-pressure cracking. Hydrogen collected at the crack sites until it built up enough pressure to fracture the material. Repeated over time, McHenry says, this process extended the 1- to 2-millimeter-deep cracks until they penetrated the 25-millimeter-thick vessel walls. Conventional wisdom would suggest that such wall-penetrating cracks would leak indefinitely; instead, McHenry's team found that the embrittlement caused the vessel to split in two.
On the basis of these findings, the NBS team recommends:
using low-hydrogen welding techniques for repair work because, McHenry says, cracks can form--even if the weld metal wasn't embrittled--if too much hydrogen is used in welding.
preheating weld areas (to slow weld cooling) and tempering completed welds.
surveying vessel walls periodically with techniques like "magnetic particle inspection' to find cracks.
In a letter sent to regional Occupational Safety and Health Administration (OSHA) administrators earlier this year, the agency's field operations director, John B. Miles, Jr., noted that preliminary results of a survey by the National Association of Corrosion Engineers "indicate that approximately 60 percent of 24 amine absorbers exhibited cracking.' A Japanese survey he cited found that 72 percent of amine gas treatment facilities showed cracking. Miles recommends that warnings on the "potentially hazardous circumstances relative to amine-absorber units' be shared with local OSHA offices, state health and safety agencies and refinery owners.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||investigation of explosion at petroleum refinery|
|Date:||Oct 25, 1986|
|Previous Article:||Late-night legislation: new laws include revised Superfund.|
|Next Article:||Conflict and cliff-hangers.|