Shortcuts and temporary remedies can cause damage and worse. Yet, when engineers meet an immovable object like a turnaround deadline, design mistakes seem inescapable. This stress often is amplified during emergencies. Ad hoc measures often get implemented when management is pressing to get a problem affecting production resolved quickly.
Inadequate solutions fall into two general categories: process and maintenance. Process weaknesses include poor procedures, workarounds (see: "Work on Workarounds,") and bad design. Maintenance miscues stem from corrosion, inadequate design, a poor job plan, ad hoc repairs, and deviation from job scope.
The consequences can be tragic. Here's a list of just a few of the fatalities caused by taking shortcuts or treating interim measures as permanent solutions:
• On October 6, 2010, a maintenance worker died of H2S exposure while repairing a clamped pipe leaking H2S and hydrocarbons at ExxonMobil's refinery in Chalmette, La.
• On January 23, 2010, a stainless-steel-braided polytetrafluoroethylene (PTFE) hose at DuPont's pesticide plant in Belle, W.V., ruptured, killing an operator. The hose was supposed to be replaced monthly; the burst hose had been in service for seven months. (I have seen numerous serious and near-fatal accidents involving the use of stainless-steel-braided PTFE hose where piping was needed as a permanent solution.)
• On August 28, 2008, an explosion at the Bayer CropScience Institute, W.V., pesticide plant killed two workers sent to investigate reactor problems. The U.S. Chemical Safety Board (CSB) described operators using ad hoc procedures to bypass critical interlocks because safety equipment hadn't been installed before startup. Management wanted to get the plant up as soon as possible and pushed operations to find a workaround. The accident very nearly involved a tank of methyl isocyanate, the chemical responsible for thousands of deaths in Bhopal, India (see: "Grasp All the Lessons of Bhopal"). The CSB also noted the original procedures were overly complex; had the accident not occurred during startup, Bayer might well have continued to use them until the safety equipment was installed.
• On February 23, 1999, an inspector was burned in an accident involving temporary valves at the Citgo refinery in Lemont, Ill. These valves were installed during a 1983 HF alkylation fire and were left operating. Citgo's records indicated they had been removed.
• On June 1, 1974, a temporary bypass pipe at Nypro UK's Flixborough caprolactam plant ruptured from mechanical stress, instantly killing all 18 people in the control room and nine others in the plant. The bypass was installed two months earlier around a leaking reactor.
Fortunately, good operator habits often catch problems like these before they lead to disaster. For instance, in 2008, an operator at Citgo's Lemont refinery thought he'd better check out a valve replaced the night before. His suspicions were justified when he found an ANSI 150 valve installed in 300-psi line. Operations isn't the only culprit when it comes to ill-conceived efforts to keep production going!
If I were a corrosion engineer I'd wear a tee shirt emblazoned with "I told you so." After witnessing a June 1985 pipeline fire at the Elk Hills Reserve in California, a production site half owned by the U.S. Dept. of Energy (DOE), I asked the unit corrosion engineers about the fire. They showed me charts they regularly reviewed with unit management, i.e., DOE, Bechtel and Chevron. Operations told them to slap a clamp on whatever leaked. So, the fire was no surprise to them; I suspect pipeline and maintenance engineers around the country face similar situations. In 2007, Occidental Petroleum, the new owners of Elk Hills, suffered another pipeline fire. Risks from corroded pipe and temporary clamping afflict many sites. In 2010, five different refineries in Louisiana submitted reports about accidents involving temporary clamped piping.
Part of the problem stems from the way work is done. In the U.S., I've observed that maintenance and construction often deviate from a job plan. Very little thought seems to go into considering the implications and risks. One tragic result was the death of a welder at DuPont's polymer plant in Tonawanda, N.Y., on November 9, 2010.
This sharply contrasts with the approach I've seen in Australia. The Ausies create a methodical and thoughtful job plan and stick to it.
In my opinion, the U.S. Occupational Safety and Health Administration focuses too much on plant operations while largely ignoring the way work is to be completed, i.e., the job plan. I strongly believe as much thought should go into preventing ad hoc solutions to maintenance problems as goes into improving operational safety.
