How to Keep a Lid on the Process: Carbide Industries Incident

Maintenance activities at Carbide Industries couldn’t keep pace with equipment deterioration. The failure was ultimately catastrophic.

While no single cause was pinpointed for the March 21, 2011 explosion at Carbide Industries, most fingers point back to the furnace lid. This was a complex structure of steel and ceramic construction, designed to take the brunt of the hugely hot process. The lid was fabricated in six pie-shaped sections that bolted together using flanges radiating out from the center. The side facing the process had a ceramic lining designed to withstand the heat from the electric arcs.

To keep the steel from getting hot enough to sag or even melt, internal passages carrying cooling water ran throughout the structure. Where the section faces were bolted together, the passages interconnected and carried water from one to the next. This was not easy service in any respect, and the lid had been subject to all sorts of damage.

Cooling water leaks were a constant and dangerous problem, and they had a variety of causes:

• Refractory and ceramic materials deteriorate over time and can chip or flake off. This leaves the steel exposed directly to the high furnace temperatures causing rapid corrosion.

• Where bare steel is exposed, it tends to sag and the flexing action causes more ceramic material to fall off.

• Boil up incidents where the liquid carbide overheats can fling molten carbide up to the lid causing damage to the ceramic lining.

• When liquid thrown up by boil up incidents hits bare metal, it can melt.

• All of these working together cause cooling water leaks.


How Many Leaks are Too Many?

Maintenance technicians knew there were leaks because they were fixing them every time the furnace was shut down. There had been 26 work orders for repairs in the four months prior to the March incident. Workers at the facility would add oats and boiler solder to the cooling water to clog small leaks from the inside. (As the solid material flows through the system, it lodges in the leaks and forms a blockage on the inside.) Larger holes had to be patched by welding on pieces of steel. This stopped the flow, but hurt the heat transfer by adding more material between the cooling water and heat source causing local hot spots. There was no practical way for workers to repair the ceramic material, so as it flaked away it was gone.

Given that there had been leaks on probably 35% of the surface of the lid, Carbide Industries management had finally ordered a new lid which was scheduled to be installed in May 2011. At the time of the incident, the steel structure had been completed but was still having the ceramic liner installed.

There was no question that the lid had long passed its safe service life. The order for the new one should have been issued at least a year earlier than it had. But, regardless of the condition of the lid, the company continued production.