Reader Followup to Bayer Pesticide Plant Disaster, 2008, West Virginia

Dear Sir/Madam,

I just finished reading your on-line article, "Bayer Pesticide Plant Disaster, 2008, Institute, West Virginia" ( http://www.semp.us/publications/biot_reader.php?BiotID=619 ).  I was a former engineer at that site and wish to share additional information, not generally known to the public, of an additional incident that has too many similarities.  It distresses me that the "lessons learned"  did not take hold, albeit there were several changes in site ownership.

The incident I'll tell about is not listed in your section "VI. Institute Pesticide Plant’s Rocky Safety History".  It was circa 1989 when an explosion occurred on the recirculation line of an ethylene oxide (EtOH) polymerization reactor due to safety interlocks that were manually over-ridden, not long after a new computer control system was installed.  This was at the Allethrin (a misnomer) Unit at Institute, where EtOH was polymerized to make polyethylene glycols and polyols, and similar products.  We had recently completed computerizing this facility with the addition of a Taylor MOD300 distributed control system and Triconex safety system, which was my project.  One particular product required the addition of an organic peroxide, which was done at the inlet to the recirculation pump.  Maintenance was recently done, and a manual valve at the bottom of the reactor was accidentally left closed.  The pump was running, but the flow switch indicated no flow, so the safety interlocks would not permit the feed of the peroxide into the loop, which would normally provide through and vigorous mixing with the reactor's contents.  The engineer on call (not I) did not believe the new safety interlocks were working properly, so he used pliers to manually force open the peroxide feed valve, and several hundred pounds were fed to the pump inlet. As the pump was running but there was no circulation, the peroxide was heated by the operation of the pump until it explosively decomposed and detonated.  No safety system could have sufficiently vented the explosion, so the pump and the 10" stainless steel piping ruptured and sent shrapnel in all directions.  Huge I-beams were drastically deformed, and a 10" manual gate valve on the top of the pump went flying about 200 feet, hitting the side of the new control building that was recently completed for this unit (where my office was).  Fortunately, this was a Saturday morning, and hardly anyone was present -- there were no fatalities or injuries, though nerves were quite shattered.  Damage at that time was about $10,000,000.  I led the investigation into the root causes of this incident and presented the report to management (Union Carbide, at that time).

Key similarities to the 2008 incident:

1. New control system
2. Safety interlocks over-ridden, manually
3. Toxic, explosive chemicals involved
4. Run-away reaction without capability to vent
5. Explosion and damage

========================
Kent Moraga