Inspection Detail

Inspection: 125753368 - General Chemical Corp.

Inspection Information - Office: Ca-E-Concord (R2-D2)

Investigation Summary

On May 1 2001, a sulfuric acid plant in Richmond, California had a release of sulfur dioxide and sulfur trioxide (SO2/SO3) from the facilities' process piping. The events leading to the release of SO2/SO3 began with a power outage at the facility from an automobile crashing into a power pole off site. Upon the resumption of power to the facility, the employer was unable to provide steam to the turbine, manufactured by Dresser, Rand, Trip and Throttle, that moves the process gases through the facilities' process pipes. The negative pressure normally found in the process piping reverted to atmospheric pressure with the turbine shut down. At the exit flue pipes, located just after boilers and next to a flange containing a blind, there were multiple large holes found under insulation that allowed SO2/SO3 vapors to escape to atmosphere. The holes in the 6-foot-diameter flue pipe were measured in feet. The escaping gases impacted the general community in Richmond, California by shutting down a freeway and requiring residents to shelter in place. The escaping gas also exposed employees to SO2/SO3 vapors while working on restarting the steam turbine. At the time of the incident employees were attempting to restart the steam turbine providing negative pressure to the process pipes. The steam turbine was restarted by maintenance workers by first isolating the problem to a Governor valve on the steam turbine, and then opening the Governor valve which failed in a closed position. The plant manager at the time of the incident directed emergency operations on site. The employees working on the turbine were downwind and off center of the SO2/SO3 exit flue pipe leak, however, at times these employees were exposed to an acidic SO2/SO3 vapor cloud. The workers attempting to restart the turbine were approximately 200 ft from the leak in the flue pipes. The facility failed to adequately inspect their process piping for corrosion as required by American Petroleum Institute document 570. Furthermore, the location of the large holes in the exit flue pipes were adjacent to a flange with a blind in place (or a location were a blind could be placed) where accelerated corrosion would be expected. The exit flue piping was put into place in 1963 and 1969, yet it was not until 1999 that the facility attempted to establish corrosion rates of their process piping. The calculated corrosion rates were not valid due to the inadequate frequency. Visual inspections were also inadequate for the process piping. The failure to detect the large holes in the exit flue pipes played a prominent role in the SO2/SO3 release. The absence of steam to the turbine also played a pivotal role in the incident by removing the negative pressure on the exit flue pipes. Over the course of approximately two hours the employer isolated the absence of steam to the turbine to a stuck Governor valve upstream of the turbine. A third party expert, hired by the employer, stated that the most likely cause for the stuck Governor valve was do to a lack of lubrication of the actuator arm on the Governor valve. The valve expert also stated in other facilities the Governor valve's bushings on the actuator arm have fittings to allow employees to conveniently grease the bushings. This was not the case at this facility. No maintenance records were provided showing the actuator arm on the Governor valve had been lubricated. There were no documented injuries to workers as a result of the acidic release.

Keywords: GAS, SULFUR