Case Study: BP Texas City Explosion 23 March 2005
Panel findings, the average rate of overtime for operators and maintenance personnel at the Texas City refinery in the past four years was 27 percent, with several employees exceeding 68 percent (Baker et al., 2007). The Baker Panel Report concluded that such overtime rates were excessive, would likely compromise safety, and were symptomatic of understaffing (Baker et al., 2007). By degrading judgment and causing cognitive fixation, fatigue likely contributed to the overfilling of the raffinate splitter tower.
Inadequate operator training for abnormal and startup conditions.
Inadequate training for operations personnel, particularly for the board operator position, contributed to causing the incident. The hazards of unit startup, including tower overfill scenarios, were not adequately covered in operator training. The ISOM unit operator training program did not include
training for abnormal situation management, the importance of material balance calculations, and how to avoid high liquid level in towers;
effective verification methods of operator knowledge and qualifications; and
a formal program for operations crews to discuss potentially hazardous conditions, such as startup or shutdown, to enhance operator knowledge and define roles.
Failure to establish effective safe operating limits.
the ISOM operating limits did not include limits for high level in the raffinate splitter tower.
Placement of Temporary Structures
BP’s placement of occupied trailers close to the ISOM unit was a key factor leading to the fatalities from this disaster. All 15 who died during the March 23, 2005, incident were in or around trailers. To address potential safety and health hazards of the trailers that had been placed near the ISOM unit since 2002, the refinery required them to be evaluated under its Management Of Change (MOC) procedure, which also required that a Process Hazard Analysis (PHA) be conducted. None of MOC team members had been trained to use the Amoco workbook and thus did not understand how to do the building siting analysis. In January and February 2005, nine other trailers were sited between the ISOM and NDU without conducting a MOC. For siting purposes, the siting workbook specifically instructs users to consider clusters of buildings, such as turnaround trailers, as one building, which the team did not do. By conducting only one MOC, the occupancy load of the other trailers was never considered to increase risk.
In addition, the MOC procedures clearly state that the proposed change – in this case the siting of the first trailer – cannot be initiated until all action items identified in the PHA have been resolved. Although two action items were still pending from the MOC at the time of the March 2005 explosion, this trailer had been occupied by contractor personnel since