Statement of Board Member John S. Bresland Updating the Status of the CSB's Investigation at the BP Texas City Refinery

April 28, 2005

[The following statement was issued to update the public in Texas City, Texas, April 28, 2005, 11 a.m. For further information contact Daniel Horowitz, CSB Director of Public Affairs, at (202) 261-7613 / (202) 441-6074.]

I am John Bresland, Board Member of the U.S. Chemical Safety and Hazard Investigation Board the CSB. We are the independent federal agency that has been investigating the cause of the tragedy at the BP Texas City refinery for the past five weeks.

With me at the podium is CSB investigator-in-charge Don Holmstrom, who is leading the field investigation.

We currently have a staff of 10 investigators and experts at the site. We expect this field phase of our root cause investigation to continue for a number of additional weeks. In our first five weeks, we have conducted detailed interviews of about 120 witnesses and collected more than 10,000 pages of documentary evidence.

Yesterday I had to opportunity to enter the isomerization unit along with Mr. Holmstrom, and it was a very informative but sobering experience.

The Board's investigation is continuing to focus on all the principal safety issues raised by the case including the design of the unflared blowdown system, the causes of the overpressure in the raffinate splitter, and the proximity of the trailers to an operating process unit. As most of you know, those trailers were 100-150 feet away from the blowdown drum when a geyser-like release of flammable hydrocarbons occurred on March 23.

The trailers suffered extreme and catastrophic damage. The debris pattern at the site is complex and is not indicative of a single vapor cloud explosion. We believe that there were a number of distinct explosions in rapid succession, possibly as many as five. We are reviewing a variety of factors -- including the prevailing wind and topographic conditions -- to understand just how flammable vapors traveled from the blowdown drum to the area of the trailers.

In particular, we are looking at the possibility of an explosion underneath the J.E. Merit double-wide trailer where a number of personnel were meeting. We are also examining whether this particular trailer had skirting around the base. The lack of skirting would have allowed flammable vapors to travel under the trailer.

We hope to complete our field blast evaluation in the next one to two weeks. Thereafter there will be a period of computer-aided analysis. Ultimately, the information from this analysis will feed into our overall root-cause investigation.

We are continuing to focus on the raffinate splitter column which was the original source of the flammable liquid and vapor that was released to the atmosphere. Specifically, we have been reviewing computer records from the control system equipment. Preliminary indications are that there was diminished flow of liquid out of the column. This column normally operates with 20-30 vertical feet of liquid at the bottom. In fact, we believe that the column likely became flooded, a highly abnormal condition.

Over a period of several minutes, the pressure inside the column rose from about 20 pounds per square inch to about 60 pounds per square inch, which was also highly abnormal. This overpressure was more than sufficient to open three pressure relief valves on the column, which were set to open at 40 to 42 pounds per square inch. These valves remained open for six minutes, according to computerized records. During this period, it is assumed that sufficient liquid and vapor were discharged to overwhelm the blowdown system and cause the geyser-like release and the subsequent explosion. We will be adding several additional experts to our investigative team. These include experts in the design and operation of distillation columns and pressure-relief systems.

Our investigation continues as we seek to further understand what precisely caused the flooding condition and the overpressure.

In the course of our investigation, we also determined that there was a six-inch diameter drain line leading from the isom unit blowdown drum to a plant sewer. The manual valve on this line was found chained and locked open, and thus it is likely that a significant volume of flammable hydrocarbon traveled into the sewer from the blowdown drum. There is evidence that some flammable material was later released from the sewer at another location resulting in a fire, but at this time we cannot connect that fire with any of the fatalities.

Finally, we have received reports of past instances where flammable vapor clouds were vented from unflared blowdown drums at the refinery. We are following up closely on these reports and hope to have further information at a future briefing.

BP has continued to cooperate with our investigation.

Mr. Holmstrom and I have time for some questions.

Back to news