September 9, 2005 | Mark Paradies

BP Texas City Refinery Explosion Past e-News

Would you like to see all the coverage on the BP Texas City Refinery explosion that has been published in the e-Newsletter? Then click on this link. The coverage is in order of occurrence.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

From June e-Newsletter:

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _



Fatalities are perhaps the most difficult investigations.

Those who were killed aren’t talking.

Those who survived are often afraid and filled with guilt. They may be fired or disciplined because of mistakes they made. They worry about what their family, peers, supervisors, and management may think of them. They may even worry about civil or criminal charges because of mistakes that were made. And they grieve for their lost co-workers who were often good friends.

The Texas City BP Refinery Explosion seems to have all the worst aspects of a fatality investigation.

Fifteen people have died. They will never be returning to work or to their families.

And now the disciplinary actions have started.

As a result of the accident and the emergency response, it seems that three employees and one Fire Chief (who was not employed at BP but who responded from another refinery per their mutual assistance plan and who later quit over a controversy about unauthorized pictures) may lose their jobs. And this may only be the tip of the disciplinary iceberg.

BP released an interim report in a pdf format. To download it see:

The initial reaction by the press was that BP was blaming the workers for the explosion.

On May 17, Roiss Pillari, the President of Products North America for BP, and Pat Gower, Vice President of Refining for Products North American for BP, were quoted by several press sources and in a press release as having said the following:

“The mistakes made during the startup of this unit were surprising and deeply disturbing.”

“The result was an extraordinary tragedy we didn’t foresee.”

“… we have begun disciplinary actions against both supervisory and hourly employees directly responsible for operations of the Isomerization Unit …”

“The core issue here is people not following procedures…”

“If they had … followed procedures, the accident wouldn’t have happened.”

For the complete comments of Mr. Pillari and Mr. Gower in a pdf format, see:

This provoked a response from the plant’s union. Gary Beevers, Director of the United Steelworkers said: “Blaming workers doesn’t solve the problem of unsafe conditions in that refinery.”

Sam Munn, President of the Galveston County AFL-CIO, said: “I think they inherited a situation from Amoco.” Mr. Munn, who worked at a chemical plant near the refinery said: “When I worked at Union Carbide, the only thing that separated the two plants was a two lane road. They had so many problems, we used to joke that if they didn???t have a fire by ten in the morning, someone would start one.”

The press coverage seemed to imply that senior management had “accepted responsibility” but that they had shifted blame to the operators and supervisors.

To read the press coverage for yourself, see:



As someone interested in root cause analysis and investigation reports, I find the report and the reaction extremely interesting.

From my initial review of the report, I noticed that the exact causes of the overpressurization have not been identified. There is still much more testing left to be done. Problems that could have lead to the release and explosion include:

Operators overfilling the unit.

Improper venting trapping Nitrogen in the unit.

Improper draining trapping water in the unit.

Introducing water or light ends in the feedstock.

Overheating the unit.

Improper venting and feeding of the unit.

Corrosion of the unit causing operation to be modified (lower operating pressure).

I looked at the graphs provided in the interim report and saw a sharp pressure spike that caused the unit to overpressurize, the relief valves to lift, amounts of liquid and vapor hydrocarbons beyond the design capacity to be sent to the poorly maintained Blowdown Drum and Stack (F20 referred to in the report), and to result in an uncontrolled overflow of the Blowdown Drum and Stack. This loss of containment of hot, highly flammable material caused the explosion and loss of life.

The report was not clear enough about the causes so that someone not familiar with the Isomerization Unit could put the details they provided together. So I have asked Ken Turnbull, retired Safety Manager from Texaco and TapRooT(R) Instructor, to put together a SnapCharT(R) of the accident so that we at SI could understand the sequence and the relationship to the causes better. If this goes well (provides a clear understanding of the issues involved using the information available from official sources), I’ll post the SnapCharT(R) on our web site and provide a link so that everyone can see it. Then we all can learn more.

What I did spot in the report? Here’s a sample list of problems:

An old risky design (Blowdown Drum and Stack) that probably should have been upgraded to reduce the risk.

Poor maintenance of the Blowdown Drum and Stack (no service water to cool discharges and collapsed baffles).

Poorly understood risk because of the mistake belief that the operators would do the right thing and always follow procedures.

Inexperience personnel. “Upgraded – _____” seems to be how things are run. (It made me wonder, is an “upgraded” operator, supervisor, or superintendent fully qualified? Is an “upgrade” shorthand for “new” or does it mean that they really aren’t qualified for the job but they are filling in?)

Inexperienced operators are unsupervised.

Training deficiencies and misunderstanding of operating requirements for the unit being started up.

Management and supervisors who can’t agree WHO was the supervisor at the time of the accident. (Was there no standard? Could there be no supervisor?)

The supervisor calling in from off-site to give the inside operator advice about reducing pressure in the unit.

No shift turnover for the supervisor. (In fact, no on-site supervisor for the unit for at least twice during the startup.)

Inaccurate instrumentation that might have fooled the operator.

Inadequate shift turnover for plant’s inside and outside operators.

Superintendent doesn’t even know startup is taking place.

Rusty vessel (splitter) de-rated but procedure not changed.

“Most” of the operators aware of de-rated splitter.

Two different startup procedures – Inside and outside operators using procedures that don’t agree. (How are they coordinating their work?)

Procedures inaccurate and not referred to by the inside operator.

Checklists not used or misused by night and day-shift outside operators.

“Local” ways to operate things that don’t agree with the procedure.

BP accident investigators who seem to think it is OK not to follow the procedure when the procedure is wrong (pressure test of system to lower pressure than procedure requires because procedure was not updated when unit was de-rated) but then think it is bad not to follow procedures when “local practices” have diverged from the practices called for in the procedure (venting system).

Automatic controls not used.

Emergency procedures not followed.

History of un-safe siting of temporary buildings.

History of fires, injuries, and fatalities.

With so many violations and shortcuts, my guess is that these problems are more deeply rooted than the particular operators and supervisors that were on shift that night and day. My guess – and without doing a complete investigation it is only a guess – is that there are serious Management System problems that need to be addressed. Thus the discipline for supervisors and hourly workers that is referred to in the press release won’t solve these problems and is probably counter-productive.

I also had questions about the investigation. It seemed to stop at the obvious answers (procedures not used) and not address other areas. For example, I didn’t see:

Data about work hours or an evaluation of operator or supervisor fatigue.

A detailed analysis of the human actions required to successfully start up the unit or the human factors of the plant’s controls, displays, and alarms.

Why operators were NOT using procedures or why they checked off steps they didn’t perform or checked off steps they did not perform as they were written in the procedure.

Why it had become a custom to allow “local practice” to be different than the start up procedure steps.

An analysis of the communications and coordination between team members (supervisor, inside operator, and outside operator).

Why were so many safety positions “open” (un-staffed) at the refinery (according to the organization chart in the report, those positions “open” included the unit’s HSE position and the Process Safety Management positions at the West Plant – where the problem occurred).

The recommendations provided in the report seem to point to some of the areas I’ve mentioned but the information needed to analyze these areas was not included in the report. This lack of information in the report leaves one to wonder if all the specific and generic (especially the generic) problems that caused the explosion are being addressed.

What do you – the readers of the e-Newsletter – think? I would like to hear your evaluation of the report and the actions taken by management. Let me know by e-mailing me at and I’ll pass them along in the next issue.

– – –

One last item.

The press release quoted Mr. Pillari as saying: “Our goal is to provide fair compensation without the need for lawsuits or lengthy court proceedings.”

Anyone who has lost a parent, child, brother, sister, or spouse can tell you that there is no “fair” compensation when you lose a loved one. What you want is the person not to be gone. No amount of monetary compensation based on economic value can compensate for the loss suffered. What everyone wants is for BP to operate the Texas City Refinery safely, WITHOUT fatalities.

Perhaps Mr. Pillari should read the “Stop the Sacrifices” article at the TapRooT(R) web site (click on this link to download a pdf version: ) before he talks about fair compensation. Because if we went back to the code of conduct and compensation of ancient times (an eye for an eye and a tooth for a tooth), 15 BP executives would be choosing which of their children would be sacrificed to provide “fair compensation” for the loss of the 15 people who died at the refinery. Yes – this is a harsh comment. But I saw no self critical evaluation of senior management’s role in this disaster. Perhaps with personal consequences senior management would be more proactive in preventing fatalities? Perhaps procedure compliance, procedure accuracy, training, staffing, fatigue, corrosion, equipment designs and upgrades, and other problems would be fixed quickly – BEFORE a major accident occurs.

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

From July e-Newsletter:

– – –


News about the Texas City BP Refinery Explosion has slowed since the last e-Newsletter, but several people had comments that I thought I would list here.

Because of the controversial nature of this explosion and fire, I decided NOT to list the names and companies of the readers who sent me comments. Rather, I will just list them under the titles of Reader A, B, C, …, with a description of their job type.

– – –

Reader A (Equipment reliability and root cause analysis professional working at a refinery in the US. This response also included a word document with some analysis using TapRooT which isn’t included here because of the preliminary nature of the analysis.):


The level of detail in the report makes it difficult to determine exactly what TapRooT categories apply. I tried analyzing it and several times I had to make a call between training, communication, or work direction. Therefore, by no means is my analysis necessarily going to agree with someone else’s interpretation. Unfortunately the report itself is not presented in TapRooT language (you likely noticed this also.)

Something that bothers me about the report is that there is no explanation (or interest?) into what caused the splitter to pressure up at least twice before the incident. Seems like: pressure high -> opened 8″ bypass valve -> pressure high -> opened 1.5″ bypass valve -> pressure high -> kaboom. Looking back I am left wondering if there were two missed signals that valves were not lined up properly (consistent with “local practice”) that remained un-addressed as things continued to warm up.

I look forward to reading more about this on your website.

And my response to Reader A:

Thanks for the information – it was helpful to see the insights of someone more familiar with the operation of these types of units.

Also, it was good to hear that you had doubts about what categories apply because I think they haven’t come close to giving me enough information to reach root causes.

One interesting “tidbit” that I saw was that the Supervisor who left the site for “personal” reasons, called in on his cell phone and gave the board operator advice on how to reduce pressure. This led to the opening of the 1.5″ bypass which did temporarily lower pressure.

The other thing that I noticed looking at the graphs was that the previous pressure increases were gradual – perhaps due to heating and feeding the unit. But the final pressure increase was a “spike” – very rapid. Did some chemical reaction, onset of boiling, introduction of water or lighter feed, or other rapid reaction take place to cause the rapid pressure spike? It seems like the report didn’t explain this very well.

I don’t think the operators had a good idea what was going on in the plant. I wonder about their instrumentation/indications and what they “thought” was happening.

I also wonder about old operating practices and new practices since the since the unit was modified and the unit was de-rated (lower pressure).

It doesn’t seem likely that they would “purposefully” violate procedures, overfill and overheat the unit, and then ignore critical indications when their relief system went to atmosphere with no flare.

It seems more likely that they had become accustomed to operating things by the seat of their pants (even though they weren’t very experienced) and that on this day, they didn’t understand what was happening, got distracted and confused, didn’t have anyone with “experience” to help them get back on the right track, and then the events overwhelmed them.

That’s my guess from what I read. But an investigation should not rest on guesses. I think the investigation is a long way from the actual sequence of events – much less the root causes.

Thanks again for your information.


– – –

Reader B (US government accident investigator):

I saw your comments about the BP Amoco matter in you June 05 newsletter and I found them quite interesting.

– – –

Reader C (equipment reliability and root cause analysis expert with extensive worldwide experience in petrochemicals/refining):

Very outstanding analysis of BP/AMOCO Texas City disaster.

Question for those executives about results from “testing”: What exactly needs to be tested? That massive amounts of volatile hydrocarbon gases and/or liquids will ignite when they escape?

Where was the quench system? Who deleted it? When was it deleted? Where was their Management Of Change?

When quench system was deleted, added operator vigilance and uncompromising adherence to procedures should have taken its place.??There are no excuses here.

Mark, keep preaching TapRooT. Folks that don’t believe in the validity of the TapRooT approach don’t belong in sensitive industrial environments.

– – –

Reader D (root cause analysis and incident investigator from the chemical industry):


Your insights on the BP interim report and subsequent press releases were discerning, as usual.

My reaction is one of compassion in the broadest terms: compassion for the family and friends of the victims; compassion for employees and management who did not intend for this to happen; compassion for an impatient citizenry who need transparency and closure; and compassion for pressured investigators who must be efficient, accurate and thorough lest they make matters worse.

I believe that, separate from this investigation, the purpose of interim reports should be clarified for the benefit of all future investigation sponsors and participants.??

Based on involvement in many investigations, my opinion is that the purpose of periodic status reports should be to apprise stakeholders of progress, or lack of it, during lengthy investigations. The nature of the investigation process is such that the deepest factors are the last to be identified and validated. For this reason, investigators should be wary of offering ???provisional??? results and ???proposed??? recommendations as part of interim reports. For the same reasons, readers of interim reports should be wary of grasping shallow symptoms or extrapolating to deeper, but as yet unsubstantiated, claims.

Specifically, the BP interim report did not recommend any disciplinary action. The recommendations that it did ???propose??? were only proposals, and were all aimed at improving management controls for future operations, not selectively punishing individuals for errors that the organization set up and allowed to happen. I regard the management decision to discipline as a misapplication of an overly-ambitious interim report.

– – –

Reader E (root cause analysis and incident investigator from the Australian oil/chemical industry):

My colleagues and I have been reading the available information on the BP

Texas City incident with keen interest. We have been quite amazed by the

release of an “interim” report when there seems to be a lot of

investigation work still to be done, and also by accompanying suggestions

that “heads will roll” (an oldie but a goodie – blame the operator). I

wonder whether these are driven by BP’s PR department in an effort to be

seen to be responsive/open/proactive/willing to take firm & decisive

action/etc? It gels with their comments about compensation for victims’

families and avoiding “unnecessary” court action. Sounds like damage


I have a few observations on the incident report itself. I think that what

the incident report has done is to identify causal factors – what they have

termed “provisional critical factors”. Clearly each one of these factors –

the overpressure of the splitter; the design of the relief system; the

location of the trailers; and the use of startup procedures – had a major

role to play in the incident. But in themselves they are not root causes.

There seems to have been a lack of further analysis of each of the causal

factors in terms of what the underlying reasons (and ultimately the root

causes) were, and a lack of willingness to get down to systemic problems

rather than stopping with symptoms.

The report is very wordy and contains a lot of description but is very

light on in terms of clearly illustrating the flow of events from a causal

analysis point of view (they really should have used a SnapCharT!). It is

not clear from the analysis of each causal factor what they consider the

key root causes to have been – ie there are no conclusions! I also was

surprised to notice that there are many recommendations which appeared to

address issues which are not raised in the causal analysis discussion (eg

“staffing plans for turnarounds and high workload periods must show

explicit consideration for fatigue”).

Some general comments on the recommendations. I find it interesting that

the team feels able to make recommendations without completing their causal

analysis. Are these only preliminary recommendations that are likely to be

changed at a later date? If so, why bother making them so public at this

stage? How do they know that their recommendations are really addressing

the root causes of the incident? Or are the recommendations really just a

PR exercise? In general the recommendations themselves are quite generic &

vague. Hard to know whether they would pass the test of being specific,

measurable etc (SMARTER).

There are issues (including Management of Change, basic Hazard

Identification/Risk Management especially of unusual operations) that do

not seem to have been addressed in the report.

I note also that the recommendations from this report don’t mention

anything about disciplining anyone. So on what basis would BP’s management

be commencing disciplinary action?

Waiting with interest for the Final report!

– – –

Reader F (equipment reliability expert at a South American paper manufacturing site):

First of all Mark, congratulations for this excellent and motivating report.

I’m glad that that this e-Newsletter reported on big disaster that we hear about around the world (for example, the Texas City BP Refinery explosion). Many of the 9,400 TapRooT(R) Friends/Experts are interested and want to know more but don’t have time to find the information.

I personally see that it would be beneficial if TapRooT(R) was used as the investigation process for all these fatal accidents. We would all be better off if investigators of these major accidents found root causes and developed recommendations to avoid their recurrence in all major investigations around the world.

I will be looking forward to more information excellent material to study and to use for training.

My best regards to all on the TapRooT(R) staff and I look forward to the Friends/Experts promptly reaching 10,000.

– – –

Reader G (root cause analysis and incident investigator from the Canadian oil/chemical industry):

Mark, from my reading of the BP Texas City Report that you sent out with this newsletter, it seems to me that in the case of “Local Practices” it smacks of Normalization of Deviance, where we do things because it is always the way we do things. There is not usually enough thought or Risk Assessment put into these deviations.

Plant management (not only BP’s but others as well) empower their employees for this type of behavior as well, by giving out incentives or rewards for getting a job done quickly and a unit up and running in a shorter period of time.

The other thing that causes NOD to happen is the cut back in staffing levels, we are asking our people to do more with less resources (other bodies) or when people leave they are not replaced and the work is distributed among those that are left. This and the fact that every day a unit is down it is costing the company money. So it is implied that operations and maintenance should find ways to get the unit back into production as quickly as possible.

Companies talk about safety but what is the price? By allowing NOD we are setting employees up for Failure! You can only get away with “Local Practices” for so long until it comes back to bite you and you have a more costly outage or a loss of life incident happens.

If management would have the guts to not reward bad habits and give people the resources and time to do a job by the book, not by “This is the way we have always done it!” attitude, then these incidents would be totally preventable!!!

Thanks for the great information and newsletter.

– – –

Reader H (oil industry safety engineering services manager):

Your comments on the recent events at Texas City are interesting. I find

that very many people are (independently) comparing it to the Esso Longford



– – –

On the on-going investigation front, the CSB held a press conference on June 28 (after the items above were written) about their continuing investigation. To see the statements that were released, go to:

One key item to note is that the CSB investigation has discovered that key alarms did not come in to warn the operator of a problem with the unit.

– – –

I mentioned last issue that Ken Turnbull is working on a SnapCharT(R) of the accident. This continues, but, as you can see if you read the CSB news releases, new information about the explosion continues to be released by various investigations and both Ken and I agree that the chart is still too “preliminary” to release for publication. Perhaps next month…

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

From August e-Newsletter:

– – –


We continued to receive comments about the Texas City BP Refinery Explosion. Two more (one with an article attached) are provided below. Also, all the original comments and the original article I wrote are attached below as well.

The interest expressed by readers made me think … “Shouldn’t we have a talk about this investigation at the TapRooT(R) Summit?”

I decided that we should and asked Mark Kaszniak, investigator at the CSB for help. It turns out that he is on the investigation team and he was willing to share his experiences at the Summit. So if you are interested in talking to one of the CSB investigators, attend the Regulatory Investigations breakout and hear Mark talk about “BP Texas City Refinery Explosion Investigation“.

Now for the new comments from readers:

– – –

Reader I (Safety Coordinator at Canadian Chemical Manufacturing Site):


I work for a petro-chemical company at a chemical manufacturing site. We operate similar vessels and equipment as BP would in Texas.??

I have been following the BP incident with great sadness knowing that this event occurred in a day and age when the technology and knowledge from lessons learned are completely capable of avoiding this kind of loss. And yet 15 people have lost their lives to say nothing??of the overall wake of loss caused by this unacceptable event.

I agree with your assessment of probable root causes??to this point.????I know the list will be much longer by the time all the facts??are??revealed.

It is interesting what you??mentioned about the??pressure spike happening shortly after they managed to depress the??vessel somewhat.????It sounds as if the depressing may have been the wrong thing to do at that time. If the vessel temperature was rising along with the pressure, the sudden depressurizing would almost certainly cause the??liquids in the vessel to suddenly flash off which would do two things. The boil up action would cause sudden??rise in the liquid level and the flashing would cause a massive pressure spike as the liquids are converted to vapor. This is purely conjecture as I would need to see the all the measurement variables. The same data??which the supervisor may not have had a strong grasp of before he recommended using the valve to depress the vessel.

The fact that there was no supervisor on site during start-up is??unbelievable. This is indicative of complacency of the worst kind and the fact that the supervisor thought it was alright demonstrates that management has not been communicating or enforcing SPAC very well.??That’s if SPAC required him to be on site in the first place.

I look forward to seeing the SnapCharT(R) on this. As it is I am already in awe at all the red flags that were ignored leading up to this event.

– – –

Reader A sends more info:

Some of these questions look familiar to those seen on the TapRooT(R) e-Newsletter . . .

From: Roland Goodman

Sent: Wednesday, July 27, 2005 9:31 AM

Subject: Fatal Ignition: An Oil Giant Faces Questions About a Deadly Blast in Texas

Fatal Ignition: An Oil Giant Faces Questions About a Deadly Blast in Texas —

BP Acknowledges Troubles At Key Refinery Unit But Blames Employees — Explosion in March Killed 15

TEXAS CITY, Texas — After a huge explosion at a BP PLC plant here in March killed 15 workers in the deadliest U.S. petrochemical-industry accident in 15 years, the company put most of the blame on some of its employees. But now the search for the cause is raising some unsettling questions.

London-based BP acknowledges faulty equipment and other troubles at the unit at the Texas City refinery where the explosion occurred, although it doesn’t directly connect those problems to the accident. In addition to the deaths, more than 170 people were injured.

Two federal agencies have pointed to safety problems at Texas City. And all this comes against the backdrop of cost-cutting in an industry with very old equipment.

Yesterday, BP said that it had set aside $700 million to cover compensation for victims’ families and those injured in the March 23 blast.

But the company could still face regulatory action. Managers at the Texas City refinery “don’t have a culture of safety that looks at all the details,” said John Miles, Southwest regional director for the Occupational Safety and Health Administration. OSHA is investigating the deadly explosion. Mr. Miles notes that over the past 16 months, there have been two other major accidents at Texas City. In September 2004, two employees were scalded to death while removing a valve from a hot-water line. In March 2004, the refinery had a big, nonfatal fire.

Three significant accidents in that span of time are “a concern,” Mr. Miles said. In addition, in May 2004, a contractor was killed in a fall at the refinery, although the contractor’s employer, not BP, was cited.

In April, OSHA put the BP refinery on a national watch list of safety violators in all industries. OSHA doesn’t publicly release the list but confirmed that BP is one of only two refiners among some 700 companies identified. OSHA said the other is Coffeyville Resources Refining & Marketing LLC, a much smaller operation in Coffeyville, Kan. Coffeyville Resources said that it believes its inclusion on the list reflects the accidental death of an employee in May 2004. The company said that overall it “enjoys a good safety record.”

Another federal body, the Chemical Safety and Hazard Investigation Board, is conducting its own probe. Spokesman Daniel Horowitz said the board so far has identified a number of safety problems at Texas City.

BP publicly issued a report in May that accused some of its mid-level plant managers and hourly workers of “surprising and deeply disturbing” mistakes. These errors allegedly included a failure to provide supervision and disregard of company procedures. BP has removed the refinery’s top manager and fired a handful of other employees — but won’t say how many.

A lawsuit filed in state court in Galveston claims that the ultimate blame lies with BP. The suit alleges corporate “gross negligence” and seeks unspecified damages on behalf of victims’ families and several people injured in March at the sprawling refinery near Galveston Bay.

One of the suit’s original plaintiffs was David Crow, a 58-year-old outside maintenance contractor. In the early afternoon on March 23, he heard a thundering noise and then was buried beneath debris. He fractured six vertebrae and shattered his foot and ankle. “People lost their lives over something that could have been avoided,” said Mr. Crow, who now uses a walker to get around. After being interviewed for this article, he agreed to settle his claim against BP on confidential terms.

The explosion occurred when workers restarted a small gasoline-processing unit, an operation that is inherently tricky because many controls are adjusted at once and combustible substances begin moving rapidly.

BP is walking a fine line in its response to the accident. It has denied any negligence and is cooperating with federal investigators. But it also has apologized and promised to continue settling with survivors and victims’ families. Many settlements have already been agreed to, the company said, without providing details.

BP has denied any connection between cost-cutting and plant fatalities. It contends that overall safety at its American refineries has improved since it acquired them.

“I think the culture of safety, in terms of policies and procedures, was there,” said Ross Pillari, president of BP Products North America. “But the implementation of these policies and procedures was clearly not there, because if it was, the accidents wouldn’t have happened.”

BP has five refineries in the U.S. Two others that, like Texas City, were acquired during a buying spree started in the late-1990s have also had worker deaths recently.

On New Year’s Day 2004, a technician at a plant in Whiting, Ind., fell and cracked his skull after a corroded handrail gave way. BP investigators concluded in an internal report, separate from the one issued in May, that there hadn’t been a procedure to inspect and repair the facility’s handrails, which date to the 1940s. BP said that it has since inspected all handrails at its refineries. Indiana regulators fined BP $1,625 over the incident.

In May, a contractor was found dead at BP’s refinery in Cherry Point, Wash. An initial company investigation has found no evidence that the death was related to an accident, a person familiar with the inquiry said. The death is under investigation by the county coroner and state safety regulators.

Even excluding the Cherry Point death and the 15 fatalities in March, BP’s four other deaths since January 2002 are more than the number recorded by its main rivals in the U.S., according to federal data and information provided by the companies. BP is America’s third-largest refiner. No. 1 ConocoPhillips and No. 2 ExxonMobil each had one death during that period.

There hasn’t been a new oil refinery built in the U.S. since 1976, and most plants show their age. New construction has been discouraged by stricter environmental rules and other factors. Regulators say that old facilities don’t necessarily pose unusual dangers, as long as effective safety and maintenance programs are in place.

The world’s second-largest publicly traded oil company, BP operates internationally and does everything from drilling for oil to running gasoline stations. Its global safety record — measured by hours workers miss because of injury and accidents that merit reporting to regulators — is one of the industry’s best.

Using the motto “Beyond Petroleum,” the company has promoted its image for social responsibility. Chief Executive John Browne has championed attempts to reduce global warming by lowering carbon-dioxide emissions at BP facilities.

The company grew suddenly when it purchased Amoco in 1998 and later scooped up Atlantic Richfield Company (ARCO). BP’s success in improving efficiency and lowering costs at its newly acquired American refineries has been shown in a periodic survey done by the Dallas consulting firm Solomon Associates. Widely used for comparing refinery performance, the survey showed that in 2000, BP’s U.S. plants ranked in the middle of the industry on measures such as return on investment. By 2004, BP rose to the top 25%. Last year, BP reported overall profit of $15.7 billion on revenue of $294.9 billion.

The municipality of Texas City has long been an industrial hub. In 1947, it suffered one of the country’s worst industrial accidents ever when a ship full of fertilizer exploded, killing nearly 600 people.

BP acquired the Texas City refinery from Amoco. In the 1990s, Amoco had reduced the plant’s unionized work force by 19%, to 1,300 people, according to Sonny Sanders, a former Texas City employee and longtime labor-union official. Under Amoco, major maintenance overhauls, called “turnarounds,” became less frequent, said Mr. Sanders, now a United Steelworkers representative. BP said it wasn’t in a position to comment on Amoco’s actions. The steelworkers union, which represents BP employees, has challenged the company’s findings on the blast and is conducting its own probe.

As it absorbed its American acquisitions in 1999 and 2000, BP cut its work force of U.S. refinery employees and contractors by 10%, largely by means of buyouts in Texas City and Whiting, Ind. At Texas City, the staff of unionized maintenance craftsmen and operators fell by 213, or about 18%, the company said in written answers to questions for this article. The reductions were partially offset by greater use of outside contractors, BP added.

“The approach to reducing costs was well thought out and systematic,” BP’s Mr. Pillari said. It “does not appear, in so far as we have seen, to have had anything to do with the fatalities” at Texas City or anywhere else, he adds. The company said in written answers that it has steadily increased overall spending on maintenance in the U.S. At Texas City in 2003-2004, BP said that it spent 40% more per barrel of oil it refined than was spent in 1997-1998 under Amoco. BP declined to disclose dollar figures.

Current and former Texas City employees and contractors paint a different picture. Under BP, the refinery deferred some routine maintenance inspections because of staff shortages, according to three former employees and one current worker. In addition, certain safety procedures have been ignored at the plant, according to seven people who have worked there. Contractors and BP employees sometimes work high above ground without proper safety gear, according to four of these people. BP said that it requires strict compliance with its policies for working at an elevation.

But OSHA’s regional director, Mr. Miles, said that managers at other Texas refineries he has inspected, including one nearby owned by Valero Energy Corp., are more actively involved in safety issues. “You don’t see that down the street” at BP, he said.

Mr. Crow, the veteran maintenance contractor who was injured in March at Texas City, said disrepair at the plant was worse than what he has seen at comparable refineries. He said he was particularly troubled by corroding metal springs that hold refinery pipes in place. “Everything out there is rusty,” he said.

Glenn Alexander, a 45-year-old electrical contractor who suffered shoulder and back injuries in the March blast, said corrosion plagued much of the refinery. Last year, he said, a metal structure supporting power and communications lines high above ground collapsed because of corrosion. No one was injured. Another section of the same sort of structure “was wobbly and could have fallen any minute,” he said. Mr. Alexander was a plaintiff in the negligence suit against BP but agreed to a confidential settlement after he was interviewed.

The company said that it aggressively addresses corrosion, sometimes shutting down certain operations to do so. Citing the large size of the plant, BP said it couldn’t confirm the incidents Mr. Alexander described.

Last year, before the March explosion, BP shook up its maintenance-inspection program at Texas City, replacing the top inspectors. But the company emphasized that inspection-staffing levels weren’t “a critical issue.”

BP executives said that overall, the company’s U.S. operations are sound. Still, it is auditing safety procedures at all of its U.S. plants, trying to determine if any are “resource constrained,” BP’s Mr. Pillari said.

The blast in March occurred in a part of the refinery known as an isomerization unit, or isom. It processes gasoline to boost octane, which prevents engine knocks. BP’s report on the accident in May details a series of past problems at the isom but doesn’t link that history to the accident.

Since 1986, employees had recorded seven fires at the isom, not including the March blast, the company said. This small section of the refinery consisted of a control room, a 164-foot processing tower and venting equipment, which included a 113-foot “blowdown” stack. The blowdown stack released buildups of dangerous liquid or vapor in emergencies.

BP’s report said that a water system used to cool the venting equipment had been out of service for some time before the huge blowup. Baffles, which are metal plates that aid the condensation of vapor, had corroded badly.

In April 1997, Steven Adams, an isom-unit operator, filed a suggestion to Amoco management to replace a faulty pump connected to the venting system. If the pump failed, combustible liquids could overflow into the refinery’s sewer system and possibly ignite. Amoco apparently never acted on the suggestion. Six months after BP took over the plant, the idea was rejected as “not cost effective,” according to an internal company memo.

Mr. Adams was one of the employees fired after the March blast. Last month, he filed a defamation suit against BP in Texas state court, alleging that the company ruined his reputation by unfairly blaming him and others for the deadly explosion. BP declined to comment on the Adams firing or the suit.

The decision not to replace the isom unit’s pump could prove significant. In its report, BP said it was possible that combustible material escaped from the venting system into the sewer and ignited there, setting off the blast. But in its written responses to questions for this article, BP said there is no indication the pump malfunctioned.

In February, operators of the isom unit shut it down for maintenance. Mr. Crow and dozens of other outside contractors were working out of trailers nearby. BP said in its report that the contractors weren’t warned about the potentially dangerous restarting of the unit, as they should have been.

The isom unit’s daytime supervisor, Scott Yerrell, arrived late the morning of March 23, according to BP’s report, which didn’t identify him by name. Then, just before 11 a.m., he left the plant for personal reasons, the report said. BP fired him after the accident.

Mr. Yerrell is a co-plaintiff in the defamation suit against BP. His attorney, Charles Dunkel, said that his client left to attend emergency surgery for his son, who had broken his arm the day before. Mr. Yerrell lined up a replacement to supervise the start-up and shouldn’t have been fired, the lawyer said.

By 12:40 p.m., pressure and heat had climbed above normal levels in the isom unit’s processing tower, where gasoline components are separated, the BP report said. A worker outside the unit called the control room to report steam-like vapor coming from the unit’s smoke stack, an indication that something might have been going wrong.

At about 1:20, a small explosion occurred, followed by a much louder one. Some people on the scene said that they heard a series of explosions. The blasts shook windows in houses miles away. BP said the most likely cause was that explosive vapor and liquid ignited after operators overfilled the processing tower and the volatile mix vented out of the stack.

While continuing to press its investigation, the federal Chemical Safety Board has said that some of the isom unit’s alarms and indicators appear to have failed to notify workers about dangerous levels of liquid building up in the unit’s processing tower. The board is also studying why BP placed the contractors’ trailers so close to the isom plant.

BP said that the alarms weren’t a critical factor in the accident. In its May report, it acknowledged that the trailers were too close to the isom.

Mr. Alexander, the injured electrician, watched the destruction of the trailer in which his wife, Lorena Cruz-Alexander, was working as an administrator. He identified her remains several days later. “If I had known a unit was starting up, Lorie wouldn’t have been anywhere near this site,” he said. “I wouldn’t have been anywhere near the site.”<span style=”font-fami

– – –

Next, CSB has issued more information since our last e-Newsletter. Here is the link:

and a summary press release from CSB:

Houston, Texas, August 17, 2005 ??? The U.S. Chemical Safety and Hazard Investigation Board, (CSB) today issued an urgent recommendation calling on the BP Global Executive Board of Directors to commission an independent panel that would review a range of safety management and culture issues in the wake of recent chemical accidents at BP???s south Texas facilities.

The independent panel would focus on BP???s oversight of its five North American refineries in Texas City, Texas; Carson, California; Whiting, Indiana; Cherry Point, Washington; and Toledo, Ohio. Four of the refineries were acquired through BP???s mergers with the former Amoco and ARCO corporations.

The CSB is currently investigating an explosion and fire which occurred at BP???s Texas City refinery on March 23, 2005, killing 15 workers and causing about 170 injuries. It is the first safety recommendation in the agency???s eight-year history that has been designated as ???urgent??? and issued in advance of a completed investigation. In voting to adopt the recommendation, the Board said that identified safety management lapses pose an issue that ???is considered to be an imminent hazard and has the potential to cause serious harm unless rectified in a short time frame.???

Citing a series of serious safety incidents at the Texas City facility over the past two years, the Board recommended that BP commission and fund a diverse panel of experts, including employee representatives. The panel would review corporate safety oversight, safe management of refineries obtained through mergers and acquisitions, corporate safety culture, and management systems such as near-miss reporting and mechanical integrity programs.

CSB Chairman Carolyn Merritt announced the action at a news conference in Houston, near the Texas City refinery, which experienced two fatal safety incidents last year and has had two additional serious incidents since the fatal March explosion, including a hydrogen fire and a gas release. She also praised BP???s recent commitments to relocate nonessential personnel from its Texas City refinery and to eliminate hazardous atmospheric vents at its refineries in Texas City and Whiting.

Chairman Merritt said, ???Today, the Chemical Safety Board is issuing an urgent safety recommendation to BP America and BP???s Global Board of Directors. This is the first urgent safety recommendation in the Board???s eight-year history. The Chemical Safety Board recommends that BP immediately convene an independent panel of experts to examine BP???s corporate safety management systems, safety culture, and corporate oversight of its refineries. The panel should report its findings and recommendations to the BP workforce and the public.???

Merritt said the work of the panel would dovetail with CSB???s investigation of the March 23 tragedy, which will continue to focus on uncovering the specific root causes of the incidents as well as generate recommendations for national changes to prevent a recurrence.

The panel should include safety experts from a wide variety of sectors, such as aviation, space exploration, nuclear energy, and the undersea navy, as well as the process industries. The panel must be independent from BP and have an external chairperson as well as labor representation.

Chairman Merritt cited the Columbia Accident Investigation Board, which investigated in the 2003 Space Shuttle reentry disaster, as one of the models for the independent panel. She said the CSB was requesting that BP develop an implementation plan for the recommendation within 30 days and complete all work within six to twelve months. The CSB will not serve on the panel but will track and evaluate progress in implementing the recommendation, with periodic reporting to the public.

The March 23 incident involved a sudden release of flammable hydrocarbon liquid and vapor from an atmospheric vent stack in the refinery???s isomerization, or isom, unit. Workers in nearby trailers were killed and injured in the subsequent explosions.

The 114-foot tall stack, which dated from the 1950s and was not tied in to a safety flare system, was overfilled with hydrocarbons during the startup of a raffinate splitter tower, a 164-foot tall distillation column that became flooded with at least 120 vertical feet of liquid. Normal operating levels in the tower are less than 10 vertical feet. The flooded tower experienced a sudden pressure increase, opening relief valves and venting hydrocarbon liquid and vapor that overwhelmed the vent stack and its associated blowdown drum.

The urgent safety recommendation was accompanied by new and more detailed CSB findings that were reported at the news conference. The findings included:

– Key alarms and a level transmitter failed to operate properly and to warn operators of unsafe and abnormal conditions within the tower and the blowdown drum.

– The startup of the raffinate splitter was authorized on March 23 despite known problems with the tower level transmitter and the high-level alarms on both the tower and the blowdown drum; for example, a work order dated March 10 and signed by management officials, acknowledged that the level transmitter needed repairs but indicated that these repairs would be deferred until after startup.

– The majority of 17 startups of the raffinate splitter tower from April 2000 to March 2005 exhibited abnormally high internal pressures and liquid levels ??? including several occasions where pressure-relief valves likely opened ??? but the abnormal startups were not investigated as near-misses and the adequacy of the tower???s design, instrumentation, and process controls were not re-evaluated.

– Written startup procedures for the raffinate splitter were incomplete and directed operators to use the so-called ???3-lb.??? vent system to control tower pressure, even though the pressure-control valve did not function in pre-startup equipment checks and also failed to operate effectively during post-accident testing.

The full text of the Board???s urgent safety recommendation will be available on the web at after 10 a.m. CDT, on Wednesday, August 17, 2005.

The CSB is an independent federal agency charged with investigating industrial chemical accidents. The agency???s board members are appointed by the president and confirmed by the Senate. CSB investigations look into all aspects of chemical accidents, including physical causes such as equipment failure as well as inadequacies in safety management systems.

The Board does not issue citations or fines but does make safety recommendations to plants, industry organizations, labor groups, and regulatory agencies such as OSHA and EPA. Please visit our website,

Show Comments

Leave a Reply

Your email address will not be published. Required fields are marked *