JORDAN BARAB
Deputy Assistant Secretary
For Occupational Safety and Health

National Petrochemical and Refiners Association
National Safety Conference
San Antonio, Texas
Wednesday, May 19, 2010


"Learning From Industry Mistakes"

Good morning. I sincerely want to thank NPRA's Director, Lara Swett, for inviting me to join you this morning in what we all know is going to be a difficult but necessary conversation about a crisis in our nation's refineries.

I appreciate the opportunity to speak to so many of you at one time, about a subject that's important to all of us, and particularly now. The timing of this conference has turned out to be astoundingly fortunate -- fortunate because this conversation and the remedies we all need to talk about can't wait any longer.

The ringing in your ears an hour or two ago wasn't just the alarm clock by your bed. It was a wake-up call for everyone in your business.

The headlines of refinery worker injuries and deaths on the job and of OSHA's stepped-up inspections are sounding an alarm about an industry-wide problem -- a problem that we are obliged to address.

Bluntly speaking: Your workers are dying on the job and it has to stop.

Speaking a little more gently, I want you to know that Labor Secretary Hilda Solis and OSHA Assistant Secretary David Michaels asked me to convey their appreciation to the safety and health professionals attending this conference. They... and I... realize that you strive every day to make sure your fellow workers go home safe and sound.

While you don't always get credit for doing your job -- how often do you hear "thank you" when no one gets hurt? -- OSHA recognizes that you are America's quiet heroes and you deserve our thanks.

After BP-Texas City, Have We Learned Anything?

Now, in spite of your efforts, we have to acknowledge that something is desperately wrong. The status quo isn't working.

In the past three months alone, 58 workers have died in explosions, fires and collapses at refineries, coal mines, an oil drilling rig, and a natural-gas-fired power plant construction site.

OSHA is particularly concerned about the recent number of serious incidents at refineries that have scalded, burned or struck down your fellow workers. We are tracking these catastrophes and looking for trends -- including problems resulting from aging facilities.

Since the BP Texas City explosion in 2005, OSHA has counted over 20 serious incidents in refineries across the country.

Last year, OSHA completed an investigation of a naphtha piping failure and release, in which the resulting explosion and fire seriously injured three workers; two other workers, relatively young at 49 and 53 years old, died. One of these two workers was killed in the explosion; the other struggled for 13 days in the hospital before dying from severe burns. Within the unit where this rupture occurred, OSHA discovered multiple pipes that were operating below their retirement wall thickness. In fact, the very line that ruptured had previously ruptured and had to be replaced a decade earlier. As this tragedy makes clear, this type of breakdown maintenance is simply unacceptable. Good mechanical integrity programs are absolutely essential to safe refinery operation.

In 2007, water freezing in liquid propane piping resulted in a jet fire and a rapid evacuation of the entire refinery. Three workers, aged 33, 35, and 42 were seriously burned and hospitalized. Investigators found that a Process Hazard Analysis team had recommended installing remotely operable shut-off valves, yet the recommendation was improperly closed as "complete" by the previous owner. In fact, the valves had not been installed at all. The lack of these shut-off valves impeded workers' ability to control the propane release before it ignited. The refinery learned a hard lesson: It is essential to rigorously follow up on PHA findings to ensure that hazards are adequately controlled. Failure to abate serious hazards can have deadly consequences.

In 2008, at another facility, an explosion in a hydrocracking unit blew the head off a process water filter. The debris struck and killed a foreman; he was 53 and had been with the company for 30 years. OSHA's investigation revealed that an inadequate start-up procedure had allowed hydrogen gas and air to accumulate in the top of the filter where it was likely ignited by pyrophoric deposits. OSHA learned that some operators had recognized the hazard and used an undocumented alternate approach that was actually safer, but the procedure had never been updated to incorporate the safer practice. The result of following the faulty procedure was a violent explosion and the needless death of a refinery worker - and a reminder that having safe, complete, and accurate operating procedures is essential to safe operations in process units.

What do these incidents have in common? They point to process safety-related problems and, most likely, systemic safety and health problems -- in the company and in the entire industry.

Three Concepts for Change

In the brief time we have together this morning, I'm going to present you with three concepts that I believe can help you, as safety professionals, to save more workers' lives.

First: Effective process safety programs and strong workplace health and safety culture are critical for success in preventing catastrophic events

Second: This industry needs to learn from its mistakes. We know the major causes and we know the remedies. Systemic reform is needed now.

Third -- and I'm not telling you anything you don't already know: Numbers don't tell the whole story. Focusing on low DART rates alone won't protect you from disaster. New metrics are needed.

Now, let's explore these ideas a bit more.

Let's look at Concept Number One: Effective process safety management systems and workplace safety culture are critical for success in preventing catastrophic events.

I'm sure you are familiar with the Baker Panel Report issued after the BP Texas City explosion. The panel dedicated considerable space to the importance of effective process safety systems and the need to build a strong corporate safety culture.

Process safety failures are typically low-frequency but high-consequence events. Our PSM systems have to be strong, and we can't wait until we have an incident to discover that they were not.

To ensure strong PSM systems, we need to do a better job of identifying useful leading indicators. We all know the warning that "past performance is no guarantee of future success." This is particularly true of the low-frequency, high-impact events that process safety programs guard against.

The petrochemical industry must continue to develop and track leading indicators to measure the performance and continuously improve process safety management systems. Recent work by the Center for Chemical Process Safety and API is a good start.

However, real progress will come when companies and industry groups can talk to the press about their progress on valid metrics instead of holding up their injury and illness rates after a disaster.

What it comes down to is organizational culture. To paraphrase Professor Andrew Hopkins (whose work you should all be reading), workplace culture is not just an educational program that gets everyone to be more risk aware and think "safety first." Hopkins and the Center for Chemical Process Safety have defined culture simply as "the way we do things around here."

What I'm talking about is a set of practices that define the organization and influence the individuals who make up the organization. It goes without saying, but I'll say it anyway: Organizational safety culture must come from the top.

Next: Concept Number Two: This industry must learn from its mistakes.

OSHA has had in place a Refinery Process Safety Management (PSM)-National Emphasis Program (NEP) for nearly three years -- since July 2007. Consequently, we are deeply troubled by the significant lack of compliance we are finding in our inspections, and with the number of serious refinery problems that continue to occur.

Time and again, our inspectors are finding the same violations in multiple refineries, including those with common ownership -- a clear indication that concerns and findings are not being communicated across corporations or throughout the industry or even within different units in the same refinery.

Consistently throughout the course of the Refinery NEP,
we have found that over 70 percent of the citations fall into the top four PSM elements:

• Mechanical Integrity
• Process Safety Information
• Operating Procedures
• Process Hazard Analysis

Let's talk about these top four elements:

1. In MECHANICAL INTEGRITY, problems include failure to perform inspections and tests, and failure to correct deficiencies in a timely manner. This is a particular concern given the aging of refineries in the United States.
2. PROCESS SAFETY INFORMATION, including failure to document compliance with Recognized and Generally Accepted Good Engineering Practices to keep process safety information up to date, and to document the design of emergency pressure relief systems.
3. OPERATING PROCEDURES: failure to establish and follow procedures for key operating phases, such as emergency shutdowns, and using inaccurate or out-of-date procedures.
4. PROCESS HAZARDS ANALYSIS, including lack of attention to human factors and facility siting, and failing to address PHA findings and recommendations in a timely manner-- or, all too often, failing to address them at all.

A year ago, OSHA sent a letter to every petroleum plant manager in the country, informing them of these frequently cited hazards. Yet, a year after that letter went out, our inspectors are still finding the same problems in too many facilities.

I urge you to take advantage of this information, and to treat our stepped-up inspections under the NEP as an opportunity -- to increase your focus on Process Safety Management to reduce the number and severity of process incidents in your facilities.

Finally, Concept Number Three: The problem with numbers.

In any business or organization, one of the problems we find when trying to measure performance is determining how and what we measure. Companies have good tools for measuring and managing personal, or "hard hat" safety, and the refining and chemical sectors have generally done well in this area.

Unfortunately, as we've discovered, having good numbers on your OSHA 300 logs doesn't correlate with having an effective process safety program. The classic example of this is BP-Texas City, which had very good injury and illness numbers prior to the 2005 explosion. That tragedy, of course, revealed serious process safety and workplace culture problems at the facility.

Don't misunderstand me: We need to keep reporting and tracking the numbers -- DART rates are useful -- but you must not let those numbers lull you into a false sense of security. Looking only at these numbers doesn't warn us about pending doom from cutting corners on process safety.

Now, this is not new information for any of you, nor is it new to us or the press. But it is extremely upsetting. I cannot say too strongly to industry leaders: Stop boasting about your safety records when you're literally putting out fires. You're only undermining your credibility.

In the real world, people are connecting the dots and the picture they have formed of the petroleum industry treatment of their workers and the environment isn't pretty.

It hasn't escaped the notice of the press that BP executives were on the Deepwater Horizon celebrating its excellent safety record, nor that many of the workers killed at BP Texas City had just finished a meeting touting their safety record. There's a message here.

Faced with the bashing this industry is getting daily in the news, with continual replays of graphic images of fires and explosions, you all need to understand: Continuing to tout favorable safety rates while workers are dying doesn't make you look like serious employers and doesn't make NPRA look like a serious organization.

Boasting about the great safety record of refinery industry while widows and children are planning funerals doesn't make you sound like a serious organization.

And giving awards to your members based solely on a lack of slips, trips and falls doesn't make you look like a serious organization.

If you want to be taken seriously, you need to act like an organization that recognizes there are serious problems and an organization that is seriously addressing those problems.

Playing the circular "blame game" with contractors isn't earning your industry any respect, either.

Dismissing a steady string of similar, deadly catastrophes as isolated, rare, or unpredictable incidents isn't credible -- not to OSHA, not to Congress, and not to the public.

When your company has become a nightly punching bag for late night talk shows gags, making your entire industry look like a joke -- you know you're in trouble. Ask Toyota.

Now I feel that I can be frank with you because OSHA isn't completely innocent either. Until recently, OSHA has based its refinery targeting system on DART rates as well. Clearly we need to change this. We need to find a better way to target problem refineries so that we aren't wasting our time or your time inspecting refineries that don't have major problems.

But to the extent we continue to use DART rates as one factor in our targeting mechanism, we need to make sure that they're accurate. That's why we're paying special attention to safety incentive and discipline programs that have been shown to discourage workers from reporting injuries and illnesses.

Here's another problem: Our targeting system was based only on the injury and illness rates of refinery employers; it did not include refinery contractors who may make up a large part of the workforce and employ workers who do some of the most dangerous work in the refineries. Unless we take the safety of the entire refinery workforce in account -- no matter who the employer is or what industry code they fall under -- we will miss many of the most important indicators.

And unless we are looking at the entire refinery workforce, we will not be able to accurately assess the safety of the refinery industry.

We want to work with you and other stakeholders like unions and experts to find a better way to target problem refineries for more attention. Identifying problems before they become tragedies is the right thing to do for refinery workers and refinery owners.

Similarly, be careful not to let risk assessments convince you that something can't happen. We've witnessed several recent incidents caused by events that companies believed were not "credible" -- the current oil spill in the Gulf of Mexico, for example.

Because the consequences of a single failure in your facilities can be catastrophic, management must establish systems that ensure that every process operates safely. Process Safety is all about solving systems problems.

And watch out for the small things -- the "tip of the iceberg" principle. The few problems you do see, particularly at higher management levels, are probably a fraction of the problems you don't see below the surface. Follow up on close calls and unusual circumstances; these can point to underlying problems that, if not addressed, could lead to tragedy.

And look at your number of close calls; we will. If your near-miss incidents do not significantly outnumber your actual incidents, your incident investigation program is probably ineffective.

Ask yourselves: Do your workers feel they can report close calls without repercussions? This is a tough one, but it's a real measure of the culture at your facilities.

Take a broader perspective when performing your risk assessments. Good numbers and seemingly long odds of experiencing injuries or fatalities lead to complacency and failing to adequately mitigate risks.

Monitor any complacency that you see, anywhere along your company's hierarchy. Beware the words "that'll never happen" or "the risk is too small to worry about." Believe me: When you hear these words, that's the time to worry.

Think about all the PHAs you've conducted at your facilities. Ask yourselves: Did your teams conducting these analyses look to identify, evaluate and control ALL hazards, or did they look over their shoulders -- maybe hedging on some of the difficult decisions because they believed that management would not support their conclusions? Are you confident in your answer to this question? -- because the answer to this question is one of the foundation blocks of a strong PSM program.

I challenge you to go back to your plant personnel, have a frank discussion about your expectations and ask this question. You might be surprised by what you hear.

Finally on this point, I'm not just speaking to BP or their Texas City Refinery. The Baker Panel made no findings about companies other than BP, but the Panel stated that it was "under no illusion that the deficiencies in process safety culture, management, or corporate oversight identified in the Panel's report are limited to BP."

The Baker report went on to say that when refining and chemical companies understand and apply these principals -- to their safety cultures, to process safety management systems and to corporate oversight mechanisms -- the safety of the world's refineries, chemical plants, and other process facilities will be improved and lives will be saved.

Working with Industry

Dr. Michaels recently met with the leadership of NPRA, the American Petroleum Institute and the United Steelworkers. They discussed OSHA's concerns with the poor process safety performance in the refining sector and initiated the process of addressing the mutual problems that I just discussed.

I'm pleased that NPRA is planning a workshop this summer to share data and ratchet up industry commitment to improve refinery safety.

We are looking forward to working with you, other industry associations, labor unions and experts in this field.

You can expect to see OSHA collaborating more with NIOSH, EPA and other agencies to address the worker health and safety problems in your industry -- and in other industries as well. The chemical industry has many of the same widespread problems that boil down to a few pervasive compliance violations across multiple facilities and across the industry.

Together, we can develop a more effective system for targeting problem hazards and problem worksites, and addressing the problems that we have identified.

Historic Turning Point

You can make this conference a historic turning point for your industry by resolving to practice prevention at every site, at every level. I urge you:

• Learn from your worksites' mistakes and others' mistakes
• Share best practices and success stories with each other, with NPRA, with OSHA and other agencies.
• Develop and track new metrics that give a truer picture of your problems and progress.
• Transform the safety and health culture of your worksites. Make PSM count.

We all know that the worst time to develop workable, effective solutions is in the midst of a crisis. Let's work together now to find workable solutions before any more people get hurt -- and let's not take another six months or a year to fix what's fixable now.

Thank you very much -- and good luck.