Los Alamos Workers Debate the Lab’s Safety, Morale, and Leadership

As the division leader for health, safety, and radiation protection at Los Alamos National Laboratory (LANL), I am well aware of public discussion about the laboratory's safety record and the reasonableness of last summer's decision to suspend activities here (see Brad Lee Holian's Opinion piece, Physics Today, December 2004, page 60).

Director G. Peter Nanos said that he suspended operations because he had little confidence that, as an institution, we had sufficiently identified and addressed our risks and potential vulnerabilities. Critics have argued that LANL's safety record was good enough, and they therefore questioned the logic underlying the director's actions.

In my opinion, LANL's safety record is not good enough. The laboratory collectively, and all employees individually, must redouble their efforts to embrace a safety mindset, reduce safety incidents, and strive for a best-in-class record that is immune to debate.

Like most statistics, those relating to safety can be presented in many ways to support just about any message, and a number of attendant complexities are difficult to completely analyze.

The Occupational Safety and Health Administration's standardized total recordable injury rate, the number of injuries per 100 person-years worked, establishes uniform categories of injuries that allow for comparison of safety rates of businesses that represent the same type of industry and organizational size.

In July 2004, the Department of Energy's average injury rate for its 27 research contractor organizations was 1.7, compared to LANL's rate of 1.9. While these data indicate that Los Alamos accident rates hover around the mean for DOE research contractors, it also indicates that we are far from best-in-class. Also, though LANL's injury rate improved dramatically between 1996 (6.0 injury rate) and 2001 (1.5 injury rate), over the past few years our rate of improvement has not just stagnated, but actually reversed.

As a nuclear laboratory, LANL bears an enormous public trust. Society tends to tolerate accidents resulting from familiar causes such as construction or driving; at the same time, society is intolerant of accidents at a place where the hazards are unfamiliar and potentially catastrophic. The public holds the laboratory to a very high standard of safety, and it's our job to meet that standard.

In scientific research, we content ourselves with nothing less than best-in-class. Why would we settle for anything less in safety when the stakes-the health and lives of our employees-matter even more?

In hindsight, the statistics paint a revealing picture about safety at Los Alamos. But in the midst of July's crises and turmoil, what drove Nanos's decision was a very real concern: his regard for each and every employee, and his knowledge of the human toll that any safety incident takes.

Since 1996, I have served as a safety consultant to Los Alamos National Laboratory. In that capacity, I have provided guidance to senior executive team members, two past laboratory directors, and the current director. I have also provided safety training for laboratory staff and management.

The laboratory exists within a complex nuclear industry with extraordinarily high consequences for error; as a result, the public demands nothing less than this institution's total commitment to achieving the best safety and operational records possible.

There is a simple rationale for Director Nanos's standard and expectation of excellence in all things: The higher the hazards and risks of an operation, the more important it is to develop the highest standards for operating procedures and performance, and to implement those standards consistently.

It's easy to become complacent about safety. "Experts" fall into the trap of expertise: As they become more familiar with safety hazards, they perceive the risks as being lower than they really are. Such underestimation leads, in turn, to a false sense of confidence and the gradual erosion of standards. The result? Safety loses its prominence and preeminence, and individuals fail to focus on the big-picture priority of safety.

The consequences of anything less than a full commitment to excellence in safety are clear and often tragic. Certainly, in a culture of excellence, there is no room for carelessness with, or willful disregard of, important standards and processes. But on a more fundamental level, data points on an injury and illness chart represent human beings and pain and suffering for them and their families.

Organizations that succeed in achieving greatness do not shy away from looking at the hard facts, but instead confront those facts and use them to drive continuous improvement. Here, in my view, are the hard facts about safety at Los Alamos:

• The laboratory's safety performance is good. Everyone working at the laboratory, and those who support it, should be proud of that record. However, 250-300 significant injuries still occur each year. That situation must improve, and I cannot imagine a reason for not working hard to make that improvement happen. Safety excellence requires it.
• For about four years, the laboratory's safety performance has been on a plateau. This stagnation is inconsistent with the continuing performance improvement achieved by both private industry and the Department of Energy during the same time period. One hallmark of an excellent organization is continuous improvement in results. Because results follow systems, the laboratory must improve its safety systems if it is to improve its safety results.
• There have been several recent serious incidents and injuries. The laboratory concluded that a pattern of near misses and security incidents over the past year has caused increasing concerns at DOE, the National Nuclear Security Administration, the University of California, and LANL. Some of those have had the potential to cause serious permanent injury or death. One significant root cause in a few of the incidents was failure to follow and implement accepted industry standards. A failure to implement management systems that ensured uniform high quality was another significant root cause for losing focus on safety.

Confronting these hard facts about safety is the first step in achieving the lab's goal of excellence in not just science, but also operations and safety. We have to face the present before we can look toward our future.

Excellence in science, operations, and safety is not exclusionary. Rather, the three areas of excellence are interdependent, and Los Alamos must pursue all of them to fulfill its national security mission. That task requires the wholehearted commitment, dedication, attention, and awareness of every single individual working at the laboratory. It demands a culture of excellence-the product of not one great decision, but a million correct decisions made every day. The stakes are too high to permit anything less.

Holian replies: Spokespeople from Los Alamos National Laboratory argue that in shutting down the lab, G. Peter Nanos showed that he really cares about the individuals who are injured or nearly killed, and their families, rather than only the statistics. But LANL scientists care far more-arguably more than management-about the human costs, because we are the troops in the trenches. Obviously, it is in our self-interest to strive continually for a safe work environment, an essential component of good science.

Do the safety data show that behavior at the laboratory is so bad that we scientists and workers deserve the public humiliation and opprobrium heaped upon us by our own director?

In my Opinion piece, I made sure that the accident rates I reported placed all the labs and industries on equal footing: The rate is the number of accidents requiring medical attention, for everyone at the site-including outside contractors and maintenance and construction workers- divided by 100 person-years, so as to normalize institutions for their size and work done. Averaged over a year, the rate is a rough measure of the percent likelihood that someone would have needed medical care for an injury.

I focused on the national labs that perform work similar to LANL's, and did not discuss the average over the entire Department of Energy complex, which would have also included offices that only process paper and places that have been totally shut down, apart from guards at the gates. If Los Alamos were in that category, it too might have a very low accident rate, but that would not be a very good outcome for national security.

With the ground rules outlined above, all DOE labs had comparable average total-site accident rates at the end of 2003-for the four major nuclear weapons labs (LANL, Lawrence Livermore, Sandia, and Oak Ridge), the rates were 1.9, 3.3, 3.2, and 2.3, respectively.¹ Following the successful implementation of a safety program at LANL, the trend in its yearly accident rates for the years 1997-2003 was downward: 5.6, 3.5, 2.6, 1.9, 1.8, 2.0, and 1.9. During this time, LANL consistently led the other three weapons labs (apart from the first year only, when rates for Sandia and Oak Ridge were better). This level of attention to safety can hardly be characterized objectively as "stagnation."

Were the lessons learned by managers during the shutdown so critical that the laboratory's scientific work could be suspended for three months, experimental work stopped for more than six months, customers disappointed, students discouraged from coming to LANL, and staff driven to contemplate leaving? From the taxpayers' perspective, the annual cost of doing business at LANL is more than $2 billion. While salaries were being paid, benefits were being given out, and retirement plans were proceeding as usual, scientists' livelihoods were put on hold. By livelihood, I mean the reason that scientists are eager to get up in the morning and go to work. The morale at Los Alamos has been thoroughly devastated by Nanos's unprecedented, unwarranted action.

Did the shutdown result in a dramatic drop in the labwide accident rate, as one might reasonably suspect? Surprisingly, the LANL rate went up dramatically in the first three months of the shutdown, from 2.0 for January- June 2004 to 2.5 for January- September 2004, although the rates for the four nuclear-weapons labs ended up closely comparable, nevertheless.^1,2 One likely contribution to the remarkable rise in the LANL rate was the intense stress from the rush to meet artificial deadlines during the early chaos of the shutdown. The director's threat to close the lab for any future safety or security infraction put a punishing psychological burden on the staff. His decision was a classic top-down fiat. As any safety expert knows, you improve safety by getting buy-in from the workers-by valuing them and the work they do-and by listening to them.