The Wrong Question

Who would drain hot naphtha to the ground?

That was my first question when the news broke: Two operators, brothers, died while draining hot naphtha directly to the ground at the BP-Husky refinery near Toledo, Ohio, on September 20, 2022.

It was the whole industry’s first question. By the time the investigation finished, nobody who read the record was asking what those brothers did wrong.

Two years later, two workers died at the PEMEX refinery in Deer Park, Texas, when they opened the wrong line. Same question: how could somebody open that line without checking? The same reversal after the incident details were made public.

In my four decades in this business, I have never heard the tone in the industry so raw. The professionals I respect most are asking what it will take to get people to pay attention. I have asked the same question myself.

It is the wrong question.

The Better Questions

The better question is what in the system made the action possible. Neither tragedy had a single cause. Strengthening training and competency is one part of a much larger system, and I believe it is among the most important. That is the part I want to examine here: how the industry builds, verifies, and sustains the judgment of the people who touch process equipment.

When similar failures recur across companies and regions, we have to look beyond the people involved and examine the systems that prepared them for the work.

Over the past few decades this industry has invested heavily in hardware and management systems. Better metallurgy, safety instrumented systems, updated codes and standards. That investment worked. The layer we never fully formalized was judgment. A generation of operators, supervisors, and engineers carried it and passed it on informally. The people were the redundancy.

Many of the operators who built the postwar equipment and systems came to the gate from the military. Procedure-following and verification were formed into them before we hired them. The next generation absorbed that culture by standing next to it for twenty years. We told ourselves we had a training system. What we had was an inheritance.

The veterans retired, and the generation they formed retired with them. Much of the apprenticeship went with them. Our current workforce is different, and the systems these workers have been given have aged and grown stale.

Staff cuts and organizational changes reduced in-person training personnel and budgets and replaced that instruction with computer modules.

Meanwhile, the procedures swelled to a hundred pages. After incident investigations, too often we took the same actions: revise the procedure, train on the revision, get the signature. The signature sat in a file, waiting to assign fault after the next event. Training became part of the system for distributing blame, and then we wondered why nobody learned from it.

The current workforce is not the problem. These are intelligent and capable learners. They arrive at work ready to learn, and we sit them in front of a learning management system and call it qualification.

The risk never went away. It migrated to the layer we never resourced: the people executing the work, on shift, under pressure.

Asking what it will take to get people to pay attention is asking the new generation to supply the judgment we stopped transferring. They cannot. Nobody could.

I know something about knowledge walking out the gate. In April, I was part of it.

I spent the last seven years as a corporate process safety manager. Alongside that role I was an active member and part of the leadership of the AFPM Human and Organizational Performance (HOP) subgroup. One project stays with me. A team of industry HOP experts from the leading oil companies rebuilt James Reason’s culpability model into a practical job aid that moves an organization from “who did this” to “what produced this.” I had the privilege to lead that work.

The human error job aid resonated with everyone it reached. I want the thinking to reach more of the front line. I am building my own version of James Reason’s published model, and in a coming edition I will put it in your hands.

That model is one tool. Training and competency are one part of the larger process safety system. The book, Process Safety Excellence: A Practical Reference for Operating Companies, comes out this fall and contains more than 300 actionable items organized in over 30 chapters covering risk-based process safety, technical safety, and safety leadership. It translates complex technical and regulatory process safety management topics into a format usable by the frontline workforce. Nearly every chapter pairs the material with an incident we have already paid for and puts it in plain language for the leader on shift. If you lead people who touch process equipment, I wrote it for you.

Every two weeks, this newsletter will bring you one idea for the people who run the unit and everyone whose decisions reach them.

Separate from this newsletter, I will send direct email updates on the book as it moves to launch. If you want those, go to my website at hse2.com/subscribe. You will hear the release date there first, and any reply comes straight to me.

I spent four decades learning this work, some of it the hard way. The people on shift tonight should not have to learn it that way. See you in two weeks.