My experience as a process engineer/production engineer on the plant floor, close to operations, is not only dynamic but also incredibly valuable. It shapes you, makes you stronger, and sometimes even changes you as a person. You grow not only as an engineer but also as an individual.
And trust me, when something happens, everyone has a different version of the story.
Once, I had to conduct a Root Cause Analysis (RCA) for a tank that had overflowed through the pressure relief valves. Not exactly the kind of fountain you'd want to show visitors.
The first operator said:
“Yeah, I saw the pressure rising in the tank. To relieve the pressure, I climbed and gave the pressure relief valve a few taps.”
I could picture it immediately. A high-tech process installation, millions invested, advanced control systems… and in the end, someone solves the problem using the universal engineering technique we all know: giving it a good tap.
The first operator said:
“Volgens mij was de bypassslang van het reduceerventiel vol met vloeistof.”
Weer een andere collega zei:
“Yeah, I saw the pressure rising in the tank.”
And there I was, with three explanations that all sounded reasonable, yet none of them told the same story...
The beauty of assumptions
One thing I've learned over the years is that people are not lying. They are sharing what they saw, heard, felt, or believed.
- The operator in the control room saw the pressure increasing.
- The operator walking outside noticed liquid where it shouldn't have been.
- The mechanically skilled operator grabbed his tools and rushed to the tank to provide support.
Everyone was looking at the same incident, but through a different window. And that is exactly why incident investigations are so fascinating. As an engineer, it's tempting to hear the first explanation and think:
“Aha! Found it. Root cause identified.”
But in reality, that's usually where the investigation begins.
Detective of the Plant
Sometimes my job feels like a detective story. You start with a process that you know was stable.
- What happened at 09:15?
- Which alarms were triggered?
- Which valves were open?
- Which process parameters changed?
- Who was where?
- What actions were taken?
Meanwhile, you're gathering process data, trend charts, logbooks, and interview notes. More often than not, you discover that the incident wasn't caused by one major mistake. It was instead the result of a bunch of little things happening at the wrong time. Just like a row of dominoes falling one after the other.
The lesson I keep learning
What's the biggest lesson from investigations like these?
Listen first. Judge later.
- The operator who says he tapped the valve is providing you information.
- The operator who thought the tank was full is providing you information.
- The operator who noticed something unusual in the bypass hose is providing you information.
None of them needs to know the root cause to hold an important piece of the puzzle. In fact, the truth is often hidden somewhere between all the different stories.
Do not let other people make you doubt yourself; trust your own judgment.
- Process data tells you what happened.
- Operators help you understand why people did what they did.
For a solid RCA, you need both. Ultimately, we don't solve problems by looking at graphs alone. We solve them by understanding both processes and people. And sometimes, it all starts with an overflowing tank, a pressure relief valve that received a few strategic taps, and a group of colleagues who each have their version of the story. And honestly? That's what makes this work so interesting and different every single time.
The main message
In incident investigations, the first explanation is rarely the complete story. Gather facts, listen to everyone involved, and build an objective timeline. The best engineers are not the ones who jump to conclusions the fastest. They are the ones who remain curious the longest.
What's your key lesson?