Dr Helen Edmonds was a newly minted graduate when she found herself in unfamiliar territory performing quantitative risk analysis for offshore oil and gas platforms in the North Sea.
As told to Joe Ennis
I was working in Aberdeen about 10 years after the Piper Alpha disaster that killed 167 workers. The regulations for offshore rigs had shifted dramatically. Every platform operator had to submit a safety case proving that individual risk was below the risk of fatality of a deep-sea fisherman. I landed a job with DNV (Det Norske Veritas) doing quantitative risk analysis on some of the biggest offshore projects.
My job was to model all the hazards: from helicopter travel to dropped objects to the consequences of a gas leak igniting.
I first trained as a civil engineer back in the 90s. During my degree I did some work experience in water engineering and hydraulic modelling, so when I found it hard to get work on graduation, I decided to go back to study for a PhD in fluid mechanics. This made me desirable to DNV who liked to hire “smart, numerate people” and skill them up on the job.
I had confidence in my abilities and assumed that this would neatly translate into any engineering field. But suddenly I was assessing fatality risk to the people working on these rigs. The stakes were high and I found that the translation of skills wasn’t as smooth as I expected.
While I could grasp the physical principles behind things like gas dispersion and ignition, I didn’t know enough to tell if my modelling outputs were sensible.
It all came to a head when I was working on a risk assessment for the expansion of a Conoco gas complex. I made some basic but critical mistakes, such as inputting data in the wrong data units.
The models ran, but the outputs were clearly way off.
The numbers didn’t make sense, deadlines slipped and tempers frayed. With these projects, every day delayed came at a huge financial cost. I doubled down, looking to myself to try to solve the problem, but eventually it became obvious I needed help – so I approached the project manager. He was understandably frustrated at first, but he then helped me trace the problem back to incorrect inputs.
That experience taught me a foundational lesson: start by questioning your inputs, not your outputs. Rubbish in, rubbish out. You can’t just trust a number. You need to understand where it came from and what it means. And making sure you use the correct units might help too.
It’s a lesson I still carry with me. Now, leading a wastewater utility in local government, I often ask my team: “What is that assumption based on? Where does the underlying data come from?” And if they can’t tell me, we dig deeper.
I’ve also learned the value of building a network. I’m naturally introverted, but over time I’ve developed a circle of peers I can call on when I need a second opinion. We’re part of a profession. But more than that, it’s a community of knowledge. As an engineer, you’re never as alone as you think.
Key lessons
- Always check your inputs. Get them right before you model anything.
- Interrogate assumptions. Don’t rely on instinct alone.
- Lean on your network. Engineering is a team sport.
Dr Helen Edmonds FIEAust CPEng EngExec is Head of Wastewater at Mount Barker District Council in South Australia, and a member of the Engineers Australia College of Leadership and Management.
This story was originally published in the August 2025 edition of create.
Elevate your career and skills by becoming Chartered. Learn how to prepare for and begin the application process.
Excellent article Helen,
As a civil engineer with over 50 years of experience I am extremely concerned at how computer modelling is being used to predict flooding and inundation. I have seen examples of predicted land flooding that are physically impossible but could not get any sense from local authorities as they accepted the modelling without question. Modelling used data from satellite imagery and there no statement of assumptions yet rules were being imposed that have an affect on property developments which had economic impact on projects and property values.