By Jan Willem Flamma – Training Director
Alright well engineers, let's get real about risk. We spend our days immersed in data, probabilities, and statistical models. We calculate average failure rates for tools, estimate average NPT for operations, and project average production profiles. It's how we design, how we plan, and how we budget. And for the most part, it works.
But what if I told you that sometimes, relying only on these averages can be dangerously misleading, especially when we're talking about the truly critical stuff? I'm talking about those low-probability, high-impact events that can sink a well, a project, or even a career. We're about to dive into something called ergodicity, and while it might sound like a physics lecture, it’s actually a fundamental concept that directly impacts how you assess risk.
Ergodicity: The two ways we look at the world
Let's simplify. Imagine you're trying to figure out the chance of a critical well component failing.
1. The “ensemble” view (the averages we use)
This is our comfort zone. You pull up industry data, look at hundreds or thousands of similar components installed in different wells, and calculate an overall average failure rate – “Across all our wells, this packer has a 0.5% failure rate”. This view is fantastic for:
- Industry benchmarking.
- Portfolio-level risk assessment.
- Setting general design guidelines.
It tells you what happens on average across a large group of wells at a specific point in time.
2. The “time” view (the reality of your well)
Now, shift your focus to the single well you're working on right now – your baby. This well is on a unique journey. It’s encountering specific geological conditions, dealing with unique operational challenges, and experiencing a one-of-a-kind sequence of events. The risk of that component failing in your specific well, at this very moment, isn't just the overall industry average. It's influenced by your well's unique history, the specific personnel on site, the quality of that particular component, and even the weather. It's about what happens to a single well over its lifetime.
Here's the core idea of ergodicity:
- If the average behaviour across many wells (the “ensemble” view) is the same as the average behaviour of one single well over a long period (the “time” view for that one well), then the system is ergodic. In this case, looking at averages works just fine.
- But if these two views are different, the system is non-ergodic. And that's where it gets interesting – and critical – for us.
Why non-ergodicity is your problem
Most of the truly catastrophic risks we manage in well engineering are, in fact, non-ergodic. Why? Because when things go wrong in a big way on a well, the consequences are often irreversible and path-dependent. You can't just hit a reset button and try again for that specific well.
Think about it:
- Irreversible events & “ruin”
- The average view: “Our company's average well control incident rate is X per 100 wells.” Sounds manageable.
- The non-ergodic reality: If your well experiences a catastrophic blow-out, it's not just a statistical blip. That well is often permanently damaged or lost. There's no “averaging out” that failure for your project, your budget, and your team's performance review. The impact on you and your project is 100% loss for that specific well. You can't drill another 99 wells to make up for the one you just lost.
- Path dependency & compounding failures
- The average view: “The average risk of stuck pipe is Y%.”
- The non-ergodic reality: That small fluid loss you had last week, combined with that slight hole cleaning issue you've been battling, now puts you at a much higher risk of differential sticking today. The outcome for your well isn't a random roll of the dice; it's a sequence of events. One small issue can snowball into a massive, irreversible problem. The “average” risk doesn't account for the unique, cumulative stress your well has already endured. Your past decisions on this well directly impact its future.
- Survival bias & statistical blind spots
- The average view: When we look at “average well performance”, we're lumping together wildly successful wells with catastrophic failures.
- The non-ergodic reality: If a particular design choice, on average, works for 99 wells but carries a small, non-recoverable 1% risk of total well loss for your specific well, that average isn't good enough. You are designing for the survival of your individual well, not for the average success of a theoretical population. You can't just walk away from your well if it fails catastrophically.
What does this mean for you?
Understanding non-ergodicity isn't about ditching statistics; it's about using them smarter and complementing them with a deep, personal understanding of your well's unique journey. It’s about remembering that for your well, you have skin in the game.
- Focus on consequences, not just probabilities: Instead of just “what's the average chance of this happening?”, ask “if this happens to our well, what are the irreversible consequences, and what’s our plan B (and C, and D)?” For those ruinous, non-ergodic risks, even a tiny probability demands extreme mitigation.
- Embrace dynamic risk assessment: Your well's risk profile isn't static. It changes with every foot drilled, every decision made, every piece of data. Constantly re-evaluate based on the current state and history of your well, not just generic averages.
- Prioritise robust contingencies: Because you can't “average out” a catastrophe on your single well, spend more time on robust contingency planning for high-impact, low-probability events. These are often the non-ergodic risks that keep us up at night.
- Learn from every incident (and near miss): Every little anomaly, every near miss on your rig, is a direct data point for your well's unique journey. Don't just tick a box; deeply understand why it happened and how it impacts this well, for the rest of this well's life.
- Value experience and intuition: Those seasoned veterans on the rig or in the office often have an almost uncanny feel for non-ergodic risks. Their “gut feelings” are often based on a deep, time-series understanding of how things can unravel. Listen to them.
The bottom line: Your well, your responsibility
Averages are crucial for strategic planning and industry trends. But when you’re the engineer making decisions for your well, remember that its journey is unique. Understanding non-ergodic risks – those irreversible, path-dependent events – is key to making better, more proactive, and ultimately safer decisions at the wellsite. It's about taking true ownership, because when it comes to your well.
