A few years ago, I learned a painful lesson about energy systems — not in a laboratory, but on a marathon course.

It happened at the Portsmouth Coastal Marathon. I had trained hard, pushed my pace, and was on track to run a personal best. Everything felt strong through most of the race: the miles were ticking by, my stride felt smooth, and my pace was faster than I’d ever sustained before.

Then, with only about 500 meters left to the finish line, my body simply stopped cooperating.

I hadn’t fuelled properly.

I had focused on speed and endurance during training, but I’d underestimated the importance of energy input during the race itself. My glycogen stores were depleted, my blood sugar crashed, and my legs turned to stone. The next thing I remember was waking up on the ground surrounded by medics after passing out just short of the finish.

Ironically, I had run the fastest marathon pace of my life — but I hadn’t supplied the system with enough fuel to sustain it.

That experience stuck with me, because it perfectly mirrors what I’ve seen time and time again in anaerobic digestion plants.

An anaerobic digester is, in many ways, like a marathon runner.

Microbial communities work continuously to convert organic matter into biogas. Like a marathon runner, they rely on a delicate balance of energy supply, nutrient availability, and system stability.

When everything is balanced — feedstock composition, trace elements, loading rate — the microbial community performs incredibly efficiently.

But when the “fueling strategy” is wrong, performance collapses.

Operators often focus on throughput or biogas yield in the same way runners obsess over pace. Yet the most important factor is whether the microbial system has the nutrients, conditions, and stability it needs to sustain that output.

Just like my body on the marathon course, digesters can run well for a long time even when something is missing — until suddenly they can’t.

In the anaerobic digestion world, collapse rarely happens gradually.

Instead, a plant might run seemingly well for weeks or months before problems suddenly appear: volatile fatty acids accumulate, methane production drops, and the microbial community becomes stressed or inhibited.

From the outside, it can look like the system failed unexpectedly.

In reality, the digester was under-fuelled or imbalanced for some time — the warning signs just weren’t visible early enough.

My marathon collapse was exactly the same phenomenon. The energy deficit had been building mile after mile before my body finally shut down.

The lesson from endurance sport — and from digestion systems — is that performance depends on understanding energy flows in real time.

As operators, engineers, and scientists, we often rely on periodic lab tests or retrospective analysis. But microbial ecosystems operate continuously and dynamically. By the time problems appear in conventional monitoring data, the system may already be heading toward instability.

That realization was one of the sparks behind founding BiofuelAI.

In marathon training, athletes use data to optimize performance: heart rate, pace, carbohydrate intake, recovery times. These metrics allow runners to predict fatigue before it happens and adjust strategy accordingly.

BiofuelAI applies the same philosophy to anaerobic digestion.

By using machine learning and advanced analytics, we can interpret plant data in ways that reveal early signals of microbial stress, nutrient limitation, or process imbalance — long before they result in lost gas production.

In other words, we help operators fuel the microbial “athletes” in their digesters properly.

In the end, I never did finish that marathon. Instead, I spent the next few hours in the medical tent recovering, followed by several weeks undergoing medical tests to confirm it was safe for me to run again.

It was a frustrating experience, but an unforgettable one. The lesson was simple: performance systems — whether human or microbial — only work when energy supply is managed correctly. Ignore that, and even the strongest system can collapse just before the finish line.

Alan Beesley
CEO of BiofuelAI