With the 2026 harvest drawing to a close, anaerobic digestion (AD) operators are now beginning to see what this year’s feedstock availability and quality look like.

The Met Office has provisionally confirmed summer 2026 as the UK’s hottest on record, with England recording its driest July. The prolonged dry weather has resulted in disappointing and highly variable crop yields, according to the Agriculture and Horticulture Development Board (AHDB).

By late August, wheat yields were around 12% below the five-year average, spring barley was 19% below average, and oats were 17% below average. Winter barley performed better, while oilseed rape was a notable exception, recording yields around 19% above its five-year average.

These figures may not necessarily translate to energy crops AD plants usually rely on. Crops grown specifically for AD, such as maize, have different growing and harvesting cycles.

However, the wider agricultural picture provides a clear warning: hotter, drier and more variable growing conditions can affect both the quantity and characteristics of the biomass available to AD operators.

Agricultural AD plants convert crops, crop residues, manures and other organic materials into biogas through a biological digestion process. The methane in that biogas can be used for electricity and heat or upgraded into biomethane for injection into the gas network.

Today, biogas represents only around 1% of Great Britain’s gas consumption, compared with more than 40% in Denmark in 2025. But its role could increase substantially. NESO’s Future Energy Scenarios model a biomethane supply of up to 36 TWh by 2035 under its Holistic Transition pathway.

Increasing renewable gas production, however, does not have to mean simply building more digesters or consuming more feedstock.

There is another, often overlooked opportunity:

Extract more methane from every tonne of organic material already available.

This becomes particularly valuable following a difficult harvest.

AD operators have traditionally paid close attention to tonnes: tonnes harvested, tonnes stored and tonnes fed into the digester.

But the real value lies in the recoverable energy within those tonnes.

Feedstocks differ in dry matter, organic matter, fibre structure, degradability and methane potential. Growing and harvesting conditions can also change these characteristics.

Consequently, two tonnes of apparently similar material can produce very different amounts of methane.

When feedstock is plentiful and relatively inexpensive, some of this potential can be overlooked. When availability tightens or quality becomes more variable, improving conversion efficiency becomes increasingly valuable. There are two established ways to respond, and they work on different parts of the problem.

One lever is to physically treat material before it enters the digester. A good example of this is the SMASH pretreatment technology developed by ALPS Ecoscience. This increases the bioavailability of feedstock, which means that the digester’s microbial community has access to a greater proportion of the organic material. In a ring-fenced commercial trial, this approach delivered a 19% increase in biomethane production. The achievable gain depends on the feedstock itself and the economics of treatment in the context of the plant.

The other lever is to make better decisions about how feedstock is fed and managed.

AD plants already generate the data needed for this, from feedstock composition and loading rates to gas production, methane concentration, temperature, pH and ammonia. Operators use it to understand what’s already happened, and draw on experience to decide what to try next.

BiofuelAi’s platform shifts that process from hindsight to foresight. Using physics-informed, uncertainty-aware AI models, it enables operators to better understand the relationship between feedstock, biology and plant performance, and predict how a feeding or operating decision is likely to play out before it’s made.

Instead of simply asking “What happened yesterday?”, digital optimisation allows operators to ask:

• What happens if I change tomorrow’s feedstock mix?
• Can I maintain gas production using less or different material?
• How much additional methane should I expect from a change in feeding strategy?

And, crucially:

Which operating decision offers the best economic return without compromising biological stability?

The goal is better conversion of available feedstock into valuable renewable energy, not AI for its own sake.

The 2026 harvest highlights a wider challenge for the AD industry.

If agricultural production becomes more variable, operators cannot assume that tomorrow’s feedstocks will behave like yesterday’s. Simply purchasing or growing additional tonnes may become increasingly expensive, and in some cases may just not be possible.

The alternative is to become better at extracting value from what is available.

Physical optimisation technologies such as SMASH can increase the accessibility of organic material. Digital optimisation technologies such as BiofuelAi can help determine how that material should be combined and fed, while predicting the biological and economic consequences of different decisions.

Together, these technologies point towards a different way of thinking about AD plant performance, moving from how many tonnes can we process, to how much renewable energy can we reliably and profitably extract from every tonne available.

At BiofuelAi, we believe the industry’s next gains will come as much from decisions as from tonnes, from getting more out of the feedstock already on site and not just sourcing more of it. That’s the problem we’re focused on solving, and a harvest like this one is a reminder of why it matters.

Alan Beesley is CEO and co-founder of BiofuelAi. He combines commercial leadership with hands-on experience in anaerobic digestion and biological process optimisation. His focus is turning complex modelling and AI into practical tools that help operators make better decisions and deliver stronger performance from real-world assets.