Drax just reported record renewable generation — and is now exploring data centre colocation on the same site where it’s trying to build one of the world’s largest BECCS installations.

Sit with that combination for a second.

Drax is the UK’s largest renewable generator by output. Its North Yorkshire plant burns biomass to generate electricity, captures the CO₂ from combustion, and plans to store it permanently — bioenergy with carbon capture and storage. If BECCS works as advertised, every unit of electricity generated is carbon-negative: you pulled CO₂ from the atmosphere (as trees grow), burned it for power, and buried the emissions. Net negative.

Now they want to power data centres from the same site.

Data centres are enormous power consumers. A hyperscale facility can run 100-500 megawatts continuously. Colocation with a power generator eliminates transmission losses, can enable direct power purchase agreements, and gives a tech company a clean story about where their electrons come from. For Drax, it’s a revenue diversification play — BECCS development is expensive and years from first commercial carbon credits, so anchoring a data centre customer provides near-term cash flow and load factor.

The interesting question is whether this becomes a template. Carbon-negative power + compute-intensive industrial load, co-sited, integrated economically. There are other potential versions: a DAC facility collocated with a nuclear plant, or enhanced geothermal powering an industrial cluster that includes CO₂ removal. The logic is the same — match high-reliability, low-carbon power generation with loads that need reliability and want the clean credential.

One caveat that can’t be avoided with Drax: the biomass sustainability questions are real and unresolved. Their sourcing practices — particularly around what qualifies as “waste wood” — have been contested by environmental groups for years. BECCS’ carbon-negativity depends entirely on the lifecycle accounting of the biomass supply chain. If the wood sourcing is questionable, the carbon math is questionable.

That doesn’t make the colocation model uninteresting. It does mean the model only works if the underlying CDR claim holds up. Watch both.