Drax wants to feed Britain's AI habit with £20B datacenter at biomass power station

Drax, Brookfield and its UK AI infrastructure business Radiant proposed a datacenter development at Drax Power Station, with a £20 billion headline potential and a planned £2.5 billion first phase delivering up to 100MW by early 2028. Later phases could add 500MW between 2028 and 2031, with the site eventually exceeding 1GW, but the project still needs a tenant, planning and network approvals, commercial agreements and final investment decisions. Drax also wants to supply up to 500MW from its existing generation, subject to government and regulatory approval; community concerns include electricity bills, water use and local impacts.

Analysis

The marketable asset is scarce, permitted power-and-grid access—not the headline development value. But reported grid capacity is not equivalent to firm, deliverable capacity for a tenant, and the economics to Drax depend on the eventual lease, power-price formula, connection costs and allocation of construction risk. Until those are disclosed, the project is better viewed as a long-dated option than as contracted earnings. Applying data-center or AI-infrastructure multiples to Drax now would risk capitalizing an unfinanced, unpermitted ambition.

For DRX, the immediate catalyst is narrative support; over the next 1–3 months, evidence of a tenant, planning progress and regulatory treatment of the proposed power diversion matters more than further political endorsements. Over 6–18 months, a signed offtake and final investment decision could validate a new use for the site. Conversely, consumer-cost scrutiny, sustainability objections to biomass, or failure to secure approvals could strand the option and intensify pressure on the existing generation model. A long-term PPA could also mute Drax’s exposure to any wholesale-price upside from added data-center demand.

BN’s exposure is not quantifiable from the announcement: a JV involving Brookfield does not establish the amount, timing or returns of capital attributable to Brookfield Corporation. For the wider UK market, successful projects with genuinely usable connections could strengthen the value of grid-ready sites and pressure developers without them; do not assume this project establishes a replicable template.

Key Decisions for Investors

• DRX: do not chase the announcement as a near-term earnings upgrade. Treat any price strength as optionality until a creditworthy tenant, commercial power terms and a final investment decision are disclosed.
• Set a catalyst watch for planning permission, network approvals and consent to redirect generation under the existing support arrangements. A signed tenant contract plus FID would strengthen the case for a measured long; loss or material delay of approvals would falsify it.
• BN: avoid attributing the project’s headline value to Brookfield Corporation. Reassess only when the JV’s capital commitment, funding structure and expected returns are disclosed.
• Monitor UK data-center connection approvals, power-contract terms and local opposition as read-throughs for grid-ready infrastructure. A sustained rise in DRX without progress on tenancy or approvals would make the risk/reward less attractive.