Green Party wants to slam the brakes on UK datacenter construction until water and energy use are sorted

The Green Party of England and Wales is pressing the government to revoke the 2024 critical national infrastructure (CNI) blanket designation for new “data barns,” arguing it grants priority for water and power over households under drought restrictions. The group cites water use of ~1.5 million liters/day for a large-scale datacenter (vs. ~10,000-person town needs) and warns growth is concentrated in already water-stressed regions like London and the South East, potentially raising energy-bill pressures. The government counters that pausing datacenter builds would move investment/jobs abroad and harm national security, while urging responsible water use and clean power rather than a “blank cheque” for Big Tech.

Analysis

This is less a demand shock than a permitting-tax on future supply. The first-order winner is any incumbent with already-energized, water-secure land banks and operational campuses; the real margin expansion comes from scarcer approvals lifting lease pricing and reducing competitive intensity in the UK pipeline. By contrast, greenfield developers and landowners banking on rapid “power-first” site conversion face a longer cash-conversion cycle and a higher discount rate on planned capacity.

The second-order effect is geographic leakage: if the UK becomes a slower place to add AI capacity, hyperscalers and colo customers will route incremental workloads to Ireland, the Nordics, Spain, and US regions with easier grid access. That helps non-UK data-center platforms and utilities in power-abundant jurisdictions, while pulling future ancillary spend—construction, fiber, switchgear, cooling equipment—out of the UK. It also makes grid-constrained London assets more valuable because the bottleneck is not compute demand, it is entitlement to connect.

The contrarian point is that the political attack is directionally right but operationally incomplete: newer direct-to-chip cooling materially reduces water intensity, so the “water-guzzling bit barn” narrative may overstate the incremental constraint on next-gen facilities. If policymakers end up distinguishing between legacy and high-efficiency designs, the actual impact could narrow to a slower approval process rather than a ban. That would hurt sentiment for days/weeks, but the structural effect over 6-18 months is mostly a higher hurdle rate for UK AI infrastructure, not a collapse in demand.

Key Decisions for Investors

• Long UK incumbent infrastructure owners with existing power and planning rights vs short greenfield data-center developers: favor assets already connected to the grid and de-risked on water. Highest payoff over 6-18 months if planning friction persists; invalidate if the government carves out fast-track approvals for high-efficiency campuses.
• Buy call spreads on European colo/infra names with diversified non-UK exposure; this is a relative-value hedge against UK-specific permitting tightening, with the spread working over 1-3 months as capital is reallocated to friendlier jurisdictions.
• Watch-list, not a trade yet: long utility/grid-equipment beneficiaries in regions that can absorb displaced demand (Nordics/Ireland/Spain). Only act if there is evidence of UK project deferrals in earnings commentary or backlog conversion metrics.
• Avoid chasing UK REITs tied to speculative data-campus land options until there is clarity on water and grid criteria; the downside is multiple compression if approval timelines stretch beyond one planning cycle.