Summary
- Ireland’s CRU policy requires new data centre connections to address generation, storage, renewable procurement, and locational constraints.
- Data centres accounted for 22% of Ireland’s electricity demand in 2024, according to the regulator.
- The framework makes power delivery central to whether Irish data centre expansion can proceed.
The Commission for Regulation of Utilities has set a more demanding connection framework for data centres in Ireland, as large digital loads continue to reshape the country’s electricity system.
The regulator’s Large Energy Users Connection Policy, published in December 2025, creates a pathway for data centre connections while taking account of grid constraints, security of supply, generation adequacy, network delivery, and national renewable energy targets.
The policy has been described during the current news cycle as a “bring your own power” model. The phrase is simplified, but the direction is clear. New data centres connecting to the electricity network will continue to be required to provide generation or storage capacity, either on site or locally, to match their requested maximum import demand.
CRU figures show how quickly the issue has moved from sector debate into national infrastructure policy. Data centres accounted for 22% of Ireland’s electricity demand in 2024, up from 5% in 2015. EirGrid has forecast that, based on currently contracted demand, data centre electricity consumption could rise from 9.4TWh in 2025 to 14.6TWh in 2034.
Connections now depend on energy strategy
Ireland’s data centre market was built on cloud investment, fibre connectivity, international business links, and a deep technology base. The constraint is now the electricity system. Dublin and its surrounding region have already experienced pressure around connection availability, while national adequacy and emissions targets have made large-load approvals more politically exposed.
The CRU framework changes the practical development test. A new site must show more than the ability to buy electricity. It needs a workable position on generation, storage, renewables, location, and participation in the wider electricity market. That moves energy strategy from a supporting workstream into the centre of the project.
Annual renewable procurement will not be enough by itself. A data centre can be matched with clean electricity on paper while still requiring firm grid support during periods of low renewable output or local congestion. The Irish approach tries to close that gap by linking connection policy with dispatchable support, system adequacy, and the pace of network delivery.
The added requirements will favour developers with capital, power-market expertise, and strong utility relationships. Larger hyperscale and colocation platforms are better placed to underwrite generation, storage, offtake agreements, and compliance reporting than smaller speculative projects. Even for them, the commercial model becomes more complex.
A clearer but harder route
The CRU has avoided a simple moratorium. Instead, it has set out a route that keeps data centre development possible while requiring projects to solve more of the energy problem themselves.
That clarity may help serious projects move forward, but the delivery burden is heavier. On-site or proximate generation requires land, permits, fuel or renewable inputs, grid arrangements, market participation, and operational coordination. Storage can support flexibility and adequacy, but it adds cost and technical complexity. Matching data hall ramp-up with energy-asset delivery will be difficult across phased campuses.
The policy also creates a useful comparator for other European markets. The UK, the Netherlands, Germany, Spain, and the Nordics are all dealing with rising data centre demand, constrained grid capacity, and public concern over large electricity users. Ireland’s approach puts the burden on developers to demonstrate system compatibility rather than leaving the issue to grid operators alone.
Irish data centre growth is therefore no longer only a question of land, customers, and construction. The first commercial conversation now sits much closer to power procurement, generation development, storage strategy, and grid adequacy.
The next test will come through real connection applications. Projects that can combine a strong site with credible local generation, storage, renewable procurement, and transparent reporting will have a route forward. Projects that depend on constrained grid capacity without bringing new system support will face a much narrower path.
