The CRU’s new decision on data centre connections is notable because it lays bare a deeper problem in our climate governance.

To its credit, the decision explicitly acknowledges that the growth of data centres poses a material risk to Ireland’s ability to meet its carbon budgets. It also openly states that the CRU believes there is a regulatory gap: the Climate Act, in its view, does not give it sufficient authority to impose greenhouse gas limits through electricity connection policy. That admission alone should give policymakers pause.

The CRU does, however, consider that it has the legal remit to require renewables in line with government policy. Its solution is to require large data centres to meet 80 per cent of their annual electricity demand with renewable electricity generated in Ireland — eventually. Crucially, this is not an emissions limit. Renewable energy development is not the same thing as emissions reduction, particularly in a system where electricity demand is rising rapidly and fossil generation still fills the gaps.

The physical and legal basis of our climate targets is to meet binding cumulative emissions limits, which we are already on track to exceed significantly. Higher overshoot in earlier periods is carried over to later periods - it doesn’t vanish. The CRU decision is likely to significantly exacerbate this overshoot, and concerningly, this risk is not quantified.

The decision requires new data centres to provide on-site or proximate dispatchable generation or storage capacity to ensure they can meet demand during periods of system stress. In theory this could include zero-carbon storage, but in practice the most readily available and scalable option is on-site fossil-fuel generation, particularly natural gas. This requirement shifts risk away from the electricity system and onto individual sites, but it also locks in new fossil infrastructure at precisely the moment Ireland is meant to be rapidly reducing its reliance on gas.

The decision allows a six year “glide path” from the moment a data centre is energised before this 80 per cent renewable requirement must be met. There is no requirement for renewables at the outset, no interim milestones, and no accounting for the emissions that arise during this period. Those emissions are real, cumulative, and irreversible. Later renewables do not undo them. Under this policy, data centres under development now could be operating largely on gas-fired power generation well into the 2030s.

Even after the glide path, the policy relies on annual energy accounting, not physical and temporal matching of renewable generation and demand. Wind and solar do not match data centre demand hour by hour. When renewables are unavailable, fossil generation fills the gap. The decision does not require data centres to invest in the grid infrastrucutre and storage to ensure that renewables, even when they are abundant, actually reaches customers. The grid is already struggling to absorb a growing share of renewable electricity because of network and system constraints, and limited storage and interconnection. In practice, that means a data centre could be powered by fossil fuels for significantly more than 20 per cent of the time, even if it meets the 80 per cent annual renewable requirement on paper.

Additionality is also a concern. Even though there is a requirement that these renewables must be procured outside of auctions, data centres could end up mopping up significant shares of the limited pool of viable renewable projects that would otherwise have displaced fossil fuels outside auctions.

There are system-level consequences too. By encouraging large corporate buyers with deep pockets to secure renewables through private contracts, the policy risks undermining the competitiveness of public renewable auctions, potentially pushing up costs for everyone else.

Concerningly, the CRU’s policy has not quantified these risks for emissions or energy prices.

I repeat again - more renewables is not the point of our Climate Act - they are just a means to cut cumulative emissions. The CRU has interpreted that it has statutory power only in relation to the promotion of renewable electricity (Electricity Regulation Act 1999), and not in relation to regulating GHG emissions. However, that does not mean that renewables obligations can be interpreted without refrence to the policy framework they are meant to serve. Under the Climate Action Plan, the target for 80 per cent renewable electricity by 2030 is not a free-standing objective, and is very clearly a means to replacing existing fossil fuel use in existing electricity generation, heat, transport and industry, and to enable zero-carbon demand growth, not unconstrained demand growth, or to facilitate growth met with fossil fuels. The policy decision itself makes reference to the risk of fossil gas lock-in, the potential requirement for new gas infrastructure, and the risk of stranded assets.

The CRU could have interpreted it existing powers in this context and added conditions that minimised fossil fuels, including requiring earlier renewable delivery, time-matching of renewables, and requiring evidence that renewable procurement would not materially divert resources from existing decarbonisation pathways. It could also have required, at a minimum, an 80 per cent renewable target to be met by 2030, rather than 6 years following energisation of the data centre.

Stepping back, the most important takeaway is institutional. The CRU’s decision shows how essential Section 15 of the Climate Act is to ensuring that the statutory remits of public bodies are in line with essential decarbonisation. If the Climate Act cannot meaningfully constrain a significant source of fossil fuel demand, then we need to urgently ask how to address this.

This is not unique to energy regulation — similar tensions exist across agriculture, transport, enterprise and land use policy. Unless Section 15 is enforcable, the climate imperative will lose when it confronts the imperative of economic growth.