This is part of the paid GridteamAI Quarterly Sector Report.
£295 single volume · £995/yr annual · Team £2,495 · Enterprise £4,995 · Corporate £9,995. Secure checkout coming soon; email to purchase today.
A battery boom, one old reservoir — and no long-duration plan until 2028.
Ireland has built roughly a gigawatt of grid-scale batteries in five years — and almost all of it stores power for one or two hours. The long-duration storage its high-wind grid actually needs barely exists: the only asset that can shift energy across a day was built in 1974.
Six observations stand out: a gigawatt of batteries, built in five years — a genuine, fast success driven by system-services revenue; half the 2030 target, and the wrong shape (one to two hours, when the system needs hours); one reservoir, fifty years — Turlough Hill is still the only long-duration asset; the long-duration route to market doesn’t arrive until 2028; the revenue floor is moving — DS3 gives way to competitive auctions from late 2026; and the market may move before the policy does — Silvermines and a Google-backed CO₂ battery are arriving ahead of the State’s own procurement.
The consequence: for short-duration storage the coming repricing is the test; for long-duration, the opportunity is real but gated by policy — and the design of EirGrid’s tender is the pivotal decision for the sector.
Ireland’s storage story is a study in incentives. Where the market sent a signal — the DS3 system-services payments — the country built a gigawatt of batteries in five years, fast and largely private. Where it sent no signal — for long-duration storage — it built nothing in fifty years except a single reservoir. The lesson is not that Ireland cannot build storage; it plainly can. It is that the storage that matters most for a high-wind island grid is exactly the storage the market has not been paid to build.
Storage is the firming-side of an integrated picture: it is the missing shock-absorber behind the 11.3% of wind thrown away on the grid of IE001, the tool that lets the renewables of IE003 run the system at 80%, and the flexibility the data-centre load of IE002 will increasingly pay for. You cannot run a grid on 80% wind and solar without something to store the surplus — and Ireland’s fleet is half the size and the wrong shape for that job.
The shape of this volume. § 1 sets out the headline figures. § 2 develops twelve findings, each confidence-annotated. A chapter-weight pull at p. 14 anchors the argument. § 3 decodes the boom and the gap — the fleet, the shape, the reservoir and the 2028 promise. § 4 sets out the pipeline and the projects. § 5 decodes the repricing and the 80% test. § 6 sets out recommendations, with a one-page tear-out at p. 25.
Island battery storage reached ~1,052 MW / 1,314 MWh in 2025, from a standing start in 2020 — one of Europe’s fastest build-outs. EirGrid’s 2030 target is at least 2 GW. Almost all of the fleet stores one to two hours; the first four-hour battery only opened in February 2026, and the only asset that shifts energy across a day dates from 1974.
The fleet (steel) against the dashed target — and the 1974 reservoir (ink) that is still the only long-duration asset. Half the megawatts, the wrong duration.
HIGH — anchored to a named Irish primary source (EirGrid, ESB, DECC, the SEM Committee). MEDIUM — direction of travel confirmed but a material figure is unpublished or contested. Every forward position is labelled as a projection.
Findings 01 – 12 · summary
Findings 01 – 09 follow in detail · pp. 11 – 13 · 10 – 12 carried on the § 6 matrix
Island grid-scale battery storage reached roughly 1,052 MW / 1,314 MWh in 2025, up from 969 MW / 1,074 MWh in 2024 (Green Collective, January 2026). ESB is the largest owner — Aghada 2 (150 MW / 300 MWh, Co. Cork, November 2024) and Poolbeg (75 MW / 150 MWh, Dublin, February 2024) — with VPI, Statkraft and others building fast. From a standing start in 2020, this is real infrastructure, and it was built without a dedicated support scheme, on the strength of the payments the grid makes for fast frequency and reserve services. It is one of the fastest storage build-outs in Europe and a genuine Irish success. The caveat, which the rest of this report unpacks, is that it is the right technology for the wrong problem: excellent at steadying the grid second by second, largely irrelevant to the hours-long wind surpluses that Ireland actually needs to capture.
EirGrid’s target is at least 2 GW of storage by 2030; the fleet is about half that and almost entirely short-duration, one to two hours. Ireland’s first four-hour battery, Statkraft’s Cushaling (22.8 MW / 91.2 MWh, Co. Offaly), only opened in February 2026. On a system that curtails wind for hours at a stretch, a fleet of one-hour batteries is the wrong tool: it can shift a little energy for a short time, but it cannot store an afternoon of surplus wind to release in the evening peak. The gap is therefore both quantitative — half the megawatts — and qualitative — the wrong duration. Closing it means not just more batteries but longer ones, and the market has, until very recently, had no reason to build them, because the revenue streams that paid for the boom rewarded fast response, not long duration.
Turlough Hill (292 MW / 1.59 GWh, Co. Wicklow, 1974) is still Ireland’s only pumped-storage and only genuine long-duration asset. ESB has refurbished all four units and is running a structural digital twin aiming to extend its life by up to 20 years. Nothing of comparable duration has been built in half a century. That is a remarkable fact for a country that has spent that half-century adding wind: the more wind Ireland built, the more it needed something to store it across hours and days, and it built nothing. The reasons are site scarcity, long lead times and the absence of a market signal — pumped hydro needs specific geography and a decade to build, and no support scheme rewarded it. Turlough Hill remains the workhorse of Irish long-duration storage, fifty years old and, until Silvermines, without a successor.