In the first quarter of this year, day-ahead electricity markets across the European Union recorded 1,223 hours of negative prices. In practical terms, that means generators were paying to produce electricity rather than being paid for it — because more power was flowing into the grid than anyone could use, and the system had nowhere to put the excess. Germany and Spain were among the hardest hit. Portugal and the Netherlands were not far behind. The surplus was, in almost every case, coming from wind and solar installations that Europe had spent hundreds of billions of euros building.
The number that best captures what this costs is €7.2 billion. That is the value of renewable electricity that was simply discarded across seven European countries in 2024 alone, according to a joint analysis by Beyond Fossil Fuels, E3G, Ember and IEEFA. In July 2025, Spain’s curtailment rate — the share of renewable output that was generated and then thrown away because the grid could not absorb it — reached almost 11 percent of total renewable production. In Germany, curtailment events have become structural rather than occasional. Roughly 1,700 gigawatts of clean energy projects sit stranded in connection queues across the continent, more than three times the capacity Europe needs to meet its 2030 climate goals. Grid congestion costs the EU an estimated €4.2 billion annually. These are not rounding errors in a successful transition. They are the central operational failure of a system that has expanded one component — generation — far faster than it has expanded the rest.
The architecture of Europe’s electricity grid was not designed for this. It was built, over decades, around the logic of dispatchable generation: coal plants, gas turbines and nuclear reactors that produce power when you tell them to and stop when you tell them to stop. The grid’s transmission lines, its switching infrastructure, and its market rules were all calibrated for a world in which supply followed demand. Wind and solar work the other way around. They produce when the resource is available — when the wind blows, when the sun is overhead — and the system has to absorb whatever arrives. When generation is geographically concentrated, when it arrives in large volumes at the same time, and when there is insufficient flexibility in the system to shift that power to where it is needed or store it for when it is, the grid operator has only one option: tell the wind farm or solar array to stop generating. The power that could have been produced is lost. The infrastructure that produced it sits idle. The carbon that could have been avoided continues to be emitted from the gas peaker that runs instead.
The scale of the problem has been building since 2023, but the first quarter of 2026 made it impossible to ignore. Negative prices at that frequency are not a market signal that something unusual has happened. They are a signal that the market’s structural design is failing to match supply with demand across time and space. ENTSO-E, the body that coordinates Europe’s transmission system operators, has recognised in its Ten-Year Network Development Plan that grid expansion has become the primary bottleneck of the energy transition. Additional cross-border power lines, it estimates, could cut system operating costs and reduce renewable curtailment by 30 terawatt-hours per year by 2030 — more than Slovakia’s entire annual electricity consumption. Most of the reinforcement projects that would deliver this will not materialise until after 2027 or 2028 at the earliest.
In the meantime, the agreement signed in Luxembourg on 26 June by twenty-two EU energy ministers represents the most significant collective response to the storage crisis so far. The tripartite pact, brokered by the Commission and bringing together national governments, financial institutions and industrial storage developers, sets a target of 200 gigawatts of storage capacity by 2030 — compared to the 55 gigawatts currently installed. The twenty-two signatories have committed to adding around 30 to 35 gigawatts of new capacity by 2028, aiming for at least 20 percent more than the annual installed capacity recorded in 2025. The Commission is funding part of this through the Innovation Fund: €2.5 billion is currently being disbursed to more than fifty energy projects across the bloc, with a focus on renewable generation, grid modernisation and energy efficiency.
Storage is the obvious answer to curtailment, but it is not the only one, and in isolation it is not sufficient. The “hybridisation” that European regulators and industry bodies have been pressing for this week refers to something more systemic: the co-location of generation and storage assets, the integration of flexible demand into grid management, and the use of excess renewable electricity to produce green hydrogen through electrolysis rather than simply disposing of it. Green hydrogen produced from curtailed solar and wind power is not the most efficient use of those electrons — electrolysers running on cheap surplus power will necessarily operate intermittently, which raises questions about capital utilisation and per-kilogram production costs — but it is categorically more useful than the alternative of generating nothing. Portugal and Spain, two of the countries with the highest curtailment rates, are also two of the countries with the most ambitious green hydrogen export strategies. The connection between those two facts has not yet been fully exploited in either policy or infrastructure terms.
The grid package that the Commission has been developing — and that the Council is expected to reach agreement on by the end of 2026 or early 2027, after Cyprus made it a presidency priority — is the legislative vehicle through which most of this will need to move. It covers permitting acceleration for transmission infrastructure, market design reforms to better compensate flexibility providers, and the framework for demand-side response. The details matter enormously. Market rules that were written for a world of dispatchable generation create perverse incentives for renewable operators: when negative prices arrive, the rational response is to curtail production rather than pay to export power to a market that does not want it. Redesigning those rules to reward flexibility and punish unnecessary curtailment requires changes to market architecture that member states have historically resisted because they constrain national dispatch decisions.
The data centre sector adds a further dimension that barely appears in most policy discussions but is reshaping the operational challenge in real time. The rapid growth of AI-driven computing load is creating a new category of large, relatively flexible industrial demand that could, in principle, absorb surplus renewable electricity during the periods when curtailment would otherwise occur. In practice, data centres in Europe are concentrating geographically in specific hubs — Ireland, the Netherlands, Sweden — rather than distributing to where the renewable surplus is. The result is that the fastest-growing source of electricity demand in Europe is not being planned in coordination with the fastest-growing source of electricity supply. Grid operators in several markets are already responding to this with stricter connection conditions and longer connection timelines, which delays both data centre deployment and renewable integration simultaneously.
What connects all of these strands is a basic structural mismatch that the EU’s climate targets did not anticipate and that its regulatory architecture is only now beginning to address seriously. Europe has built — and is continuing to build — enormous quantities of renewable generation capacity. It has done this quickly, successfully, and at rapidly falling cost. What it has not done, at anything like the same pace, is build the transmission infrastructure, the storage capacity, the market mechanisms, and the flexible demand that a generation-heavy system requires in order to function. The consequences of that imbalance are now arriving in the form of negative electricity prices, discarded clean power, stranded grid connection queues, and €7.2 billion in annual waste.
The Commission knows this. ENTSO-E knows this. The energy ministers who signed the Luxembourg storage agreement know this. The gap is not in diagnosis but in the pace and ambition of response. Grid reinforcement projects that will take until 2028 to materialise were approved in planning processes that began years ago; the projects entering planning today will not be operational until well into the 2030s. Storage targets that sound large — 200 gigawatts by 2030 — remain roughly four times the current installed base, to be achieved in four years, in a sector that has never deployed at that rate. Permitting reform that member states have been discussing for years continues to move more slowly than the renewable capacity it is supposed to enable.
The wind was blowing this morning over the North Sea. The solar arrays across the Iberian Peninsula were generating at close to peak output. Some of that electricity reached homes and factories and electric vehicles. Some of it was simply turned off, because the infrastructure to carry it, store it, or use it in another form was not there. That is not a market failure in the technical sense. It is a policy failure in the plainest sense: the product of decisions made slowly, resources deployed unevenly, and a transition managed at a pace that has been consistently outrun by the physics of the problem it is trying to solve.
Subscribe to EuroLuminant for independent European journalism.
