Loading market view
Scanning European outage signalsFrench nuclear units account for 19.9 GW of offline capacity — 2.6× the five-country median — while Spain's hydro fleet adds another 13.9 GW. Normalised for installed capacity, however, Sweden's 173.7 MW/GW makes it the densest outage concentration of the five countries reviewed.
Data source notice: This analysis is based on publicly available ENTSO-E Transparency Platform data. Latency patterns described here are observations from public disclosure records and do not constitute regulatory findings or determinations of wrongdoing.
French nuclear generators reported 19.9 GW of unavailable capacity in the most recent 14‑day window—2.6 times the five‑country median of 7.6 GW and 40.6% of all offline megawatts across the FR‑ES‑DE‑SE‑BE set. Normalised for fleet size, however, Sweden’s 173.7 MW per installed GW makes it the densest concentration of outage capacity, a position that better reflects actual system tightness than the raw tally. The gap between the two rankings is explained almost entirely by the seasonal rhythm of French reactor maintenance and the footprint of a handful of large Swedish nuclear units, not by an unexpected supply shock.
Active unavailability across the five countries in the most recent 14‑day window totalled 48.9 GW. France alone contributed 19.9 GW, placing it first in the peer group by a wide margin. Spain follows at 13.9 GW, Germany registers exactly the median at 7.6 GW, and both Sweden and Belgium sit below 5 GW. The declarations—reported by TSOs for all generation types—are dominated by nuclear capacity in France.
Machine-readable sidecar (JSON) - confidence n/a · 0/0 claims backed
The French figure is driven by a handful of large units. The single largest outage in the data set is a French nuclear asset of 1,500 MW, recorded from 12 through 23 June. The table of the largest active outages confirms the top entries are weighted toward French nuclear generators, with the next tier comprising a Swedish nuclear unit of 1,172 MW and a two‑unit Swedish station whose two 1,065 MW blocks were both offline simultaneously. The size of these individual events—a few thousand‑megawatt class reactors in planned maintenance—inflates the aggregate, rather than any broad fleetwide issue.
Ranking by raw megawatts rewards fleet size. France has roughly 148 GW of installed generation capacity; Sweden’s fleet is about 31.9 GW. Normalising by installed gigawatts flips the league table. Sweden moves to rank 1 at 173.7 MW/GW—1.93 times the five‑country normalised median of 90.1 MW/GW. France drops to second at 119.3 MW/GW.
Sweden’s density stems largely from two nuclear outages: the 1,172 MW unit and the two 1,065 MW blocks together remove over 3.3 GW from a relatively small fleet. Additional unavailability across hydro and wind pushes total offline capacity to roughly 5.5 GW. With only 31.9 GW installed, that translates directly into a high per‑GW ratio. From a system‑tightness perspective, the density measure is the more informative gauge: it expresses how many megawatts are missing for each gigawatt of nominal supply, a number that correlates more directly with residual margin than an un‑normalised sum.
The spring‑to‑autumn nuclear refuelling window concentrates planned French maintenance between May and August. Every outage in this snapshot with a forward‑looking end date fits that pattern. The longest‑running event began on 1 May and is scheduled to end on 31 August—a span of nearly four months typical of an extended maintenance programme involving fuel exchange and major component work. Other units carry end dates of 31 August and 4 July, timelines that align with standard 30‑ to 90‑day refuelling and maintenance schedules.
That the book is dominated by such events suggests the current 19.9 GW figure is largely cyclical. The absence of a simultaneous spike in unplanned outages is notable. The 12–23 June outage of the 1,500 MW unit, for instance, is short enough to possibly be a planned partial refit or a brief extension of earlier work, though the ENTSO‑E Transparency labels do not distinguish between planned and forced unavailability in the raw feed.
Spain’s 13.9 GW headline number is inflated by a single‑day concentration that carries little energy‑at‑risk significance. On 17 June 2026, six separate unavailability declarations were made on a single Spanish hydro reservoir unit, together summing to roughly 8.1 GW of unavailable capacity. All six events fell within a single nine‑hour window; none persisted beyond that day. The declarations likely represent either staggered turbine outages or a coordinated set of operating limits across multiple generators at the same facility, consistent with reservoir management or short‑term maintenance.
Contrasted with French nuclear outages that run for weeks or months, the Spanish hydro cluster adds substantial megawatt‑hours to the raw snapshot but imposes negligible energy curtailment. Removing those one‑day events would drop Spain’s 13.9 GW total substantially, reinforcing the need for a duration‑weighted lens when assessing tightness.
Cross‑border capacity data from a single‑hour snapshot (12 June 2026, 23:00 UTC) shows the German–French and Swiss–French borders are among the most constrained in the flow‑based set. The minimum MACZT on DE–FR stood at 38.2% with three critical network elements below the 70% threshold; CH–FR dropped to 57.6% with two elements below 70%. An internal French network element (FR–FR) showed an even lower minimum of 19.99%, also with three elements under 70%.
The timing is consistent with a period when reduced French nuclear output limits available export capacity, tightening both cross‑border and internal constraints. The data represent a single hour, however; without a longer time series it is impossible to say whether these bottlenecks are persistent. The MACZT figures are, nevertheless, among the lowest observed in the region for that hour, and they warrant monitoring alongside the outage book.
Track whether the French nuclear units with late‑summer end dates—31 August and 4 July—return to service as scheduled. Any extension into the third quarter would shift the signal from seasonal to structural and could tighten capacity margins further during the late‑summer demand period. Separately, the absence of day‑ahead price data across the zones tested means this analysis cannot tie the supply picture to price outcomes; watch for that data to become available, as it is the natural next link in the evidence chain. Finally, revisit the density‑normalised outage league on a monthly cadence: it strips out fleet‑size effects and provides a fair cross‑country benchmark that will become more valuable as the refuelling season progresses and the distribution of unavailability shifts.