Europe’s Missing Electricity: 100 TWh a Year, and the Weather Isn’t to Blame
Four years after the 2022 gas shock, European power consumption is still about 5% below its pre-crisis level. We weather-adjusted 22 countries’ grid load with heating-degree-day data to test the popular explanation — mild winters. It fails. The missing electricity is structural, and it is concentrated in the industrial core.
WorldPulse Research·
Germany · daily load
1,016 GWh
Aug 8, 2026
France · daily load
953 GWh
Aug 8, 2026
Spain · daily load
691 GWh
Aug 8, 2026
Portugal · daily load
133 GWh
Aug 8, 2026
Latest daily electricity load (ENTSO-E Transparency Platform), updated daily. Explore all 25 countries on the EU electricity monitor.
Here is the finding up front. In the twelve months to January 2026, electricity consumption across the 18 European countries we can measure consistently was 4.9% below its 2017–19 average — roughly 106 terawatt-hours of missing demand, about what the entire Netherlands consumes in a year. Weather explains almost none of it: adjusted for heating and cooling degree days, the structural gap is 5.0%. The last twelve months were in fact slightly colder than the pre-crisis baseline.
The question matters because the mild-winter alibi is comforting. If Europe’s power demand were only sleeping through warm Decembers, it would come back by itself. If instead the demand was destroyed — factories curtailed after the 2022 gas-price shock, output moved elsewhere — the gap is a measure of lost industry. Since 2022, the cumulative shortfall across our 25-country panel adds up to about 541 TWh (2022–2025) — more than a full year of France’s consumption.
WorldPulse holds both halves of the answer: daily grid load for 25 European countries from the ENTSO-E Transparency Platform (see the German series), and monthly heating and cooling degree days computed from the Copernicus ERA5 reanalysis (de-hdd). Regressing pre-crisis load on degree days gives each country’s weather sensitivity; whatever the weather cannot explain is structural.
1. A twentieth of Europe’s power demand went missing — and stayed missing
By 2021 European consumption had fully recovered from COVID — our panel was just 0.1% below the 2017–19 baseline that year. Then came the gas shock: the gap widened to 3.7% in 2022 and peaked at 7.6% in 2023, before narrowing only slowly to 6.4% in 2024 and 5.9% in 2025. The chart shows each country’s trailing-12-month consumption relative to its own 2017–19 average — note how Portugal escapes upward while Germany and France stay pinned 6–8% down.
Trailing-12-month electricity consumption vs 2017–19 baseline
Percent deviation of rolling 12-month grid load from the country’s own 2017–19 average, computed live from daily ENTSO-E data. Months with incomplete daily coverage are excluded.
2. The mild-winter alibi doesn’t survive the degree-day data
Weather genuinely moves European power demand — in France, our baseline regression puts the sensitivity near 49 GWh per heating degree day, the legacy of electric heating, and the fit is tight (R² = 0.95). So we let weather make its best case. It still loses. In 2024, the mildest year in the sample — ERA5 heating degree days ran 8.1% below the 2017–19 baseline — weather explained just 0.7 points of a 6.4% gap. In the twelve months to January 2026, degree days were 2.6% above baseline, so the weather adjustment makes the structural gap slightly larger than the raw one: −5.0% against −4.9%. Cold came back; demand didn’t.
Copernicus ERA5 reanalysis, base 18°C. The winter 2025–26 heating season was near-normal — the mild-winter era of 2023–24 is over, yet consumption has not returned.
3. A two-speed Europe: the industrial core lost it, the periphery got it back
The league table below covers the twelve months to January 2026, each country against its own 2017–19 baseline. The pattern is stark: every large industrial economy sits 5–8% below — France (−7.9%), Germany (−7.2%, a shortfall of 36 TWh in one year), Belgium (−6.8%), Spain (−6.2%), Austria (−6.0%), Poland (−5.8%) and Czechia (−4.7%). Meanwhile the smaller economies of the periphery are back at or above baseline, led by Portugal at +6.2%.
Spain deserves a caveat that cuts both ways. Its economy has outgrown the euro area since 2022, yet measured grid load is still 6.2% below baseline — weather-adjusted, −6.3%. Part of the answer is that ENTSO-E measures electricity delivered by the grid: rooftop solar consumed behind the meter never appears in these numbers, and Iberia’s self-consumption boom means true underlying demand is higher than grid load suggests. The same logic tempers the readings for the Netherlands and Poland. That is efficiency and substitution, though — not evidence for the weather story.
Electricity consumption, 12 months to January 2026, vs each country’s 2017–19 average. “Weather effect” is the part of the gap attributable to degree-day deviations; “structural” is the remainder. Belgium and the Netherlands lack usable degree-day history and show the raw gap only. Sweden, Norway, Denmark, Romania and others are excluded for partial bidding-zone coverage or data gaps — see the methodology note.
4. Is 2026 the year the electricity comes back?
The freshest data offers the first real hint of convergence since 2021. January 2026 consumption came in 2.2% above baseline across the panel — a cold month helped, but even weather-adjusted, January (+1.0pp) and February (+0.5pp) 2026 were the first months at or above baseline in four years. Germany’s January 2026 load averaged 1,496 GWh per day, within 1% of its pre-crisis January norm. Then again, July 2026 — the latest complete month — was still 2.2% below its baseline across 22 countries, with Germany down 7.2% and Poland down 7.0%. Summer load is where industry shows and heating doesn’t: the industrial core is not back.
ENTSO-E Transparency Platform, daily actual total load. Gaps reflect missing upstream data.
The verdict: demand destruction, not mild winters
Europe’s electricity gap is real — about 5% of pre-crisis consumption, over 100 TWh a year as of the twelve months to January 2026 — and it is not a weather story. Degree-day adjustment attributes roughly zero of the current gap to temperature, and at most 0.7 points even in record-mild 2024. What the weather gave in warm winters, it has now taken back — and the demand still hasn’t returned.
The gap lives where the 2022 gas shock hit hardest: the industrial economies of the European core, all still 5–8% below baseline on a weather-adjusted basis. Some of the missing demand is genuine loss — curtailed and relocated energy-intensive industry. Some is benign — efficiency and behind-the-meter solar that shifts consumption off the measured grid. Our data cannot fully separate the two, but their sum is stubbornly large. The early-2026 readings — weather-adjusted consumption back at baseline in January and February — are the first evidence of a turn. Whether electrification, data centres and reshoring can outrun deindustrialisation is now the live question; July 2026’s −2.2% says not yet.
Frequently Asked Questions
Why is Europe’s electricity consumption still below pre-crisis levels?
Mostly for structural reasons, not weather. In the 12 months to January 2026, consumption across 18 European countries was 4.9% below the 2017-19 average, and weather (heating and cooling degree days) explained roughly zero percentage points of that gap. The decline is concentrated in the industrial economies — France, Germany, Belgium, Spain, Austria, Poland and Czechia — and reflects a mix of industrial output lost after the 2022 gas-price shock, energy-efficiency gains, and rooftop solar that is consumed behind the meter and therefore never shows up in grid load.
How much of the drop in EU electricity demand is explained by mild winters?
Very little. Even in 2024, the mildest year in the sample — heating degree days ran 8.1% below the 2017-19 baseline — weather accounted for only about 0.7 points of a 6.4% consumption gap. In the 12 months to January 2026 the weather was actually slightly colder than the baseline, yet consumption was still 4.9% below it.
Which European countries have recovered their electricity demand?
Portugal is the clearest case: consumption in the 12 months to January 2026 was 6.2% above its 2017-19 average, essentially none of it weather. Latvia, Finland, Croatia, Switzerland, Hungary, Bulgaria and Lithuania were also at or above baseline. The big industrial economies were not: France (-7.9%), Germany (-7.2%), Belgium (-6.8%), Spain (-6.2%), Austria (-6.0%) and Poland (-5.8%) all remained well below.
Is European electricity demand growing again in 2026?
Early 2026 shows the first genuine convergence since 2021. In January 2026 the panel’s consumption rose 2.2% above the 2017-19 baseline (helped by a cold month), and on a weather-adjusted basis January and February 2026 were roughly at baseline. But July 2026, the latest complete month, was still about 2.2% below its baseline, with Germany down 7.2% — so the industrial core has not yet recovered.
Methodology. Daily actual total load per country (ENTSO-E Transparency Platform, via the WorldPulse electricity monitor) is aggregated to months as mean daily load × days in month; months with fewer than 20 reported days are dropped. The baseline is each country’s 2017–19 average for the same calendar month. Weather sensitivity comes from an OLS regression of monthly load on heating and cooling degree days (Copernicus ERA5, one representative point per country, SEAS5 forecast points excluded) over the 36 baseline months; the weather effect in any month is the sensitivity times the degree-day deviation from baseline, and the structural gap is the remainder. Baseline fits range from R² = 0.95 (France) to 0.14 (Italy North); most countries sit near 0.8. Known limitations, stated plainly: the baseline carries no trend; grid load excludes behind-the-meter solar; Sweden, Norway and Italy series cover a single bidding zone (only Italy North is shown, labelled); Slovakia is excluded after October 2025 and the Netherlands’ July 2026 reading is excluded due to upstream reporting breaks; Ireland and Serbia have no load data; Belgium, Greece and the Netherlands lack degree-day history and get no weather adjustment; Austria, Czechia and Romania are adjusted for heating only. Headline figures use the twelve months to January 2026, the latest window with complete weather data for all decomposed countries. Numbers are reproducible from the analysis workspace (analysis_runs/20260809_europes-missing-electricity). This article is analysis, not investment advice.
Sources. ENTSO-E Transparency Platform (daily actual total load); Copernicus Climate Change Service / ECMWF ERA5 reanalysis (heating and cooling degree days). Series are served live from the WorldPulse knowledge graph and update as new observations arrive; prose figures are as of publication (August 9, 2026).