During the worst of the European heatwave last month, German wholesale electricity prices moved from €86/MWh at midday to €566/MWh by 8pm. A 6.5x swing, inside a single day, inside a system most people assume runs on stable, predictable pricing.
It wasn't an isolated wobble. The UK system operator paid £1,379/MWh to import emergency power. That’s roughly fifteen times a typical day-ahead price. Five UK gas plants had to cut output because they couldn't cool themselves at 36°C. In Brittany, France, a heat-stressed transformer failed and took 70,000 customers with it. EDF throttled nuclear generation because the rivers that normally cool those plants had become too warm to use safely.
This was about a mismatch in timing, not a generation failure. Germany had plenty of solar at midday and prices were low precisely because supply was abundant. The problem arrived six hours later, when the sun went down and demand didn't.
That gap between when power is cheap and when it's needed is one of the biggest challenges for a cleaner grid. We’re grappling with it right now, not in the future. Germany, the UK and France – three of Europe’s largest electricity markets – were all hit in the same week, at the same time, for the same reason.
The pattern repeats
The late June heatwave coincided with London Climate Action Week in the British capital. The heat pushed UK grid balancing costs above £11m in a single evening (five times the daily average). Batteries stepped in to stabilise frequency and gas plants failed under the heat they were supposed to be managing. Britain currently pays around £1.46bn a year to curtail clean power it has already generated and burn gas instead, because the grid can't move that electricity to where and when it's needed. That's a coordination problem.
Step back further and the same shape appears in the policy debate. The argument currently playing out over AI and data centre power (gas versus clean) assumes the constraint is generation: build more firm capacity, fast, however you can. The IEA's actual finding says otherwise: 80%+ firm renewables paired with comparatively small amounts of demand flexibility is the best-performing path for integrating new large loads, including data centres. The hyperscalers running cooling and compute schedules to manage their own peak demand have already arrived at the same conclusion that European grid operators are relearning in 36°C: the constraint isn't supply, it's timing.
Why this is the year it stopped being theoretical
Europe has spent the past eighteen months building the policy architecture for exactly this problem. The European Commission's grid roadmap puts smart, demand-responsive infrastructure at its centre. COP31's electrification target of 35% of global energy demand electrified by 2035 only works if the electricity arriving on the grid can actually be absorbed and used instead of wasted. IRENA's January 2026 analysis underpins both as it found that the power sector will need three times more flexibility by 2030 simply to manage daily variability in solar and wind. The consensus across regulators, system operators and the IEA is no longer contested: electrification is the strategy, and flexibility is what makes it deliverable rather than aspirational.
What the heatwave did was compress that argument into six hours of price data that anyone can read on a chart. €86 to €566 doesn't need a policy briefing to make its point.
The part still missing
Here's the part that doesn't get said enough: none of this is solved by building more of anything. More generation doesn't fix a timing mismatch. Germany had abundant solar and still hit €566/MWh. More grid capacity doesn't fix it either, not on the timeline the next decade demands. What closes the gap between cheap daytime power and expensive evening demand is the tech that can see millions of flexible assets (batteries, EVs, heat pumps, smart tariffs) and move demand to meet supply in real time, automatically, at a scale no control room can manage manually.
Every expert quoted in the coverage converged on flexibility and storage as "the missing piece." But no one talked about the intelligence layer that underpins that flexibility. Here at Kaluza, we’re building that intelligence layer to power an energy system that is more resilient to the heat waves that will only get longer and more intense in Europe and across the world.
This summer’s heat waves will fade from the news cycle. The price data won't. Every degree of warming adds roughly a gigawatt of cooling demand in a market the size of France alone, and there's no version of the next decade where this kind of swing becomes rarer. The question worth asking isn't whether Europe's grid will face this again. It's whether the intelligence layer to manage it gets built in time.
