The Danube has run dry. Hungary's nuclear dream should cool too

WATER is supposed to be a reliable input for a nuclear power plant. It is a fixed, plentiful resource - or so engineers assumed when they built the Soviet-era Paks facility next to the Danube in the 1970s. This week that assumption evaporated. On 31 July Hungary's prime minister, Péter Magyar, announced that Paks would shut down entirely for the first time in its 44-year history. The Danube has fallen so low that the plant's suction nozzles can no longer reach sufficient cooling water. The Soviet-era plant supplies nearly half of Hungary's electricity. Its four reactors, with a combined capacity of 2GW, have been running at less than half capacity in recent days. A total blackout of the plant is now imminent.

The water-level crisis at Paks is more than an operational footnote. It exposes a structural tension at the heart of an energy model built around a single technology on a single river, while climate change steadily undermines that river's reliability. And it arrives at a moment when the Hungarian government is deciding whether to double down on that same model with a hugely expensive expansion project built by Russia.

The plumbing problem

Nuclear power is not truly location-independent. Reactors need vast quantities of water to remove waste heat. Paks draws from the Danube and discharges warmed water back into it. Environmental regulations cap the temperature of discharged water to protect river ecosystems. So the constraint is not merely about water volume. It is about water volume at an acceptable temperature.

Both conditions have failed. The Danube's gauge level at Budapest fell to 23cm on 29 July, breaking the previous record low of 33cm set in the drought of 2018, according to the national water authority. At Paks itself, Mr Magyar said the level could fall to minus 144cm relative to the gauge's zero reference - far beyond the minus 134cm at which a full shutdown becomes mandatory. The plant's suction infrastructure was designed for a river that does not recede this far. It cannot adapt to conditions that have become the new normal.

River flow across the Danube basin has collapsed to roughly a third of its average for July. The problem is compounded by temperature: prolonged heat has raised water temperatures as well as lowered volumes, tightening the thermal discharge window even before the river runs shallow. The result is a pincer movement. The river cannot absorb the reactor's heat without exceeding environmental thresholds, and it cannot supply enough volume to cool the reactors in the first place.

Hungary is not alone. Romania, downstream on the same river, has shut down one of its two reactors at Cernavoda and is considering a second. France has seen repeated forced shutdowns and output cuts at plants along the Garonne and other rivers. The pattern is becoming systemic: as European summers grow hotter and drier, nuclear plants sited on rivers become less reliable at the precise moment when cooling demand pushes electricity consumption up.

The political complication

A reactor shutdown is a technical problem. The Paks II expansion project, which has just begun construction, is a political one. In February the "first concrete" was poured for the first of two new VVER-1200 reactors designed by Rosatom, Russia's state nuclear corporation. The project was agreed under the previous government of Viktor Orbán in 2014, awarded without a tender and financed 80% by a Russian loan. Its stated price was €12.5bn. Unofficial estimates now put the total cost closer to €25bn, according to analysis from the Heinrich Böll Foundation, an independent research institute.

The new Magyar government, which took office in May after the April election, has already flagged the project for review. Its energy-minister nominee, István Kapitány, has said that the financing and implementation terms of Paks II must be examined, adding that the contracts are classified. Mr Magyar himself has called the project's cost "over-inflated". Since taking office, however, he has offered no definitive decision.

This is an awkward position. The shutdown of Paks makes the case for more domestic nuclear capacity look more urgent. Nuclear currently provides 42% of Hungary's electricity - 16TWh out of 38TWh generated domestically in 2024, according to data compiled by the World Nuclear Association. Hungary is a net importer of electricity, bringing in 22.7TWh and exporting only 12TWh. To replace Paks's output during a multi-week shutdown, Hungary must rely on interconnections with Slovakia, Austria, Romania and other neighbours. Mr Magyar said import capacity of 3,600-3,800MW could cover the shortfall. But when the same drought is crippling generation across the region, interconnections become a shared vulnerability rather than a safety valve.

The counterargument

To be sure, nuclear energy remains one of the few genuinely low-carbon sources of baseload power. The argument for expanding it is not that it is perfect but that the alternatives - gas, coal, or an intermittent mix of wind and solar without sufficient storage - carry their own trade-offs. Coal generation in Hungary is already set to be phased out by 2030. Gas plants are subject to supply and price shocks. Solar and wind have improved dramatically but still require backup capacity and grid-scale storage that remain expensive. In that context, a 2.4GW addition from Paks II would shift Hungary's nuclear share from 42% towards the government's target of 60%, reducing both carbon emissions and import dependence.

Yet the trouble is not nuclear in principle but nuclear concentrated in a single form on a single river. Paks II would add two more reactors that cool in the same Danube that is already failing to sustain four. The expansion would increase the very vulnerability that the shutdown has exposed, rather than addressing it. It is the energy equivalent of rebuilding a house on a flood plain after the flood.

What the system should do

A better strategy would diversify. The first task is to extend and modernise the existing reactors at Paks with cooling systems that reduce reliance on direct river water. Closed-loop cooling towers, which circulate heated water and release it as steam, would insulate the plant from river level and temperature. They carry an efficiency penalty and require chemical treatment, but they would make the difference between a forced shutdown and a continued operation. Existing reactors can be retrofitted. New ones should be designed with them from the start.

The second task is to diversify Hungary's generation mix away from a single-source dependency. Solar already contributes 24% of Hungary's generation - the second-largest source after nuclear. That share can grow. The country also has room for more gas-fired peaking capacity, storage, and demand-side flexibility to weather periods when nuclear or hydro output is constrained.

The third, and most politically difficult, task is to renegotiate or abandon Paks II in its current form. A €12.5bn-to-€25bn commitment to a Russian contractor during a period when Russia is at war with a neighbour, financed by a loan that locks in a long-term geopolitical dependency, is a bad bargain on its own terms. The new government is right to review it. But the review should not be an excuse for inaction. Suspending construction while alternatives are assessed is the responsible course. Hungary needs nuclear power. It does not need to buy it from the wrong partner at the wrong price in the wrong configuration.

The broader lesson

The shutdown of Paks is a small country's crisis but a continent's warning. Europe's bet on nuclear as a cornerstone of decarbonisation is real: several European countries, including France, rely on reactors sited on rivers. A French Court of Auditors report from 2023 found that forced shutdowns due to low water would become three to four times more frequent by 2050. The heatwave of 2022 was a preview. The drought of 2026 is a confirmation.

Nuclear energy will be a casualty of climate change even as it is marketed as its cure. The thermodynamics do not bend to the politics. The solution is not to abandon nuclear but to build it better: with cooling systems resilient to drought, with sites chosen for water security, and with a generation mix that does not put all of a nation's electricity on the reliability of a single river. Hungary's new government has a chance to learn that lesson before it pours more concrete.