Italy and nuclear power: from the 1987 referendum to the 2025 reopening

At 01:23:40 on the night of 26 April 1986, the explosion of Reactor 4 at Chernobyl sent a radioactive cloud into the sky that within days would reach the Po Valley. Eighteen months later, on 8 and 9 November 1987, Italians went to the polls to vote on three referendum questions about nuclear power. All three passed by overwhelming majorities. From that day onward, Italy stopped producing atomic energy.

Thirty-eight years later, in 2025, Parliament is debating an enabling law to reopen it.

1963-1986: the Italian nuclear era

Few remember it now, but for a brief period Italy was one of the world's major nuclear powers. The Latina plant entered service in 1963: a 153 MW Magnox reactor built by Britain's Nuclear Power Group, at the time the most powerful in Europe. Garigliano followed in 1964 (160 MW GE BWR) and Trino Vercellese in 1965 (270 MW Westinghouse PWR). In 1981 Caorso came online—an 882 MW BWR, the largest in the Italian fleet.

In that same 1981, Italy was the world's fifth-largest producer of nuclear energy by installed capacity. Nuclear covered roughly 5% of the national electricity mix, and the freshly approved National Energy Plan called for 12,000 MW of new atomic capacity by 2000. New sites were being discussed in Puglia, Lombardy, Piedmont. Montalto di Castro was under construction.

Then came Chernobyl. And a year later, the referendum.

November 1987: three questions that changed everything

It is worth recalling what Italians actually voted on in the autumn of 1987. None of the three questions asked directly to close nuclear power. They concerned:

1. The repeal of the compensation mechanism for municipalities hosting plants (approved by 80.6%).
2. The repeal of ENEL's ability to participate in nuclear plants abroad (79.7% yes).
3. The repeal of CIPE's substitute powers for siting plants (71.9% yes).

In other words, voters did not shut down nuclear power: they took away its toolbox. Without compensation to host municipalities, no mayor would accept a plant. Without CIPE's substitute powers, any local resistance became a veto.

The De Mita government read the vote as a mandate for phase-out. Construction stopped in 1988: Montalto di Castro, already 70% complete with GE-Hitachi reactors ready, became the most expensive monument to Italian energy second-guessing (roughly one billion euros in then-currency, never recovered). In 1990 the last operating reactors were shut down, Caorso and Trino. In 1999 Sogin was founded, the state-owned company in charge of decommissioning—still at work today.

1988-2008: twenty years of silence

For two decades nuclear power in Italy was a bipartisan political taboo. The country imported ever more electricity from France (paradoxically, produced by EDF reactors just over the border, some installed at sites geologically indistinguishable from the Italian side of the Alps), burned Russian and Algerian gas, built combined-cycle gas plants. By 2008, 12% of Italy's electricity was imported, against a European average of 3%. Technical debate stayed confined to ENEA conferences and a handful of university departments, while the generation of Italian nuclear engineers trained in the 1970s retired without direct heirs at home—many ended up working for EDF, Westinghouse or the IAEA.

2008-2011: the Berlusconi attempt

In 2008-2009 the fourth Berlusconi government tried a turnaround. A bilateral agreement with France was signed in Rome in February 2009, the Sviluppo Nucleare Italia joint venture between Enel and EDF was set up, and a plan was drafted for four third-generation EPR reactors. The Development Law 99/2009 and Legislative Decree 31/2010 rebuilt in a few pages the regulatory framework that had been missing for twenty years. Preliminary site surveys were done—never made public—and a first operational reactor was discussed for 2020.

Then, on 11 March 2011, the Fukushima tsunami overwhelmed the Japanese reactors. Three months later, on 12-13 June, Italians returned to the polls. Referendum to repeal the nuclear law: 94.0% yes, with 54% turnout. Stop, again. Final, it seemed. The joint venture was quietly wound down in 2012.

2022-2026: back on the agenda

The wind shifted thanks to a combination of three factors: the 2022 energy crisis after Russia's invasion of Ukraine and the gas cut-off, the 2050 decarbonization targets requiring non-emitting baseload, and the surge in electricity consumption forecast by AI data centers.

In October 2023, Italy's Ministry for Environment and Energy Security, led by Gilberto Pichetto Fratin, launched the National Platform for Sustainable Nuclear (PNNS), a technical roundtable involving universities, industry, unions and environmentalists. In September 2024, the PNNS operational plan indicated a range between 11% and 22% of the 2050 electricity mix from new-generation nuclear—SMRs (Small Modular Reactors) and Gen IV reactors, not the large EPRs of the Berlusconi attempt.

Through 2024 and 2025, parliamentary work began on a sustainable nuclear enabling law, intended to rebuild the regulatory framework: authorizations, safety, spent fuel management. There is discussion of strengthening ISIN, the National Inspectorate for Nuclear Safety, currently under-resourced for a country that seriously intends to build reactors.

Polling shows a divided but shifting public. Censis 2024: 51% of Italians favor SMRs, 38% are opposed. Support drops when large traditional reactors are mentioned, but the picture is radically different from 2011. Among under-35s, the share in favor rises to 58%—a number that reflects a generation that has lived through the climate crisis more than Chernobyl, and that weighs nuclear risk against global warming rather than in absolute terms.

Who builds, who is thinking about it

Italy's nuclear industry never fully disappeared. Newcleo, an Italian-British scale-up founded by Stefano Buono (ex CERN, ex Advanced Accelerator Applications), develops lead-cooled Gen IV reactors fueled by MOX from existing waste. Series A round of $310 million in 2023, London headquarters but technical core in Italy and France. Ansaldo Nucleare, a subsidiary of Ansaldo Energia, has been supplying components for AP1000, EPR and other international reactors for decades: Italy never stopped exporting nuclear engineering—it just stopped consuming it.

Edison has signed various exploratory memoranda. Enel and Eni maintain more cautious positions, waiting for regulatory clarity. On the academic side, the Politecnico di Milano with the LENA reactor, the University of Pavia with Triga, Turin and Pisa have kept training pipelines alive—miraculously—throughout the silent decades.

The real challenges

Reopening nuclear power in Italy is not a matter of signing a law. There are at least five knots.

Political: the Five Star Movement and part of the PD are opposed, the centre-right coalition is in favor. Stability of choices across legislatures is the most uncertain variable. Regulatory: ISIN has limited staff and limited operational expertise after thirty years without working reactors. Siting: no preliminary selection has been made, and local resistance (NIMBY) is predictable in any candidate region. Timing: in the best scenario, a first Italian SMR would be operational no earlier than 2032-2035. Waste: the National Repository for radioactive waste, in the works for over twenty years, has yet to be built; without it, any new program limps.

And then there is the question nobody quite wants to ask out loud: who pays? A 300 MW SMR costs between 3 and 6 billion euros per first-of-a-kind unit, with historical over-budget risk in the 50-100% range on the first units of a series (see Olkiluoto in Finland and Vogtle in the United States). Regulated capacity market, dedicated PPAs with large industrial consumers, contracts for difference on the British Hinkley Point model: all viable models, none yet chosen in Italy. Pure market logic, today, does not finance nuclear in any OECD country.

Why it matters, now

The choice is not ideological. It is infrastructural. Italy today hosts 0.4% of European data center capacity—against Germany's 25% and Ireland's 13%. One reason is the relative scarcity of abundant, dispatchable, competitively priced energy: Italian wholesale electricity prices have historically been among Europe's highest, on average 30-40% above the EU mean. The gigawatt a modern AI campus needs is exactly what a reactor can supply twenty-four hours a day, with a capacity factor above 90% and a land footprint of a few hectares.

The question is not whether Italy will return to nuclear. It is whether it will have an option in ten years, or just an observation—of electricity it keeps importing, of compute that lands elsewhere, of industrial supply chains rebuilt without us.

Italian nuclear power in 2035 is being decided today, in the 2025 parliamentary committees. It is not a bet on the past, but on the physical shape of the country ten years from now.