Seated near Bologna with 155 metric tons of molten lead and no uranium inside, a full-scale reactor vessel began generating real electricity using electric heaters in place of a fission core

The vessel weighs just over 20 metric tons empty and sits in a research hall tucked into wooded hills south of Bologna.

Fill it with what comes next, and it tips the scale at 155 metric tons of metal that has to stay above 621 degrees Fahrenheit or it locks solid.

No uranium goes in.

The vessel itself is only slightly smaller than the one the planned commercial reactor will need, which puts this far outside the category of tabletop demonstration.

No fission happens.

And yet, when the testing program runs, that vessel will spin a real turbine and push electricity into the grid, which raises an obvious question about how that is even possible.

Electric heaters stand in for the uranium the coolant will eventually surround

The machine is called PRECURSOR, built at the ENEA Brasimone Research Centre in the Apennine foothills about 50 miles south of Bologna. Its designers placed banks of high-powered electric heaters exactly where fuel assemblies would sit in a live unit, and those heaters deliver roughly 10 thermal MW into a pool of molten lead. The lead cannot tell the heaters from a fission core. It absorbs the heat, circulates through the vessel under the same fluid dynamics a fueled reactor would produce, and carries the energy toward a steam generator.

That steam generator drives a commercial-grade turbine supplied by Fincantieri, the Italian shipbuilder better known for cruise liners than power plants. The turbine is rated to produce approximately 2 MW of electricity, and a portion of that output feeds back to offset the power the heaters consume. The loop closes on itself, from electric heat to molten metal to steam to current, without a single atom of nuclear fuel involved anywhere in the chain.

Lead melts at 327 degrees Celsius, which means the entire system must be kept hot even during maintenance pauses. Getting it wrong means the coolant solidifies inside the vessel like concrete setting in a pipe. That physical reality is, by itself, one of the things the project is designed to stress-test at full scale.

A research center in the Italian hills becomes an unlikely stand-in for a commercial reactor hall

The ENEA Brasimone site has studied lead-cooled systems for decades, running smaller loop experiments and component tests. What changed with PRECURSOR is scale. The installation team seated the main reactor vessel on June 21, 2026, completing a milestone that moved the project from fabrication into commissioning. The vessel itself is only slightly smaller than the one the planned commercial reactor will need, which puts this far outside the category of tabletop demonstration.

Three separate support systems had to be built alongside the vessel: a melting tank to liquefy solid lead ingots, a storage vessel to hold the metal molten and chemically conditioned, and a transfer system to move it safely into the main pool. None of those pieces exist in any water-cooled plant, because water does not arrive as ingots and does not freeze if the temperature dips. Lead requires what amounts to an entirely separate logistics chain just to reach its working state.

The facility is expected to reach full operation before the end of 2026. Its completion would make it the only installation in the world to close the entire loop from lead-cooled heat production through steam generation to grid-connected electricity output, without any radioactive material on site. The company’s founder described it as potentially “the only facility in the world to do such a complete demonstration.”

The evidence for why this approach matters comes from what water-cooled plants cannot do

Lead-cooled fast reactors are one of six Generation IV designs that international energy agencies have designated as worth pursuing past the concept stage. Their appeal is specific. Molten lead does not pressurize as water does, so the primary circuit runs at near-atmospheric pressure, which removes the explosive rupture risk that drives the heaviest containment requirements in conventional plants. A fast neutron spectrum also means the reactor can consume certain forms of nuclear waste as fuel rather than generating more of it.

The efficiency case is equally direct. Water-cooled plants convert roughly 30 percent of their thermal output to electricity. Lead-cooled designs can push that figure above 40 percent in real operating conditions, because the higher operating temperature allows a more efficient thermodynamic cycle. That gap compounds across the lifetime of a commercial unit and shows up directly in the cost per unit of power delivered.

But lead is corrosive to steel at high temperatures, and controlling that corrosion across decades of operation is exactly what no one has validated at the scale a commercial plant requires. PRECURSOR’s job is to run that corrosion chemistry, alongside pumps, heat exchangers and welds, for long enough to generate real data. You can see internal links to related coverage of nuclear fuel back-end challenges and the recent progress of privately sited reactors reaching criticality in the United States.

The complication is that PRECURSOR proves the plumbing, not the neutronics

Running a reactor with electric heaters instead of fuel is genuinely useful, but it tests a specific subset of the engineering. Thermal hydraulics, corrosion chemistry and the power conversion chain all get real data. What does not get tested is the behavior of the fuel itself under fast-neutron bombardment, the interaction between lead and irradiated cladding materials, or the control dynamics of an actual fission chain reaction inside a lead pool. Those questions require a fueled machine.

The planned next step is the LFR-AS-30, a 30-MW commercial demonstrator the developer expects to begin operation in 2031. After that comes a 200-MW commercial unit, with a first US deployment targeted for 2032 according to company filings. Each step introduces the radioactive variables that PRECURSOR deliberately excludes. So the facility is a prerequisite, not a proof.

Even so, the engineering community has tended to treat lead-cooled designs as permanently a decade away, partly because no one had run molten lead through a full-scale power conversion loop long enough to find out what breaks first. PRECURSOR is the machine that will finally run that test, which makes the Italian hills near Bologna an unlikely dateline for one of the more consequential experiments in advanced reactor development.

What the vessel seated in the Apennines means for where lead-cooled power goes next

The commercial path depends on what PRECURSOR finds. If the corrosion data comes back within the modeled range, the 2031 target for the first commercial demonstrator becomes more credible. If the lead chemistry produces surprises, those surprises are far cheaper to resolve in a fuel-free facility than inside a licensed reactor with spent fuel on site. That is the honest logic behind building the full loop first and putting the uranium in second.

What the project has already settled is a narrower but real question: whether a full-scale lead-cooled vessel can be fabricated, transported to a research site and seated without the metal or the structure failing on the way in. The answer from the June installation is yes, it can. The vessel is in the ground, the turbine is wired to the steam circuit, and the lead ingots are waiting for the melting tank to begin its first heat cycle.

Whether the machine eventually leads to a commercial reactor or becomes a very expensive lesson in what molten lead does to steel over time, the answer will now be a measured one rather than a modeled guess, which is the most the industry could honestly ask of a machine with no fuel inside.