A 1 to 20 megawatt microreactor went critical in Idaho with no pumps and no moving parts in its cooling, while the record it actually set is narrower than the coverage suggests

A concrete bay at a national laboratory complex in eastern Idaho, early on a June morning.

Inside a rig roughly the size of a large household furnace, a chain reaction starts and holds.

There are no coolant pumps in it. There is no circulating loop and no forced flow of anything.

The fuel is high assay low enriched uranium in coated particle form, enriched to just under 20 percent, with each particle around three hundredths of an inch across.

The heat leaves the core through sealed metal tubes and nothing inside the machine turns.

Cooling here has no moving parts.

It runs on boiling sodium.

How a sealed tube moves heat better than a pump does

A heat pipe is a closed tube with a working fluid inside and a wick lining its wall, and that is the entire mechanism.

Heat at one end boils the fluid. The vapor rushes to the cold end because vapor moves toward lower pressure, which is where the condensing is happening.

At the cold end it gives up its latent heat and turns back to liquid, and capillary action in the wick draws that liquid back along the wall to the hot end.

The loop runs on phase change and surface tension. No pump, no valve, no electricity, and nothing that can be switched off by accident or fail to start.

In this reactor the fluid is sodium, chosen because it stays liquid across the temperature band a reactor core works in and carries heat extremely well.

Pumps can stop. Boiling cannot.

What went critical and what is inside it

The machine reached first criticality on the fourth of June at the Idaho site.

Microreactors in this class are aimed at 1 to 20 megawatts of thermal output, which is heat for a base or a remote site rather than power for a city.

The fuel is high assay low enriched uranium in coated particle form, enriched to just under 20 percent, with each particle around three hundredths of an inch across.

That particle is its own containment. The coatings around the kernel are designed to hold fission products in even if everything outside them fails.

The whole unit is small enough to move on a large truck, which is the entire point of the category.

Twenty percent enrichment, a grain of fuel.

The deadline that shaped the entire schedule

None of this speed is accidental, and the reason is a date rather than a breakthrough.

A federal pilot program set out to have several privately designed reactors reach criticality by the fourth of July, and the companies in it built to that calendar.

Four separate machines made it, at the same Idaho complex, inside a few weeks of each other. That is more new reactor designs going critical in one summer than the country managed in the previous four decades.

What changed was not the physics but the paperwork. Authorization ran through the energy department rather than the civilian regulator, which is a different and much faster route.

It is also a route that does not end in a commercial license, so none of these machines can sell power to anybody yet.

The bottleneck was never the reactor. It was the approval.

The claim that needs trimming

Coverage has framed this as the first genuinely new reactor design to go critical in more than half a century. That is not what the record says.

The accurate version is narrower in two ways. It is the first advanced reactor to reach criticality under a specific federal pilot program.

And it is the first new reactor in the United States that is not cooled by water to go critical in more than 40 years, which dates the gap to the early 1980s rather than the early 1970s.

Forty years is still a long silence and the milestone survives the correction. It is not fifty, and it is not every reactor design.

The same laboratory has another microreactor in fabrication, and a criticality date still ahead.

The criticality date, the fuel and the cooling method are described by an engineering title.

Forty years is the real gap, not fifty three.

What a critical reactor has and has not shown

Criticality means the physics works. The geometry, the fuel loading and the reflector all behave the way the model said they would.

It does not mean the machine makes useful power, and it does not mean it can hold output under thermal load for months without drifting.

Heat pipes have a failure mode nobody has tested at this scale in a reactor. Push one past its limit and the wick dries out at the hot end, and a dried pipe stops moving heat entirely.

Electricity generation is targeted for next year and a military deployment the year after, an aggressive schedule for a machine that went critical this summer.

The load these reactors are really chasing is industrial, which is why small reactors keep turning up beside data center plans.

The pilot program framing and the criticality sequence are set out by a nuclear service.

The physics is demonstrated. The duty cycle is not yet.