Idaho's nuclear history

Idaho has a claim no other state can make: the first electricity ever produced by a nuclear reactor was made on its high desert. It also has the only fatal reactor accident in U.S. history. Both happened on the same federal reservation, and that reservation is still where the country tests new reactors. If you follow nuclear power in Idaho, this history is the local context your neighbors half remember.

The National Reactor Testing Station

In 1949 the Atomic Energy Commission picked a stretch of sagebrush plain in eastern Idaho, much of it a former Navy gunnery range, for a reactor testing site. It needed remoteness, a large aquifer and room to put reactors miles apart in case something went wrong. It was called the National Reactor Testing Station. INL's history page notes the site's name changed several times over the decades; it took the name Idaho National Laboratory in 2005. INL puts the site at 890 square miles, larger than Rhode Island.

INL says 55 reactors have been built and operated there since 1949. Most were experiments: prototypes for Navy submarines and aircraft carriers, safety tests designed to push reactors to failure, and early commercial concepts.

EBR-I and the four light bulbs

Experimental Breeder Reactor-I was a small sodium-potassium cooled fast reactor built to test a bold idea: a reactor that creates more fuel than it burns. On December 20, 1951, its heat drove a small generator that lit four 200-watt light bulbs. INL calls it the first reactor in the world to make usable amounts of electricity. The next day it powered the building.

EBR-I later confirmed that breeding works, producing more fissile plutonium than the uranium it consumed. On November 29, 1955, during a test of how the reactor responded to changes in coolant flow, part of its core melted. No injuries were reported, and the unexpected behavior was traced to fuel rods and their support plates bending as they heated, which taught designers a lesson about thermal feedback. The building is now a National Historic Landmark and museum open to visitors in summer.

A related first is often credited to EBR-I by mistake. On July 17, 1955, power from BORAX-III, a boiling water test reactor about half a mile away, lit the nearby town of Arco for about an hour. INL credits BORAX-III. The BORAX series also deliberately destroyed a reactor in 1954 to study what happens in a power excursion. That experiment fed directly into the design of commercial boiling water reactors.

SL-1, January 3, 1961

SL-1 was a small Army reactor, about 3 megawatts thermal, built as a prototype for remote military bases in places like Greenland. On the night of January 3, 1961, three young military operators were reassembling control rod drives after maintenance. The procedure called for lifting the central control rod about four inches to reconnect it.

Investigators later concluded the rod was pulled about 20 inches. With one rod out that far, the small core went prompt critical. In about four milliseconds, power spiked to nearly 20,000 megawatts. The fuel vaporized, the water flashed to steam, and the pressure wave drove the reactor vessel up more than nine feet. All three operators died: two at the scene, one shortly after. Their bodies were so contaminated that parts of them were buried as radioactive waste.

Why the rod was pulled so far has never been established. Theories range from a stuck rod freed with a jerk to deliberate action. The official findings did not settle it, and anyone describing the accident should say so.

The engineering consequence was clear. SL-1 depended too heavily on a single rod. Every U.S. reactor since has had to meet the "one stuck rod" criterion: the reactor must be able to shut down even if its single most powerful rod is stuck fully out. Procedures became far more detailed, and emergency radiation meters were built to read much higher levels.

The workhorse reactors

EBR-II deserves a second look, because today's advanced reactor companies cite it constantly. In April 1986, the same month as Chernobyl, operators deliberately cut power to EBR-II's cooling pumps without dropping the control rods. The reactor's physics and natural circulation brought power down safely on their own. Oklo's Aurora design is explicitly modeled on EBR-II, and its first fuel is meant to come partly from material recovered from EBR-II's used fuel.

The ATR, operating since 1967, has a cloverleaf-shaped core that lets different lobes run at different power levels at once, so many experiments share one reactor. INL describes the Navy as a key customer; the fuel and materials in today's submarine and carrier reactors were tested there.

MARVEL and the 2026 test reactor rush

MARVEL (Microreactor Applications Research Validation and Evaluation) is a DOE microreactor rated at 85 kilowatts thermal and about 10 kilowatts electric, cooled by a sodium-potassium alloy with natural circulation and fueled with HALEU in TRIGA-type uranium zirconium hydride fuel. INL's page puts installation in the TREAT facility starting in late 2026, initial criticality in 2027 and full power in 2028. Its job is to let companies test how a microreactor connects to microgrids, data centers and industrial heat users.

The bigger news in 2026 came from private companies. Under Executive Order 14301's goal of three test reactors critical by July 4, 2026, Antares reached zero-power criticality at INL on June 4, and Deployable Energy announced its criticality at INL on July 1, both under DOE's program. Aalo Atomics followed at INL on July 4. Valar Atomics did it at a Utah site on June 18. Oklo's 75 MWe Aurora is under construction at INL with DOE approval of its preliminary safety analysis in June 2026.

The waste side of the history

Decades of reactor testing, plus spent fuel from the Navy and other DOE programs, left Idaho holding high-level waste and spent fuel. In 1995 Idaho and the federal government signed a settlement agreement that limits fuel shipments into the state and sets a 2035 deadline to remove spent fuel. Lesson 7 covers it in detail, because it still shapes what INL can and cannot accept, including research quantities of commercial fuel for advanced reactor work.