Laser weapons, advanced radar arrays, electronic warfare systems and a constantly-evolving web of data feeds pouring into field command operation centers may be changing the ways we fight wars, but the methodology we use to power all the tech hasn’t changed much since the early days of digitizing warfare. Diesel generators, often dependent on convoy-heavy supply lines, continue to provide the vast majority of America’s military needs in austere environments, but the Army has a plan to change all that.

Called “Project Dilithium,” the Army has been seeking proposals to build small, transportable nuclear reactors capable of sustaining field command elements for up to three years without the need for refueling. According to the Army’s specifications, these “portable” reactors “should be less than 40 tons total weight; small enough to be transported by truck, ship, and aircraft; able to run for at least three years without refueling; and capable of semi-autonomous operation.”

Those requirements are far from insignificant. While nuclear energy is widely seen as a largely safe enterprise under the proper supervision, most experts would agree that the technology for a self-sustaining or semi-autonomous nuclear reactor isn’t quite there yet. Assuming, however, these technical issues are worked out through the research and development process, Project Dilithium still has a pretty big problem to contend with: planting a largely stationary nuclear reactor in the middle of U.S. military bases in combat environments just sounds like a recipe for disaster.

In a recent piece published by the “Bulletin of Atomic Scientists,” the same group responsible for maintaining the nuclear “doomsday” clock that regularly makes headlines, the group pointed out that a failure to properly cool a nuclear reactor would inevitably lead to a meltdown that would put American troops at immediate risk. The Army’s requested requirements, however, call for a design that would render a meltdown all but impossible under most critical failure scenarios. The Bulletin contends that would be nearly impossible to deliver. That leaves the question, is the Army so desperate for a better source of energy that it would be willing to forgo some of these safety requirements?

It’s also important to consider that the current model of power production is also not a safe one. A 2009 Army study found that an average of one U.S. soldier is killed for every 24 convoy fuel deliveries. Considering fuel consumption for the sake of energy production hit 900,000 barrels in the same span of time, one can see the need for a new source of reliable energy. The Bulletin doesn’t propose a different solution, so much as contend that nuclear power should not be that source.

Edwin Lyman, a senior scientist and acting director of the Nuclear Safety Project at the Union of Concerned Scientists, wrote:

An energy startup in Oregon may have just changed everything about nuclear power

Read Next: An energy startup in Oregon may have just changed everything about nuclear power

An operating nuclear reactor is essentially a can filled with concentrated radioactive material, including some highly volatile radionuclides, under conditions of high pressure and/or temperature. Even a reactor as small as 1 megawatt-electric would contain a large quantity of highly radioactive, long-lived isotopes such as cesium-137—a potential dirty bomb far bigger than the medical radiation sources that have caused much concern among security experts.”

The Army’s requirements call for a minimum of a one megawatt reactor, but the request goes up to ten megawatts.

Of course, nuclear reactors aren’t unheard of in combat applications. America’s fleets of aircraft carriers and submarines are notably nuclear powered — allowing for unparalleled force projection capabilities. They are also, however, constantly moving and rarely operate within any proximity to indigenous populations that could suffer the long-term effects of a meltdown in an Army reactor placed in areas American troops consider to be an austere environment.

Whether Project Dilithium manages to make it off the drawing board will ultimately come down to two simple things: can someone meet the Army’s criteria, and will the Army be willing to budge on its demand for safety?