By Adam Minter / Bloomberg Opinion

What’s a government space agency like the National Aeronautics and Space Administration supposed to do if private companies like SpaceX get all the spacefaring glory? One option is to double down on investments in leading-edge advancements that may not pay off for years.

Super-fast and maneuverable nuclear-powered rocket engines are one such technology. Last month, NASA, partnering with the Defense Department, gave Lockheed Martin nearly $500 million to build and test one by 2027.

Without this collaboration, two things could be in jeopardy: NASA’s dream of putting boots down in more parts of the solar system and the U.S.’s upper hand in outer-space warfare.

For nearly a century, rockets have operated in a fundamentally similar manner: A tank stores fuel that, when ignited, spits out a nozzle at high speed and creates thrust. Problem is, anything that you might want to do with a spacecraft, such as maneuver toward Mars, requires lots of fuel, and because there are no gas stations in space (yet), a craft must carry as much fuel as its operators expect it will need for the duration of its mission.

That can be a lot: Nearly half the mass of GOES-U, the 5.5-ton weather satellite that NASA plans to launch next year, will be fuel. Last year, Canada’s Telesat announced it would take a big financial hit due to insufficient fuel in a key communication satellite.

Scientists have long recognized the need for more efficient alternatives.

In the 1950s, they came up with an explosive one: Use a small nuclear reactor to heat up a propellant, such as liquid hydrogen, to even higher temperatures than what can be achieved in a chemical rocket. Such an engine would be more than twice as efficient as a traditional rocket and much faster; in part because its engines can run nonstop for weeks, accelerating faster and faster. A chemical engine would simply burn out.

Nuclear thermal propulsion, or NTP, was actively researched by NASA and other government agencies until the early 1970s. Such rockets weren’t intended to be launch vehicles (an NTP system lacks the thrust to leave Earth’s surface); instead, an NTP rocket would be carried into space on a traditional rocket and operate there. Though no reactor was ever flown, there were many successful ground tests demonstrating the concept could work; on Earth. Safety concerns, especially over what might happen if such a rocket crashed back to Earth, and political pressures ended the program.

But NTP was never entirely forgotten, and in recent years, advancements in space technology have placed it on the agenda of civilian and military space authorities.

For NASA, the goal is Mars. The agency aims for a human mission to the Red Planet in the 2030s. Traditional rockets can reach Mars in as little as seven months, with a round-trip mission lasting perhaps two to three years. An efficient nuclear rocket could get astronauts to the planet in as little as 45 days under one scenario, boosting their well-being — psychologically and physically — and potentially enabling more frequent trips.

For the military, it’s a race against China and Russia. The U.S. operated spy satellites and other military spacecraft largely uncontested for decades. But in recent years, China and Russia have advanced their technologies and are actively taking steps to neutralize the U.S. space advantage via jammers, antisatellite weapons and other techniques. The military would like to move satellites out of the way, but those built with traditional technologies are either too slow or will run out of fuel if they are relocated too often. Those limitations won’t be significant factors with nuclear-powered engines.

So what’s the holdup? Nobody has ever tested a nuclear-powered rocket in space, and serious questions exist about how one would perform in extreme conditions. The public must also be assured that an accident during launch won’t result in an environmental and health catastrophe.

The good news is that such a rocket is easier to build in 2023 than it was the last time the U.S. tried. Materials science has advanced considerably, which should help engineer a system that can withstand space and a nuclear reactor’s heat. Likewise, modern computing power will allow complex reactor designs to be subjected to simulations and redesigns rapidly.

All of this sets the stage for the federal government’s new effort. And while success is far from guaranteed, with a little luck and continued funding and commitment from Congress, the partnership between NASA and the Defense Department will help the U.S. maintain and widen its lead in a new space-race era.

Adam Minter is a Bloomberg Opinion columnist covering Asia, technology and the environment. He is author, most recently, of “Secondhand: Travels in the New Global Garage Sale.”