Nuclear-Thermal: Our Squandered Opportunity

As promised, here is my not-to-lengthy whine about our idiotic abandonment of nuclear-thermal fifty years ago and my hope that we get it right this time around. But before we proceed further, let's do a quick flip through Rocket Propulsion 101.

Chemicals, either solid or liquid, power all of our current stable of rockets. The chemicals combine to produce expansion (for example 'burning' kerosene in oxygen), which is directed out the rear of the vehicle. The higher the combined volume and velocity of the expelled gases, the greater the thrust, and the greater the acceleration of the vehicle. Each combination of chemicals has a different efficiency, and we commonly measure that as ISP, or "specific impulse". For example, one kilogram of fuel that can generate one kilogram of thrust for 200 seconds has an ISP of 200. ISP and exhaust velocity are directly related; if you multiply the ISP by 9.82 (gravitational acceleration on Earth's surface) you get the exhaust velocity (in this case 200 * 9.82 = 1962 meters per second). Below is a list of commonly known rocket fuels and their ISPs.

Space Shuttle Solid Rocket Booster (Polybutadiene acrylonitrile) = 250

Saturn V First Stage (Kerosene/Liquid Oxygen) = 265

Russian RD-80 Vacuum (Kerosene/Liquid Oxygen) = 315

Falcon 9 Merlin Vacuum (Kerosene/Liquid Oxygen) = 350

Starship Raptor 2 Vacuum (Methane/Liquid Oxygen) = 380

Saturn V 2nd/3rd Stages (Liquid Hydrogen/Liquid Oxygen) = 420

But it almost didn't play out this way. Seventy years ago (yes, 70 with a 7!) the US military launched Project Rover to explore the possibility of putting a nuclear-thermal upper stage on Inter-Continental Ballistic Missiles (ICBMs). The Soviet Union was a serious adversary back in the early-to-mid 1950s, the USA didn't have rockets with sufficient "throw weight" to carry large, heavy nuclear bombs to all parts of the USSR, and nuclear-thermal had the potential to double, even triple, the ISP and hence substantially increase payloads. They debated the concept for several years before finally launching the project in 1957 at Los Alamos, New Mexico, to develop a nuclear rocket that could deliver 2,700 megawatts of power. It was very promising, but chemical rockets were getting better, and nuclear weapons were getting smaller, so that initial rocket morphed into a solution in search of a problem.

Then, just as politicians began openly considering pulling the plug on Project Rover, the Soviet Union launched a payload into space, and a year later they put a human into orbit. The race was on. Nuclear-thermal upper stages were a very promising option for the Apollo Program and development proceeded for several years under Project Kiwi (ground-testing, hence the flightless bird reference) and then NERVA (Nuclear Engine for Rocket Vehicle Application), which included the pumps, engines, etc to make an operating vehicle. The goal was to develop a 1000 MW (that's 1.3 million horsepower!) nuclear-thermal rocket with a thrust of 25 tons and an ISP of 825 or more. NASA had big plans for nuclear-thermal, including a moon base, a manned mission to Mars, missions to the outer planets, and a space tug for moving low-orbit spacecraft into higher orbits.

Drawing of a generic NERVA rocket

The NERVA program proceeded from 1964 until it was terminated by President Richard Nixon in 1973. A failure, one would think. But no, NERVA was a resounding success, with the program meeting or exceeding every stated goal. The final rocket, XE Prime, weighed 18 tons, generated 1137 megawatts of power and 27 tons of thrust, and had an ISP of 840+. NASA built six iterations of NERVA between 1964 and 1969 that all achieved similar or greater performance, ran at full power for over two hours, and shut down and restarted dozens of times. Here's a short list of the NERVAS that were built and a few of their specs.

Engine Date Tested Run time Power ISP

Nerva A2 September 1964 40 seconds 1096 MW 810

Nerva A3 April 1965 16.5 minutes 1093 MW 840+

NRX EST February 1966 3.8 minutes 1144 MW 840+

NRX A5 June 1966 9.7 minutes 1120 MW 840+

NRX A6 November 1967 60.4 minutes 1199 MW 870

XE Prime March 1969 28 minutes 1137 MW 840+

They built that last engine 55 years ago, and it would perform alongside any chemical rocket we have right now. In our last post, we considered sending a Crew Dragon/Dragon XL combo to Mars, a nominal 35-ton payload, with a Falcon 9 upper stage. Trotting out the Rocket Equation, we can see that if we replaced the methalox engines with a NERVA XE Prime it could deliver 35 tons at 4.85 kilometers per second using LNG as the propellant. In other words, it could deliver the same payload as the best chemical rocket we have today. (Note: We used methane as a propellant because liquid hydrogen is a pain in the butt to work with. It is very light, so you can't get much into a fuel tank, and it leaks out of just about everything. LNG has a 25% lower ISP, but it's easy to work with.)

https://www.omnicalculator.com/physics/ideal-rocket-equation

NERVA XE Prime exhaust velocity (LNG) = 6187 meters per second

Total propellant capacity (LNG is lighter than kerosene) = 63 tons

The pusher unit weighs 18 tons

Payload = 35 tons

So, the wet mass is 116 tons, dry mass is 53 tons.

That combination will deliver a delta-vee of 4.85 kilometers per second.

The "typical" delta-vee from LEO to LMO is 4.3 kilometers per second.

Imagine where we would be if we had carried on developing this technology instead of abandoning it fifty years ago.

The good news? DARPA (the US Defense Advanced Research Projects Agency) has contracted Lockeed-Martin, working in concert with BWX Technologies, to develop a nuclear-thermal rocket for the US Lunar Exploration Program and for future applications such as Mars, the outer planets, asteroids, etc. Details are sketchy, but this is very promising!

Drawing of a shiny new rocket as that's all we've got so far. Still...

Our next few posts will discuss options for getting to Mars more quickly, or more comfortably.

Thanks for being here!