[api]

I don't know for certain but I'm fairly sure that the idea is that you don't activate the reactor until it's in space. Before a reactor is turned on the fuels are less radioactive. It's once you turn it on that radioactivity increases dramatically and you get all the nasty decay products and such.

So not zero but not as much as you might think.

Personally I don't like the idea. Environmental concerns are real, but those aside it's likely more expensive than multiple refueling flights with big conventional rockets. These would be expendable and very costly to research, develop, fuel, and launch, whereas for the same cost you could probably put stages in orbit and send fuel up to them with reusable tankers. Like hydrogen this is another example of NASA chasing the sexiness of high performance in a pure sense (high iSP etc.) without doing a total cost analysis.

In general SpaceX and Blue Origin have the right approach.

[Robotbeat]

Why is this downvoted? This is informative and (in my opinion) basically correct.

And while I agree that in the near term, refueling via chemical rockets is a far cheaper (and even higher performance) way of solving this problem, I do support the research because someday we'll want to go even beyond refueling of chemical rockets. When you get REALLY high transfer times between Earth and Mars, the higher Isp makes a significant difference.

To explain: Conventionally, it takes about 6-8 months to get to Mars. Nuclear thermal rockets can shorten this time for the same mass in LEO to like 3 or 4 months. HOWEVER, agreeing with what api said, you can get the same exact speedup by using refueling with conventional rockets (and aerocapture/braking/direct-entry). It increases the required mass in LEO, but if you have cheap (especially reusable) rockets, then cost to launch more mass to LEO is not a major factor compared to the cost of a nuclear thermal rocket. And this is exactly what SpaceX has proposed: (see slides 19 through 22) http://www.spacex.com/sites/spacex/files/making_life_multipl...

But the Isp (exhaust velocity) advantage is maintained. The rocket equation is exponential: mass full = (empty mass)*e^((mission delta-v)/(exhaust velocity))

So eventually, when mission delta v is much higher than exhaust velocity, the mass ratio explodes. So a factor of 2 improvement in Isp is worth the extra cost, even if you have reusable rockets. The exponential curve eventually beats even the cheap, brute-force approach, if you want transfer times of on the order of 1 month.

It's also the kind of work NASA should be doing. Private industry is doing a really good job reducing the cost to orbit, so NASA can focus on these longer-term problems.

[baybal2]

> Why is this downvoted? This is informative and (in my opinion) basically correct.

I once wrote that Chernobyl had no chance to explode in a nuclear explosion in rebuke to some guy called Moxie Marlinspike. I had -4 for the next few days on all my posts, and somebody even bothered to find my work email, and futilely tried to troll me and my colleagues into deleting my rebuke for a week.

"That" demographic is definitely there, and working in a "tech" occupation does not preclude a person from being a part to it these days.

[SketchySeaBeast]

> "The massive amount of energy produced by these reactors could be used to sustain human outposts on other worlds and cut the travel time to Mars in half.

>“Many space exploration problems require that high-density power be available at all times, and there is a class of such problems for which nuclear power is the preferred—if not the only— option,”

It seems that nuclear reactors has more utility than simple power to weight ratio.

[mlindner]

Nuclear reactors will be needed if you're going much further out than Earth/Mars. Solar power falls off rapidly as you go further out.