Business Week reports on a nuclear electric battery (more like a reactor) that has 27 MW for 5 years worth of juice and it's "the size of a hot tub". It's patent pending (20040062340; search for "uranium hydride over at USPTO.gov). That's about 64 cubic feet. That's 1.2 terrawatt hours or 1.2 billion kwh. They say it's 70% cheaper than natural gas--maybe $30 million? If it's 54 cubic feet and pure uranium hydride (a high overestimate), it would weigh about 15 tons. Compare that with 15 tons of LOX and hydrogen with 66,000 kwh at 39 kwh per kg of hydrogen. Pretty good ISP. Thrust to weight not so good. Combine it with a reaction mass fill up on Mars?

One proposal for an artificial-gravity Mars mission used Nuclear Electric Propulsion. The rocket required 9 megawatts of electric power and the mission had a travel time of 1 year to Mars including the slow propulsive braking into low Mars orbit.

Using your estimate of the reactor mass a scaled down system adequate to power the AGNEP spacecraft would mass less than 2,000 kg. Not bad.

This looks like the same reactor Brian Wang was talking about on his Advanced Nanotechnology blog. If so he's got a pretty good round up here:

http://advancednano.++UNGOODURL.com/2007/11/nuclear-battery-can-be-used-to-help.html

and a follow-up with er...transterrestrial applications here:
http://advancednano.++UNGOODURL.com/2007/11/vasimr-engines-plus-200-mw-of-nuclear.html

where ++UNGOODURL is bl#g$pot

It is neat that it is going forward. Toshiba has a head start on small dispersed reactors and I don't THINK these guys have built one yet. I understand the little Toshiba reactor in Galena (which is much bigger than this) is actually on track for 2010.

How feasable IS this reactor? (Technically... leaving aside the regulatory hurdles and luddite NIMBY's )

As an aside, I've got a rant on (mostly) small reactors here:
http://brickmuppet.mee.nu/atomic_get

I think they'd be foolish to call it anything but a battery. Batteries are PC. Maybe this device will find its way into use before the Luddites and NIMBYs catch on. (Well, I can dream.)

We will need some sort of power supply on mars to turn over the life support and generate fuel for launching back off the surface with anything more massive than a tinfoil escape pod.

I've often been of the opinion that in the long run, to go anywhere in space and do anything interesting, we'll have to go nuclear in a big way.

Interesting - I read the patent. Figure a critical reactor with half the volume taken up with heat pipes at an output of 50 MW(thermal) and a weight of around 7500 kg. I'm assuming that the junk you have to put around the core to shield and extract power weighs about as much as the core itself, which is probably naive. Conversion efficiencies are anyone's guess, but if you could produce 5MW (electric) that'd be reasonably decent at ten percent.

It's bigger than you'd need for Mars though - at least for housekeeping and such when you got there. To make a smaller one, if I read the patent correctly, you'd need hotter uranium - maybe close to bomb-grade. Otherwise it won't go critical.

Fun idea. But....If you sell electricity at 4 cents a kWh, and run one of these babies for five years, then you've got (5000 kW * 24 h * 365 * 5 * .04 $) = 8.7 million bucks. Is that enough to cover all the costs? Initially, I'd be a little skeptical. But maybe in volume it'd be okay.

Jane: GE can convert natural gas heat to electricity at 60% efficiency. I think they are assuming about 54% if your assumption about 50 MW heat output is correct. In any case, if the efficiency is lower, that means the heat output is higher because their press says 27 MW electric output. I calculated the price of the reactor starting from their claimed output and the price of natural gas power and their claimed 70% savings.

Sam, I don't perceive that their reactor gets nearly as hot as a gas turbine combustor - barely 800C at the high end, and probably less. So I think 54% is optimistic. But I'll grant that my 10% number is too low. So maybe instead of 9M revenue per unit it might be more like 25-30M. That surely helps. If they get to build the factory they're talking about then perhaps it'll be profitable. 4000 reactors at 25M of revenue apiece provides a gross revenue of 100B. That's a piece of change worth going after.

This whole thing comes apart the first time some MSM nut says, Chernobyl, Three Mile Island, or China Syndrome.

The American public has no one to blame but themselves for high electricity prices and dwindling supplies. They let the MSM sell them a bill of goods regarding nuclear power. And this isn't just my opinion, I used to be in nuclear construction. I lost a career to idiots who thought every reactor in the country was 10 seconds from being a homegrown Hiroshima.