Water-cooled lander Options

[tanjent]

There is a recent posting on Emily's Planetary Society blog, which must be Doug's because she's not there herself, although her name is the only name on it. The subject is using water to cool a long-lived surface probe on Venus. It sounds far more practical than any of the other proposals for landing giant atomic-powered refrigerators, or developing a whole new family of high-temperature semiconductors, etc.

But I didn't understand the whispered criticism to the effect that the Ekonomov paper assumed that the water would absorb heat only from the one watt of power driving the instrument package itself. I simply can't believe that he went to the podium and presented his model without taking into account the fact that the surface of Venus is a pretty hot place, and that the proposed probe would be absorbing the ambient heat. This is an interesting proposal and I would like to understand both the original calculation of 50 days to bring the water to a boil, and the cited flaw in the calculation. I too find it hard to believe that it would take 50 days to bring water to a boil on the Venusian surface, but where exactly is the error, and what remains after we correct it?

Doug is busy of course, but I hope he will find the time to address this when he returns, if someone else hasn't done so by then.

[helvick]

..The concept is based around a 60-centimeter-wide pressure vessel, that's thermally isolated from a larger vessel using a vacuum. Inside the vessel are all the electronics you need, insulated from the otuside using 100 kilograms of water..

It appears that what is being suggested is a very small electronics package surrounded by the water within the 60cm internal vessel. A 60cm diameter sphere would hold ~110kg of water so you probably could get 100kg of water + a few kg of smarts into such a package.

The assumption is that you can effectively thermally isolate this internal block from the outer shell where you have the ultra high temperature power components and the sensors. I assume the objection came from someone who had some idea about how hard it is to build a thermal isolation system that keep the heat transfer across a 400deg+ temperature gradient into a 60cm core at <<1 watt. to be fair it does seem a bit fanciful but I'm no thermal expert.

Apart from that the numbers don't seem to add up (to me at any rate).

100kg of water. Specific heat capacity of 4.2 kJ.kg^-1.K^-1. Say we are heating that from an initial (arbitrary) temperature of 300K (25C) to 600K. That requires 4.2 * 100 * 300 kiloJoules. That's 126000 kiloJoules.
A Joule is a watt per second, or a kiloJoule is 0.2777 watt hours so 126000 is 35000 watt hours. Divide by 24 and you get 1458 days. Maybe he's talking about Venusian days?

The good news though is that the error is all in the right direction. This means that you could support an electronics package that produced 30 watts of power instead and still last 50 days before you had to start venting coolant. The other number they talked about then actually makes sense too. The latent heat of vapourisation of water is 2260kJ.kg^-1. Conveniently enough 30watts for one day is close to that at 2592kJ so you lose about 1.1kg of coolant per day once boiloff has started.

So maybe the system is viable for an electronics package that generates 1 watt of heat and a thermal isolation system that leaks 29 watts? That may still be impossible but it's certainly a lot easier to build than a perfect thermal isolation system.