On This Page
The Problem With Heat
Now in some subsequent pages in this website, there will be unpalatable scientific truths. But of them all, there are two in particular that give me more hate mail than all the rest put together. The fact that there ain't no stealth in space, and the heat radiator "Achilles Heel".
Both problems are caused by heat. In one case, it is the waste heat of a spacecraft that becomes a stealth-destroying beacon. In the other, it is the fragile nature of heat radiators.
About once a month, I get an email from some eager fan who believes they have the miracle solution to both problems. "Eureka! I've got it! All you have to do is use the Peltier-Seebeck effect to convert the heat into electricity! Then you can store it or use it as is convenient. A simple thermocouple and the problem is gone!"
This solution is elegant, marvelous -- and totally wrong. People who come up with this are to be congratulated on their brainpower, but they need to study their physics a bit closer.
What's the problem? Well, the general problem is that pesky Second Law of Thermodynamics. In this context, it tells you that it is impossible to destroy heat, the best you can do is move it around. So using a thermocouple to convert heat into electricity is impossible.
The specific problem is that a thermocouple does NOT convert heat into electricity. It converts a heat gradient into electricity. The original heat is still there. In fact, the conversion process adds even more waste heat to the original total.
But don't take just my word for it:
Using useful energy creates heat. You can't get rid of this heat; you can only move it around. You can in some cases use differences in temperature (i.e., heat) to generate useful energy, but it also generates even more heat. That can be helpful if you have somewhere to dump the heat, but if your problem is that you have a lot of heat and need to get rid of it, that doesn't fix your problem; it only makes it worse.
Peltier devices only make electricity by using a heat gradient. They get power by moving heat from an area of high heat to an area of low heat. They don't get rid of heat, they just move it around. Once the gradient is collapsed, no more electricty can be generated.
The thermoelectric effect converts a temperature difference into electricity; like all heat engines, it does not work if no temperature difference exists. In addition, like all heat engines, only a portion of the heat is converted into energy; the remaining heat is conducted from the heat source to the heat sink; specifically, like all heat engines, the upper limit on conversion efficiency is equal to the temperature change divided by the initial temperature.
In order to keep producing power, you must keep the heat sink cooler than the heat source. To keep the heat sink cooler, you must either refrigerate it (which consumes more power than is generated by the thermoelectric effect) or you must run a radiator at the temperature of the heat sink. If the heat sink is reasonably cool, this radiator will be extremely large.
Heat, by definition, is energy that comes with associated entropy. It is a well verified physical law that entropy cannot be destroyed, it can be created but the only way to decrease your amount of entropy is to move it somewhere else. To avoid having entropy build up on your spacecraft until something breaks, you need to get rid of it by moving it off the craft. The thermoelectric effect (Seebeck effect, Peltier effect, Thompson effect) moves heat through a device -- meaning it moves both energy and entropy. Some of the energy is shunted aside as entropy-free electric work that you can use for doing other things, but the remaining energy still has at least as much (and probably more) entropy with it. You have just moved the problem around, not gotten rid of any of it. In fact, in doing this, you made it harder to get rid of that entropy.
To explain why requires a bit of simple math. If you transfer an amount of heat Q at constant temperature T, the entropy S that moves along with the heat is S=Q/T. If you take in the heat energy Q_h and entropy S at a high temperature T_h, you can get rid of the entropy at low temperature T_l with less associated heat energy Q_l = S * T_l < S * T_h = Q_h. All the remaining energy Q_h - Q_l is entropy free, and you can use it for other things. To get rid of the entropy now at T_l, however, you can only discharge it into something even colder (or do work to move it someplace warmer). This is how all heat engines work, from gas turbines to automotive piston engines to Peltier or Seebeck thermocouples.
On a spacecraft, the only way you can move energy and entropy off the spacecraft is by putting it in something else and ejecting that something else, or radiating it away as electromagnetic waves. Since trying to put all your entropy in matter and dumping it is wasteful for your mass budget, radiation is the usual trick. However, the required radiator area scales as 1/temperature to the fourth power. The colder you are storing your entropy, the larger the radiator you need to get rid of it, and the size increases very rapidly with decreasing temperature. Using a thermocouple to extract electricity means you end up with lower temperature entropy and thus need even larger radiators to get rid of it.
The joke version of the Three Laws of Thermodynamics are:
- You can't win, you can only break even.
- You can only break even at absolute zero.
- You can never reach absolute zero.