# What methods are used to extract energy from low energy heat sources?

Most of what I was able to find pointed towards thermoelectric generators which are very inefficient, stirling engines which have a low torque output at low temperatures, or generators that use the organic rankine cycle. What are other methods of getting power from a low temperature heat source? And are any of them currently viable or potentially viable in the near future?

• The vapour-compression cycle - aka heat pump and viable if the source of energy is abundant cf the cost of other sources of energy... Jun 24 '18 at 7:56
• Turboexpanders. You need to be able to boil and condense something and overcome pumping costs if you want to run closed loop. Jun 26 '18 at 22:46

All the approaches to designing a heat engine that runs off a relatively low-temperature heat source suffer from severe limits on their thermodynamic efficiency which cannot be circumvented by clever design. For any of them to become viable requires that energy costs rise to much higher levels than they are at present.

• What about recovering waste heat? Isn't there some research going on towards that end? Jun 25 '18 at 3:56
• certainly- so far the heat pump is the best way to make use of "low-grade" heat sources but the energy thus extracted is in the form of heat and not shaft horsepower. Nonetheless, it is very effective for that purpose. Jun 25 '18 at 4:15

On the Carnot Theorem, which is based on the Second Law of the Thermodynamics, the maximal effectivity of any heat engine is $1-\frac{T_C}{T_H}$. $T_C$ and $T_H$ are the temperatures of the cold and warm heat tanks.

For example, if we burn gas ($1000K$ burning temperature) in a $\approx 300K$ environment, then the theoretical maximum is $70\%$. If you want to power your watch with your body heat, then $T_H \approx 20C^\circ$, $T_C \approx 36C^\circ$, thus the maximal effectivity is $\approx 3\%$.

In the engineering practice, there is no try to extract too much energy from little temperature differences. Because it is practically impossible.

If it is unavoidible, then the rule of thumb is, that the thermomechanical conversions are the best (for example, Stirling-engine), but they have many disadvantage (cost, complexity, high and unpredictable fault rate of the moving parts). Here very complex (=costly) solutions aren't too far away from the theoretical limit. On obvious reasons, there is still intensive development here.

Thermochemical, thermoelectrical conversions are far away in effectivity, but they are simple and cheap with a better predictable wear rate.