Thermoelectric (TE) materials and devices, which directly convert temperature differences to electric voltage and vice versa.. Thermoelectric devices are based on the fact that when certain materials are heated, they generate a significant electrical voltage. Conversely, when a voltage is applied to them, they become hotter on one side, and colder on the other. The process works with a variety of materials, and especially well with semiconductors.
Now the energy system throws away vast amounts of energy as waste heat. The energy now lost as waste heat just from U.S. power generation exceeds the energy used by Japan for all purposes. And that doesn’t even include the massive amount of waste heat from much smaller scale engines, like those in your car, where some 80% of the fuel’s energy is lost.
TE materials might also play a role in improving the efficiency of photovoltaic cells, harnessing some of the sun’s heat as well as its light to make electricity. The key will be finding materials that have the right properties but are not too expensive to produce.
But [TE] always had one big drawback: it is very inefficient. The fundamental problem in creating efficient thermoelectric materials is that they need to be very good at conducting electricity, but not heat. That way, one end of the apparatus can get hot while the other remains cold, instead of the material quickly equalizing the temperature. In most materials, electrical and thermal conductivity go hand in hand. So researchers had to find ways of modifying materials to separate the two properties.
The key to making it more practical, Dresselhaus explains, was in creating engineered semiconductor materials in which tiny patterns have been created to alter the materials’ behavior. This might include embedding nanoscale particles or wires in a matrix of another material. These nanoscale structures — just a few billionths of a meter across — interfere with the flow of heat, while allowing electricity to flow freely.
“TE devices are solid-state heat engines. Unlike today’s air conditioners, which use two-phase fluids such as the standard refrigerant R-134A, TE devices use electrons as their working fluid.”,
++++
But we can’t beat Carnot.
Once you’ve turned primary energy from fuel into heat (as in any standard electric power generator) you have already lost roughly 60+% of that energy to a form that cannot be used for work. Thermo-electric devices simply cannot recover that irrecoverable 60+%, no matter what nanotech techniques you apply.
The thermodynamic efficiency of a thermal power plant is determined by the ratio between the temperatures of the two heat baths it has available, the high-temperature bath being from the primary fuel consumption (usually applied to make high temperature steam), and the low-temperature bath of coolant (usually ambient water). Your thermoelectric solution would have the same two heat baths available, and the same maximum efficiency. No getting around the basic thermodynamics.
Gas turbines can beat Carnot by turning some of that fuel energy into mechanical energy (the expanding burning gas turns the turbine) rather than heat. Fuel cells can beat Carnot by using electrochemical processes that avoid the thermal step altogether. But thermoelectrics depends on the thermal step, and so is ultimately not going to help, at least with major power generation.
Now, thermoelectrics can help in situations where you naturally have a high-temperature waste stream and you’re at too small a scale to do something like a steam turbine. For example, pulling heat from automobile exhaust, it could well make sense to do this. Small industrial facilities that generate waste heat similarly.
Here are the basic physical limits.
When a quantity E of primary fuel energy (say oil), is burned, the energy E becomes thermal energy of the combustion products, raising their temperature to some high value T_high. In order to extract useful work from this thermal energy E, you need some process that takes it from the T_high system and eventually exhausts the energy through a waste stream at some lower temperature T_low. The fraction of the energy E that can be extracted as work, W, is physically limited by the laws of thermodynamics according to the two temperatures:
W <= (1 – T_low/T_high) * E
or a maximum efficiency for the system of (1 – T_low/T_high). It doesn’t matter if the process is a steam boiler or a thermocouple, once you’ve burned the fuel to thermal energy the efficiency limit is the same. If T_low is around room temperature (300 K) and T_high is around 500 K (227 degrees C) the maximum efficiency you can possibly get is 40%. That is, as soon as your primary energy E has been burned to thermal energy at 500 K, 60% of that original energy is no longer available to perform work, and at least that 60% is bound to leave your plant as waste heat.
You can get more work out of the same primary energy by increasing T_high – that’s an element of several proposals for advanced nuclear power reactors, to run them at higher temperatures and thus extract more of the primary energy from the fission fuel. For coal you can do something similar by gasification, which allows more direct use of the combustion products (Kirk above is of course right that these don’t “beat Carnot”, they just beat the normal limits for steam turbine plants by making more direct use of the waste stream at its hottest). But there are capital and maintenance cost issues – high temperature operation is more dangerous and more likely to lead to corrosion and failure, so the efficiency improvement may not be worth the extra costs.
It’s possible that replacing steam turbines with thermo-electrics could be a cost effective solution, I’ve not seen the numbers on these systems at all. If they could couple to a higher-temperature part of the process they might even have some potential for efficiency improvements. But most of that 60% wasted by power plants is a simple consequence of burning fuel, and thermoelectrics can’t avoid the same fundamental physical limits.
If we’re talking about waste streams at not much above room temperature, the efficiency limits there are far lower – for a 40 C differential you couldn’t get more than 12% of the energy out as work (or electricity). Imagine a two-step process where the steam turbine extracts the energy at T_high of 500 K, leaving a waste stream at T_low of 340 K. Then add a thermo-electric element between the 340 K waste stream and a 300 K bath (a local river, say). Of the original primary energy E, at most 0.32 E can be extracted by the steam turbine, and 0.68 E goes into the waste stream. The maximum possible efficiency of the thermo-electric system is then a little under 12%, which applied to the 0.68 E waste stream gives us an additional 0.08 E, for a total of 0.40 E energy extracted. But that is exactly the same useful work we could have extracted from the system if we could have had the original process exhaust at 300 K instead of 340 K – all that’s required to do that is the exact same larger volume of coolant water that we’d need to keep the thermo-electric system cool.
The long and short of it is, there’s no way to trick the thermodynamics into letting you have more useful energy once you’ve burned the fuel; to improve efficiency you need to work on the front-end of the process, not the back-end.
– from climateprogress