I’ve always been fascinated by the thermocouple and its ability to generate electricity from heat. I’ve often wondered if we could put millions of thermocouples into places where heat is a byproduct of some other operation and capture it as electricity. That may soon happen.
From light.sci-toys.com: In 1821, physicist Thomas Johan Seebeck discovered an interesting effect when he heated a junction of two different metals. He found that they generated an electric current. Actually, he thought it was a magnetic effect, because what he had noticed was that a compass needle was deflected as the current flowed through the wire.
Today, we call the thermoelectric effect he discovered by his name — the Seebeck effect. He must have had a very sensitive compass, because the currents that are generated are very small. Unless the wires are formed into a coil around the compass, to increase the magnetism of the current carrying wire, it would be difficult to notice any movement.
But a modern voltmeter is sensitive enough to show the effect, as shown in the photo. Now that concept has literally taken a quantum leap.
From University of Arizona News – Quantum physicists turn waste heat into power
University of Arizona physicists have discovered a new way of harvesting waste heat and turning it into electrical power. Taking advantage of quantum effects, the technology holds great promise for making cars, power plants, factories and solar panels more efficient.
What do a car engine, a power plant, a factory and a solar panel have in common? They all generate heat – a lot of which is wasted.
University of Arizona physicists have discovered a new way of harvesting waste heat and turning it into electrical power.
Using a theoretical model of a so-called molecular thermoelectric device, the technology holds great promise for making cars, power plants, factories and solar panels more efficient, to name a few possible applications. In addition, more efficient thermoelectric materials would make ozone-depleting chlorofluorocarbons, or CFCs, obsolete.
The research group led by Charles Stafford, associate professor of physics, published its findings in the September issue of the scientific journal, ACS Nano.
“Thermoelectricity makes it possible to cleanly convert heat directly into electrical energy in a device with no moving parts,” said lead author Justin Bergfield, a doctoral candidate in the UA College of Optical Sciences.
“Our colleagues in the field tell us they are pretty confident that the devices we have designed on the computer can be built with the characteristics that we see in our simulations.”
“We anticipate the thermoelectric voltage using our design to be about 100 times larger than what others have achieved in the lab,” Stafford added.
Catching the energy lost through waste heat has been on the wish list of engineers for a long time but, so far, a concept for replacing existing devices that is both more efficient and economically competitive has been lacking.
Unlike existing heat-conversion devices such as refrigerators and steam turbines, the devices of Bergfield and Stafford require no mechanics and no ozone-depleting chemicals. Instead, a rubber-like polymer sandwiched between two metals acting as electrodes can do the trick.
Car or factory exhaust pipes could be coated with the material, less than 1 millionth of an inch thick, to harvest energy otherwise lost as heat and generate electricity.
The physicists take advantage of the laws of quantum physics, a realm not typically tapped into when engineering power-generating technology. To the uninitiated, the laws of quantum physics appear to fly in the face of how things are “supposed” to behave.
The key to the technology lies in a quantum law physicists call wave-particle duality: Tiny objects such as electrons can behave either as a wave or as a particle.
“In a sense, an electron is like a red sports car,” Bergfield said. “The sports car is both a car and it’s red, just as the electron is both a particle and a wave. The two are properties of the same thing. Electrons are just less obvious to us than sports cars.”
Bergfield and Stafford discovered the potential for converting heat into electricity when they studied polyphenyl ethers, molecules that spontaneously aggregate into polymers, long chains of repeating units. The backbone of each polyphenyl ether molecule consists of a chain of benzene rings, which in turn are built from carbon atoms. The chain link structure of each molecule acts as a “molecular wire” through which electrons can travel.
“We had both worked with these molecules before and thought about using them for a thermoelectric device,” Bergfield said, “but we hadn’t really found anything special about them until Michelle Solis, an undergrad who worked on independent study in the lab, discovered that, low and behold, these things had a special feature.”
Using computer simulations, Bergfield then “grew” a forest of molecules sandwiched between two electrodes and exposed the array to a simulated heat source.
“As you increase the number of benzene rings in each molecule, you increase the power generated,” Bergfield said.
The secret to the molecules’ capability to turn heat into power lies in their structure: Like water reaching a fork in a river, the flow of electrons along the molecule is split in two once it encounters a benzene ring, with one flow of electrons following along each arm of the ring.
Bergfield designed the benzene ring circuit in such a way that in one path the electron is forced to travel a longer distance around the ring than the other. This causes the two electron waves to be out of phase once they reunite upon reaching the far side of the benzene ring. When the waves meet, they cancel each other out in a process known as quantum interference. When a temperature difference is placed across the circuit, this interruption in the flow of electric charge leads to the buildup of an electric potential – voltage – between the two electrodes.
Wave interference is a concept exploited by noise-cancelling headphones: Incoming sound waves are met with counter waves generated by the device, wiping out the offending noise.
“We are the first to harness the wave nature of the electron and develop a concept to turn it into usable energy,” Stafford said.
Analogous to solid state versus spinning hard drive type computer memory, the UA-designed thermoelectric devices require no moving parts. By design, they are self-contained, easier to manufacture and easier to maintain compared to currently available technology.
“You could just take a pair of metal electrodes and paint them with a single layer of these molecules,” Bergfield said. “That would give you a little sandwich that would act as your thermoelectric device. With a solid-state device you don’t need cooling agents, you don’t need liquid nitrogen shipments, and you don’t need to do a lot of maintenance.”
“You could say, instead of Freon gas, we use electron gas,” Stafford added.
“The effects we see are not unique to the molecules we used in our simulation,” Bergfield said. “Any quantum-scale device where you have a cancellation of electric charge will do the trick, as long as there is a temperature difference. The greater the temperature difference, the more power you can generate.”
Molecular thermoelectric devices could help solve an issue currently plaguing photovoltaic cells harvesting energy from sunlight.
“Solar panels get very hot and their efficiency goes down,” Stafford said. “You could harvest some of that heat and use it to generate additional electricity while simultaneously cooling the panel and making its own photovoltaic process more efficient.”
“With a very efficient thermoelectric device based on our design, you could power about 200 100-Watt light bulbs using the waste heat of an automobile,” he said. “Put another way, one could increase the car’s efficiency by well over 25 percent, which would be ideal for a hybrid since it already uses an electrical motor.”
So, next time you watch a red sports car zip by, think of the hidden power of the electron and how much more efficient that sports car could be with a thermoelectric device wrapped around its exhaust pipe.
Reference: Giant Thermoelectric Effect from Transmission Supernodes. Justin Bergfield, Michelle Solis, and Charles Stafford. ACS Nano Sept. 2010.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
AnonyMoose says:
September 29, 2010 at 2:27 pm
“Coat the radiator with this material also.”
Radiator might not be hot enough (typically 190F) to get any appreciable efficiency. If it is then we got a real winner. Holy grail of solar electric generation is to do it with water at or just over boiling point (low pressure) because it’s so cheap and easy to handle at that temperature & pressure (store it in an off the shelf hot-water heater) plus you can store a buttload of energy in a small volume due to latent heat of vaporization.
This seems to be an answer looking for a problem. The easiest way to extract the energy from a car’s exhaust system is to use current technology, to wit, an exhaust turbine commonly called a turbocharger. Instead of connecting the turbine to an air compressor connect it to an alternator, basically a small portable turbo-alternator electrical generation plant. I vaguely remember that the (non turbo) supercharger on a Mercedes McClaren (5.4 litre V8) requires over 200 bhp from the engine at full power, so an equivalent turbocharger, measured in output gas flow, should need about the same. 200bhp equals approximately 150 kilowatts, or 1500 100 watt lightbulbs.
By designing cars with a small ICE, to be used basically as a heat generator but connected to the drive wheels to use its 35% efficiency, passing the exhaust gases through an exhaust turbine connected to an alternator which then drives an additional electric drive motor (both at 90+% efficiency) most of the “waste” heat could be retrieved and usefully used. Even if only 1/2 of the waste heat could be extracted the overall efficiency of the engine could be increased to ~70% of the input energy. Simple and doable now without needing computer models to generate a possible hypothesis.
Are there obvious problems? Certainly, but these are merely simple engineering problems that our clever mechanical and electrical engineers could work on now, not wait for some possible breakthrough in “nano-technology”.
This article is a splendid demonstration of the strong force of scientific illiteracy- it’s just bloody thermocouple.
There is a lot of wasted energy in autos period. Much of it, as several pointed out, is low level heat energy that is hard to harvest.
One place being looked at is the car’s kinetic energy. Hitting the brakes and converting 4,200Lbs@60mph into undesirable waste heat via friction is primitive at best and an enormous waste of energy. High mass rotors harvesting the energy and air compression are being looked at, and the rotors seem promising.
This is an interesting theory. There could be a lot of uses for it if it overcomes a few challenges.
No kiddin’ I had a neighbor that let me in on his big secret. A generator and an electric motor plugged into each other. He was serious. He figured that as it picked up speed, he could harvest energy off of it. What? Do I want in on it?
I noticed a few comments about how the heat-difference drops to 0 at small distances and figured it would be best to clarify: The direction of the potential difference seems to be determined by the relative phases of the parts of the electron-wave. That, in turn, looks like it is determined by the direction in which the electron which interacts with this thermocouple is moving as it reaches the rings. Summing these potential-differences, that means that equal numbers of electrons from both sides will cancel, giving 0 V total. More heat on one side is faster random motion of particles on that side. This implies that those particles will hit the thermocouple more often.
Any given molecule will have a very small heat-difference across it, but that is to be expected: You don’t get much power out of a single molecule. Stacking them, though, you can get something. Still, I would really like to see the efficiency and price for these things.
From Richard on September 29, 2010 at 5:39 pm :
Except what makes turbines spin is the speed of the gases moving through them. Four stroke engines have a piston stroke just for pushing out the exhaust gases. Anything that restricts the exhaust flow will reduce the engine efficiency, which includes catalytic converters thus that’s an additional trade-off of efficiency for pollution control. Restrict the flow with a turbine and the engine will burn more fuel to power the exhaust stroke, creating more heat getting dumped at the radiator and a net efficiency loss.
George E. Smith says:
September 29, 2010 at 8:03 am
So the space probes are still working; that’s wonderful.
I think I have said several times here at WUWT, that using “electricity” to create “heat” should be a felony. [–snip rest for brevity–]
Well, I dunno, George.
Have you tried sucking on a piece of coal, coke, charcoal, wood, peat, or a hunk of dung lately? :o)
Don’t get me wrong here, but radiant heat panels are quite efficient over other means of heat energy production.
See this: http://www.sshcinc.com/
Now, rubbing two sticks together to create a fire is, well, using electricity to create heat. The electricity of your nervous system, which causes your muscles to contract and relax is using electricity to create heat.
Will you now submit yourself to the U.S. court system for being a felon in fact all these long years?
A more effective design is the new generation of Stirling engines. The reason they never caught on was because the pressure of the working fluid had to be low to avoid blowing the seals on the connecting rod power was transmitted through. The new designs use piston mounted magnets reciprocating inside cylinder mounted coils. No connecting rod. A combined heat and power unit for domestic used is being trialed by Eon in the UK. It uses gas to run the system, heating the home and additionally generating about 1kw which is fed back onto the grid, reducing the electricity bill. They claim a 25% saving over traditional combi-boilers.
The problem with adapting this to automotive use is the comparitively low temperature differential of most of the heat coming off the engine, as pointed out by several people on this thread.
George E. Smith says:
September 29, 2010 at 8:03 am
So yes going either way from electricity to heat or back is stupid, and should be a felony
Since all methods of converting traditional fuels to electricity use that method, it sounds like you are advocating wind power and PV.
😉
Buy your stirling engine here (and thermoelectric fans etc)
http://gyroscope.com/StirlingEngines/
kadaka (KD Knoebel)
Thank you for your reply. I find that there are a number, of umm, misapprehensions in your post.
1. The primary source of scavenging in a four stroke ICE is the release of high pressure when the exhaust valve opens. this also generates a shock wave that continues to extract the combustion products even when the pressure has dropped significantly. The exhaust strokehelps to a small degree by allowing the exhaust valve to remain open longer, and the to permit more complete filling of the chamber during the induction stroke. 2 stroke engines work successfully without an exhaust stroke, albeit with reduced efficiency. 2 stroke engines can also be turbo-charged quite successfully.
2. Efficiency reduction through greater back pressure in the exhaust system. A standard turbo-charger does this already, and all it does is permit greater power to be produced from a very inefficient power plant. The overall efficiency is not increased but I would hazard a guess it is slightly (somewhat) reduced.
3. Efficiency reduction of some smallish amount is immaterial provided that the extra power produced by the turbo-alternator/electric motor combination is significantly more e.g. lose 10bhp but gain 20bhp, a net gain of 10bhp from the same amount of fuel. Incidentally CO2 reduction is a distinct possibility }:-)
The use of heat engines to power turbines in their exhaust is already commonly done e.g. the auxiliary power units in large aircraft and the related so called turbo-prop engine. It is also done on a larger scale in gas turbine electrical generating plants and for (primarily military) ship propulsion systems. Even the “pure” jet engine needs a power turbine in the exhaust to drive the compressors and these turbines become larger and larger as the by-pass ratio increases as in fan-jet engines.
A lot of people sound like they know what they’re talking about in these comments but no one has addressed my simple question, so I’ll try again. Repeating:
Rod Everson says:
September 29, 2010 at 1:50 pm
oeman50 says:
September 29, 2010 at 12:28 pm
Yes, the driving force for any thermocouple is the delta T between the hot and cold junctions, so even thermocouples have to obey the Second law of Thermodynamics. It is implied from the way this article is written we can just directly convert any heat that is lying around into electricity, not true.
Rod asks: I don’t understand. Are we talking about two different processes here? Look at the picture at the beginning of the article where a candle is directly heating the junction between two metals and is creating a voltage difference. That doesn’t seem to hinge on “the delta T between the hot and cold junctions.” Is the Seebeck Effect different from a thermocouple? Can someone clarify?
“”” Rod Everson says:
September 30, 2010 at 8:01 am
A lot of people sound like they know what they’re talking about in these comments but no one has addressed my simple question, so I’ll try again. Repeating: “””
Rod, If you have two different metals welded together to make one of these “thermo electric junctions”, you of course have to have a complete circuit to get any current flow. I’m sure that is self evident to you.
So if you have a complete circuit; and you also have a junction between tow DIFFERENT metals; common sense sys that you must have AT LEAST ONE other junction between two different metals.
Simplest case would be two wires, say iron, and copper welded together at two different junctions.
A current will flow; if, and only if, those two junctions are at DIFFERENT TEMPERATURES.
So yes the Seebeck effect can convert the DIFFERENCE between two juntion Temperatures to electric power.
Now if their are more than two differnt metals in the circuit, then the resulting current flow will depend on the Temperature of each junction.
If all junctions in the circuit have the exact same Temperature, then the sum of the thermal EMFs around the circuit is zero.
And no the seebeck effect is NOT different from a thermocouple.
Here’s a spin on this I haven’t seen here yet:
Can these be used to eliminate waste-heat and cool down spacecraft? This gets me excited for two reasons:
First, keeping the temperature low is very important for satellite-based particle-detectors, and those directly impact on my work.
Second, I was looking through ideas for space-lift and travel, and one idea that really stuck was the Gaseous Nuclear Light Bulb engine. It looks workable, except for the waste-heat problem which looks like it would fry everyone and everything aboard. Mix that with the recent discovery of a possibly habitable Earth-like planet nearby and you can see my excitement about a possible way to cool down in space.
“”” tallbloke says:
September 30, 2010 at 1:33 am
George E. Smith says:
September 29, 2010 at 8:03 am
So yes going either way from electricity to heat or back is stupid, and should be a felony
Since all methods of converting traditional fuels to electricity use that method, it sounds like you are advocating wind power and PV.
😉 “””
Well hell no Tallbloke; but you have perhaps captured me in the act of exaggeration !
I plead that in the case of converting combustible fuels through inefficient heat engines to good clean real energy like electricity; the initial combustion Temperatures are so high compared to exhaust Temperatures, that even with less than the Carnot efficiency; a respectable conversion can take place.
Wind of course is actually a gas turbine engine, where sunlight provides the energy to heat the working fluid (air) over a gigantic volume, and direct a very small portion of it through the turbine blade to extract a minuscule amount of the original solar energy.
Well you look at the Temperature difference that may exist between say land and sea air masses, that cause winds; and you see how small the Carnot Efficiency is going to be.
PV in principle could convert 100% of solar energy to electricity; but as a practical matter; we don’t have enough choice of solar cell band gaps to actually do that.
Silicon has a band gap (300K) of 1.106 Volts versus 0.67 for Germanium. GaAs, is 1.47, GaN is 3.3 volts.
Einstein’s E.lambda product is 1.23980 Electron Volt Microns.
So a one micron wavelength Photon is not sufficient energy to activate a Silicon photo diode (solar cell) but GaAs would work.
So you need more than the band gap photon energy to get electricity; but unfortunately if you have more than the band gap photon energy; you don’t get any more electrons liberated, and the extra photon energy ultimately ends up as waste heat.
If my somewhat rusty memory serves, it has something to do with something called the Lamb Shift.
So you have to use a multijunction cell to extract energy from more of the solar spectrum; and even then you have a problem, because you have to effectively separate the correct parts of the spectrum and send them to the correct junction. It does you no good having a Gallium Nitride junction, if the shorter wavelength higher energy photons from the solar spectrum get stopped by some lower band gap material.
So MJPV cells are very tricky technology.
But I’ll let you go from heat to electric; so long as you are burning something that is hot enough. Steam turbines are very well developed technology.
But you understand how wasteful it is to go the other way; so electricity ought to be preserved for things electric; rather than cooking. Gas is much better for cooking than electric anyway. [George . . ]
Rod Everson says:
September 30, 2010 at 8:01 am
“I don’t understand. Are we talking about two different processes here? Look at the picture at the beginning of the article where a candle is directly heating the junction between two metals and is creating a voltage difference. That doesn’t seem to hinge on “the delta T between the hot and cold junctions.” Is the Seebeck Effect different from a thermocouple? Can someone clarify?”
To extract electrical power there has to be an electrical circuit with (at least) two junctions, one at a higher temperature (T1) than the other (T2). The delta T is T1-T2. The maximum possible thermodynamic efficiency is 1-T2/T1. So for a car radiator at 90C (363K) down to ambient 20C (293K) the theoretical maximum efficiency of recovery of the waste heat is 19%. Though small, this could nevertheless be a worthwhile fraction of the engine’s useful work. But there’s an obvious snag. The waste heat from this process (>81% of the original waste heat) still has to be got rid of somehow at this lower temperature. The radiator can’t do it; if it could it would already have run that much cooler, and the engine could have been made more efficient to start with, dumping its waste heat at that lower temperature. A thermopile can’t do anything that a mechanical heat engine couldn’t do already (though it might be able to do it more conveniently or more cheaply).
The suggestion being made here is that the hot and cold junctions could be at the top and bottom of a rather thin film (with the “forest” in between. Nothing wrong with this in itself, but what people are pointing out is that unless the “forest” has an astonishingly low thermal conductivity, such a thin film will have a very low thermal resistance and be unable to maintain more than a uselessly small temperature difference without a huge parasitic heat flow, which provides no useful power.
“”” 899 says:
September 30, 2010 at 12:49 am
George E. Smith says:
September 29, 2010 at 8:03 am
So the space probes are still working; that’s wonderful.
I think I have said several times here at WUWT, that using “electricity” to create “heat” should be a felony. [–snip rest for brevity–]
Well, I dunno, George.
Have you tried sucking on a piece of coal, coke, charcoal, wood, peat, or a hunk of dung lately? :o)
Don’t get me wrong here, but radiant heat panels are quite efficient over other means of heat energy production. “””
Well 899, you are completely missing the point. I AGREE with you that converting electricity to heat can be extremely high efficiency; the problem is that going the other way is not nearly as efficient so going back to heat crap from pristine electric is to be avoided if possible.
The less ELECTRICITY we WASTE by using it to make HEAT, the less EXPENSIVE ELECTRICITY we will NEED.
HEAT we already have PLENTY of; so lets use it as HEAT ! (whenever possible).
George E. Smith says:
September 29, 2010 at 11:29 am
“”” A C Osborn says:
September 29, 2010 at 10:08 am
Eric Gisin says:
September 29, 2010 at 7:55 am
David Socrates says:
September 29, 2010 at 11:38 am
I am surprised that nobody has yet commented that the above claim is complete and utter nonsense.
You both obviously didn’t bother reading my 1st post, which has a link to an article showing that BMW have worked on this.
http://www.i-sis.org.uk/harvestingWasteHeat.php
and as I said they are not the only ones, one Truck company has retrieved 1Kw of electrical power without any engine losses.
A C Osborn said on October 1, 2010 at 6:22 am
You … obviously didn’t bother reading my 1st post, which has a link to an article showing that BMW have worked on this. http://www.i-sis.org.uk/harvestingWasteHeat.php and as I said they are not the only ones, one Truck company has retrieved 1Kw of electrical power without any engine losses.
1kW of electrical power without any engine losses? Wow! … All I can do is to expand on what I have said briefly previously (September 29, 2010 at 11:38 am).
An ICE is a heat engine. It draws heat from a higher temperature source (burning fuel) and delivers heat to a lower temperature sink (the environment, primarily via the vehicle’s radiator and exhaust systems). In doing so it generates mechanical power. A thermopile is just another kind of heat engine. It too must draw heat from a higher temperature source and deliver heat to a lower temperature sink. In doing so it generates electrical power.
Neither of these heat engines nor any other heat engine that could ever be constructed can possibly ever be 100% efficient. This is not because of engineering limitations that might one day be overcome. It’s because the Second Law of Thermodynamics limits the % efficiency of a heat engine to an absolute maximum given by the very simple formula 100 x (T2-T1)/T2 where T2 is the higher (source) temperature and T1 is the lower (sink) temperature, both temperatures being measured in degrees Kelvin. You can see immediately from this formula that a heat engine could only be 100% efficient at converting heat to mechanical energy if T1 was 0 degrees Kelvin. This may be very nearly achievable in space but obviously not in the world of transportation!
For an ICE in our real world, the typical upper temperature T2 is around 1000K (burning fuel) and the lower exhaust gas temperature is around 350K. Putting those two numbers into the above formula gives a maximum theoretical efficiency for such an engine of 100 x (1000-350)/1000 = 65%.
Now let’s examine the proposal to use a thermopile as the heat sink of an ICE so as to tap in to the ICE’s ‘wasted’ heat and increase the overall efficiency of the vehicle’s power system. Yes of course the thermopile will generate some additional (electric) power but (and it’s a huge ‘but’) it has only got the much lower temperature waste heat from the ICE to play with, the temperature of its heat source being the ICE’s sink temperature (only 350K in the above example). If the thermopile sinks heat to an ambient air temperature of 288K (15degC), the maximum theoretical efficiency of the thermopile would be 100 x (350-288)/350 = 18%.
So in this ideal world of a perfect ICE delivering 65% of the input fuel’s power and a perfect thermopile delivering a further 18% there would indeed be a useful gain in engine system efficiency. But in the real world right now we have an ICE efficiency of around 33% (due to limitations of materials and compromises in construction, making the engine as light as possible and minimizing cost) and a thermopile efficiency of around (1% to 3%). On the basis of this analysis, it would almost certainly be better to work on ICE improvement to get the ICE nearer to its theoretical maximum and this is, of course, exactly what the world’s manufacturers are doing all the time.
Talk of thermocouples is just plain marketing hype centered around the ‘green agenda’. I guess it must be doing wonders for vehicle manufacturers’ sales.