Car engines generate plenty of heat. According to Technology Review, researchers at MIT have a bright idea: use “thermophotovoltaics” to convert heat into light, then convert the light into electricity. The prototype system uses gasoline to heat tungsten to illuminate a photovoltaic cell to generate electricity. The hope is that systems based on this research could eventually replace the current gas-wasting alternators and air conditioning compressors. Of course, this technology hasn’t escaped the attention of the Department of Defence. "The military has had a lot of interest in it for portable power supplies in the field. Because there are no moving parts, there wouldn't be any noise, so you couldn't detect it," says NASA researcher Donald Chubb. It’ll be a few years before we see any practical applications, but as one of the key research sponsors, Toyota would be the first automaker in line.
Find Reviews by Make:
Read all comments
The company I used to work for makes thermoelectric devices (they refer to them as “coolers” since that is most of what they do). These devices use the Peltier effect; you pass a current through the device and it transfers heat.
The converse is true as well; if you heat the device, it generates current. They are working with some automakers to research the feasibility of using this type of setup to power some of the systems in a vehicle. I don’t know where the research is now, but it is interesting…
TexasAg03: The inverse of a Peltier device is a Thermocouple. You are right that when you heat a thermocouple, it generates current. However, it’s not some magic thing that generate current out of heat, it requires a temperature difference. The current generated in a thermocouple is proportional to the difference in temperature between two separate junctions (http://en.wikipedia.org/wiki/Thermocouple). So, that’s great if you get one junciton of the thermocouple in the engine block, and the other junction out in front of the radiator. You’ll have one that’s hot and another that’s cold. The trouble is that you don’t get much current for the temperature difference, so you couldn’t replace an alternator.
The same thing applies in the above article. What they’re doing is heating tungsten to get light to convert back into electricity. Again, you’re not going to be getting much current from this because the temperatures aren’t very high. They say they’re heating tungsten with gasoline (~2400 degrees), whereas you’ll never (safely) see those temperatures in the engine block where you’re trying to get the heat. Even exhaust temperatures are “only” about ~1000 degrees.
One of the issues, is that the OPTIMUM efficiency you could ever get out of any “heat engine” (i.e. something that converts heat into work) is 1-(T_cold)/(T_hot) (http://en.wikipedia.org/wiki/Heat_Engine#Efficiency). So the key is to get the heat from the hottest place you can, leave it as close to cold as you can get.
So here’s an example: A turbo diesel (where I know the data). EGT pre turbo is 1200 degrees, EGT post turbo is about 900 degrees. In absolute units (kelvin) that’s 922 and 755 degrees, respectively. So the theoretical maximum efficecy of your turbo is 1-(755/922) = .18 = 18%. That means that your turbo is 18% efficient converting heat into work (that’s better than an NA car which is 0% efficient).
The trouble is that the High and Low temperatures aren’t that far apart, so these fancy new things aren’t going to be able to be very efficient at converting heat into work (electricity in the case of this article), so you won’t be getting that much electricity. Which makes it really hard to use it as a replacement for an alternator.
One thing I remember a physics professor telling me is that there’s a big difference between “Hot heat” and “Cold heat”. The discussion was started when I asked him why we don’t recover the heat from brakes when we stop and use that as a heat engine to get more power to the car (Sort of like what hybrids try to do with regenerative breaking). And his response was that the temperature of the brakes wasn’t high enough compared to the surroundings to do anything. The Carnot efficiency of converting brake heat into useful work is to low.
These heat -> electricity converters have been around for a long time, but they’re really only good in low-power applications. There are fundamental physical limits (the Carnot cycle) that prevent us from having them in high-power applications.
I’m sure the guys at MIT are aware of thermoelectric devices – they also know that they are miserably inefficient, and that the low-grade waste heat from the engine is not suitable for significant power generation. That’s why they’re focusing on this clever method to make electricity directly from fuel so that you can have electrical power even if the engine is off without relying on a large battery. If it works out, size and efficiency wise, it would solve some problems. Maybe they could even divert some of the light via fiber optics to use for lighting.
Any object with a temperature higher than 0 Kelvin radiates electromagnetic waves. Visible light is but a small spectrum of that continuum. If the researchers can, instead of using a photovoltaic cell, develop another semiconductor that is excitable by radiation of the infrared wavelengths, then it would not be necessary to first convert waste heat into light for it then to be converted into electricity. The bandgap of such a material would be equal to the frequency of the infrared radiation X Plank’s constant. I’m not sure if such a material exists (about 1.2eV).
TonyH, the Blackbody radiation (http://en.wikipedia.org/wiki/Blackbody) of which you speak is only valid when the objects are in both thermal and radiative equilibrium. Something sitting around at room temperature, is surely radiating radiation in a proportion to what Plank said, but at the same time it’s absorbing an equal amount of radiation from it’s surroundings. If the absorption and radiation are not in equilibrium then the object will either heat up or cool down. So if you had a semiconductor that would absorb in the 1.2eV range (and turn it into usable current), the stuff emitting the radiation would cool down since you’re using the radiation for something, and then it’d be farther down in the IR and you’d need a new semiconductor with a smaller band gap…and so on.
Also, we are talking about electronic transitions (otherwise we wouldn’t be generating electricity), and electronic transitions are by nature high energy transition, which requires short wavelength light (visible, and more commonly UV). There are only a few semiconductors in the IR (InGaAs, and the family: http://en.wikipedia.org/wiki/Indium_gallium_arsenide) that are in the IR, but they’re expensive and inefficient.
People I work with are working on understanding the absorption and emission of quantum dots (http://en.wikipedia.org/wiki/Quantum_dots) which have quite a range of tunable wavelengths, and fairly high quantum yields (efficiencies), but we’re still a long way off from having them used in electricity production on any scale.
How many letters do you guys have after your names?
Good discussion!
I don’t see the efficiency of the heat/light/electricity equation — why not use a specially designed catalytic converter (with a built-in thermopile) to generate electricity (Deep space probes power their t’piles with plutonium). You would be utilizing the existing waste heat generated — just add a little fuel/O2 to keep thermal equilibrium.
By the same token, you could use a similar cat to heat a fluid to drive an air conditioning system.
Of course, these things would have already been done in a “rational” world; the limited future of petrol makes it a wasted engineering investment.
The Department of Defense (AFRL & NASA) have been working on this for quite a while to replace the radioisotope thermoelectic generators they’ve been using.
shaker, that’s a decent idea to get heat, but thermopiles need a cold side to generate power. On a space probe, you have a whopper of a cold side to use (the outside). Underneath a car, it’s a lot harder to generate a decent temperature differential. As for using heat to drive an AC system, which way to you propose to do it? If you use the heat to vaporize a fluid to spin a turbine that turns the AC compressor, you’re adding too much complexity to an already cramped space. If you use the heat to drive an absorption-type refrigeration system, you not only have the complexity of all that plumbing (check out the back of an RV refrigerator sometime), but you’ve also got much lower efficiency than a vapor compression cycle as well as glacial response times.
I know it is probably a dumb, inefficient idea, but could you mount a small Stirling cycle engine on the exhaust manifold and use it to generate electricity?
What about a small turbine in the exhaust flow? Would it create too much back pressure for the engine?
I appreciate MIT’s research into these types of projects, but really, this reminds me of NASA’s project to make a space pen. Seems some people make things harder than they need to be.
Along the same lines of NeonCat93’s thinking,I was thinking add a turbine in the exhaust that spins a magnet inside a coil. Instant free electricity.
Low-tech, but proven concept and low maintenance. Or is that too easy?
I know it is probably a dumb, inefficient idea, but could you mount a small Stirling cycle engine on the exhaust manifold and use it to generate electricity?
Just like shaker’s idea and bolhuijo’s responce, for a Stirling engine (or any engine) to work, you need to have a hot side and a cold side. In an IC engine, the hot side is the explosion the pushed the piston down, and the cold side is the exhaust. What you’re trying to do is use the exhaust as your hot side and the atmosphere as your cold side. There’s just not much of a temperature difference there to actually get any useful work.
Along the same lines of NeonCat93’s thinking,I was thinking add a turbine in the exhaust that spins a magnet inside a coil. Instant free electricity.
It ain’t free. That would increase back pressure which would make the engine work harder. You never get anything for free. Even taking heat out of the exhaust to run a Stirling engine would not be free. The fact that the exhaust gasses are hot makes them flow easier and have less back pressure, cooling the gasses would put more strain on the motor.
One thing people don’t realize is that (most) engineers and scientists are actually pretty smart and have thought of most of these ideas. The reason that many of the ideas are not in place is because they don’t actually help. Believe me, especially with such high gas prices, if any car company could instantly increase efficiency with “simple” ideas, you better bet they’d be doing it.
These problems are actually quite hard, that’s why it takes teams of smart people to solve them.
Beautifully said, miked.
“There is no free lunch.” – Robert Heinlein