Slashdot Mirror
A Flexible Way To Convert Waste Heat To Electricity (asianscientist.com)
A research group in Japan has developed an inexpensive, large-scale and flexible thermoelectric generator (FlexTEG) that has high mechanical reliability and can convert heat into electricity efficiently. The findings are published in the journal Advanced Materials Technologies.
From a report: Thermoelectric conversion is one of the most attractive techniques for converting low-temperature (150C or lower) waste heat into electric power. However, widespread adoption of this technology has been hampered by a lack of suitable packaging techniques for thermoelectric generation modules that can operate in the 100-150C range. In addition, the production cost of modules for generating power at room temperature was prohibitive.
In the present study, scientists at Osaka University, Japan, have developed a method to manufacture thermoelectric generation modules in a cost-efficient manner while preserving the conversion efficiency of the modules. They mounted small thermoelectric semiconductor chips on a flexible substrate and were able to achieve reliable and stable adhesion of the electrical contacts between the chips and the flexible substrate. They called their invention FlexTEG.
In the present study, scientists at Osaka University, Japan, have developed a method to manufacture thermoelectric generation modules in a cost-efficient manner while preserving the conversion efficiency of the modules. They mounted small thermoelectric semiconductor chips on a flexible substrate and were able to achieve reliable and stable adhesion of the electrical contacts between the chips and the flexible substrate. They called their invention FlexTEG.
No reason to do anything now :)
The industries I can see adopting this quickly:
Automotive industry: There is a continuing push to improve overall mileage of a car. If waste heat can be captured as electrical power, It will get adopted.
Power industry: These guys are already trying to up their efficiency and the competition is stiff. If the power industry can implement this it will get implemented real fast.
Industrial processes: These guys vent so much waste heat it isn't funny. But they do have space normally. I can see them intentionally re-tuning their discharge temperatures to take advantage of this. Many of these industries are (also) very competitive. If they can shave a couple tenths of a percent of their costs, they will do it.
I expect most people in the industries I mentioned will not immediately notice it. But one of those industries will see it and get it implemented. And then Ford or Chevy or Subaru will come out with a car that doesn't have an alternator in it. All of the electrical power is generated from a rebuilt radiator and exhaust system that recycles the power from the heat. A lot of alternators on a car can draw 1-2 hp. So that is now either top end power or additional MPG. Either way everyone else in the industry is now saying "how do we do that?"
I expect the same thing to occur in industrial process, only the average consumer will never hear of it. In metal refining their price per ton will drop by a penny or two consistently and everyone else will start asking how they managed to shave that much cost off without reducing anymore many power.
In either case, once this gets adopted in a particular industry, everyone in that industry will stampede to adopt it so they can stay competitive.
Architectural plans are like computer source code with a couple of differences: You only compile once.
Less than 2% if I read the article right. Which is almost 10x worse than other existing systems.
Sure it bends, but the efficiency is so crappy it's only useful for novelty applications.
How does the heat turn into electrons flowing. Can someone explain in straightforward terms how it happens?
Helped design a teg system awhile ago, that was a pain.
.. I think it's trivial, that its just a different version of the well known peltier effect based heater/cooler units you use when buying a cooler for keeping beverages cool when underway.
Because of the "250 pân" pairs from the abstract this is how peltier cooler/heater pads are built up, and I think you can already get around 0,8 % effeciency with commonly availiable off the shelf components.
You can use these modules yourself to extract electric energy from a temperature differential.
> A lot of alternators on a car can draw 1-2 hp. So that is now either top end power or additional MPG.
You're correct there. The alternator needs to be able to produce about 1 HP of electricity. At 1.8% conversion efficiency, a TEG would need 56 horsepower (42,000 watts) of heat in order to generate enough electricity, in the lab. Do you think your engine wastes 56 HP as tailpipe heat while cruising around? Even at full throttle?
The car uses about 20 HP to maintain cruising speed. Is it wasting three times that amount as heat? Probably not. Let's say it's wasting 5 HP as tailpipe heat.
Obviously blowing the exhaust through "radiator" (heat exchanger) isn't going to make the exhaust cold. It'll be out almost as hot as it went in. Perhaps we can recover 1HP, on a brand new vehicle in the lab. We need 56HP, we only got 1HP. Oops.
Now go look in your tailpipe. See the soot? Notice all the rust and everything on the bottom of the car after a few years of driving? That'll probably cut efficiency in had again, so we end up getting about 1% of the power we need.
Thanks though.
Another way, perhaps simpler than the math and making assumptions about waste heat, is to try this simple experiment:
Walk by your tailpipe with the car running.
Does it feel like a 42,000 watt heater to you? In other words, did you get cooked when you walked by?
According to the research paper describing the new thermoelectric generator, it resulted from a joint effort among researchers at the University of Osaka, the Technical University of Denmark, and E-ThermoGentek Co., Ltd.
These researchers deserve the IEEE Junichi Nishizawa Medal
They don't reduce emissions. They reduce the perception of emissions by the public.
The cars didn't reduce emissions the last 20 Years in real mileage, only on paper.
aaaaaaa
"The module exhibits a maximum output power density of 158 mW cm2 at dT = 105 K, corresponding to an efficiency value of 1.84%"
What more do you need to know ?
I'm just guessing that someone along the line never learned about Carnot's law and absolute zero and efficiency and such. As another wild guess, the efficiency of this device is gonna be under 5 percent., which means it likely can't even pay the cost of money it took to buy it, ever.
Automotive industry: There is a continuing push to improve overall mileage of a car. If waste heat can be captured as electrical power, It will get adopted.
A far better solution is to use EVs which don't generate anywhere near the amount of waste heat in the first place and already get vastly better fuel economy. It is almost always a better idea to not generate the waste in the first place than to generate it and try to recapture the waste. Instead of trying to reclaim a few extra percent waste heat, work on fast charging for EVs which has a FAR greater long term ROI.
And then Ford or Chevy or Subaru will come out with a car that doesn't have an alternator in it.
Don't need an alternator in an EV so that's already done.
Let's say we are going to use this for generating electricity from home water heating and furnace exhaust. Let's assume it can put out 100 watts at a cost of $200 installed. That is about $0.003 dollars of electricity per hour, wholesale rate. Let's say you need heat or hot water 1/3 of the year, that is about 3,000 hours. So this thing makes 3,000 times 0.003 dollars per year, that's $9 dollars. If you borrowed the $200 at 5% interest, that's $10 per year. You are losing $1 a year. Probably more, as feeding the power back to the utility is going to require at least $200 in wiring and controllers and permits. So you're losing like $11 a year. Not an economical idea.
I can't wait to charge my phone with my oven! Who knew frozen pizzas could save the wortld?!?
Package it right and put it on every cpu heatsink and psu in a datacenter. As long as the parts are cheap enough, cutting power usage in a large datacenter by 2% could be huge.
Your assumptions on engine usage and efficiency are way way off. By your estimate: 20HP to the wheels and 5HP rejected as waste (you didn't mention the radiator so I am assuming you are including the radiator in this number), the engine of the car would be 80% efficient. Not only would you be upsetting known engine design and tip toeing up to (or over) the line on known modern physics and metallurgy, most engine manufacturers would be knocking on your door asking what the secret is.
Reality: An ICE engine is roughly 25% efficient (with variations per manufacturer and their patented goodies), and an ICE car has a tank to wheel efficiency around 17%-19% (to account for other losses- transmission, parasitic losses, wheel contact losses, etc). I'll use the 25% for this exercise.
Using the 20HP to maintain highway speeds. The engine produces 20HP at the shaft and this goes to the wheels. So the thermal input to the engine is 80 HP: 20HP goes to the shaft and 60HP gets rejected through the radiator, direct radiation from the engine or down the tailpipe. The majority of the heat rejection is through the radiator: This is a controlled point of heat rejection. The only heat lost at the tail pipe is the heat carried by the ejected mass of that is left over from the products of combustion (CO2, H20 and some un-combusted hydrocarbons that made it into the exhaust pipe). So if this new heat-to-electricity generator is 1.84% efficient. It draws power off the rejected heat. In this case the car is rejecting 60HP in heat and would be getting 1.1HP in electrical power out of it. This is assuming the waste heat could be channeled into a useful shape. But automotive engineers are clever, I think they could work something out.
So this would be an at a glance improvement of 1%. It is actually more though. By removing the alternator, some of the parasitic losses (that contribute to the total of 17-19%) are removed. So the car gets more power at the wheel and a higher fuel efficiency. Even if the automotive industry doesn't jump on this, you can bet the trucking industry will leap on a 1% engine efficiency improvement. And this savings will be bigger for the refrigerator trailers that draw electricity off the trucks. For them, every shipping dollar counts.
Architectural plans are like computer source code with a couple of differences: You only compile once.
The intent is not to use this as the primary power source, but as part of the heat recovery system after the primary systems have used the heat.They are getting 2% efficiency on the waste heat rejected. So this heat recovery works on industrial processes that generate lots of heat as well: Foundry work, forges, HVAC applications, molding, and industrial process where there there is waste heat and steam. So if I have a power plant that is 35% efficient (pretty standard, some can get as high as 40% now), that power plant is rejecting 65% of the energy to the atmosphere. Usually this heat is rejected by cooling towers. The intent here is to recover 2% of the 65% that is being lost. So that recovers roughly 1.3% of the original power input.
Here is an example (this is simplified):
I have a 1 MW power plant that is 35% efficient. In order to power this power plant I have to inject 2.85 MW of heat into the power plant. 1MW goes to the generators and the other 1.85 MW goes to the cooling towers. If I could recover 2% of the 1.85 MW going to the cooling towers, I produce an additional 37,000 KW (0.037 MW) of power and that much less has to go to the cooling towers reducing cost to operate the cooling towers.
Architectural plans are like computer source code with a couple of differences: You only compile once.
The operating range of this system is 100C-150C (200-300F), so we are well above the boiling point of water. This system is more likely to get picked up in industrial fields first: Power generation, heavy trucking, shipping. equipment with big engines and big generators where the heat recovery will make a noticeable dent.
Since most data centers run in the 65-85F range, the cooling systems run chilled water in the 40-45 range. The work required to get the temperature of the rejected heat up to something useful would be prohibitive.
Architectural plans are like computer source code with a couple of differences: You only compile once.
You didn't mention the catalytic converter, which gets very hot. You did mention several places that get hot - where heat is dissipated. Unless you put the entire car in a bubble of transducer material, you're not going to capture most of that heat.
You're right you'd probably get the most electricity by wrapping the radiator in this material. Of course then the radiator would stop working and the engine would overheat. You'd need to make the radiator about 50 times larger to offset that.
Where this technology might be useful, according to the people who invented the technology, would be sensors and such would require only a tiny amount of power. In an automotive context, consider for example tire pressure sensors. You can't very well run a power wire to them.
Perhapss more useful, sensors in the exhaust stack of a factory, where it's rather inconvenient to put plastic wiring inside a hot chimney. Easier to let the wireless sensor be powered by the heat.
So you have pushed the problem upstream to the power generation plants.
You say that like its a bad thing. Power generation plants are far more efficient right out of the gate than even the best car engine.
And I'm ignoring the fact that you can put solar panels out and cut the power plant out of the loop altogether.
(solar power doesn't have the efficiency yet to drive cars around full power on solar cells).
What are you talking about? There already are people powering EVs with 100% solar power. That's not even a question. A typical large roof can provide more than enough energy to power an EV for typical driving needs. Larger solar farms can do even better.
The power plants currently generate lots of waste heat
Power plants are FAR more efficient than the engines in cars. It's not even close. Yes they generate waste heat which can be recaptured but what do you think is easier or more sensible? Capturing waste heat from millions of inefficient cars or capturing waste heat from a few thousand already more efficient power plants?
Both Power generation and trucking are highly competitive with low profit margins and anything they can do to improve their bottom line will be done.
You think power generation is a low margin business? You might want to look closer because their margins are pretty solid on average. Gross margins of around 60% and Operating margins of around 10-15% on average. In the US most power generation is a regulated monopoly. Not software company margins but I wouldn't describe them as a low margin business either. And again, most trucking would benefit FAR more from utilization of EVs (or hybrids) than some marginal gains in thermal recapture from inefficient ICEs. You're proposing stepping over a dollar to pick up a few pennies.
> Unless you put the entire car in a bubble of transducer material, you're not going to capture most of that heat.
The TFA is talking about a system that works on 100-150C differentials. You don't have that differential between the vast majority of the car and any other part, so putting the entire car in a bubble would be nonsensical.
> Of course then the radiator would stop working and the engine would overheat.
Again, you don't understand that most of the heat is NOT CAPTURED by the transducer material. It's not an insulator. Almost all the heat energy gets through, which is part of why the efficiency is so low. You'd need a radiator with maybe 5 to 10% larger cooling ability, which could be achievable with 5 to 10% larger area, not 50 times.
You could extract energy from places like the catalytic convertor which would be more efficient than the system in TFA, but currently you'd end up using a Stirling engine, which is more complexity and more of a mechanical liability than an alternator, so it's done. If there was a transducer that worked well at 500C differential, and had a higher efficiency, then I think it would be viable, although the underside of the car where it might be damaged isn't my first choice for location.
Auto companies, for decades, have been playing around with thermoelectric recovery of power from internal combustion exhaust heat as a replacement or supplement for the alternator. They'd LOVE to replace that pack of moving parts, wearing bearings and slip rings, and fan-belt wear, with a quiet, solid state, black box that pulls the necessary power "for free" from otherwise wasted heat, rather than sucking down a horsepower or two (when the battery needs its starting and pre-start energy consumption replaced) and lasts the life of the car - or at least the muffler or catalytic converter.
But this won't do it. It's less efficient than the peliter cells they've already tried, and its big advantage is that it is thin and can flex, which isn't needed in a car.
The last two improvements in electric generation for cars were the result of semiconductor technology: The replacement of the double-wound, commutator-rectified generator with the diode-rectified alternator, an the relay-and-buzzer electromechanical regulator with a semiconductor regulator (currently built into the alternator).
The main advantage, which got these deployed as soon as the semiconductors were up to it, was that the alternator could generate enough at idle to immediately pick up the operating load and start replacing the battery power used to start the engine. Generators needed the higher RPMs of driving to replace the starting power. This drastically reduced the depth of discharge on the battery, lengthening its life, and also avoided the scenario of running the battery down if you have an in-city driving cycle composed of just short trips. (The other advantage was that commutators and relay-regulators wore out, requiring regulator replacement and generator rebuilds occasionally during the life of the vehicle, while slip rings on an alternator can last until the bearings also fail.)
A waste-heat scavenger doesn't generate substantial power until the exhaust system is up to temperature, or close to it. That would put you back to the lots-of-short-trips-kills-the-battery scenario if you replaced the alternator with a heat scavenger. It would be even worse, because you'd still be powering the running loads off the battery even as you drive away. So you still need the alternator.
By the time the scavenger is putting out, the alternator would typically have replaced the starting power and any pre-starting energy use, and be just running the operating loads. 3/4 horsepower is close enough to a kilowatt as not to bother with the difference, and your typical vehicle's run power is well below that, so (even with its slight inefficiencies) it's not putting enough running load on the engine to make much difference So far, peltier cells (apparently better than this invention) haven't been attractive enough as a fuel-saver to justify their expense.
There are other issues too. (Like increased complexity and opportunity for failure, whether they can stand the heat levels needed to scavenge significant power, and the vibration to last the life of the car without an expensive repair.) But the above lack of improvement on an alternator seems to me to be enough of a killer in itself.
Having "free" energy available doesn't mean the value of the-amount you can collect and put to good use will exceed the direct and reliability risk costs of collecting it.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Package it right and put it on every cpu heatsink and psu in a datacenter.
That's nuts. You're burning power to pump heat to cool those CPUs.
Even if this thing, the air conditioning system, and the electrical system, were a perfect carnot cycle prime mover, refrigerator, and supercondicting wiring with 100% efficient electronics, you'd be burning every watthour generated by the device to cool the cold end of the thermal scavenger to a temperature lower than that of the CPU's heatsink. It would be a wash (except for buying the extra gadgetry and paying for a bigger heat pump).
But they're not perfect. In fact, this thing is a drop in the bucket compared to the carnot cycle. So you'd be buying several times more extra power from the grid than the amount generated by the gadget.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Yeah I see they've been doing PR pieces on it for over a decade. Hopefully in another decade or two it'll be an option on one of their vehicles. That would be cool.
Delta T for this generator is 105 K.
In other words, to get that 1.84% you need to put one side of the TEG on ice, and then boil the other side of it.
The module exhibits a maximum output power density of 158 mW cm-2 at dT = 105 K, corresponding to an efficiency value of 1.84%, which is comparable to a conventional bulk TEG.
On the other hand...
Average surface area for adult men IS 1.9 m2 or 19000 cm2.
Meaning that human body wearing a full-body TEG-suit would produce ~3000 Watts if doused with gasoline and set on fire.
I only hope Apple users record themselves powering their devices that way. So the rest of us could laugh at them longer.
Mit der Dummheit kämpfen Götter selbst vergebens
> but from what I can tell they've been at the prototype stage for decades.
Yeah they've been a hot new thing for at least 50 or 60 years.
Maybe one day they will actually have practical application.
too pointless didnt read.
First off, new ICE cars will probably stop selling in about 2 more years
--
WindBourne
> I think I I can help you find more waste heat than you have found. The Bugatti Veyron (base model) ...
Darn, so close! It would be fun to tinker and see how much waste heat we can actually capture, but unfortunately mine isn't the base model. :)
On a more serious note, such a car can produce a lot.of horsepower *for a few seconds*. After a few seconds it either just goes to roughly 0 HP, or the driver dies and it goes to zero.
Either way, you can only put about 20HP to the wheels for an appreciable amount of time with a car on the highway.