Why aren't we using Peltier modules?

[greasemonkee]

I admit, I get behind technology while entrapped in the slave grid, but occasionally a window of opportunity to research does arise and the 21st century hits me like a ton of bricks.

So these Peltier or thermoelectric devices are gaining popularity. I was oblivious to the technology until someone gave me a wine cooler and noticed there was no cooling system, nothing but heat sinks. With all of the heat of our ICE's being blown out into the air I'm surprised to find the alternatorless diehards haven't taken advantage of the technology - which appears to be quite affordable by the way.

Could one not dump a few hundred bucks on several dozen of these things and produce a surplus of electrical energy? I'm intrigued.

...And why the heck are we still using fans on radiators to dump engine heat as if it were waste?


Nevermind. I know the answer - any idea that works must be executed covertly in this world...

[serialk11r]

First of all, heat from the radiator is being rejected at ~90C or so. The ideal heat engine will make squat from that, so a better place to be looking is perhaps the exhaust.

The problem with Peltier units is that they have absolutely horrendous efficiency, and most of the ones being produced nowadays are meant for lower temperatures and thus cannot take advantage of the high temperatures found in automotive exhaust (I don't know what the typical temperatures look like, but it gets as hot as 900C at the catalytic converter, and I remember seeing a chart that showed 450C as the lowest temperature). You can certainly buy a bunch of them right now and rig them up to your exhaust, there's a company that someone here gave a link to a while ago in another thread that wants to make a commercial plug and play automotive exhaust heat regeneration unit. They happen to also sell 300C TEGs (which are the best you can get these days apparently) for 100 dollars each, which put out 15W maximum.

So you start to see the problem, to power ignition, lights, ECU, etc. you'll probably need 1000+ dollars for the thermoelectric devices alone (a DIY heat exchanger and cooling system might not be that hard though). Meanwhile, the BMW turbosteamer is I think well under 50kg, and is producing up to 10kW. So that's kinda what TEGs are up against.

[2000mc]

bmw is working on incorporating tegs too
https://www.press.bmwgroup.com/press...?id=T0119738EN

[yugomodder]

I bought some Peltier coolers, and while I haven't done any serious experimentation with them, my limited experience with them may be of use.

The ones I got were $5 ones off of ebay. They have a maximum draw of 136.8W, with 15.2V, 9A. They're 40mm by 40mm, and have a maximum temperature rating of 67 degrees C (152.6 F).

One of the things that struck me about them is that they create a large temperature differential very quickly after applying a voltage. One side gets fairly cold, like the inside of a refrigerator, and the other gets hot actually pretty hot. As soon as power is disconnected both sides feel slightly warm. I think this is because it acts as a generator as soon as you stop applying voltage. I have not experimented with using them as generators, but might be able to.

Now looking from a broad standpoint, thermoelectric generators are very inefficient. That being said, they are already efficient enough to generate lots of power, just not the ones that are commercially available.
KOMATSU : Komatsu to Launch Sales of the World's Highest Efficiency Thermoelectric Generation Modules Developed In-house
These are experimental thermoelectric generators, with a conversion efficiency of 7.2% and an output of 24W each for a 50mm x 50mm device. They can withstand temperatures up to 280C (536F). Now the problem is that they have an announced price of ¥ 30 000 per piece (approx 373USD). This is a cost of $15.54/W, for a device that only generates power while the engine is running. The cost is definitely the prohibitive factor in this case.

If we look at a standard engine, it's efficiency is approximately 18%-20%. This means that out of the 100% of the energy released, about 80% is released as heat or sound. So if we take that 80% and could convert it at 7.2% efficiency (now I realize we cannot take all of the heat and convert it, this is just for the math) we get 5.76%. Now this may sound like a 5.76% increase in FE, but it's not, it's 5.76% of the total energy released. For example if the motor produces 20kw of power, then 80kw are put into heat and sound, and 5.76kw could potentially be harvested from the heat at 7.2% efficiency. This is an increase of about 28.8% in the energy we get from the engine. That is massive! Now unfortunately we cannot harvest near 5.76kw from this engine at 7.2% efficiency, as we cannot trap all the heat and exchange it with the devices perfectly.


In summary, the devices are very cool, and look promising. They however at this point they are too expensive for use in vehicles that only use them for a few minutes or hours at a time. The conversion efficiency is already more than enough to make a significant impact on fuel economy, but not at their current cost.

If the cost of Peltier coolers is reduced to a reasonable level, like even $6/W they would have huge potential in power plants converting excess heat into electricity. They could also be used to harvest waste heat from blacktops. To reach the point that they could be used in cars however, they would have to be cheap enough that they could pay for themselves in the few minutes or hours a day that they are used. (note for the 136.8W $5 Peltier coolers off of ebay, 136.8W is the maximum load, the output would be far less, but as to how much I do not know)


TEC1-12709 TEC Thermoelectric Cooler Peltier Module New | eBay
Here a link if anybody would like to experiment with them. I would be happy to perform an experiment using the ones I have, if anybody has parameters they want tested please ask me.

[Ryland]

Yugomodder is dead on, Solar is down to $1.35 per watt and TEG's are around $15 per watt and are very fussy, to cold and their output is pretty low, to hot and they burn up and are destroyed.
Heat from the radiator would be near the ideal temp but it takes many miles for it to warm up and in that time they are going to produce very little power and these short trips make up most of the average persons trips in a car.
Only 5-7% of the energy that is passing thru a TEG is being turned in to electricity and if you don't get rid of that heat then the performance drops as well! so while the hot side needs to be around 180-200F the cool side needs to be kept as cool as possible and the spec numbers tend to be around 80F, so you need some massive heat sinks to dissipate that heat!

In the peltier coolers, the reason that both sides get hot after you turn it off is because so much of the energy that you are putting in to the unit is being turned in to heat, lets say you put 100 watts of energy in and you get 300 watts of heat out because you moved 200 watts of heat away from the cold side to make it cold and in doing so you produced 100 watts of heat that also had to be pulled away from that cold side, so as soon as you stop pulling that heat from one side to the other the whole thing gets hot, now don't quote me on those figures I mentioned, because I don't remember if a 100 watt really can move 200 watts of energy or not but I from what I remember it seems pretty close.

[oil pan 4]

Originally Posted by Ryland

Solar is down to $1.35 per watt and TEG's are around $15 per watt and are very fussy, ....to hot and they burn up and are destroyed.

That is why.

The "bolt on" exhaust powered TEG that is still in development is going to cost well over $3000 and has a target power out put of 500w.

[jamesqf]

If it's simply a matter of recapturing energy from waste heat to improve IC engine efficiency, it'd be so much more effective - and cheaper - just to replace it with a Stirling engine as generator in a hybrid system. Or for just capturing exhaust energy, look up the "TIGER" device, which drives a generator with the exhaust side of a turbocharger.

[greasemonkee]

I see, a couple more hours of research reveals the technology is still in its infant stages - well, what is available to the general consumer; neat gizmos nonetheless. Perhaps I could make a seat cooler for my non-AC civic at a minimal cost in economy.

[oil pan 4]

Originally Posted by greasemonkee

Perhaps I could make a seat cooler for my non-AC civic at a minimal cost in economy.

Thats one of the best ideas yet.

[ConnClark]

Originally Posted by greasemonkee

Perhaps I could make a seat cooler for my non-AC civic at a minimal cost in economy.

Just remember that the heat is going to be dumped on the other side of the Peltier and its going to be about ten times as much as you remove from the cool side.