How to use TEGs to create power with your exhaust heat

[Daox]

PART 1




So, you'd like to use a thermal electric generator (TEGs) to eliminate, or reduce the load on your alternator? We have actually talked about it quite a bit over the years here on EM. Here are just a few examples:

http://ecomodder.com/forum/showthrea...ator-4538.html
http://ecomodder.com/forum/showthrea...les-22368.html
http://ecomodder.com/forum/showthrea...tegs-3124.html

But, this isn't a thread discussing if its possible. This is thread that shows you how to do it! So lets see what it would take to get there.


First off, what the heck is a TEG / Seebeck / Peltier generator anyway?

This is what wikipedia has to say about them:

A thermoelectric generator (TEG), also called a Seebeck generator, is a solid state device that converts heat flux (temperature differences) directly into electrical energy through a phenomenon called the Seebeck effect (a form of thermoelectric effect). Thermoelectric generators function like heat engines, but are less bulky and have no moving parts. However, TEGs are typically more expensive and less efficient.

So, we need a temperature differential because that is what makes these TEGs thingies work. You put heat on one side, and cool on the other an they generate a voltage. Pretty neato stuff. The larger the temperature difference, the more power they create. So, what is the largest temperature differential we can find on our vehicles? That would be heat from the exhaust for the hot side, and ambient air temperature for the cool side. This gives us several hundred degrees Fahrenheit of temperature differential. Yes, we could use coolant, but that only has at the very most about a 200°F differential. 600F vs 200F is going to give us a lot more power. Sadly, TEGs are not crazy powerful or efficient, so we need all the temperature differential we can get.



Okay, where the heck can I find TEGs?

Now that we know where we are getting our heat from, we can go try to find some TEGs. The biggest challenge now is finding TEGs that hold up to exhaust temperatures. The closest thing I've been able to find are TEGs sold at TecTeg.com and CustomThermoelectric.com. Both of these sites have TEGs capable of handling 320C (608F). This may still be on the low side for heat resistance. Exhausts can definitely get hotter than 600F, so you would have to be sure to position the generator downstream far enough so that it would not see above 600F. This will require testing by you on your vehicle in the middle of summer as that'll give the worst case scenario.



Thats great, which one do I choose?

Now, we have a whole bunch of TEGs to choose from, but which one will work the best for us? Well, first we have to figure out what size we want to look at. I chose to look at the 40mm (1.57 inches) units as I think that would fit best on the exhaust of a smaller car. More on this later though.

To compare them, I made the spreadsheet you see below. It contains all the 40mm TEGs that the two companies have listed. Ideally, we want about a 13.5-14.5V output from the TEGs. This will ensure that the 12V battery gets fully charged. To get these specs, I used the "matched load" numbers, not the open circuit numbers on their spec sheets.





Out of all of these TEGs, I like the second one the best. Here are the full specs on it. We will need 2 of them wired in series to get 13.4V output which should be great for getting a battery charged up. This doesn't mean it won't charge above that. It just means that 13.4V is where its most efficient (putting out the most amps). So, I will be using this one as the example going forward.



PART 2

How many of these TEG thingies do I actually need?

Great, we have a TEG selected that we'll use! Now, how many do we need? As you can see, the TEG listed above puts out 7.5W. Two of them in series will create 15W. This is not an amazing amount of power by any means. A standard alternator puts out 13.5V and lets say 80 amps for a total of 1080W. To match that output, we would need 144 TEGs! That is a bit much, and actually probably not required the vast majority of the time. The alternator is sized to be able to handle every electrical load on the car at the same time. The liklihood of this happening is small and if you're aware of it, you can ensure it never happens. So, the next step really is to figure out how much power you want your TEGs to generate. The best way of doing this is to figure out what kind of electrical load your car has while its in operation. As a rough guide, MetroMPG has posted some information on typical electrical loads. The best way is to measure it with an ammeter though.

Once you know what your typical electrical load is, you can figure out how many TEGs you need. For my example, I am going to say that I need 15A (200W) of charging. We already know that we need 2 TEGs in series, and that series of 2 TEGs gives us 15W. So, the math is simple at this point. We take 200W of generation divided by 15 which gives us 13.3 TEGs in parallel. I'm going to round that up to 14 to keep the numbers even. 14 in parallel times 2 in series gives us 28 TEGs. That sounds a lot more doable.





Nope, wait! You need more.

BUT, this is their rated output at maximum temperature differential. We surely won't be operating at maximum temperature all of the time. So, we will need to derate them. By how much is a total guess unless you've done testing to see what kind of exhaust temperatures you'll see consistently. Then you can reference the data sheet for the TEG and guess what kind of power output you'll get. For this example I am going to take a total guess and say we will derate them by a third. Now we need 36 TEGs to get our 200W.



PART 3

How do I put this jumble of TEGs together to make some powah?

Yay, we're getting 36 TEGs to generate around 210W and significantly reduce the load on the alternator. But, how do we slap the TEGs on the car and make it work? Lets get to that now.

Note, this is just a quick way I came up with to slap the TEGs on the exhaust. There are many other ways you could possibly do it.

You'll notice I've kept the number of TEGS divisible by four. This is intentional as the mounting of them will be on a square tube. Four sides to the tube, four TEGs keeps things all neat and tidy. This is also why I chose 40mm TEGs as they fit the square tube that kind of matches the pipe size the tube will be replacing. The square tube will replace a section of the exhaust. It would look something like this:





Then, you would need to add heat sinks to the cold side of all the TEGs to help cool them off and maintain the highest temperature differential possible. It would look something like this:





And beyond that, you have to find a way to bolt those heat sinks down to the tubing. The TEG modules are supposed to be under a fairly specific amount of pressure to ensure long life and good performance. Here is some of the guide from TecTegs.





PART 4

OMGz the wires!

Next up we need to take care of the electrical wiring. In this example, you would wire 2 TEGs in series, and 18 of them in parallel. It'll look something like this. Yep, its a bit messy, but really its nothing compared to an EV with BMS wires haha.





Controlling the heat!

Beyond this, if we've done things right, we'll also need some way to control the generator. We can't just hook it up to the battery and let'r rip. Its possible that the generator could overcharge the battery. We also need to stop the TEGs from discharging the battery when they aren't producing power. If you put power through a TEG, it will create heat on one side, and cool on the other. While this is neat, its not really what we're looking to do.
Therefore, we need a charge controller of some sort. This could be as simple as a switch if you want to keep things extremely simple. You'll also want a voltmeter to monitor things if you go that way. However, a much better idea is a purpose made charge controller. Something that is made for a solar panel or wind turbine would be what we're looking for. It will protect the battery from overcharging and from discharging as well. The charge controller will have to handle the max amps of this setup of 15A. Here is a cheap example of what I'm talking about. I just searched amazon real quickly and found one.





PART 5


image by BA1969

Holy cow, it cost how much?!?!

Now, assuming you've done all these things, which is totally possible. I've specifically left out one very important detail to this all, the price. If we need 36 TEGs, those TEGs are going to cost us $31 a piece (that is with their volume discount). That brings us to a total of $1116 just for the TEGs. That doesn't include the pipe, welding work that will need to be done, the mounting system, heat sinks for the cold side, or charge controller and wiring. So, this is where things sadly really fall apart. I've bought entire cars for less than what this system would cost. Granted, you could do this, and transfer the TEG modules from vehicle to vehicle assuming they last that long. In a automotive application I have my doubts though. In these high temperature situations, they estimate the TEGs to last about 4 years.

So, while this is all completely doable, its insanely expensive. But, that is how you do it folks! Hope you enjoyed the engineering exercise.



===============================

Here is my analysis of using TEGs powered by your coolant:

How to use TEGs to create power with your coolant heat

[California98Civic]

I like this thread idea a lot. I swear I saw on digikey.com a tile that generated 14.5 volts on its own and more amps. And there seemed to be lots of them. Have you looked at that site?

Anyway, I think an ideal spot for placing such a system would be near the rear wheel well along the exhaust pipe but in front of the rear wheel well. It'll be a little hotter there than it would be down by the muffler but mufflers often rot out on this you get a really good stainless steel one and I think all of that compromising of the structure of the muffler and the leaking of exhaust can't be good for the tegs.

One or 2 tiles just to trickle charge a battery is a small system and just shouldn't be all that hard to produce.

[Daox]

Thanks. It is pretty fun going through the design requirements for me. Of course I'm a mechanical designer and quite enjoy it too, so that make sense.

TEGs are really cool little devices.

[teoman]

Tec is a thermoelectric cooler and a teg is a thermoelectric generator.

Tecs are relatively cheap compared to tegs. Their major difference is their heat tolerance. I think creating a setup with tec's and coolant is a more controllable.

I bought quite a bit of materials for this a long time ago. I really should finish the project at some point.

[mwebb]

most new er cars need
minimum of 10 to 20 amps to run the engine and DRL with out much else

at full loads with AC and wipers and high beams most cars are approaching
100 amps
and many of the big SUVs are considerably over that

which is why the ECM has control of charge rate ... so it can minimize alternator load on the engine when not needed

so
you have chosen a pathway that will present a considerable challenge

.................
fresnel lenses can produce upwards of 2000f at the focal point .... that would be dependent on the sun and its relative position to the TEG and lens ... which would be in constant flux in a car

[Daox]

PART 2

How many of these TEG thingies do I actually need?

Great, we have a TEG selected that we'll use! Now, how many do we need? As you can see, the TEG listed above puts out 7.5W. Two of them in series will create 15W. This is not an amazing amount of power by any means. A standard alternator puts out 13.5V and lets say 80 amps for a total of 1080W. To match that output, we would need 144 TEGs! That is a bit much, and actually probably not required the vast majority of the time. The alternator is sized to be able to handle every electrical load on the car at the same time. The liklihood of this happening is small and if you're aware of it, you can ensure it never happens. So, the next step really is to figure out how much power you want your TEGs to generate. The best way of doing this is to figure out what kind of electrical load your car has while its in operation. As a rough guide, MetroMPG has posted some information on typical electrical loads. The best way is to measure it with an ammeter though.

Once you know what your typical electrical load is, you can figure out how many TEGs you need. For my example, I am going to say that I need 15A (200W) of charging. We already know that we need 2 TEGs in series, and that series of 2 TEGs gives us 15W. So, the math is simple at this point. We take 200W of generation divided by 15 which gives us 13.3 TEGs in parallel. I'm going to round that up to 14 to keep the numbers even. 14 in parallel times 2 in series gives us 28 TEGs. That sounds a lot more doable.





Nope, wait! You need more.

BUT, this is their rated output at maximum temperature differential. We surely won't be operating at maximum temperature all of the time. So, we will need to derate them. By how much is a total guess unless you've done testing to see what kind of exhaust temperatures you'll see consistently. Then you can reference the data sheet for the TEG and guess what kind of power output you'll get. For this example I am going to take a total guess and say we will derate them by a third. Now we need 36 TEGs to get our 200W.

[t vago]

I would suggest placing these TEGs in the immediate vicinity of your upstream O2 sensor, and just behind and downstream of your downstream O2 sensor. If the O2 sensors can take the heat, the TEGs should also be fairly happy there.

Edit: And are you sure you want to rely on ambient air to generate your temperature differential? Seems like you're going to want to put some cooling fins on to take away this heat, and use some sort of forced airflow, at a minimum. Maybe a coolant loop would be better.

[California98Civic]

Thanks for continuing the thread. That's a lot of TEGs at $20-$25 each!

I think this description so far suggests significant complexity in using TEGs to replace an alternator entirely. A much simpler set up would contribute to reducing alternator need. It could be smaller and cheaper.

To tvago's good points, I assume we'd use cooling fins if we put a TEG unit into the under car airstream. Daox is talking about a lot of TEGs, and I think they would not easily fit around the manifold. Better to be lined-up along the straight middle sections of exhaust piping. Some sort of forced airflow seems possible under the car, along the piping, quite easily. Personally, I would sacrifice temp differential for system simplicity. Coolant circulation makes for complexity.

Any system would need some kind of controller, yes? And there would be a need for a way to prevent the TEGs from drawing power when the car is cool, maybe? There might be situations in which ambient air is warmer than exhaust piping, no?

[Daox]

Originally Posted by California98Civic

Thanks for continuing the thread.

No problemo. I have it mostly written out, its just a matter of rolling out all the parts of it. There are 5 parts in total.

I think this description so far suggests significant complexity in using TEGs to replace an alternator entirely. A much simpler set up would contribute to reducing alternator need. It could be smaller and cheaper.

This is possible, but is there much point in saving 5A from an alternator? The gains will be fairly small.

As Oil Pan mentioned in one of the other threads, doing a temporary alt delete will give you a good idea of the gains you'll get from doing a mod such as this.

[gone-ot]

Originally Posted by Daox

Thanks. It is pretty fun going through the design requirements for me. Of course I'm a mechanical designer and quite enjoy it too, so that make sense.

TEGs are really cool little devices.

The engineers' version of Cuba Gooding, Jr's famous Jerry McGuire movie line: "...show me the DATA..."