Electricity from Exhaust info - Seebeck cells

[oil pan 4]

I like the idea of getting power from exhaust heat.
Wonder how much its gona cost?
I could do $500 maybe a little more.

[Ladogaboy]

Probably a lot more than $500. I'd also need four of these units to equal the output of my stock alternator. Granted, I might not NEED that much power, but I'd want to assess how much power I am actually using first.

I'm curious if the price would be reduced for a group buy...

[mans]

Originally Posted by serialk11r

That 136 watts is how much power you can pump into it to make it act as a horrendously bad heat pump. If you wired it in reverse, you'd probably see only a few watts come out.

The 62x62 one you linked is 180C max temperature, again, useless. It'll produce at best a few watts.

Efficiency is a concern because loading the cooling system significantly will be a problem.

i think were still ok:

even if what you write is true, were still ok cuz even worst case scenario lets say you only get 20% of it's capacity in electricity when the unit is heated (and its probably substantially better then that) then if you had 20-25 of them then you'd get 540-675 watts of power which is enough to replace the alternator on a small vehicle any day.

at $5 a piece that would run you $100-$125 plus a way to install them.

I'd say put half of them on the coolant and half on the exhaust

the trick would be getting them the proper distance away from the exhaust pipe so they dont fry.

I welcome all feedback. lets remember: alt delete can equal 10-12% gain in mileage, so lets try to get an idea how to make an idea work for us

[Air-Hybrid]

Originally Posted by mans

i think were still ok:
at $5 a piece that would run you $100-$125 plus a way to install them.

I'd say put half of them on the coolant and half on the exhaust

the trick would be getting them the proper distance away from the exhaust pipe so they dont fry.

Put as many on the exhaust as room allows. Coolant temp is too low.
If these things have a max temp rating of 180degC then the efficiency will drop off exponentially away from that (course the actual efficiency is defined by the differential in heat - cold side to hot side, but for this application cold-side is ambient & won't vary enough to change the equation).
So on the coolant system the temps conducted to the device will be 100degC at best, which for ambient of 20C equals bad efficiency. Hot side 179degC versus 20degC will be far better.

To control the heat transferred from exhaust - you could use a bi-metal strip (or an electronic equivalent) to actuate a plate (give it some thermal mass, ie. don't make it not too thin or you'll get hot spots!) half an inch to make/break contact with the exhaust. The Seebeck devices can be mounted on the back of the plate.

[deathtrain]

personally I dont see this working at all. but haveing a positive attitude to anyones ideas and willingness to try it. if I read this right the bigger the heat differance the higher the out put. OK heat side against the exhaust right we get that. How about installing a CPU heat sink (with out fan) on the coolside. then it will "air cool" while the car is moving.

[ProDarwin]

Originally Posted by deathtrain

personally I dont see this working at all. but haveing a positive attitude to anyones ideas and willingness to try it. if I read this right the bigger the heat differance the higher the out put. OK heat side against the exhaust right we get that. How about installing a CPU heat sink (with out fan) on the coolside. then it will "air cool" while the car is moving.

This is the method I was thinking about trying... mount it on the exh manifold with a big CPU tower heatsink with air directed across it and right out the hood.

Anyway, these things are terribly inefficient. My company makes TE Refigerators for aircraft applications (No moving parts). They are astoundingly inefficient compared to a compressor cycle, and astoundingly expensive as well. They can't deal with the ~800-1500 deg temperature delta you'd get from the "hot" parts of your exhaust system.

Several manufacturers have tried to create a thermoelectric catalytic converter, but it hasn't proven to be cost effective at all.

[Air-Hybrid]

Originally Posted by deathtrain

personally I dont see this working at all. but haveing a positive attitude to anyones ideas and willingness to try it. if I read this right the bigger the heat differance the higher the out put. OK heat side against the exhaust right we get that. How about installing a CPU heat sink (with out fan) on the coolside. then it will "air cool" while the car is moving.

Re. the heatsink for the TEG's cold-side: if the car is at low speed or stationary, the devices are likely to become even less efficient without some forced air (like a fan) - because the thermal differential across the device will shrink if the heat isn't removed effectively.

I should add:
Maybe my exploring a 'best' setup implied I -do- think these devices would work ... I don't. At least I think the cost would mean a payback about 10 times the life of the car! Give it another decade for the tech to improve, I reckon.

=====
The best way to recover power from the exhaust is not to make so much of it in the first place by fitting an Atkinson cycle engine!

[mans]

Originally Posted by Air-Hybrid

Re. the heatsink for the TEG's cold-side: if the car is at low speed or stationary, the devices are likely to become even less efficient without some forced air (like a fan) - because the thermal differential across the device will shrink if the heat isn't removed effectively.

I should add:
Maybe my exploring a 'best' setup implied I -do- think these devices would work ... I don't. At least I think the cost would mean a payback about 10 times the life of the car! Give it another decade for the tech to improve, I reckon.

=====
The best way to recover power from the exhaust is not to make so much of it in the first place by fitting an Atkinson cycle engine!

do you believe that one of these devices rated at 136 watts is likely to be unable to deliver even 13.6 watts of electricity when at full operating temp?

i'm guessing it should put out at least 10% of its rated capacity, and with enough of them on an exhaust system at operating temp it should do the trick.... no?

[Air-Hybrid]

Originally Posted by mans

do you believe that one of these devices rated at 136 watts is likely to be unable to deliver even 13.6 watts of electricity when at full operating temp?

i'm guessing it should put out at least 10% of its rated capacity, and with enough of them on an exhaust system at operating temp it should do the trick.... no?

I was going to write an answer, but I found Serialk11r has covered this question already

Originally Posted by serialk11r

That 136 watts is how much power you can pump into it to make it act as a horrendously bad heat pump. If you wired it in reverse, you'd probably see only a few watts come out.

The 62x62 one you linked is 180C max temperature, again, useless. It'll produce at best a few watts.

Like I said the commercially available ones that are any good at producing electricity are rated up to 300+ and cost a fortune.

They're working on developing better materials that should increase efficiency a good bit, but don't expect that to show up on the market anytime soon. There are TECs that can run at 700C or whatever but they're probably not for sale to the general public, and probably cost a crapload, not to mention still have poor efficiency. Efficiency is a concern because loading the cooling system significantly will be a problem.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^

So $5 for a couple of Watts might not sound too bad on the face of it, but if these devices are even 1% electrically efficient that would mean that potentially 99% of the heat would have travelled across the device to the 'cold' side. That's alot of extra heat to get rid of and that is almost certainly going to mean indirect load on the engine itself - running cooling fans, pumps, etc.
Gut feeling is the net result would be a negative... unles I'm missing something.

[Ecky]

If I'm not mistaken, isn't this what NASA uses in its deep space probes for power? (heated by a chunk of something radioactive)