Sciencemadness Discussion Board

Thermo electric generators (peltier devices) - potential uses & designs

RogueRose - 14-5-2018 at 23:00

So I was trying to figure out a way to power some things and was basically stuck with either solar or heat differential and that led me to make some calculations for TEG's and was really amazed at the potential electricity generation of these things for a relatively small size.

The device is 4cm^2 and 3.6mm thick. They will run at 12 - 15.4v at 6A and have been spec'd to run at 60, 72 and 92 watts. Operating temp are stated -50C to 83C so a temp differential of 133C or about 240F.

Now there would be 625 of these units in a square meter and if you could have the max temp differential then I would suspect the units would produce close to the max numbers on the specs (though those are for producing cold or heat). The amount of energy produced for the following unit wattage ratings are:
60w = 37.5 Kw
72w = 45Kw
92w = 57.5Kw

Now those are ideal maximums for the device but it is to show that there is a lot of potential there for either energy generation or for heating/cooling in a relatively small area. For comparison, a solar panel can get about 1-1.8Kw per sq meter (depending upon location on earth, efficiency of panel and other factors). Using the max solar # the TEG's can do 20-30x energy generation and 38-58x using the lowest solar number.

So I was wondering if it would be possible to make these in a curved fashion so as to fit something like a soda can inside? Or possibly a flexible unit?

Also is there a way to speed up heat loss by radiation without using air flow or a fluid to move anything on the surface of the radiant surface?

Ubya - 15-5-2018 at 02:32

mhh you didn't thought of a few things. after 10 sec of googlig:
Quote:

First 60 watts means nothing, it translate to 60 watts maximum transfer of heat when temperature difference is zero!

Second the generated voltage is proportional to the Seebeck effect, and the cell is made of many couples in series parallel.

For an estimation take a cell made of 126 couples, Seebeck constant 0.05 volts/K (SM), then you need a difference of 100 °K to get a useful voltage from a cell with 3 ohm internal resistance (RM). But let's use 50 °K (DT).

Use Thermoelectric Technical Reference - Power Generation formulas.

Maximum Power PM= (SM x DT)^2/(4 x RM)= (0.05 x 50)^2/(4 x 3)= 0.52 watts.

Even worse when you observe that max power load means 3 ohm with a voltage of 1.25 volts! at 0.4 amperes.

Maybe for experimental reasons you go ahead with 50 °K, then you need to pump some heat (Qc) into the module with hot surface at say 323 °K with thermal conductivity 0.7 watts/°K (Kc)

Qc= SM x Th x I- 0.05 x I^2 x RM+ Kc x ΔT, the minus sign for self heating

Qc= 0.05x323x0.04-0.05x0.04^2x3+0.7x50= 35.6 watts of heat in

And heat out on the cold plate of 35.6-0.52 or 35.1 watts at 273 °K

Try to scale this to a useful power of say only 10 watts!

You are stuck with an efficiency below 1.5%, and in the best case below 3% with ΔT of 100 °K

The real problem is not the low power output, is the big waste of heat that must get out on the cold plate.




copied from quora


[Edited on 15-5-2018 by Ubya]

Melgar - 15-5-2018 at 06:24

They're a lot more efficient if you use a heat sink and a fan to keep the hot side cooler. That's how they're intended to be used. I'd have two 60-watt TECs cooling the liquid for a distillation with 14/20 glassware, and they certainly do help. They can easily keep the coolant 15˚C or so below whatever you're distilling. Running with no distillation, it'll keep it cool enough that there's condensation on the tubing.