# Peltier Thermal Calculations

#### newothegreat

Joined May 3, 2012
26
Hello,
I am working on a project involving a Peltier (TEC) module, and I need to calculate the total heat load for the hot side so I can spec out a heatsink. I'm sure it's pretty straight forward, but I'm having trouble finding an exact formula that shows that value.

For a given electrical input on the TEC (Vtec [volts], Itec [amps]), and a given temperature differential (deltaT [deg C]), there is a specified amount of heat transferred from one side to the other (Qtransfer [Watts]). I can find that information very easily from the device datasheet. And as far as I understand it, Qtransfer will be subtracted from the cold side and will be added to the hot side, in an ideal case. So my question is: what is the thermal load on the hot side?
Is it:
1) Qtransfer + (Vtec x Itec)
or
2) Qtransfer + ( (Vtec x Itec) x efficiency constant), where the efficiency constant is between 0 and 1, and it accounts for the work being done due to the peltier effect.

So another way to ask the question is, does all the electrical power that goes into the TEC get converted to heat, or does some of it get used up by the peltier effect? If not all the electrical power is converted to heat, what are some typical values for the "efficiency constant"?

Thank you,
Owen

#### wayneh

Joined Sep 9, 2010
16,124
So another way to ask the question is, does all the electrical power that goes into the TEC get converted to heat...
Yes

.. or does some of it get used up by the peltier effect? If not all the electrical power is converted to heat, what are some typical values for the "efficiency constant"?
Energy is conserved and doesn't get "used up". You can be absolutely certain of that and use that fact to calculate your energy balance.

The piece you are missing is the efficiency of transferring heat versus the operating conditions. It's worst when you run the TEC to the rated specs, and might be one unit of energy moved for every ten consumed from the power supply. I think you can achieve close to one-to-one efficiency (one moved for one consumed, two appear on the on the hot side) if you operate at low current, lower ∆T and so on. In other words you can get better efficiency for a given heat flow if you have more TEC capacity.

The data sheets for the TEC should provide the efficiency curves you need. If not, there are generalized references out there. I might have some links if you can't find them.