[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: NFC: [Fwd: Peltier Junctions]

"D. Martin Moore" wrote:
> OK, call me lazy.  How much cash would you have to spend in
> order to cool, say, a 20 long, uninsulated, by 15 degrees using
> Peltier Junctions?

First you need to calculate the rate of heat gain (Q) for the tank,
which is the same as the amount of heat that the cooling device will
have to remove in order to keep the aquarium temperature constant. 
Generally we can use the formula Q = U*A*(Ta - Tw), where

Q = heat gain
U = overall heat transfer coefficient
A = total surface area (sides and top)
Ta = ambient air temperature
Tw = water temperature

From an experiment, I determined that U = 0.45 W/ft^2/F.

The dimensions of a 20 gallon long tank are 30"(L) x 12"(W) x 12"{H). 
For our purposes, A = L*W + 2*L*H + 2*W*H.  To obtain the area in square
feet for this tank, use A = (30*12 + 2*30*12 + 2*12*12)/144 = 9.5 ft^2.

We already made the assumption that Ta - Tw = 15 F.

So the total heat that must be removed is this:
Q = 0.45*9.5*15 = 64 W.  This could be done with one unit, or with
several units in parallel.

Based on the information at the Melcor web site, the Peltier junctions
consume about 1.7 watts of electricity (I*V) for every 1 watt of heat
transferred.  So the power consumption required for this application is
64*1.7 = 109 W.  This will probably require a heat sink and fan on the
hot side of the Peltier junction.  I haven't yet figured out how the
heat might be drawn out of the water on the cold side.  Any ideas?

Unfortunately, the Melcor page doesn't give prices for the hardware
itself. I would be surprised, however, if it came close to the
$500-$1000+ price tag for a typical compression chiller.  Compression
chillers, by the way, are much more efficient, especially for high heat
transfer applications.  (The company admits this on their web site.)  So
even if Peltier junctions are practical for cooling a small aquarium,
anyone with a 100+ gallon beast is certain to be stuck with a
conventional chiller.

Andrew Dalton


Follow-Ups: References: