Entropy

[Paul Sears <psears at nrn1_NRCan.gc.ca>]

I never thought I would use that title for a posting here!

Scott asked about the entropy involved in a chilling process.

I'm going to avoid _all_ the nineteenth century units (F, BTU,
hp, BTU/h, SEER).  I'll stick to watts (joules/second), joules,
degrees C and K.  Anyone who wants can do conversions!

A chiller might realistically remove 3 kW of heat from a coil at
5C, and dump heat to another coil at 40C, and use 1 kW of electrical
power in the process (COP 3).  (I'll ignore motor losses).

3 kW heat is absorbed at 5C (278K), so the entropy lost here is:

3000/278 = 10.79 (joule/K) (entropy unit) per second

4 kW heat is emitted at 40C (313K) so the entropy gained here is:

4000/313 = 12.78 (joule/K) (entropy unit) per second

The electrical energy is pure work and has no entropy, so the
net increase in entropy is 1.99 joule/K per second.  This is O.K.
by both the first and second laws of thermodynamics (energy balance,
and _increase_ in entropy).

One can see that the temperatures involved place a limit on the COP
possible.  One can always get a _lower_ one!

Peltier effect coolers are neat, but have very low COP's


-- 

Paul Sears
psears at nrn1_nrcan.gc.ca, 613-996-4171, facsimile / télécopieur 613-996-9400
Natural Resources Canada, 1, Haanel Drive, Nepean, Ontario K1A 1M1
Ressources naturelles Canada, 1, Haanel Drive, Nepean, Ontario K1A 1M1
Government of Canada / Gouvernement du Canada