X-Message-Number: 2051
From: R. Ettinger 
Subject: Thermoelectric Cooling

I have appended below Robert Ettinger's replies to the March 20 - 26
CryoNet messages concerning thermoelectric cooling.
					Kevin Q. Brown
					
					
-----
March 30, 1993

From: Robert Ettinger

To: Wowk & Correspondents

Subject: Thermoelectric Cooling

Brian Wowk has had some good ideas for using thermoelectric effects in
regulation of a cold room, and then has proposed it for the actual
cooling, and asked whether his cost projections are not too good to be
true.  I'm afraid they are.

First, a minor quibble or two.  Mr. Wowk uses "Peltier effect" for
both the actual Peltier effect (using a potential difference to
produce a temperature difference) and its inverse, the Seebeck effect
(using a temperature difference to produce a voltage).  If you want a
generic term, you could say "Thomson effect," which includes the
others.  Also, he uses the term "efficiency" of a refrigerator when he
means "coefficient of performance." The efficiency of any system is
defined as the useful work output divided by the energy input, and can
never be more than 100%, while the coefficient of performance of a
refrigerator, the ratio of the heat extracted to the energy input, can
be more than 100%.

Now: Mr. Wowk uses a heat engine formula to show--correctly--that in
the ideal case, operating between 143 K and 283 K, the coefficient of
performance is greater than 1.  He then guesses that, in an actual
case, the coefficient might be only half of that, which would still be
good.  Unfortunately, that guess is almost certainly much too
optimistic.  I don't have hard numbers, but I'm almost certain of this
on a common sense basis, as follows.

First, Peltier kitchen refrigerators are not competitive, although
they are said to be getting closer.  Second, heat pumps for home
heating are apparently sold almost exclusively in the warmer states
where there isn't much heating load, and even there they are premium
priced compared even to electric baseboard heating, which in turn is
maybe three times as expensive (ongoing) as fuel heating.  I think
this indirect evidence is nearly conclusive--that we would have to pay
a high premium price for thermoelectric cooling.

Eventually--as I pointed out in The Prospect of Immortality and at
various times since--we may be willing to pay a premium price to get
both energy (thermopiles) and refrigeration (Peltier cooling) without
moving parts or purchased supplies, a high capital cost but virtually
no maintenance or ongoing cost, giving us cherished reliability and
independence.  That time is getting closer, but I don't think it is
here yet.  We probably still need better transistor junctions and
maybe better superconductors (since Joule heating is one of the major
problems in Peltier cooling).  I wish I had something more positive to
say about this, but this aspect of Mr.  Wowk's cold room probably
won't be realized for a while.
-----

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=2051