X-Message-Number: 2031
Date: Fri, 26 Mar 93 01:40:04 CST
From: Brian Wowk <>
Subject: CRYONICS: More Peltier Effect
The -130'C Cold Room will require about 50 watts of electric
power for fans and the active temperature control system. We should use
the Peltier effect to provide this power. A heat sink should be built
into the concrete surrounding the vault (and perhaps extending into the
earth as well). By attaching thermopiles to this heat sink and the LN2
reservoir, the 210'K temperature difference can be used to passively
generate electricity to power everything. Thermodynamics predicts that
producing 50 watts in this manner will heat the LN2 at a rate of 20
watts, increasing LN2 boiloff by only 4%. Of course the 50 watts itself
is eventually dissipated in the room, also increasing boiloff, but this
power would otherwise be piped in from the outside anyway.
I dare say this is a great idea. With this system you don't
have to worry about external connections to power supply lines, battery
backups, or ever switching between the two. Heck, you don't have to
worry about anything other than faithfully replenishing your LN2 once a
week. Now your cryonics facility can explode, burn, and fall on top of
your vault (covered by fire-resistant panels) and your self-contained
patient care system will never know the difference.
There is another problem the Peltier effect may solve for us. A
Cold Room must be "failsafe." I interpret this to mean that the
consequences of total mechanical failure, though suboptimal, must be
acceptable. The only mechanical parts remaining in the present room
design are the fans. We must consider what new equilibrium will
establish itself in the room if the fans are stopped. (In fact, during
patient loading operations they may be deliberately stopped for extended
periods of time.)
The design I introduced yesterday would conduct heat toward the
insulated LN2 reservoir through the aluminum walls between storage
cells. Metal heat conductors attached to the -130'C aluminum walls
would pass through the reservior insulation, making direct contact with
-196'C interior. This represents a steep temperature gradient, with the
icy -196'C temperature constantly "fighting" to get out of the reservoir
and into the rest of the room. If the air circulation stopped, I'm
afraid it would succeed. I ran some numbers today and concluded that in
absence of air circulation, unnacceptably large temperature gradients
would arise within the cell walls between the center of the room and the
periphery. The inner cells would cool dangerously. The only way to
prevent this would be to make the aluminum walls cetimeters thick
(ouch!).
I suggest we make the walls a more reasonable thickness, and use
the savings to go out and buy a couple of dozen state-of-the-art
thermoelectric heat exchangers (thermopiles). We insulate the LN2
reservoir heavily and completely, line the inside with thermopiles, and
rely on thick electric cables to carry heat from thermopiles at the room
periphery to their counterparts in the reservoir. The goal is to move
400 watts of heat across the room without temperature gradients or air
flow, and thermoelectricity is the only way to do this. Such a system
would be inherently stable and self-regulating, making the fans almost
redundant (although we will still have fans anyway).
I ran some numbers based on a copper/constantan therocouple and
concluded that passively conducting 30 watts between a 60'K temperature
difference would require moving thousands of amperes through copper
cable thicker than my fist. Since commercial refrigeration systems
exist based on the Peltier effect, I infer that they must use something
other than copper/constantan (probably semiconductors). When time
allows, I will investigate this further and see what hardware is out
there available commercially.
I am also now thinking that there is really no need for a
"utility" cell large enough to lower a man into. Other than the fans,
what is there to really service anyway? We simply stick the 2000 litre
LN2 reservoir in three central cells, stuffing them full of insulation
right up to the very top where the fans are. This now leaves 22 cells
for patient storage, increasing total capacity to 132 patients.
By the way, my "can within a can" design for the LN2 reservoir
addresses the LOX buildup problem nicely. The inner can can be
periodically hoisted out of the unit and dumped to get rid of LOX
residue at the bottom. This is a far simpler maintenance procedure than
purging dewars with patients in them. Chalk up another one for -130'C
storage!
--- Brian Wowk
Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=2031