The most common desiccants are silica gel, calcium chloride (CaCl2), activated carbon, zeolites, lithium chloride (LiCl) and lithium bromide (LiBr).

Some desiccants can adsorb other liquids than water. Of interest are liquids which boil at temperatures somewhat below water's freezing point: methanol, ammonia, ethanol, methylene chloride.

The most obvious use of desiccants in the HVAC area is to dehumidify intake air. Systems which do this are usually set up as "desiccant wheels". If the heat produced during adsorption is excluded from the conditioned space, this produces a form of cooling effect by removing enthalpy from the air. (Warm dry air has less potential energy in it than warm wet air; if you were to add moisture back into the air using an evaporative cooler, you could produce an open-cycle cooling effect.)

A closed-cycle cooling effect can be produced by desiccants, by using their affinity for the working fluid to "draw down" its partial pressure and cause a container of working fluid to boil, absorbing heat from the environment.

An open-cycle effect also seems feasible -- a form of evaporative cooling in which humidity is actively added inside the conditioned space, and is simultaneously removed by a desiccant outside the conditioned space.

faq@solarmirror.com

 NREL desiccant cooling page
   http://www.nrel.gov/desiccantcool/

 EREC brief: zeolites in solar energy applications
   http://www.eren.doe.gov/consumerinfo/refbriefs/ba3.html

 Solar Adsorption Refrigeration Using Zeolite and Water
   http://www.fh-luebeck.de/an/pt/solar/publish/euros-00.pdf

 Kai Oertel's homepage, methanol/silicagel based adsorption cooling systems.
   http://infoserv.kp.dlr.de/koertel

 Zeo-Tech, makers of zeolite-water adsorption coolers and heat pumps
   http://home.t-online.de/home/zeolith/e_index.htm

faq@solarmirror.com

The "Zeolite and water" link above is an 8 page paper by
Kreussler & Bolz with lots of specific numbers.

One concern I have is the volume of zeolite required -- K&B seem to be
using about 0.1 to 0.2 cubic meters of zeolite per kWh of cooling power,
which is pretty humongous.  Is this stuff going to be sort of like handling
blocks of styrofoam?  And if I wanted to keep my house cool for a day
(say, 50 kWh of cool for which I'd usually pay $3 on my electric bill)
I'd have to bake, what, 5 or 10 cubic meters of zeolite???

Another concern is the efficiency (a COP of 0.08).  A concentrating
collector could probably double that efficiency pretty easily, but
even a COP of 0.16 is going to require several times more collector area
than the solar-electric or ammonia-absorption alternatives, with a
sun-to-cooling COP of 0.5 or more.

Marc Ringuette

From the solar-ac archives, re: solid desiccants

Calcium chloride is by far the best choice for a desiccant system because of
its cheapness, strong affinity for water, and relative ease of regeneration.
In contact with metals (even copper), it causes the metals to react with
carbon dioxide and corrode, so anything in contact with the wet calcium
chloride needs to be made of glass or plastic.

Concentrated calcium chloride solutions don't freeze as easily, unlike
magnesium chloride brines (another possible desiccant available in vast
natural deposits).

If you need a solid desiccant, there is a possibility 
that desiccant materials like calcium chloride might
be integrated into cements.  I don't know if anyone has tried this, but it
should get a look.  Cements are, in some cases, pretty porous.  Imagine a
honeycomb of cement and calcium chloride, extruded through a forming process
to form a high surface area material like a honeycomb.  Ever hear of
desiccant wheels? They are basically this same concept: a ceramic binder
with an absorptive material integrated into it.  I think the current
desiccant wheels are made with titania and cement somehow.  The desiccant
wheels out there now have very high surface areas, and are quite a piece of
technology.  But here, I'm talking about a lower-cost, bulk material that
would do the job.

Then again, there is simply sawdust and calcium chloride: a cheap, high
surface area, solid/liquid desiccant.  Or, for a purely "organic" approach,
it might be possible to integrate the calcium chloride desiccant somehow
into the construction of a straw bale houses. The trick would be to find a
particular type of straw that would be saturable with calcium chloride, but
not loose its strength.

Another solid desiccant you might be interested in is activated charcoal.
It is interesting in that it is black, and would be ideal in a combined
desiccant/solar collector.

People have tried solid desiccants with other "working fluids".  For
example, instead of using water vapor, you can make a desiccant cooling
cycle by replacing the water with methanol, ethanol, or methylene chloride.
The big advantage: in comparison to water vapor, these working fluids have
higher partial pressures.  The higher the partial pressure, the higher the
heat transfer ratio.

A little appreciated fact about boiling heat transfer in all liquids: the
heat transfer coefficient depends on the partial pressure of the boiling
fluid!!  Systems using lithium bromide/water vapor  must work with very low
pressures (substantially below atmospheric).  The result: heat exchanger
surfaces must be made very large (that's part of the reason these systems
are so expensive).

-David Wells

Liquid desiccant systems have certain advantages over dry desiccants.  One 
of the biggest is the fact that as the desiccant is heated up to drive off 
the moisture, significant heat may remain in the hot, dried solution.

In the case of solid desiccant systems, such as systems employing solid des=
iccant wheels, it is hard to recover this heat.

With liquid desiccants,the hot, dried solution may be circulated into a cou=
nter-flow heat exchanger, using the heat of the hot and dried solution to p=
re-heat the incoming moisture loaded solution.

Calcium chloride is a cheap, non-toxic, widely available material that has =
certain properties that make it a good candidate for this application.

I've got a few tables from my Lange's Handbook of Chemistry that have data =
that may be useful.

First, I'll take a few points off of the twelfth edition's Table 10-20, "Co=
mposition of Aqueous Antifreeze Solutions", which is based on Dow Chemical'=
s 1929 data sheet. The (%CaCl2) is percent anhydrous Calcium Chloride.  Bag=
s typically don't contain pure anhydrous, so keep this in mind.

Freezing Point(C)    % CaCl2=20=20=20=20
0      (32F)                  0
-4.8  (23.4F)             9.2
-9.9  (14.4F)           14.6
-15.6(4.1F)             18.6
-21   (-5.8F)            21.5
-30.8(-34.9F)           26.9
-41.8(-41F)              30.5

The above table shows that the calcium chloride solutions will not freeze, =
and this is important.

Next, here is  some excerpts from Lange's Table 10-25, which shows the humi=
dity level above various concentrations of calcium chloride solutions.  Dat=
a is for 25 C.

Humidity    %CaCl2
100%        0
95%        9.33
90%        14.95
80%        22.25
70%        27.4
60%        31.73
50%        35.64
40%        39.62
30%        44.36%

At very high concentrations, you end up with basically a pure Calcium Chlor=
ide Hexahydrate tank.  Calcium Chloride Hexahydrate melts at almost room te=
mperature, 29.9 C.  So, in effect, it could act as its own phase change sto=
rage material.

As for the heats of reaction, it is generally the case that the heat of dec=
omposition of most desiccants is about equal to the heat of boiling of the =
water in the desiccant.

So, as a "phase change storage material", desiccants are pretty amazing.

-David Wells

You can store heat in a concentrated LiCl solution.  It's also possible
to cool with this kind of desiccant using an open-loop system.

You might have a 180 F water tank and another tank that's filled
with a 180 F 50% LiCl/CaCl2 solution on an average day, and mix them over
5 days to make a 33% solution. Diluting a 50% LiCl solution from 50% to 0%
releases 262 Btu/lb of heat... 50 to 33% might release (50-33)/50x262
= 87 Btu/lb, so starting with P pounds of LiCl solution and P pounds of
water, 210K = 2P(180-80)+87P makes P = 732 pounds, approximately, ie 2
91 gallon tanks, or 4 8'x12" PVC pipes in a closet.

To get cooling, you need to drive an evaporation process.
You might have a pipe full of LiCl solution and a pipe full of
rocks to act as a "packed column" to absorb water vapor in LiCl, and
a pond full of rocks under the slab or outdoors to supply the vapor.

Each pound of vapor absorbed releases about 1000 Btu of heat, but should
make a similar amount of cooling available.

We might turn on a fountain in the house when cooling is needed and
trickle some LiCl solution into and circulate house air through the
packed column, keeping the house at 50% RH. If the house becomes more
humid, we run the column without the fountain.

House air exits the column warmer than when it entered, but with less
moisture, so the net effect on the house is cooling, because it takes
more heat to raise the exit air back to 50% RH than to cool it to 70 F.

The heat in the dry return air can be reduced via an air-to-air counterflow
heat exchanger (swapping heat with the intake air).

Meanwhile, the LiCl solution is getting warmer and more dilute. It might
warm from say, 100 to 140 F during the day, as it dilutes from 50 to 25%. 


These papers have more:

   "Unglazed collector/regenerator performance for a solar assisted open
   cycle absorption cooling system" by M. N. A. Hawlader, K. S. Novak,
   and B. D. Wood of the Center for Energy System Research, College of
   Engineering and Applied Sciences, Arizona State University, Tempe,
   AZ 85287-5806 USA, in Solar Energy, Vol. 50, pp 59-73, 1993 and

   "Effectiveness of heat and mass transfer processes in a packed bed
   liquid desiccant dehumidifier/regenerator" by Viktoria Martin and 
   D. Yogi Goswami in HVAC&R Research, Vol. 6, No. 1, pp 21-39,
   January, 2000 and

   "A review of liquid desiccant cooling" by Viktoria Oeberg and D. Yogi
   Goswami, chapter 10 in Advances in Solar Energy, Vol. 12, pp 431-470,
   1998, American Solar Energy Society Publishers.

Nick Pine

Source: http://www.ece.villanova.edu/~nick/usenet/00001775