Continuous Ammonia Absorption Air Conditioner / Heat Pump

The following discusses a possible way to configure an intermittent ammonia absorption system for continuous use by alternately heating and cooling two independent systems. See a description of an intermittent ammonia absorption system here: http://knowledgeableideas.blogspot.com/2013/05/a-solar-ammonia-absorption-icemaker.html

I've considered various possibilities on how to devise a practical micro absorption air conditioning system, and even spent some time working with a small lithium bromide system that uses water as the refrigerant by operating at a high vacuum. A problem with the vacuum systems that use water as the refrigerant is that they are bulky, and maintaining vacuum can be a problem. Lithium bromide systems can see problems with crystallization, and the units that use solid adsorbents are difficult to regenerate and cool without extensive heat exchange equipment. I now believe that a modest system can be had by operating two small intermittent ammonia absorption systems in tandem. What I like about the ammonia systems more than anything is that they can (1) see much lower temperatures in the evaporator, and (2) they can be compact. Unfortunately, ammonia is nasty stuff (however, note that there are systems that use an ammonia evaporator to chill water, and the chilled water can be distributed to a fan coil unit for space cooling - this will allow for isolating ammonia away from living spaces). The problem in configuring an intermittent system for continuous use is figuring out how best to alternate between heating and cooling the two absorbers so that the system is always cooling (or heating). As a side note, this system could also provide refrigerant to an evaporator used to cool a freezer (or refrigerator), and that makes it even more interesting, .

About the absorber - the absorber of these systems is always the heart, and they tend to be the bottleneck in the system. I had considered one way to devise an ammonia absorber that might increase the rate of absorption. Consider two lines entering the top of the absorber (consider a vessel that is oriented vertically as opposed to the one in the link that is horizontal). One line extends into the top of the vessel ending in a perforated coil of tubing at the bottom of the vessel. The other takes off directly from the top of the vessel. The second line includes a check valve that prevents ammonia gas from entering the absorber through that line. The idea here is that the pressure in the absorber will be lower than the evaporator at all times except when it's heated. Therefore, ammonia vapor will be driven off the evaporator and enter the absorber through the line that is perforated at the end. This will penetrate the absorbent with ammonia vapor for good absorption.

Now, I'm open to suggestions on how to devise a practical system to automatically heat and cool the absorbers to achieve a continual operation. Until then, I'll throw out a suggestion. Consider two systems of absorbers (side by side) where each is made up of a cylindrical vessel positioned vertically. Each vessel is contained in an insulated shroud. A system can be devised to direct either high temperature combustion gases OR the exhaust from a fairly high capacity blower fan through the system. I suggest a wood gasifier furnace as the heat source. Basically, a wood gas furnace is operated continually to supply high temperature combustion gases, and a blower fan is operated to supply cooling air. A baffle is positioned periodically to switch the flow path of these two fluids. I'm not interested in the specifics right now as I just considered this. Please offer feedback.

HEAT PUMP MODE: This system also has the ability to provide space heating. This configuration entails using the evaporator to pick up heat from outside the home. The heat from the furnace is provided by the condenser during regeneration, and the heat harvested from outside is provided by the absorber during absorption. I'm not concerned about how to configure the system for this, I'm only pointing out that it's possible. This configuration can provide heating in excess to the heat provided by the furnace (i.e. COP > 1.00).

Ammonia is hard to get these days. Where can you find it?

Abe Connally wrote:Ammonia is hard to get these days. Where can you find it?

I don't know about restrictions on the purchase of ammonia, but it is widely available... but yes, these restrictions are a problem. It's probably illegal to own, so this idea is likely not viable even though I'm sure it can work. So far the best prospects for powering a modest off grid home I've considered is based in photovoltaics with a wood gas engine system for back up. Air conditioning and heat pumps can be had as an opportunity load on a large PV array. Also, a wood gas engine system could be operated intermittently for battery charging and with heat recovery to regenerate a desiccant that can provide heating and cooling applications (see the open desiccant systems described in the threads "Desiccants for Thermal Storage" and "Intermittent Adsorption Refrigerator").

Basically, I'm discussing the physics of this possibility. I'm not focusing on the legislation that affects it. Although, I acknowledge this problem.

NOTE: 29% aqueous ammonia. This can be distilled to yield ammonia of sufficient purity. Actually, a better way might be to allow dry calcium chloride to absorb the ammonia out of the solution directly. So, ammonia can be had from industrial strength ammonia based cleaning agents.
http://www.amazon.com/Ammonium-Hydroxide-High-Purity-1000ml/dp/B009J5NRQQ/ref=sr_1_1?s=hpc&ie=UTF8&qid=1387389273&sr=1-1&keywords=ammonium+hydroxide

I'll speculate more on how one might harvest ammonia from these products. The basic idea is to connect an absorbent vessel to the dilute ammonia solution vessel and allow the absorbent to take in ammonia vapor. However, the amount of water taken in should be minimized. This can be done by keeping the dilute ammonia solution at a low temperature to discourage water evaporation (but don't let it get too cold where it freezes). I suggest placing it in an ice water bath as this will keep it above the freezing point which is less than that of water due to the ammonia, but it will be kept low enough to discourage water evaporation (note that the solution will cool during absorption, so the ice water actually provides heat to the ammonia solution). Also, the absorbent should be kept cold, so place the absorbent vessel in a freezer along with a long steel tubing coil. The idea is that keeping the absorbent cold will encourage absorption as the heat of absorption will be removed quickly - and the absorbent has a higher capacity for absorption at a lower temperature. Providing the cold steel tubing will freeze water vapor on the tubing and keep it from getting into the absorbent vessel. Applying a vacuum pump to remove air will speed the process, but perhaps the ammonia vessel should be isolated with a shut valve while the rest of the system is evacuated to prevent pulling ammonia out of the system. Open the valve to start the process. The absorbent vessel should remain in the freezer at all times until one desires to construct a working system since allowing it to heat up will cause some ammonia loss (assuming the absorbent is saturated with ammonia at the freezer temperature). So, in theory anyway, one could have the absorbent vessels ready to go... just get the evaporators and condenser set up, get a tight system, draw a vacuum, then connect the absorbent vessel to the system, and open the isolation valve. Heat the absorbent vessel, then the ammonia will fill the system, condense in the condenser coil, then collect in the evaporator.

It's possible in principle to regenerate a solid adsorbent such as zeolite efficiently when high temperatures are available by using this same set up, and using water as the refrigerant. Some differences include the use of larger tubing to transfer the low density water vapor to the absorber. Also, no check valves should be used in these lines as there is not enough differential pressure available. With high temperature available for regeneration, then it could be made to work. Unfortunately, it would be very difficult to cool the zeolite during absorption. Better absorption and regeneration might be had by placing a central core in the cylindrical absorber vessel as a fine mesh cylinder where the zeolite is on the outside against the cylinder wall of the vessel, and the central mesh core will facilitate water vapor exchange. One advantage is that a store of water could be chilled directly and the chilled water distributed to a fan coil unit directly using the magnetic drive pumps I discuss in the thread "Intermittent Adsorption Refrigerator". Still, I really think ammonia is the way to go if it can be had.

Maybe just look at open desiccant cooling if heat powered a/c is desired.

Maybe a very small continuous ammonia absorption unit like I describe here can be used with a large freezer. In that case a lot less ammonia is required.

This comment follows directly from the previous. Perhaps a large chest freezer can be had using the continuous ammonia absorption concept?... that is, alternately heating and cooling two independent intermittent systems. The big advantage here is that each system can be very small. It could likely work, but won't likely be practical. Really, conventional freezers are the practical option. Still, it's an interesting prospect.

Here's another wacky idea on how this kind of system might be configured (I'm really bored guys). A major problem with the first configuration is the electricity consumed in the blower fan for cooling the absorber. A system might be devised to heat and cool the absorber with liquids. For example, consider an oil used for both heating and cooling, and with the oil distributed with very low power DC pumps. This approach would introduce opportunities for heat regeneration to boost efficiency. Another possibility I imagined for a small system like a freezer is suspending absorber vessels inside combustion chambers with a cable (or a rack and pinion set up). The cables are attached to pulleys on a rotating shaft such that when the shaft is turned, then one vessel drops while the other raises. When all the way up, one vessel will be exposed to outside air for cooling and the lower part of the vessel will seal the top of the flue. This will direct all combustion gases to the other side. This approach would consume almost no electricity (just a low power gear motor controlled by a timer or thermostat - and, in fact, it is possible in principle to cycle the system with a mechanical device using gravitational potential energy like elevated water). The furnace could also be devised to operate without a fan. So, a zero electricity freezer is possible.

Just for grins, I'll discuss a conceptual way to devise a purely mechanical actuator to automatically shift the device last described. The shaft in question is counterweighted with the two absorbers that are the same weight, so there is no net torque. Two additional counter weights may be added in a way that the weight of one may be varied to generate a torque. This can be had with water. So, imagine that these two new counterweights are buckets. Also, consider that the shaft is latched into the two extreme positions (*), and the shaft also positions a water valve to direct the flow of water from an elevated tank into one of the two buckets. When enough water has entered a bucket, then the weight overcomes the latch, the weight falls to rotate the shaft, and the momentum carries the system over to get latched into the new position. There is a valve in the bottom of the bucket that gets opened when the bucket hits bottom, so the water drains out. The water from the tank then fills the other bucket to start another cycle, just back and forth until the tank runs out of water. (*) An approach better than a latch can be had by securing the system with friction using an adjustable spring mechanism. Since static friction is greater than dynamic, then finding the right spring tension will secure the system until the weight of water overcomes the static friction (which varies with spring tension and can be adjusted). When rotation starts, then the lesser dynamic friction will allow increasing momentum to force the system over until it stops where static friction takes hold again as the water drains from the vessel and the other vessel starts to fill and start the cycle over.

... then again, just use a small gear motor with limit switches on a timer. Of course, the other idea is more fun. I wonder if an air conditioning system could be used to chill a large tank of water in a small home, and would this dehumidify and provide sufficient cooling? Maybe connect several vessels in such a way that thermosiphon could keep the water flowing around for good cooling. It sure would be interesting to devise a 100% electricity free system of air conditioning and food refrigeration using a system like this, and be able to configure it to heat the water as well. Anyway, just speculating.

http://www.achrnews.com/articles/aluminum-evaporators-for-ammonia

It seems ammonia has good compatibility with aluminum. Good to know, but hot pressurized ammonia seems a bad idea with ammonia (the article does specify the use of aluminum evaporators which remain at low temps and pressures).

I've been focusing on a continuous ammonia absorption system based on operation two intermittent systems in tandem. Here, I'll describe a system that operates on a single circuit by using a liquid pump and using water as the absorbent. Note one advantage here is that it's possible to use 28-30% ammonium hydroxide (mixture of water and ammonia) directly as the working fluid for this system, and this can be purchased directly as industrial cleaning agents (*).

A low power piston or plunger pump is required. I recommend a HyPro piston pump with teflon seals, and driven by a dc permanent magnet gear motor. The pump work is roughly 1/20 - 1/10 that of a compressor used to power an efficient conventional air conditioning system of the same cooling capacity.

The pump sends the water/ammonia solution to a heater where the solution is partially vaporized. The solution then enters the top of a vertical section of pipe that serves as a separator. Here primarily ammonia vapor leaves the top of the vessel and primarily liquid water remains. A few points here: it's advantageous to introduce the solution to the vessel through a vortex to aid separation (just point the incoming tube against the top of the pipe wall so it spins when it comes in - often done in "cyclone separators" - to help lessen carryover of water with the ammonia). Also, the liquid level in the pipe cannot be allow to get too high or too low. This might require a small float valve at the base. There are available small stainless steel floats designed for this purpose that thread onto rods. This is a good idea for this application where the float is connected to a valve at the base of the pipe section - just cut the end of the rod to a long taper and set inside a restriction at the base that leads to the liquid drain.

There are now two flow paths: ammonia vapor and liquid water. The ammonia vapor goes to a condenser where the ammonia returns to liquid state. From there, the liquid ammonia reaches a restriction (might be a needle valve, orifice, or thermostatic expansion valve). The liquid ammonia passes the restriction to enter the low pressure evaporator where the liquid ammonia flashes to vapor and provides the cooling effect. Now, let's consider the water side of the system. The water in the separator vessel is at a high temperature. This water has to be cooled after leaving the separator (it's possible to regenerate this heat back into the system for a boost in efficiency if desired). The water now enters the low pressure side of the system through a restriction.

Ok, now we have liquid water and ammonia vapor in the low pressure side. We have to get these together in the absorber to reconstitute the liquid ammonia/water solution. Once this is done, then the solution returns to the pump to start the cycle over. There are various ways to go about this. I suggest using a water spray nozzle as the restriction for the water side of the system to help mix water with the ammonia vapor. Things must be kept as cool as practical here to remove the heat of absorption. I expect the overall electricity requirements for this kind of unit to be 1/5 - 1/4 that of an equivalent vapor compression system (considering ammonia/water pump, cooling water pump, and fan coil unit - an additional small pump could distribute chilled water to a copper fan coil unit to keep an ammonia evaporator outside a living space). Also, note that the ammonia condenser can be used for water heating purposes.

I just wanted the reader to be aware that these systems exist.

(*) In principle, one can use water and ammonia in the intermittent cycle. However, there will be carryover of water to the evaporator over time which will likely degrade performance. However, I wonder that a purge valve might be added to the evaporator to bleed water that might accumulate. Either way, it's clear that water/ammonia systems work for the intermittent system (see Crosley Icy Balls), so one should not dismiss the use of industrial ammonium hydroxide here. Also, it seems to me that a tubing coil used as the evaporator here would see water settle at a low point in the tube which should be the end of the tube. If water is allowed to settle, then it seems to me that it would just stay frozen there and not present a problem during operation.

This is interesting, and it has some potential in my opinion. I considered that it's possible to use the high pressure ammonia vapor in a continuous ammonia absorption system to power a small plunger pump used to send the ammonia/water solution into the generator. Well, it can absolutely be done, and I have considered a simple and direct way to configure it. The pump is hermetically sealed and starts automatically once pressure builds to a sufficient level. This main benefit of this configuration over the intermittent ice maker system that uses calcium chloride is that the continuous system requires far less ammonia. Such a system can be more compact, and in particular the receiver on the collector can be smaller and therefore more easily insulated with a vacuum tube. Since the thermal losses from the receiver are the major loss, then such a system used with a collector of the same size as the one used in that system could increase performance on the order of ten fold. Yes, the system is more complex, but the higher efficiency and more compact design seems to justify it. Finally, with efficiency so much higher, it makes sense even in the "third world" to make use of biomass as a heat source here. A compact system with superior performance and not limited to one cycle per day (at most) could be extremely useful in that setting.

Such a system could be scaled up so that a furnace could freeze a large thermal mass to provide a/c, and it would require no electricity as long as the condenser and absorber can be sufficiently cooled by means that do not require electricity... but the ammonia/water pump would be heat powered. Actually, there is likely sufficient work available from the system to drive a water pump for cooling as well. Interesting.

Switch to Carbon Dioxide -CO2 and get your heat needs down to above -40*C.
CO2 long ago replaced Ammonia! CO2 not attack metal fittings.

Peter Mckinlay wrote: Switch to Carbon Dioxide -CO2 and get your heat needs down to above -40*C.
CO2 long ago replaced Ammonia! CO2 not attack metal fittings.

CO2 replaced ammonia? What absorbent is used with CO2 refrigerant in absorption systems? What pressure range do you expect in the regenerator?

Since the heat of vaporization of ammonia is 6 times that of CO2, then ammonia seems the better candidate from an efficiency standpoint. Please explain if I am mistaken.

Carbon steel fittings with teflon seals do very well with ammonia - inexpensive and reliable.

For further information R744-CO2, R744 web site, and Coles possibly of greater assistance than I.