The Problem
Two essential requirements for comfortable and healthy indoor environments are adequate ventilation and good humidity control. Unfortunately in humid climates, which includes much of the densely populated regions of the world, it is difficult to meet both these requirements without using a lot of electricity.
The fundamental problem is that a conventional cooling coil (whether using chilled water or direct-expansion refrigerant) cannot effectively meet the latent loads from ventilation on very humid days. All conventional chillers and air conditioners are essentially sensible cooling devices that dehumidify by lowering air temperature below its dewpoint so that moisture condenses. These systems must run with a wet cooling coil, and the air that leaves this coil will be close to saturation.
The limitation of conventional chillers and DX air conditioners becomes evident when one tries to use them in an advanced HVAC system. Technologies such as displacement ventilation, chilled beams, and radiate panels can be part of a low-energy HVAC system that eliminates the fan energy used in a conventional system that recirculates large volumes of air. However, these advanced systems will not work with a conventional chiller or DX air conditioner that supplies relatively cold air (e.g., 50 to 55 F) that is saturated with moisture (i.e., 100% rh). What is required is a cooling system that supplies drier, but warmer air.
Typical supply air conditions for displacement ventilation might be 65 F and 50% relative humidity. This supply air has an absolute humidity of 45.5 grains moisture per pound of dry air (which is equivalent to a humidity ratio of 65 g/kg and a dewpoint of 46.0 F). Any cooling system that dehumidifies air by reducing its temperature to condense the moisture must first cool air to below 46 F and then reheat the air to 65 F. As shown on the neighboring psychrometric chart, a cooling system that process 6,000 cfm of warm, humid outdoor air must do 54.0 tons of cooling and then 10.8 tons of reheating. This air conditioner is doing 25% more cooling than is required to meet the load. Furthermore, this percent excess cooling becomes much larger during cooler, damp weather, e.g., if it were 70 F and raining, overcooling would be 42% of the required cooling.
The Solution: A Liquid Desiccant Air Conditioner (LDAC)
Liquid desiccants are solutions that have a high affinity for water vapor. This property is the key to creating cooling systems that dehumidify air without over-cooling. (Dr. Lowenstein's review paper presents a more detailed explanation of the theory, operation and status of liquid desiccant air conditioners than can be included here.)
Since the 1930s, liquid desiccants have been used in industrial dehumidifiers. The liquid desiccants used in these systems commonly are very strong solutions of the ionic salts lithium chloride and calcium chloride. These ionic salts have the attractive characteristic that the salts themselves have essential zero vapor pressure, and so vapors of the desiccant will not appear in the air supplied by the LDAC. However, zero vapor pressure comes with a price: as with seawater (a chemically similar salt solution), solutions of lithium and calcium chloride are very corrosive. This corrosiveness requires that all wetted parts within the LDAC be protected and that no droplets of desiccant are entrained by the supply air.
A desiccant has the ability to dry air without cooling because it forms a relatively strong bond with water molecules (i.e., a stronger bond than that between molecules in pure liquid water). Whereas the heat released when water condenses (i.e., the latent heat of condensation) is approximately 1,000 Btu/lb, more heat--typically an additional 50 to 100 Btu/lb--will be released when water vapor "condenses" into a liquid desiccant due to the stronger bonds between the molecules.
At this point it is important to recognize that a desiccant's ability to dry air decreases as its temperature increases. If a 43% solution of lithium chloride at 80 F were to absorb an amount of water vapor that diluted it to 42%, its temperature would increase to 123 F (assuming the desiccant is not cooled). Whereas the desiccant initially could dry air to 23.2 grains, after increasing in temperature and becoming slightly more dilute, its ability to dry air decreases dramatically to 107.6 grains. Most of this loss in drying potential is caused by the increase in temperature: if cooled back to 80 F, the 42% solution of lithium chloride would dry air to 25.7 grains.
One approach to limiting the impact of the heat released when the desiccant absorbs water vapor is to flow desiccant at a sufficiently high rate that its temperature rise is limited (i.e., in the preceding example, if the desiccant's concentration changed only 0.1 point, its temperature would rise only be 4 F). This approach, which was first used in the liquid-desiccant systems of the 1930s and is still used in many LDACs today, requires that the liquid desiccant is first cooled before it is delivered to the top of a bed of porous contact media. The air that is to be dried then comes in direct contact with the liquid desiccant as the air is drawn through the porous bed. Typically, a ratio of the desiccant-to-air mass flow ratio that is on the order of one will limit the temperature rise of the desiccant.
The preceding "high flow" liquid desiccant system has the following disadvantages: (1) a large volume of desiccant must be circulated requiring large pumps with relatively large power draws, (2) the air flowing through the highly flooded porous beds has a relatively high pressure drop which increases fan power, (3) a separate heat exchanger is required to cool the desiccant before it is delivered to the porous bed, and (4) the air may entrain droplets of desiccant as it flows through the highly flooded porous bed. This last disadvantage is particularly important because of the corrosiveness of the desiccant. Carryover of desiccant droplets can be eliminated by droplet filters, but at the expense of additional pressure drop.
In 1994 AIL Research received a U.S. patent for an LDAC in which the flow of desiccant is more than an order of magnitude less than that used in flooded-bed systems. This low desiccant flow was implemented by replacing the bed of porous contact media of the high-flow system with a plastic heat exchanger that continually cools the desiccant as it absorbs water vapor. The continuous cooling of the desiccant insures that the desiccant maintains its drying potential despite the heat released as it absorbs water.
As shown in the neighboring figure, a "low flow" LDAC has three main components: the conditioner, the regenerator and the interchange heat exchanger (IHX). The conditioner is a parallel-plate liquid-to-air heat exchanger. A coolant, typically cooling tower water (but possibly water from a geothermal well, lake or chilled water loop), flows within the plates and a very low flow of liquid desiccant flows down the outer surfaces of the plates. Thin wicks on the plate surfaces create uniform desiccant films. The air to be processed flows horizontally through the gaps between the plates. As this humid air comes in contact with the desiccant, water vapor is absorbed. The heat released by this absorption is transferred to the coolant. The air leaves the conditioner much drier, although its temperature may not significantly change.
The dilute desiccant that leaves the conditioner is pumped to the regenerator. The regenerator is configured similarly to the conditioner: a parallel-plate liquid-to-air heat exchanger. Again, very thin films of desiccant flow in wicks on the outer surfaces of the plates, and air flows in the gaps between the plates. For the regenerator, however, a hot heat transfer fluid flows within the plates. This hot fluid can be supplied by a gas-fired boiler, solar thermal collectors, recovered heat from an engine or fuel cell, or other energy source. As the temperature of the desiccant increases, it releases water into a "scavenging" air stream that is discharged outdoors. (To avoid possible confusion that might be created by the close arrangement of the conditioner and regenerator in the neighboring figure, we note that the scavenging air that enters the regenerator is almost always drawn from outdoors. The processed air that leaves the conditioner would never enter the regenerator--although the discharge air arrows from the conditioner are directed towards the regenerator.)
The hot, concentrated desiccant that leaves the regenerator and the cool, dilute desiccant that flows to the regenerator exchange thermal energy in the interchange heat exchanger. This exchange increases the efficiency of the regenerator and decreases the cooling load on the conditioner.
The efficiency of the regenerator can also be increased by adding an air-to-air heat exchanger to preheat the air that enters the regenerator using the warm, humid air that leaves it. (This air-to-air heat exchanger is not shown in the preceding figure.)
In both the regenerator and conditioner, the flow rate of desiccant is so low that the falling films on the plates are contained completely within the thin wicks. The air velocity over the films is too low to entrain desiccant droplets. Since both the desiccant delivery to and collection from the plates are done without creating droplets, desiccant does not carryover during normal operation. (However, as protection against abnormal operating conditions, droplet filters are commonly installed downstream of a low-flow conditioner.)
AILR's earliest prototypes for low-flow LDACs used plastic-plate heat exchangers for both the regenerator and the conditioner. More recent prototypes have used wicking-fin technology.
How does a low-flow LDAC solve the "overcooling/reheat" dilemma? A low-flow LDAC that uses 43% lithium chloride will supply air at about a 20% relative humidity and a temperature that is 6 F to 8 F above the temperature of the cooling water supplied to the conditioner. On a typical humid summer day, an LDAC that processes outdoor air and is cooled with water from a cooling tower will come close to isothermally drying the air (i.e., it performs almost 100% latent cooling). This LDAC could meet the requirements of the earlier example (i.e., condition air from 86 dry-bulb and 78 F wet-bulb to 65 F and 50% rh) by directly performing 31.7 tons of latent cooling and then relying on a separate sensible cooling device for the remaining 11.5 tons of sensible cooling. In this application, the low-flow LDAC eliminates the 10.8 tons of overcooling. Of the remaining 43.2 tons of required cooling, only 11.5 tons has to be handled by a compressor-based refrigeration system, the remaining 31.7 tons being rejected to ambient by the cooling tower. Of course the latent load served by the LDAC and cooling tower imposes an energy cost: the thermal energy that is needed to regenerate the desiccant.
The Implementation of Low-Flow Liquid-Desiccant Technology
Low-flow liquid-desiccant technology can be the basis for several innovative HVAC products. As described elsewhere on this website, a low-flow LDAC can convert a conventional compressor-based air conditioner into a cooling system that efficiently supplies latent cooling (i.e., a liquid-desiccant DX air conditioner). In the first section of this tutorial, low-flow liquid-desiccant technology was introduced in the context of a thermally driven LDAC. The most efficient gas-fired version of a thermally driven LDAC will use a two-stage regenerator for removing water from the desiccant. All thermally driven LDACs should use an interchange heat exchanger (IHX). For LDACs that use halide salts such as lithium chloride and calcium chloride, expensive corrosion-resistant metals can be avoided by implementing the IHX as a plastic-plate heat exchanger.
For thermally-driven LDACs that use a single-stage regenerator, the regenerator is typically supplied with a hot fluid above 160 F, but preferably above 180 F. As described in the section for the single-stage, scavenging-air regenerator, higher fluid temperatures will significantly increase the water-removal capacity of the regenerator.
The ability of a scavenging-air regenerator to efficiently operate at relatively low temperatures creates interesting opportunities to provide latent cooling with minimal use of fossil fuels--and minimal greenhouse gas emissions. Several types of solar thermal collectors will provide water sufficiently hot to make a solar LDAC practical. The on-site production of electricity using either engine-generator sets or fuel cells, referred to as combined heat and power, often provides thermal energy at a temperature that can effectively drive an LDAC.
The heat rejected by an LDAC can sometimes be used to offset energy for other end-uses. When an LDAC is used to dehumidify a pool, natatorium or water park, the pool water can be used as the LDAC conditioner's coolant. Since the pool's need for heating is created mostly by evaporation of water from the pool's surface there will be a close match between the heat rejected by the LDAC and the heat required to maintain the pool's temperature.
The dewpoint of the air leaving a scavenging-air regenerator increases as the temperature of the hot fluid increases and the desiccant concentration decreases. The dewpoint of the scavenging air leaving a regenerator that is supplied with 210 F hot water and 35% LiCl would be on the order of 135 F. This hot, humid air can provide some or all of the thermal energy required by a domestic hot water system.
Perhaps the greatest impact of low-flow liquid desiccant technology will occur when it is integrated with a dewpoint indirect evaporative cooler (DIEC) to produce an air conditioner that efficiently delivers both sensible and latent cooling with minimal electrical demand. This integration of an LDAC and a DIEC has been named DEVap by the researchers at the National Renewable Energy Laboratory who first studied the concept.