A Desiccant-Enhanced Evaporative Cooler (DEVap) is one of the more promising technologies for dramatically reducing on-peak electrical demand for air conditioning.  In place of a compressor-based refrigeration cycle that pumps heat out to ambient, DEVap rejects heat to ambient via evaporative cooling.  

     DEVap combines a deep drying, liquid-desiccant stage that provides latent cooling with an indirect evaporative cooler (IEC) that provides sensible cooling.  The IEC is the key to cooling without a compressor.  Using principles first patented by William Niehart in 1939, DEVap's IEC "super charges" sensible cooling by first cooling the air that drives the evaporation.

     Similar to a conventional device, Niehart's IEC is an air-to-air heat exchanger with alternating wet and dry passages.  The wet passages have films of water on their surfaces.  The cooling air flowing through these passages evaporates water, reducing the temperature of the water films towards the cooling air's wet-bulb temperature.  The process air flowing through the dry passages is then "indirectly evaporatively cooled" by heat exchange with the low-temperature water films of the wet passages.  

      Niehart's innovation is to divert some of the cooled process air into the wet passages.  With this cooling air itself being cooled, it enters the wet passages with its wet-bulb temperature significantly lower than its initial state at the inlet to the dry passages.

     There is a lower limit to the wet-bulb temperature of the cooling air that enters the wet passages, and this limit occurs when the process air leaves the dry passages at 100% rh.  But at 100% rh, the air's dry-bulb, wet-bulb and dewpoint are all the same, so the lowest temperature that a Niehart device can supply equals the air's initial dewpoint (which is the one characteristic temperature of the three that does not change as the process air is cooled towards 100% rh). 

      Niehart's design for his Dewpoint Indirect Evaporative Cooler (DIEC) is shown in the original patent drawings at the left.  The drawing of Woods and Kozubal for DEVap that appear on the right illustrates the operation of both the first-stage, liquid-desiccant dehumidifier and the second-stage DIEC.  

       The energy efficiency of DEVap will be degraded by large "cooling fractions", i.e., the fraction of process air leaving the second-stage IEC that turns to become cooling air.  Unlike a stand-alone DIEC with the ambient as the source of the process air, for DEVap, the process air for the second-stage is the deeply dried air supplied by the liquid-desiccant first stage. This is expensive air in that  there is an energy cost to produce the strong desiccant that does the deep drying in the first stage.

      The three key components of DEVap—the liquid-desiccant dehumidifier, the DIEC and the desiccant regenerator—are all commercially available.  However, the available technologies all have limitations that prevent the DEVap cycle from coming close to its performance potential.  Our Deep Dive pages for an Electrically Driven Desiccant Regenerator (EDDR) and for a Wicking-Fin Heat and Mass Exchanger describe technology that could be the desiccant regenerator and the liquid-desiccant dehumidifier for DEVap.  Our patented design for a DIEC—which is next described—is the final essential component.

      We start our discussion of the DIEC by describing its "ideal" performance, i.e., its operation at the lowest cooling fraction at which the DIEC supplies air at a dry-bulb temperature equal to its dewpoint temperature.  This ideal DIEC operates with process air and cooling air flowing counter to each other.   The thermodynamic state of the process air leaving the dry passages is known—its dry-bulb temperature equals its dewpoint temperature.  Assuming a specified flow of process air, the rate of sensible cooling that reduces the process air's temperature to this final state can be calculated.  

     On the "wet" side of the DIEC, the cooling air gains the heat lost by the process air.  With the initial thermodynamic state of the cooling air known (i.e., it equals the state of the process air leaving the dry passages), the mass flow of cooling air will be a minimum, when its final state leaving the wet passages has maximum energy (i.e., maximum enthalpy).  Since the maximum dry-bulb temperature for the existing cooling air is the process air's entering dry-bulb temperature, maximum enthalpy occurs at the air's maximum humidity (i.e., 100% relative humidity).  

     Applying the preceding requirements for ideal performance, we calculate a minimum cooling fraction of  0.25 when the process air enters at 96.5/55°F DB/DP (i.e., representative conditions of air supplied by a wicking-fin liquid-desiccant first-stage that processes outdoor air on a summer design day.)  This 0.25 ideal cooling fraction is clearly unattainable—but what is a reasonable target in practice?

      There are only a few commercial sources for DIECs.  DIECs based on the Maisotsenko-cycle are sold by at least two vendors (Seeley International and DRI), and they probably now (2025) have the largest share of what is a small market.  In typical applications for Maisotsenko-cycle DIECs, outdoor air is processed, and so the cooling fraction is of secondary importance (but not insignificant since it does effect product size and water use).  With cooling fractions reported in product literature on the order of 0.45, a Maisotsenko-cycle DIEC is ill-suited to be part of DEVap.

     The rate and manner that water is supplied to a DIEC can have a profound effect on the cooling fraction required to meet target performance.  Most evaporatively cooled technologies specify a minimum water flooding rate that insures complete and  adequate wetting of surfaces where evaporation occurs.  Furthermore, since evaporatively cooled technologies typically are supplied water with dissolved minerals, they must purge (or blowdown) some of the recirculating water so that mineral concentrations do not exceed levels where scale forms on the evaporating surfaces.

      The effectiveness of a counter-flow heat exchanger can be degraded by parasitic mechanisms that weaken the wall temperature gradient in the direction of the flows.  For a conventional heat exchanger, thick, conductive walls can "leak" heat that flattens the wall temperature gradient. 

     In U.S. Patent 8,636,269,  James recognized that flooding a DIEC from a single sump could also flatten the wall temperature gradient and degrade performance.  James addressed the problem by dividing the DIEC into segments along the direction of air flow with each segment having its own water recirculation sump and pump.  During operation of James' DIEC, the water temperature in the sumps closely match the local wet-bulb temperature of the air, which then suppresses flow-wise heat transfer from the water.

      All heat and mass exchangers are penalized by low heat and mass transfer coefficients—more surface area is required to reach target effectiveness, units get larger and more expensive, pressure drops increase.  The problem will be particularly acute for a DEVap DIEC, where the higher volume flow of air through the dry passages require wider passages.  The problem is further complicated by the practical requirement to keep all spacing elements on the dry side where they do not interfere with the water films on the walls of the wet passages.  

      In our U.S. Patent No. 10,739,079 for a DIEC, we convert the spacing elements in the dry passages into extended surfaces for heat transfer, i.e., sheets of  4-mil, aluminum, split fins (which are shown in the banner photo for this page) are installed in the dry passages of the DIEC.  However, a design that simply bonds the fins to aluminum skins, as shown in photo at the right, will be very fragile and easily damaged.  Our patent claims a much more robust, but equally effective thermally, design in which the aluminum fin sheets are protected within tough polymeric frames.  This design approach is shown both in an isometric drawing and a photograph at the right.

      As noted earlier, DEVap has three main components: a liquid-desiccant dehumidifier, a DIEC and a desiccant regenerator.  In a DEVap application, there is an interesting synergy between our Electrically Driven Desiccant Regenerator and our DIEC.  This is illustrated in a DEVap that processes 100% outdoor air on a summer design day (95/78°F DB/WB). 

       Our first stage for this DEVap device is a 3,600-cfm Wicking-Fin Liquid-Desiccant Conditioner (WF-LDC) that is cooled by water from a cooling tower.  When supplied with 39.4% LiCl and operating with a 295 fpm face velocity, this WF-LDC with an 8-tube deep core will condition outdoor air from 95/78°F DB/WB to 96.5/55°F DB/DP. The liquid desiccant absorbs 125 lb/h of water.

      We project a regeneration efficiency of 12.9 lb/kWh for an EDDR using our diffusion-gap technology that returns the 35.4% LiCl discharged by the WF-LDC to full concentration.  The refrigerant for the EDDR is R1233zd(E), and it operates with a 90°C​ condenser and 55°C​ evaporator.  Fundamental to the operation of our EDDR is 100% recovery of water as mineral-free liquid—125 lb/h.

      The second-stage for this DEVap is a DIEC with a design consistent with the 10,739,079 patent. There are trade-off to be made when designing the DIEC.  One of the more important is between the physical size of the DIEC and the amount of cooling air required to meet a target performance. 

     In our DEVap design study, the DIEC cools the warm air leaving the WF-LDC down to 75°F (which avoids a post cooling penalty under AHRI Standard 920 for DOASs).  Because the exhausted cooling air is a direct parasitic penalty on DEVap (i.e., the cooling air must be deeply dried by the WF-LDC), we've specified a low 0.20 cooling fraction for the DIEC*.  With this cooling fraction, we predict that a DIEC operating at a face velocity of 340 fpm (based on air supplied to the building) with plates that have an active depth of 27 inches will supply air at the required 75/55°F DB/DP.  At this operating point (the A Rating Point for a DOAS under AHRI Standard 920), the MRE based on EDDR performance will be 10.8 lb/kWh.

    The water evaporation rate on the wet side of the DIEC is 77 lb/h.  With 125 lb/h or mineral-free water available from the EDDR, the DIEC is designed to operate with essentially no over-flooding of its wetted surfaces, i.e. the flooding rate equals the evaporation rate.  This low flooding rate will have almost no effect on the wall temperature gradient.      

     In the psychrometric chart on the right, we compare the state-point performance of DEVap with a conventional DX-DOAS and the LD-DOAS that uses condenser air to regenerate the desiccant.  Focusing on the main demand for power—the compressor—the MRE for each alternative after applying a penalty for post cooling supply air to 75°F is 10.8, 6.67 and 4.51 lb/kWh for DEVap, LD-DOAS and DX-DOAS respectively.  (In applications where supplying very warm, but dry ventilation air is acceptable, the MRE for DEVap would increase to 12.9 lb/kWh, and for the LD-DOAS, 8.69 lb/kWh.)

​​​