CHAPTER 2 LITERATURE REVIEW

CHAPTER 2 LITERATURE REVIEW Desiccant cooling systems have been studied by many researchers in last two decades. Liquid desiccants can either absorb moisture from or add moisture to the air which depends upon the difference of vapour pressure between the air and the desiccant solution. The vapour pressure of the solution depends on its temperature and concentration. The higher the concentration and lower the temperature, lower will be the vapour pressure and higher will be the moisture absorbed from the air. Absorber and regenerator are the most crucial components of the liquid desiccant cooling system. Absorber is the component where dehumidification of air takes place and moisture is absorbed by the desiccant. Desiccant solution becomes weak in the absorber and it is to be made strong in the regenerator for reuse where the air picks up moisture from the hot solution due to vapour pressure difference. A good contact of air and desiccant is necessary to absorb maximum moisture from air or to desorb the same from desiccant in the air-desiccant contact equipment i.e., absorber and regenerator. Liquid desiccant systems are designed to carry large amount of process air and low flow rates of the desiccant. Packing density in dehumidifier/regenerator should be high with small pressure drop. 2.1. LIQUID DESICCANT MATERIALS The liquid desiccants materials are characterized by their low vapour pressure which leads to moisture transfer due to vapour pressure difference between air and desiccant solution. The criteria for the choice of desiccant material for a specific application are determined by a number of factors which includes cost, chemical properties, energy storage density and regeneration temperature. Liquid desiccants should have less viscosity and good heat transfer characteristics. It is desirable that these materials should be non-corrosive, non-inflammable, stable, readily available and inexpensive. The surface tension of the liquid desiccant effects static hold up and surface wetting in the dehumidifier. 15

Different liquid desiccants generally utilized are calcium chloride, tri-ethylene glycol, lithium chloride, lithium bromide and mixture of these solutions. Other desiccants include KCOOH [Longo & Gasparella (2005)], glycols like MEG, DEG and propylene glycol. Vapour pressure of these desiccants is function of temperature and concentration as compiled from Dow Chemical Company (Dow, 2003) and M. Conde Engineering (2004). The desiccants selected by different researchers are tabulated in table 2.1. Table 2.1 Researchers for various liquid desiccants systems Sr. Name of Liquid No. Desiccant Researchers 1 Lithium Bromide Factor and Grossman (1980), Andberg and Vliet (1983), Patnaik et al. (1990), Jain et al. (1994), Jain et al. (2000a,b), Lazzarin et al. (1999), Oliveira et al. (2000), Liu (2006), Salarian (2011), Bouzenada (2014) etc. 2 Lithium Chloride Marsala et al. (1989), Chung et al. (1993), Chung (1994), Chung and Ghose (1996), Kessling et al. (1998), Yin et al. (2006), Huang et al. (2012), Bakhtiar et al. (2012) etc. 3 Calcium Chloride Rahamah et al. (1998), Alizadeh & Saman (2002), Fumo and Goswami (2001), Sultan et al. (2002) Bassuoni (2011), Hammad et al. (2008), Bouzenada (2014a), Seenivasan et al. (2014) etc. 4 Tri-ethylene Glycol Chung (1994), Chung and Wu (1998), Oberg and Goswami (1998), Abdul-Wahab et al. (2003), Zurigat et al., (2003), Esam Elsarrag (2005), Esam Elsarrag (2006), Asati (2007) etc. 5 Mono-ethylene Glycol Factor and Grossman (1980) 6 Propylene Glycol Chung et al. (2000), Niagara Blower (2006) 7 KCOOH Longo & Gasparella (2005) 16

Factor and Grossman (1980) have reported a practical problem with mono ethylene glycol. After several weeks of experimentation, they observed that the parts of apparatus had been coated with oily droplets of mono ethylene glycol, which had condensed out from air. So they recommended the use of TEG as a desiccant, which has a lower vapour pressure. Chung et al. (1995) and Chung et al. (1993) have reported that some desiccants are also capable of absorbing inorganic and organic impurities in the air and thus, refined quality of the indoor air. Their experimental findings show that a 95% by weight tri ethylene glycol solution purifies the air containing pollutants of formaldehyde (0.02 ppm), toluene (3 ppm), 1,1,1-trichloroehane (24 ppm) and carbon dioxide (1000 ppm) by removing all toluene and 1,1,1-trichloroethane, 56% of the carbon dioxide and 30% of formaldehyde [Chung et al. (1995)]. Some of the desiccants have safety issues that need to be clearly understood. Tri ethylene glycol has a low vapour pressure and low surface tension and it is not expected to present inhalation hazards at ambient conditions. It is also non-toxic but vapour from warmed or heated surface may be a mild respiratory irritant. So, caution has to be taken to prevent it from forming aerosols or mist. Ertas et al. (1992) explored mixture of CaCl 2 and LiCl with different weight combinations to reduce the high cost of lithium chloride. Data for vapour pressure, density, viscosity and solubility of CELD (Cost effective liquid desiccant) solution has been reported. The vapour pressure of CELD solution has been found to be lower than that of pure calcium chloride with low viscosity and high solubility over a considerable temperature range. Ameel et al. (1995) compared the various liquid desiccants viz. LiBr, LiCl, ZnCl 2, Li 2 ZnCl 4, Li 2 CaCl 4 and CaZnCl 4 on the basis of predicted performance and cost. The absorber model of Grossman (1983) was employed to simulate simultaneous heat and mass transfer procedures. They reported that the most promising absorbent was a blend of two parts of LiCl and one part of ZnCl 2 ; which was eight times cheaper than lithium bromide. The capacity of absorption per unit dehumidifier area for reported mixture was 50-70% less than that of lithium bromide but at the same time pumping power was also less. The mixture reduced the risk of solidification of the solution in the absorber. 17

The performance of absorber and regenerator using LiBr as desiccant has been studied by Jain et al. (1994). They pointed out the highly corrosive nature of LiBr and suggested the use of other desiccant materials like TEG. Bharti (1994) carried out a comparative study of different liquid desiccants using overall system cost, corrosion, safety and desiccant cost as parameters. Different weightages were assigned to various parameters and each desiccant was rated for each of the selected parameter. Table 2.2 summarizes the results of the study, which concludes that TEG is the best desiccant followed closely by CaCl 2. Table 2.2 Relative weightage given to various desiccants from Bharti (1994) Characteristics Weighting Factor LiCl LiBr CaCl 2 TEG Overall system cost excluding desiccant 1.0 7 6 10 10 Corrosion 0.8 10 10 9 10 Safety 1.0 7 8 9 10 Desiccant cost 0.5 5 5 10 8 Total 3.3 24.5 24.5 31.2 32 Oberg (1998) reported that the salt solutions have higher surface tension compared to glycols. Hence adequate wetting of mass transfer surfaces will be more difficult when salt solution is used. The vapour pressure of pure TEG is not zero, resulting in its little evaporation into air. Chung and Luo (1999) experimentally measured the vapour pressures of potential desiccant solutions and their mixtures at typical operating concentrations (40% for salt solutions, 95% for glycols and mixtures 1 to 1 volume ratio) and in the temperature range of 25 o C to 40 o C. Gandhidasan at al. (2002) used CaCl 2 as liquid desiccant in hybrid desiccant cooling system and studied process parameters of the desiccant cooling system. Abdul-Wahab et al. investigated the performance of dehumidification system using TEG as liquid desiccant and stacked plates as packing. Moisture removal rate and 18

effectiveness of the dehumidifier were studied as the performance parameters. Zuirghat et al. (2004) also used try ethylene glycol as liquid desiccant to compare the performance of dehumidification system with two different structured packing materials i.e., wood and aluminium. Elsarrag (2007) studied the liquid desiccant regeneration system with try ethylene glycol as desiccant. MRR and effectiveness were investigated as the performance parameters of the structure packed dehumidification system. Hammad et al. (2008) used calcium chloride as liquid desiccant to study the reduction ratio of specific humidity of the air and the dehumidifier effectiveness. A decrease in both parameters was observed with increase in inlet parameters; cooling water inlet temperature, the desiccant solution temperature and the air flow rate. Bakhtiar et al. (2012) utilized LiCl as the liquid desiccant to discuss three performance assessment review factors. These were moisture removal rate, dehumidification effectiveness and COP of the system. They proposed a method which was also appropriate to assess the efficiency of system with low humidity effectiveness. Bouzenada et al. (2014) presented an experimental study of dehumidification/regeneration processes using LiBr as liquid desiccant in direct contact with the air at different operating conditions. Bouzenada et al. (2014a) also used CaCl 2 and calcium chloride dehydrates as desiccant. It was reported that the partial vapour pressure of process air was higher than the vapour pressure of desiccant solution at packing surface area, the moisture of air was absorbed into the desiccant solution and vice versa in case of lower vapour pressure of air and higher vapour pressure of desiccant. 2.2. PACKING MATERIALS Packing material surface is a medium for interaction of air and liquid desiccant in the absorber and regenerator. The desirable properties of different types of packings used in these columns are listed below: (1) large void volume to avoid excessive pressure drop (2) chemically inert 19

(3) large surface area per unit volume of packing (4) light weight and strong (5) good distribution of liquid (6) good wettability (7) low cost Mainly four types of packing materials are used in liquid desiccant cooling systems. These are structured type packing, random packing, spray type packing and falling film type packing. A brief study of these packing materials has been studied to seek better packing material for the desiccant-air contact equipments. 2.2.1. Structured Packing The more recent designs of packed columns use structured packing in counter/cross flow of the desiccant solution and air [Chung and Ghose (1996), Wahab et al. (2004), Elsarrag et al. (2005), Al-Farayedhi et al. (2002), Elsarrag (2006), Bassuoni (2011), Salarian (2011), Huang et al. (2012), Seenivasan et al. (2014)]. Chung and Ghose (1996) compared the efficiency of random and structured packings for dehumidification of process air in a packed column using LiCl as desiccant solutions. Both packings were found to be equally efficient in the dehumidifier to dehumidify the process air under same working environments. Pontis and Lenz (1996) compared random and structured packings with varying bed heights in a solar-based liquid-desiccant system. Mass-transfer rates for the randomly packed bed were about 300% and 130% larger in the regenerator and approximately 60% and 45% bigger in the dehumidifier than those for the structured packed beds with bed heights of 0.03 m and 0.55 m respectively. Gandhidasan et al. (2002) used structured packing with calcium chloride as liquid desiccant for 5 TR hybrid desiccant cooling system. They observed that when packing height is kept constant, mass flow rate and the absolute humidity of air at the exit decreases with increasing flow rate of desiccant solution and concentration at inlet of the desiccant solution. 20

Low density structured packing (77 m 2 /m 3 ) was used by Zurigat et al. (2004) using TEG as desiccant. The packings were made of wood or aluminum. An increase in humidity effectiveness was observed with increase in desiccant flow rate and inlet temperature. Wahab et al. (2004) extended the study to three structured packings with densities 77, 100 and 200 m 2 /m 3. The performance was studied in the terms of MRR and dehumidifier effectiveness. He also made statistical model to validate the performance of system by using the multiple regression method and the principal component Elsarrag et al. (2005), Elsarrag (2006) and Elsarrag (2007) reported studies on cellulose rigid media pads as structured packing with tri ethylene glycol as a desiccant. Elsarrag et al. (2005) found the reduction of carryover of TEG with the use of cellulose structured packing. They also developed and compared theoretical model with the experimental results. Elasarrag (2007) expanded the validity of results by using different values of liquid to air flow ratios. Liu et al. (2006) and Liu et al. (2007) used celdek packing in a cross flow dehumidifier and regenerator using LiBr as desiccant. Bakhtiar et al. (2012) used a lithium chloride as desiccant material. The packing height of structured packings in the column was 0.4 m with the cross section area of 0.16 m 2. 2.2.2. Random Packing The random packings have been studied extensively by various researchers [Factor and Grossman (1980), Patnaik et al. (1990), Chung et al. (1993), Pontis and Lenz (1996), Kinsara et al. (1997), Lazzarin et al. (1999), Oberg and Goswami (2000), Fumo and Goswami (2002), Longo and Gasparella (2005), Yin et al. (2006), Martin and Goswami (2000), Bakhtiar et al. (2012), Bouzenada (2014)]. Even random packings offer good contact surface between air and desiccant but high desiccant flow rate is required for good wetting in packings [Factor and Grossman (1980), Patnaik et al. (1990)] and therefore the air pressure drop and the electric power requirement for blowing the air is also large. Factor and Grossman (1980) performed experiments using mono-ethylene glycol and lithium bromide with 3/4-inch ceramic intalox saddles packing. Some 21

practical problems and inaccuracy in results with mono-ethylene glycol have been reported. Gandhidasan et al. (1986) studied the heat and mass transfer coefficient for the absorber and regenerator using ceramic raschig rings (1 inch) and berl saddles (1 & 1/2 inch) as packing and calcium chloride as liquid desiccant. Patnaik et al. (1990) designed and installed 10.5 kw open cycle LDCS, using tripack no. 1/2 polyethylene spheres as packings and lithium bromide as solution, with two types of distributor systems. The capacity of regenerator has been reported to increase from a maximum 6.7 kw with tray distributor arrangement to 10.5 kw with a nozzle distributor arrangement. Increase in absorber capacity was also observed with increasing mass flow rate of desiccant solution. Pressure drop is reported 0.52 mm of H 2 O per cm of packing depth, at the air mass flux of 4611 kg/m 2 /hr. Chung et al. (1993) and Chung (1994) used 1.6 cm polypropylene flexi rings packings with LiCl and TEG as desiccants. Elsayed et al. (1993), Radhwan et al. (1993), Elsayed (1994) and Kinsara et al. (1997) developed finite difference models to study the heat and mass transfer in packed beds. The effectiveness of heat and mass transfer for a number of bed heights, air flow rate, desiccant flow rate, inlet temperatures of the air and desiccant solution, inlet concentration of CaCl 2 solution was predicted by the model, thus prepared charts of effectiveness. The COP and COP max of liquid desiccant system were expressed in terms of operating variables based on the effectiveness method. Oberg and Goswami (1998) used 2.54 cm polypropylene rauschert hiflow rings packing using TEG as desiccant and identified the key design variables such as air flow rate, specific humidity, temperature and concentration of desiccant solution, and height of packing which have the greatest impact on the performance of the dehumidifier. Lazzarin et al. (1999) investigated the chemical air dehumidification by H 2 O/LiBr and H 2 O/CaCl 2 in a packed tower for air conditioning applications. The ranges of variables used for experimentation were air flow rate of 220 m 3 /hr, ṁ s ranging from 0.0183 kg/s to 0.1297 kg/s, T ai from 23.6 o C to 35.4 o C, ω ai from 10.4 g/kg to 18.7 g/kg d.a., T si from 16.1 o C to 13.1 o C and C si (LiBr) from 53% to 57%. The maximum and minimum moisture reduction were found to be 12 g/kg da and 3 g/kg d.a. respectively. The problem of crystallization at higher concentration and at 22

low inlet temperature of desiccant was observed. The value of pressure drop in the absorber was not reported. A theoretical simulation model was developed that over predicted the experimental data within absolute mean deviation of around 20%. Doan and Fayed (2000) investigated the effect of bed height and entry segment under the packing support on mass transfer in a 1.2 m diameter packed bed of 50 mm ceramic intallox, with gas flow rates varying from 1957 kg/h.m 2 to 7828 kg/h.m 2 and liquid flow rates varying from 12200 kg/h.m 2 to 46700 kg/h.m 2. Two packing heights of 1.8 m and 0.91 m were studied by them. The entrance effect was observed to be independent from the packing height. It was found that entrance effect was 17% of the overall water vapour transported from the air to desiccant solution. For the packing height of 1.8 m, average mass-transfer coefficient was found to be directly proportional to the gas flow rate raised to the power of 0.89 and was found to be independent of the flow rate of desiccant flow rate. For the packing height of 0.91 m, average mass-transfer coefficient was found to be directly proportional to the gas flow rate and desiccant flow rate was raised to the power of 0.24. Fumo and Goswami (2002) used lithium chloride with packings similar to Oberg and Goswami (1998) and identified the design variables which have the greatest impact on the humidity effectiveness and moisture removal rate of the dehumidifier. The performance of system was observed to be dependent on mass flow rate, inlet temperature and concentration of desiccant solution. Longo and Gasparella (2005) tested random packings of 25 mm plastic pall rings with two conventional desiccants viz. LiCl, LiBr and a new desiccant KCOOH. LiCl and LiBr were shown to have similar absorber performance (dehumidification rate 6-7 g/kg) which was higher than KCOOH (dehumidification rate 5 g/kg) whereas KCOOH solution showed a better regenerator performance compared to LiCl and LiBr solutions. Dehumidification effectiveness was observed to be varied from 30% to 90% with ratio of air to desiccant flow rate within the range 0.2 to 3.0, while regeneration humidity effectiveness was found to be varied between 20% and 75% with a flow rate ratio within the range 0.3 to 3.5. Salarian (2011) studied the size and performance of a dehumidifier by simulating changing working conditions. Depending upon the experimental observations, he presented the performance of a packed tower absorber for a LiCl desiccant dehumidification system. 23

2.2.3. Spray Type Packing Liquid desiccant solution is converted into very fine droplets by nozzles in the spray compartment which is absorber/regenerator tower. In fact there is no separate packing material but it is full of fine droplets and outer surface of droplets of desiccant provide contact area to process air. So, entire space of tower can be supposed as packing, therefore, named as spray type packing. Spray type towers are famous for low cost, simplicity and low pressure drop on air side. The effectiveness of spray towers is low when compared with other types of towers. Pressure drop on liquid side in these types of towers is high. Sprayed coils [Scalabrin and Scaltri (1990) and Chung and Wu (2000)], cross flow heat exchangers and heat pipes [Pesaran et al. (1995)] are some of the schemes utilized in spray towers. Gandhidasan and Gupta (1977) and Gupta and Gandhidasan (1978) developed a liquid desiccant system of 3 TR capacity using CaCl 2 solution (47 %) as absorbent. The system had a forced flow solar collector cum regenerator [Mullick and Gupta (1974)] and a spray chamber as absorber. The reported COP was 0.2 at ambient condition of 33.3 o C DBT and ω a = 22.3 g/kg of d.a. The conditioned room was maintained only at 29.8 o C DBT and ω a = 18 g/kg of d.a. Pesaran et al. (1995) used finned tube heat pipe recovery unit as desiccant/ air contacting device with TEG and LiCl as liquid desiccants. Exhaust air from room was used to remove the heat of absorption released during dehumidification process. Heat pipe with liquid desiccant provided 20-40% more cooling in comparison to pipe with no desiccant, whereas LiCl provided 10% more cooling than TEG. Chung and Wu (1998) reported the efficiency of spray tower in the range of 64 to 86%, while using TEG as desiccant. The maximum efficiency of 86% was achieved for the operating conditions given as: ṁ a = 1.94 kg/min, ṁ sol = 3.31 kg/min, T ai =25.4 o C, T ao =26.5 o C, ω ai = 17.2 g/kg of d.a., ω ao = 6.5 g/kg of d.a., T si = 23.3 o C, T so =28.3 o C; and C s = 95.2%. This set of data showed that the liquid desiccant to airflow ratio was high i.e. 1.7. Chung et al. (1999) fabricated a U-shaped stripping tower and successfully tested it for regeneration of TEG. They proved that the MRR increases with increase in air flow rate and desiccant flow rate. The increase in overall mass transfer 24

coefficients was observed with increase in both air and desiccant mass flow rates. The performance of stripping tower was also increased with desiccant temperature. Chung and Wu (2000b) compared efficiency of spray towers dehumidifiers with and without finned cooling coils with the use of TEG solution. The maximum and minimum efficiencies without the finned cooling coil have been reported to be 0.58 and 0.41, whereas with the finned cooling coil the same have been reported to be 0.82 and 0.60 respectively. It has been concluded that the absorption efficiency was directly proportional to solution flow rate and indirectly proportional to air flow rate. The overall mass transfer coefficient increases with increasing mass flow rate of the air and liquid. 2.2.4. Falling Film Type Packing In wetted wall or falling film columns, thin film of liquid desiccant flows down the inside of a vertical tube or plate, with air flowing in either co-current or counter-current direction. Actually, falling film is considered as fictitious packing corresponding to shape of falling film surface. The wetted wall tower may be surrounded with flowing cooling water. Falling film column has little pressure drop [Treybal (1981)], less preliminary price. These columns deliver high contact surface area per unit volume. It is difficult to obtain stable film on the entire surface area of big dehumidifying towers [Jain et al. (2000)]. With proper distribution system in absorber and regenerator, these systems appear to be preferably suitable for liquid desiccant based cooling system. Falling film designs have been studied by Rahamah et al. (1998), Jain et al. (2000a,b) etc. Jain et al. (1994) carried out experimental and simulation studies on falling film tubular type dehumidifier and falling film plate type regenerator using LiBr as desiccant. The maximum moisture removal in the absorber and the maximum moisture picked up in regenerator were experimentally found to be 5 g/kg d.a. and 8 g/kg d.a. respectively. A comparison with simulated results showed that the predicted heat and mass transfer were considerable greater than the experimental results, which was attributed to inadequate wetting of dehumidifier tubes and regenerator plates. Two different wetness factors f m and f h were introduced to accomplish the 25

difference between simulated and experimental results. f m was the fraction of total area available for mass transfer and f h is the fraction of total area available for heat transfer. The values of these wetness factors for mass and heat transfer were estimated by matching experimental results with computer simulations. These varied from 0.45 to 0.8 for the absorber and 0.6 to 1.0 for the regenerator. Thus there is a scope for considerable improvement in the design of these columns. Deng and Ma (1998) investigated the features of a falling film dehumidifier. They concluded that the mass transfer increases with increase of spray density; though increase in heat transfer coefficient was observed only within lesser spray density range (0.005-0.055 kg/m 2 -s). Alizadeh and Saman (2002) designed and tested a solar collector cum regenerator for Adelaide summer conditions using CaCl 2 as desiccant. The experiment proved that the performance of the forced flow collector/regenerator was increased with increase in mass flow rate of air and with decease in solution mass flow rate. The evaporation rate was increased with increasing the air and solution temperatures at the inlet of the regenerator. Kim et al. (2003) investigated the effect of micro scale treatment on the smooth tube. The wettability and vapour absorption rate of micro-hatched tube were found to be 10% and 21% more than smooth tube respectively. Islam et al. (2003a) calculated the heat and mass transfer coefficients for falling film tubular (horizontal) dehumidifier with LiBr-H 2 O solution. It has been reported that both heat and mass transfer coefficients increase with increasing temperature of cooling water at inlet of dehumidifier. Kabeel (2005) explored the regeneration of desiccant (CaCl 2 ) in cross flow of air stream and desiccant solution using a solar collector/regenerator. Kumar et al. (2008) simulated and analyzed the steady-state performance of stand-alone liquid desiccant systems. Falling film designs of absorber and regenerator had been selected for the study due to their lower pressure drops. Hammad et al. (2011) used a block of twenty copper tubes with 1220 mm length, 16 mm outer diameter and 1 mm thickness were welded in each front and back movable flanges and calcium chloride was used as liquid desiccant. Table 2.3 gives a brief review of various packings used by different researchers in past. 26

Table 2.3 Researchers of various packing materials Name of Packing Structured Packing Random Packing Falling Film Type Spray Type Packing Researchers Chung and Ghose (1996), Sultan et al. (2002), Abdul-Wahab et al. (2003), Zurigat et al., (2003), Esam Elsarrag (2005), Esam Elsarrag (2006), Liu and Jiang, Liu (2006), Bassuoni (2011), Salarian (2011), Huang et al. (2012), Seenivasan et al. (2014) etc. Factor and Grossman (1980), Patnaik et al. (1990), Chung et al. (1993), Chung (1994) Pontis and Lenz (1996), Kinsara et al. (1997), Lazzarin et al.(1999), Oberg and Goswami (2000), Fumo and Goswami (2001), Fumo and Goswami (2002) Gandhidasan et al. (2002), Longo and Gasparella (2005) Yin et al. (2006), Martin and Goswami (2000), Bakhtiar et al. (2012), Bouzenada (2014) etc. Jain et al. (1994), Jain (1999), Jain et al. (2000a,b), Alizadeh and Saman (2002), Islam et al. (2003a), Islam et al. (2003b) Kim et al. (2003), Kabeel (2005), Asati (2007), Hammad et al. (2008), Bouzenada (2014a) etc. Scalabrin and Scaltri (1990) and Chung and Wu (2000), Pesaran et al. (1995), Gandhidasan and Gupta (1977), Chung and Wu (1998), Chung et al. (1999), Chung and Wu (2000b) etc. 2.3. EXPERIMENTAL STUDIES ON LIQUID DESICCANT SYSTEMS Dehumidifier and regenerator are required as desiccant-air contact equipment in liquid desiccant cooling systems. In dehumidifier, air is brought in contact with the liquid desiccant and water is absorbed from the air into the desiccant. As the water is absorbed, the latent heat of condensation of the water, as well as the heat of mixing, is evolved. The desiccant solution absorbs the moisture and its concentration reduces. 27

During the process of absorption, water vapour in air gets condensed and becomes a part of desiccant solution, releasing the latent heat of condensation and heat of mixing. Patnaik et al. (1990), Chung (1994), Oberg et al. (1998), Gandhidasan et al. (2002), Elsarrag et al. (2005), Gommed and Grossman (2007) etc. reported studies on absorbers without any cooling. Deng and Ma (1999), Islam et al. (2003a & b), Jain et al. (2000a), Mesquita et al. (2006) provided simultaneous cooling in absorber by circulating cooling water. A regenerator is used to re-concentrate the weak desiccant solution flowing out of the absorber. Here, the desiccant solution is heated so as to reverse the mass transfer potential. Due to the higher water vapour pressure of desiccant, it loses moisture to the scavenging air passed through the regenerator. In adiabatic designs, the solution or air is heated before it enters the regenerator. Gandhidasan (1990), Patnaik et al. (1990), Sultan et al. (2002) reported studies on regenerators where the air is heated before its entry to regenerator. Many researchers have used adiabatic regenerators where the solution is heated before it enters the regenerator e.g. Factor et al. (1980), Chung et al. (1993), Ertas et al. (1994), Khan et al. (1994), Chung et al. (1995), Oberg et al. (1998), Chung et al. (2000), Sultan et al. (2002), Ghaddar et al. (2003). Simple trickle solar collectors have been used for the regeneration of desiccant by Gandhidasan and Al-Farayedhi (1995). Factor and Grossman (1980) reported that bed performance was get enhanced markedly by precooling and preheating of the air stream, in absorber and regenerator. It has been mentioned that satisfactory regeneration is accomplished at 80 o C regeneration temperature and effective dehumidification is accomplished with desiccant temperature 25-30 o C. The pressure drop has been measured to be 0.8 mm of H 2 O per centimeter of packing at an air flux of 10000 kg/hr.m 2. Patnaik et al. (1990) designed a 10.5 kw open-cycle liquid desiccant dehumidifier. A tray distribution system and a spray nozzle distribution were used to distribute liquid desiccant over the packing. 30-40% decrease in pressure drop and 40-50% increase in capacity were observed with spray distribution system. Cooling capacities of 3.5-14.0 kw were attained for regenerator and dehumidifier. Correlations for moisture removal rate for dehumidifier and regenerator were acquired by statistical method. 28

Oberg and Goswami (1998) studied that mass flow rate of solution (ṁ s ) and the inlet temperature of air (T ai ) do not have significant effect on the dehumidifier performance, as long as m s is sufficient to confirm wetting of the packing. The reported pressure drop across the packed bed varies in the range 30 to 210 Pa/m of packing, depending on the air flow rate. Sultan et al. (2002) investigated the operation of packed tower regenerator. A theoretical model was studied to demonstrate the effect of inlet parameters of air and desiccant. It was observed that regeneration process was extremely dependent on the inlet air temperature, specific humidity and flow rate of air. Other parameters studied were liquid temperature, concentration and flow rate of desiccant. Fumo and Goswami (2002) proposed a solar based liquid desiccant cooling system with packed tower dehumidifier and regenerator and LiCl as desiccant solution. The system provided low pressure drop and good heat and mass transfer characteristics for compact design. The performance of system in terms of moisture removal rates and effectiveness of dehumidifier and regenerator was studied by varying inlet parameters; air flow rate, desiccant flow rate, temperature and specific humidity of air, temperature and concentration of desiccant solution. Gandhidasan et al. (2002) investigated a structured packed dehumidification system which was suitable for 5 ton hybrid cooling system. Two elements of structured packing with height of 17 cm each and aqueous solution of CaCl 2 was used in the cooling system. They presented a complete experimental study of the heat and mass transfer processes between a liquid desiccant and air by considering the effects of various inlet parameters on the performance of dehumidification system. Zurigat et al. (2004) studied the influence of design parameters on the performance of dehumidifier with two low density structured packing of wood and aluminum using TEG as desiccant. It has been observed that the MRR is increased with increasing air flow rate, concentration and flow rate of desiccant solution for both packings. The increase in MRR was observed with increasing air inlet temperature for aluminum packing only. Elsarrag et al. (2004), Elsarrag et al. (2005), Elsarrag (2006) and Elsarrag (2007) studied the effect of air flow rate, solution flow rate, specific humidity of air and the vapour pressure of desiccant solution on the humidity effectiveness, humidity ratio reduction and the wet bulb temperature reduction of the column. It has been 29

found that greater temperature of desiccant solution is required for the regeneration of desiccant in moist environments compared to dry environments. Further, correlations for heat and mass transfer coefficients have also been proposed for the given design. Longo and Gasparella (2005) compared the performance of absorption/desorption columns (having random packings) using conventional LiCl and LiBr solutions with the new desiccant KCOOH. The new solution H 2 O/ KCOOH was found to perform better during regeneration, but poorer during dehumidification, when compared with the solution of lithium chloride and lithium bromide. The solution H 2 O/ KCOOH was also reported to be less corrosive and expensive than traditional desiccants and ecofriendly. Liu et al. (2006) and Liu et al. (2007) investigated that the MRR in dehumidifier increased with increasing air flow rate, solution flow rate, air inlet specific humidity and inlet concentration of desiccant. The MRR was observed to be decrease with increasing inlet temperature of desiccant solution and altered very little with inlet temperature of air. The dehumidifier effectiveness was found to be directly proportional to flow rate, inlet temperature of desiccant solution and inversely proportional to air flow rate. The effect of concentration of desiccant, inlet temperature and specific humidity of air was very little on MRR. Kumar et al. (2008) simulated falling film designs of dehumidifier and regenerator by using mass and energy balance equations. Fourth order finite difference Runge-Kutta method was used to find out solution of these non-linear first order differential equations. Warner s technique has been used for simulation of the system. They proposed two new stand-alone liquid desiccant cycles using the potential of desiccant fully through multiple absorbers. Improve in COP was observed for both the new cycles. Hammad et al. (2008) studied the reduction ratio of air humidity ratio and dehumidifier effectiveness for a dehumidification system. Increase in both parameters was observed with desiccant solution flow rate, ratio of desiccant solution flow rate and air flow rate, desiccant concentration and heating water temperature. The effect of inlet air temperature was almost negligible on the performance parameters. When air inlet specific humidity was increased, reduction ratio of air humidity ratio was slightly increased and no change in effectiveness was observed. 30

Salarian (2011) studied the moisture removal rate by varying air flow rate, liquid desiccant flow rate and inlet temperature of the air. The optimum height of packing in dehumidifier was found with the help of a finite difference model using MATLAB. Bakhtiar et al. (2012) evaluated coefficient of performance of desiccant air dehumidification system. They offered a solution over lacunaes of common methods with a proposed novel method to calculate the performance of liquid desiccant dehumidification system. Bouzenada et al. (2014) made analysis of the mass transfer in order to prove that material was the best liquid desiccant for LDCS. The experimental results have shown the effect of air conditions on mass transfer. It was observed that LiBr was able to absorb moisture and could be regenerated at low temperature. Bouzenada et al. (2014a) experimentally studies the mass transfer during the dehumidification/regeneration operation using the hygroscopic material calcium chloride and calcium chloride dehydrates as desiccant. These desiccants were compared on the basis of the same operation conditions. Main influencing factors were relative humidity of the air and temperature of regeneration process. It was observed that CaCl 2.2H 2 O was more rapidly diluted than the CaCl 2 at the same time during dehumidification, while the CaCl 2 solution can be more dried than CaCl 2.2H 2 O solution during regeneration. CaCl 2 was able to absorb more moisture in the cycle of Liquid desiccant cooling system. 2.4. PERFORMANCE OF LIQUID DESICCANT SYSTEMS The performance of absorber and regenerator involves both heat transfer and mass transfer processes, therefore their effectiveness is also studied in terms of humidity ratio and enthalpy. The performance of the dehumidifier and regenerator is assessed in form of effectiveness. The dehumidifier effectiveness, ε d, is an air-side characteristic parameter and it is related to the mass transfer effectiveness during the dehumidification process. It is defined as the ratio of the actual difference in moisture content of the air leaving the dehumidifier to the maximum possible difference in moisture content in specified operating environments as described by Ullah et al. 31

(1988), Chung (1994). The dehumidifier effectiveness, ε d, is expressed mathematically as: d i o (2.1) i eq The regeneration effectiveness, ε r, is also an air-side characteristic factor and is correlated to the mass transfer effectiveness through regeneration process. It is defined as the ratio of the actual difference in moisture content of the air leaving the regenerator to the maximum possible difference in moisture content in specified working circumstances. The regeneration effectiveness, ε r, is expressed mathematically as: r o i (2.2) eq i The vapour pressure of desiccant solution at inlet of the regenerator depends upon operating conditions of desiccant solution. It can be greater/lower than its vapour pressure at outlet of the regenerator. Therefore, evaluation of the regeneration operation may also be based on desiccant concentration instead of its vapour pressure. The regenerator effectiveness using heated air is defined as the ratio of actual difference in the concentration of the desiccant to the maximum possible difference in the concentration of the desiccant [Al-Mutairi (2000)] as given by equation 2.3. He has taken the maximum concentration of the desiccant in the equation, as a fixed value independent of air and solution operating conditions. For CaCl 2 and LiBr, the maximum concentration values are taken as 50% and 45% respectively, to avoid crystallization. As these values do not depend on any theoretically achievable limit, it will be better to express the maximum concentration corresponding to the inlet specific humidity of air. C C o i c (2.3) Cmax Ci 32

Gandhidasan (2005) expressed the humidity effectiveness in terms of the partial pressures of air in place of its specific humidity as: p p ai ao p (2.4) pai psi Enthalpy effectiveness, ε, is defined as the ratio of actual difference in H enthalpy of the air in the dehumidifier/regenerator to the maximum possible difference in enthalpy of air. The maximum possible difference corresponds to equilibrium with the inlet desiccant solution [Oberg (1998)]. H H ai ao H (2.5) H ai H eq Here, the equilibrium enthalpy, H eq, is the air enthalpy in equilibrium with the desiccant at the inlet temperature and concentration. 2.5. MODELLING AND SIMULATION OF LIQUID DESICCANT SYSTEM Peng and Howell (1981) carried out mathematical modelling of each component and optimization of design for PRM (process recirculation mode) and ERM (exhaust recirculation mode) cycles. The optimum design of PRM had an absorber height of 1.2 m with the feeding position at the middle of the tower. For ERM, the optimum absorber height was only 0.9 m. The performance of the two cycles using 96.6% concentrated TEG solution at 60 o C was studied. The ERM was found to be better in terms of thermal performance with (lower) blower power requirements, while the PRM produced cooler and drier air. Scalabrin and Scaltriti (1990) evaluated LDCS performance with maximum possible heat recovery. An analysis of system performance was carried out by varying the supply air conditions, mass flow rate and the latent load on the system. The COP reported by them was in the range of 0.7-0.99. 33

Sadasivam and Balakrishnan (1992) also derived an expression for effectiveness based on NTU and heat capacity ratio for design of packed beds where simultaneous heat and mass transfer process occurs. Their definition of NTU was based on minimum heat capacity rate instead of heat capacity of air as used by Steven et al. (1989). The model was found to compare well with finite difference models and experimental data. Mahmoud and Ball (1992) presented a mathematical model for a commercial liquid desiccant (Kathabar spray-cel system). The numerical solution of the model accurately predicted the manufacturer s performance data. Khan and Ball (1992) developed a model for performance assessment of packed beds and estimated its annual energy consumption. In another paper, they presented a mathematical model to predict the year round performance of a commercially available coil type liquid desiccant system. Park et al. (1994) developed a numerical model for investigation of the combined heat and mass transfer between laminar liquid desiccant films and air in cross-flow. The findings of the numerical calculation compared fairly well with experimental results. Khan (1998) offered heat and mass transfer analysis of an internally-cooled liquid desiccant absorber (ICLDA). The ICLDA was assumed as a spray finned tube type heat exchanger. A 2D steady state model was developed and solved mathematically. It had been reported that performance of ICLDA was a strong function of carried over regeneration heat, water to air mass flow ratio, coolant water inlet temperature, outside NTU based on air side heat capacity, inside NTU based on water side heat capacity and solution operating concentration. Based on the NTUeffectiveness method, a performance forecast model was developed to evaluate the annual energy requirements of an ICLDA using an hour by hour analysis. Yigit (1999) carried out numerical study of heat and mass transfer in falling film absorber. Absorption of a gas or vapour in a laminar falling film on the surface of vertical tube was considered. The simultaneous partial (parabolic) differential equations based on diffusion and energy equation characterizing absorption process were solved by Crank-Nicholson method. It is reported that the coolant side flow rate has little effect on mass of water vapour absorbed and the outlet film temperature. 34

Lazzarin et al. (1999) developed a computer model of a packed tower to determine heat and mass transfer between air and desiccant, with H 2 O/CaCl 2 solution. This model was validated with experimental data. A parametric study was carried out to determine optimum operating conditions. Jain et al. (2000b) carried out design optimization study of a liquid desiccant system. The system components considered were falling film absorber, falling film regenerator, two tube-fin air to air heat exchangers, a solution heater, and a double pipe solution to solution heat exchanger. The total cost (capital+running) of the system over its life span was taken as the objective function to be minimized. Gandhidasan (2004) described a model for the initial design of a packed bed air dehumidifier through dimensionless vapour pressure and temperature difference ratios. A mathematical relation was derived using the above ratios to calculate the condensate rate in terms of known working parameters. The results of model were found to be very good agreement with the experimental observations existing in the literature. Islam et al. (2004) developed a simplified linearized coupled model for the heat and mass transfer in falling-film absorbers. They did comparison of the model with a non-linear model and also numerical simulations. The linearized model produces analytical expressions that are used to find heat and mass transfer coefficients from the experimental data for a horizontal tubular absorber and a vertical tubular absorber under certain conditions. Yin et al. (2006) studied the packed bed liquid desiccant dehumidification and regeneration system. They made one-dimensional model to study the combined heat and mass transfer operations in the dehumidifier and regenerator. A linear estimate to correlate the equilibrium humidity ratio and solution temperature has been prepared. Reorganizing the main equations by considering new parameters, two coupled ordinary differential equations have been derived. Analytical expression for the tower efficiency was further developed based on the analytical solution. Finally average overall heat and mass transfer coefficients from experimental data have been obtained. Mesquita et al. (2006) developed mathematical and numerical models for internally cooled falling film absorbers by utilizing heat and mass transfer correlations. Numerical modeling was based on a finite difference method. The film 35

thickness was assumed constant in one model and variable in another. Fully developed laminar flow for the liquid and air streams was assumed in all approaches. The results of model were validated with the experimental records existing in the literature. Chen et al. (2006) developed analytical model for a packed-bed liquid desiccant equipment for both parallel-flow and counter-flow configurations. The desiccant mass flow rate was assumed high enough so that its concentration may be taken unaltered during dehumidification process. The optimum ratio of air-to-solution mass flow rate also suggested keeping the driving force constant throughout the column. Xiaohua et al. (2006) compared numerical solutions with analytical solutions and proved that analytical solutions are more useful rather than numerical solution. The analytical solutions have been developed for the operating parameters in dehumidifier and regenerator in three manners of flow; parallel flow, counter flow and cross flow of air and desiccant solution. The analytical solutions of the enthalpy and moisture efficiencies were found to be good in agreement with experimental results. Audha et al. (2011) fabricated solar powered liquid desiccant dehumidifier and regenerator using CaCl 2 as desiccant solution for cooling of building and drinking water. They developed a model for the solar based cooling system and predicted the MRR from regeneration process. An optimization problem was formulated for selection and operation of liquid desiccant system to meet fresh water requirement and air conditioning load at minimal energy cost for conditioned area of 80 m 2 with the objective of producing fifteen liters of fresh drinking water a day. Bassuoni (2011) investigated experimentally the performance of the structured packed liquid desiccant dehumidification system using CaCl 2 as liquid desiccant and celdek pads as packing material. Cross flow of desiccant solution and air was used in the cooling system. Mass transfer coefficient, moisture removal rate, effectiveness and the coefficient of performance were used as performance parameters of the system. The significant increase of mass transfer coefficient and MRR for both the dehumidifier and regenerator were detected with increasing air flow rate and solution flow rate. 36

Huang et al. (2012) utilized membrane plates to form parallel channels in the dehumidifier. A semi permeable membrane was used to separate the air and liquid desiccant. Heat and moisture were transferred through the membrane during cross flow of two fluids then membrane permits the transport of water vapour and heat. Momentum, energy and concentration equations for the both fluids were used to adopt the real boundary conditions on membrane surfaces. These equations were coupled with the surface of membrane. The local and mean Nusselt and Sherwood numbers were calculated by using boundary conditions. Mohammad et al. (2013) investigated a computer simulation with MATLAB in a cross flow dehumidifier with 14 parallel plates, surface density of 80 m 2 /m 3 using CaCl 2 as liquid desiccant. Specific humidity and temperature of the air, and temperature and concentration of the desiccant solution were the key inlet parameters in the dehumidifier. The simulation results specified the effect of solution mass flow rate on the MRR and dehumidifier effectiveness. Bassuoni (2014) presented the performance outcomes of an analytical model of a LDCS in cross flow arrangement of air and liquid desiccant i.e., CaCl 2. Software, the engineering equation solver, was used to get analytical solution. Analytical and experimental results have found good agreement with each other with deviation in the range of +6.63% and -5.65% in the MRR. Seenivasan et al. (2014) have conducted experiments to study the effect of temperature, concentration and flow rate of the desiccant, and mass flow rate and specific humidity of the air on the dehumidifier effectiveness. Dehumidification process was completed in two steps using CaCl 2 as desiccant solution. The evaporative cooler has been used to sensibly cool the process air in the first step and then air dehumidification was carried out. The optimum dehumidification condition has been established by using Taguchi technique. It has been confirmed that the optimum condition achieve better effectiveness for the dehumidifier. Wang et al. (2014) described a simple and high accurate model for a packed column liquid desiccant regeneration system. Heat and mass transfer rate for the regeneration process were obtained by making two equations with seven identified parameters. Once the parameters of the projected model were determined, no iterative computations are required by the model. 37