Theoretical analysis of a liquid desiccant based indirect evaporative cooling system

Energy 95 (2016) 303e312

Contents lists available at ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

Theoretical analysis of a liquid desiccant based indirect evaporative cooling system X. Cui a, M.R. Islam b, B. Mohan a, K.J. Chua a, b, * a b

Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore Engineering Science Programme, National University of Singapore, 9 Engineering Drive 1, Singapore 117575, Singapore

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 August 2015 Received in revised form 17 November 2015 Accepted 10 December 2015 Available online xxx

A compact desiccant-evaporative HMX (heat and mass exchanger) has been proposed by combining the benefits of the regenerative indirect evaporative cooling and the liquid desiccant dehumidification. In this design, the compact HMX was able to cool and dehumidify the product air simultaneously in a single unit. A computational model has been developed and validated using experimental data. The model displayed good agreement with the experimental findings with maximum discrepancy of 8%. The heat and mass transfer behavior was theoretically investigated to illustrate the detailed air treatment performance of the HMX. Simulations were performed to study the effect of several key parameters on the HMX's performance. Due to the effect of pre-cooling and pre-dehumidification, the working air showed improved cooling potential in the working channel. Consequently, the temperature of the product air could be reduced below the dew-point temperature of intake air. Simulation results showed that the outlet temperature of the product air was affected by the working-to-intake air flow rate ratio and the dimensionless channel length, while the outlet humidity ratio of the product air was influenced by the length of the liquid desiccant film and the dimensionless channel length. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Air-conditioning Indirect evaporative cooling Dehumidification Heat and mass transfer Mathematical model

1. Introduction Air-conditioning systems should fulfill their duties to meet the cooling requirements due to sensible and latent loads. In conventional vapor compression systems, the latent cooling load is handled by cooling the process air to below its dew-point temperature in order to condense water vapor. The dehumidified air may be reheated thereafter to meet the required indoor condition. The over-cooling and reheating processes are key disadvantages of a conventional vapor compression system due to its inefficient method to dehumidify air. To overcome the drawback of the conventional vapor compression system, novel energy-saving green air-conditioning techniques are imperative. Indirect evaporative cooling is considered an effective and sustainable method for sensible cooling. Indirect evaporative cooling system produces cool air by taking the advantage of the large latent heat of water evaporation, therefore, it is suitable for hot and arid regions. As a potential alternative to the mechanical vapor compression system, it has been studied in numerous research works [1e3]. Possible improvements on the IEHX (indirect

* Corresponding author. Tel.: þ65 6516 2558; fax: þ65 6779 1459. E-mail address: [email protected] (K.J. Chua). http://dx.doi.org/10.1016/j.energy.2015.12.032 0360-5442/© 2015 Elsevier Ltd. All rights reserved.

evaporative heat exchanger) have been proposed for better cooling effectiveness. Regenerative IEHX, based on M-cycle, is able to provide cool air with an outlet temperature approaching to its dewpoint temperature [4e6]. In a regenerative IEHX, the working air employs the pre-cooled air which is redirected from the product channel [2]. As a result, the product air is cooled without absolute humidity change. Researchers have studied several types of the IEHX with counter-flow [7e10] and cross-flow arrangements [11,12]. It is reported in previous studies that the regenerative IEHX is able to achieve a dew-point effectiveness up to 0.9. The concept of evaporative cooling has been widely used in a variety of airconditioning systems [13e16]. However, the IEHX is unable to effectively handle latent cooling load which limits its application in humid climates. The use of LD (liquid desiccant) for dehumidification is one promising method to deal with latent cooling load [17e19]. A number of theoretical and experimental studies have been conducted to investigate liquid-desiccant's dehumidification performance. For example, Mesquita [20] et al. developed mathematical models for parallel-plate type liquid desiccant dehumidifiers. Dai and Zhang [21] numerically studied the simultaneous heat and mass transfer in a cross-flow liquid desiccant dehumidifier. Analytical models were developed based on the overall heat and mass transfer coefficients determined from using dimensionless

304

X. Cui et al. / Energy 95 (2016) 303e312

Nomenclature

a c D hfg H k L m M P r R T u v V

u ε

thermal diffusivity [m2/s] molar concentration [mol/m3] diffusivity [m2/s] specific latent heat of water evaporation [kJ/kg] height of the channel [m] thermal conductivity [W/(m  C)] length of the channel [m] mass flow [kg/s] molar mass [kg/mol] pressure [kPa] working-to-intake air flow rate ratio ideal gas constant [kJ/(K mol)] temperature [ C] velocity in x direction[m/s] velocity in y direction [m/s] volume flow rate of air [m3/s] humidity ratio [g moisture/kg dry air] efficiency

parameters [22,23]. In the dehumidifier, the liquid desiccant directly contacts the moist air. One of the commonly selected dehumidification equipment is packed towers [24,25]. Compared with a random packed configuration, the structured packing provides a smaller pressure drop on the air side [26,27]. Since the moisture removal process releases latent heat, the increased temperature may depreciate the capability of the liquid desiccant. Consequently, improvements were proposed by incorporating internal cooled elements such as cooling water [28e30]. Several studies have proposed combined systems in which the liquid desiccant dehumidifier is integrated with other cooling equipment. Saman and Alizadeh [31] investigated a dehumidifier which was indirectly cooled by a secondary air stream. It was also reported that the performance of the conventional vapor compression system could be improved by utilizing liquid desiccant dehumidification [32,33]. In some combined systems, the liquid desiccant dehumidifier could also be operated in tandem with IEC (indirect evaporative cooling) devices [34e37], and/or DEC (direct evaporative cooling) devices [38,39]. In such systems, the process air was first dehumidified by the liquid desiccant, and then cooled in the IEC or DEC devices. The review of previous studies indicates that the indirect evaporative cooling system and the liquid desiccant dehumidification system have been extensively investigated as shown in Table 1. In most of the previous related works, the IEC device or the LD dehumidifier was examined as a stand-alone unit. Thus far, few works have focused on the development of a compact HMX (heat and mass exchanger) which incorporates liquid desiccant to realize a regenerative IEHX based on M-cycle. The present study aims to investigate the thermal process in a novel compact HMX which is able to dehumidify and cool the product air simultaneously in one single unit by combining the benefits of regenerative indirect evaporative cooling and liquid desiccant dehumidification. The advantages of the HMX are as follows. (i) The compact unit can save space. (ii) The arrangement makes full use of the working air to cool the liquid desiccant before the working air is finally exhausted. Cooling of liquid desiccant using precooled working air enhances the moisture absorption potential of liquid desiccant from the product air. (iii) The air can be dehumidified and cooled simultaneously in one unit. (iv) The

m n r d G

zw

dynamic viscosity coefficient [Pa s] kinematic viscosity [m2/s] density [kg/m3] thickness [m] mass flow rate of liquid desiccant solution per unit width [kg/(m s)] concentration (mass fraction) of desiccant in the solution concentration (mass fraction) of water in the solution

Subscript a D in out w dew s sat equ

air desiccant inlet outlet water dew-point temperature solution saturated equilibrium

zD

compact HMX could be a potential alternative to the conventional vapor compression air handling unit. Therefore, considering these aspects, we believe it is essential to conduct a study on this compact design. The key objectives of this work are as follows: (1) introduce a desiccant-evaporative cooling HMX design in which the product channel is partially covered with liquid desiccant; (2) present a mathematical model to theoretically study the performance of the HMX; (3) investigate the influence of several key parameters such as the working-to-intake air flow rate ratio (r), the liquid desiccant film length (LD), the dimensionless channel length (L/H), and the inlet conditions (Ta,in, ua,in). Table 1 also indicates the differences between the present study and selected representative studies from literature. In this paper, we first describe the configuration of the desiccant-evaporative HMX, followed by the mathematical formulation of the heat and mass transfer phenomenon in the HMX. The mathematical model is then validated against two sets of experimental data. Finally, the theoretical performance of the HMX is analyzed via the validated mathematical model. 2. Description of the desiccant-evaporative HMX The desiccant-evaporative cooling HMX combines the benefit of the regenerative IEHX with the liquid desiccant dehumidification process. Fig. 1 is a schematic of a one-unit channel pair of the desiccant-evaporative cooling HMX. The HMX comprises numerous channel pairs stacked together. In the working wet channel, similar to the regenerative IEHX, water is injected to maintain a wet condition. In the product channel, liquid desiccant is supplied and covers part of the active surface. The length of the liquid desiccant film (LD) is one of the key parameters that impact dehumidification performance. At the end of the product channel, part of the product air is diverted into the working wet channel and is subsequently exhausted to the atmosphere. As shown in Fig. 1, the diverted product air is first dehumidified and pre-cooled before channeling to the wet channel. Therefore, the air at the starting point of the working channel has significant cooling potential. The product air usually experiences an increase in temperature due to the release of heat during the dehumidification process, and it is then gradually cooled by the indirect evaporative cooling. As a

X. Cui et al. / Energy 95 (2016) 303e312

305

Table 1 Comparison between the present study and selected representative studies. Study

Analysis method

Air treatment process

Key features

Operation performance

[31]

Experiment

One compact unit: LD þ IEC

Studied the thermal and dehumidification behavior; proposed an arrangement allows the primary air to be indirectly evaporative cooled by the secondary air and dehumidified by the liquid desiccant.

[34]

Simulation; Experiment

Two stages: LD (1st stage) þ IEC (2nd stage)

[38]

Simulation; Experiment

Two stages: LD (1st stage) þ DEC (2nd stage)

[36]

Simulation

Three stages: LD (1st stage) þ IEC (2nd stage) þ DEC (3rd stage)

[37]

Simulation

Three stages: LD (1st stage) þ IEC (2nd stage) þ DEC (3rd stage)

Present study

Simulation

One compact unit: LD þ Regenerative IEHX

Proposed a two-stage liquid desiccant air-conditioner; developed numerical models for the fluid and thermal performance; presented experimental results of prototypes for each stage. Developed mathematical model adapted from existing theories that were applied in similar designs; tested the system in the laboratory using MgCl2 as the desiccant; predicted the theoretical performance during hot summer months. Integrated a liquid desiccant system into a 100% outdoor air system; conducted simulation to estimate the energy saving potential of the system compared with a conventional variable air volume system. Utilized lithium chloride solution as liquid desiccant in the dehumidifier before the IEC and DEC; analyzed the effect of key parameters on the performance by using first and second laws of thermodynamics. Proposed a compact HMX to cool and dehumidify the product air simultaneously; investigated the air treatment process in the unit; studied the impacts of several key parameters such as the working-to-intake air flow rate ratio, the liquid desiccant film length, and the inlet conditions.

Tin ¼ 29.8e38.0  C; uin ¼ 17.0 g/kg; min ¼ 0.16e0.47 kg/s; PHE effectiveness: 0.38e0.81 Tin ¼ 26.7e35.0  C; uin ¼ 13.2e18.6 g/kg; min ¼ 0.16e0.47 kg/s; Tout ¼ 13.7e28.8  C Tin ¼ 34.7e35.2  C; uin ¼ 16.0e27.0 g/kg; DT ¼ 5.5e7.6  C;

Tin ¼ 31.6  C; uin ¼ 21.5 g/kg; Tout ¼ 15.1  C; uout ¼ 10.2 g/kg Tin ¼ 25.0e43.0  C; RHin ¼ 37e90%; Tout ¼ 15.1e25.0  C; uout ¼ 5.2e12.2 g/kg Tin ¼ 27.5e37.5  C; uin ¼ 12e20 g/kg; L/H ¼ 140e280; LD ¼ 0e0.6 m; Tout ¼ 13.5e32.4  C; uout ¼ 6.1e13.3 g/kg

result, using the proposed HMX design, it is possible to cool and dehumidify the product air simultaneously in a single compact HMX. The present design of the compact HMX has several unique features. For example, the HMX combines the liquid desiccant with the indirect evaporative cooling in order to create a compact unit which is able to save space for the air handling unit. The present HMX makes full use of the exhausted working air in our compact design by employing the working air to further cool the liquid desiccant before the working air is finally exhausted. This air flow arrangement will probably improve the performance of dehumidification by incorporating cooling element for the liquid desiccant. Based on the concept of the HMX, following suggestions are made in order to address potential issues in practice. The carryover problem of the liquid desiccant may be eliminated by reducing the air flow velocity, using laminar nonwavy flow of the liquid desiccant, and installing drift eliminators. The utilization of wicking material is able to maintain the wet condition due to capillary effect even at low flow rate of liquid desiccant. 3. Mathematical formulation To establish a mathematical model describing the simultaneous heat and mass transfer in the HMX, the following assumptions are made: (1) both the liquid and air flows are steady and laminar; (2) thermodynamic equilibrium is attained at the airewater and the air-solution interfaces; (3) the thickness of the liquid desiccant film is constant; (4) the buoyancy force is negligible; (5) the working channel contains wicking material which is evenly saturated with water so that the water film thickness is assumed as zero. 3.1. Liquid desiccant film Fig. 1. Schematic of the liquid desiccant-evaporative HMX. (a) One-unit channel pair (b) Plan view.

The flow rate of the liquid desiccant is larger compared with the water absorption rate. As a result, the solution flow rate, its mean

306

X. Cui et al. / Energy 95 (2016) 303e312

velocity, and the film thickness are considered as constants [40]. Based on the above assumptions and the geometry of the heat exchanger, the following governing equations for the liquid desiccant film are described. The momentum equation is given as [20]

mD

d2 uD þ rD g ¼ 0 dy2

(1)

PD ¼ b1 þ b2 TD þ b3 zD þ b4 TD2 þ b5 z2D þ b6 TD zD

The film thickness can be expressed as [21]

dD ¼

3GD mD r2D g

!1=3 (2)

The energy conservation equation is written as

uD

vTD vT v2 T þ vD D ¼ aD 2D vx vy vy

(3)

The species diffusion equation is written as

uD

vzD vz v2 z þ vD D ¼ DD 2D vx vy vy

The interface vapor pressure of liquid desiccant is influenced by its temperature and concentration. A lower interface vapor pressure can be achieved by either reducing the temperature or increasing the concentration of the solution. According to the thermodynamic properties acquired from literature [42,43], the interface vapor pressure of the liquid desiccant is obtained through an algebraic fitted equation:

(4)

where b1 ¼ 1.004, b2 ¼ 0.1265, b3 ¼ 5.982, b4 ¼ 0.003644, b5 ¼ 11.13, b6 ¼ 0.7975, TD is the desiccant solution temperature, and zD is its concentration (mass fraction). The temperature range in Eq. (13) is from 20 to 50  C, and the concentration ranges from 0.20 to 0.40. An equilibrium condition is assumed at the air-solution interface. The airevapor mixture is also assumed to be an ideal gas mixture. In other words, the water vapor pressure of the moist air at the air-solution interface can be evaluated by the surface vapor pressure of the liquid desiccant. Therefore, the equilibrium vapor concentration of the moist air at the air-solution interface c*a is calculated as

3.2. Moist air The working fluid is the moist air. The moist air flowing in both the product channel and the working channel is governed by the following equations [41]. The continuity equation:

vua vva þ ¼0 vx vy

(5)

Momentum conservation equation is written as

ua

vua vua 1 dp v2 ua þ na 2 þ va ¼ ra dx vx vy vy

(6)

Energy conservation equation is given below:

ua

vTa vTa v2 Ta þ va ¼ aa 2 vx vy vy

(11)

c*a ¼

PD ðTD ; zD Þ ; RTD

x < LD

(12)

where PD ðTD ; zD Þ is the surface vapor pressure of the liquid desiccant determined by using Eq. (11) under the specific temperature (TD) and concentration (zD ) condition. Therefore, the boundary conditions between the desiccant film and the moist air are explained as follows. The species balances at the air-solution interface is given as

rD DD

vzw vca ¼ MH2 O Da vy vy

(13)

The energy balance at the interface is expressed as

(7)

Equation of species diffusion for water vapor is expressed as

kD

dTD dTa vca ¼ ka þ MH2 O hfg Da dy dy vy

(14)

The velocity of the liquid desiccant at the interface requires

2

ua

vca vca v ca þ va ¼ Da 2 vx vy vy

(8)

3.3. Boundary and interfacial conditions The inlet boundary conditions of the moist air in the product channel are specified as

ua ¼ ua; in ; va ¼ 0; Ta ¼ Ta;in ; ca ¼ ca;in

(9)

The inlet boundary conditions of the desiccant solution are given by

uD ¼ uD; in ; vD ¼ 0; TD ¼ TD;in ; zD ¼ zD;in

(10)

The interfacial conditions are explained as follows. In this study, the aqueous lithium chloride solution is employed as the desiccant material. The driving force for dehumidifying the moist air is the vapor pressure difference between the vapor pressure at the air-solution interface and the vapor pressure of the main moist air stream.

vuD ¼0 vy

(15)

As the thickness of the liquid desiccant and the plate wall is small, the temperature difference across the liquid desiccant film and the internal wall (including the water film in the working channel) is negligible. The local temperature of the liquid desiccant and the water film can be assumed the same. Therefore, the boundary conditions between the desiccant film and the internal plate wall are written as:

uD ¼ 0; vD ¼ 0;

vzD ¼ 0; TD ¼ Tw vx

(16)

In the working channel, the stagnant water film is evenly distributed on the wet surface. The moist air at the airewater interface is assumed to be saturated at the temperature of the water film. The saturated vapor pressure can be determined as a function of temperature through the equation [44]:

X. Cui et al. / Energy 95 (2016) 303e312

ln Psat ¼

C1 2 3 þ C2 þ C3 Tw þ C4 Tw þ C5 Tw þ C6 ln Tw Tw

(17)

where C1 ¼ 5.800 220 6 E3, C2 ¼ 1.391 499 3, C3 ¼ 4.864 023 9 E2, C4 ¼ 4.176 476 8 E5, C5 ¼ 1.445 209 3 E8, C6 ¼ 6.545 967 3, and Tw is the absolute temperature of water film. The water vapor concentration at the water film surface is obtained as

ca ¼

Psat ðTw Þ RTw

(18)

where Psat(Tw) is the saturated vapor pressure determined by using Eq. (17) at the absolute temperature of the water film. The evaporation rate of water is governed by the gradient diffusion as:

 mv ¼ MH2 O Da

vca vy

 (19) w

As a results, the interfacial condition at the water film surface in the working wet channel is given as

ua ¼ 0; va ¼ 0 kw

(20)

dTw dTa vca ¼ ka þ MH2 O hfg Da dy dy vy

(21)

3.4. Simulation condition The heat and mass transfer process in the desiccantevaporative cooling HMX is evaluated using the above presented mathematical model. The governing equations and boundary conditions are established and numerically solved using the software COMSOL Multiphysics [45]. Three physical modes (i.e. Conjugate Heat Transfer, Transport of Diluted Species, and Transport of Concentrated Species) are fully coupled to model the fluid flow, the heat transfer, and the mass transfer of the working fluids. The pre-set parameters for the HMX are shown in Table 2. The working-to-intake air flow rate ratio ðr ¼ ðVworking air =Vintake air ÞÞ, the liquid desiccant film length (LD), inlet temperature and humidity ratio of air are key parameters which influence the cooling and dehumidification performance. Simulations were carried out based on the pre-set parameters while changing one of these parameters in order to investigate the sensitivity of the selected parameters. To estimate the cooling performance, a key factor named dewpoint effectiveness is defined as:

εdew ¼

Ta;in  Ta;out Ta;in  Ta;in;dew

εD ¼

ua;in  ua;out ua;in  us;equ

307

(23)

where us,equ is the humidity ratio of the air in equilibrium with the desiccant solution. 4. Simulation results and analysis 4.1. Validation When no liquid desiccant exists in the product channel, the desiccant-evaporative cooling HMX is considered as a regenerative IEHX. On the other hand, when the working-to-intake air flow rate ratio equals to zero (i.e., ðVworking air =Vintake air Þ ¼ 0), the HMX functions purely as a dehumidifier. Therefore, the model validation process is conducted in two parts: (1) regenerative IEHX without the usage of liquid desiccant; and (2) liquid desiccant dehumidifier system. For the first part, experimental data on a regenerative IEHX [3] was employed to validate the model's predictive accuracy of the HMX without the usage of liquid desiccant. Hsu et al. [3] has experimentally investigated a counter-flow regenerative IEHX. The intake air was pre-cooled in the dry channel and was separated into two parts at the end of the dry channel. One part of the pre-cooled air was diverted into the wet channel and acted as working air. By vaporizing water, the working air absorbed heat from the dry channel so that the product air was cooled without the change of humidity ratio. To validate the model, the experimental condition was precisely replicated in the simulation. The simulated air temperature distribution in the dry product channel was compared with the experimental data as shown in Fig. 2. It is observed that the mathematical model is able to predict accurately within the discrepancy of ±8%. For the second part, the mathematical model was validated against experimental data acquired from a literature studying on a liquid desiccant system for dehumidification. Yin et al. [28,47] conducted experiments on a dehumidifier with internal cooled elements. Fig. 3 compares our simulated results with experimental data in terms of the outlet air humidity ratio under varying desiccant temperature. The inlet air temperature was kept at 30.5  C, while its humidity ratio was 13.4 g/kg. The lithium chloride solution had a mass fraction of 0.377, and a mass flow rate of 0.1036 kg/s. The

(22)

To evaluate the dehumidification ability, the dehumidification effectiveness is given as [21,46]:

Table 2 Pre-set parameters for the HMX. Parameters

Value

Unit

Channel length (L) Channel gap (H) Solution film length (LD) Plate thickness Solution film thickness

1.00 0.005 0e0.6 0.0003 0.0005

m m m m m

Fig. 2. Validation 1: comparison between the simulated results and experimental data on a regenerative indirect evaporative heat exchanger.

308

X. Cui et al. / Energy 95 (2016) 303e312

Fig. 4. Temperature and humidity ratio profiles in the product channel. Fig. 3. Validation 2: comparison between simulated results and experimental data on a liquid desiccant dehumidifier.

mathematical model displayed a good agreement with the experimental data with a maximum discrepancy of ±5%. It can be inferred that the computational model is capable of evaluating the performance of the desiccant-evaporative HMX through two respective validations. 4.2. Psychrometric description of the air treatment process The state of air flow in the HMX was predicted to describe the temperature and humidity distribution. The inlet air assumed a dry bulb temperature of 30  C, and a humidity ratio of 16 g/kg. The product air inlet velocity was 1.0 m/s, while the working air velocity was 0.5 m/s. The initial mass fraction of the lithium chloride solution was 0.40. The desiccant film length was 0.2 m. Fig. 4 illustrates the temperature and humidity ratio profiles for the air flowing through the product channel. At the beginning section of the product channel which is covered by liquid desiccant (x  0.2 m), the water vapor in the moist air is absorbed by the liquid desiccant. As shown in Fig. 4, the humidity ratio decreased from 16 g/kg to 10.7 g/kg. During this mass transfer process, the product channel experienced a rapid release of latent heat due to the absorption of moisture resulting in the temperature being raised from its initial temperature to 31.9  C. After the interaction with the liquid desiccant, the humidity ratio of product air remained stable at 10.7 g/kg. As the remaining section of channel which did not contain the desiccant solution, the temperature of product air was gradually lowered since sensible heat was transferred to the adjacent wet working channel where evaporation of water occurred. The simulated air treatment process is illustrated on the psychrometric chart as shown in Fig. 5. The intake air flowed into the product channel from a state at the point P1. Once the air came into contact with the liquid desiccant film, it was dehumidified due to the moisture absorption by the desiccant solution. At the same time, latent heat was released during this dehumidification process. As a result, in this region of the product channel, the product air temperature increased slightly while the humidity ratio decreased. Subsequently, the product air was further cooled by the adjacent working air without absolute humidity change, resulting in the product air reaching a condition at point P2 (the outlet condition of product air). At the end of the product channel, a part of the product air was redirected into the working channel acting as the working air. In the

Fig. 5. Psychrometric illustration of the air treatment process (T1,in ¼ 30 w1,in ¼ 16 g/kg).

 C,

wet channel, the working air absorbed heat due to the evaporation of water. The condition of the working air was regulated from point P2 to point P3 (the outlet condition of working air) since it is heated and humidified. As illustrated on the psychrometric chart, the temperature of the product air can potentially be reduced below its inlet dew-point temperature. For example, in this operating condition, the dewpoint temperature of the product air was 21.4  C at the inlet (point P1). After the dehumidification process, the humidity ratio of the product air decreased to 10.7 g/kg and the corresponding dewpoint temperature of the product air was lowered to 15.1  C at x ¼ 0.2 m. When a portion of the product air was diverted to the wet working channel, it had greater potential to absorb water vapor. In other words, the pre-dehumidified working air had a higher cooling potential in the working channel. Therefore, at the outlet of the product channel, the air was eventually cooled to the dry-bulb temperature of 20.7  C (a temperature which was lower than its inlet dew-point temperature of 21.4  C). Compared with previous related studies, it can be found that the HMX demonstrates following differences in terms of the

X. Cui et al. / Energy 95 (2016) 303e312

309

psychrometric illustration due to the compact configuration: (i) no extra outdoor air is used for cooling the dehumidification process; (ii) the temperature increase in the dehumidification region is not significant due to the fact that the HMX effectively makes the use of the cooling potential of the working air. In general, by combining the advantage of both evaporative cooling and liquid-desiccant dehumidification in the compact unit, the product air is simultaneously cooled and dehumidified. 4.3. Influence of air flow rate ratio The working air acts as the cooling source in the HMX. The working-to-intake air flow rate ratio ðr ¼ ðVworking air =Vintake air ÞÞ is one of the influential parameters controlling the cooling effectiveness. Fig. 6 presents the outlet air temperature and humidity ratio impacted by the air flow rate ratio. The inlet air was assumed at a condition having dry bulb temperature of 30  C, and humidity ratio of 16 g/kg. The liquid desiccant film length was 0.2 m. The intake air velocity in the product channel was maintained at 1 m/s, while the working air velocity was changed from 0 m/s to 0.8 m/s. When the working-to-intake air flow rate ratio is set zero, the HMX becomes a purely dehumidifier in which the product air is dehumidified without taking the advantage of evaporative cooling. As apparent from Fig. 6, the product air was dehumidified from a humidity ratio of 16 g/kg to a humidity ratio of 9.86 g/kg. In addition, the temperature of the product air increased from 30  C to 32.43  C due to the release of heat in the dehumidification process. When the working-to-intake air flow rate ratio increased from 0 to 0.8, the outlet temperature of the product air was significantly lowered from 32.43  C to 19.39  C. This marked temperature reduction can be attributed to a greater capacity of the working air to absorb water vapor in the wet channel. Therefore, a larger sensible heat was transferred from the product channel to the working channel. In addition, it was observed that the outlet humidity ratio decreased marginally from 9.86 g/kg to 9.48 g/kg with a higher air flow rate ratio. The observation is attributed to the slight reduction of the solution temperature owing to a lower product air temperature which translates to a lower surface vapor pressure of the liquid desiccant. In sum, the working-to-intake air flow rate ratio predominantly impacts the outlet temperature of the product air. In contrast, its influence on the outlet humidity ratio is marginal. A lower outlet temperature can be effectively achieved by increasing the air flow rate ratio.

Fig. 6. Influence of the working-to-intake air flow rate ratio on the outlet condition of product air.

Fig. 7. Influence of the liquid desiccant film length on the outlet condition of product air.

Fig. 8. Influence of the liquid desiccant film length on the cooling effectiveness and dehumidification effectiveness.

4.4. Influence of the liquid desiccant film length Fig. 7 shows how the outlet temperature and humidity ratio affected by the liquid desiccant film length. Simulation results were obtained by varying the liquid desiccant film length from 0 m to 0.6 m. Inlet conditions of the air were maintained constant as follows: Ta,in ¼ 30  C, ua,in ¼ 16 g/kg, ua,in ¼ 1.0 m/s, and r ¼ 0.5. Based on the simulated results, the cooling effectiveness and the dehumidification effectiveness were evaluated and displayed in Fig. 8. When the liquid desiccant film length equals to zero (i.e., no liquid desiccant exists in the channel), the HMX is considered as a regenerative indirect evaporative heat exchanger. In this case, the product air was cooled to an outlet temperature of 23.56  C while its humidity ratio was being kept constant. Accordingly, the HMX achieved a dew-point effectiveness of 0.75. When the liquid desiccant film length was increased from 0 m to 0.6 m, the product air was dehumidified effectively evidenced from the reduction in the outlet humidity ratio from 16 g/kg to 7.42 g/kg. The result can be attributed to a longer contact time and a larger contact area between the product air and the liquid desiccant. Consequently, the dehumidification effectiveness increased from 0 to 0.78 as depicted in Fig. 8.

310

X. Cui et al. / Energy 95 (2016) 303e312

Fig. 7 further shows that the outlet temperature of the product air first decreased from 23.56  C to 20.40  C by increasing the liquid desiccant film length from 0 m to 0.3 m. Thereafter, the outlet temperature began to increase to 21.12  C when the length of liquid desiccant film was further increased to 0.6 m. This outlet temperature variation is a consequence of the pre-dehumidification process. The dehumidified working air is capable of absorbing more water vapor resulting in a greater cooling capacity in the product channel. Therefore, the outlet air temperature may be reduced up to a specific length of the liquid desiccant film. The dehumidification process is coupled with a release of heat. A longer liquid desiccant film length leads to a greater humidity ratio reduction which is associated with a higher degree of heat release. The released heat in turn increases the temperature of the product air particularly when the absorbed heat by the working air is insufficient to balance the released heat. For instance, in Fig. 7, the outlet temperature of product air increased when the liquid desiccant film length was larger than 0.3 m. Based on the outlet temperature of the product air, the corresponding dew-point effectiveness is calculated to determine the cooling performance of the HMX. As illustrated in section 4.2, the product air is capable of being cooled to below its inlet dew-point temperature. The maximum dew-point effectiveness of 1.1 was observed when the length of the liquid desiccant film was about 0.3 m. 4.5. Influence of the channel dimension The dimension of the airflow passage is another influential parameter on the HMX's performance. Fig. 9 indicates the outlet condition under varying channel dimensions. Simulation results were obtained under three types of the channel height. The dimensionless channel length (L/H) was changed from 142 to 280. Other parameters were maintained constant as: Ta,in ¼ 30  C, ua,in ¼ 16 g/kg, ua,in ¼ 1.0 m/s, LD/L ¼ 0.3, and r ¼ 0.5. Simulation results illustrate that the outlet temperature can be reduced with increasing dimensionless channel length. It is probably a consequence of the increase in the contact time and area caused by the longer channel length. This trend is consistent with the results in a previous related study on the indirect evaporative heat exchanger [9]. A similar impact is observed that the longer channel length leads to a lower outlet humidity ratio. In addition, the channel height has a great influence on the outlet condition. A lower outlet temperature and humidity ratio can be effectively

Fig. 9. Influence of the dimensionless channel length on the outlet condition of product air.

obtained by decreasing the channel height. For example, when the L/H was 200, the outlet temperature was 15.9  C (for H ¼ 3 mm), 20.4  C (for H ¼ 5 mm), and 23.06 (for H ¼ 7 mm), respectively. The reason can be explained as follows. Firstly, the heat and mass transfer coefficient may be enhanced by reducing the channel height. Secondly, the improved dehumidification process may provide a greater cooling capacity to both the product air and the film of liquid desiccant which will further decrease the product air temperature and moisture content at the outlet. In sum, it can be inferred from Fig. 9 that a lower outlet temperature and humidity ratio can be achieved by reducing the channel height and increasing the dimensionless channel length. 4.6. Influence of inlet condition The performance of the HMX has been theoretically studied under varying temperature and moisture content of inlet air. The intake air temperature was varied from 27.5 to 37.5  C, while the intake air humidity ratio was changed from 12 to 20 g/kg. The inlet air velocity was 1.0 m/s, and the working-to-intake air ratio was 0.5. Fig. 10 illustrates the performance of the HMX for a selected configuration (i.e. H ¼ 0.005 m, L ¼ 1 m, LD ¼ 0.3 m). Due to the dehumidification process, the outlet temperature may be cooled to below its inlet dew-point temperature. This demonstrates a better cooling performance compared with the conventional regenerative indirect evaporative cooling. For instance, in Fig. 10, when the inlet humidity ratio was 20 g/kg, the outlet temperature could potentially be lowered than its inlet dewpoint temperature of 24.95  C. For a specific inlet temperature condition, a lower outlet temperature was obtained by decreasing the inlet humidity ratio. This finding could be attributed to the benefit of employing evaporative cooling. The working air with a lower humidity ratio was able to absorb more heat due to the evaporation of water. The outlet humidity ratio was also observed to increase with higher inlet temperature. It is probably the result of a smaller driving force for dehumidification. The driving force is evaluated by the difference of vapor pressure between the moist air stream and the air-solution interface. The partial pressure at the solution surface is greater at higher temperature, resulting in a reduced driving force. Fig. 10 also illustrates the capability of the HMX to promote air dehumidification. The dehumidification effectiveness is determined based on the outlet humidity ratio of the product air. For the simulated operating conditions with a constant liquid desiccant film length of 0.3 m, the dehumidification effectiveness spans 0.55 to 0.60.

Fig. 10. Thermal performance of the HMX under varying inlet conditions.

X. Cui et al. / Energy 95 (2016) 303e312

This HMX could be used as a pre-conditioning unit installed in series with the chiller based air handling unit. The proposed HMX will dehumidify and precool the product air so that the chiller based air handling unit may only provide necessary sensible cooling to maintain the preset supply air temperature. Consequently, the cooling load for the chiller plant will drop remarkably. Moreover, as the product air will be dehumidified by the HMX, it is not necessary to cool the product air below the dew-point temperature in the chiller based air handling unit for dehumidification. Therefore, chiller plant can be operated at higher chiller water supply temperature. Generally, chiller plant efficiency is improved by about 3%e4% for every 1  C increase of chiller water supply temperature. As a result, energy efficiency of the chiller plant will be improved significantly as well. 5. Conclusions We have introduced a desiccant-evaporative cooling HMX which combines the advantage of both the liquid desiccant dehumidification system and the regenerative indirect evaporative cooling system. Theoretical study on the HMX has been carried out using an experimentally validated computational model. The simulated air treatment processes indicate that the product air can be cooled and dehumidified simultaneously in the HMX. It can be inferred that the working air has a greater capacity to absorb water vapor due to the pre-cooling and the pre-dehumidifying processes. The outlet air temperature could potentially be reduced to a temperature lower than its inlet dew-point temperature. The present work investigates the influence of several key parameters on the performance of the HMX. These key parameters include the working-to-intake air flow rate ratio, the liquid desiccant film length, the dimensionless channel length, and the inlet air condition. A lower outlet air temperature can be effectively achieved by increasing the working-to-intake air flow rate ratio or increasing the dimensionless channel length. In addition, a lower outlet humidity ratio can be attained by increasing the liquid desiccant film length or increasing the dimensionless channel length. Simulation results clearly illustrate the two purposes of the HMX. First, the HMX demonstrates its capability of being used as a potential stand-alone air handling unit. As explained in Section 4.3e4.6, a desired outlet condition could be obtained by adjusting those key parameters. Second, the HMX could be employed as a pre-conditioning unit installed in series with the chiller based air handling unit. As a result, the cooling load for the chiller plant will drop remarkably and the energy efficiency of the chiller plant will also be improved significantly. References [1] Duan Z, Zhan C, Zhang X, Mustafa M, Zhao X, Alimohammadisagvand B, et al. Indirect evaporative cooling: past, present and future potentials. Renew Sustain Energy Rev 2012;16:6823e50. http://dx.doi.org/10.1016/ j.rser.2012.07.007. [2] Hasan A. Going below the wet-bulb temperature by indirect evaporative cooling: analysis using a modified ε-NTU method. Appl Energy 2012;89: 237e45. http://dx.doi.org/10.1016/j.apenergy.2011.07.005. [3] Hsu ST, Lavan Z, Worek WM. Optimization of wet-surface heat exchanger. Energy 1989;14:757e70. [4] Chua KJ, Chou SK, Yang WM, Yan J. Achieving better energy-efficient air conditioning e a review of technologies and strategies. Appl Energy 2013;104:87e104. http://dx.doi.org/10.1016/j.apenergy.2012.10.037. [5] Maisotsenko V, Gillan LE, Heaton TL, Gillan AD. Method and plate apparatus for dew point evaporative cooler. US Patent 6,508,402 2003. [6] Caliskan H, Hepbasli A, Dincer I, Maisotsenko V. Thermodynamic performance assessment of a novel air cooling cycle: Maisotsenko cycle. Int J Refrig 2011;34:980e90. http://dx.doi.org/10.1016/j.ijrefrig.2011.02.001. [7] Zhao X, Li JM, Riffat SB. Numerical study of a novel counter-flow heat and mass exchanger for dew point evaporative cooling. Appl Therm Eng 2008;28: 1942e51. http://dx.doi.org/10.1016/j.applthermaleng.2007.12.006.

311

[8] Riangvilaikul B, Kumar S. An experimental study of a novel dew point evaporative cooling system. Energy Build 2010;42:637e44. http://dx.doi.org/ 10.1016/j.enbuild.2009.10.034. [9] Cui X, Chua KJ, Yang WM. Numerical simulation of a novel energy-efficient dew-point evaporative air cooler. Appl Energy 2014;136:979e88. http:// dx.doi.org/10.1016/j.apenergy.2014.04.040. [10] Cui X, Chua KJ, Islam MR, Yang WM. Fundamental formulation of a modified LMTD method to study indirect evaporative heat exchangers. Energy Convers Manag 2014;88:372e81. http://dx.doi.org/10.1016/j.enconman.2014.08.056. [11] Zhan C, Duan Z, Zhao X, Smith S, Jin H, Riffat S. Comparative study of the performance of the M-cycle counter-flow and cross-flow heat exchangers for indirect evaporative cooling e paving the path toward sustainable cooling of buildings. Energy 2011;36:6790e805. http://dx.doi.org/10.1016/ j.energy.2011.10.019. [12] Anisimov S, Pandelidis D, Jedlikowski A, Polushkin V. Performance investigation of a M (Maisotsenko)-cycle cross-flow heat exchanger used for indirect evaporative cooling. Energy 2014;76:593e606. http://dx.doi.org/10.1016/ j.energy.2014.08.055. [13] Kim M, Ham S, Park J, Jeong J. Impact of integrated hot water cooling and desiccant-assisted evaporative cooling systems on energy savings in a data center. Energy 2014;78:384e96. http://dx.doi.org/10.1016/j.energy.2014.10.023. [14] Islam MR, Jahangeer KA, Chua KJ. Experimental and numerical study of an evaporatively-cooled condenser of air-conditioning systems. Energy 2015;87: 390e9. http://dx.doi.org/10.1016/j.energy.2015.05.005. [15] Ge TS, Dai YJ, Wang RZ, Li Y. Performance of two-stage rotary desiccant cooling system with different regeneration temperatures. Energy 2015;80: 1e11. http://dx.doi.org/10.1016/j.energy.2014.12.010. [16] Yuan F, Chen Q. A global optimization method for evaporative cooling systems based on the entransy theory. Energy 2012;42:181e91. http://dx.doi.org/ 10.1016/j.energy.2012.03.070. [17] Mei L, Dai YJ. A technical review on use of liquid-desiccant dehumidification for air-conditioning application. Renew Sustain Energy Rev 2008;12:662e89. http://dx.doi.org/10.1016/j.rser.2006.10.006. [18] Yin Y, Qian J, Zhang X. Recent advancements in liquid desiccant dehumidification technology. Renew Sustain Energy Rev 2014;31:38e52. http:// dx.doi.org/10.1016/j.rser.2013.11.021. [19] Luo Y, Yang H, Lu L, Qi R. A review of the mathematical models for predicting the heat and mass transfer process in the liquid desiccant dehumidifier. Renew Sustain Energy Rev 2014;31:587e99. http://dx.doi.org/10.1016/ j.rser.2013.12.009. [20] Mesquita LCS, Harrison SJ, Thomey D. Modeling of heat and mass transfer in parallel plate liquid-desiccant dehumidifiers. Sol Energy 2006;80:1475e82. http://dx.doi.org/10.1016/j.solener.2006.03.003. [21] Dai YJ, Zhang HF. Numerical simulation and theoretical analysis of heat and mass transfer in a cross flow liquid desiccant air dehumidifier packed with honeycomb paper. Energy Convers Manag 2004;45:1343e56. http:// dx.doi.org/10.1016/j.enconman.2003.09.006. [22] Liu XH, Jiang Y, Qu KY. Heat and mass transfer model of cross flow liquid desiccant air dehumidifier/regenerator. Energy Convers Manag 2007;48: 546e54. http://dx.doi.org/10.1016/j.enconman.2006.06.002. [23] Ren CQ. EffectivenesseNTU relation for packed bed liquid desiccanteair contact systems with a double film model for heat and mass transfer. Int J Heat Mass Transf 2008;51:1793e803. http://dx.doi.org/10.1016/ j.ijheatmasstransfer.2007.07.038. [24] Longo GA, Gasparella A. Experimental analysis on desiccant regeneration in a packed column with structured and random packing. Sol Energy 2009;83: 511e21. http://dx.doi.org/10.1016/j.solener.2008.08.016. [25] Abdul-Wahab SA, Zurigat YH, Abu-Arabi MK. Predictions of moisture removal rate and dehumidification effectiveness for structured liquid desiccant air dehumidifier. Energy 2004;29:19e34. http://dx.doi.org/10.1016/ j.energy.2003.08.001. [26] Jain S, Tripathi S, Das RS. Experimental performance of a liquid desiccant dehumidification system under tropical climates. Energy Convers Manag 2011;52:2461e6. http://dx.doi.org/10.1016/j.enconman.2010.12.052. [27] Zhang LZ, Zhang N. A heat pump driven and hollow fiber membrane-based liquid desiccant air dehumidification system: modeling and experimental validation. Energy 2014;65:441e51. http://dx.doi.org/10.1016/j.energy. 2013.10.014. [28] Yin Y, Zhang X, Wang G, Luo L. Experimental study on a new internally cooled/ heated dehumidifier/regenerator of liquid desiccant systems. Int J Refrig 2008;31:857e66. http://dx.doi.org/10.1016/j.ijrefrig.2007.10.004. [29] Zhang T, Liu X, Jiang J, Chang X, Jiang Y. Experimental analysis of an internallycooled liquid desiccant dehumidifier. Build Environ 2013;63:1e10. http:// dx.doi.org/10.1016/j.buildenv.2013.01.007. [30] Qi R, Lu L, Yang H. Development of simplified prediction model for internally cooled/heated liquid desiccant dehumidification system. Energy Build 2013;59:133e42. http://dx.doi.org/10.1016/j.enbuild.2012.12.028. [31] Saman WY, Alizadeh S. An experimental study of a cross-flow type plate heat exchanger for dehumidification/cooling. Sol Energy 2002;73:59e71. http:// dx.doi.org/10.1016/S0038-092X(01)00078-0. [32] Dai YJ, Wang RZ, Zhang HF, Yu JD. Use of liquid desiccant cooling to improve the performance of vapor compression air conditioning. Appl Therm Eng 2001;21:1185e202. http://dx.doi.org/10.1016/S1359-4311(01) 00002-3.

312

X. Cui et al. / Energy 95 (2016) 303e312

[33] Liu X, Li Z, Jiang Y, Lin B. Annual performance of liquid desiccant based independent humidity control HVAC system. Appl Therm Eng 2006;26: 1198e207. http://dx.doi.org/10.1016/j.applthermaleng.2005.10.043. [34] Woods J, Kozubal E. A desiccant-enhanced evaporative air conditioner: numerical model and experiments. Energy Convers Manag 2013;65:208e20. http://dx.doi.org/10.1016/j.enconman.2012.08.007. [35] Kozubal E, Woods J, Burch J, Boranian A, Merrigan T. Desiccant enhanced evaporative air-conditioning (DEVap): evaluation of a new concept in ultra efficient air conditioning. National Renewable Energy Laboratory; 2011. TP5500-49722. [36] Kim M-H, Park J-S, Jeong J-W. Energy saving potential of liquid desiccant in evaporative-cooling-assisted 100% outdoor air system. Energy 2013;59: 726e36. http://dx.doi.org/10.1016/j.energy.2013.07.018. [37] Tu M, Ren CQ, Zhang LA, Shao JW. Simulation and analysis of a novel liquid desiccant air-conditioning system. Appl Therm Eng 2009;29:2417e25. http:// dx.doi.org/10.1016/j.applthermaleng.2008.12.006. [38] Lychnos G, Davies P. Modelling and experimental verification of a solarpowered liquid desiccant cooling system for greenhouse food production in hot climates. Energy 2012;40:116e30. http://dx.doi.org/10.1016/j.energy. 2012.02.021. [39] Heidarinejad G, Pasdarshahri H. Potential of a desiccant-evaporative cooling system performance in a multi-climate country. Int J Refrig 2011;34:1251e61. http://dx.doi.org/10.1016/j.ijrefrig.2011.03.008.

[40] Ali A, Vafai K, Khaled A-RA. Analysis of heat and mass transfer between air and falling film in a cross flow configuration. Int J Heat Mass Transf 2004;47: 743e55. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2003.07.017. [41] Cui X, Chua KJ, Islam MR, Ng KC. Performance evaluation of an indirect precooling evaporative heat exchanger operating in hot and humid climate. Energy Convers Manag 2015;102:140e50. http://dx.doi.org/10.1016/ j.enconman.2015.02.025. [42] Conde MR. Properties of aqueous solutions of lithium and calcium chlorides: formulations for use in air conditioning equipment design. Int J Therm Sci 2004;43:367e82. http://dx.doi.org/10.1016/j.ijthermalsci.2003.09.003. [43] Wang X, Cai W, Lu J, Sun Y, Ding X. A hybrid dehumidifier model for real-time performance monitoring, control and optimization in liquid desiccant dehumidification system. Appl Energy 2013;111:449e55. http://dx.doi.org/ 10.1016/j.apenergy.2013.05.026. [44] ASHRAE. ASHRAE handbook-fundamentals. Atalanta, GA. 2009. [45] COMSOL 4.3 COMSOL multiphysics reference manual. 2013. [46] Martin V, Goswami DY. Effectiveness of heat and mass transfer processes in a packed bed liquid desiccant Dehumidifier/regenerator. HVAC&R Res 2000;6: 21e39. http://dx.doi.org/10.1080/10789669.2000.10391248. [47] Yin Y, Zhang X, Peng D, Li X. Model validation and case study on internally cooled/heated dehumidifier/regenerator of liquid desiccant systems. Int J Therm Sci 2009;48:1664e71. http://dx.doi.org/10.1016/ j.ijthermalsci.2008.12.017.