Solar-powered open absorption cycle modeling with two desiccant solutions

Energy Conversion and Management 52 (2011) 2768–2776

Contents lists available at ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

Solar-powered open absorption cycle modeling with two desiccant solutions Ayman A. Aly a,b, El-Shafei B. Zeidan a,c, Ahmed M. Hamed a,c,⇑ a

Department of Mechanical Engineering, Faculty of Engineering, Taif University, P.O. Box 888, Al-Hawyah, Saudi Arabia Department of Mechanical Engineering, Faculty of Engineering, Assiut University, Assiut, Egypt c Department of Mechanical Power Engineering, Faculty of Engineering, Mansoura University, Mansoura, Egypt b

a r t i c l e

i n f o

Article history: Received 23 May 2010 Received in revised form 14 February 2011 Accepted 15 February 2011 Available online 31 March 2011 Keywords: Solar-powered Open-cycle Absorption cooling Lithium chloride Calcium chloride Regeneration

a b s t r a c t This paper presents the modeling and simulation of solar-powered desiccant regenerator used for open absorption cooling system. The input heat, which is used to re-concentrate the desiccant solution, is estimated via a real-time solar radiation model in terms of the location, day of the year and time of the day. Lithium chloride (LiCl) and calcium chloride (CaCl2) solutions are applied as the working desiccants in this investigation. To compute the thermo-physical properties, a state equation is used for the calcium chloride desiccant while tabulated data along with an artificial neural network (ANN) model is used for the lithium chloride desiccant. A finite difference method is used to simulate the combined heat and mass transfer processes that occur in the liquid desiccant regenerator using the Matlab–Simulink platform. Using the proposed model, the effect of the important parameters, namely the regenerator length, desiccant solution flow rate and concentration, and air flow rate, on the performance of the system is investigated. It has been found that the vapor pressure difference has a maximum value for a given regenerator length with higher values obtained with the calcium chloride desiccant than those obtained with the lithium chloride desiccant. The proposed model can be successfully used for investigating the effect of different operating parameters under different ambient conditions and for predicting the overall performance of the system. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Liquid desiccant air-conditioning systems have been proposed as alternatives to the conventional vapor compression cooling systems with relatively less energy for both industrial and residential applications [1]. Liquid desiccants are solutions, such as lithium chloride (LiCl) and calcium chloride (CaCl2) solutions, that can easily absorb water vapor from air (dehumidify it) due to their high affinity for water. The gain of using these systems is increased by using waste heat or renewable heat sources, such as solar energy, to dry the desiccant solution (regenerate it) at relatively low temperature [2]. The systems which use direct solar heat, termed as solar-powered desiccant systems, are classified into closed and open-cycle systems. A closed cycle system contains a chiller which provides chilled water that can be used either directly via a water network or in air handling units. While in an open-cycle system, water which is supplied from an external source is used as refrigerant and the condenser is eliminated. Since its introduction by Kakabaev and Khandurdyev [3], solarpowered open-cycle liquid-desiccant absorption system has been ⇑ Corresponding author at: Department of Mechanical Engineering, Taif University, Al-Haweiah, P.O. Box 888, Zip Code 21974, Saudi Arabia. Tel.: +966 599121850, fax: +966 02 7274299. E-mail address: [email protected] (A.M. Hamed). 0196-8904/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.enconman.2011.02.021

investigated extensively. The feasibility of the system and the advantages it can offer in terms of energy and cost savings have been proved in different climates (see for example [2,4–10]). A schematic of an open solar-powered absorption cooling system is shown in Fig. 1. The weak absorbent solution is heated and subsequently concentrated in the solar collector. The strong regenerated solution leaves the collector and passes through a liquid column, to allow the strong solution to go from atmospheric pressure to reduced pressure efficiently. The strong solution then passes through a regenerative heat exchanger on its way to the absorber, where it absorbs water from the evaporator, maintaining the reduced pressure required with the energy supplied by heat from the cold space. The resultant weak solution is pumped from the absorber back to atmospheric pressure through the regenerative heat exchanger and the collector, completing the cycle. The advantages of this system would include a simpler collector, which also acts as a regenerator, and a reduction in thermal losses. The overall performance of the system is governed by the thermo-physical properties of the used desiccant solution and operating parameters which dictate the flow rate at which water is driven from the solution in the collector. This in turn determines the flow of water that can be introduced into the evaporator as refrigerant. The rate of water evaporation from the regenerator gives a direct measure of the system cooling capacity. Many desiccant solutions with different thermo-physical characteristics

2769

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

Nomenclature A B b C Cs c Fss g H h I i L m n p Q R s T UL wij x

apparent solar radiation at air mass zero, W/m2 atmospheric extinction coefficient constant diffuse radiation factor and solution concentration solution concentration, % constant angle factor between the surface and the sky gradient of performance function specific enthalpy, J/kg hour angle and heat transfer coefficient, W/m2 °C solar radiation intensity, W/m2 incidence angle latitude angle and latent heat of evaporation, J/kg rate of water evaporation, kg/s day number vapor pressure, mm Hg heat rate, W tilt factor surface tilt angle temperature, °C overall heat loss coefficient, W/m2 °C weighting factor coordinate, input to the ANN model

have been used for dehumidification such as calcium chloride and lithium bromide. Some researchers tried to lower the desiccant solution cost and assure dehumidification performance by using alternative low cost desiccant solutions [11] or mixing desiccant solutions such as lithium chloride and zinc chloride for open-cycle absorption solar systems [12,13]. Siddiqui [14] performed an economic analysis of absorption systems using four types of fluids, operated either by solar, biogas, or liquefied petroleum gas. To optimize the various operating parameters, the costs for various compounds were presented graphically for different operating conditions. Those data could be useful as guidelines for similar designs and operating conditions. Liu and Yi [15] presented a comparative analysis of two commonly used liquid desiccants, LiBr

z

zenith angle

Greek symbols solar altitude angle and surface absorptivity b mass transfer coefficient, kg/s m2 mm Hg d declination angle qg reflectance of the Earth’s surface s transmittance

a

Subscripts a air and absorbed B beam Bn beam at normal incidence c collector d diffuse i inlet and index j index k computation level L liquid O outside max maximum s solution and surface t total, tilted, and target output

and LiCl solutions, based on mass transfer experimental data with different operating conditions. It was concluded that the COPs of the liquid desiccant systems using these two desiccants are similar while LiCl solution cost is lower than that of LiBr solution by about 18%. Mathematical modeling has been used to complement experimental studies to assess performance and optimize complicated open solar-powered absorption systems. Simplified analytical procedures for calculating the mass of water evaporated from the weak solution in the solar regenerator in terms of climatic conditions and solution properties at the regenerator inlet has been presented in [3]. Open-cycle absorption refrigeration system was simulated and analyzed for five cities using actual weather data

Solar collector/ regenerator

Vacuum pump

Absorber Water supply

Strong solution collection tank

Regenerative heat exchanger

Chilled water

Cooling water

Reducing Valve Refrigerant

Weak solution Liquid column

Solution pump

Refrigerant (water) pump

Fig. 1. Schematic of the solar-powered open absorption cooling system.

2770

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

by Collier [16]. It was found that the relationship between the collector length and the solution mass flow rate was tied to environmental parameters such as humidity and wind characteristics. Energy and exergy analyses of open-cycle desiccant cooling systems along with an attempt to minimize exergy losses were developed by Kanoglu et al. [17]. Hawlader et al. [18] presented an iterative solution of the equations describing the conservation of mass and energy that led to the concentrations of the absorbent and heat and mass transfer coefficients. Then they developed correlations for nondimensional heat and mass transfer parameters. Haim et al. [19] performed an analysis of absorption systems with different fluids and varying cycle configurations based on specific design features with the possibility to accommodate the features of open-cycle systems. Yang and Yan [20], and Yang and Wang [21] performed a computer simulation for the collector/regenerator using a radiation processor which makes use of the statistical meteorological data for the summer season at Kaohsiung, Taiwan. Alizadeh and Saman [22] developed a computer model using calcium chloride as the working desiccant to study the thermal performance of a forced parallel flow solar regenerator. A parametric analysis of the system has been performed to calculate the rate of evaporation of water from the solution as a function of the system variables and the climatic conditions. However, the solar radiation intensity was assumed constant in the analysis. More recently, graphical programming approaches such as Simulink/ Matlab, widely used in systems modeling and control, have been used in solar energy applications [23–26]. These approaches provide modular modeling environment to model system components and integrate them with external processing tools and can be applied to model and optimize complex systems. In this paper, real-time simulations of solar-powered desiccant regenerator for open absorption cooling cycle using calcium chloride or lithium chloride solutions as the working desiccant has been carried out using the Simulink/Matlab modeling environment. The regenerator model is augmented with a model to simulate the solar radiation, and the system is assumed to be in operation during day time from sunrise to sunset.

2.1. Solar radiation model The total radiation incident on a tilted surface could be evaluated in terms of the location, day of the year and time of the day. The total perceived solar radiation can be estimated by the following relationship [27]:

ð1Þ

where IB is beam radiation on a horizontal surface, RB is beam radiation tilt factor, IBn is beam radiation at normal incidence, W/m2, Id is the diffuse sky radiation, W/m2, C is diffuse radiation factor, s is the surface tilt angle, qg is solar reflectance of the Earth’s surface. The three terms in the above equation represent the direct, diffuse, and reflected components, respectively. The terrestrial beam radiation within the atmosphere and on the Earth’s surface on a typical clear day is calculated using the following relation:

IBn ¼ A exp



 B sin a

sin a ¼ sin L sin d þ cos L cos d cosh

ð2Þ

where A is an empirically determined constant which represent the apparent solar radiation at air mass zero, W/m2, B is an apparent atmospheric extinction coefficient and a is the solar altitude angle.

ð3Þ

where L, d and h are the latitude, declination and hour angles, respectively. The declination angle d can be calculated as a function of the day number, n as:

d ¼ 23:45 sin



360 ð284 þ nÞ 365

 ð4Þ

The hour angle is defined by

1 h ¼  ðnumber of min from local solar noonÞ 4

ð5Þ

where the value of h is assumed positive in the after noon period. In Eq. (1), the diffuse solar radiation is estimated from:

Id ¼ CF ss IBn

ð6Þ

where Fss = 0.5(1 + cos s) is the angle factor between the surface and the sky and s is the tilt angle of the solar collector. The beam radiation tilt factor RB is defined by:

RB ¼

IBt cos i ¼ IB cos z

ð7Þ

where IBt, IB are the beam radiation on a tilted surface and on the horizontal surface, respectively. The incidence angle, i, and zenith angle z are calculated from the following expressions,

cos i ¼ sinðL  sÞ sin d þ cosðL  sÞ cos d cos h

ð8Þ

cos z ¼ sin a

ð9Þ

The day length, which is the period from sunrise to sunset, can be evaluated from,

day length ¼

2 cos1 ð tan L tan dÞ 15

ð10Þ

The solar radiation absorbed by the solution in the collector/ regenerator can be obtained from,

Ia ¼ It ð1  qÞsa

2. Mathematical model

    1 þ cos s 1  cos s þ ðIB þ Id Þqg It ¼ RB IB þ CIBn 2 2

The altitude angle a can be evaluated from the following expression:

ð11Þ

where q and s are the reflectance and transmittance of the glass cover on the collector surface, respectively, and a is the regenerator absorptivity. 2.2. Collector/regenerator (C/R) model A schematic representation of the forced flow C/R is shown in Fig. 2. The C/R employs an inclined flat blackened surface over which the absorbent solution to be concentrated trickles down as a thin liquid film. In order to reduce top losses and eliminate contamination of the solution with dust, the C/R is covered by a single or double glazing. Due to absorption of solar energy by the plate, water evaporates from the liquid surface and is removed by a forced air stream. The air stream may flow parallel or counter to the liquid film. The channel is divided into a large number of equal segments of width dx with the assumption of constant properties within the segment (air vapor pressure, pa, and temperature, Ta, and vapor pressure on the solution surface, ps, temperature, Ts, and concentration, Cs). The main equations include the energy balance and mass balance for each segment of the open-cycle regenerator. These equations are summarized as follows [26]: Energy balance for the regenerator-segment can be expressed in the following form,

Ia dx ¼ ms dHs þ ma dHa þ U L ðT s  T 0 Þ þ mhfg

ð12Þ

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

2771

Fig. 2. Open-cycle solar collector regenerator.

Qe It Ac

where ma , and ms are the mass flow rates of air and solution kg/s, respectively, Ha , and Hs are the specific enthalpies of air and solution J/kg, respectively, m is the mass of evaporated water, kg/s, hfg is latent heat of evaporation of water, J/kg, T 0 is the outside temperature, and U L is the overall heat loss coefficient, W/m2 °C. Similarly, energy balance for the air stream passing through the regenerator-segment is expressed as,

where Ac is the collector area and Q e is the cooling rate of the desiccant cooling system, which can be evaluated by multiplying the rate of water evaporation, m, by the latent heat of water L at the evaporator pressure, i.e.,

ma dHa ¼ ha ðT s  T a Þdx  hs ðT a  T 0 Þdx

Q e ¼ mL

ð13Þ

where ha and hs are heat transfer coefficients for the air and solution sides, each in W/m2 °C. The amount of water evaporated from the weak solution is evaluated as:

m ¼ 0:622

ma ðp  pai Þ pb a

ð14Þ

where pb and pai are the barometric pressure and initial vapor pressure in air at the regenerator inlet, in mm Hg, respectively. The rate of mass transport of water vapor is given by,

dm ¼ bðps  pa Þ dx

ð15Þ

where b is the mass transfer coefficient, kg/s m2 mm Hg, and the relation between the mass of evaporated water and solution flow rates is given by:

  m C s ¼ C si = 1  ms

ð16Þ

where C si is the initial concentration of the solution at regenerator inlet. For calcium chloride solution, the relationship between the solution temperature, concentration and vapor pressure is given by,

ps ¼ a þ bT s þ

c Cs

COP ¼

ð19Þ

ð20Þ

3. Simulation procedure 3.1. Solving the equations The theoretical model consists of coupled algebraic and nonlinear ordinary differential equations which link the characteristic parameters of air and desiccant solution. An analytical solution is rather difficult and could only be obtained for simplified situations that allow the reduction of the basic equations. In the current study, a numerical solution is obtained by the finite difference method. The method is implemented using the Matlab–Simulink platform. Simulink allows the system to be modeled by drawing a block diagram directly on the screen. The Simulink representation of the system of equations is presented in Fig. 3. Each block represents a calculation unit and may be composed of other more detailed sub-systems. The sub-systems for calculating the mass of evaporated water is shown in Fig. 4. A Matlab computer code is written to perform the computations and visualize the results. The ordinary differential equations are solved using the fourth-order Runge–Kutta scheme with variable time steps. As the time step decreases, the total time for running the simulation increases. The time step was set as 1 min in the solar radiation calculations.

ð17Þ

where a, b and c are empirical constants [22]. For lithium chloride solution, an artificial neural network (ANN) model is used to extract the required parameter from tabulated data available from the literature [28] as follows,

T s ¼ f ðps ; C s Þ

ð18Þ

where T s , ps , and C s are the solution temperature, vapor pressure, and concentration, respectively The overall coefficient of performance of the system, COP, can be evaluated from the following expression:

Fig. 3. Matlab–Simulink representation of the model.

2772

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

Fig. 4. Sub-blocks for calculating the system.

2773

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

3.2. Solar radiation module

(a)

The sunshine period is evaluated for the specified location (defined by the latitude angle, L) by knowledge of the day number of the year, n, evaluating the declination angle, d (Eq. (4)), then substituting in Eq. (10). Eq. (1) is used to evaluate the instantaneous value of total radiation on the tilted surface during the sunshine period. The parameters in Eq. (1) are calculated from Eqs. (2)–(9). 3.3. Collector/regenerator module The system of equations from (12)–(17) have six unknowns; which are: m; T a ; T s ; pa ; ps and C s . Given the input values of mass flow rates of air and solution, air temperature, and solution concentration at the inlet of any segment, along with the physical properties of the working desiccant, the output values are obtained using the above equations by a step-by-step analysis up to the outlet. An iterative procedure is used to obtain a numerical solution with the criterion set priory. The heat and mass transfer coefficients are evaluated by using available correlations from the literature [29].

Ts

(b)

3.4. Artificial neural network module In general, ANNs are simply mathematical techniques designed to accomplish a variety of tasks. Neural networks can be configured in various arrangements to perform a range of tasks including pattern recognition, data mining, classification, forecasting and process modeling. ANNs are composed of attributes that lead to perfect solutions in applications where we need to learn a linear or nonlinear mapping. Some of these attributes are: learning ability, generalization, parallel processing and error endurance. These attributes would cause the ANNs solve complex problem methods precisely and flexibly [30,31]. As stated above, the thermo-physical properties of LiCl are obtained by the artificial neural network (ANN). The ANN is a computing system, made up of a number of simple, highly interconnected processing elements, each of which processes information by its dynamic state response to external inputs. The input nodes are the previous lagged observations while the output provides the forecast for the future value. Hidden nodes with appropriate nonlinear transfer functions are used to process the information received by the input nodes. If we denote the ith input as xi = [Cs, ps] and the output as y = [Ts], then we can write the mapping from the inputs to the output performed by the processing elements in this case as [31,32]:

y¼f

X

xi wij þ b



ð21Þ

where y and xi are the output and inputs of the ANN model and wij are weighting factors, and b is a constant. Layers of neural nodes are connected together composing an ANN as shown in Fig. 5. Inputs could be connected to many nodes with various weights, resulting in a series of outputs, one per node. The connections are multiplied by the weights associated with that particular node with which they interconnect. The network consists of input layer, hidden layer and output layer with many more connections than nodes. The network is said to be fully connected if every output from one layer is passed along to every node in the next layer. Many activation functions used in ANNs nowadays produce a continuous value rather than discrete. One of the most popular activation functions used is the logistic activation function or more popularly referred to as the sigmoid function. This function is semilinear in characteristic, differentiable and produces a value be-

Cs

Ps

Fig. 5. Schematic of neural network: (a) the role of Neural Network and (b) multilayer of neural network.

tween 0 and 1. The mathematical expression of this sigmoid function is:

f ðnetj Þ ¼

1 1 þ ecnetj

ð22Þ

when c is large, the sigmoid becomes like a threshold function and when is c small, the sigmoid becomes more like a straight line (linear). When c is large learning is much faster but a lot of information is lost, however when c is small, learning is very slow but information is retained. Because this function is differentiable, it enables the backpropagation algorithm to adapt the lower layers of weights in a multilayer neural network. Where c controls the firing angle of the sigmoid [33]. Activation function for the inner connection between input layer and hidden layer is linear whereas between the hidden layer and output layer it is sigmoid. For activation function from the input layer, c = 0.05 while for the activation function to the output layer, c = 1.5. In the current study, the technique used for modifying weights in the appropriate direction is the back propagation learning scheme. Several studies have found that a three-layered NN with one hidden layer can approximate any nonlinear function to any desired accuracy [33]. The method is based on an analysis of how a change in any particular weight influences the output of the network. After such analysis is done, the designer understands how to change the weights to achieve the specified values for the outputs. For designing ANN model, the effort is made to identify the best fitted network for the desired model according to the characteristics of the problem and ANN features. First of all, one must construct a chain similar to that presented in Fig. 5. This chain examines the influence of any weight factor on the output value and, hence, on the error value. At the learning stage, all weights in the network are initialized to small random values. The algorithm uses a learning set, which consists of input–desired output pattern pairs. Each input–output pair is obtained by the offline processing of historical data.

We consider as an error the difference between the actual output and the desired one. The learning rule gradually adjusts the weights until the performance function, which is the sum square error (SSE), i.e. difference between target, t, output (expected) and the actual, y, output from the ANN falls below a certain threshold or minimized, m X SSE ¼ ðtk  yk Þ2

(a)

ð23Þ

k¼1

Back propagation learning updates the network weights and biases in the direction in which the performance function decreases most rapidly, i.e., the negative of the gradient.

wkþ1 ¼ wk  ak g k

ð24Þ

where wkþ1 are updated weights, wk are current weights, g k is current gradient performance function, and ak is the learning rate. In this work the network has two input elements (ps ;C s ) and one output (T s ) and 10 element hidden layer, as shown in Fig. 5b. More details about the verification of the ANN model are given by Aly et al. [26]. 4. Results and discussion The performance of the solar collector/regenerator is influenced by design parameters (regenerator length, solution flow rate, working solution concentration and air flow rate) and ambient conditions (air temperature and vapor pressure in the flowing air). A sensitivity analysis is performed by varying the parameters of interest one at a time, while keeping all others fixed at given values. These key parameters are investigated in the following paragraphs. The variation of the solution temperature during the regeneration period, which lasts from sun rise to sun set, is demonstrated for calcium chloride solution in Fig. 6. As the solution flows from the regenerator inlet to exit, it is expected that the solution temperature will be maximum at the end of the regenerator. On the other hand, the maximum temperature of the solution will follow the solar radiation intensity along the day time. It can be observed from Fig. 6 that the solution temperature increases with the decrease in the flow rate. In order to analyze the effect of air mass flow rate on the regeneration process, the solution mass flow rate, ms , is settled at 20 kg/ h and the air mass flow rate, ma , is investigated for the two solutions at two different values (100 kg/h and 200 kg/h). The vapor pressure difference between the regenerated solution and flowing air is plotted versus the regenerator length. For a given regenerator length, the vapor pressure, which is the mass transfer potential, is directly proportional with the rate of water evaporation, when the mass transfer coefficient is assumed constant. As shown in Fig. 7, the vapor pressure difference has a maximum at a given length of the regenerator for the two solutions. The length, at which the

Fig. 6. Variation of solution temperature during the day time for different values of calcium chloride solution flow rate.

Vapor pressure difference, mmHg

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

14 200

12 10

Ma=100kg/hr

8 Ms=20 kg/hr To=40C Tai=40C Tsi=40C Csi=40% pai=20mmHg Day=21June

6 4 2

0

1

2

3

4

5

Regenerator Length,m

(b)

Vapor pressure differnce, mmHg

2774

3.5 3 2.5

1.5 1 0.5

200

Ms=20 kg/hr To=40 C Tai=40 C Tsi=40 C Csi=40% Pai=20 mmHg Day=21 June

2

0

1

Ma=100 kg/hr

2

3

4

5

Regenerator Length, m Fig. 7. Variation of vapor pressure at regenerator exit with different values of air flow rate, at noon time, (a) calcium chloride, (b) lithium chloride.

maximum rate of evaporation occurs, increases with the air flow rate. Concerning the effect of solution inlet concentration on regeneration process, the decrease of solution concentration can effectively improve the regenerator performance, though it sacrifices solution outlet concentration. It is also evident that the vapor pressure difference for the specified operating conditions is higher for CaCl2 than LiCl solutions. This is mainly due to the difference in thermo-physical properties of the two solutions, where the vapor pressure on the surface of CaCl2 is higher than LiCl at the same temperature and solution concentration. Variation of vapor pressure difference with the regenerator length at different values of solution mass flow rate, for the two solutions, is depicted in Fig. 8. These data are evaluated at solution flow rate ranging from 10 kg/h to 40 kg/h. It can be noted that a maximum rate of water evaporation occurs at a specified regenerator length, for the given operating conditions. The location at which the maximum vapor pressure difference occurs, as well as the value of the vapor pressure difference, is also dependent on the type of solution. Fig. 8b shows that the decrease in the vapor pressure difference for LiCl solution seems to be rapid compared with CaCl2 solution, however the drop in the solution vapor pressure in the two cases is nearly the same but the main difference between the two cases is in the operating range of the vapor pressure difference. The variation of solution concentration at regenerator exit for different values of solution flow rates of the two solutions is shown in Fig. 9. It can be observed that the solution concentration increases with regenerator length due to the continuous evaporation of water. At the end of the regenerator, it can be noted that the solution concentration is higher for the cases with higher vapor pressure difference. The solution concentration at the end of regenerator is highly dependent on the rate of water evaporation, whereas an increase in water evaporation increases the solution concentration. In general, CaCL2 has higher potential for mass transfer compared with LiCl (see Fig. 8) and consequently the rate of evaporation of water as well as the solution concentration will

(a)

Vapor pressure difference, mmHg

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

14 12 20

10

30

Ms=40kg/hr

10

8 Ma=100kg/hr To=40 C Tsi=40 C Csi=40% pai=20 mmHg Day=21 June

6 4 2

0

1

2

3

4

5

(b)

Vapor pressur difference, mmHg

Regenerator length,m 3.5

2775

liquid desiccant regenerator. The model is implemented using the Matlab–Simulink platform and its feasibility is established. The proposed model can be successfully used for investigating the effect of different operating parameters under different ambient conditions and for predicting the overall performance of the system including the ease of examining the design parameters and ability to include the varying solar radiation. The simulation has shown that the maximum temperature of the solution follows the solar radiation intensity along the day time with the maximum solution temperature increasing with the decrease in the flow rate. It has been also shown that the vapor pressure difference has a maximum value for a given regenerator length. For specified operating conditions, the thermo-physical properties of the two solutions have shown a pronounced effect on the performance of the regeneration process.

10

3

Ma=100 kg/hr To=40 C Tsi=40 C Csi=40% Pai=20 mmHg Day= 21 June

20 30

2.5

Ms=40 kg/hr

2

References

1.5 1 0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Regenerator length, m Fig. 8. Variation of vapor pressure at the regenerator exit for different values of solution flow rate, at noon time, (a) calcium chloride, (b) lithium chloride.

Fig. 9. Variation of solution concentration at regenerator exit for different values of solution flow rate: (a) calcium chloride and (b) lithium chloride.

be higher as shown in Fig. 9. On the other hand, lower values of solution flow rate results in significant change in the solution concentration, where the solution concentration increases rapidly when compared with higher values of solution flow rate.

5. Conclusion The modeling and simulation of solar-powered desiccant regenerator used for open absorption cooling cycle is presented. The source of input heat is assumed to be the total radiation incident on a tilted surface. A finite difference method is used to simulate the combined heat and mass transfer processes that occur in the

[1] Mei L, Dai YJ. A technical review on use of liquid-desiccant dehumidification for air conditioning application. Renew Sustain Energy Rev 2008;12(3): 662–89. [2] Gommed K, Grossman G. Experimental investigation of a liquid desiccant system for solar cooling and dehumidification. Sol Energy 2007;81(1):131–8. [3] Kakabaev A, Khandurdyev A. Absorption solar refrigeration unit with open regeneration of solution. Gliotekhnika 1969;5(4):28–32. [4] Li X, Yang H. Energy and economic performance analysis of an open cycle solar desiccant dehumidification air-conditioning system for application in Hong Kong. Sol Energy 2010;84(12):2085–95. [5] Dieckmann JRK, Brodrick J. Liquid desiccant air conditioners. ASHRA J 2008:91–5. [6] Krause M, Saman W, Vajen K. Open cycle liquid desiccant air conditioning systems-theoretical and experimental investigations. In: Proceeding of the ANZSES conference, Dunedin, New Zealand; 2005. [7] Halliday SP, Beggs CB, Sleigh PA. The use of solar desiccant cooling in the UK: a feasibility study. Appl Therm Eng 2002;22(12):1327–38. [8] Maroudaki P, Beggs CB, Sleigh PA, et al. The potential for solar powered singlestage desiccant cooling in southern Europe. Appl Therm Eng 2002;22(10): 1129–40. [9] Grossman G. Solar-powered systems for cooling, dehumidification and airconditioning. Sol Energy 2002;72(1):53–62. [10] Yang R, Wang PL. Experimental study of a forced convection solar collector/ regenerator for open cycle absorption cooling. Trans ASME Sol Energy Eng 1994;116:194–9. [11] Ameel TA, Gee KG, Wood BD. Performance predictions of alternative, low cost absorbents for open-cycle absorption solar cooling. Sol Energy 1995;54(2): 65–73. [12] Xiu-Wei L, Xiao-Song Z, Geng W, Rong-Quan C. Research on ratio selection of a mixed liquid desiccant: mixed LiCl–CaCl2 solution. Sol Energy 2008;82(12): 1161–71. [13] Donate M, Rodriguez L, De Lucas A, Rodriguez JF. Thermodynamic evaluation of new absorbent mixtures of lithium bromide and organic salts for absorption refrigeration machines. Int J Refrig 2006;29(1):30–5. [14] Siddiqui MA. Economic analysis of absorption systems: part B – optimization of operating parameters. Energy Convers Manage 1997;38(9):905–18. [15] Liu XH, Yi XQ. Mass transfer performance comparison of two commonly used liquid desiccants: LiBr and LiCl solutions. Energy Convers Manage 2011;52(1): 180–90. [16] Collier RK. The analysis and simulation of an open-cycle absorption refrigeration system. Sol Energy 1979;23(4):357–66. [17] Kanoglu M, Carpinlioglu MO, Yildirim M. Energy and exergy analyses of an experimental open-cycle desiccant cooling systems. Appl Therm Eng 2004;24(5–6):919–32. [18] Hawlader MNA, Novak KS, Wood BD. Unglazed collector/regenerator performance for a solar assisted open cycle absorption cooling system. Sol Energy 1993;50(1):59–73. [19] Haim I, Grossman G, Shavit A. Simulation and analysis of open cycle absorption systems for solar cooling. Sol Energy 1992;49(6):515–34. [20] Yang R, Yan WJ. Simulation study for an open-cycle absorption solar-cooling system operated in a humid area. Energy 1992;17(7):649–55. [21] Yang R, Wang PL. A simulation study of the performance evaluation of singleglazed and double-glazed collectors/regenerators for an open-cycle absorption solar cooling system. Sol Energy 2001;71(4):263–8. [22] Alizadeh A, Saman W. Modeling and performance of a forced flow solar collector/regenerator using liquid desiccant. Sol Energy 2002;72(2):143–54. [23] Shen C, He Y-L, Liu Y-W, Tao W-Q. Modeling and simulation of solar radiation data processing with Simulink. Simul Model Pract Theory 2008;16(7):721–35. [24] Nia FE, van Paassen D, Saidi MH. Modeling and simulation of desiccant wheel for an air conditioning. Energy Build 2006;38(10):1230–9. [25] da Silva RM, Fernandes JLM. Hybrid photovoltaic thermal (PV/T) solar systems simulation with Simulink/Matlab. Sol Energy 2010;84(12):1985–96.

2776

A.A. Aly et al. / Energy Conversion and Management 52 (2011) 2768–2776

[26] Aly AA, Zeidan E-SB, Hamed AM. Performance evaluation of open-cycle solar regenerator using artificial neural network technique. Energy Build 2011;43(2–3):454–7. [27] Hsieh K. Solar energy engineering. Prentice-Hall; 1986. [28] Zaetsev EZ, Aseev GG. Physical–chemical properties of binary non-organic solutions. Khemia, Leningrad, USSR; 1988. [29] McCabe WL, Smith JC, Harriott P. Unit operation of chemical engineering. McGraw-Hill; 1985.

[30] Farrell J, Baker W. Learning control systems. In: Antsaklis PJ, Passino KM, editors. An introduction to intelligent and autonomous control. Norwell, MA (USA): Kluwer Academic; 1992. [31] Peng TM, Hubele N, Karady GF. Advancement in the application of neural networks for short-term load forecasting. IEEE Trans Power Syst 1992;7(1):250–6. [32] Kung AY. Digital neural networks. Prentice Hall: Englewood Cliffs, NJ (USA); 1993. [33] Hinton GE. How neural networks learn from experience. Sci Am 1992;105–9.