Sage Journals: Discover world-class research

Abstract

An attempt has been made to improve the performance of dehumidifier using liquid desiccant in this study. Experiments have been conducted based on L₁₆ orthogonal array in order to experimentally investigate the influence of dehumidification process parameters like desiccant temperature, desiccant concentration, desiccant flow rate, air flow rate, and relative humidity on the effectiveness of dehumidifier. The optimum dehumidification condition has been found through Taguchi method. It has been confirmed that the optimum condition exhibited better effectiveness for the dehumidification process.

1. Introduction

Desiccant dehumidifiers are emerging as a promising alternative to achieve humidity control in variety of applications with high latent loads and low humidity requirements such as in supermarkets and pharmaceuticals. Both solid and liquid desiccants are being used in practice, but liquid desiccant absorbers are becoming more attractive because of the possibility of simultaneous cooling during the process of dehumidification. Liquid desiccants also have the capability to absorb inorganic and organic air contaminants and microorganisms (Chung et al. [1]; Öberg and Goswami [2]).

Most of the systems already developed employ solid desiccants, with relatively high regeneration temperatures. One alternative is the use of liquid-desiccant systems. In these systems, lower regeneration temperatures can be employed, allowing for a more efficient use of heat from low temperature sources, for example, flat plate solar collectors [3].

In the air conditioning systems, the use of liquid desiccant systems has become more popular in the past decades due to the need for reduction in the consumption of energy [4, 5]. The capability of these systems in handling latent heat in the space that will be conditioned by a dehumidifying process also allows control of the humidity without the overcool/reheat scheme as done in a regular ventilating and air conditioning system [6].

The vapour pressure difference between the air and the desiccant is the moisture transfer driving force, which makes desiccant to absorb the moisture or deliver the moisture to the air. In the liquid desiccant air conditioning system, air is dehumidified by directly contacting the concentrated desiccant solution in the dehumidifier and the diluted desiccant solution from the dehumidifier will come to the regenerator in which it is reconcentrated. Low grade heat can be utilized to regenerate the desiccants [7 –10] such as solar energy and waste heat.

The dehumidifier is one of the essential parts of the system, where the air is dehumidified by a liquid desiccant. Its performance greatly influences the performance of the whole system. Compared to other types of dehumidifier, the packed dehumidifier has drawn more attention due to its compactness [11]. The heat and mass transfer process in the packed dehumidifier is affected by many parameters, such as the relative flow direction of the air to the desiccant, the type and material of the packing, and the inlet parameters of the air and the desiccant. The dehumidification process is so complex that pure theoretical study usually fails to give satisfactory results [12].

Bansal et al. [13] studied the performance comparison of an adiabatic and an internally cooled structured packed-bed dehumidifier. The performance of this absorber has been evaluated at varying desiccant flow rates. The moisture removal from the air, the effectiveness of the dehumidifier, and the mass transfer coefficients between air and solution have been compared for the dehumidifier operation with and without internal cooling.

Xiong et al. [14] experimentally demonstrated a two-stage solar powered liquid-desiccant dehumidification system, for which two kinds of desiccant solution (lithium chloride and calcium bromide) are fed to the two dehumidification stages separately. In this study, air moisture (latent) load is separately removed by a predehumidifier using cheap calcium chloride (CaCl₂) and a main dehumidifier using stable lithium bromide (LiBr). Side-effect of mixing heat rejected during dehumidification process is considerably alleviated by an indirect evaporative cooling unit added between the two dehumidification stages.

In this study, dehumidification process using CaCl₂ liquid desiccant has been carried out in two stages. In between dehumidification stages, process air is sensibly cooled using indirect evaporative cooler and then further dehumidification process is carried out. The process parameters affecting the effectiveness of dehumidifier, namely, desiccant concentration, temperature, flow rate, air flow rate, and relative humidity, have been considered in this study. A L₁₆ (4⁵) orthogonal array is used for the conduct of experiments and Taguchi method is employed to obtain the optimum parametric condition in this study.

2. Description of Experimental Set Up

The schematic diagram of experimental set up is shown in Figure 1. It consists of three important components, namely, dehumidifiers, indirect evaporative cooler, and regenerator. The dehumidifier (20 cm deep, 10 cm wide, and 20 cm long) is made with acrylic sheet (thickness 5 mm) to see inside during operation and inner core of dehumidifier is made of corrugated PVC sheets (200 mm × 5 mm × 200 mm). In the inner core, the gap 5 mm is maintained between corrugated PVC sheets and nonwoven cloth is wounded over the corrugated sheet surface to promote wetting by air and desiccant. A spray head is used to distribute liquid desiccant over the structured packing. Process air is made to meet with liquid desiccant in crosswise direction. The weak solution is collected at the bottom of the dehumidifier.

Figure 1:

Schematic of experimental set up.

The outer casing (50 cm in diameter and 75 cm in height) of regenerator is made of PVC and packed with balls (1.5″ diameter). The weak desiccant solution from dehumidifier is pumped to heater and the hot weak solution is then sprayed over the regenerator through nozzle. The atmospheric air is made to meet with weak solution in counter clockwise direction and the strong desiccant solution is collected at the bottom of the regenerator.

Indirect evaporative cooler (50 cm in length and 20 cm in height, 10 cm in width), in which outer body is made of acrylic sheet, is used to cool the process air coming from the dehumidifier. The process air is passed through the channels over which evaporative cooling takes place due to the contact of atmospheric air and water, thereby cooling effect is transferred to process air.

The dry bulb temperature of process air is measured at various points using k-type thermocouple. Relative humidity of air, desiccant flow rate, concentration of desiccant, and air flow rate are measured with humidity sensors, rotameter, hydrometer, and anemometer, respectively.

Dehumidification process is carried out with four steps by using the experimental set up shown in Figure 2.

The process air is passed through dehumidifier and required flow rate of air which is obtained by controlling fan speed.

The solution pumps are started and flow rates are adjusted to required level using control valves.

Temperature of diluted desiccant solution is adjusted with heating system.

Indirect evaporative cooler is started to cool process air.

Figure 2:

Photographic view of experimental set up.

Process air is passed through first dehumidifier where moisture is reduced from state 1 to state 2. Then dehumidified air is sent through indirect evaporative cooler where temperature is reduced from state 2 to state 3 at constant specific humidity. Afterwards, the process air is passed through second dehumidifier, where moisture is again reduced from state 3 to state 4. All processes are indicated in Figure 3. The effectiveness of dehumidifier (ε) is determined as follows.

ε = \frac{w_{i} - w_{o}}{w_{i} - w_{e}},

(1)

where w_i is specific humidity of inlet air; w_o is specific humidity of outlet air; w_e is equilibrium specific humidity of solution and air; and w_e is the humidity ratio of air in equilibrium with calcium chloride solution at the interfacial area. It is calculated from the following ordinary equation:

w_{e} = \frac{0.622 P_{vs}}{P_{a} - P_{vs}} .

(2)

Figure 3:

Psychrometric process.

Vapour pressure is important properties which determine the air humidity in equilibrium with liquid desiccant at surface. Fumo and Goswami [15] developed second order polynomial and coefficient which were obtained from a curve fit [16]:

P_{vs} = (a_{o} + a_{1} T + a_{2} T^{2}) + (b_{o} + b_{1} T + b_{2} T^{2}) X + (c_{o} + c_{1} T + c_{2} T^{2}) X^{2},

(3)

where the constant is given for dehumidification process as a_o = 4.58208, a₁ = −0.159174, a₂ = 0.0072594; b_o = −18.3816, b₁ = 0.5661, b₂ = −0.019314; c_o = 21.312, c₁ = −0.666, c₂ = 0.01332.

3. Taguchi Method

This method is a powerful statistical tool for improving the performance of the design, process, and product with a considerable reduction in investigational time and cost, Wu and Chang [17]. It employs the concepts of orthogonal array which is a set of well-balanced experiments and signal-to-noise (S/N) ratio which is the ratio of the mean (signal) to the standard deviation (noise). This method defines three categories of quality characteristics such as lower-the-better, larger-the-better, and nominal-the-best and involves five main steps as follows.

Identify process parameters influencing output response of the process.

Choose a suitable orthogonal array to carry out experiments.

Examine the results to find out optimum parametric condition.

Run a confirmatory test using the optimum parametric condition.

Five process control parameters, namely, desiccant concentration (A), desiccant temperature (B), desiccant flow rate (C), air flow rate (D), and air relative humidity (E), each at four levels, were considered in this study and are listed in Table 1. An orthogonal array L₁₆ (4⁵) was selected for the conduct of experiments. Effectiveness of dehumidifier was treated as output response with the category of quality characteristics “larger-the-better.” The S/N ratio for this response can be estimated by using

S / N (dB) = - 10 \log_{10} (\frac{1}{n} \sum_{i = 1}^{n = 4} \frac{1}{ε_{i}^{2}}),

(4)

where ε_i is response value for an experimental condition.

Table 1:

Process control parameters and their levels.

Factor	Control factor	Level 1	Level 2	Level 3	Level 4

A	Desiccant concentration (kg/kg)	0.30	0.35	0.40	0.45
B	Desiccant temperature (°C)	25	30	35	40
C	Desiccant flow rate (kg/m²'s)	1.25	1.75	2.25	2.75
D	Air flow rate (kg/m²'s)	1	1.33	1.66	2
E	Air relative humidity	0.70	0.75	0.80	0.85

4. Results and Discussion

4.1. S/N Ratio Response

The S/N ratio of effectiveness was calculated for each process parametric setting and their values are given in Table 2. In order to find ranking and optimum level of the process parameters, average S/N ratio response was estimated and the corresponding details are given in Table 3. The response graph shown in Figure 4 described the variation of each process control parameter on the performance of the dehumidification process [18, 19]. The percentage contribution of each parameter to dehumidifier effectiveness was shown in Figure 5. The optimum process condition A₃ B₁ C₃ D₁ E₄ (desiccant concentration of 0.4 kg/kg of solution, desiccant temperature of 25°C, desiccant flow rate of 2.25 kg/m²'s, air flow rate of 1 kg/m²'s, and air relative humidity 0.85) was noted in this study.

Table 2:

Experiments and S/N ratio.

Ex. number	A	B	C	D	E	Effectiveness, ε (%)					S/N ratio (dB)
Ex. number	A	B	C	D	E	ε₁	ε₂	ε₃	ε₄	ε_avg	Y _j

1	0.30	25	1.25	1.00	0.70	59	57	59	58	58	34.88
2	0.30	30	1.75	1.33	0.75	53	53	54	53	53	34.95
3	0.30	35	2.25	1.66	0.80	54	55	54	55	55	35.07
4	0.30	40	2.75	2.00	0.85	53	52	53	53	53	35.13
5	0.35	25	1.75	1.66	0.75	65	64	64	63	64	34.71
6	0.35	30	1.25	2.00	0.80	56	56	56	57	56	34.51
7	0.35	35	2.75	1.00	0.85	46	45	44	45	45	34.36
8	0.35	40	2.25	1.33	0.70	59	58	57	58	58	34.67
9	0.40	25	2.25	2.00	0.80	62	63	64	63	63	34.52
10	0.40	30	2.75	1.66	0.85	65	66	67	65	66	34.40
11	0.40	35	1.25	1.33	0.70	65	64	63	65	64	34.64
12	0.40	40	1.75	1.00	0.75	68	68	68	68	68	34.83
13	0.45	25	2.75	1.33	0.85	61	60	60	59	60	35.07
14	0.45	30	2.25	1.00	0.70	67	68	68	68	68	34.67
15	0.45	35	1.75	2.00	0.75	53	53	52	53	53	34.82
16	0.45	40	1.25	1.66	0.80	46	44	45	45	45	35.03
								Mean			35.196

Table 3:

Average S/N ratio response.

	A	B	C	D	E

Level 1	34.7393	35.7348	34.8806	35.3962	35.3367
Level 2	34.8595	35.6169	35.4124	35.382	34.1588
Level 3	36.2825	34.5866	35.6458	35.0585	34.6939
Level 4	34.9029	34.846	34.8455	34.9475	35.8078
Max.–Min.	1.54326	1.14818	0.80022	0.44865	1.64899
Rank	2	3	4	5	1
Optimum condition	A ₃	B ₁	C ₃	D ₁	E ₄
% Contribution	27.6111	20.5425	14.3169	8.0269	29.5026

Figure 4:

Response curve.

Figure 5:

Percentage of contribution of each parameter.

4.2. Confirmation Experiments

Two test runs have been made for the optimum parametric setting. The values of effectiveness were found to be 70, 72, 70, and 71% and their S/N ratio was 36.99 dB. This S/N ratio value was found within the confidence interval limits.

4.3. Future Scope

Research is still required to improve the performance of dehumidification system through various combinations of parameters and levels by including other process parameters like air temperature, wettability, mass flow ratio, heat, and mass transfer coefficient. The optimization techniques like genetic algorithm and simulation analysis can be used to find the global optimum parametric condition by which dry air can be produced.

5. Conclusion

Taguchi method was applied to find the optimum parametric setting for air dehumidification system with high effectiveness. The optimum setting was found as follows:

solution concentration: 0.4 kg/kg;

solution temperature: 25°C;

solution flow rate: 2.25 kg/m²'s;

air flow rate: 1 kg/m²'s;

relative humidity: 85%.

This setting was checked through the confirmation experiments. From the percentage contribution analysis, it was noted that solution flow rate was the most important parameter for the improvement of effectiveness of dehumidifier system.

Footnotes

Nomenclature

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

The authors are thankful to Aspiration Energy Pvt. Ltd.,Chennai,India,for providing research facilities to carry out this research work.

References

Chung

T. W.

Ghosh

T. K.

Hines

A. L.

, and Novosel

, “Dehumidification of moist air with simultaneous removal of selected indoor pollutants by tri ethylene glycol solution in a packed-bed absorber,” Separation Science and Technology, vol. 30, pp. 1807–1832, 1995.

Öberg

and Goswami

D. Y.

, “A review of liquid desiccant cooling, advantages in solar energy,” American Solar Energy Society, vol. 12, pp. 431–470, 1998.

Öberg

and Goswami

, “A review of liquid-desiccant cooling,” in Advances in Solar Energy, Böer

, Ed., American Solar Energy Society, Boulder, Colo, USA, 1998.

Feyka

and Vafai

, “An investigation of a falling film desiccant dehumidification/ regeneration cooling system,” Heat Transfer Engineering, vol. 28, no. 2, pp. 163–172, 2007.

Jain

and Bansal

P. K.

, “Performance analysis of liquid desiccant dehumidification systems,” International Journal of Refrigeration, vol. 30, no. 5, pp. 861–872, 2007.

Katejanekarn

and Kumar

, “Performance of a solar-regenerated liquid desiccant ventilation pre-conditioning system,” Energy and Buildings, vol. 40, no. 7, pp. 1252–1267, 2008.

Öberg

and Goswami

D. Y.

, “Experimental study of the heat and mass transfer in a packed bed liquid desiccant air dehumidifier,” Transactions of the ASME, Journal of Solar Energy Engineering, vol. 120, no. 4, pp. 289–297, 1998.

Gandhidasan

Rifat Ullah

, and Kettleborough

C. F.

, “Analysis of heat and mass transfer between a desiccant-air system in a packed tower,” Transactions of the ASME, Journal of Solar Energy Engineering, vol. 109, no. 2, pp. 89–93, 1987.

Lowenstein

A. I.

and Gabruk

R. S.

, “The effect of absorber design on the performance of a liquid desiccant air conditioner,” Transactions of the ASHRAE, vol. 98, no. 1, pp. 712–720, 1992.

10.

Jain

Dhar

P. L.

, and Kaushik

S. C.

, “Experimental studies on the dehumidifier and regenerator of a liquid desiccant cooling system,” Applied Thermal Engineering, vol. 20, no. 3, pp. 253–267, 2000.

11.

Öberg

and Goswami

D. Y.

, “Experimental study of the heat and mass transfer in a packed bed liquid desiccant air dehumidifier,” Journal of Solar Energy Engineering, vol. 120, no. 4, pp. 289–297, 1998.

12.

Liu

X. H.

Zhang

K. Y.

, and Jiang

, “Experimental study on mass transfer performances of cross flow dehumidifier using liquid desiccant,” Energy Conversion and Management, vol. 47, no. 15–16, pp. 2682–2692, 2006.

13.

Bansal

Jain

, and Moon

, “Performance comparison of an adiabatic and an internally cooled structured packed-bed dehumidifier,” Applied Thermal Engineering, vol. 31, no. 1, pp. 14–19, 2011.

14.

Xiong

Z. Q.

Dai

Y. J.

, and Wang

R. Z.

, “Investigation on a two-stage solar liquid-desiccant (LiBr) dehumidification system assisted by CaCl₂ solution,” Applied Thermal Engineering, vol. 29, no. 5–6, pp. 1209–1215, 2009.

15.

Fumo

and Goswami

D. Y.

, “Study of an aqueous lithium chloride desiccant system: air dehumidification and desiccant regeneration,” Solar Energy, vol. 72, no. 4, pp. 351–361, 2002.

16.

Uemura

, “Studies on the lithium chloride-water absorption refrigeration machine',” Technology Reports of Kansai University, vol. 9, pp. 71–84, 1967.

17.

D. H.

and Chang

M. S.

, “Use of Taguchi method to develop a robust design for the magnesium alloy die casting process,” Materials Science and Engineering A, vol. 379, no. 1–2, pp. 366–371, 2004.

18.

Belavendram

, Quality by Design, Prentice Hall, Singapore, 5th edition, 1999.

19.

Ross

P. J.

, Taguchi Techniques for Quality Engineering, McGraw-Hill, Singapore, 4th edition, 1999.

Optimization of Liquid Desiccant Dehumidifier Performance Using Taguchi Method

Abstract

1. Introduction

2. Description of Experimental Set Up

3. Taguchi Method

4. Results and Discussion

4.1. S/N Ratio Response

4.2. Confirmation Experiments

4.3. Future Scope

5. Conclusion

Footnotes

Nomenclature

Conflict of Interests

Acknowledgment

References