An Experimental Study On the Dehumidification Performance of a Low-flow Falling-film Liquid Desiccant Air-conditioner

Abstract

The dehumidifier is one of the main components in open-cycle liquid desiccant air-conditioning systems. An experimental study was carried out to evaluate the performance of a solar thermally driven, low-flow, falling-film, internally-cooled parallel- plate liquid desiccant air-conditioner in Kingston, Ontario at Queen’s University. A solution of LiCl and water was used as the desiccant. Unlike high-flow devices, the low-flow of desiccant solution flowing across the unit’s dehumidifier and regenerator sections produces large variations in solution concentration. In this study, a series of tests were undertaken to evaluate the performance of the dehumidifier section of the unit.
Results presented are based on mass flow and energy transport measurements that allowed the moisture transport rate between the air and liquid desiccant solution to be determined. Based on these results, a relationship between the desiccant concentration and the rate of dehumidification rate was found and the effect of inlet-air humidity on the dehumidification effectiveness identified. The moisture removal rate of the system was found to range from 1.1 g/s to 3.5 g/s under the conditions evaluated. These result corresponded to an average dehumidification effectiveness of 0.55.

Full text

Procedia Computer Science 52 ( 2015 ) 796 – 803
1877-0509 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Conference Program Chairs
doi: 10.1016/j.procs.2015.05.135
ScienceDirect
Available online at www.sciencedirect.com
The 5th International Conference on Sustainable Energy Information Technology
(SEIT 2015)
An experimental study on the dehumidification performance of a
low-flow falling-film liquid desiccant air-conditioner
S. Bouzenadaa*, C. McNevinb, S. Harrison c, A. N. Kaabid
a,d Laboratory of Energy and Environment, Department of Architecture, University 3 Constantine, 25000, Algeria
b,c Solar Calorimetry Laboratory, Department of Mechanical and Materials Engineering, Que en’s University, Kingston, Ontario, Canada
Abstract
The dehumidifier is one of the main components in open-cycle liquid desiccant air-conditioning systems. An experimental
study was carried out to evaluate the performance of a solar thermally driven, low-flow, falling-film, internally-cooled parallel-
plate liquid desiccant air-conditioner in Kingston, Ontario at Queen’s University. A solution of LiCl and water was used as the
desiccant. Unlike high-flow devices, the low-flow of desiccant solution flowing across the unit’s dehumidifier and regenerator
sections produces large variations in solution concentration. In this study, a series of tests were undertaken to evaluate the
performance of the dehumidifier section of the unit. Results presented are based on mass flow and energy transport
measurements that allowed the moisture transport rate between the air and liquid desiccant solution to be determined. Based on
these results, a relationship between the desiccant concentration and the rate of dehumidification rate was found and the effect of
inlet-air humidity on the dehumidification effectiveness identified. The moisture removal rate of the system was found to range
from 1.1 g/s to 3.5 g/s under the conditions evaluated. These result corresponded to an average dehumidification effectiveness of
0.55.
© 2015 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Conference Program Chairs.
Keywords: liquid desiccant; air-conditioning; dehumidification effectiveness; lithium chloride; falling film; low flow; HVAC
1. Introduction
The current energy crisis, climate change, and increased air-conditioning demand have generated a need to
develop new space cooling technologies based on renewable energy sources.
_________
* Corresponding author. Tel.: +213 663 330 227; Fax: +213 31 904 580.
E-mail address: saliha_610@yahoo.fr
© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Conference Program Chairs

797
S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
Nomenclature
Subscripts
h
specific enthalpy (kJ/kg)
a
air
hf
enthalpy of water vapour (kJ/kg)
a-d
air to desiccant
hfg
latent heat of condensation (kJ/kg)
abs
absorbed
ṁ
mass flow rate (kg/s)
c
conditioner
P
atmospheric pressure (Pa)
cw
cooling water
p
partial pressure (Pa)
d
desiccant solution
Q
̇
power (kJ/s)
d-cw
desiccant to cooling water
X
mass fraction (-)
dil
dilution
ε
effectiveness (-)
eq
equilibrium
ω
absolute humidity (kgw/kga)
in
inlet
out
outlet
Liquid desiccant air-conditioning systems (LDAC) have emerged as a potential alternative to conventional
vapour compression systems for air-conditioning. Desiccants are materials that have high affinity for absorbing
water vapour. This technology, which can efficiently serve large latent loads, can greatly improve indoor air quality
by allowing for easier humidity control while reducing the electrical energy consumption of traditional vapour
compression systems. Air dehumidification is an important process not only in industrial applications, but also in
space cooling for occupant health and comfort. In the case of desiccant based dehumidification systems, much of the
thermal energy required to operate the systems can be drawn from sustainable sources such as solar thermal
collectors. The present work studies one such desiccant system. The dehumidifier component (also referred to as the
conditioner) of this system is the principle topic under investigation as it is directly responsible for the
dehumidification of the air and the other components of the system exist to support its operation. Its performance
greatly influences the overall performance of the complete system.
In the dehumidifier of the system studied, the concentrated desiccant solution is brought into direct contact with
the process-air to absorb the moisture from the inlet-air (i.e., process air stream)1. During this process, the solution
becomes diluted by the water removed from the air-stream and its ability to further absorb moisture is reduced. In
order to reuse the desiccant solution in the process, the solution must be pumped through the regenerator section of
the unit. Regeneration occurs when the desiccant is exposed to a scavenging air-stream (separate from the
dehumidified process-air stream) while being heated. The heat drives off the absorbed water from the solution,
increasing its salt concentration and regenerating its capability to remove moisture. It is the heat consumed in the
regeneration process that can be provided from low-grade waste heat processes or renewable energy sources such as
solar thermal energy2.
An early liquid desiccant system was developed and experimentally tested by Löf3, who used triethylene glycol
as the hygroscopic solution. In this system, air was dehumidified and simultaneously cooled in an absorber and then
evaporatively cooled. Recently, the concept of air dehumidification by a liquid desiccant has again been investigated
by numerous researchers. Mesquita et al.1 developed mathematical and numerical models for falling-film liquid
desiccant dehumidifiers using three approaches. The first approach was based on heat and mass transfer correlations.
The second was numerically solved by the finite-difference method assuming constant film thickness. The third
approach introduced a variable film thickness to the numerical model. All approaches assumed fully developed
laminar flow for the liquid and air streams. Jain and Bansal4 proposed an analysis of packed bed dehumidifiers for
three commonly used desiccant materials; triethylene glycol, lithium chloride, and calcium chloride, using
effectiveness values drawn from the literature. The analysis revealed significant variations and anomalies in trends
between the predictions of various correlations for the same operating conditions. This highlighted the need for
benchmarking the performance of desiccant dehumidifiers. Sreelal and Hariharan5 studied the effect of air velocity
on the heat and mass transfer in a falling film type liquid desiccant dehumidifier using a two-dimensional simulation
of a two-phase system. The results indicated that the air velocity had a significant impact on the dehumidification
process.
In the present work, an experimental study was carried out using a solar thermally driven LDAC to evaluate the
dehumidification effectiveness of an internally cooled, low-flow, falling-film, parallel plate desiccant dehumidifier,
designed and constructed by AIL research6. A liquid solution of lithium chloride and water was used as the
desiccating agent. This particular system has been studied previously in several works with a focus on the complete

798 S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
system’s performance and modelling2,7,8. In the present work, the process parameters affecting the performance of
the dehumidifier, namely the desiccant inlet concentration, and the inlet-air relative humidity, were analysed. The
unit investigated had also recently suffered freezing damage in the internal cooling channels associated with the
conditioner plates. The freezing damaged of some of the flow channels resulting in the leakage of cooling water into
the desiccant solution. Most of the channels were repaired prior to this study but a small volume of continuous
leakage could not be stopped. It was feared that the slow leakage of cooling water into the desiccant solution would
reduce the performance of the unit, as the additional water might be diluting the desiccant solution faster than the
regenerator could reject the moisture. The performance of the dehumidifier was also analyzed to evaluate the
effectiveness of the repair.
The resulting performance data for the dehumidifier is presented in terms of moisture removal rates and
dehumidification effectiveness as determined from process-air measurements as described below.
2. System Description and Experimental study
The liquid desiccant air-conditioning system shown in Fig. 1(a) is principally composed of a dehumidifier and
regenerator, as well as: blowers; pumps; a desiccant sump; boiler; evaporative cooling tower; data acquisition; and
control equipment. Another major component is the evacuated tube solar thermal collector array shown in Fig. 1(b).
The array has a total area of 95 m² and is automatically controlled and fully instrumented. The array stores heat in a
pair of 435 litre insulated storage tanks. The capacity of these tanks is insufficient to store large quantities of energy
for the LDAC as their primary function is to operate as buffers between the solar array and the LDAC as both the
operating times and the water flow rates are different for each system. During times of low solar availability, the
natural gas boiler provided additional thermal energy to the system to maintain the hot water set-point temperature2.
Fig. 1. (a) Liquid desiccant air handling unit and cooling tower as installed at Queen's University; (b) Photo of the solar thermal collector array
The LDAC uses an internally cooled parallel-plate heat and mass exchangers for the conditioner, shown in Fig. 2
(a). The cooling water flows inside narrow channels within each plate, Fig. 2(b), while a thin film of desiccant
solution falls down the outer faces of each plate. A source of cooling water is required for continuous operation of
the unit. The water reduces the temperature of the desiccant, which improves its latent cooling rate, as well as
reducing the temperature of the air stream, allowing for a degree of sensible cooling. During the dehumidification
process, the latent heat of condensation heats up the desiccant as it flows down the plate, and if uncooled, will
reduce its ability to absorb moisture. Consequently, the unit tested uses integral cooling channels in the dehumidifier
plates to reduce the temperature of the desiccant solution as it runs down the entire length of the plate.
During normal operation, air is blown in-between the plates and across the desiccant flow, (i.e., in a crossflow
configuration). The complete dehumidifier consists of multiple (curved) plates separated by a 2.5 mm air gap. Each
conditioner plate was 2.5 mm thick, 305 mm wide, and 1250 mm from top to bottom. The plates had a thin (0.5 mm)
coating to promote uniform wetting by “wicking” the desiccant solution across the surface of the plates. A sump
pump was used to deliver concentrated desiccant solution to the top of the conditioner and the solution returned to
the sump via gravity flow over the plates.

799
S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
Fig. 2 (a) Internally cooled low-flow parallel plate falling-film conditioner; (b) top-view of a single plate showing the internal channels6
The system used a lithium chloride and water solution as the desiccant. The cooling cycle is driven by the
dilution and re-concentration of the liquid desiccant solution. In addition to the heat of condensation being absorbed
into the desiccant stream (due to the conversion of latent energy into sensible heat); the chemical reaction caused by
the addition of water to the desiccant solution also releases the heat of dilution energy into the desiccant stream.
The conditioner and regenerator both return to and draw from a shared sump located beneath the unit. The
regenerator contains an additional heat exchanger to preheat the incoming desiccant solution from the sump using
the hot desiccant solution leaving the regenerator. The energy used to heat the water used in the regenerator was
supplied from the solar system and supplemented by an auxiliary natural gas boiler. Figure 3 shows a simplified
schematic of the system.
Fig. 3. Simplified system schematic8
The system was operated for four days during July and August 2014 at the Kingston, Ontario test site. During
this period, the temperature and humidity of the ambient air was monitored to identify the properties of the inlet air
streams for the conditioner and regenerator. The properties of the air exiting the regenerator and conditioner were
continuously measured with commercial air-temperature and humidity sensors. The air flow rate through the
conditioner was recorded with pitot-tube array positioned in a long uniform-cross-section air duct. The inlet and
outlet temperatures of the cooling and heating water were monitored as well as their flow rates. Desiccant
temperatures and flow rates were recorded throughout the experiment. Further details on the operation and
instrumentation of this system can be found in the original work on this system by Jones7. The data system scanned
the instruments every five seconds and recorded the averaged data every minute. Samples of desiccant solution were
regularly tested by a density meter. Correlations from Conde9 were then used to convert these density values to mass
fractions. A separate data logging system monitored the performance of the solar array as described by Crofoot et al.

800 S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
2. The estimated accuracy of the measurement instrumentation is shown in Table 1 based on the manufacturer’s
specifications. The accuracy data was used to estimate the relative uncertainty in the calculated values based on a
“root mean square” propagation of the component measurement errors. These values were used to set the magnitude
of the error bars shown in the data plots.
Table 1. Estimated measurement accuracy values
Instruments
Accuracy
Cooling Water
inlet and outlet thermistors
±0.3 °C
flow meter
±0.5 LPM
Desiccant Sol ution
inlet and outlet thermistors
±0.3 °C
flow meter
±1.5 % of reading
density meter
±0.0001g/cm3
Process Air
inlet and outlet temperature
±0.2°C
inlet and outlet relative humidity
±2%
flow meter
±3% of reading
3. Performance analysis
The driving force for the mass transfer of moisture between the air stream and the liquid desiccant solution is the
difference between the water vapour-pressure caused by the desiccant solution and the partial-pressure of water
vapour in the process-air stream. The vapour-pressure of the desiccant solution is a function of its concentration and
temperature, whereas the partial-pressure of water-vapour in the air is a function of the air’s moisture content.
In the conditioner, there are three working fluids: moist air, liquid desiccant solution, and cooling water. Heat
transfer occurs between all three fluids, and mass transfer occurs between the air and the liquid desiccant. The mass
and energy balance for the moist air stream are given by Eq. (1) and Eq. (2).
,, , ,, ,a in c in c abs a in c out c
mmm
ZZ

(1)
 
,, , ,, ,, ,, , ,, ,,a in c in c f in c a in c a out c out c f out c a out c abs fg a d
mhhm hhmhQ
ZZ

   
(2)
The mass and energy balance equations for the desiccant stream are given by Eq. (3) and Eq. (4). The energy
balance is only valid if it is assumed that the moisture transferred into the desiccant (ṁabs) is very small compared to
the total desiccant flow rate (ṁd,in, c).
,, , , , , ,, ,
()
d in c in c d out c out c d in c abs out c
mX m X m mX 
(3)
,, ,, , , ,
()
d in c d in c a d d outc d out c abs dil d cw
mh Q mh mh Q

  
(4)
The energy balance on the cooling water stream is given by Eq. (5). A mass balance is not needed for this stream
because the flow rate was effectively fixed. The amount of water lost due to leakage (caused by the freezing
damage) was considered to be is negligible as it was small compared to the total cooling water flow rate.
,,cw cw in d cw cw cw out
mh Q m h


(5)
The mass transfer performance of the dehumidifier was evaluated in terms of the moisture removal rate,
calculated using Eq. (1), and the dehumidification effectiveness (εde) as calculated by Eq. (6). The effectiveness
value was defined to describe the conditioners ability to dehumidify the air [2]. The dehumidifier effectiveness
represents the ratio of experimental humidity ratio change to the theoretical maximum possible change.

801
S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
eqin
outin
de
ZZ
ZZ
H


(6)
The maximum possible change is dictated by the difference between the inlet air humidity ratio (ωin) and the
humidity ratio of air in equilibrium (
Z݁
q) with lithium chloride solution at the inlet cooling-water temperature and
inlet desiccant-stream mass fraction. This is calculated from the standard psychrometric equation shown in Eq. (7)
where peq is determined using correlations by Conde9.
0.622
eq
eq
eq
p
Pp
Z

(7)
4. Results and discussion
The objective of the study was to determine the dehumidification rate and impact of several factors affecting the
dehumidification effectiveness in order to evaluate the conditioner’s performance. The parameters studied were the
difference between the desiccant solution’s inlet and outlet concentration as well as the effects of the inlet desiccant
concentration on the water condensation rate. The effect of the difference in the relative humidity of the process-air
at the inlet and outlet of the dehumidifier was studied and finally, the dehumidification effectiveness was analysed.
The LDAC ran for eight to ten hours each test day. The ambient temperature ranged from 15 °C to 26 °C during
the unit’s operation. Figure 4 shows the variation between inlet and outlet desiccant concentration during the
experiment on July 25th. As expected, this data indicated that the system was absorbing moisture into the desiccant.
In addition, the inlet concentration values were maintained at reasonably high levels and this was due to the
regeneration of desiccant solution. The data indicates that the small leaks in the cooling water channels in the
dehumidifier did not adversely affect the operation or capacity of the system.
A transient warm-up period is evident during the first hour of operation and is also reflected in the results, with
the desiccant solution rapidly becoming diluted until the regenerator reaches its operating temperature and capacity.
The data in Fig. 4 shows several outliers at 9:00, 13:00 and 14:00 hours. These outliers are attributed to
sampling error associated with the collection and analysis of the desiccant samples. The desiccant solution was
drawn from the working machine with syringes at thirty minute intervals and transferred to the density meter for
analysis. Bias measurement errors may have occurred if the samples were not well-mixed, or contained any
contaminants such as air bubbles. As these points were outside the bounds of the predicted measurement error a
smooth data line was chosen to represent the test data.
Fig.4. Inlet and outlet desiccant concentration during July 25
Fig. 5. Influence of inlet desiccant concentration on moisture
absorption rate
0.3
0.31
0.32
0.33
0.34
0.35
0.36
0.37
0.38
0.39
0.4
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17: 00 18:00
Mass Fraction (-)
Time
X in
X out
X out adjusted
1
1.25
1.5
1.75
2
2.25
2.5
2.75
3
3.25
3.5
3.75
0.355 0.36 0.365 0.37 0.375 0.38
Moisture Abs. Rate (g/s)
Desiccant Solution Inlet Mass Fraction (-)

802 S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
The graph shown in Fig. 5 illustrates the effect of inlet desiccant concentration on the moisture removal rate. It
can be seen that the dehumidification rate increased with the increasing inlet concentration. This increase was due to
the increase in the magnitude of the difference in the vapour pressure between the air and the desiccant solution.
This was due to the reduced vapour pressure of the desiccant solution at higher concentrations as shown in Fig. 6.
Fig. 6. Vapour pressure of lithium chloride at different concentrations
and temperatures as calculated using the Conde9 correlations
Fig.7. Dehumidifier inlet- and outlet-air relative humidity over the
course of July 25
Figure 7 clearly shows the variation between the relative humidity of the process-air stream before and after the
dehumidification process on July 25th. The drop in relative humidity throughout the day can be attributed to the
increasing ambient temperature. From this figure, it can be seen that the inlet-air relative humidity does not have a
large effect on the outlet air relative humidity. The outlet-air relative humidity was primarily dependant on the
desiccant concentration, and the cooling water temperature (as would be expected according to Fig. 6). An important
note regarding this particular installation is that the cooling water temperature was closely linked to the ambient-air
temperature and humidity due to the use of an evaporative cooling tower to reject heat. High humidity levels should
increase the dehumidification rate in the conditioner, but higher ambient relative humidity will lower the capacity of
the cooling tower, leading to higher cooling water temperatures and reduced system capacity.
Figure 8 shows the results from August 1st during which the ambient air’s relative humidity changed from a high
to low value over the course of the test period. The spikes evident in the data records can be attributed the different
firing stages of the boiler as it cycled between low/high/off while attempting to maintain the unit’s set-point. Also,
the opening of the unit’s access doors for brief periods to obtain desiccant samples shut down the unit’s fans briefly
(for safety reasons) and this caused some fluctuations. The overall effect of these incursions was considered to be
negligible.
Fig.8. Dehumidification effectiveness of system over August 1st shown with the inlet and outlet dehumidifier outlet air relative humidity
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Vapour Partial Pressure (kPa)
Temperature (oC)
36%
38%
40%
42%
10
12
14
16
18
20
22
24
26
20
30
40
50
60
70
80
90
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16: 00 17:00 18:00
Ambient Temperature (oC)
Relative Humid ity (%)
Time
RH IN
RH OUT
T amb.
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
35
40
45
50
55
60
65
70
75
80
85
90
95
100
08:00 09:00 10:00 11:00 12:00 13:00 14:00 15: 00 16:00 17:00 18:00
Dehumidif ication E ffectivnes s
Relative Humid ity (%)
Time
RH In
RH out
Dehumid. Eff.

803
S. Bouzenada et al. / Procedia Computer Science 52 ( 2015 ) 796 – 803
The large change in inlet-air relative humidity was only marginally reflected by a slight increase in outlet-air
relative humidity. This caused an increase in the dehumidification effectiveness. This is an interesting result, typical
of LDAC systems. In locations with high latent cooling loads, a LDAC system such as the one studied, would be a
suitable choice to maximize HVAC system efficiency. However, the lack of significant sensible cooling capacity
would make a desiccant system a poor choice in areas with low humidity and high temperatures.
Across all experimental days, the average value of conditioner effectiveness reached 0.55. A previous study2 on
the unit achieved an average value of 0.59. This value does vary depending on the ambient conditions so the
difference between the two is marginal.
5. Conclusion
This paper experimentally studied the performance of a liquid desiccant dehumidifier using aqueous lithium
chloride. Reliable sets of data for air dehumidification were obtained. The experimental results showed the effect of
the difference between the inlet and outlet relative humidity and the inlet and outlet desiccant concentration. It was
found that dehumidification rate increased with increasing desiccant inlet concentration. The inlet concentration of
the desiccant solution was relatively consistent throughout the day, indicating adequate regeneration capacity, even
with the additional dilution of desiccant caused by the minor cooling-water leak into the unit’s sump. The average
value of dehumidifier effectiveness was measured at 0.55, consistent with previous measurements and was shown to
be influenced by the inlet air’s relative humidity.
Acknowledgements
The authors would like to thank Natural Resources Canada (NRCan), the Smart Net-Zero Energy Buildings
Strategic Research Network (SNEBRN), and Natural Sciences and Engineering Research Council of Canada
(NSERC) for their contributions to this research. This work was completed in support of Canada’s contribution to
the International Energy Agency’s (IEA) Task 48.
References
1. Mesquita L, Harrison S, and Thomey D. Modelling of heat and mass transfer in parallel plate liquid desiccant dehumidifiers. Solar Energy.
2006; 80(11):1475-1482.
2. Crofoot L, McNevin C, and Harrison S. Performance evaluation of a liquid -desiccant solar air-conditioning system. Proceedings of ISES
EuroSun International Conference on Solar Energy and Buildings. Aix-les-Bains, France; 2014.
3. Löf G. Cooling with solar energy. Congress on Solar Energy. Tucson, AZ; 1955; 171-189.
4. Jain S, and Bansal P. Performance analysis of liquid desiccant dehumidification systems. International Journal of Refrigeration. 2007; 30:
861-872.
5. Sreelal B, and Hariharan R. The effect of air velocity in liquid desiccant dehumidifier based on two phase flow model using computational
method. International Journal of Emerging Engineering Research and Technology. 2014; 2(7): 142-152.
6. Lowenstein A, Slayzak S, and Kozubal E. A zero carryover liquid-desiccant air conditioner for solar applications. Proceedings of ISEC 2006
ASME international solar energy conference, Denver, USA; 2006.
7. Jones M. Field evaluation and analysis of a liquid desiccant air handling system, M.A.Sc. Thesis, Queen's University, Kingston, ON; 2008.
8. McNevin C, and Stephen J. Harrison, Performance improvements on a solar thermally driven liquid desiccant air-conditioner. Canadian
Society of Mechanical Engineers International Congress. Toronto, Canada; 2014.
9. Conde M. Aqueous solutions of lithium and calcium chlorides: property formulations for use in air conditioning equipment design.
International Journal of Thermal Sciences. 2004; 43(4): 367-382.