international journal of refrigeration 31 (2008) 857–866

available at www.sciencedirect.com

w w w . i i fi i r . o r g

journal homepage: www.elsevier.com/locate/ijrefrig

Experimental study on a new internally cooled/heated

dehumidifier/regenerator of liquid desiccant systems

Yonggao Yin*, Xiaosong Zhang, Geng Wang, Lei Luo

School of Energy and Environment, Southeast University, 2# SiPaiLou Road, Nanjing, Jiangsu 210096, China

article info abstract

Article history: For providing good performance of dehumidifier and regenerator with certain dimensions,
Received 11 June 2007 a new type of internally cooled/heated dehumidifier/regenerator based on the plate–fin
Received in revised form heat exchanger (PFHE) was designed. To investigate the behavior of the new equipment,
10 October 2007 an experimental setup was established in an environment chamber with regulable temper-
Accepted 11 October 2007 ature and humidity air. By the internally cooled dehumidification testing, effects of the
Published online 22 October 2007 cooling water temperature, the air flow rate and the desiccant temperature on the dehu-
midification performance and the cooling efficiency were presented. The behavior of
Keywords: internally cooled dehumidification process was compared with that of the adiabatic dehu-
Air conditioning midification process. The results suggested that the cooling efficiency decreased with the
Dehumidifier increasing of the cooling water temperature and desiccant with low temperature could
Desiccant bring more mass transfer coefficients. There is an optimal air flow rate to achieve the max-
Design imum absolute humidity decrease of the air. By the internally heated regeneration testing,
Heat exchanger effects of the air flow rate and the desiccant inlet temperature on the regeneration perfor-
Regenerator mance and air outlet parameters were discussed and also compared with those of the
Finned tube adiabatic regeneration process. It was concluded that the regeneration efficiency of
Experiment internally heated regeneration was more than that of the adiabatic regeneration, and the
internally heated regenerator could offer better thermal performance.
ª 2007 Elsevier Ltd and IIR. All rights reserved.

Nouveau déshumidificateur/régénérateur refroidi/chauffé de

façon interne, utilisé pour les systèmes à déshydratant
liquide: étude expérimentale
Mots clés : Conditionnement d’air ; Déshumidificateur ; Déshydratant ; Conception ; Échangeur de chaleur ; Régénérateur ; Tube aileté ;
Expérimentation

* Corresponding author. Tel./fax: þ86 25 83792722.

E-mail address: [email protected] (Y. Yin).
0140-7007/$ – see front matter ª 2007 Elsevier Ltd and IIR. All rights reserved.
doi:10.1016/j.ijrefrig.2007.10.004
858 international journal of refrigeration 31 (2008) 857–866

Nomenclature Greek letters

h regeneration efficiency
A mass transfer area (m2)
e cooling efficiency Subscripts
Ga flow rate of air (m3 s1) a air
Gs flow rate of solution (kg s1) c cooling
hD mass transfer coefficient (m s1) equ equilibrium
m vapor removal rate (kg s1) deh dehumidification
Qh heating power (kW) in inlet
r vaporization heat (kJ kg1) out outlet
T temperature ( C) reg regeneration
W humidity ratio of the air (kg kg1) s desiccant solution
X mass concentration (%) w cooling water

1. Introduction and suggested that the improper wetness had important ef-
fect on the performance of dehumidifier and regenerator
Liquid desiccant systems show great energy-saving potential and gave the wetness factor. Saman and Alizadeh (2002) pre-
so that much research on them has been conducted recently sented the experimental results using cross-flow type plate
(Factor and Grossman, 1980; Gandhidsan, 2005). Liquid desic- heat exchanger (PHE) as a dehumidifier/cooler. Many flow pas-
cant systems can be driven by low-grade heat (about 70– sages were separated by thin plastic plates, and one side of
80  C), such as solar energy, industrial waste heat (Pohl each thin plastic plate was for the desiccant solution/air pas-
et al., 1998), and have no pollution to the environment. In ad- sage; the other side is for water/air passage. So the indirect
dition, hybrid liquid desiccant systems combining with tradi- evaporation process in the water/air passage provided cooling
tional vapor compression chilling units can bring much better to the dehumidification process. The experimental results
thermal performance and obviously increase the coefficient of indicated the effects of inlet parameters of air and solution
performance (COP) of the refrigeration systems. Dehumidifi- on the PHE effectiveness. Yin et al. (2006) presented a model
cation and regeneration processes are very important to liquid of internally cooled and adiabatic dehumidifier and carried
desiccant systems, and determine the thermal performance out the parameters’ comparative study on the two types of
of the systems. In earlier experimental studies on liquid desic- dehumidification processes.
cant systems, most of them (Fumo and Goswami, 2002; However, none of these studies were carried out on the in-
Gommed and Grossman, 2004; Liu et al., 2006) focused on ternally heated regeneration processes, and generally experi-
packed bed dehumidifiers and regenerators which could pro- mental studies on the internally cooled dehumidification
vide much contacting area for air and liquid desiccant, but processes were very limited and more investigation and rela-
the heat and mass transfer processes happening in them tive data support are required for developing liquid desiccant
were adiabatic and with the progress of dehumidification/ systems. This study presented an experimental apparatus
regeneration process the heat and mass transfer gradients which could be used as an internally cooled dehumidifier
would decrease greatly. Yin et al. (2007) conducted parame- and also as an internally heated regenerator depending on
ters’ analysis on packed bed dehumidifier and regenerator ex- the operation conditions, and parameters’ investigation on
perimentally and the results indicated that with the process the internally cooled dehumidifier processes and internally
progress the bottom of the regenerator showed very wick re- heated regeneration processes was carried out.
generative performance because the desiccant temperature
at the bottom of the regenerator was very low. But internally
cooled dehumidifiers and internally heated regenerators 2. Experimental apparatus and
could provide more heat and mass transfer gradients and instrumentation
have the possibility to make dehumidifiers and regenerators
miniaturization. Khan and Martinez (1998) studied numeri- A plate–fin heat exchanger (PFHE) made of stainless steel is
cally the performance of an internally cooled counter flow ab- proposed to use for internally cooled/heated element of the
sorber using thin plate heat and mass exchanger cooled by dehumidifier/regenerator. The schematic diagram of the
direct evaporation with lithium chloride as desiccant. En- PFHE is shown in Fig. 1. There are seven water passages and
thalpy and humidity effectiveness were brought out to define six desiccant solution/air passages in every PFHE as shown
the thermal performance of the absorber. It was found that in Fig. 1(b). Each solution passage is 298 mm in length and
the number of mass transfer units had great effect on the en- 12 mm in width. The cooling/heating water can enter into
thalpy and humidity effectiveness. Khan (1998) investigated and leave the PFHE at the same side of the PFHE. Between
numerically on an internally cooled dehumidifier using tube– two neighboring plates, there are three layers of fin side by
fin exchanger with the air crossing flow. Jain et al. (2000) inves- side intercrossing, which can provide more contacting area
tigated a liquid desiccant system which was made up of a for desiccant solution and air. The distance between fins is
packed bed regenerator and an internally cooled dehumidifier 4 mm. The dimensions of the PFHE are shown in Fig. 1(a)
 international journal of refrigeration 31 (2008) 857–866 859

(a) 298
(c) Solution In

312
Air Out

Water Solution and Water

in air channel out
Top view of the plate-fin heat exchanger

(b) Solution

Water Out
100

Water In
Water Water
In Out Air In

Air
Schematic diagram of the plate-fin heat exchanger Solution Out

Fig. 1 – A new type of internally cooled/heated dehumidifier/regenerator.

and (b). The height of the PFHE is 100 mm. The desiccant solu- be heated by the hot water during the regeneration process
tion is sprayed to the fins between the plates from the upside to keep higher temperature. If the hot water is inactivated,
and flows down by the gravity, at the same time the air is the regenerator would be the same as the packed bed, too.
blown from the underside counter-flowing with the desiccant In a word, the equipment shown in Fig. 1(c) would be used
solution. as an internally cooled dehumidifier when cooling water en-
The internally cooled/heated dehumidifier/regenerator is ters into it, and also used as an internally heated regenerator
made up of six PFHEs stacking up along the vertical direction when the heating water substitutes the cooling water.
as shown in Fig. 1(c). For the convenience of installing these Fig. 2 shows the schematic diagram of the experimental
PFHEs the direction of water inlet changes every other PFHE, setup for testing the internally cooled/heated dehumidifier/
and therefore three of them have their water inlet/outlet on regenerator. The experimental system is made up of two
one side and remnant three PFHEs have their water inlet/out- pumps, two tanks, a water tank with regulable temperature,
let on the other side. The desiccant solution enters into the in- flow meters, valves, an internally cooled/heated dehumidi-
ternally cooled/heated dehumidifier/regenerator from the top fier/regenerator, a heater, a cooler, PPR pipe and some control-
and contacts with the countercurrent air blown from the bot- ling equipments, etc. Two liquid desiccant tanks (Tank 1 and
tom. The solution would be dispersed to the fins by a distribut- Tank 2) are cylinders with the same dimensions, 70 cm in
ing unit between the plates due to the gravity. The cooling/ height and 55 cm in diameter. Before starting the experi-
heating water in the water passages flows horizontally and ments, valves 5, 6, 7 used for drainage were closed and liquid
carries out heat transfer with the crossed air/solution. desiccant solution with the weight concentration of about
As known, during the desiccant solution dehumidification, 39% was confected by 59 kg of granular lithium chloride dis-
because of the water vapor partial pressure difference the solved into 92 kg pure water and poured into Tank 1. After
vapor in the air would become liquid water to enter into the closing valves 1, 4 and opening the valves 2, 3, dehumidification
solution, and release the vaporization heat which results in experiments started, and the liquid desiccant solution was
increasing temperature of the solution and air. The cooling transported from the Tank 1 to the dehumidifier by Pump 1
water entering into the dehumidifier would bring away the and collected in the Tank 2. Until the Tank 1 was emptied
heat to restrain the increase of the solution temperature. So out, the liquid desiccant solution was pumped again via the
the internally cooled dehumidifier could hold the low temper- dehumidifier from the Tank 2 to the Tank 1 by opening the valves
ature of the solution to show good dehumidifying perfor- 1, 4 and closing the valves 2, 3, and so repeated. The inlet tem-
mance. If the cooling water is inactivated, the dehumidifier perature of the solution entering into the dehumidifier was
would be an adiabatic one, just the same as the packed bed. controlled by the cooler and the electric heater which was con-
During the desiccant solution regeneration, the water in the trolled by a temperature controller. Temperature of the water
desiccant solution would become vapor to join the air. The in the water tank was controlled by a chilling unit and an elec-
process requires absorbing vaporizing heat from the solution tric heater to achieve required water state. The flow rate of the
and air, and so the temperatures of the solution and air would cooling/heating water was conditioned by the valve 8. The flow
reduce. The solution and air with low temperature would rate of the solution was conditioned by the valve 3 or 4.
show weak regenerative ability. If the hot water with high In order to obtain steady-going air state, the experimental
temperature enters into the PFHEs, the solution and air can setup was placed in a psychrometric room as shown in
860 international journal of refrigeration 31 (2008) 857–866

T1
Dehumidifier/Regenerator

Exhaust air
Flowmeter

T4 Ts2
Pump

Valve

T3 Ts1
Heater

Air

Tw2
Tw1

T2
8 Water tank
Cooler

Tank 1 2 Tank Chilling

or
1 2
3 4 P2 heating
P1

Valve
5 6 7
Drainage

Fig. 2 – Schematic diagram of the experimental setup.

Fig. 3, which was often used to achieve constant temperature recorded by Tw1 and Tw2. The mass concentration of the
and absolute humidity of the air for testing the performance of aqueous lithium chloride was indirectly tested by measuring
room air-conditioners. The dry-bulb and wet-bulb tempera- the density and temperature of the solution, and the proper-
tures of the air in the chamber should be conditioned to the re- ties of the aqueous lithium chloride could reference the liter-
quired values and be in a stable atmospheric condition before ature (Conde, 2004). According to the density and temperature
dehumidification/regeneration experiment. Usually the pro- of the solution, its weight concentration could be calculated.
cess needed a long time. The environmental chamber pro- The solution density was measured by three different specific
vided the measuring instruments for the flow rate, dry-bulb gravity hydrometers with the accuracy of 1 kg m3, whose
temperature and wet-bulb temperature of the air, respec- measure ranges were from 1000 to 1100 kg m3, from 1100 to
tively, recorded by the measuring points T3 and Ts1. The 1200 kg m3 and from 1200 to 1300 kg m3. All the experiments
dry-bulb and wet-bulb temperatures at outlet point were mea- were carried out under the steady-state of the air and desic-
sured by T4 and Ts2. Temperatures of the desiccant solution at cant and every experiment would last about 15–20 min.
inlet point and outlet point were measured by T1 and T2. The In order to study liquid desiccant solution dehumidifica-
flow rate of the air was measured by three standard nozzles tion processes, two sets of experiments were conducted, re-
with the accuracy of 1%, whose diameters were 70, 80 and spectively, under water cooling and no water cooling, which
110 mm, respectively, corresponding measure ranges from intended for comparing the difference between internally
208 to 485 m3 h1, from 424 to 990 m3 h1 and from 513 to cooled dehumidifier and adiabatic one, and finding out the
1197 m3 h1. The flow rate of the air was conditioned by effects of inlet parameters on dehumidification processes.
a transducer controlling the rotation speed of the fan. The Similarly, for the same purpose to study the desiccant solu-
temperatures of the air were measured by some Pt100 RTDs tion regeneration processes, two sets of experiments were
(resistance temperature detectors) with the accuracy of conducted under water heating and no water heating,
0.1  C and recorded by the Agilent 34970A. Temperatures of respectively.
the solution and water were measured by some thermocou-
ples with the accuracy of 0.1  C. The flow rate of the desiccant
solution was measured by an anticorrosive rotameter with the 3. Experimental results and discussions
measure range from 40 to 400 l h1 with the accuracy of
10 l h1 and it was the same as the measure of the water For the internally cooled desiccant solution dehumidification
flow rate. Temperatures of the water at inlet and outlet were processes, the flow rate of the air and the temperature of the
 international journal of refrigeration 31 (2008) 857–866 861

Fan

Humidifier
Air

Evaporator
Compressor Dehumidifier

Nozzle Air Sampler

Heater

Environmental Chamber

To condenser From throttle

Fig. 3 – Environmental chamber with constant temperature and humidity.

cooling water have great effect on the processes. For the Here, A is the total mass transfer area between the solution
internally heated desiccant solution regeneration processes, and processed air; hD is the mass transfer coefficient; Ws,equ is
the flow rate of the air and the temperature of the solution the absolute humidity of surface air in equilibrium with the
have important effect on the regeneration processes. So fol- solution at its temperature and concentration. So under the
lowing experiments were intended to find out the effects of experimental conditions (constants Win, Ga and A), the mass
these parameters on the dehumidification and regeneration transfer coefficient entirely depends on the parameters DW
processes. and Ws,equ.
From Fig. 4, it was concluded that the absolute humidity
3.1. Dehumidification process test difference (DW ) of the processed air decreased with increase
of the desiccant temperature. The reason for this behavior is
3.1.1. Internally cooled dehumidification process with explained by the fact that desiccant with lower temperature
different desiccant temperatures can bring more mass transfer coefficient and lower vapor par-
In this set of experiment, the internally cooled and adiabatic tial pressure, which would result in more moisture removal,
dehumidifier made of the PFHE was, respectively, tested under
two groups of experimental conditions as shown in Table 1. 3.5
The experimental results are displayed in Fig. 4. Adiabatic-1
In order to declaim effects of the desiccant temperature on Internally cooled-1
Adiabatic-2
the mass transfer coefficient, the mass transfer coefficient be- 3 Internally cooled-2
tween the processed air and solution may be defined as fol-
lows according to the literature (Saman and Alizadeh, 2002):
W/g kg-1

Ga Win  Wout 2.5

hD ¼ (1)
A Win  Ws;equ

Table 1 – Dehumidification experimental conditions with

different temperatures of desiccant
1.5
Group Ta W Ga Gs Xs Gw Tw 20 22 24 26 28 30 32
( C) (g kg1) (m3 s1) (kg s1) (%) (kg s1) ( C) Ts / ºC
1 30.5 13.4 0.061 0.1036 37.7 0.151 22
Fig. 4 – Effect of desiccant temperature on the air absolute
2 30.9 12.6 0.061 0.1036 38.8 0.151 22
humidity change.
862 international journal of refrigeration 31 (2008) 857–866

5.2 the other was that the cooling water was not activated, which
Adiabatic-1 was adiabatic one.
Adiabatic-2 Fig. 6 displayed contrastively the effects of the processed
4.8
air flow rate (Ga) on the absolute humidity change (DW ) under
the internally cooled and adiabatic dehumidification pro-
hD × 10-3/m.s-1

4.4
cesses. The figure showed that the absolute humidity differ-
ence (DW ) between the inlet and outlet air increased with
4.0
the increase of flow rate of the processed air in both internally
cooled dehumidification and adiabatic dehumidification, the
3.6 reason for which was that the mass transfer coefficient in-
creased greatly by increasing the processed air flow rate. At
3.2 low air flow rate, the DW increased rapidly and at high flow
rate the DW nearly remained constant, which was explained
20 22 24 26 28 30
that at low flow rate of the air the rate of increase of the
Ts / ºC
mass transfer coefficient was more than the rate of increase
of the air flow rate, but at high flow rate of the air both rates
Fig. 5 – Effect of desiccant temperature on mass transfer
of increase were balanceable. As described by Saman and
coefficient.
Alizadeh (2002), if the air flow rate continued to increase to
exceed the increase rate of the mass transfer coefficient, the
value of DW would be decreased. So the PFHE dehumidifier
as shown in Fig. 5, the mass transfer coefficient decreased could export better dehumidification performance at the air
with increase of the solution temperature under both experi- flow rate of about 0.07 m s1 during low air flow rate range
mental condition group 1 and group 2, which indicates that under the experimental conditions. Higher rate of moisture
the desiccant solution temperature has important influence removal was achieved in the internally cooled dehumidifica-
on both mass transfer coefficients. From Fig. 5, it was found tion, and the DW which was 0.5 g kg1 more than that of the
that the mass transfer coefficient under experimental condi- adiabatic dehumidification.
tion group 1 was more than group 2 although the desiccant so-
lution in group 2 had more concentration than group 1, the 3.1.3. Internally cooled dehumidification process with
absolute humidity of the air in group 1 was more than group different cooling water temperature
2, which indicated that the air with higher absolute humidity The objective of the dehumidification is to remove the water
could bring more mass transfer coefficient. In addition, under vapor from the processed air by the thick solution. So the
the same experimental conditions the internally cooled dehu- amount of removed vapor from the air can be calculated as
midification process produced lower absolute humidity of the follows:
processed outlet air, and from Fig. 4 it was observed that with
the increase of inlet desiccant temperature, the rate of de- mdeh ¼ Ga ðWin  Wout Þ (2)
crease of DW in the adiabatic dehumidification was more
In the earlier studies on the traditional packed bed dehu-
than that in the internally cooled dehumidification. So both
midifiers, the values of Ga/Gs were slightly more. (Factor and
low inlet desiccant temperature and simultaneous cooling
Grossman, 1980, Ga/Gs ¼ 1.25–2; Fumo and Goswami, 2002,
during the dehumidification can provide better dehumidifica-
Ga/Gs ¼ 1.3–3.3; Saman and Alizadeh, 2002, Ga/Gs ¼ 3.2–9.4.) Al-
tion performance.
though the removed water vapor (mdeh) can reflect the perfor-
mance of the dehumidifier, for liquid desiccant evaporative
3.1.2. Internally cooled dehumidification process
with different flow rate of processed air
In the environment chamber shown in Fig. 3, the flow rate of
3
the processed air is easily conditioned and controlled by
regulating the rotate speed of the fan. In this set of experiment
the flow rate of the processed air was at the lowest rate of
0.042 m3 s1 and up to the maximum of 0.075 m3 s1, and the 2.5
experimental conditions are set in Table 2. Two dehumidifica-
W/g kg-1

tion models were investigated: one was that the cooling water
was activated, which was internally cooled dehumidification;
2

Internally cooled
Table 2 – Dehumidification experimental conditions with
Adiabatic
different flow rate of processed air 1.5
0.04 0.05 0.06 0.07 0.08
Ta W Gw Tw Gs Ts Xs Ga/m3/s
( C) (g kg1) (kg s1) ( C) (kg s1) ( C) (%)
Fig. 6 – Effect of the processed air flow rate on the absolute
30.9 10.6 0.151 24 0.1036 25.3 38.4
humidity change.
 international journal of refrigeration 31 (2008) 857–866 863

cooling systems the supply air status parameters (including 26 0.8

humidity and temperature of the processed air) are required
to meet; that is to say, absolute humidity change of the pro- 25 0.7
cessed air (Win  Wout) is more important to liquid desiccant
evaporation cooling system rather than the amount of 24 0.6
removed vapor (mdeh). Therefore in order to provide required

Tout / ºC

ec
absolute humidity of the air leaving dehumidifier, usually 23 0.5
during the dehumidification low flux ratio of processed air
to desiccant solution Ga/Gs is preferred. So here the value of 22 0.4
Ga/Gs was chosen as 0.8. During the internally cooled dehumid-
ification, the cooling efficiency of the desiccant solution is 21 0.3
defined as Ts Tw ec
20 0.2
19 20 21 21 22 22 23 23 24 25
Ts;out  Ts;in
ec ¼ (3) Tw / ºC
Ts;out  Tw;in

Fig. 7 – Effects of Tw on the cooling efficiency, outlet

Under the operation conditions shown in Table 3, the dehu-
temperature of solution and cooling water.
midification experiments were conducted with different
temperatures of the cooling water from 19 to 25  C. Every
experiment would last 8 min after reaching the steady-state.
partial pressure and provided stronger mass transfer force
Fig. 7 demonstrated effects of the cooling water tempera-
gradient; the other was that there was more mass transfer co-
ture Tw on the outlet temperatures of desiccant solution and
efficient between the air and solution with lower temperature,
air and the cooling efficiency. The figure showed that the out-
as described in Section 3.1.1. The air absolute humidity differ-
let temperature of the desiccant solution increased with the
ence (DW ) between the inlet and outlet was about 2.8 g kg1
increase of the temperature of cooling water, and the outlet
under the cooling water temperature of 19.4  C and about
temperature of the solution was very close to the outlet tem-
2.5 g kg1 under the cooling water temperature of 24.5  C,
perature of the cooling water. It was easily concluded that
which means that the outlet absolute humidity of the air
the outlet temperature difference between the solution and
was about 10 g kg1 and was remarkably higher than the
cooling water was less than 1  C, which indicated that the
saturation air absolute humidity at equilibrium with the des-
PFHE had good performance in cooling the solution. But the
iccant solution about 4 g kg1. It may be resulted from not pro-
cooling efficiency went down with the increase of the cooling
viding enough contacting area of the desiccant and air by the
water temperature, which was very different from the only
dehumidifier. Therefore, in order to obtain good dehumidifica-
heat transfer process in the PFHE. Because of the dehumidifi-
tion performance, two very important factors have to be taken
cation in the solution side, the removed vaporization heat
into account: one is to provide more contacted area between
from the air caused by the mass transfer made the solution
solution and air as possibly, and the other is to keep low tem-
temperature obviously increasing, and according to Eq. (3),
perature of the solution and air.
the cooling efficiency decreased. In addition, from the figure
it was concluded that the internally cooled dehumidification
process was very different from the one happening in packed 3.2. Regeneration process test
bed dehumidifiers (adiabatic dehumidification). Usually in
adiabatic dehumidification processes the temperature of the 3.2.1. Internally cooled regeneration process with different
solution or the processed air increased rapidly, here the inlet flow rate of air
temperature of the solution was 27  C in the internally cooled This set of experimental conditions is shown in Table 4. The
dehumidification process but the outlet temperature of it was desiccant solution was heated to is a high temperature for the
less than the inlet value, and changed from 22 to 25.8  C under regeneration to enter into the PFHE. The ratio of flow rate of
different inlet temperatures of the cooling water; moreover
from Fig. 8 the outlet air temperature increased slightly, about 30.8 2.85
0.4  C higher than the inlet value. 2.8
30.7
Fig. 8 also indicated that the removed vapor from the air
2.75
decreased with the increase of the cooling water temperature, 30.6
W/g kg-1

which resulted from two main reasons: one was that the des- 2.7
Ta / ºC

30.5
iccant solution with low temperature had low surface vapor 2.65
30.4
2.6
30.3 2.55
Table 3 – Dehumidification experimental conditions with
different temperatures of water 30.2 2.5
20 21 22 23 24
Ta W Ga Gs Ts Xs Gw Tw / ºC
( C) (g kg1) (m3 s1) (kg s1) ( C) (%) (kg s1)
Fig. 8 – Effects of Tw on the absolute humidity change (DW )
30.3 12.6  0.05 0.072 0.1036 27.0 38.65 0.151
of the processed air.
864 international journal of refrigeration 31 (2008) 857–866

14
Table 4 – Regeneration experimental conditions with
different flow rate of processed air
(a)
12
Ta W Ts Gw Tw Gs Xs
( C) (g kg1) ( C) (kg s1) ( C) (kg s1) (%)

W/g kg-1
10
26.5 11.5 70 0.14 70 0.1 38

processed air to desiccant solution Ga/Gs was small in earlier 6 Internally heated
Adiabatic
studies on the regeneration using packing towers. (Martin
and Goswami, 1999, Ga/Gs ¼ 0.077–0.33; Gandhidsan, 2005, 4
0.03 0.04 0.05 0.06 0.07 0.08
Ga/Gs ¼ 0.1–0.32.) Here the Ga/Gs changed in a wide range
Ga /m3/s
from 0.36 to 0.96. It was very similar to the dehumidification
experiment that the regenerator operated as an internally 42
heated regenerator if the heating water was activated, other-
wise as an adiabatic one.
(b)
Fig. 9 displayed contrastively the effects of the air flow rate
on the outlet parameters of the air of the internally heated and 40

Ta,out / ºC
adiabatic regeneration processes. The purpose of regeneration
processes is to shift the water of desiccant solution to the air.
So the absolute humidity change of the air (DW ) can suggest
38
the amount of removed water from the desiccant solution.
Fig. 9(a) showed the effects of the air flow rate on its absolute Internally heated
humidity change of internally heated and adiabatic regenera- Adiabatic
tion. The DW increased with the increase of flow rate of air 36
0.03 0.04 0.05 0.06 0.07 0.08
both in the internally heated regeneration and in the adiabatic
Ga /m3/s
regeneration. The reason is that more air flow rate can bring
more mass transfer coefficient during the low air flow rate 50
range. Fig. 9(b) showed that the outlet temperature of the air (c)
increased slightly with the flow rate of the air and there was
very little difference between the outlet air temperature of 45
the internally heated regeneration and that of the adiabatic
ha / kJ kg-1

regeneration. The temperature of the air increased from about

40
26.5 to 39  C during the two regeneration processes when the
air flow rate was 0.06 m3 s1. Here the regeneration efficiency
is defined as follows: 35
Internally heated
DWr Adiabatic
hreg ¼ (4)
Qh
30
0.03 0.04 0.05 0.06 0.07 0.08
If in the adiabatic regeneration the heating power is Q1, the
Ga /m3/s
absolute humidity change of the air is DW, the regeneration
efficiency h1 may be calculated as
Fig. 9 – Effect of the air flow rate on the (a) absolute
DW1 r humidity, (b) temperature, and (c) enthalpy of the air.
h1 ¼ <1 (5)
Q1

Considering that the outlet temperatures of air in both re-

generations were of no difference, the more heating power In other words, the regeneration efficiency of internally
consumed by the internally heated regeneration was divided heated regeneration was more than that of the adiabatic re-
into two parts according to the energy balance; one part generation and the internally heated regenerator could offer
(DQ2) was to provide for more water vaporization from the better thermal performance.
desiccant solution, and the other (DQ3) was to provide for Fig. 9(c) demonstrated that the enthalpy change of the air
higher outlet temperature of desiccant solution. Suppose increased with the increase of the air flow rate. And the outlet
that the heat DQ3 could be recovered by the desiccant solution, enthalpy of the air of the internally regeneration was dis-
the regeneration efficiency of the internally heated regenera- tinctly about 8 kJ kg1 higher than the adiabatic one.
tion could be expressed as
DW1 r þ DQ2 3.2.2. Internally cooled regeneration process with different
h2 ¼ (6)
Q1 þ DQ2 desiccant temperatures
This set of experiments was conducted under the conditions
Combining Eqs. (5) and (6), it was easy to conclude:
shown in Table 5. The inlet temperature of desiccant solution
h1 < h2 (7) varied from 62 to 73  C by controlling the heating power of the
 international journal of refrigeration 31 (2008) 857–866 865

transfer process in the PFHE. During the low air flow rate
Table 5 – Regeneration experimental conditions with
different desiccant temperatures range, more air flow rate can achieve better dehumidifica-
tion performance.
Ta W Ga Gw Tw Gs Xs
(2) By calculating the mass transfer coefficient between the
( C) (g kg1) (m3 s1) (kg s1) ( C) (kg s1) (%)
air and solution, it is found that the desiccant with low
26.7 11.3 0.052 0.14 72 0.1 36 temperature can bring more mass transfer coefficient.
Therefore, both desiccant with low inlet temperature and
simultaneous internally cooling during the dehumidifica-
electric heater. Fig. 10 indicated the effects of inlet tempera- tion can offer better dehumidification performance.
ture of the desiccant on the air absolute humidity change. (3) From the regeneration experiment tests, it was pointed
From the figure, it was seen that the air absolute humidity out that the increase of the air flow rate could bring
change DW increased rapidly with the increase of inlet tem- better regeneration performance, and the air absolute hu-
perature of the desiccant solution, and during the inlet midity change increased rapidly with the increase of inlet
temperature of desiccant varying from 65 to 69  C the air abso- temperature of the desiccant solution. Contrastively
lute humidity change of the internally heated regeneration studying the internally heated and adiabatic regeneration
was obviously higher than that of the adiabatic regeneration. processes, it was suggested that there was very little differ-
Also the data suggested that the inlet temperature of the des- ence between their outlet air temperatures which deduced
iccant solution had very important effect on the regeneration that the regeneration efficiency of the internally heated
performance. regeneration was more than that of the adiabatic
regeneration.

4. Conclusions

A new type of internally cooled/heated dehumidifier/regen-

Acknowledgments
erator based on the PFHE unit was constructed and experi-
This research was supported by the grants from the fund of
mental study on its operation behavior under different
National Natural Science Foundation of China under the
conditions was carried out. According to the cooling/heating
contracts No. 50676018, the key grant project of Chinese
water activated or inactivated, the new equipment operated
Ministry of Education (307013), Foundation for Excellent
as the internally cooled/heated or adiabatic dehumidifier/
Doctoral Dissertation of Southeast University and the grant
regenerator. The study in the paper led to the following
of China Scholarship Council. These supports are gratefully
conclusions:
acknowledged.
(1) The dehumidification performance could be promoted by
increasing the air flow rate or decreasing the cooling water references
temperature. The outlet temperature of the desiccant solu-
tion increased with the increase of the temperature of
cooling water, and the outlet temperature of the solution Conde, M.R., 2004. Properties of aqueous solutions of lithium and
was very close to the outlet temperature of the cooling wa- calcium chlorides: formulations for use in air conditioning
ter, which indicated that the PFHE had good cooling perfor- equipment design. International Journal of Thermal Science
mance for the desiccant. During the internally cooled 43, 367–382.
dehumidification process, the cooling efficiency of desic- Factor, H.M., Grossman, G., 1980. A packed bed dehumidifier/
cant decreased with the increase of the cooling water tem- regenerator for solar air conditioning with liquid desiccant.
Solar Energy 24, 541–550.
perature, which was very different from the only heat
Fumo, N., Goswami, D.Y., 2002. Study of an aqueous lithium
chloride desiccant system: air dehumidification and desiccant
regeneration. Solar energy 72, 351–361.
Internally heated Gandhidsan, P., 2005. Quick performance predictions of liquid
15 desiccant regeneration in a packed bed. Solar Energy 79,
Adiabatic
47–55.
Gommed, K., Grossman, G., 2004. A liquid desiccant system of
W/g kg-1

14 solar cooling and dehumidification. Transaction of ASME:

Journal of Solar Energy Engineering 126, 879–885.
Jain, S., Dhar, P.L., Kaushik, S.C., 2000. Experimental study
on the dehumidifier and regenerator of a liquid desiccant
13 cooling system. Applied Thermal Engineering 20, 253–
267.
Khan, A.Y., 1998. Cooling and dehumidification performance
12 analysis of internally-cooled liquid desiccant absorber.
62 64 66 68 70 72 74 Applied Thermal Engineering 18, 265–281.
Ts / ºC Khan, A.Y., Martinez, J.L., 1998. Modelling and parametric
analysis of heat and mass transfer performance of a hybrid
Fig. 10 – Effects of inlet temperature of the desiccant on the liquid desiccant absorber. Energy Conversion Management 39,
air absolute humidity change. 1095–1112.
866 international journal of refrigeration 31 (2008) 857–866

Liu, X.H., Zhang, Y., Qu, K.Y., et al., 2006. Experimental study on mass Saman, W.Y., Alizadeh, S., 2002. An experimental study on
transfer performances of cross flow dehumidifier using liquid a cross-flow type plate heat exchanger for dehumidification/
desiccant. Energy Conversion and Management 47, 2682–2692. cooling. Solar Energy 73, 59–71.
Martin, V., Goswami, D.Y., 1999. Heat and mass transfer in packed Yin, Yonggao, Zhang, Xiaosong, Lai, Rongbing, 2006. Modeling
bed liquid desiccant regenerators – an experimental and comparative study on two types of dehumidifiers with
investigation. Transactions of the ASME: Journal of Solar liquid desiccant. Journal of Chemical Industry and
Energy Engineering 121, 162–170. Engineering 57, 2828–2833.
Pohl, J.P., Hellman, H.M., Grossman, G., 1998. Investigation and Yin, Yonggao, Zhang, Xiaosong, Chen, Zhenqian, 2007.
comparison of two configurations of a novel open-cycle Experimental study on dehumidifier and regenerator of liquid
absorption chiller. International Journal of Refrigeration 21, desiccant cooling air conditioning system. Building and
142–149. Environment 42, 2505–2511.