Energy Conversion and Management 266 (2022) 115867

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Studying the energy efficiency feasibility of composite superabsorbent

coated heat exchangers in open-cycle heat transformation applications
Vivekh Prabakaran , Duc Thuan Bui , Md. Raisul Islam , Kian Jon Chua *
Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore

A R T I C L E I N F O A B S T R A C T

Keywords: Composite superabsorbent polymer desiccants have emerged as a more suitable choice for open-cycle heat
Desiccant coated heat exchangers transformation applications, namely, heat pumps and dehumidifiers as they record remarkably high sorption
Superabsorbent polymers capacity at high relative humidity conditions. Existing mathematical models that predict the desiccant coated
Heat pumps
heat pump’s (DCHP) performance exhibit up to 60% discrepancy with experimental findings and are not
Dehumidifiers
Solar collectors
applicable for variations in experimental conditions/desiccant materials. A simplified mathematical model has
Photovoltaic thermal been developed in this study to predict DCHP’s performance, which records ± 12% and ± 10% discrepancies in
outlet air specific humidity and temperature, respectively. In conducting a parametric analysis, the DCHP
demonstrated 30–40% improvement in energy efficiency compared to conventional heat pumps. Additionally,
the composite superabsorbent polymer desiccant regenerates at 40–50 ◦ C. Integrating either a solar evacuated
tube (ETC) and photovoltaic thermal (PV/T) system to provide for thermal regeneration marks an extraordinary
potential to achieve significant improvement in energy efficiency. The dynamic performance experiments
revealed that although the PV/T was capable of producing hot water at 40 ◦ C, around 10 ◦ C lower than the ETC,
the moisture removal ability of the desiccant coated heat exchanger (DCHE) remained identical. Further, the PV/
T recorded 11% improved thermal efficiency when compared to the ETC, while providing the electrical power
that is required to operate the auxiliary components of the dehumidifier.

principle remains similar to other solid desiccant dehumidifiers, DCHE is

1. Introduction capable of isothermal dehumidification as opposed to isenthalpic/
adiabatic dehumidification; realizing superior heat transfer efficiency
The air dehumidification process in mechanical vapor compression [6]. Further, sensible heat in the air can also be independently handled
(MVC) chillers is governed by the moisture condensation. The supply air due to the simultaneous cooling process [7].
is first cooled 15–20 ◦ C below its dew point to condense moisture, and an Recent developments in DCHEs are classified into four categories:
additional reheating step is often necessary to supply air at acceptable new desiccant synthesis, performance evaluation, advanced engineering
thermal comfort conditions [1]. This inherent need to overcool and analysis, and energy-related applications (Fig. 1). A comprehensive re­
reheat air consumes significant electricity, contributes to excessive view of the existing development in DCHEs along with the future rec­
emissions, and achieves around 10% second law efficiency [2]. There­ ommendations for research is documented in the literature [8,9].
fore, it is indispensable to decouple the simultaneous handling of sen­ Conventional silica gel, zeolite, and activated carbon-based materials
sible and latent cooling loads, raise the conventional chillers’ evaporator along with their hygroscopic salt-based derivatives exhibit low adsorp­
temperature, and improve air-conditioning process efficiency [3]. tion capacity of around 20–50% of its initial dry mass. Metal-organic
Desiccant coated heat exchanger (DCHE) is a promising technology that frameworks (MOFs) are emerging to be a new class of organic adsor­
facilitates the decoupling of cooling loads and eliminates the need to bents with their equilibrium capacities reaching up to 150–160% of dry
sustain a low chiller temperature [4]. In DCHEs, the desiccant (coated on their dry mass [10]. Yet, most of the MOFs synthesized are only suitable
metallic fins) captures moisture from the supply air due to the vapor for small-scale applications owing to their synthesis challenges [11].
pressure difference between the desiccant and the air. When the desic­ Additionally, the conventional desiccants need high temperature for
cant gets saturated, an appropriate heat source is employed to regen­ regeneration between 60 and 80 ◦ C, have poor coating stability due to
erate it for subsequent operating cycles [5]. Although its operating deliquescence and carryover effects, and reduce operational duration

* Corresponding author.
E-mail address: mpeckje@nus.edu.sg (K. Jon Chua).

https://doi.org/10.1016/j.enconman.2022.115867
Received 23 March 2022; Received in revised form 4 June 2022; Accepted 6 June 2022
Available online 11 June 2022
0196-8904/© 2022 Elsevier Ltd. All rights reserved.
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Nomenclature Xt transverse tube pitch, m

y derived parameters
A area, m2
cp specific heat capacity, kJkg-1K− 1 Greek Symbols
d diameter, m α ratio of heat transfer area on the air side to the volume of
dc fin-collar outside diameter, m the heat exchanger
D diffusivity, m2/s η efficiency
Dh hydraulic diameter, m φ relative humidity
f mathematical relationship between measured and derived ρ density, kg/m3
parameters σ ratio of minimum free flow area to the frontal area
h air enthalpy, kJ/kg ω humidity ratio, g/kg
hm mass transfer coefficient, m/s Subscripts
ht heat transfer coefficient, kWm-2K− 1 a air
Hd desiccant layer thickness, m atm atmosphere
Hf fin thickness, m cw cooling water
i specific enthalpy, kJ/kg d desiccant/dehumidification
j Colburn’s j-factor da between desiccant and air
k kinetic constant, 1/s e equilibrium
Lx length of the fin along the air flow direction, m el electrical
Ly width of the fin, m f saturated liquid
Lz height of the heat exchanger, m fr frontal
ṁ mass flow rate, kg/s g saturated vapor
n number of parameters, - hw hot water
Nf number of fins, - in inlet
Nr number of tube rows, - out outlet
Nt total number of tubes, - r refrigerant/regeneration
Nu Nusselts number rd between refrigerant and desiccant
P pressure, kPa sat saturated
Pf fin pitch, mm or m sol solar
Pr Prandtl number th thermal
q instantaneous uptake capacity of the desiccant, g/g v vapor
Q̇ heat transfer rate, kW w water
qst heat of sorption, kJ/kg wet wetted
Re Reynolds number, -
St Stanon Number, - Abbreviations
T temperature, oC COP coefficient of performance
t time, s COPth thermal coefficient of performance
tcyc cycle time, min DCHE desiccant coated heat exchanger
Umax maximum air velocity, m/s DCHP desiccant coated heat pump
V volume, m3 ETC evacuated tube collector
Win,compressor work input to the compressor, kW MPPT maximum power point tracker
x measured parameters/vapor quality MRC moisture removal capacity, g/kg
Xl longitudinal tube pitch, m PV/T photovoltaic thermal

due to frequent cycle switching (20–30 times per hour) [12]. To produce and condensers with DCHEs [20,21]. In addition, in hybrid de­
a desiccant that displays excellent sorption capacity, high stability, fast humidifiers, DCHEs work alongside conventional MVC chillers in a
kinetics, and low regeneration temperature, composite superabsorbent central air-conditioning system. The working fluid (ambient air) in the
polymers were synthesized [13]. These desiccants are able to achieve 12 open-cycle applications does not undergo a cyclic operation but operates
times improvement in the water sorption capacity and can be regener­ under open loop conditions. Since the composite superabsorbent poly­
ated effectively between 40 and 50 ◦ C. Further, when they are coated on mer desiccants have recorded large sorption capacity for RH that is
the surface of a heat exchanger, they reduce three-times higher cooling greater than 60%, they are often the first-choice consideration for heat
load than silica gel based DCHEs and have recorded two-times higher transformation applications [22]. Therefore, the potential of the com­
thermal efficiency. Also, these DCHEs have exhibited their potential to posite superabsorbent polymer desiccants integrated with heat pumps
cut down 50% of operating electrical consumption and realize short and dehumidifiers is to be studied comprehensively, which is one of the
payback periods of less than 2 years when integrated with conventional key objectives of this paper (Fig. 1).
MVC air-conditioners [3,14,15]. Table 1 summarizes the key merits of In the first application as DCHPs, the desiccant coated evaporator
employing composite superabsorbent polymer coated heat exchangers becomes a standalone component for cooling and dehumidification
over silica gel-based DCHEs. while the compressed refrigerant at high temperature and high pressure
The merits of the composite polymer desiccant have promoted the regenerates the desiccant coated condenser. Tu and co-researchers [23]
suitability of DCHEs in two open-cycle heat transformation applications, built a DCHP with composite silica gel-LiCl-based DCHEs. Energy
namely standalone desiccant coated heat pumps (DCHPs) [16] and analysis was carried out, and a coefficient of performance (COP) of
hybrid desiccant dehumidifiers [17–19]. A conventional direct expan­ around 7 was recorded, roughly two times the average COP of conven­
sion heat pump is converted to a DCHP by retrofitting the evaporators tional air-conditioners. In another proof-of-concept study, COP of

2
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

around 8.0 was achieved by employing a metal–organic framework Table 1

(MOF)-based heat pump [24]. It is noteworthy that these COP values are Comparison of material-, component-, and system-level performance of com­
specific to certain special experimental cases involving short cycle times posite superabsorbent polymer based DCHEs with silica gel coated heat ex­
of 2–3 min, low refrigerant charge in the system, and the use of return air changers [2,3,14,15].
for regenerating the heat exchangers. Under different climatic condi­ Categories Properties Silica gel based Composite
tions and multiple cooling capacities, a more realistic COP range of DCHEs superabsorbent
polymer based
3.8–6 has been reported [25]. Table 2 reviews the different types of
DCHEs
desiccant materials used for heat pumping applications, and to the best
of our knowledge, there exists no analysis in the literature on employing Material- Maximum 28–30% 250–450%
level isothermal
composite superabsorbent polymers for heat pumping applications. equilibrium capacity
Tu and co-researchers [23] simulated the performance of a com­ Desiccant kinetics at 0.27 g(H2O)/g 0.6 g(H2O)/g
posite silica gel-LiCl based DCHP with the lumped parameter model 30 ◦ C and 80% RH (desiccant) (desiccant) at silica
using the ε-NTU method. Their model did not consider the transient gel’s saturation
point.
variation of desiccant temperature, adsorption heat, and the degree of
Component- Desiccant coating Needs 3.3w% No requirement of
superheating/subcooling. Further, the effects of sorption/desorption level hydroxyethyl additional binder is
switching were also ignored. Ge and co-workers [32] developed a cellulose binder necessary as the
mathematical model tailored for the DCHP system by integrating the to achieve superabsorbent
lumped parameter model of DCHEs with a steady-state thermodynamic uniform coating polymer possesses
with Al fins. binding properties.
model of the two-phase refrigerant. The model’s major drawback was
Cycle time 5 min 10–15 min
that it assumed the heat transfer correlations to be constants while Regeneration 60–80 ◦ C 40–50 ◦ C
ignoring their transient variation. Hua and co-researchers [30,31] temperature
addressed this issue by designing a mathematical approach that in­ Average moisture 2–3 g/kg in 5-min 6–7 g/kg in 10-min
removal capacity cycle cycle
corporates the combined transient effect of the two-phase refrigerant,
COPth 0.1–0.2 0.4–0.8
solid desiccant, and air. However, the simulated model poorly predicted Specific entropy 0.8 Jg-1K− 1 0.38 Jg-1K− 1
the dynamic profile of both refrigerant and air inlet/outlet conditions production rates
with discrepancies of around 60% [26,28]. As a result, the inferences System- Energy savings when A specific power SPC of 1.85 Wh/g is
made in their detailed study cannot be generalized for DCHPs regardless level employed for a consumption recorded implying
ventilation air (SPC) of 4Wh/g is over 54% energy
of the desiccant/refrigerant types.
dehumidification necessary. SPC is savings compared to
It is also worthy to note that the existing mathematical models ignore system. defined as the silica gel based
the effects of switching between the dehumidification/regeneration ratio of energy DCHEs.
cycle and assume that the desiccant layer does not contribute to any heat consumption to
the moisture
transfer resistance. The refrigerant flow is considered to be single-phase
removal rate.
for simplicity, albeit the actual process occurring in two-phases.
Therefore, there is a current research gap in the development of a
mathematical model that is capable of (1) accurately predicting the differential equation (ODE) to sufficiently describe the transport phe­
transport phenomenon of multiple desiccants, (2) adapting to both nomena drastically minimizes the need for computational resources. In
single/two-phase refrigerant systems, and (3) evaluating the perfor­ addition, the ODE model permits greater flexibility for the DCHE,
mance of DCHPs under various operating conditions. A DCHP comprises enabling easier simulations using dynamic modeling software platforms
multiple thermal components such as compressors, expansion valves, such as TRNSYS, Mathematica, Simulink, and Dymola. As a result, a
and DCHEs. Theoretical performance prediction involves solving the mathematical model involving only ODEs will suffice in reducing the
inputs and outputs of these components simultaneously. Although a solution complexities due to spatial variation and offer a pragmatic
partial differential equation (PDE) model is arguably more sophisti­ integration with existing air-conditioning control systems.
cated, very often, the spatial aspects of the DCHE can be ignored when As a secondary application, DCHEs are also known to be retrofitted
performance prediction is concerned. Incorporating an ordinary upstream of a cooling coil in an air handling unit (AHU) for the purpose

Fig. 1. A schematic of the advances in composite superabsorbent polymer DCHEs and this study’s objectives.

3
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Table 2 and generated higher entropy that increased life cycle operating costs
A review of different desiccants used for heat pumping applications [9,26]. [39,40]. Nevertheless, using concentrated solar collectors raised the
Literature Desiccant type Type of study and key results water temperature to 120 ◦ C while substantially increasing the regen­
eration cost and generating high irreversibility [2]. Therefore, a desic­
[16] Composite silica gel (16.2w • Experimental dynamic
%) with emulsion glue performance evaluation was cant comprising composite superabsorbent polymer is highly desirable
binder conducted. for energy-efficient regeneration by simply employing a commonly used
A loosely coupled heat and mass solar water heating system. This study further investigates the perfor­
transfer approach model was mance of a composite superabsorbent polymer coated heat exchanger
developed.
System COP of 6.2 could be
coupled with two types of solar energy systems: (1) evacuated tube
achieved. collectors (ETC), and (2) photovoltaic-thermal (PV/T) collectors and
[27] AQSOA Z02 zeolite • A DCHP system for electric vehicles compares their regeneration performance.
was proposed.Simulation of Adopting both experimental and theoretical performance evaluation
heating mode at low ambient
approaches, this paper systematically evaluates the potential of com­
temperatures was
performed.De posite superabsorbent polymer desiccants in two key open-cycle heat
-fogging of the windshield under transformation applications. The novelty and potential key contribu­
interior-air circulation and a tions of this study vis-à-vis works found in the literature, can be sum­
largest cruising range could be marized as follows:
achieved.
[28,29] Silica gel • Both experimental and theoretical
performance evaluation were 1. Developing a simplified lumped parameter mathematical modeling
conducted. platform that evaluates the performance of DCHPs irrespective of
The novel system can supply air desiccant and refrigerant types;
with lower temperature and
2. Conducting parametric analysis of the DCHP to quantify the energy
temperature variation comparing
to traditional heat pump. efficiency improvements realized by a composite polymer DCHP
[25,30,31] Composite silica gel -LiCl • Experimental performance studies over a traditional heat pump system;
(38w%) with emulsion glue were conducted under ARI and 3. Integrating ETC and PV/T with DCHE dehumidifiers and experi­
binder Shanghai summer conditions. mentally evaluating their regeneration feasibility;
A three-dimensional mathemat­
4. Comparing the regeneration performance of ETC and PV/T to
ical model was developed.
Maximum system COP of 7.0 determine the most effective solar water heating design for low-
could be reached. temperature regeneration.
[24] MIL-100(Fe) with a water- • A temperature difference of around
borne silica sol binder 30 ◦ C between the evaporator and
the condenser was sufficient to
2. Experimental methodology
exploit the S-shape isotherm of
MOFs. 2.1. Heat exchanger coating
A system COP of 7.9 could be
attained when the condensation
Table 3 lists the geometric specifications of the fin-tube heat ex­
temperature dropped to 45 ◦ C.
Present Composite superabsorbent • A mathematical model is changers used in this study. These heat exchangers were first dip-coated
study polymer desiccant developed to predict the with the composite superabsorbent polymer desiccant solution based on
performance of composite the coating process described in our previous works [3,12]. The coating
superabsorbent polymer desiccant process was continued until the average desiccant layer thickness of
coated heat pumps.
A parametric analysis is
around 120–150 μm was obtained.
conducted, and the energy
efficiency improvements realized
by DCHP over traditional heat 2.2. DCHP experimental facility
pumps is quantified.
The comparative performance A schematic cum photographic perspective of the DCHP experi­
between a superabsorbent DCHE mental setup is shown in Fig. 2. This setup is a standalone system and
integrated with either a solar
evacuated tube (ETC) or a
can be easily deployed at different locations for performance testing.
photovoltaic thermal (PV/T) SAP-LiCl (40w%) heat exchangers were employed, and these DCHEs
system is judiciously evaluated. have been labelled as “DCHE L” and “DCHE R” indicating left and right,
respectively; based on their relative position in the setup. In the first half
of the cycle, a low-pressure refrigerant flowed through DCHE L (acting
of central air-conditioning [33]. This arrangement ensures that the hot
and humid ambient air first experiences energy-efficient dehumidifica­
Table 3
tion in DCHEs before mixing with the room return air to undergo sen­
Geometrical specifications of the fin-tube heat exchangers used in DCHP and
sible cooling in the active cooling coil. DCHEs can potentially be coupled
DCHE experimental facilities.
with cooling towers to supply the needed cooling water and also be
Parameters Units DCHP DCHE
regenerated by integrating with low-cost solar energy collectors. Since
facility facility
the demand for a higher cooling load often coincides with the solar
irradiation intensity, driving the DCHEs via solar energy offers excellent Length of the fin in the air flow direction mm 80 22
(Lx)
potential to reduce the peak electricity demand and promote energy
Width of the fin (Ly) mm 650 150
savings [34]. In the previous studies involving desiccant cooling sys­ Height of the heat exchanger (Lz) mm 800 200
tems, the need for desiccant’s high regeneration temperature limits the Fin thickness (Hf) mm 0.12 0.12
use of non-concentrated solar collectors [35,36]. Accordingly, auxiliary Fin pitch (Pf)
( )
mm 2.5 1.6
Inner tube diameter dt,i mm 7.9 9.5
electrical water heaters or natural gas boilers were employed to raise the ( )
Outer tube diameter dt,o mm 9.5 10.5
water temperature to the desired levels of 80–120 ◦ C [37,38]. This
No. of transverse tube passes – 25 6
process resulted in lower energy savings, increased system complexity, No. of longitudinal tube passes – 4 1

4
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 2. (a) and (b) Schematic of DCHP experimental system showing two different operating cycles of DCHEs and directional switching of the refrigerant via a four-
way valve, and (c) a picture of the experimental setup.

as an evaporator). The ambient air, passing through the DCHE L, Two differential pressure gauges with integrated temperature scales
experienced cooling and dehumidification. At the same time, the su­ were also placed before and after the compressor to measure the
perheated vapor refrigerant was supplied to the DCHE R (acting as a refrigerant pressure and temperature. Table 4 lists the measuring in­
condenser) at high pressure to facilitate regeneration, where the struments used in the DCHP testing facility.
absorbed moisture was purged out to the airstream. Temperature, hu­
midity sensors, and air velocity transmitters were placed before and 2.3. Dches integrated with solar collector experimental facility
after the DCHEs to evaluate the respective inlet and outlet air conditions.
It is worth noting that since the sensors were installed at the face of the Fig. 3 portrays the schematic and photograph of the testing facility
heat exchangers, their inlet and outlet humidity ratios could be employed to evaluate the performance of a solar ETC and PV/T inte­
measured accurately and fitting bulky ducts before/after the coil to grated DCHE system. The geometric and thermal specifications of ETC
achieve stable readings was not necessary. Further, surface mounted and PV/T are listed in Table 5. Like DCHP testing facility, this setup was
resistance temperature detectors (RTDs) and pressure transducers were also integrated with SAP-LiCl (40w%) coated heat exchangers. Howev­
placed appropriately to measure the refrigerant states. Since the desic­ er, the DCHEs make 30◦ with the airflow direction. Outdoor air was
cant has a finite moisture capturing ability, a four-way valve was drawn into the system during the dehumidification process and was
employed to change the direction of refrigerant flow when the desired directed to pass through the by-pass valve until a stable temperature and
switching period was attained. A bi-directional expansion valve was humidity state could be achieved. Thereafter, the humid air was allowed
used for throttling, and an accumulator was incorporated before the to interact with DCHEs. When the pre-determined operating time for the
compressor to prevent liquid refrigerant flowing into the compressor. dehumidification cycle was attained, the outdoor air was switched to the

5
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 2. (continued).

ensures that the DC output voltage from solar PV/T is efficiently con­
Table 4
verted to a lower voltage suitable for charging batteries. The measuring
Specifications of the measurement instruments used in the DCHP testing facility.
instruments used in the system and their specifications, ranges, and
Parameter Sensors Range Accuracy accuracies are listed in Table 6. A class C Pyranometer with a measuring
Air temperature Kanomax − 20–70 ◦ C ±0.5 ◦ C ranging of 0–2000 W/m2 was used to measure solar irradiance across the
Climomaster full spectrum. Its accuracy is at ± 10% (full-scale) and can operate be­
6501 tween − 40 to 80 ◦ C. Dynamic experiments were carried out and the
Air humidity Kanomax 2.0–98.0% RH ±2% (at range of 2
Climomaster to 80%)±3%
dehumidification performance, cooling capacity, and thermal energy
6501 (at range of 80 to efficiency were consequently computed. The operating parameters
98%) controlled during the study are listed in Table 7.
Air flow velocity Kanomax 0.01–50 m/s ±2% of reading or The ambient air conditions are selected to represent the typical
Climomaster ±0.015 m/s
tropical climate. Since the DCHE-based hybrid dehumidifier will be
6501 (whichever is
greater) employed in indoor central air-conditioning applications, 10–11 g/kg
Tube surface Surface RTD − 50–200 ◦ C ±0.15 ◦ C (1/10 humidity ratio is set for regeneration air to simulate the effect of room
temperature DIN) return air. Based on the detailed parametric analyses carried out in our
Pressure 0–25 bar ±0.5% (Full Scale) previous experimental studies [3,14,15], the values of cycle time,
transmitter
Analog refrigerant Refco pressure − 1–10 bar Class 1.6 accuracy
cooling water temperature, and air flow rate are selected to achieve
pressure gauge gauge (Suction side) (±1.6%) enhanced performance of DCHEs. The air flow rate of 35 kg/h achieves
− 1–30 bar the ideal desiccant-air contact time of 0.1–0.12 s, the cycle time of 7 min
(Discharge side) is selected to enhance dehumidification performance and thermal effi­
ciency while minimizing the effect of thermal inertia. Lastly, the water
flow rate of 4 kg/min is controlled to sufficiently exchange heat within
room return air (30 ◦ C and ~ 11 g/kg) to facilitate the regeneration
the heat exchangers and minimize pumping costs/inertia.
process.
The cooling water flow was achieved by regulating a cooling water
bath, while solar ETC and PV/T supply the required regeneration heat. 2.4. Uncertainty analysis
The electricity generated by the PV/T can be used for operating auxiliary
components of the system, such as pumps or blowers. A maximum power The uncertainties of the measured parameters are directly obtained
point tracker (MPPT) was employed to optimize and match the voltage from the sensor accuracies, and the error propagation method (Eq.) is
between the solar array (PV panels) and the battery bank. The MPPT used to compute the uncertainties of derived parameters [41]. They are

6
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 3. Schematic of the DCHE system coupled with (a) evacuated tube solar collector; and (b) photovoltaic-thermal solar collector, and (c) picture of the solar
integrated DCHE experimental setup.

7
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 3. (continued).

Table 5 Table 7
Geometric and thermal specifications of ETC and PV/T. Operating parameters controlled during the DCHE’s performance evaluation
experiments with ETC and PV/T.
Evacuated tube collector
Parameters Units ETC PV/T
Geometric and thermal specifications Units Values
o
Ta,in C 31.8 29
Heat transfer fluid in the tube Water
ωa,in,d g/kg 18–19 18–19
Collector area m2 2.83
ωa,in,r g/kg 10.5–11 10.5–11
No. of tubes – 20 o
Tcw,in C 26.4 26.4
Tube length m 1.73 o
Thw,in C From solar collectors
Tube inner/outer diameter m 0.047/0.055
ṁa kg/h 35 35
Spacing between the tube centers m 0.08
ṁw kg/min ~4 ~4
Collector efficiency – 0.5
tcyc min 7 7
Tube absorptance – greater than94%
Tube emittance – less than6%
Photovoltaic thermal collector
Geometric and thermal specifications Units Values number of parameters directly measured using the sensors; and f rep­
Heat transfer fluid Water resents the mathematical relationship between the derived and
Overall length m 1.59 measured parameters.
Overall width m 1
Effective collector area m2 1.53
3. Theoretical methodology

In this section, a simplified theoretical approach is adopted to readily

Table 6 compute the performance of both DCHEs and DCHPs. This approach is
Specifications of sensors and measurement instruments used in the DCHE-solar
based on a 1D lumped parameter methodology and is governed by
energy system performance testing setup.
fundamental mass and energy conservation principles. The dynamic
Parameter Sensors Range Accuracy models comprising key governing equations are applied in the three key
Air Temperature 10 k Thermistors − 10–100 ◦ C ±0.15 ◦ C domains of a DCHE – air, desiccant (including fins and tubes), and
Surface 10 k Thermistors − 10–100 ◦ C ±0.15 ◦ C refrigerant. Additionally, a steady-state thermodynamic model is
temperature
established to compute the thermodynamic states of refrigerant oper­
Water RTD temperature − 29–100 ◦ C 1/10 DIN
temperature sensor ating under closed-loop cycle in heat pump conditions. The governing
Air flow rate Thermal mass flow − 10–100 ◦ C ±0.5% (Full equations are simultaneously solved with appropriate initial conditions,
meter Scale) input parameters, and auxiliary equations on the MATLAB Simulink
Water flow rate Compact vortex 3.2–22 L/ ±2.5% (Full (v2021b) software platform.
flowmeter min Scale)
The following assumptions are made to simplify the mathematical
representation of the complex heat and mass transfer phenomena during
then plotted as error bars on the graphical presentations in the results desiccant dehumidification:
and discussion section.
√[ ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ i. The desiccant is uniformly coated on the metallic fins of the heat
√ n ( )2 ]
√ ∑ ∂f exchanger, and comparatively, the coating on tubes is considered
Δy = √ (Δxi ) 2
(1) negligible [42].
∂xi
ii. Sorption/desorption processes are thermodynamically reversible
i=1

and sorption/desorption heat is released in the desiccant layer

where Δy is the uncertainty associated with the derived parameters; Δx
and transferred to the air and refrigerant domains via convection
is the uncertainty associated with the measured parameters; n is the
only. The effect of radiation heat transfer is considered negligible.

8
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

iii. The thermodynamic properties of both air and desiccant are area of heat transfer between the refrigerant and the desiccant, Tr is the
assumed to be constants in the operating range of the DCHE/ instantaneous temperature of the refrigerant; and cp,v is the specific heat
DCHP system. capacity of water vapor.
iv. The effect of thermal contact resistance between the fins and Since the refrigerant does not interact with the desiccant and
tubes is neglected [43–45]. contribute to the moisture transfer, only the energy conservation
v. Evaporation and condensation temperature remain constant equation is used. The energy conservation equation for the two-phase
throughout the heat exchanger and the evaporation/condensa­ refrigerant flow is not as straightforward as the energy conservation
tion heat is distributed evenly. principle for a single-phase flow. However, by adopting a simplified
assumption of validity of single-phase governing equation for the two-
phase zone, Eq. is derived [46,47]. The meaning and the significance
of the terms is identical to those in the previous equations. However, the
3.1. Governing equations
variables correspond to the average value of corresponding liquid and
vapor properties of the refrigerant, and these values can be obtained
The mass and energy conservation equations for air are given by Eq.
from CoolProp software platform [48].
and, respectively. The terms on the left side of both equations represent
the respective moisture and energy storage in air. The first term on the d(ρr ir ) ( )
Vr = ṁr ir,in − ir,out + ht,rd Ard (Tr − Td ) (7)
right side indicates the mass/energy transfer due to the airflow, and the dt
second term represents the mass/energy transfer between the air and the
desiccant. Further, the third term of Eq. records the sensible heat where ρr is the density of the two-phase refrigerant; Vr is the volume of
transfer associated with the transfer of moisture between the air and the the refrigerant in the system; ṁr is the mass flow rate of the refrigerant;
desiccant. and ir,in and ir,out are the respective inlet and outlet refrigerant
enthalpies.
dωa ( )
ρa,dry Va = ṁa ωa,in − ωa,out + ρa,dry hm Ada (ωd − ωa ) (2)
dt
3.2. Auxiliary equations
dT ( )
ρa,dry Va cp,a,dry a = ṁa cp,da Ta,in − Ta,out + ht,da Ada (Td − Ta ) + ρa,dry hm Ada (ωa
dt The five governing equations contain 12 unknown quantities that
− ωd )cp,v (Td − Ta ) need to be solved simultaneously. These unknowns are in addition to the
(3) thermo-physical properties of air, desiccant, water, and the geometric
specifications of the heat exchangers. To obtain a well posed solution,
where ρa,dry is the density of dry air (taken as a constant within the additional equations are needed to satisfy the number of unknowns.
operating range); Va is the volume of air in the heat exchanger; ωa is the While approximations on the instantaneous temperature and humidity
humidity ratio of air; ṁa is the desired mass flow rate of air; ωa,in and ratio can be made as shown in Eqs. -, several correlations for deter­
ωa,out are the respective inlet and outlet air humidity ratios; hm is the mining the heat and mass transfer coefficients are needed.
mass transfer coefficient between air and desiccant; Ada is the area of ωa,in + ωa,out
ωa = (8)
mass transfer between air and desiccant; cp,da is the specific heat ca­ 2
pacity at constant pressure for dry air; Ta is the absolute temperature of
air; Ta,in and Ta,out are the inlet and outlet temperature of moist air, Ta =
Ta,in + Ta,out
(9)
respectively; ht,da is the heat transfer coefficient between the air and the 2
desiccant; and Td is the desiccant temperature.
ir,in + ir,out
The mass and energy balance equations for the desiccant are pre­ ir = (10)
2
sented in Eqs. and, respectively. In Eq., the term on the left side accounts
for the mass stored in the desiccant, and the term on the right side in­ Air-desiccant side heat transfer coefficient is calculated based on the
dicates the transfer of moisture between the air and desiccant layer. extensive experimental correlations of the fin-tube heat exchangers
Likewise, the left side of Eq. represents the energy storage rate of the [49,50].
desiccant, while the first term on the right side denotes the sorption heat ( )c1 ( )− 1.084 ( )− 0.786 ( )c2
Xt Pf Pf Pf
released/absorbed by the desiccant during dehumidification/regenera­ j = 0.108Re−dc0.29 (11)
Xl dc Dh Xt
tion processes, respectively. The second and third terms signify the heat
transfer rates of the desiccant layer with the air and the refrigerant, c1 = 1.9 − 0.23lnRedc
respectively. The last term pertains to the sensible heat transfer associ­ (12)
c2 = − 0.236 + 0.126lnRedc
ated with the transfer of moisture between the air and the desiccant.
ρa Umax dc
ρ d Vd
dq
= ρa,dry hm Ada (ωa − ωd ) (4) Redc = (13)
dt μa

dTd dq where j is the Colburn’s j-factor, Redc is the Reynold’s number, dc is the
ρeff Veff cp,eff = ρd Vd qst + ht,da Ada (Ta − Td ) + ht,rd Ard (Tr − Td ) fin-collar outside diameter, Xt is the transverse tube pitch, Xl is the
dt dt (5)
+ρa,dry hm Ada (ωa − ωd )cp,v (Ta − Td ) longitudinal tube pitch, Pf is the fin pitch, Dh is the hydraulic diameter,
µa is the dynamic viscosity of air, Umax is the maximum air velocity.
ρeff Veff cp,eff = ρd Vd cp,d + ρAl VAl cp,Al + ρCu VCu cp,Cu (6) The Colburn’s j-factor is related to the air-desiccant heat transfer
coefficient via the following equations:
where ρd, ρCu, and ρAl are the respective densities of the desiccant, Cu (
ht,da
)
tube, and Al fin; Vd, VCu, and VAl are the volumes of the desiccant layer, j = StPr2/3 = Pr2/3 (14)
Gcp,a
Cu tube, and Al fin, respectively; q is the instantaneous uptake capacity
of the desiccant; cp,d is the specific heat capacity at constant pressure for ρa Ufr Afr
desiccant; qst is the heat of sorption released by the desiccant; ht,da is the G = ρa Umax = (15)
A0
heat transfer coefficient between air and desiccant; ht,rd is the heat
transfer coefficient between the refrigerant and the desiccant; Ard is the where St is the Stanon number; Pr is the Prandtl number; Ufr is the

9
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

frontal air velocity; Afr is the frontal area; A0 is the minimum free flow
qe = f (φd ) (28)
area. The geometrical specifications of the heat exchanger required to
evaluate the heat transfer coefficient are listed in Eqs. -. The details of
0.62188φd
the equations are well documented in the literature and can be referred ωd = Patm (29)
− φd
to [43,49,50]. Pv,sat

Awet A0 Lastly, the desiccant-refrigerant heat transfer coefficient (hrd) is

V = Lx Ly Lz ; Afr = Ly Lz ; α = ;σ = (16) determined based on phase condition (single-phase or two-phase
V Afr
refrigerant) and the nature of the fluid flow in the tubes. Under
Lx Lx Ly Lz laminar flows, the evaluation of the heat transfer coefficient is
Nr = ; Nt = ; Nf = (17) straightforward as the Nusselt’s Number assumes a constant value [52].
Xl Xl Xt Pf
However, when the flow is turbulent and a single-phase refrigerant
(
( ( )) (water) is used, hrd is evaluated via the Dittus Boelter correlation [52].
Awet = πdt,o Lz − Nf 2Hd + Hf Nt + 2Nf Lx Ly ( )
0.023Re0.8 n
w Pr w kw
( )) ht,rd = (30)
2
Nt πdt,o ( ) Dw,i
− + 2Ly Nf 2Hd + Hf (18)
4 4ṁw
where Re is the Reynolds Number evaluated usingRew = πμw Dw,i ; Prw is
((
Ly
)
( ) (
(
) Nf Hf + 2Hd
)) the Prandtl number; kw is the thermal conductivity of water; Dw,i is the
A0 = Lz − 1 c′′ + Xt − dt,0 − Xt − dt,0 (19) inner diameter of water pipe; μw is the dynamic viscosity of water; and
Xt Lz
the value of the constant n is taken as 0.4 during dehumidification and
c′′ = min(2a′′ , 2b′′ ) (20) 0.3 during regeneration.
For two-phase mixtures, the Nusselt’s number for turbulent flow
( )
( ) ( ) Nf Hf + 2Hd follows the Cavallini and Zecchin correlation [52], where subscripts g
2a′′ = Xt − dt,0 − Xt − dt,0 (21) and f denote the saturated vapor and saturated liquid states of the two-
Lz
phase mixture respectively; and × is the quality of the mixture.
⎛√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ))⎞
( )2 ( ( [ √̅̅̅̅̅ ]0.8
X ( ) Nf Hf + 2Hd vg
(22) (31)
t
2b = 2⎝
′′ 2
X l − dt,0 − Xt − dt,0 ⎠ Nu = 0.05 1 − x + x Ref Pr0.33
2 Lz vf f

dc = dt,0 + 2Hf (23) 3.3. Modeling software

4σ The governing and auxiliary equations are dynamically programmed

Dh = (24)
α on MATLAB (2021b)-Simulink graphical interface with each physical
system modeled in the Simscape environment. The use of Simscape
where V is the volume of the heat exchanger; Lx is the length of the fin enables modeling thermal components as block diagrams that can be
along the air flow direction; Ly is the width of the fin; Lz is the height of integrated with each other using physical graphical connections, as
the heat exchanger; α is the ratio of heat transfer area on the air side to presented in Fig. 4. The compressor, expansion valve, and two DCHEs
the volume of the heat exchanger; Awet is the wetted area; σ is the ratio of operating as evaporator and condenser can be assembled, and their
minimum free flow area to the frontal area; Nr is the number of tube dynamic models are solved using MATLAB’s ODE solver.
rows; Nt is the total number of tubes; Nf is the number of fins; dt,o is the
outer tube diameter; Hd is the desiccant layer thickness; and Hf is the fin 4. Performance parameters evaluation
thickness.
The mass transfer coefficient is obtained by employing the heat and The desiccant dehumidifier’s moisture removal capacity (MRC) is
mass transfer analogy for air–water mixture. For air–water mixture, defined as the average amount of moisture removed per kg of the dry air
since the Lewis Number is approximately 1, the mass transfer coefficient and is represented as.
can be obtained via.
∫
1 tcyc ( )
ht,da Da MRC = ωa,in − ωa,out dt (32)
hm ≃ (25) tcyc 0
ka
where ωa,in and ωa,out are the air humidity ratio values measured at inlet
where ka is the thermal conductivity of air, Da is the moisture diffusivity
and outlet of the DCHE chamber.
in air, which is also dependent on temperature and can be evaluated by
The energy efficiency of a dehumidifier is quantified by its thermal
employing the Hall and Pruppacher correlation, which is displayed in
coefficient of performance (COPth), which is defined as the de­
Eq. [51]. ( )
( )1.94 humidifier’s average cooling capacity Q̇a to the average heat ex­
Ta
Da = 2.11 × 10− 5 (26) ( )
273.15
change rate of hot water during regeneration process Q̇w , and is
To obtain the equivalent air humidity ratio at the desiccant(ωd ),
computed via.
desiccant kinetics data is used, as portrayed in Eqs. -. The value of k is a
constant whereas the value of qe is a function of desiccant’s RH(φd ). Q̇a
COPth = (33)
These are desiccant-dependent properties which are adopted from the Q̇w
material characterization experimental results of our previous works
[3,14]. Q̇a is calculated using Eq., where ha,in and ha,out represent the
respective air-on and air-off the desiccant. Further, Q̇w is computed via
dq
= k(qe − q) (27) Eq..
dt
( )
Q̇a = ṁa ha,in − ha,out (34)

10
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 4. Schematic of the graphical mathematical model in MATLAB-Simulink interface using the Simscape environment.

( ) 5.1.1. Validation with results from DCHEs

Q̇w = ṁw cp,w Thw,in − Thw,out (35)
Fig. 5 (a) and (b) illustrate the transient behavior of the air humidity
Similarly, the energy efficiency of the DCHP system is computed ratio and temperature obtained by the mathematical model and the
using its COP, as shown in Eq.. Since electrical energy is used to drive the experiments conducted on silica gel coated heat exchangers [42]. Water
compressor, the denominator is represented as the work input to operate was used as the refrigerant, and the DCHEs were angled at 30◦ with the
the compressor, which is measured using a power clamp meter with an airflow direction. The airflow rate, water flow rate, cooling water tem­
accuracy of ± 0.2% in V and ± 0.5% in A. perature, and regeneration temperature were maintained as 55 kg/h, 4
kg/min, 25 ◦ C, and 70◦ , respectively. The model predicts the overall
COP =
Q̇a
(36) transient profile of the air humidity ratio and temperature during
Win,compressor dehumidification and regeneration cycles with good accuracy. The
normalized root mean squared errors (NRMSE) range between 6 and 9%.
5. Results and discussion Specifically, the accuracy is high while calculating the peak values of ω
and T during both cycles. However, towards the end of the regeneration
5.1. Validation of the mathematical model process, the model overestimates the air humidity ratio and under-
predicts the air temperature. This observation is attributed to the
In this section, the results from the mathematical model are validated following reasons. Firstly, an approximation is relied on to compute the
with the experimental results obtained from the dehumidification and overall volume of the air in contact with the DCHE since estimating the
regeneration cycles of both DCHE and DCHP testing facilities. This exact volume of air is complex and difficult. Secondly, potentially there
extensive validation exercise ensures that the developed mathematical exist some time delays at the sensor location causing these discrep­
model is independent of the experimental setup, desiccant and refrig­ ancies. Lastly, non-uniform desiccant coating and unavoidable heat
erant used, and airflow configurations. The types of desiccant employed, losses during the experiments contribute to further disparities. Despite
their coating amount, and the isothermal characteristics are highlighted these unavoidable experimental uncertainties, the maximum discrep­
in Table 8. ancy in predicting the humidity ratio and temperature is ±12%

Table 8
Desiccant type and isothermal characteristics for model validation.
Desiccant Experimental setup Coating mass Isotherm Curve

Silica gel DCHE setup in Section 2.3 50 g qe = − 0.0035φ2 + 0.6113φ + 1.3582

( )
SAP-LiCl (50w%) DCHE setup in Section 2.3 60 g qe = 1.68881 × 10− 5 φ4 − 0.00191φ3 ...
... + 0.06545φ2 + 1.50439φ - 3.02843
( )
SAP-LiCl (40w%) DCHP setup in Section 2.2 1.3 kg qe = 1.23794 × 10− 5 φ4 − 0.00131φ3 ...
2
... + 0.05087φ + 0.30402φ + 1.99531

11
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 5. (a) and (b) Experimental results and model validation of the transient response of outlet air humidity ratio and outlet air temperatures; (c) and (d) accuracy of
the numerical simulation with respect to the experimental data for silica gel coated heat exchangers.

and±9%, respectively as shown in Fig. 5 (c) and (d). Further, the highest discrepancies between the model and the experi­
Fig. 6 (a) and (b) compare the transient behavior of the air humidity ments are shown in Fig. 7 (c) and (d), and they range between 10 and
ratio and temperature simulated by the mathematical model and the 12%.
experiments conducted on SAP-LiCl (50w%) coated heat exchangers. In
this experiment, water is used as the heat transfer fluid while the DCHEs
are aligned perpendicular to the air flow direction. The airflow rate, 5.2. Parametric studies on DCHPs
water flow rate, cooling water temperature, and hot water temperature
were maintained at 27 kg/h, 2.5 kg/min, 20 ◦ C, and 40 ◦ C, respectively In the ensuing sections, the validated mathematical model is
[3]. The operating parameters for validation are judiciously chosen to be employed to simulate the effect of different experimental conditions,
different from the previously obtained values to ensure the model’s namely, inlet air temperature, evaporator temperature, and cycle time.
applicability across a wide range of operating conditions. Albeit the Additionally, the COP improvements realized by DCHP over a tradi­
change in desiccant type, operating parameters, and the airflow tional heat pump (THP) is also judiciously computed.
configuration, the model is able to predict humidity ratio and temper­
ature profile with an NRMSE of 4–11%. The accuracy of the numerical 5.2.1. Effect of inlet air and evaporator temperatures
simulation is highlighted in Fig. 6 (c) and (d). The maximum discrep­ Fig. 8 shows the results of different inlet air and evaporator tem­
ancies in predicting the humidity ratio and temperature are ±11% peratures on DCHP’s performance. The inlet air temperature is regulated
and±9%, respectively. from 28 to 36 ◦ C whereas the evaporator temperature is increased from
12 to 21 ◦ C. It is worthy to note that the impact of the inlet air tem­
5.1.2. Validation with results from DCHPs perature on the latent load removal is observed to be marginal. This is
The mathematical model is further validated with experiments car­ because the change in air temperature does not affect the desiccant’s
ried out on the DCHP facility. SAP-LiCl (40w%) coated heat exchangers moisture sorption ability under a cycle time of 10 min. Its effect is more
are used as desiccant-coated evaporators and condensers, and R-134a is pronounced when the cycle time is increased beyond 30 min [15]. In
the operating refrigerant. The evaporator and condenser temperature contrast, the evaporator temperature has a significant effect on DCHP’s
were chosen at 20 ◦ C and 45 ◦ C, respectively, and the degree of super­ latent cooling performance. A lower evaporator temperature promotes
heating and subcooling is reasonably maintained in the range spanning the heat and mass transfer rates and improves cooling and dehumidifi­
5–7 ◦ C. The transient behavior of the outlet air conditions of the DCHP cation potential. As a result, the outlet air humidity ratio is reduced by
system is portrayed in Fig. 7 (a) and (b), and the model predicts the 33% when the evaporator temperature is lowered to 12 ◦ C. While the
outlet air humidity ratio and temperature with an NRMSE of 8–13%. cooling effect is enhanced at lower evaporator temperatures, a higher
compressor power is necessary to maintain the system, thereby

12
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 6. Experimental results and model validation of the transient response of (a) outlet air humidity ratio and (b) outlet air temperature under the dehumidification
and regeneration modes of SAP-LiCl (50w%) coated heat exchanger; Accuracy of the numerical simulation with respect to the experimental data for (c) outlet
humidity ratio and (d) outlet air temperature of SAP-LiCl (50w%) coated heat exchanger.

negatively affecting the DCHP’s COP. From Fig. 8 (b), around 3.5% of 200 s. After that, a gradual rise is seen in both outlet air humidity ratio
power savings can be realized for every degree rise in the evaporator and temperature profiles. The cooling load handled by the heat pump
temperature. also rises significantly from 0 to 0.38 kW at 220 s and drops gradually by
13% until 840 s. Fig. 9 (d) shows that when the cycle time is increased
5.2.2. Effect of cycle time from 120 to 240 s, the COP appreciates by approximately 7% across
The desiccant must undergo periodic regeneration cycles since there various evaporation temperatures. After that, the COP falls linearly
is a limit imposed on its sorption capacity. Before studying the effect of between 2 and 4%. If to attain the highest COP is the most desirable
cycle time on the DCHP’s cooling performance and COP, it is critical to outcome in DCHP, an optimal switchover period in the range of around
understand the role of thermal mass due to frequent switching and 240 s (4 min) is then recommended. However, considering the practical
identify a lower limit of tcyc for DCHP’s practical operation. Fig. 9 (a) and constraints of operating the DCHP system for 10–15 min as portrayed in
(b) illustrate the time taken for the refrigerant temperature to attain Fig. 9 (a) and (b), a trade-off is borne out. A minimal performance
stability. Around 150 s is required for the heat exchangers to overcome reduction of 3–5% in terms of COP is reasonable in exchange for higher
the thermal mass of its previous cycle. When the DCHP’s tcyc is 300 s, the thermal stability of the system.
second half of the cycle (i.e., between 150 s and 300 s) is available for
achieving the desired performance. However, when the cycle time is 5.2.3. Comparison with conventional heat pumps
increased to 900 s, a duration spanning 700–750 s is utilized to realize Fig. 10 (a) compares the outlet air humidity ratio achieved by THPs
the desired cooling effect. As a result, operating the DCHP at 15 min and DCHPs operating at different evaporator temperatures. Accounting
cycle markedly reduces the effect of thermal mass per cycle. Further, tcyc for the contributions of dehumidified condensation and desiccant
of less than 200 s has been frequently adopted in the literature due to sorption mechanisms, the DCHP is observed to remove around 40%
silica gel’s lower sorption capacity and its potential to undergo deli­ higher moisture than the THP. As the condensation mechanism domi­
quescence. Operating DCHPs at this cyclic period is deemed to be nates at lower evaporator temperatures, a 30% higher cooling load is
impractical based on the current experimental observation. In contrast, removed by both systems. Nevertheless, Fig. 10 (b) conveys that the
composite superabsorbent polymer desiccant can hold significantly higher evaporator temperature translates to an improved COP. This is
more moisture; therefore, achieving a tcyc of 15 min is reasonable. attributed to the compressor’s lower power assumption to sustain the
Fig. 9 (c) depicts the simulation results of the outlet air conditions respective refrigerant pressure differential. It is apparent that to achieve
and cooling load removal as a function of the cycle time. At the onset of similar latent load removal, THP will have to operate at lower evapo­
the cycle, the outlet air humidity ratio and the temperature drop rator temperatures than DCHP. Therefore, operating at a comparatively
significantly from their initial values and reach a minimum at around higher evaporation temperature via the employment of DCHPs will lead

13
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 7. Experimental results and model validation of the transient response of (a) outlet air humidity ratio and (b) outlet air temperature under the dehumidification
and regeneration modes of SAP-LiCl (40w%) coated heat pump with R134a refrigerant; Accuracy of the numerical simulation with respect to the experimental data
for (c) outlet humidity ratio and (d) outlet air temperature.

Fig. 8. The effect of inlet air temperature rise from 28 to 36 ◦ C and evaporator temperature variation from 12 to 21 ◦ C and on DCHP’s (a) outlet air humidity ratio
and (b) COP.

to attaining higher energy efficiency. approximately 10 ◦ C higher than the PV/T system. This is because the
vacuum sealing in the tubes negates any convective heat losses through
the collector. Additionally, a portion of the solar irradiation incident on
5.3. Performance of ETC and PV/T integrated DCHEs the PV/T collector gets converted to useful electrical work, which also
contributes to the reduced water temperature from the PV/T.
5.3.1. Dehumidification performance Fig. 12 shows the transient results of the outlet air humidity ratio
Fig. 11 (a) highlights that the dynamic solar irradiation intensity on during dehumidification and regeneration processes of the SAP-LiCl
ETC and PV/T on different experiment days are comparable. Accord­ (40w%) coated heat exchanger system when integrated with ETC and
ingly, Fig. 11 (b) illustrates the hot water temperature generated by the PV/T. The cyclic averages of key parameters are presented in Table 9. In
respective solar collectors. The water outlet temperature from the ETC is

14
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 9. (a) and (b) Experimental results of the transient refrigerant temperature of a DCHP when cycle time is 300 s and 900 s; (c) simulated results of outlet air
humidity ratio, outlet air temperature, and cooling load removal by the DCHP at 18 ◦ C evaporator temperature and cycle times spanning 0–840 s, and (d) simulated
results on the effect of evaporator temperature variation from 12 ◦ C to 21 ◦ C and the cycle time increase from 0 to 840 s.

Fig. 10. Comparison of (a) outlet air humidity ratio and (b) COP between traditional heat pumps and DCHPs under the variation of evaporation temperature from
12 ◦ C to 21 ◦ C.

both regeneration cases, the outlet air humidity ratio first drops rapidly dehumidification and regeneration processes, the quality of the regen­
by about 30% from 18 g/kg to 12.5 g/kg at the onset of the dehumidi­ eration process is verified. If the condition of this equation is satisfied,
fication process. This is due to the strong driving force available for then the amount of moisture absorbed by the desiccant during dehu­
moisture transport to take place from air to the dry desiccant layer. As midification process is fully desorbed during the regeneration process.
the desiccant continues to absorb, ωa,out rises by 20% and attains 15 g/kg The observations made in this experimental study are consistent with
at the end of the cycle time of 420 s, when the cycle is switched to the results observed for composite polymer DCHEs in [3].
regeneration mode. Hypothetically if the desiccant is permitted to cap­ ∫ tcyc ∫ tcyc
( ) ( )
ture moisture sufficiently longer than 420 s, the difference between ωa,in ṁa ωa,in − ωa,out = ṁa ωa,out − ωa,in (37)
and ωa,out would continue to fall and negate the beneficial dehumidifi­ 0
⏟̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅⏞⏞̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅⏟ 0
⏟̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅⏞⏞̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅⏟
cation effect. Therefore, upon switching the cycle at 420 s for regener­ dehumidification regeneration

ation, the converse effect for dehumidification is perceived: ωa,out Despite ETC achieving a higher regeneration temperature, the MRC
immediately rises to 18 g/kg whent > 420s. Subsequently, a steady drop difference recorded between the DCHEs integrated with ETC and PV/T is
in ωa,out is observed until ωa,out approaches a value closer toωa,in . By considered to be marginal (less than1.5%). This is attributed to the low-
employing the mass balance equation, Eq. (37), between temperature regeneration capability of the superabsorbent polymer

15
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Fig. 11. (a) Dynamic solar irradiation during the experimental duration as measured by the pyranometer and (b) hot water temperature inlet to DCHE produced by
ETC and PV/T.

Fig. 12. Dynamic performance results of a DCHE integrated with (a) ETC and (b) PV/T solar collectors.

desiccant. The RH of the air at the desiccant temperatures of 40 ◦ C and in MRCs. On the other hand, COPth of the DCHE-PV/T system is
50 ◦ C is 22% and 14%, respectively. Since the equilibrium isothermal computed to be 10% higher than that of the DCHE-ETC system. This is
capacity does not vary markedly between 10 and 30% RH, the desic­ attributed to the rise in regeneration temperature, which translates to a
cant’s moisture removal potential is maintained. As a result, the higher significant upswing in the thermal component of exergy [2]. Accord­
Thw,in obtained by ETC does not bring about any effective improvement ingly, the entropy production rates are higher and the potential to do

16
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

Table 9 predictability of the model is independent of the desiccant employed,

Time-averaged outlet air conditions and performance metrics of DCHE inte­ type of refrigerant used, airflow configurations, and operating
grated ETC and PV/T systems. parameters;
Parameters Units ETC PVT 2) The DCHP removes about 40% higher moisture than the traditional
Ta,out o
C 28.3 26.6
heat pump units operating under similar conditions. Further, a 30%
ωa,out g/kg 15.1 15.2 higher latent cooling load is removed when the evaporator temper­
MRC g/kg 3.43 3.38 ature is reduced from 21 to 12 ◦ C, while the inlet air temperature has
COPth – 0.53 0.59 a marginal effect on the DCHP’s performance;
3) The DCHP’s switching cycle demonstrates that the heat pump needs
at least 150 s to achieve steady refrigerant conditions after the
useful work is also reduced. Consequently, PV/T achieve better thermal
switching process. As a result, a cycle time of at least 4–5 times
efficiency despite producing lower regeneration temperature.
higher should be considered for operating the DCHP system;
4) Both types of solar water heaters, namely, ETC and PV/T, are capable
5.3.2. Effect of regeneration water flow rate
of effectively regenerating DCHEs. However, under similar solar
The water flow rate is a key parameter to be considered as a higher
irradiation and comparable collector area, PV/T produces hot water
flow rate promotes heat transfer rate and enhances MRCs at the expense
at 40 ◦ C, which is around 10 ◦ C lower than the ETC;
of consuming higher power. Its effect on the dynamic dehumidification
5) Despite a lower water temperature provision from the PV/T, MRC
performance of DCHEs is highlighted in Fig. 13 (a). There is no
does not vary significantly between PV/T and ETC. PV/T achieves
observable change in the dynamic profile ofωa,out . This effect is due to
higher energy efficiencies by taking the advantage of the DCHE’s
two key reasons: (1) a higher water flow rate does not directly impact
lower entropy generation rates; and
the driving force for moisture sorption, and (2) ΔThw drops propor­
6) The regeneration water flow rate does not markedly impact the
tionally at higher flow rates, resulting in ṁhw cp,w ΔThw being relatively
DCHE’s performance. As a result, selecting a lower flow rate is ad­
constant. It is, therefore, apparent that, Fig. 13 (b) shows that both MRC
vantageous in terms of achieving pumping energy savings.
and COPth of the system remain stable at 3.5 g/kg and 0.5, respectively.
Since the regeneration water flow rate has a negligible effect on the
DCHE’s performance, selecting a lower flow rate is beneficial in over­
coming the necessary head loss in order to realize higher pump energy CRediT authorship contribution statement
savings.
Vivekh Prabakaran: Conceptualization, Methodology, Investiga­
6. Conclusions tion, Software, Validation, Visualization, Data curation, Writing –
original draft, Writing – review & editing. Duc Thuan Bui: Validation,
In this study, the feasibility of employing composite superabsorbent Formal analysis, Writing – review & editing. Md. Raisul Islam: Vali­
polymer DCHEs in two open-cycle heat transformation applications has dation, Supervision, Methodology, Funding acquisition, Writing – re­
been evaluated comprehensively. Firstly, the DCHE is capable of view & editing. Kian Jon Chua: Funding acquisition, Project
replacing the evaporators and condensers of a conventional heat pump administration, Resources, Conceptualization, Formal analysis, Super­
system and effectively remove both sensible and latent cooling loads. A vision, Writing – review & editing.
simplified mathematical model has been developed to predict the per­
formance of DCHPs. Secondly, a detailed parametric study has been Declaration of Competing Interest
conducted to understand the factors affecting DCHP’s performance.
Thirdly, experiments were carried out to compare the regeneration The authors declare that they have no known competing financial
performance of DCHEs using solar ETC and PV/T. Several key findings interests or personal relationships that could have appeared to influence
that emerged from this study include. the work reported in this paper.

1) A 1D lumped parameter model predicts the performance of both Data availability

DCHEs and DCHPs with the maximum discrepancy of ± 12% in
outlet air humidity ratio and ± 10% in outlet air temperature. The Experimental and modeling data are already provided in the paper.
The modeling schematic, equations used, assumptions, parametric

Fig. 13. (a) Dynamic results of the outlet air humidity ratio of DCHE-ETC system, and (b) MRC and COPth of the DCHE system under the effect of variation of
regeneration water flow rate is increased from 4 kg/min to 7 kg/min.

17
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

constants, and operating conditions are also provided. [25] Hua LJ, Jiang Y, Ge TS, Wang RZ. Experimental investigation on a novel heat pump
system based on desiccant coated heat exchangers. Energy 2017;142:96–107.
https://doi.org/10.1016/j.energy.2017.07.141.
Acknowledgements [26] Prabakaran V, Chua KJ. Advances in desiccant dehumidification : from
fundamentals to applications 2021.
[27] Zhang L, Hashimoto K, Hasegawa H, Saikawa M. Performance analysis of a heat
This research is supported by National Research Foundation (NRF), pump system with integrated desiccant for electric vehicles. Int J Refrig 2018;86:
Singapore, and Building Construction Authority (BCA) under its Green 154–62. https://doi.org/10.1016/j.ijrefrig.2017.10.036.
Buildings Innovation Cluster (GBIC) Funding Scheme (BCA 94.23.1.11). [28] Yang TY, Ge TS, Lu FL, Dai YJ, Wang RZ. A novel semi-coupled solid desiccant heat
pump system - Part 1: Simulation study. Int J Refrig 2020;120:150–60. https://doi.
org/10.1016/j.ijrefrig.2020.09.005.
References [29] Ge TS, Yang TY, Lu FL, Dai YJ, Wang RZ. A novel semi-coupled solid desiccant heat
pump system Part 2: Experimental investigation. Int J Refrig 2021;121:86–94.
https://doi.org/10.1016/j.ijrefrig.2020.10.008.
[1] Institution of Engineers. Mechanical Ventilation (ACMV) Systems 2016.
[30] Hua LJ, Ge TS, Wang RZ. A mathematical model to predict the performance of
[2] Vivekh P, Bui DT, Islam MR, Zaw K, Chua KJ. Experimental performance
desiccant coated evaporators and condensers. Int J Refrig 2020;109:188–207.
evaluation of desiccant coated heat exchangers from a combined first and second
[31] Hua LJ, Ge TS, Wang RZ. Extremely high efficient heat pump with desiccant coated
law of thermodynamics perspective. Energy Convers Manag 2020;207:112518.
evaporator and condenser. Energy 2019;170:569–79. https://doi.org/10.1016/j.
[3] Vivekh P, Islam MR, Chua KJ. Experimental performance evaluation of a composite
energy.2018.12.169.
superabsorbent polymer coated heat exchanger based air dehumidification system.
[32] Ge TS, Dai YJ, Wang RZ. Analysis on integrated low grade condensation heat
Appl Energy 2020;260:114256. https://doi.org/10.1016/j.apenergy.2019.114256.
powered desiccant coated vapor compression system. Appl Therm Eng 2018;138:
[4] Jagirdar M, Lee PS, Padding JT. Performance of an internally cooled and heated
307–18. https://doi.org/10.1016/j.applthermaleng.2018.04.044.
desiccant-coated heat and mass exchanger: Effectiveness criteria and design
[33] Prabakaran V. Studying the transport phenomena of superabsorbents: an
methodology. Appl Therm Eng 2021;188:116593. https://doi.org/10.1016/J.
experimental and theoretical approach 2020.
APPLTHERMALENG.2021.116593.
[34] Vivekh P, Chua KJ, Islam MR. A comparison between the regeneration
[5] Chen WD, Vivekh P, Liu MZ, Kumja M, Chua KJ. Energy improvement and
performance of solar evacuated tube and photovoltaic thermal collectors
performance prediction of desiccant coated dehumidifiers based on dimensional
integrated desiccant coated heat exchangers. Energy, Bangkok/Virtual: Int. Conf.
and scaling analysis. Appl Energy 2021;303:117571. https://doi.org/10.1016/j.
Appl; 2020.
apenergy.2021.117571.
[35] Eicker U, Schneider D, Schumacher J, Ge T, Dai Y. Operational experiences with
[6] Saeed A, Al-Alili A. A review on desiccant coated heat exchangers. Sci Technol
solar air collector driven desiccant cooling systems. Appl Energy 2010;87(12):
Built Environ 2017;23:136–50. https://doi.org/10.1080/
3735–47.
23744731.2016.1226076.
[36] Beccali M, Finocchiaro P, Nocke B. Energy and economic assessment of desiccant
[7] Amani M, Bahrami M. Greenhouse dehumidification by zeolite-based desiccant
cooling systems coupled with single glazed air and hybrid PV/thermal solar
coated heat exchanger. Appl Therm Eng 2021;183:116178. https://doi.org/
collectors for applications in hot and humid climate. Sol Energy 2009;83:1828–46.
10.1016/j.applthermaleng.2020.116178.
https://doi.org/10.1016/j.solener.2009.06.015.
[8] Vivekh P, Kumja M, Bui DT, Chua KJ. Recent developments in solid desiccant
[37] Heidari A, Roshandel R, Vakiloroaya V. An innovative solar assisted desiccant-
coated heat exchangers – A review. Appl Energy 2018;229:778–803. https://doi.
based evaporative cooling system for co-production of water and cooling in hot and
org/10.1016/j.apenergy.2018.08.041.
humid climates. Energy Convers Manag 2019;185:396–409. https://doi.org/
[9] Prabakaran V, Chua KJ. Advanced energy-related applications of desiccants. Adv
10.1016/j.enconman.2019.02.015.
Desiccant Dehumidification 2021:127–52. https://doi.org/10.1007/978-3-030-
[38] White SD, Lin S, Guo J, Sproul AB, Bilbao JI. A review of photovoltaic thermal (PV/
80843-3_5.
T) heat utilisation with low temperature desiccant cooling and dehumidification.
[10] Ashraf S, Sultan M, Bahrami M, McCague C, Shahzad MW, Amani M, et al. Recent
Renew Sustain Energy Rev 2016;67:1–14. https://doi.org/10.1016/j.
progress on water vapor adsorption equilibrium by metal-organic frameworks for
rser.2016.08.056.
heat transformation applications. Int Commun Heat Mass Transf 2021;124:105242.
[39] Jani DB, Mishra M, Sahoo PK. A critical review on application of solar energy as
[11] Zu K, Qin M, Cui S. Progress and potential of metal-organic frameworks (MOFs) as
renewable regeneration heat source in solid desiccant – vapor compression hybrid
novel desiccants for built environment control: A review. Renew Sustain Energy
cooling system. J Build Eng 2018;18:107–24. https://doi.org/10.1016/j.
Rev 2020;133:110246. https://doi.org/10.1016/j.rser.2020.110246.
jobe.2018.03.012.
[12] Prabakaran V, Chua KJ. Current state-of-the-art in desiccant dehumidifiers. Adv
[40] Jani DB, Mishra M, Sahoo PK. Performance studies of hybrid solid desiccant-vapor
Desiccant Dehumidification 2021:23–48. https://doi.org/10.1007/978-3-030-
compression air-conditioning system for hot and humid climates. Energy Build
80843-3_2.
2015;102:284–92. https://doi.org/10.1016/j.enbuild.2015.05.055.
[13] Vivekh P, Bui DT, Chua KJ. Experimental performance investigation of
[41] Coleman HW, Steele WG. Uncertainty Analysis for Uncertainty Analysis for
superabsorbent polymer and potassium formate coated heat exchangers. Energy,
Engineers. 2009.
Västerås, Sweden: Int. Conf. Appl; 2019.
[42] Vivekh P, Bui DT, Kumja M, Islam MR, Chua KJ. Theoretical performance analysis
[14] Vivekh P, Bui DT, Wong Y, Kumja M, Chua KJ. Performance evaluation of PVA-LiCl
of silica gel and composite polymer desiccant coated heat exchangers based on a
coated heat exchangers for next-generation of energy-efficient dehumidification.
CFD approach. Energy Convers Manag 2019;187:423–46.
Appl Energy 2019;237:733–50. https://doi.org/10.1016/j.apenergy.2019.01.018.
[43] Jagirdar M, Lee PS. Mathematical modeling and performance evaluation of a
[15] Vivekh P, Bui DT, Islam MR, Zaw K, Chua KJ. Experimental performance and
desiccant coated fin-tube heat exchanger. Appl Energy 2018;212:401–15. https://
energy efficiency investigation of composite superabsorbent polymer and
doi.org/10.1016/j.apenergy.2017.12.038.
potassium formate coated heat exchangers. Appl Energy 2020;275:115428.
[44] Li Z, Michiyuki S, Takeshi F. Experimental study on heat and mass transfer
https://doi.org/10.1016/j.apenergy.2020.115428.
characteristics for a desiccant-coated fin-tube heat exchanger. Int J Heat Mass
[16] Tu YD, Wang RZ, Ge TS, Zheng X. Comfortable, high-efficiency heat pump with
Transf 2015;89:641–51. https://doi.org/10.1016/j.
desiccant-coated, water-sorbing heat exchangers. Sci Rep 2017;7(1).
ijheatmasstransfer.2015.05.095.
[17] Kumar A, Yadav A. Experimental investigation of solar driven desiccant air
[45] Tu YD, Wang RZ, Hua LJ, Ge TS, Cao BY. Desiccant-coated water-sorbing heat
conditioning system based on silica gel coated heat exchanger. Int J Refrig 2016;
exchanger: Weakly-coupled heat and mass transfer. Int J Heat Mass Transf 2017;
69:51–63. https://doi.org/10.1016/j.ijrefrig.2016.05.008.
113:22–31. https://doi.org/10.1016/j.ijheatmasstransfer.2017.05.047.
[18] Jani DB, Mishra M, Sahoo PK. Solid desiccant air conditioning - A state of the art
[46] MacArthur JW, Grald EW. Unsteady compressible two-phase flow model for
review. Renew Sustain Energy Rev 2016;60:1451–69. https://doi.org/10.1016/j.
predicting cyclic heat pump performance and a comparison with experimental
rser.2016.03.031.
data. Int J Refrig 1989;12:29–41. https://doi.org/10.1016/0140-7007(89)90009-
[19] Ge TS, Dai YJ, Wang RZ. Performance study of desiccant coated heat exchanger air
1.
conditioning system in winter. Energy Convers Manag 2016;123:559–68. https://
[47] Koury RNN, MacHado L, Ismail KAR. Numerical simulation of a variable speed
doi.org/10.1016/j.enconman.2016.06.075.
refrigeration system. Int J Refrig 2001;24:192–200. https://doi.org/10.1016/
[20] Zu K, Qin M. Experimental investigation of a novel metal-organic framework
S0140-7007(00)00014-1.
(MOF) based humidity pump under high humidity conditions. J Phys Conf Ser
[48] Bell IH, Wronski J, Quoilin S, Lemort V. Pure and pseudo-pure fluid
2021;2069(1):012049.
thermophysical property evaluation and the open-source thermophysical property
[21] Na SI, Chung Y, Kim MS. Performance analysis of an electric vehicle heat pump
library coolprop. Ind Eng Chem Res 2014;53:2498–508. https://doi.org/10.1021/
system with a desiccant dehumidifier. Energy Convers Manag 2021;236:114083.
ie4033999.
https://doi.org/10.1016/j.enconman.2021.114083.
[49] Wang C-C, Chi K-Y. Heat transfer and friction characteristics of plain fin-and-tube
[22] Karmakar A, Prabakaran V, Zhao D, Jon K. A review of metal-organic frameworks
heat exchangers, part I: New experimental data. Int J Heat Mass Transf 2000;43
(MOFs) as energy-efficient desiccants for adsorption driven heat-transformation
(15):2681–91.
applications. Appl Energy 2020;269:115070. https://doi.org/10.1016/j.
apenergy.2020.115070.
[23] Tu YD, Wang RZ, Ge TS, Zheng X. Comfortable, high-efficiency heat pump with
desiccant-coated, water-sorbing heat exchangers. Sci Rep 2017;7:40437. https://
doi.org/10.1038/srep40437.
[24] Cui S, Qin M, Marandi A, Steggles V, Wang S, Feng X, et al. Metal-Organic
Frameworks as advanced moisture sorbents for energy-efficient high temperature
cooling. Sci Rep 2018;8(1). https://doi.org/10.1038/s41598-018-33704-4.

18
V. Prabakaran et al. Energy Conversion and Management 266 (2022) 115867

[50] Wang C-C, Chi K-Y, Chang C-J. Heat transfer and friction characteristics of plain through nucleation and impaction scavenging by growing cloud drops. J Atmos Sci
fin-and-tube heat exchangers, part II: Correlation. Int J Heat Mass Transf 2000;43 1985. https://doi.org/10.1175/1520-0469(1985)042<0583:ATSOTW>2.0.CO;2.
(15):2693–700. [52] Incropera FP, DeWitt DP, Bergman TL, Lavine AS. Principles of heat and mass
[51] Flossmann AI, Hall WD, Pruppacher HR. A theoretical study of the wet removal of transfer 2013. https://doi.org/10.1016/j.applthermaleng.2011.03.022.
atmospheric pollutants. Part I: the redistribution of aerosol particles captured

19