Journal of Energy Storage 24 (2019) 100797

Contents lists available at ScienceDirect

Journal of Energy Storage

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

Heat transfer enhancement of charging and discharging of phase change T

materials and size optimization of a latent thermal energy storage system for
solar cold storage application
Shaon Talukdar, Hasan Mohammad Mostafa Afroz, Md. Anowar Hossain, M.A. Aziz,
Md. Monir Hossain
Department of Mechanical Engineering, Dhaka University of Engineering & Technology, Gazipur, Bangladesh

A R T I C LE I N FO A B S T R A C T

Keywords: In this work, phase change material (PCM) is considered as thermal energy backup system for solar cold storage
Cold storage applications when there is peak power demand or power failure or no sun shines situations. A numerical study of
Latent heat thermal energy storage (LHTES) solidification (charging) and melting (discharging) of PCM validated by experimental data is performed to ex-
Phase change material (PCM) plore the performance of a unique latent heat thermal energy storage (LHTES) system. The LHTES unit (PCM
Charging and discharging
pack) occupied with PCM acts as a heat exchanger made up from evaporator tube along with rectangular metal
Heat transfer enhancement
Numerical modeling
fins which enhance the heat transfer during phase changing of PCM. Evaporator with 5, 8, 10, and 12 long-
itudinal aluminum fins, and without fin inside the PCM pack of six different thicknesses (4.5 cm, 5.0 cm 5.5 cm,
6.0 cm, 6.5 cm, and 7.0 cm) are considered to investigate the charging and discharging characteristics of PCM in
our present study. To obtain the best performance from the PCM pack, heat transfer characteristics and tem-
perature distributions of PCM, and the effect of variation of thickness of the PCM pack are studied through
computational fluid dynamics (CFD) simulation. Solidification and melting model of ANSYS Fluent 15.0 package
employing enthalpy-porosity technique is used to develop 3-D simulation models. Faster solidification, as well as
a higher energy storage capacity and heat flux during melting is found for the PCM pack of 6.5 cm thickness with
higher number of fins. A lab scale experimental set-up is fabricated to compare the present simulation results
which shows reasonably well validation during both the charging and discharging.

1. Introduction Concerning that issues, cold storage using solar energy along with a
LHTES system could be an alternative of minimizing the paucity of
The way out of energy crisis throughout the world can only be the efficacy of the existing technologies offering almost zero running cost.
proper utilization of the sources of energy available, especially the re- In this technology, the refrigeration system of the cold storage would be
newable sources. In addition, world economy is experiencing a huge energized and operated by solar energy during the solar hours. In this
loss due to lack of proper post-harvest technologies of fruits, vegetables, regard, the key question is how the operating conditions, temperature
crops, and so like. Most of the existing cold storages are less efficient and humidity, of the cooling space of a cold storage will be maintained
and running at a higher running cost. The technology of using solar during the nighttime or the time when sunshine is not available at
energy to excite a refrigeration system in a cold storage for reducing its daytime. While the conventional battery back-up (electrochemical en-
running cost is just in its beginning. To our understanding, Ming Liu ergy storage) system is widely used in solar powered devices, it ac-
[1], one of the earliest researchers of solar operated cold chain tech- counts some difficulties including damping, shorter lifetime, and energy
nologies, developed a solar powered mobile refrigeration system. PCM, loss during charging [4]. In our present work, a LHTES system is opted
a substance which melts and solidifies at a certain temperature and is as a thermal back-up that facilitates the phase changing of PCMs by
capable of charging and discharging a large amount of latent heat per charging (solidification) and discharging (melting) processes. During
unit mass, is used as a thermal back-up of the designed refrigeration the back-up time, PCMs ensure a desired specific temperature inside the
system. The outline of recent developments of this technology is de- cold storage to maintain quality of goods and commodities by absorbing
monstrated by several works [2,3]. The efficient technology of using the influx of heat from outside as well as the generation of heat from all
solar energy in cold chain applications, however, is yet to be developed. possible sources inside the cold space. Heat exchangers with

E-mail address: [email protected] (S. Talukdar).

https://doi.org/10.1016/j.est.2019.100797
Received 9 December 2018; Received in revised form 3 May 2019; Accepted 31 May 2019
2352-152X/ © 2019 Elsevier Ltd. All rights reserved.
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Nomenclature x, y, z Space coordinates [x, y, z] = m

Latin Greek

A Surface area [A] = m2 α Heat transfer coefficient, [α] = W/(m2K)

Amush Mushy zone parameter, [Amush] = (Pa-s)/m2 β Liquid fraction
Cp Specific heat [Cp] = J kg−1k−1 Є Convective enhancement factor
D Temperature deviation between simulation and experi- ρ Density [ρ] = kg m−3
ment [D] τ Shear stress of fluid [τ] = N m−2
D̄ Standard deviation Φ Diameter [Φ] = cm
g Gravitational constant [g] = ms−2
H Height of the domain [H] = cm Symbols
H Enthalpy [H] = J
h Specific enthalpy [h] = J kg−1 ▽ Vector differential operator [▽] = (∂/∂x, ∂/∂y)
K Thermal conductivity [k] = Wm−1k−1 Δ Finite difference
L Length of the domain [L] = cm
l Latent heat of fusion [l] = J kg−1 Subscripts
m Mass of PCM [m] = kg
N Number of numerical/measured values CFD Numerical
p Pressure [p] = Pa EXPT Experimental
R Radius [R] = cm j Index variable
S Energy source term [S] = W/m3 l Liquid
T Temperature [T] = ºC liquidus Melting
t Time [t] = Sec mush Mushy zone
u Velocity vector [u] = (u, v)T s Solid
u Velocity in x direction [u] = m s−1 solidus Solidification
v Velocity in y direction [v] = m s−1 ref Reference
w Velocity in z direction [w] = m s−1 1,2,…..,9 Thermocouple positions
W Width of the domain [W] = cm

rectangular shaped container, shell and tube heat exchanger, en- potential option compared to the other available heat transfer enhan-
capsulated PCM modules, heat exchanger with flat plate are the cing approaches [15,16]. Taking use of fins in consideration to enhance
common designs of the LHTES system. Several research works on design heat transfer, thermal conductivity and corrosion potential are two key
of the LHTES, PCMs, and its applications have been published [5–11], criterion for the selection of fin materials. Additionally, cost and other
and the research is still going on. Rectangular shaped LHTES accom- thermophysical properties of fin materials, and machinability are also
plishes more attention for cold storage applications due to its flexibility influential parameters [17]. Since the HTF is a refrigerant (R-410A) in
to mount on the wall of the storage, as well as offering more heat our present investigation and the refrigerant will change its phase in the
transfer surface area to conduct heat receiving or releasing on the basis cooling coil (evaporator), a nearly constant temperature on the tube
whether it is being used for cooling or heating [12]. They also illu- wall can easily be achievable. Researchers studied the effect of: fins
strated that the melting time of PCM inside a cylindrical PCM container orientation [18,19], fin spacing [20], fin geometry [21] on the charging
is almost twice that of a rectangular one of equal volume for an amount and/or discharging process of PCMs. Sparrow et al [22] investigated the
of PCM greater than 2 kg. According to the demand of the aforesaid charging of PCM in a vertical aligned finned tube showing that con-
application, a flat plate LHTES system filled by PCM along with cooling duction is the dominating medium when phase change occurs vis a vis
coils (evaporator) inside the rectangular chamber is chosen. As the li- natural convection is the driving heat transfer phenomena when the
quid PCM releases heat with the aid of a refrigeration circuit, it be- liquid temperature is greater than the solidification temperature. Pad-
comes solid and stores cold energy; this process is called charging. mabhan and Krisna [23] conducted a numerical investigation of an
Conversely, solid PCM absorbs heat from the surrounding of cold sto- LHTES system where rectangular fins are attached with the surface of
rage during the back-up period, which results melting of PCM, and it isothermal tube. Other works on heat transfer enhancement using metal
releases cold energy; this phenomenon is called discharging. For solar fins in cylindrical LHTES systems are demonstrated [24–28]. While a
cold storage application, PCM is charged by a vapor compression re- few researchers [29–32] concentrated on using metal fins to increase
frigeration (VCR) unit using available solar energy during day time. As heat transfer in rectangular thermal energy storage (TES), a research
the availability of solar hours fluctuates throughout the year, it is ob- focusing both the experimental as well as numerical investigation of
vious to complete the charging process of PCM within the lowest heat transfer enhancement of a rectangular TES for both charging and
available solar hours. Focusing on that issue, enhancement of heat discharging is still absent. In addition, the effect of size of the PCM pack
transfers between PCM and the refrigerant is considered as a vital on heat transfer performance during phase changing is deemed to a
concern because faster heat transfer can significantly mitigate the crucial issue due to a significant fluctuation of thermal conductivity of
charging period of PCM. Researchers investigated the various ways of PCM when it changes its phase. Hoseinzadeh et al. [32] developed a
improving the performance of a system involving heat transfer. As il- numerical model depicting the effect of length and thickness of PCM
lustration, the operating temperature of heat transfer fluid (HTF) plays slab on melting process of multiple PCMs. Vakilaltojjar et al. [33]
an important role, and it improves the efficiency of a heat exchanger presented a semi-analytical model to evaluate the appearance of the
when the temperature is nearly constant [13]. In addition, the use of storage systems in air conditioning and studied the impact of slab
highly conductive fins is a possible solution to boost heat transfer thickness on storage. Therefore, the present research is an attempt to
performance of a thermal energy storage unit [14]. Lower cost, sim- resolve the limitations of the forging research on rectangle shaped
plicity of fabrication, and easy maintenance of fins gain attention as a LHTES systems. The current work investigates both the solidification

2
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

and melting of PCM experimentally as well as numerically. For optimal Taker DT80) to record the temperatures at a 30-seconds interval over
heat exchanging performance, the effect of number, material and geo- the charging and discharging period. Four thermocouples are posi-
metry of fin are the key parameters, and these are intensely analyzed tioned in the refrigeration circuit to capture the refrigerant tempera-
during both the phase changing processes of PCM. Furthermore, the tures within the accuracy of ± 0.5 °C at the inlet and exit of the com-
effect of the thickness of a rectangular PCM pack is also investigated for pressor as well as the expansion device as shown in the Fig. 1. Thirteen
the size optimization of the designed TES system. of other thermocouples are placed in two different planes (horizontal
and vertical) of PCM chamber as shown in the Fig. 5. Precise posi-
2. Experimental setup tioning of the thermocouples inside the PCM pack is one of the chal-
lenging tasks due to the absence of additional mounting facilities of
The experimental setup consists of two main components namely thermocouples, which oftentimes results a significant disturbance on
PCM pack and condensing unit. PCM pack is the thermal energy storage the flow and heat transfer inside the storage. Additionally, a shift of
unit, which is rectangular shaped and filled by PCM and cooling coil of thermocouple’s tip is also noticed because of the deformation of ther-
VCR circuit. Another major component is the condensing unit, which mocouple during phase changing of storage material. The uncertainty
generally consists of a compressor, a condenser, and an expansion of temperature measurement because of positioning error is found ±
valve. Figs. 1 and 2 show the schematic diagram and the photographic 1 °C approximately. Adding this value with a statistical error of ± 0.5
view of the test loop for our present experimental investigation. °C obtained from a set of experimental test runs for each position, the
overall approximation of uncertainty is estimated as ± 1.5 °C. Finally, a
2.1. Storage system thermocouple is used to measure the ambient temperature throughout
the whole cycle Table 2
Concerning the prime purpose that is to uniformly solidify (charge)
and melt (discharge) the PCM at a satisfactory rate, the PCM pack is 3. Numerical modelling
designed with dimensions as shown in Fig. 3 and made of stainless steel
sheet. 3.1. Assumptions
Evaporator is the heat exchanger which exchanges heat between the
PCM and the HTF (refrigerant) during the charging (solidification) of The numerical model is based on the following assumptions:
PCM. The evaporator is made of copper tubes of inner and outer dia-
meters are 12 mm and 12.7 mm, respectively. The evaporator includes 1 The liquid phase of PCM is considered as a Newtonian as well as an
20 straight tubes which are connected by copper U-bend as shown in incompressible fluid.
the Fig. 4(a). In case of finned evaporator, aluminum fins are placed 2 The resulting flow of the liquid PCM inside the PCM pack is assumed
along the width of the tube arrangement as shown in Fig. 4(c). laminar and incompressible.
3 The change in volume of PCM due to phase change is neglected.
2.2. Selection and characterization of PCM 4 There is no movement of PCM when it is fully solidified.
5 The thickness of the rectangular storage is considered to be large
The selection of PCM for cold storage applications significantly enough so that the effect of boundary wall can be neglected.
depends on the temperature inside the cooling space of cold storage 6 The outer walls of PCM pack are taken into account as adiabatic
that has to be maintained. In this work, the required temperature of walls during charging.
cooling zone is considered in a range of 2 °C–5 °C. Thus, water is con- 7 The solid-liquid interface during phase change process of PCM is
sidered as best suitable PCM because of its availability, high heat of defined as a region called mushy zone where the PCM is assumed as
fusion, and a sharp phase change temperature which is a great chal- a mixture of both solid and liquid.
lenge in case of mixture of salt solutions. Besides, it is chemically stable, 8 The effect of radiation as well as viscous dissipation is neglected.
nonpoisonous, and noncorrosive over a large storage period. Table 1
depicts the thermophysical properties of water and ice. 3.2. Physical model

2.3. Experimental measurement and uncertainties To model the charging and discharging of PCM, a method, known as
enthalpy-porosity technique, is used, where the liquid fraction of a cell
To observe the phase changing phenomenon critically, eighteen volume is calculated. The domain can be regarded as a porous media
thermocouples (K-type) are placed at the pre-defined positions in the where the porosity β = 1 in the liquid phase, β = 0 in the solid phase,
system, and the thermocouples are interfaced with a data logger (data and 0 < β < 1 in the mushy zone. The theory in this section is based

Fig. 1. Schematic diagram of PCM circuit for experimental investigation.

3
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 2. Photographic view of the experimental set-up.

on the article by Voller et al. [34], but is written in the form given by solid and liquid form, u is the velocity of solid-liquid interface, and S is
ANSYS FLUENT Theory Guide [35]. a source term. This source term, given in Eq. (7), is a momentum sink
The enthalpy, H of PCM can be expressed as due to reduction of porosity in the mushy zone and written in the fol-
(1) lowing form
H = h + ΔH
(1 − β )2
where h and ΔH are the sensible and latent heat respectively that are S= Amush (u − up)
absorbed or released by PCM. (β 3 + ε ) (8)
The sensible heat, h can be expressed as where ε is a very small number to prevent division by zero, Amush is the
T mushy constant, and u is the velocity due to the pulling of solidified
h = href + ∫ Cp dT material out of the domain.
Tref (2)
3.3. Geometry specifications and discretization with grid independence
where href is the reference enthalpy, Tref is the reference temperature,
and Cp is the specific heat at constant pressure.
Numerous studies are carried out with a 2D domain, but in present
The liquid fraction, β, can be defined as
study, it is not feasible to analyze a 2D domain, because in a 2D domain,
β = 0 if T < Tsolidus, (3) the phase changing phenomena of PCM cannot be critically observed in
all directions. Therefore, a 3D model of the PCM pack is designed using
β = 1 if T > Tliquidus, (4)
SolidWorks as shown in Fig. 6. As a 3D simulation demands enormous
T − Tsolidus computing resources [17], it is more convenient to choose a segment of
β= if Tsolidus < T < Tliquidus
Tliquidus − Tsolidus (5) the model in lieu of the whole geometry as the computational domain.
This will reduce the number of control volumes as well as the compu-
The latent heat content can be written in terms of the latent heat of tational time; the selected domain, however, must capture the entire
the material heat exchanging phenomenon that are supposed to be taken place in the
ΔH = βl (6) whole PCM pack. Since the evaporator tubes are equally spaced with
each other in the PCM pack, the heat transfer between PCM and cooling
which can vary between zero (0) for a solid and a value l for a liquid. coil will vary up to the mid surface of the spacing between two con-
The energy equation for charging (solidification) and discharging secutive tubes. Fig. 6 depicts the segment of the PCM pack which is used
(melting) phenomena is written as as the computational domain.
A pressure based implicit scheme finite volume discretization
∂ (ρH )
+ ∇ . (ρ . u. H ) = ∇ . (K . ∇T ) + S method is used for meshing the domain. A mesh sensitivity analysis is
∂t (7)
performed where the solidification of a 3D model with volume of
where H is the enthalpy as given in Eq. (1), ρ is the density for both 4093.7 cm3 is observed by varying the number of elements from

Fig. 3. Illustration of the geometry of the PCM pack and the evaporator.

4
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 4. (a) Schematic diagram of arrangement of evaporator tube and configuration of evaporator- (b) without fin (c) with fin.

Table 1 Table 2
Thermophysical properties of PCM. Boundary conditions for numerical modelling of solidification and melting.
Properties Water Ice Boundary Charging (Solidification) Discharging
(Melting)
Melting Point in °C 0 0
Freezing point in °C 0 0 Adiabatic wall Heat flux = 0 Heat flux = 0
Latent heat of fusion in kJ/Kg 335 335 Cold walls Temperature, T(°C) = Heat flux = 0
−5 2
Thermal conductivity in W/m-k 0.6 2.2 −3.63-0.00728t-9.11 × 10 t
Specific Heat in J/Kg-k 4186 2060 Where, t = Solidification Time in minute.
Density in Kg/m3 998 915 Hot walls Temperature, T(°C)= 28.103 + 0.039 t-0.0004t2+9.11 × 10−6 t3
Where, t = Phase Changing Time in minute.

241218 to 1605612 to optimize the mesh size. The obtained data are
plotted into a graph shown in Fig. 7 to narrow down the number of these walls, defined as hot walls in the domain, is added during both the
elements for the optimum grid size. Fig. 7 demonstrates a modicum of phase changing period. The conditions for different boundaries of the
variation in the simulation results when the number of mesh elements numerical model for charging and discharging of PCM are listed in the
exceeds 769065. Therefore, the optimum grid size is selected corre- table below.
sponding to the number, and the CFD analysis is done at the optimized Transient studies (time step =0.05 s) were done for the melting/
grid size. solidification model using the solution of the momentum and energy
equations. In solidification model, an initial condition, the computa-
tional domain is filled with liquid PCM at 28 °C, is patched up with the
4. Validation
solver. The final condition of the model is defined as the solid PCM of
temperature −8 °C. Since the melting starts just after the solidification,
The accuracy and flexibility numerical approach is evaluated by
hence the final condition of solidification model is considered as the
comparing the simulation results to the that of experimental ones.
initial condition of the melting model. A double precision pressure
Firstly, the flow, and the thermal variables on the boundaries of the
based solver with under relaxation factor (URF) 0.8 for energy and
physical model are specified, and then the conditions are applied on
default values for others is applied in the present simulation. The re-
these boundaries. As the PCM is stored in the PCM pack, and there is no
sidual value for solution convergence criteria for energy equation is set
in and out movement of PCM during charging/discharging, hence no
as 10−10 and default values for all other parameters.
inlet or outlet boundary conditions is required for PCM. Fig. 8 illus-
trates a representation of different boundaries of the present model.
As the outer walls of PCM pack are in contact with the ambient air,
and a temperature gradient exists among them, a heat influx through

Fig. 5. Positioning of thermocouples at horizontal and vertical mid plane.

5
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 6. Illustration of the 3D and 2D views of PCM pack and the segment of the PCM pack which is used as the computational domain for the numerical analysis.

|Tj, CFD − Tj, EXPT |

Deviation, Dj = × 100%
|Tj, EXPT | (9)

j=N
¯ = 1
and Standard Deviation, D ∑ Dj
N j=1 (10)

The quantitative study of error analysis results small values of

percent of deviations in temperature measurement at different locations
during charging and discharging phenomenon. These small deviations
affirm the plausibility of the CFD model with adequate accuracy so that
the model can be useful for further scaling up/down the storage system.
The maximum values of deviation for TV3 and TH2 are found noticeable
during charging due to very frequent transient behavior of the phase
changing heat transfer at the beginning of solidification. These two
thermocouples (TV3 and TH2) experienced the effect of unstable heat
transfer behavior to a considerable extent because of their compara-
tively closed placement with respect to the heat sink (evaporator tube)
of the storage system.

5. Results and discussions

Fig. 7. Variation of simulation result of volume of liquid for various number of
mesh elements. Initially, the entire domain is a liquid PCM, and the volume of
5.1. Effect of fin for the enhancement of heat transfer of charging and
liquid PCM after a 10 min of charging time is determined for different number
of mesh elements. discharging

Heat transfer enhancement during charging and discharging of PCM

4.1. Comparison of simulation results with experimental data
using metal fin is numerically analyzed. Contours of liquid fraction of
PCM are plotted at different moments of phase changing for the fol-
Fig. 9 compares the CFD simulation results and experimental data
lowing heat exchanging conditions.
showing the variation of temperature of PCM at different locations
(TV2, TV3, TH2, and TH3) in the storage unit during both the charging
• Case I: Evaporator tubes without fin
(solidification) and discharging (melting).
• Case II: Evaporator tubes with 05 (five) identical longitudinal fins
As the initial temperature of PCM is taken about 28 °C, a significant
• Case III: Evaporator tubes with 08 (eight) identical longitudinal fins
amount of sensible heat is required to be released by the PCM to reduce
• Case IV: Evaporator tubes with 10 (ten) identical longitudinal fins
the temperature to its freezing point. During the period of sensible
cooling, the heat capacity of PCM varies with temperature, however, • Case V: Evaporator tubes with 12 (twelve) identical longitudinal fins
this effect is not taken into account in current CFD model.
The geometry, orientation of fin, and other parameters of the
Consequently, CFD model results a faster heat transfer rate during
aforementioned cases are represented in the Fig. 10 and Table 4. Size of
sensible cooling than that of experimental investigation. An error the storage system, the boundary conditions and the initial and final
analysis showing the quantitative measure of deviation of numerical
conditions of these cases are kept similar for making a performance
results from the experimental data is carried out to measure the un- comparison. Acting as a heat sink, the temperature at the wall of the
certainties for both the solidification and melting model. The percent of
evaporator tube is maintained at -10 °C for all the cases while the outer
deviation, Dj and standard deviation, D̄ is defined using the Eqs. (9) and
surfaces of PCM pack are considered at a temperature of 5 °C. At the
(10) where Tj, CFD and Tj, EXPT are measured for the same locations of the
starting of the charging, the PCM is considered in liquid phase with a
storage system for N number of times. The results of the error analysis
value of liquid fraction 1.00 at a temperature of 5 °C. On the other hand,
are summarized in Table 3.
in case of discharging, thermal energy will be absorbed by solid PCM

6
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 8. Demonstration of the computational domain of the LHTES system showing different zones and boundaries. A horizontally cross sectional view is cut across the
mid portion of the domain; the 2D view captures a detail depiction of all the segments involved with the model.

through the outer surfaces of the PCM pack to meet the cooling load of Table 3
cold zone of the cold storage. Since the temperature inside the cold Summary of deviations of CFD results from experiments.
storage should be kept within 5 °C, the outer surfaces of the PCM pack % of Deviation Solidification Melting
are needed to be set at a temperature equal to 5 °C during discharging.
Thus, in melting model, the LHTES system is initially occupied by solid TV2 TV3 TH2 TH3 TV2 TH2
PCM of -10 °C temperature with a liquid fraction of 0.00. Due to the
Maximum 1.42 3.92 3.87 0.92 1.19 1.19
insertion of metal fins in the PCM pack, a decline in volume of PCM is Minimum 0.07 0.05 0.04 0.01 0.01 0.01
observed in the finned storage systems. The volumes of PCM are Standard 0.96 1.72 1.84 0.49 0.26 0.29
4093cm3, 4063 cm3, 4045 cm3, 4033 cm3 and 4021 cm3 for the cases I
to V, respectively.
Fig. 11 depicts the contours of liquid fraction of PCM for different the effect is minimum. It is obvious that the plane located at the middle
cases at vertical mid plane of the PCM pack from the starting to 130 min in between two consecutive metal fins is the plane where the effect of
of charging. Initially, the fluid domain is fully occupied by the liquid fin on heat transfer of PCM is minimum. From the geometry of com-
PCM, therefore, the value of liquid fraction is one (1.00), and this value putational domain, the vertical plane for the cases II to V can be taken
decreases as the charging time increases and becomes zero (0.00) when at a distance of 59.18 cm 50.01 cm, 49.62 cm and 42.74, respectively
the PCM is completely solidified. As the effect of fin on heat transfer from the front end of the PCM pack as shown in Fig. 10. The contours
enhancement of PCM varies with distance between the fin and PCM, the shown in Fig. 11 illustrate a comparison of rate of phase change from
vertical plane for the cases II to V should be taken at a distance where liquid to solid phase of PCM for the five different aforesaid evaporator

Fig. 9. Temperature of PCM at different locations in the LHTES during the (a) charging (solidification) and (b) discharging (melting) process. Temperatures are
recorded experimentally at some predefined locations in the LHTES to compare with the temperatures at the same locations that calculated numerically.

7
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 10. Geometry of the computational domain with (a) 0 fin, (b) 5 fin, (c) 8 fin, (d) 10 fin, and (e) 12 fin.

Table 4
PCM Pack, fin, and evaporator tube characteristics for all the examined test cases for numerical study of heat transfer enhancement.
Case PCM pack parameters Fin parameters Evaporator tube Parameters

Size Material Number Material Thickness Spacing Material Dia. Length

(L × W×H) (cm) (cm) (cm) (cm)
(cm3)

I 100 × 7.9 × 5.5 Stainless Steel 0 Aluminum 0.15 – Copper 1.27 198.38
II 5 19.70
III 8 11.45
IV 10 8.82
V 12 7.29

Fig. 11. Contours of the liquid fraction of PCM at five different moments of the charging (solidification) process showing the heat transfer enhancement due to metal
fin insertions in the LHTES. The contours are taken at vertical mid plane defined in previous section.

configurations. It has been found that the value of liquid fraction at 120 min and150 min the liquid fraction became almost one (1.00) for
vertical plane becomes zero (00) 10 min earlier when the number of fins the cases IV and V, respectively.
increase from 10 to 12. Fig. 13 shows a comparison of rate of change of liquid fraction
In discharging process, the outer faces of PCM pack act as the heat during solidification as well as melting of PCM among the heat ex-
source to melt the solid PCM. Fig. 12 depicts the contours of liquid changing configurations considered in our analysis. It has been shown
fraction of PCM which increases as the melting time increases. As the that rate of change of liquid fraction increases when number of fins is
source of heat is not in direct contact with the aluminum fins, the en- increased for same thickness of PCM pack, and the time required to
hancement of heat transfer for melting is lower than that of solidifi- completely solidify the PCM is plummeted by 170 min and 190 min in
cation. However, it is clear from the above contours that the rate of case of evaporator with 10 and 12 identical aluminum fins, respectively
melting is faster when the number of fin increases. As a result, after compared to the evaporator arrangement without fin. Like charging,

8
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 12. Contours of the liquid fraction of PCM at four different moments of the discharging (melting) process showing the heat transfer enhancement due to the
insertion of metal fins in the LHTES. The contours are taken at vertical mid plane defined in previous section.

Fig. 13. A quantitative comparison of heat

transfer enhancement showing the rate of
change of the value of liquid fraction during
(a) charging (solidification) and (b) dischar-
ging (melting) of PCM. Heat transfer en-
hancement ratio is defined as the ratio of heat
transfer due to the presence of enhancing
medium to that of an absence of that medium.

the discharging time is also decreased by 120 min and 160 min for the Liquid fraction of PCM at different instants of charging and dis-
cases IV and V, respectively compared to case I. charging is depicted by the contours and line graphs shown in Figs. 14
and 15 for six different cases. Figs. 14 and 15 demonstrate that the time
required to solidify and melt the whole PCM pack significantly rises as
5.2. Effect of variation of thickness of PCM pack during charging and the thickness of PCM pack increases. This is because of higher thermal
discharging energy storage capacity compared to others of lower thickness. In our
current application, when the charging of PCM is done by solar energy,
In this section, the amount of PCM to be solidified within an al- it is essential to freeze optimum amount of PCM within the available
lowable time is optimized by varying the thickness of the storage solar hour. Thus, in case of charging, the focal aim of thickness varia-
system while the other dimensions are unchanged. Thus, the size and tion is to find out how the solidification rate changes due to change in
orientation of tubes, and operating conditions of evaporator are kept storage thickness which will assist to optimize the size of the PCM pack
same for all the considered cases. The optimize size of the PCM pack resulting a reduction in total number of PCM pack. Fig. 15 indicates
will buttress the back-up system reducing the total number of PCM pack that the time required for complete solidification and melting of PCM
to meet the cooling demand of cold storage. In addition, heat receiving for a PCM pack of thickness 7.0 cm is maximum compared to other
rate of the PCM pack during discharging is a key parameter to satisfy cases. Since a larger discharging time results a longer back-up period,
the demand of the cold storage during back-up period. Therefore, the the larger melting time of PCM pack is desirable. However, the charging
variation heat flux through the outer surfaces of storage is also ana- time should be in an allowable range so that the charging can be done
lyzed. Enhancement of heat transfer for phase changing of PCM using in a minimum solar hour. It requires approximately four (04) hours to
metal fin is investigated in the earlier section and found that as the charge the PCM pack considered in case F as shown in Fig. 15. The large
number of metal fins increases the phase changing time decreased due charging time required for the aforementioned case may result an in-
to an intense increase of heat conduction rate inside the PCM. In this complete charging of PCM due to the lower value of average sun hours
section, the configuration of evaporator with 10 internal fins is chosen during July and August in Bangladesh [36]. Moreover, it is crucial to
among the alternatives because of its significant performance during confirm that the PCM pack provides a desirable heat flux, the amount of
both charging and discharging. The different cases studied in the pre- heat absorbed or released per unit area, to maintain a desired
sent section are:

9
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 14. Contours of the liquid fraction of PCM at vertical mid plane of LHTES at different moments of the (a) charging and (b) discharging processes showing the
effect of variation of storage thickness.

10
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 15. A quantitative comparison of the rate of change of the value of liquid fraction during solidification (charging) and melting (discharging) of PCM when
storage thickness is different.

temperature in the cooling space. In this regard, it is indispensable to

investigate the behavior of heat flux of the PCM pack during the back-
up.

5.3. Comparison of change of heat flux with time for melting

Fig. 16 compares the fluctuation of heat flux with melting of PCM at

different conditions. The heat flux is the heat received by PCM during
melting through per unit area of upper surface of PCM pack. Due to a
high temperature difference for all cases between the heat source and
sink, the heat flux has a maximum value (approximately 2350 W/m2) at
the starting of discharging as illustrated in Fig. 16. However, the further
continuation of discharging shows a decrease of heat flux at a different
rate for each case. From the contours of liquid fraction during melting
(Figs. 12 and 14(b)), it is clear that the melting first occurs at the upper
surface, and then the solid-liquid interface moves towards the inner
surfaces of PCM pack. The increase in liquid PCM’s layer thickness on
the outer sides of PCM pack results a reduction of heat transfer because
of liquid PCM having lower thermal conductivity compared to solid
Fig. 16. Heat flux vs time graph showing effect of variation of thickness of PCM
PCM. PCM pack of 6.5 cm thickness with an evaporator configuration of
pack and number of fin used during melting.
10 longitudinal fins shows a higher heat flux throughout the dischar-
ging period compared to other cases. The temperature difference be-
tween the outer surface and the any other surfaces of PCM pack plays
significant role to result a better heat conduction. This reason is ex-
plained in the Fig. 17 where the temperature difference is defined as the
difference between the temperature of upper surface of PCM pack and
the temperature at a vertical distance of 2 cm from the upper surface for
all three cases. The temperature differences are taken from 60 min to
90 min of melting period. The graph shows a higher temperature dif-
ference for a thickness of 6.5 cm which results a higher heat transfer
rate which mentioned earlier.

5.4. Summary of heat transfer enhancement and size optimization of the

PCM pack

Heat transfer enhancement using metal fins and the analysis of size
optimization of the PCM pack is summarized by illustrating the energy
storage capacity of PCM pack and average heat transfer rate during
charging and discharging in Fig. 18.
To ascertain the thermal energy storage capacity and the average
heat transfer rate, Eqs. (11)–(13) are employed for the individual cases
Fig. 17. Decreasing rate of temperature difference with time showing the effect considered in this study.
of thickness.
Heat loss by PCM in charging = mCp, l ΔTl + ml + mCp, s ΔTs (11)

Heat gain by PCM in discharging = mCp, s ΔTs + ml + mCp, l ΔTl (12)

Average heat transfer rate during charging/discharging is given by

11
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

Fig. 18. Illustration of thermal energy storage capacity, average heat transfer rate during charging (solidification) and discharging (melting) of PCM pack with six
different sizes showing the effect of insertion of metal fin in the PCM pack.

Heat loss / gain by PCM during charging / discharging as well. Consequently, the heat transfer falls as the thickness increases
Total time required for charging / discharging of PCM (13) for a specific evaporator configurations and operating conditions. On
the contrary of charging, the heat transfer enhancement using metal
In term of energy storage capacity, Fig. 18 illustrates that the sto- fins becomes more prominent during discharging when the size of PCM
rage capacity varies with both the insertion of metal fin as well as pack increases. For discharging, it is seen from the Fig. 18 that com-
change of thickness of the PCM pack. As the metal fins occupy the space pared to an evaporator without fin, an evaporator with 12 aluminum
in the storage unit allocated for PCM, it results a decline in the amount fins enhances heat transfer by 101.72% and 118.33% for PCM pack of
of PCM that can be stored by a PCM pack. Thus, the energy storage thickness of 4.5 cm and 7.0 cm, respectively.
capacity experiences a reduction as the number of internal fins in-
creases. It is found from the Fig. 18 that the reduction is 1.73%
(21.43 kJ) and 1.76% (34.75 kJ) for a PCM pack of 4.5 cm and 7.0 cm, 6. Conclusion
respectively when the number of fins increases from 0 to 12. Although
the insertion of metal fin demonstrates a small variation of energy A numerical investigation is performed by developing a CFD model
storage capacity, the variation is prominent when the size of PCM pack to investigate the solidification and melting behavior of a LHTES system
differs. The amount of energy stored by PCM pack is increased by which is an evaporator tube inside a PCM filled rectangular box. The
47.88% and 59.73% for the PCM pack with 10 aluminum fins of model also predicts the effect of using metal fin to enhance the heat
thickness 6.5 cm and 7.0 cm, respectively compared to the that of a transfer and effect of variation of thickness of PCM chamber to optimize
4.5 cm thick PCM pack. The bar graph also depicts the enhancement of the size of PCM pack. Results predicted by the model were compared to
heat transfer using metal fins with the evaporator during both the experimental data and found that the CFD model predicts the behavior
charging and discharging of PCM. The enhancement trend, heat transfer of PCM during the charging and discharging phenomenon within an
increase as the number of fins increases, is similar for all the cases allowable deviation. Experiments show that the phase change starts at
considered. However, the rate of heat transfer enhancement demon- the cooling coil surface and ends at the outer surface of the PCM pack
strates a variation as the thickness of PCM pack varies for both charging during freezing. On the other hand, the initiation of melting of charged
and discharging. For instance, the heat transfer during charging is PCM occurs at near the outer surfaces of the PCM chamber and ends at
rocketed by almost 165%, which is maximum among the observed the near the middle of the chamber. The CFD model also demonstrates
cases, for a PCM pack of thickness 4.5 cm when the evaporator con- the same behavior of solidification and melting as experimental ana-
figuration changes from without fin to 12 aluminum fins. For the same lysis predicts.
operating conditions, a 7.0 cm thick PCM pack represents 91.8% rise of Minimization of charging time of PCM as well as better heat re-
heat transfer. In case of charging, as discussed earlier, when the ceiving rate during discharging are obtained by incorporating metal fins
thickness increases, the distance between the evaporator tubes and with the evaporator tubes. The numerical analysis of charging and
solid-liquid interface of PCM, which moves vertically from the hor- discharging of PCM are investigated for PCM pack of sizes 100 cm ×
izontal mid plane of the PCM pack to the outer surface, becomes larger 76 cm× (4.5 cm to 7.0 cm) with evaporator arrangement of 0, 5, 8, 10,

12
S. Talukdar, et al. Journal of Energy Storage 24 (2019) 100797

and 12 aluminum fins. It has been found that the heat transfers of developments, Appl. Energy 160 (2015) 286–307.
charging increases as the number of fins increases. Among the cases [11] B. Zivkovic, I. Fujii, An analysis of isothermal phase change of phase change ma-
terial within rectangular and cylindrical containers, Sol. Energy 70 (1) (2001)
considered in this study, the maximum heat transfer (260.77 W) during 51–61.
charging is obtained for the PCM pack of thickness 4.5 cm when the [12] L. Geissbühler, et al., Analysis of industrial-scale high-temperature combined sen-
evaporator is finned by 12 internal aluminum fins. Additionally, the sible/latent thermal energy storage, Appl. Therm. Eng. 101 (2016) 657668.
[13] M.C. Floros, S.S. Narine, Latent heat storage using renewable saturated diesters as
heat transfers for 6.5 cm and 7 cm thick PCM pack are increased from phase change materials, Energy 115 (2016) 924–930.
89.23 W to 182.21 W and from 83.25 W to 154.74 W when the number [14] F. Colella, A. Sciacovelli, V. Verda, Numerical analysis of a medium scale latent
of fin increases from 0 to 12, which significantly decrease the charging energy storage unit for district heating systems, Energy 45 (1) (2012) 397–406.
[15] M.H. Abokersh, et al., Review of the phase change material (PCM) usage for solar
time from 325 min to 165 min and 410 min to 210 min, respectively. domestic water heating systems (SDWHS), Int. J. Energy Res. 42 (2) (2018)
The amount of energy stored by PCM is increased by 11.95%, 23.86%, 329–357.
35.86%, 47.81%, and 59.75% for the PCM pack of thickness 5 cm, [16] M. Liu, W. Saman, F. Bruno, Review on storage materials and thermal performance
enhancement techniques for high temperature phase change thermal storage sys-
5.5 cm, 6 cm, 6.5 cm, and 7 cm, respectively compared to the energy
tems, Renew. Sustain. Energy Rev. 16 (4) (2012) 2118–2132.
that a 4.5 cm thick PCM pack stores. [17] N. Tay, F. Bruno, M. Belusko, Experimental validation of a CFD model for tubes in a
Although PCM pack of size 100 cm × 76 cm×4.5 cm shows the phase change thermal energy storage system, Int. J. Heat Mass Transf. 55 (4) (2012)
fastest freezing rate as the number of fins increase, this configuration 574–585.
[18] Y. Kozak, T. Rozenfeld, G. Ziskind, Close-contact melting in vertical annular en-
indicates a lower energy storage capacity which results an increase of closures with a non-isothermal base: theoretical modeling and application to
number of the PCM pack. On the other hand, although the PCM pack of thermal storage, Int. J. Heat Mass Transf. 72 (2014) 114–127.
7 cm thickness shows the largest energy storage capacity, it’s long [19] J.N. Chiu, V. Martin, Submerged finned heat exchanger latent heat storage design
and its experimental verification, Appl. Energy 93 (2012) 507–516.
charging time (210 min) may result an incomplete solidification of PCM [20] S. Mat, et al., Enhance heat transfer for PCM melting in triplex tube with inter-
during the days when sun hours are minimum as discussed earlier in the nal–external fins, Energy Convers. Manage. 74 (2013) 223–236.
Section 5.2. Concerning that, the 6.5 cm thick PCM pack with 12 alu- [21] E. Sparrow, E. Larson, J. Ramsey, Freezing on a finned tube for either con-
ductioncontrolled or natural-convection-controlled heat transfer, Int. J. Heat Mass
minum fins has a feasible charging time (165 min) with a satisfactory Transf. 24 (2) (1981) 273–284.
energy storage capacity. Thus, among the cases investigated, a PCM [22] P. Padmanabhan, M.K. Murthy, Outward phase change in a cylindrical annulus with
pack of thickness 6.5 cm and an evaporator with 12 aluminum fins can axial fins on the inner tube, Int. J. Heat Mass Transf. 29 (12) (1986) 1855–1868.
[23] M. Lacroix, Study of the heat transfer behavior of a latent heat thermal energy
be a best fit for a longer and necessary backup of a cold storage. storage unit with a finned tube, Int. J. Heat Mass Transf. 36 (8) (1993) 2083–2092.
[24] R. Velraj, et al., Experimental analysis and numerical modelling of inward solidi-
Acknowledgements fication on a finned vertical tube for a latent heat storage unit, Sol. Energy 60 (5)
(1997) 281290.
[25] K. Ismail, C. Alves, M. Modesto, Numerical and experimental study on the solidi-
The authors would like to acknowledge their appreciation and fication of PCM around a vertical axially finned isothermal cylinder, Appl. Therm.
gratitude to Infrastructure Development Company Limited (IDCOL), Eng. 21 (1) (2001) 53–77.
Bangladesh as well as World Bank for their administrative and financial [26] Y. Tao, Y. He, Z. Qu, Numerical study on performance of molten salt phase change
thermal energy storage system with enhanced tubes, Sol. Energy 86 (5) (2012)
support given to the project titled PCM based hybrid mini cold storage. 11551163.
[27] M.K. Rathod, J. Banerjee, Thermal performance enhancement of shell and tube
References Latent Heat Storage Unit using longitudinal fins, Appl. Therm. Eng. 75 (2015)
1084–1092.
[28] V. Shatikian, G. Ziskind, R. Letan, Numerical investigation of a PCM-based heat sink
[1] M. Liu, W. Saman, F. Bruno, Computer simulation with TRNSYS for a mobile re- with internal fins, Int. J. Heat Mass Transf. 48 (17) (2005) 36893706.
frigeration system incorporating a phase change thermal storage unit, Appl. Energy [29] M. Gharebaghi, I. Sezai, Enhancement of heat transfer in latent heat storage mod-
132 (2014) 226–235. ules with internal fins, Numer. Heat Transf. Part A Appl. 53 (7) (2007) 749–765.
[2] U. Stritih, Increasing the efficiency of PV panel with the use of PCM, Renew. Energy [30] S. Hosseinizadeh, F. Tan, S. Moosania, Experimental and numerical studies on
97 (2016) 671–679. performance of PCM-based heat sink with different configurations of internal fins,
[3] P. Sudhakar, G. Kumaresan, R. Velraj, Experimental analysis of solar photovoltaic Appl. Therm. Eng. 31 (17–18) (2011) 3827–3838.
unit integrated with free cool thermal energy storage system, Sol. Energy 158 [31] N. Sharifi, T.L. Bergman, A. Faghri, Enhancement of PCM melting in enclosures
(2017) 837–844. with horizontally-finned internal surfaces, Int. J. Heat Mass Transf. 54 (19–20)
[4] A. Chaurey, S. Deambi, Battery storage for PV power systems: an overview, Renew. (2011) 4182–4192.
Energy 2 (3) (1992) 227–235. [32] S. Hoseinzadeh, et al., Numerical investigation of rectangular thermal energy sto-
[5] F. Agyenim, et al., A review of materials, heat transfer and phase change problem rage units with multiple phase change materials, J. Mol. Liq. 271 (2018) 655–660.
formulation for latent heat thermal energy storage systems (LHTESS), Renew. [33] S.M. Vakilaltojjar, W. Saman, Analysis and modelling of a phase change storage
Sustain. Energy Rev. 14 (2) (2010) 615–628. system for air conditioning applications, Appl. Therm. Eng. 21 (3) (2001) 249–263.
[6] M.M. Kenisarin, High-temperature phase change materials for thermal energy sto- [34] V.R. Voller, C. Prakash, A fixed grid numerical modelling methodology for con-
rage, Renew. Sustain. Energy Rev. 14 (3) (2010) 955–970. vectiondiffusion mushy region phase-change problems, Int. J. Heat Mass Transf. 30
[7] A. Sharma, et al., Review on thermal energy storage with phase change materials (8) (1987) 1709–1719.
and applications, Renew. Sustain. Energy Rev. 13 (2) (2009) 318–345. [35] A. Inc, ANSYS FLUENT theory guide, Release 14 (2012) 5.
[8] H. Mehling, L.F. Cabeza, Heat and Cold Storage With PCM Vol. 308 Springer, 2008. [36] “Average Monthly Hours of Sunshine Over the Year.” World Weather & Climate
[9] J.P. da Cunha, P. Eames, Thermal energy storage for low and medium temperature Information, World Weather & Climate Information, Sunshine, Dhaka, Bangladesh,
applications using phase change materials–a review, Appl. Energy 177 (2016) 2019 (Accessed 29 April 2019), https://weather-and-climate.com/average-
227238. monthly-hours.
[10] B. Xu, P. Li, C. Chan, Application of phase change materials for thermal energy
storage in concentrated solar thermal power plants: a review to recent

13