See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332071841

Investigation of charging and discharging characteristics of a horizontal

conical shell and tube latent thermal energy storage device

Article  in  Energy Conversion and Management · March 2019

DOI: 10.1016/j.enconman.2019.03.022

CITATIONS READS

6 230

3 authors:

Gurpreet SINGH Sodhi Vigneshwaran . K

Indian Institute of Technology Guwahati Indian Institute of Technology Guwahati
7 PUBLICATIONS   15 CITATIONS    9 PUBLICATIONS   20 CITATIONS   

SEE PROFILE SEE PROFILE

P. Muthukumar
Indian Institute of Technology Guwahati
97 PUBLICATIONS   1,241 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The 23rd Conference on Process Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction - PRES'20 View project

PCM Heat Sinks View project

All content following this page was uploaded by Gurpreet SINGH Sodhi on 18 November 2019.

The user has requested enhancement of the downloaded file.

 Energy Conversion and Management 188 (2019) 381–397

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Investigation of charging and discharging characteristics of a horizontal T

conical shell and tube latent thermal energy storage device
Gurpreet Singh Sodhia, Abhishek Kumar Jaiswala, K. Vigneshwaranb, P. Muthukumara,
⁎

a
Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India
b
Centre for Energy, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India

ARTICLE INFO ABSTRACT

Keywords: In this study, a numerical model to analyse the charging and discharging characteristics of a horizontal shell and
Latent Heat Storage tube type Latent Heat Storage (LHS) prototype is presented. The system comprises of Sodium Nitrate as phase
Melting change material (PCM) in the shell side and flow of air as heat transfer fluid (HTF) in the tube side. The effective
Solidification heat capacity approach is followed while solving the fluid flow and energy interactions in the PCM. The design is
Conical shell system
optimized by modifying a cylindrical shell into a conical shell heat exchanger system. The optimized values of
Phase change material
diameters at inlet and outlet of the conical shell are found to be 98.6 mm and 54 mm, respectively having a cone
angle of 3.4°. A performance comparison is made with a cylindrical shell system of equivalent storage capacity.
3-D numerical simulations performed on the proposed system revealed that an innovation in the design leads to
enhanced heat transfer rate caused by uniform melting throughout the system. Further, the conical shell model is
numerically simulated for various operating parameters such as inlet HTF velocity and temperature. Also, the
effect of fins attached on the HTF tube on the performance of the system is analysed. With the advent of ad-
vanced heat exchanger design, there is a greater scope of obtaining higher heat transfer rates employing the
proposed conical shell and tube LHS system.

1. Introduction
In recent times, the global energy consumption trends have shown a
rapid increase due to continuous growth in industrial and domestic
demands. Renewable energy sources, especially solar energy have
drawn attention in the energy sector due to their long term sustain-
ability and ample availability. As solar energy extraction is becoming
one of the most needful utility parameter in industry, thermal energy
storage systems act as a temporary reservoir to store this energy and
assist the application system during demand [1]. Among principal
storage techniques like sensible, latent and thermochemical storage,
Latent Heat Storage (LHS) is an attractive solution to store a large
amount of energy at a nearly constant temperature in the form of latent
heat using phase change materials [2,3]. These systems enable regular
supply of heat during insufficiency of the solar radiation. Phase change
process is an imperative phenomenon involved in the heat storage
concept where the melting and solidification of the PCMs govern the
charging and discharging processes of LHS. PCMs generally have a low
thermal conductivity [4], but their energy density is high, thus en-
couraging their application to cold storage [5], solar drying [6], steam
generation [7,8] and CSP technology [9]. The heat transfer methods
[10], CFD techniques [11] and design optimization of storage system
[12] are vital aspects for the development of efficient LHS systems.
Many researchers have focussed on improving the heat transfer
characteristics of PCM by enhancing the thermal conductivity in the
PCM region [13]. Different composites of organic and inorganic PCMs
using high thermal conductivity materials like graphite, carbon nano-
tubes, metal oxides have been developed and studied [14,15]. Qureshi
et al. [14] reviewed different heat transfer enhancement techniques
such as encapsulated PCMs, adding metallic foams and nano-particle
dispersion inside PCM. Zhao et al. [16] numerically studied the LHS
performance by using graphite-foam as a thermal conductivity en-
hancement additive. Elbahjaoui and Qarnia [17] numerically in-
vestigated the melting behaviour of paraffin wax (P116) dispersed with
Al2O3 nanoparticles in a rectangular LHS unit. They found that the
dispersion of high conductivity nanoparticles in a pure PCM decreases
the melting time and enhance the energy storage rate. The sensible heat
transfer is increased due to a high volume fraction of nanoparticles,
whereas increasing the aspect ratio enhances both sensible and latent
heat transfer. Gasia et al. [18] conducted experiments to compare the
performance of shell and tube heat exchangers (HXs) namely; HX
having no fin, finned HX, HX having metallic wool distributed over fins

⁎
Corresponding author.
E-mail address: pmkumar@iitg.ac.in (P. Muthukumar).

https://doi.org/10.1016/j.enconman.2019.03.022
Received 30 January 2019; Accepted 12 March 2019
Available online 29 March 2019
0196-8904/ © 2019 Elsevier Ltd. All rights reserved.
G.S. Sodhi, et al.
Nomenclature
Amushy Mushy zone constant (kg/m3 s)
cp Specific heat of PCM (J/kg K)
d Diameter of shell at outlet section of cone (m)
do Outer diameter of HTF tube (m)
di Inner diameter of HTF tube (m)
Do Outer shell diameter of cylinder (m)
D Diameter of shell at inlet section of cone (m)
E Energy (J)
f Melt fraction
F Body force term (N/m3)
g Acceleration due to gravity (m/s2)
hph Latent heat of melting of PCM (J/kg)
k Thermal conductivity of PCM (W/m K)
L Length of the system (m)
m mass of PCM (kg)
P Pressure (N/m2)
Q Heat storage capacity (J)
T Temperature (K)
ΔT Temperature difference between HTF inlet and initial
temperature (K)
t Time (s)
V PCM volume (m3)
v Velocity (m/s)
v Velocity vector
and HX having metallic wool distributed arbitrarily. Even though the
HX with only metallic wool could not perform better than finned HX, it
proved as a cheap and effective solution to obtain melting rate en-
hancement of approximately 14% over HX with no fin.
Multi-tube configurations [19], multi-PCM [20] and use of extended
surfaces like fins and heat pipes are some of the heat transfer en-
hancement techniques which have been received significant interest by
various researchers in recent years. Various geometries and configura-
tions of PCM containers such as shell and tube heat exchangers, triplex
tube heat exchangers and configurations with fins such as longitudinal
fins, annular fins, plate fins, pin fins and tree shape fins [21] have also
been emphasized upon.
Niyas et al. [22] developed a lab-scale prototype of multi-tube heat
exchanger based LHS. A study to optimize the number of HTF tubes was
presented and the total charging and discharging capacities of the
system were 16.94 MJ and 15.29 MJ, respectively. Riahi et al. [23]
developed a numerical model of a multi-tube heat exchanger based LHS
for the concentrated solar power (CSP) application. Prasad et al. [20]
conducted a study on the multi-PCM LHS system to compare the
charging and discharging times of pure conduction and combined
conduction and natural convection models. They concluded that at
higher velocities, the relative deviation in terms of charging time be-
tween both the models was insignificant. Aldoss and Rahman [24]
compared the performance of single-PCM and multi-PCM LHS systems.
They concluded that an increase in stages of multi-PCM up to three,
improves the performance of the system, but adding further stages does
not have any significant impact. Cao et al. [25] conducted numerical
investigations to study the effect of number of fins in a cylindrical and
horizontal annulus type heat exchanger. The fins were distributed
uniformly over the inner heating tube and it was concluded that
boundary wall temperature was the crucial parameter to increase the
conduction heat transfer rate, whereas increasing fins to a higher
number slowed down the convection heat transfer rate gradually.
Khalifa et al. [26] numerically and experimentally tested the effec-
tiveness of a LHS system and compared configurations with bare heat
382
Energy Conversion and Management 188 (2019) 381–397
Abbreviations
PCM Phase Change Material
HTF Heat Transfer Fluid
LHS Latent Heat Storage
Subscripts
c charging
d discharging
eff effective
in inlet
ini initial
l liquid
m melting
out outlet
s solid
Greek symbols
α Cone angle (°)
β Coefficient of Thermal Expansion (1/K)
μ Dynamic viscosity of PCM (kg/m s)
Density of PCM (kg/m3)
pipe and axially finned heat pipe. There was an improvement in heat
transfer effectiveness by 24% for heat pipe system having four fins over
the bare heat pipe system. Further, increase in the number of fins lead
to a marginal hike in effectiveness at the cost of decreased storage rate.
Rahimi et al. [27] investigated the average temperature profile of PCM
during both the solidification and melting of Paraffin RT 35 in a fin and
tube heat exchanger. The heat transfer rate is considerable for turbulent
flow regime of the HTF or by using more number of fins. The impact of
increase in inlet temperature of HTF is more prominent only for 10 °C
rise in temperature but is insignificant with a further increase in the
value. Increasing the HTF flow rate improves the melting rate sub-
stantially as compared to the solidification rate. Solomon and Velraj
[28] analysed the heat transfer mechanism in a vertical finned cylind-
rical unit and found that a higher driving temperature potential and
higher convective heat transfer coefficient exists near the entry region
that accelerates the solidification process, whereas the addition of fins
is observed to be a hindrance for the natural convection process during
sensible cooling of liquid PCM. Sciacovelli et al. [29] presented a shape
optimization strategy by proposing a tree-shaped and Y-shaped fins
with bifurcations. They found that Y-shaped fins with wide angles be-
tween branches were preferable for short operating time and smaller
angles for long operating times. Tiari and Qiu [30] investigated three
different configurations of embedded heat pipes. They established that
using more heat pipes diminishes the thermal resistance between he-
ated surface and PCM melt front leading to faster charging process.
Hosseini et al. [31] conducted numerical and experimental analysis to
test the performance of longitudinal fins attached on the HTF tube in a
LHS system. They found that the heat transfer rate increases for fins
having greater height, but the merging of convection vortices was
blocked.
Few studies have been reported to explore the influence of different
heat exchanger designs on the charging and discharging rates in a LHS
system [32]. Ahmadi et al. [33] presented a numerical model of a spiral
HTF tube heat exchanger and developed melt fraction and temperature
contours for different tube diameters and coil diameters. The melting
G.S. Sodhi, et al.
time was found to decrease by 71.4% by changing tube diameter from
50 mm to 70 mm, whereas a coil diameter of 8 mm was selected as
optimum. Kurnia et al. [34] presented the concept of a rotating LHS
unit and concluded that the performance was superior as compared to a
stationary system.
From the previous research works, it is perceived that the perfor-
mance of thermal storage systems depends upon the heat transfer
characteristics in the PCM region. Using heat transfer augmentation
techniques, especially adding fins is advantageous, but at the same
time, it suppresses the natural convection [28,31] and adds up to the
system size and cost. Also, the heat transfer weakens along the length of
a cylindrical tube in tube LHS system because of the diminishing
thermal potential between the HTF and the PCM. So, melting is very
fast in the inlet section of storage system and becomes very slow at the
extreme end [24]. Hence, there is a need to design a simple heat ex-
changer system with the heat transfer mechanism being intensified via
natural convection and effective PCM distribution. This technique may
accelerate the charging/discharging rates and maintain the compact
size of the system.
In this study, a numerical model is established and the heat transfer
rate along the length of a horizontal shell and tube LHS system is en-
hanced by modifying the shell design from cylindrical shell to conical
shell. The heat distribution is studied with the aid of melt fraction
contours and temperature evolution plots. The diameter of the shell at
inlet and outlet section of the storage system are optimized based on
best performance and finally, the effect of inlet HTF flow velocity and
temperature on the heat transfer in the PCM region are envisioned. The
Fig. 1. (a) Physical model with dimensions and (b) computational model developed for the study.
383
Energy Conversion and Management 188 (2019) 381–397
inclusion of fins in the system is also studied to check the feasibility of
different fin designs for heat transfer enhancement. In the present
study, Sodium Nitrate with a melting temperature of 579 K is chosen as
the PCM.
2. Thermal modeling
2.1. Design of LHS system
A cylindrical shell LHS system is designed for a storage capacity of
1 MJ. The PCM volume (Vpcm) required for this capacity depends on the
heat storage capacity (Q), density (ρ) and latent heat of fusion (hph),
which is calculated using Eq. (1). Based on length (L), outer diameter
(do) of the HTF tube and the calculated PCM volume, the diameter of
the cylindrical shell (Do) is calculated using Eq. (2).
Q= Vpcm hph (1)
Vpcm = (Do2 do2 ) L
4 (2)
A conical shell model is developed by keeping the PCM volume
same as that of a cylindrical model. The diameters D and d of the
conical shell at inlet and outlet sections, respectively, can be calculated
by equating the PCM volume inside the cylindrical and conical shell
LHS systems. The physical model with dimensions is described in
Fig. 1(a). During the charging process, the air is passed through the HTF
tube at a temperature higher than the initial system temperature. The
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Table 1
Dimension values of LHS system.
Dimension do (mm) di (mm) L (mm) Do (mm) tp (mm) d (mm) D (mm) (°)

Values 28.6 24.6 750 77.37 12.7 54 98.6 3.4

energy lost by the air is absorbed by the storage system in the form of Table 3
latent heat and the PCM starts melting. Thermo-physical properties of PCM (NaNO3) and tube material.
The dimensions of the model shown in Fig. 1(a) are described in
Properties Values
Table 1.
The flow and heat transfer inside the PCM region is governed by the PCM [35] Copper [36]
following equations;
Governing equations Latent Heat of fusion (kJ kg−1) 172 n.a.
Melting Temperature (K) 579 n.a.
Continuity Equation: . v = 0 (3) Density (kg m−3) 2261(s), 1910(l) 8700
Thermal conductivity (W m−1 K−1) 0.495(s), 0.565(l) 400
Specific Heat (J kg−1 K−1) 1096(s), 1823(l) 385
v 1 2 Dynamic Viscosity (Pa s) 0.00302 n.a.
Momentum Equation: + (v. )v = ( P+µ v +F + S)
t Thermal Expansion Coefficient 0. 000365 1.7 × 10 5

(4)
Note:- s-solid, l-liquid.
DT 2T
Energy Equation: cp =k
Dt (5)
Table 4
Source Terms: F = g (T Tm )(Boussinesq Approximation) (6) Thermo-physical properties of air [37].
Property Expression
(1 f) 2
S = Amushy v (Kozeny Carman equation)
(f 3 + ) (7) Density (kg m−3) 0.0034845PT 1
Thermal conductivity 0.00227583562 + 1.15480022E 4T + 7.90252856E
cp,s for T < Ts (W m−1 K−1) 8T2 + 4.11702505E 11T3 7.43864331E 15T 4
Specific Heat of PCM:cp = cp,eff for Ts T Tl Specific Heat (J kg−1 K−1) 1047.63657 0.372589265T + 9.45304214E 4T2
cp,l for T > Tl 6.02409443E 7T3 + 1.2858961E 10T 4
(8) Dynamic Viscosity 8.38278E 7 + 8.35717342E 8T 7.69429583E
(10−5 Pa.s) 11T2 + 4.6437266E 14T3 1.06585607E 17T 4
cp,s + cp,l hph
Effective Specific Heat of PCM: cp,eff = +
2 Tl Ts (9)

T Ts T Tm + Tm
Melt Fraction : f = = The flow and energy interactions in the PCM region are governed by
Tl Ts 2 Tm
Eqs. (3)–(5) [38] described above. The source terms added in the mo-
0 for T < Ts mentum equation govern the phase change and natural convection
= 0 1 for Ts T Tl process. Eq. (6) signifies the Boussinesq approximation [39] which
1 for T > Tl (10) eliminates the variation in density between the hotter and the colder
PCM fluid and approximates to source term F , representing the effect
The boundary conditions imposed on the system are described in
of natural convection in terms of temperature variation in the PCM
Table 2.
region. During phase change of PCM, both the solid and liquid phase co-
While developing the model, the following assumptions are made;
exist. The source term mentioned in Eq. (7) is taken from Kozeny-

• PCM is homogeneous and isotropic. Carman equation [40], which accounts for smooth transitions between

• HTF flow is assumed to be laminar. the solid and the liquid velocities. For higher values of mushy zone

• Density variation in the PCM region is introduced using Boussinesq constant (Amushy ) , the convection strength inside PCM decreases, which
reduces the melting rate. Several literature works have recommended
approximation.
• Natural convection process is symmetric about the vertical axes. different values of Amushy varying between 103 and 107 [41]. The value
considered in the present study is 104. is a very small number to
prevent the division of source term by zero and its value is 10−3.
The melting range of the PCM is considered as 4 K. The thermo-
Effective heat capacity method implemented initially by Bonacina
physical properties of PCM (NaNO3) and tube material are shown in
et al. [42] has been validated by several authors [20,22,43,44] to ac-
Table 3. Air is used as the HTF, the thermo-physical properties of which
count for heat capacity variation during phase change. The heat capa-
are described in Table 4. During the discharging process, the system is
city of the PCM defined in terms of latent heat and specific heat of the
at a higher temperature initially and the PCM undergoes solidification
solid and liquid PCM is described in Eqs. (8) and (9).
process as it loses energy in the form of latent heat of fusion.

Table 2
Boundary conditions imposed on the system.
For Charging: Tin > Tm > Tini At t = 0, T = Tini (all domains)
For Discharging:Tin < Tm < Tini At t > 0,T = Tin
vHTF = vin
Adiabatic outer surface of shell T T
= 0 (Cylindrical shell) = 0 (Conical shell)
r r = Do /2 r r=D/2 d/2
No slip boundary conditions vHTF = 0 and v = 0 (walls )

384
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Thermocouples Water in

Section D / Radial
locations - a, b, c
STORAGE
DEVICE

c b a
Paraffin

Water out
Fig. 2. Experimental facility, storage device with dimensions and validated thermocouple positions [45].

2.2. Experimental validation

The effective heat capacity method described above is validated

with the experimental data reported by Longeon et al. [45]. The ex-
perimental facility, storage device with dimensions and thermocouple
locations used for validation in the current study are shown in Fig. 2.
Paraffin (RT35) having a melting point of 308 K was used as PCM in the
reported work and was enclosed in the annular region of a concentric
tube system. The inner diameter of the outer Plexiglas transparent shell
was 44 mm, whereas the inner HTF tube was made of Stainless steel
having inner diameter and thickness of 15 mm and 2.5 mm, respec-
tively. The length of the storage system was 400 mm. Water was used as
the heat transfer fluid in the study.
The experimental condition selected for validation purpose is a
charging process, with water flowing from top of the storage device at
an inlet temperature of 325 K with valves VT1 and VT2 in open con-
dition. Initially, the PCM was maintained at 295 K and the average
velocity of water at tube inlet was 0.01 m/s, thus maintaining a laminar
flow condition. In the reported work, 48 thermocouples were placed at
several angular (0°, 90°, 135°, 180°, 225° and 270°) and radial positions
(a 3 mm, b 6 mm and c 9 mm ) at seven sections (A-G). The
variation at different angular positions was found to be negligible. The
evolution of melting front computed numerically after 4000 s is shown
in Fig. 3(a), and is observed to be slightly faster than in the experi-
mental observations. This happens primarily due to the fact that the
heat transfer is considered ideal, such that there is no heat loss to the
surroundings, which is not possible during real-time experiments.
Fig. 3(b) shows the deviation of temperature profile obtained through
simulations from experimental data at axial location D with radial
distances of a, b and c from the outer surface of the inner tube. Nu-
merically predicted values are found to closely match with experi-
mental data with a maximum deviation of ± 3.5 K.

2.3. Numerical procedure

The finite element based software COMSOL Multiphysics 4.3a [46]

Fig. 3. (a) Comparison of experimental data with progression of melt front
during numerical simulations at t = 4000 s and (b) experimental and numerical
comparison of temperature at different radial positions at a = 3, b = 6 and
c = 9 mm from the pipe surface at location D [45].
has been used to solve the governing equations (Eqs. (3)–(10)). The
system of equations at each Newton-Raphson step are solved using a
direct linear solver (PARDISO). Euler Backward difference formula
(BDF) is employed for time stepping. The time step is adaptive in nature

385
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

and varies based on the error values at previous iterations. The initial 1
(a)
minimum and maximum time steps chosen are 0.001 s and 0.1 s, re-
spectively. The chosen maximum time step helps to solve the problem
0.8
in minimum computational time. The problem is highly non-linear and
the solution diverges for convergence criteria less than 10−3. Hence,

Average Melt Fraction

the convergence criterion is set as 10−3. The simulations are performed 0.6
using a Dell precision T7610 workstation, equipped with two Intel Xeon
E5-2650 v2 processors and 64 GB 1866 MHz RAM.
0.4

2.4. Meshing and grid independent test

116538 mesh elements
0.2 178132 mesh elements
In the model, the mesh is developed with the help of free tetrahedral
237173 mesh elements
and triangular elements with special consideration to effective mesh
distribution close to the walls and boundaries. The model view after 0
meshing is shown in Fig. 4. It is desired to make the numerical model 0 1 2 3 4 5 6 7 8 9 10 11 12 13
independent of the grid so as to obtain best solution quality at a Time (hour)
minimum computation cost.
1
(b)

0.8

Average Melt Fraction

0.6

0.4
Cylindrical shell LHS

102432 mesh elements

0.2 160514 mesh elements
221675 mesh elements

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Time (hour)

Fig. 5. Grid independent test of the thermal model with (a) conical and (b)
cylindrical shell system.

Conical shell LHS

2.5. Key model assessment parameters

The system performance is evaluated with the help of some im-

portant parameters concerning to LHS system and important observa-
Fig. 4. Model view after mesh development in both cylindrical and conical shell tions made from the numerical study are discussed in the following
LHS system. sections.

Fig. 5(a) shows the variation of average melt fraction of PCM with
time for the conical shell models with 116,538, 178,132 and 237,173
mesh elements. It can be observed that the thermal models with
178,132 and 237,173 mesh elements are grid independent. Fig. 5(b).
shows the average melt fraction of PCM of the cylindrical shell models
with 102,432, 160,514 and 221,675 mesh elements. It is found that the
thermal models with 160,514 and 221,675 mesh elements are grid in-
dependent. The thermal model with 178,132 mesh elements for conical
model and 160,514 mesh elements for cylindrical model has been se-
lected to minimize the computational effort.
Melt fraction
The melt fraction is defined in Eq. (10) and is an important factor to
study the performance of LHS system. It represents the percentage of
average PCM by volume which is melted and its value varies from 0
(solid state) to 1 (liquid state) in the considered melting range (Tl-
Ts = 4 K). Melting starts once the average PCM temperature is Ts, and is
finished when it reaches Tl.
Charging time
The time taken for the PCM to melt completely such that the melt
fraction attains the value of unity. The average PCM temperature after

386
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

completion of charging is Tl. The system performance is optimal for a Table 6

minimum value of charging time. This can be minimized by efficient Charging and discharging times for the conical and cylindrical shell LHS sys-
system design having enhanced heat transfer characteristics. tems.
α (°) Cylindrical 1.1 2.3 3.4 4.5 5.6
Discharging time
The discharging time is marked by the completion of the solidifi- Charging time (hr) 13.18 12.53 12.15 12.16 13.05 13.98
Discharging time (hr) 16.1 12.9 12.76 12.23 14.29 14.68
cation process. The value of melt fraction is 0 and the average PCM
temperature is Ts after the completion of the discharging process. For
effective thermal energy storage with coupled end use application, the
discharging rate has to match with energy demand rate.

Energy stored
Energy is stored in the PCM in the form of sensible and latent heat.
The total energy stored is the sum of the sensible and latent heats stored
respectively, and represented in Eq. (13). The sensible heat is stored
due to change in temperature of PCM from initial temperature condi-
tions, whereas the latent heat is stored due to phase change of PCM.
Ec,sensible = m cp,s (Tpcm (t) Tini) (11)

Ec,latent = m hph f (12)

Ec,total = Ec,sensible + Ec,latent (13)

Energy discharged
Energy is lost by the storage system to the HTF as the PCM changes
its state from liquid to solid. The sensible, latent and total energy de-
livered by the storage system are represented in Eqs. (14)–(16).
Ed,sensible = m cp,l (Tini Tpcm (t) ) (14)

Ed,latent = m hph (1 f) (15)

Ed,total = Ed,sensible + Ed,latent (16)

3. Results and discussions

3.1. Optimization of PCM shell dimensions

Numerical simulations are carried out for the optimization of dia-

meters of the conical shell system at the inlet and outlet sections, for
both the charging and the discharging processes and the results are
presented in this section.
Table 5 describes the dimensions of five different conical shell Fig. 6. Variation of average melt fraction of storage system for different cone
configurations, all having a fixed PCM mass. For charging; HTF inlet angles during (a) charging and (b) discharging.
temperature (Tin) = 599 K and initial temperature (Tini) = 559 K,
whereas for discharging; Tin = 559 K and Tini = 599 K have been
chosen. Table 6 presents the charging and discharging times for the
cylindrical and conical shell system at different cone angles and at HTF During charging, it is observed that the configuration with cone
inlet velocity (vin) of 10 m/s. The variation of melt fraction is shown in angle (α) 3.4° melts faster than α = 2.3° for 96% of the process,
Fig. 6(a)–(b). whereas complete melting for both happens in same time. Discharging
process for α = 3.4° is completed in very less time than other values of
Table 5 α. Therefore, α = 3.4° is chosen as an optimal value for both the
Dimensions of different conical shell configurations. charging and discharging processes. Fig. 7 shows the melt fraction
d (mm) D (mm)
contours for the charging process at 12.01 h and discharging process at
tp =
(d
2
do )
(mm) α = 2tan (
1 D d
2L ) (°) 12.22 h. It is observed that the cylindrical shell model is only 93.5/
91.1% charged/discharged when the optimized model (α = 3.4°) is
20.7 70 84.5 1.1
16.7 62 91.8 2.3
fully charged/discharged. As the cone angle decreases, the charging
12.7 54 98.6 3.4 and discharging rates first increase and then decrease.
8.7 46 104.9 4.5 The charging phenomenon characterized by PCM melting is ex-
4.7 38 110.9 5.6 plained as follows. For α = 0° (cylindrical shell system) and α = 1.1°,
the melt front initiates at the inlet section and travels axially in the flow
direction. There is a regular melting behaviour with a decreasing rate of

387
G.S. Sodhi, et al.
Charging
α=0°
α=1.1°
α=2.3°
α=3.4°
96.5%
α=4.5°
92.8%
α=5.6°
Fig. 7. Melt fraction contours of storage system for different cone angles at t = 12.01 h for charging (left) and at 12.22 h for discharging (right).
melting from the inlet to outlet sections of the system and hence, the
melting throughout the LHS system is delayed. For α = 4.5° and 5.6°, a
regular melt front travels axially in the flow direction. At the same time,
due to very less PCM mass there is an intensification of heat transfer
near the outlet section, such that melting starts with movement of an-
other melt front from the outlet section. Due to excessive PCM mass in
the inlet section, melting takes place at a slower rate. Therefore, both
the melt fronts meet near the inlet section, such that there is an overall
delay in melting throughout the system. Hence, it is very important to
distribute the PCM uniformly throughout the system, such that the
average rate of melting is maximum. This happens for α = 3.4°,
wherein the melt fronts initiated at the inlet and outlet section
meet almost at the middle section leading to fast and uniform melting.
During the discharging process, the solidification fronts are almost
symmetrical about both the horizontal and the vertical axes. Again, 3.4°
is found out as the optimum cone angle, where, similar reasoning
equivalent to the charging process leads to this development.
Energy Conversion and Management 188 (2019) 381–397
Discharging
93.5% 91.9%
98.5% 99%
100% 99.5%
100% 100%
94.6%
92.1%
3.2. Effect of HTF inlet conditions on the performance of conical shell
system
3.2.1. Effect of inlet velocity of HTF
Fig. 8(a)-(b) represents the average melt fraction of conical shell
system (α = 3.4°) during charging (Tin = 599 K and Tini = 559 K) and
discharging (Tin = 559 K and Tini = 599 K) at different velocities
(vin = 2 m/s, 6 m/s and 10 m/s). It is observed that the percentage re-
duction in charging time with an increase in vin from 2 m/s to 6 m/s is
25.1%, whereas, with a further increase from 6 m/s to 10 m/s, it is
10.5% only. During discharging the reduction in discharging time with
an increase in vin from 2 m/s to 6 m/s and from 6 m/s to 10 m/s are 30%
and 14.4%, respectively.
The variation in average PCM temperature during charging and
discharging at different velocities is shown in Fig. 9(a)-(b). Initially, the
PCM temperature rises sharply due to a significant temperature dif-
ference between the HTF and the PCM. As the PCM reaches the phase
change temperature, an almost constant temperature profile is
388
G.S. Sodhi, et al.
Fig. 8. Average melt fraction variation of conical shell system for different HTF
inlet velocities (vin = 2, 4 and 10 m/s) for (a) charging and (b) discharging.
observed. Once the PCM melts completely, the temperature starts rising
again due to sensible heat transfer between the HTF and the melted
PCM. From Figs. 8(a)-(b) and 9(a)-(b), it is evident that increase in
velocity to very high values, does not necessarily lead to significant
reduction in charging and discharging times.
3.2.2. Effect of inlet temperature of HTF
During charging (Tini = 559 K and vin = 10 m/s) and discharging
(Tini = 599 K and vin = 10 m/s) of the conical shell system (α = 3.4°),
the ΔT = |Tin Tini| is changed from 40 K to 50 K and 60 K. The average
melt fraction curves for different ΔT values are shown in Fig. 10(a)-(b).
With an increase in ΔT from 40 K to 50 K, there is a reduction in
charging time by 26.9% and discharging time by 36.7%. Further in-
creasing ΔT to 60 K, the percentage reduction is 21.5% and 21.2%,
respectively. With an increase in ΔT, there is a definite improvement in
389
Energy Conversion and Management 188 (2019) 381–397
Fig. 9. Average temperature variation of conical shell system at different HTF
inlet velocities (vin = 2, 4 and 10 m/s) for (a) charging and (b) discharging.
system performance because of increase in the thermal gradient for heat
transfer between the HTF and the PCM. The effect of HTF inlet tem-
perature is quite significant and can be even employed to increase the
heat transfer rate for low-cost HTF having poor thermal properties.
It can be observed that the inlet conditions of the HTF have a
considerable impact on system performance. These parameters can
change according to system requirements, and hence an optimal com-
bination of these parameters has to be selected.
3.3. Performance comparison of cylindrical and conical shell systems
By now, it is evident from the study that the introduction of conical
shell design leads to improvement in performance than cylindrical shell
system. In this section, the relative improvement is studied in terms of
charging and discharging times (described in Table 7).
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Fig. 11. Average melt fraction variation for conical and cylindrical shell sys-
tems (vin = 6 m/s).

process. Figs. 12 and 13 describe the percentage completion of melting

and solidification of the conical and cylindrical shell systems with re-
spect to time.
Initially, the melting behaviour of cylindrical and conical shell
systems are identical. The behaviour continues till t = 5.43 h. Due to
less PCM mass at the outlet section of the conical shell system, an ad-
ditional melt front is initiated. Finally, at t = 13.33 h the melting in
conical shell system is almost completed in the middle section, whereas
the average melting in the cylindrical shell system is completed by
88.5%.
The solidification contours shown in Fig. 13 are symmetrical about
the horizontal and vertical axes for both the conical shell and cylind-
Fig. 10. Average melt fraction variation of conical shell system for different ΔT
rical shell systems. Initially, the solidification behaviour of both systems
for (a) charging and (b) discharging.
is identical. After t = 5.43 h, the solidification initiates at the outlet
section of the conical shell. Finally, at t = 15.5 h the solidification is
completed by 90% in the cylindrical shell system, with no compre-
Table 7 hensive solidification at the outlet section. At the same time, the conical
Charging and discharging times of the cylindrical and the optimized conical shell system is completely solidified except the small amount of PCM
shell system at different HTF inlet velocities. left at the outer periphery of the middle section. This concludes that the
vin (m/s) 2 6 10 conical shell system is advantageous due to effective distribution of
heat in the PCM region, both during the melting and solidification.
Charging time, tc (hr) Cylindrical 20.95 16.34 13.79
Conical 18.13 13.58 12.15
% tc reduction 13.4 17.0 11.9
3.3.2. Average PCM and HTF outlet temperatures
Discharging time, td (hr) Cylindrical 31.14 19.96 15.56 Fig. 14(a)-(b) describes the variation of average PCM temperature
Conical 22.16 15.5 13.26 and HTF outlet temperature of conical and cylindrical shell system.
% td reduction 28.8 22.3 14.7 During the charging process, both the average temperature of the PCM
and the HTF outlet temperature of the conical shell system reaches the
HTF inlet temperature approximately 2 h earlier than cylindrical shell
system. This indicates that the heat transfer rate is high in case of
3.3.1. Charging and discharging time conical shell system. During discharging, the behaviour of average PCM
In case of conical shell design, the PCM mass is distributed in such a temperature and HTF outlet temperature of conical and cylindrical shell
way that the decreasing temperature difference between HTF and PCM system is identical till t = 13 h. Later, the heat transfer rate is relatively
along the length of the LHS system does not affect the heat transfer high in conical shell system and the HTF outlet temperature approaches
significantly, which results in uniform melting of the PCM. the HTF inlet temperature.
The charging and discharging rates of the conical shell system are
higher than the cylindrical shell system as shown in Fig. 11. As com- 3.3.3. Energy storage and discharge rate
pared to the charging times, the discharging times are significantly less. Fig. 15(a)-(b) shows the rates of energy stored and discharged in the
Since discharging is conduction dominated, it can be comprehended form of sensible and latent heat. The conical and cylindrical shell sys-
that by changing the design from cylindrical to conical shell system, tems are developed for a latent heat capacity of 1 MJ, whereas, there is
there is much more enhancement in the conduction rate. Therefore, a an additional 0.19 MJ sensible energy stored in both the cylindrical and
similar effect leads to enhanced heat transfer during the charging conical shell systems due to temperature difference between the HTF
and the PCM. It can be observed that the energy storage/discharge rate

390
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Fig. 12. Comparison of melting contours of cylindrical and conical shell systems at different time intervals.

Fig. 13. Comparison of solidification contours of cylindrical and conical shell systems at different time intervals.

391
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Fig. 14. (a) Average PCM temperature and (b) HTF outlet temperature varia-
tion of conical and cylindrical shell systems during charging and discharging Fig. 15. Energy stored/discharged rates of the conical shell and the cylindrical
processes. shell system for (a) charging and (b) discharging.

of the conical shell LHS system is higher than that of cylindrical shell
system.

4. Inclusion of fins in the conical shell system

In this section, the current study has been further extended to

analyse the effect of different fin designs implemented in the developed
conical shell storage system. It has been observed from the reported
research works that addition of fins on the HTF tubes lead to a faster
rate of heat transfer in the LHS system. Addition of fins leads to an Design A Design B
increase in heat transfer surface area, and hence enhances the con-
duction heat transfer which dominates during the discharging process
[47]. Whereas, the charging process involves movement of the liquid
PCM that is desired to have additional convection currents leading to an
enhancement of heat transfer rate. But, the size and shape of the fins
have to be limited to a specific shape and number, in order to avoid the
blockage of convection currents [28].
The heat transfer and the fluid flow are symmetrical about the
vertical axis during both the charging and the discharging processes.
Therefore, only half of the model geometry is developed to save the
Design C
computational effort. The height, length and width of fins is selected as
10 mm, 750 mm and 2 mm, respectively and the PCM mass is main-
Fig. 16. Different designs of conical shell system with fins.
tained as constant. Different fin designs as shown in Fig. 16 are illu-
strated as follows:

392
G.S. Sodhi, et al.
1 Therefore, the heat transfer rate between the HTF and PCM at inlet
0.8
Average Melt Fraction
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7 8 9 10
Time (hour)
Fig. 17. Variation of melt fraction with time during the charging and dis-
charging of conical shell system with different fin designs.
A. Four longitudinal straight fins.
B. Four longitudinal fins with uniformly increasing tapered cross sec-
tion from inlet to outlet.
C. Four longitudinal fins with uniformly decreasing tapered cross sec-
tion from inlet to outlet.
The numerical simulations are carried out for charging (Tin = 599 K
and Tini = 559 K) and discharging (Tin = 559 K and Tini = 599 K) at
vin = 6 m/s.
4.1. Charging and discharging
Fig. 17 illustrates the time taken for the charging and discharging
processes for all the cases namely; Design A, Design B and Design C. The
Design A melts in 12.06 h, whereas, Design B and Design C melt in
13.79 h and 11.01 h. As reported in Table 7, the charging time eval-
uated for conical shell system without fins is 13.58 h. It is observed that
the time taken for charging of Design B is more than the system without
fins, whereas Design A and C show a relative improvement over system
without fins. Hence, the use of fins does not necessarily improve the
heat transfer rate.
Further, the relative comparison of the effect of presence of fin
surfaces on heat transfer rate at inlet and outlet sections is explained by
evaluation of local temperature profile at two coordinates: Xin: x, y, z
(0 mm, 20 mm, 45 mm) and Xout: x, y, z (0 mm, 730 mm, 26 mm).
Fig. 18(a)-(b) shows the local temperature variation at Xin and Xout
during charging and discharging processes. Fig. 18(c) represents the
variation in charging and discharging rates for three different designs at
Xin and Xout. It is observed that the rate of conduction heat transfer
during discharging in Design C is higher than Design B at Xin, and
marginally lower at Xout. This happens because of presence of more fin
surface at Xin in Design C and at Xout in Design B. On the other hand, the
charging rate in Design C is higher than Design B, both at Xin and Xout.
This is possible due to relatively higher conduction heat transfer rate at
Xin because of more fin surface. Whereas at Xout, even with presence of
more fins in Design B, the charging rate is slower as the convection
currents are affected adversely.
Energy Conversion and Management 188 (2019) 381–397
and outlet sections is effected by the temperature gradient, fin surfaces
and convection currents. The presence of fins is useful in the region of
high thermal potential between HTF and PCM (near inlet) both during
charging and discharging. Whereas, with the presence of fins in the
c_Design A
region of lower thermal potential (near outlet), there is a compromise
c_Design B between conduction heat transfer and free movement of PCM.
c_Design C The melting contours for conical shell system with different fin de-
d_Design A signs are plotted in Fig. 19. Initially, fins accelerate the heat transfer rate
d_Design B in Design A and C near the inlet section. At t = 5.43 h, the melting in-
itiates at the outlet section in all the three cases. Further, at t = 10.86 h,
d_Design C
the melting is completed in half of the system starting from the inlet side
of Design A and C. Whereas, due to very less fin surfaces at inlet section
and blockage of convection currents at outlet section, melting is still
incomplete in Design B. Finally, melting is only completed by 92.5% in
11 12 13 14 Design B, whereas it has completed in Design A and C.
The discharging process is conduction dominated, thus having more
fin surfaces is always beneficial. Initially, the solidification is faster due
to presence of fins near the inlet section for Design A and C as shown in
Fig. 20. Further at t = 5.43 h, the solidification initiates in the outlet
section in all the designs. In Design B, the fins do not give much ad-
vantage as more fin surfaces are present in the region of low tem-
perature difference between the HTF and the PCM (near outlet section).
The solidification completes in 13 h for Design A, followed by Design C
and B in 13.2 h and 13.9 h, respectively.
The maximum reduction in charging time as compared to system
without fins is in Design C (approximately 19%), whereas the maximum
reduction in discharging time as compared to system without fins is in
Design A and C (approximately 16% and 15%).
4.2. Average PCM and HTF outlet temperature
Fig. 21(a) shows the average temperature of the PCM for both
charging and discharging processes. The average temperature plots are
consistent with the average melt fraction curves plotted in Fig. 17. Due
to a faster rate of heat transfer throughout the charging process, the
average temperature of the PCM is always higher in Design C followed
by Design A and B. During discharging, due to a higher rate of con-
duction heat transfer in Design A, the average PCM temperature is al-
ways lower than Design B and C. Fig. 21(b) shows the variation of HTF
outlet temperature for all the fin designs. The HTF outlet temperature
attains a temperature closer to HTF inlet temperature earlier in case of a
higher heat transfer from HTF to PCM (Design C in charging) and vice
versa (Design A in discharging).
The study of different fin designs concludes that the fins are ad-
vantageous for heat transfer, but their design is a crucial criterion to
decide their feasibility with LHS system. The improvement in heat
transfer during the discharging process is merely dependent upon in-
creasing the heat transfer surface area of the fins. At the same time,
careful measures have to be taken to select the design of fins for the
charging process.
5. Conclusions
In the present study, a numerical analysis is performed to develop a
conical shell based LHS system. The adapted geometry is optimized
after varying the cone angle, starting from a perfectly cylindrical geo-
metry. After selecting an optimal shell design, the applicability of dif-
ferent fin designs and their effect on the melting/solidification rates are
393
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

599
(a) Xin : x, y, z (0mm, 20mm, 45mm)
594

589

Local Temperature (K)

584

579
c_Design A
574 c_Design B
c_Design C
569
d_Design A
d_Design B
564
d_Design C
559
0 2 4 6 8 10 12 14
Time (hour)

599
(b)
Xout : x, y, z (0mm, 730mm, 26mm)
594

589
Local Temperature (K)

584

579

c_Design A
574
c_Design B
c_Design C
569
d_Design A
d_Design B
564
d_Design C
559
0 2 4 6 8 10 12 14
Time (hour)

(c) Xin Xout

Charging rate increases Charging rate increases
Design B Design C Design B Design C
Discharging rate increases Discharging rate increases
marginally

Design A Design A

Fig. 18. (a) Local PCM temperature at inlet section (Xin), (b) Local PCM temperature at outlet section (Xout) and (c) variation in charging and discharging rates for
different fin designs during the charging and discharging processes.

discussed. The following conclusions are drawn from the study;
• The conical shell system optimized for charging and discharging
processes has a cone angle of 3.4° and diameter at inlet and outlet
section of 98.6 mm and 54 mm, respectively.
• Due to less PCM mass near the outlet section of the conical shell
system, melting and solidification starts earlier in this region than
cylindrical shell system. The melting and solidification completes at
the middle section of conical shell system due to merging of two
melt fronts originating from inlet and outlet sections. There is a
maximum reduction in charging and discharging times up to 17%
and 28% due to better PCM distribution.

394
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

Design A Design B Design C

t=1.36 hr 9.6% 6.1% 9.4%

39.2%
t=5.43 hr 46.6% 49.2%

t=10.86 hr
92.5% 82.1% 99.5%

t=12.22 hr 100.0% 92.5% 100.0%

Fig. 19. Melting contours of the conical shell system for different fin designs during the charging process.

Design A Design B Design C

t=1.36 hr 8.5% 5.7% 8.5%

t=5.43 hr
48.7% 42.1% 49.4%

t=10.86 hr
90.8% 87.7% 92.0%

t=12.22 hr
99.2% 94.8% 98.0%

Fig. 20. Solidification contours of the conical shell system for different fin designs during the discharging process.

• HTF inlet conditions impact the system performance significantly.
By increasing HTF inlet velocity from 2 m/s to 6 m/s, both charging
and discharging rates increase comprehensively, but the influence of
velocity is insignificant thereafter. Also, by changing the
ΔT = |Tin Tini| from 40 K to 50 K and later to 60 K, the percentage
reduction in charging time are 26.7% and 21.5%, whereas, the
reduction in discharging time are 36.7% and 21.2%, respectively.
Thus, even by using an HTF (air) with poor thermal properties, a
higher ΔT is beneficial to improve the system performance.
• The fins are beneficial to increase the conduction heat transfer rate
during the charging and discharging in the region of higher thermal
potential (inlet section), but are inadequate in the region with low

395
G.S. Sodhi, et al. Energy Conversion and Management 188 (2019) 381–397

599 (a) c_Design A

References
c_Design B
594
c_Design C [1] Hasnain SM. Review on sustainable thermal energy storage technologies, Part I:
d_Design A heat storage materials and techniques. Energy Convers Manag 1998;39:1127–38.
589 https://doi.org/10.1016/S0196-8904(98)00025-9.
d_Design B
[2] Reddy KS, Mudgal V, Mallick TK. Review of latent heat thermal energy storage for
Average Temperature (K)

d_Design C
584 improved material stability and effective load management. J Energy Storage
2018;15:205–27. https://doi.org/10.1016/j.est.2017.11.005.
[3] Sharma A, Tyagi VV, Chen CR, Buddhi D. Review on thermal energy storage with
579
phase change materials and applications. Renew Sustain Energy Rev
2009;13:318–45. https://doi.org/10.1016/j.rser.2007.10.005.
574 [4] Alva G, Lin Y, Fang G. An overview of thermal energy storage systems. Energy
2018;144:341–78. https://doi.org/10.1016/j.energy.2017.12.037.
569 [5] Veerakumar C, Sreekumar A. Phase change material based cold thermal energy
storage: materials, techniques and applications – a review. Int J Refrig
2016;67:271–89. https://doi.org/10.1016/j.ijrefrig.2015.12.005.
564
[6] Rabha DK, Muthukumar P. Performance studies on a forced convection solar dryer
integrated with a paraffin wax–based latent heat storage system. Sol Energy
559 2017;149:214–26. https://doi.org/10.1016/j.solener.2017.04.012.
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 [7] Guo J, Huai X, Cheng K. The comparative analysis on thermal storage systems for
Time (hour) solar power with direct steam generation. Renew Energy 2018;115:217–25. https://
doi.org/10.1016/j.renene.2017.08.046.
600 [8] Laing D, Bahl C, Bauer T, Lehmann D, Steinmann WD. Thermal energy storage for
(b) direct steam generation. Sol Energy 2011;85:627–33. https://doi.org/10.1016/j.
c_Design A solener.2010.08.015.
c_Design B [9] Islam MT, Huda N, Abdullah AB, Saidur R. A comprehensive review of state-of-the-
595 c_Design C art concentrating solar power (CSP) technologies: current status and research
d_Design A trends. Renew Sustain Energy Rev 2018;91:987–1018. https://doi.org/10.1016/j.
d_Design B rser.2018.04.097.
HTF Outlet Temperature (K)

590 d_Design C [10] Pointner H, de Gracia A, Vogel J, Tay NHS, Liu M, Johnson M, et al. Computational
efficiency in numerical modeling of high temperature latent heat storage: com-
parison of selected software tools based on experimental data. Appl Energy
585 2016;161:337–48. https://doi.org/10.1016/j.apenergy.2015.10.020.
[11] Al-Abidi AA, Bin Mat S, Sopian K, Sulaiman MY, Mohammed AT. CFD applications
for latent heat thermal energy storage: a review. Renew Sustain Energy Rev
580 2013;20:353–63. https://doi.org/10.1016/j.rser.2012.11.079.
[12] Aceves SM, Nakamura H, Reistad GM, Martinez-Frias J. Optimization of a class of
latent thermal energy storage systems with multiple phase-change materials. J Sol
575 Energy Eng ASME 1998;120:14–9. https://doi.org/10.1115/1.2888040.
[13] Fan L, Khodadadi JM. Thermal conductivity enhancement of phase change mate-
rials for thermal energy storage A review. Renewable Sustainable Energy Rev
2011;15:24–46. https://doi.org/10.1016/j.rser.2010.08.007.
570
[14] Qureshi ZA, Ali HM, Khushnood S. Recent advances on thermal conductivity en-
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Time (hour) hancement of phase change materials for energy storage system: a review. Int J
Heat Mass Transf 2018;127:838–56. https://doi.org/10.1016/j.ijheatmasstransfer.
Fig. 21. (a) Average PCM temperature and (b) HTF outlet temperature varia- 2018.08.049.
[15] Dinker A, Agarwal M, Agarwal GD. Heat storage materials, geometry and applica-
tion for different fin designs during the charging and discharging processes.
tions: a review. J Energy Inst 2017;90:1–11. https://doi.org/10.1016/j.joei.2015.
10.002.
[16] Zhao W, France DM, Yu W, Kim T, Singh D. Phase change material with graphite
thermal potential (outlet section). Near the outlet section, presence foam for applications in high-temperature latent heat storage systems of con-
of fins leads to a compromise between conduction heat transfer and centrated solar power plants. Renew Energy 2014;69:134–46. https://doi.org/10.
convection currents in liquid PCM. Proposed Design C for charging 1016/j.renene.2014.03.031.
[17] Elbahjaoui R, El Qarnia H. Transient behavior analysis of the melting of nano-
and discharging is found to be most effective with a reduction of
particle-enhanced phase change material inside a rectangular latent heat storage
∼19% and 15% in time over the system without fins. unit. Appl Therm Eng 2017;112:720–38. https://doi.org/10.1016/j.
applthermaleng.2016.10.115.
This study provides an insight for improving charging and dis- [18] Gasia J, Maldonado JM, Galati F, De Simone M, Cabeza LF. Experimental evaluation
of the use of fins and metal wool as heat transfer enhancement techniques in a
charging performances of LHS system by alteration of simple design latent heat thermal energy storage system. Energy Convers Manag 2019;184:530–8.
aspects of conventional cylindrical systems. Such modifications can be https://doi.org/10.1016/j.enconman.2019.01.085.
more useful than selecting highly complex structures, thereby control- [19] Johar DK, Sharma D, Soni SL, Gupta PK, Goyal R. Experimental investigation and
exergy analysis on thermal storage integrated micro-cogeneration system. Energy
ling the overall size and cost incurred for designing and manufacturing. Convers Manag 2017;131:127–34. https://doi.org/10.1016/j.enconman.2016.10.
There is a possibility of extending this method to high capacity storage 075.
systems, where the advantages in terms of cost reduction will increase [20] Sunku Prasad J, Muthukumar P, Anandalakshmi R, Niyas H. Comparative study of
phase change phenomenon in high temperature cascade latent heat energy storage
substantially. system using conduction and conduction-convection models. Sol Energy
2018;176:627–37. https://doi.org/10.1016/j.solener.2018.10.048.
[21] Abdulateef AM, Mat S, Abdulateef J, Sopian K, Al-abidi AA. Geometric and design
Declaration of interest statement parameters of fi ns employed for enhancing thermal energy storage systems: a re-
view. Renew Sustain Energy Rev 2017;82:1–16. https://doi.org/10.1016/j.rser.
2017.07.009.
The authors declare that they have no known competing financial [22] Niyas H, Prasad S, Muthukumar P. Performance investigation of a lab – scale latent
interests or personal relationships that could have appeared to influ- heat storage prototype – numerical results. Energy Convers Manag
2017;135:188–99. https://doi.org/10.1016/j.enconman.2016.12.075.
ence the work reported in this paper. [23] Riahi S, Saman WY, Bruno F, Belusko M, Tay NHS. Impact of periodic flow reversal
of heat transfer fluid on the melting and solidification processes in a latent heat
shell and tube storage system. Appl Energy 2017;191:276–86. https://doi.org/10.
1016/j.apenergy.2017.01.091.
Acknowledgement
[24] Aldoss TK, Rahman MM. Comparison between the single-PCM and multi-PCM
thermal energy storage design. Energy Convers Manag 2014;83:79–87. https://doi.
The authors gratefully acknowledge the financial support from the org/10.1016/j.enconman.2014.03.047.
Department of Science and Technology (DST), Government of India, [25] Cao X, Yuan Y, Xiang B, Sun L, Xingxing Z. Numerical investigation on optimal
number of longitudinal fins in horizontal annular phase change unit at different
(Project No: DST/TMD/SERI/D12(C)).

396
G.S. Sodhi, et al.
wall temperatures. Energy Build 2018;158:384–92. https://doi.org/10.1016/j.
enbuild.2017.10.029.
[26] Khalifa A, Tan L, Date A, Akbarzadeh A. A numerical and experimental study of
solidification around axially finned heat pipes for high temperature latent heat
thermal energy storage units. Appl Therm Eng 2014;70.. https://doi.org/10.1016/j.
applthermaleng.2014.05.080.
[27] Rahimi M, Ranjbar AA, Ganji DD, Sedighi K, Hosseini MJ, Bahrampoury R. Analysis
of geometrical and operational parameters of PCM in a fin and tube heat exchanger.
Int Commun Heat Mass Transf 2014;53:109–15. https://doi.org/10.1016/j.
icheatmasstransfer.2014.02.025.
[28] Solomon GR, Velraj R. Analysis of the heat transfer mechanisms during energy
storage in a Phase Change Material filled vertical finned cylindrical unit for free
cooling application. Energy Convers Manag 2013;75:466–73. https://doi.org/10.
1016/j.enconman.2013.06.044.
[29] Sciacovelli A, Gagliardi F, Verda V. Maximization of performance of a PCM latent
heat storage system with innovative fins. Appl Energy 2015;137:707–15. https://
doi.org/10.1016/j.apenergy.2014.07.015.
[30] Tiari S, Qiu S. Three-dimensional simulation of high temperature latent heat
thermal energy storage system assisted by finned heat pipes. Energy Convers Manag
2015;105:260–71. https://doi.org/10.1016/j.enconman.2015.08.004.
[31] Hosseini MJ, Ranjbar AA, Rahimi M, Bahrampoury R. Experimental and numerical
evaluation of longitudinally finned latent heat thermal storage systems. Energy
Build 2015;99:263–72. https://doi.org/10.1016/j.enbuild.2015.04.045.
[32] Medrano M, Yilmaz MO, Nogués M, Martorell I, Roca J, Cabeza LF. Experimental
evaluation of commercial heat exchangers for use as PCM thermal storage systems.
Appl Energy 2009;86:2047–55. https://doi.org/10.1016/j.apenergy.2009.01.014.
[33] Ahmadi R, Hosseini MJ, Ranjbar AA, Bahrampoury R. Phase change in spiral coil
heat storage systems. Sustain Cities Soc 2018;38:145–57. https://doi.org/10.1016/
j.scs.2017.12.026.
[34] Kurnia JC, Sasmito AP, Ping SI. Investigation of heat transfer on a rotating latent
heat energy storage. Energy Procedia 2017;105:4173–8. https://doi.org/10.1016/j.
egypro.2017.03.887.
[35] Muhammad MD, Badr O. Performance of a finned, latent-heat storage system for
397
View publication stats
Energy Conversion and Management 188 (2019) 381–397
high temperature applications. Appl Therm Eng 2017;116:799–810. https://doi.
org/10.1016/j.applthermaleng.2017.02.006.
[36] COMSOL. Heat Transfer Module Model Library; 2016.
[37] Cengel YA, Ghajar JA. Heat and Mass Transfer Fundamentals and Applications
Solution Manual; 2015.
[38] Chung TJ. Computational Fluid Dynamics. vol. 57; 2018.
[39] Demuren A, Grotjans H. Buoyancy-driven flowsbeyond the boussinesq approxima-
tion. Numer Heat Transf Part B Fundam 2009;56:1–22. https://doi.org/10.1080/
10407790902970080.
[40] Brent AD, Voller VR, Reid KJ. Enthalpy-porosity technique for modeling convec-
tion-diffusion phase change: application to the melting of a pure metal. Numer Heat
Transf 1988;13:297–318. https://doi.org/10.1080/10407788808913615.
[41] Kumar M, Krishna DJ. Influence of mushy zone constant on thermohydraulics of a
PCM. Energy Procedia 2017;109:314–21. https://doi.org/10.1016/j.egypro.2017.
03.074.
[42] Bonacina C, Comini G, Fasano A, Primicerio M. Numerical solution of phase-change
problems. Int J Heat Mass Transf 1973;16:1825–32. https://doi.org/10.1016/0017-
9310(73)90202-0.
[43] Ogoh W, Groulx D. Effects of the heat transfer fluid velocity on the storage char-
acteristics of a cylindrical latent heat energy storage system: a numerical study.
Heat Mass Transf 2012;48:439–49. https://doi.org/10.1007/s00231-011-0888-3.
[44] Yang H, He Y. Solving heat transfer problems with phase change via smoothed
effective heat capacity and element-free Galerkin methods. Int Commun Heat Mass
Transf 2010;37:385–92. https://doi.org/10.1016/j.icheatmasstransfer.2009.12.
002.
[45] Longeon M, Soupart A, Fourmigué JF, Bruch A, Marty P. Experimental and nu-
merical study of annular PCM storage in the presence of natural convection. Appl
Energy 2013;112:175–84. https://doi.org/10.1016/j.apenergy.2013.06.007.
[46] COMSOL Inc., www.comsol.com.
[47] Rabienataj Darzi AA, Jourabian M, Farhadi M. Melting and solidification of PCM
enhanced by radial conductive fins and nanoparticles in cylindrical annulus. Energy
Convers Manag 2016;118:253–63. https://doi.org/10.1016/j.enconman.2016.04.
016.