Applied Thermal Engineering 214 (2022) 118803

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Melting dynamics analysis of a multi-tube latent heat thermal energy

storage system: Numerical study
Vikas *, Ankit Yadav, Sushant Samir
Department of Mechanical Engineering, Punjab Engineering College, Chandigarh 160012, India

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

Keywords: This numerical study investigates the melting dynamics of paraffin wax as a phase change material (PCM) in a
Latent heat storage multi-tube latent heat thermal energy storage system (MT-LHTESS). The performance of an arrangement with
Phase change material four tubes is compared with a concentric single tube arrangement keeping the same cross-section area of heat
Multi-tube array
transfer fluid tubes and the same PCM quantity. Four longitudinal fins of two types are also incorporated to
Fins
Heat storage rate
further improve the melting. Thirteen cases with different arrangements of tubes and fins are compared in this
study. The study attempts to select a suitable arrangement of the tubes. A two-dimensional enthalpy-porosity
model in ANSYS-Fluent is used to solve the governing equations. The melting results are depicted as the contours
of liquid fraction, temperature, and flow streamlines. The transient performance is assessed based on progress
plots of liquid fraction and average domain velocity. It is observed that splitting into multiple tubes and using fins
enhance the melting performance. The generation and coalescence of small vortices in the molten PCM accel­
erate the melting performance. The staggered array, larger eccentricity, and (+ type) finned tubes outperform
their counterparts. The lowest melting time is observed in Case 2(+), having four staggered tubes, at 30 mm
eccentricity, and with (+ type) fins. This case takes 75 min for complete melting, and 1411.1 kJ/m heat at a rate
of 322 W/m is stored. The melting time gets reduced by 53.3–71.1% using four tubes and by 65.78–83.33% using
finned four tubes as compared to the single tube case.

available during the peak sunshine hours and later put it to use to
1. Introduction smoothen the supply and demand gap. Among the three mechanisms of
TESS, namely- sensible, latent, and thermo-chemical, the latent heat
In the past few decades, researchers have made a lot of efforts to thermal energy storage systems (LHTESS) has many advantages because
replace conventional energy resources with renewable energy resources. of its high energy storage capacity, safe operation, chemical stability,
Solar energy is the most popular among renewable energy resources as it and nearly isothermal heat storage and release [13–15]. There are many
has abundant reserves, is inexhaustible, is little polluting while har­ commercially available phase change materials (PCM) that are being
nessing, and it is accessible to everyone around the world [1]. Sun is a used as storage materials in LHTESS [16,17]. Abokersh et al. [16] briefly
good source of heat and light, which can be utilized for many domestic studied the use of PCM in the melting range of 25-100˚C for solar do­
and industrial practices. Solar energy has many practical applications mestic water heating (SDWH).
such as space heating, solar drying, solar desalination, solar photovol­ In the past, a lot of research has been conducted by researchers on
taics, solar air conditioning, etc. [2]. Industries can utilize solar heat in LHTESS, especially for the shell and tube type arrangements. The sol­
many thermal applications such as pre-heating, drying, washing, cook­ id–liquid phase change is the most investigated mechanism for LHTESS
ing, pasteurization, bleaching, etc. [3–5]. About 30% of the heat re­ because of its high specific heat of fusion and negligible volumetric
quirements of the industry are below 100˚C [6]. Solar thermal collectors changes [18,19]. These systems can store heat during the charging cycle
can be installed to use solar heat in these applications [7,8]. Flat plate and release it during discharging cycle. The thermal conductivity of
solar thermal collectors can serve for applications up to 80˚C [8]. But the most of the commercially available PCM lies on the lower side, which
intermittency of solar insolation makes its use unreliable for continuous affects the storage and release of heat [20–23]. To overcome this limi­
applications [9,10]. To fill this intermittence thermal energy storage tation many researchers have proposed structural improvements like
systems (TESS) are used [11,12]. These systems store the surplus energy using finned tubes [24–26], providing eccentricity [27–29], embedding

* Corresponding author.
E-mail address: [email protected] (Vikas).

https://doi.org/10.1016/j.applthermaleng.2022.118803
Received 1 February 2022; Received in revised form 12 April 2022; Accepted 6 June 2022
Available online 15 June 2022
1359-4311/© 2022 Elsevier Ltd. All rights reserved.
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Nomenclature q heat stored (kJ/m)

q˙ heat storage rate (W/m)
Abbreviations Sx source term of x-momentum
CFD computational fluid dynamics Sy source term of y-momentum
FVM finite volume method t time (s)
HTF heat transfer fluid tc time taken for complete melting (s)
LHTESS latent heat thermal energy storage system T temperature (K)
MT-LHTESS multi-tube latent heat thermal energy storage system T0 reference temperature (K)
PCM phase change material THTF temperature of HTF (K)
SDWH solar domestic water heating Ti initial temperature (K)
TESS thermal energy storage system Tl liquidus temperature (K)
3-D three dimensional Tm melting temperature (K)
2-D two dimensional Ts solidus temperature (K)
u x-component of velocity (m/s)
Symbols v y-component of velocity (m/s)
Amushy mushy zone parameter (kg/m3.s)
Cp specific heat (J/kg.K) Greek Symbols
d tube diameter (mm) ρ density (kg/m3)
D shell diameter (mm) μ dynamic viscosity (kg/m.s)
e eccentricity β expansion coefficient (1/K)
g acceleration due to gravity (m/s2) λ liquid fraction
h sensible enthalpy (J/kg) η efficiency
h0 enthalpy at reference temperature (J/kg)
ΔH latent enthalpy (J/kg) Subscripts
H total enthalpy (J/kg) inst instantaneous
k thermal conductivity (W/m.K) max maximum
L latent heat (J/kg) total total
P pressure (Pa) th theoretical

nanoparticles [30], using heat pipes [31], embedding metal matrix vertical orientation. In the finned tube case, the orientation had only
[32–35], etc. LHTESS with finned heat transfer fluid (HTF) tubes were minor impacts on performance. HTF flow rate slightly affected the
the most studied arrangements for performance enhancement to the best melting performance. Kousha et al. [39] also reported faster melting in
of the author’s knowledge. Many researchers have also emphasized the the case of horizontal orientation as compared to vertical. Siyabi et al.
geometrical modifications of the shell and tube of the LHTESS. [40] also investigated the effect of orientation of concentric cylinder
Yang et al. [12] added longitudinal fins in a horizontal double pipe type LHTESS. A very small axial temperature difference was reported in
LHTESS. A 72.85% reduction in melting time was reported by increasing horizontal orientation. Thus the horizontal orientation of LHTESS is
the number of fins from 4 to 52. The study also suggested not to increase used in this study.
the number of fins blindly as the presence of fins affects the local natural Many studies on the effect of eccentricity have been reported by
convection. Hosseini et al. [36] investigated experimentally and researchers. Yazici et al. [29] investigated the effect of providing ec­
numerically the melting behavior of PCM in a double pipe type LHTESS centricity to the HTF tube in a paraffin wax based double pipe LHTESS.
to understand the natural convection effects. The study concluded that Looking at the melting behavior in the upper half being faster due to
melting was greatly influenced by natural convection. A considerable convection domination, researchers attempted to increase the PCM
reduction in melting time by increasing HTF inlet temperature was volume in the upper half by providing downward eccentricity to the
observed. Mahdi et al. [37] designed a three-dimensional (3-D) nu­ inner HTF tube. A 30 mm eccentricity to a 28.5 mm diameter tube in a
merical model of a helical-shaped double pipe LHTESS. It was designed 110 mm shell reduced the melting time by 67%. Cao et al. [41] inves­
in this manner to get the advantage of both conduction and convection tigated the effect of natural convection on the performance of a lauric
dominances of horizontal arrangement and vertical arrangement, acid based double pipe LHTESS by varying the vertical eccentricity of
respectively. The melting time was reduced by 25.7% and 60% the inner HTF tube. Experiments were performed with 10, 20, and 30
compared with horizontal and vertical pipe type LHTESS, respectively. mm eccentricities and for different HTF temperatures. Simulations were
Gorzin et al. [38] investigated the solidification performance of copper also performed using the enthalpy-porosity model. The 30 mm down­
nanoparticles enhanced PCM in a triple concentric tube type LHTESS. ward eccentricity resulted in a 57% decrease in the melting time. The
Different computational cases with different distributions of PCM in the average heat transfer coefficient was higher for the cases with down­
outer annulus and the inner tube were investigated. The HTF was ward eccentricities. It was higher in the initial stage, then it decreased
assumed to flow in the inner annulus between two PCM layers. A and became stable, and lately, it decreased as the temperature was
considerable reduction in solidification time was noticed with uniform. Pahamli et al. [42] investigated the effect of providing eccen­
increasing nanoparticle fraction and different PCM distribution. tricity to the HTF tube of a double pipe LHTESS. The enthalpy-porosity
Many researchers have studied the effect of orientation of the model was used for modelling the phase change. The study concluded
LHTESS. Mahdi et al. [23] investigated the melting behavior of PCM in a that the starting and end of the melting process were governed by
double pipe type LHTESS to study the effect of orientation and finned conduction heat transfer. The downward shifting of the inner tube
configuration. Experiments were performed by filling paraffin wax as provided space for the molten pool’s convection-dominated zone, which
PCM in the annulus and flowing hot water as HTF in the inner pipe. A 3- resulted in reduced melting time. The flow rate didn’t show a significant
D numerical model was designed and simulated using an enthalpy- effect on melting performance. Kadivar et al. [43] optimized the ec­
porosity model. The horizontal orientation performed better than the centricity of HTF tube with four different PCM in a double pipe LHTESS

2
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Table 1
Summary of the literature on MT-LHTESS.
No. Reference Year Type PCM Orientation Geometrical Consideration Key Findings
(Tm˚
,C)
Experimental Numerical Comparison Variation of Variation of
with Single Tube fin
tube arrangements parameters

1. Agyemin 2010 ✔ ✔ Erythritol (117) Horizontal ✔ A 3.5% temperature

et al. [44] gradient in the axial
direction
2. Johar et al. 2016 ✔ Erythritol (117) Vertical 45 tubes with four fins
[45] were used
3. Esapour 2016 ✔ Paraffin Wax Horizontal ✔ ✔ HTF temperature is
et al. [46] (29–35) more influential than
HTF flow rate during
melting
4. Esapour 2016 ✔ Paraffin Wax Horizontal ✔ ✔ Downward shifting of
et al. [47] (29–36) tubes reduced melting
time
5. Anish et al. 2019 ✔ Erythritol (117) Horizontal HTF temperature is
[48] more influential than
HTF flow rate during
melting
6. Joybari 2019 ✔ Paraffin Wax Vertical ✔ Increased surface area
et al. [58] (55–61) and more natural
convection in multiple
tubes case promoted
melting during charging
7✔. Sodhi et al. 2019 ✔ Sodium nitrate Horizontal ✔ Increasing number of
[13] (305.1) tubes reduced the
complete charging-
discharging time
8. Leng et al. 2020 ✔ Lauric acid Horizontal ✔ Axial temperature
[49] (40.72–43.02) gradient reduced by
using multiple tubes
9. Park et al. 2020 ✔ n-eicosane (36) Horizontal ✔ ✔ Downward shifting of
[50] tubes reduced melting
time
10. Shi et al. 2020 ✔ PCM-A16 (15–17) Vertical Magnetic field
[51] with Fe3O4 enhanced the natural
nanoparticles convection, which
shortened the melting
and solidification of
PCM
11. Qaiser 2021 ✔ ✔ Stearic acid Horizontal ✔ ✔ Generation of small
et al. [52] (54–64) vortices due to Y-Fins
and multiple tubes
accelerated melting
12. Mahdi 2021 ✔ Lauric acid Horizontal ✔ Downward shifting of
et al. [53] (44.5–48.3) tubes reduced melting
time
13. Sodhi et al. 2021 ✔ Sodium nitrate Horizontal The axial temperature
[54] (305.1) gradient was marginal
during charging
14. Mateo 2021 ✔ ✔ RT25 (18–25) Vertical Natural convection can
et al. [55] be neglected while
modelling solidification
15. Mahad 2021 ✔ KNO3-NaNO3 Horizontal ✔ ✔ Downward shifting of
et al. [56] (222) tubes and increase in
HTF inlet temperature
reduced melting time
16. Song et al. 2022 ✔ Lauric acid Horizontal ✔ Fins increased melting
[57] (42–48) and the influence area
RT58(55–60) of vortices
NaNO3(37–39)

by analyzing the enthalpy-porosity based simulation results. The performance of four HTF tubes with an equivalent single tube using
downward shifting accelerated the melting, and the upward shifting erythritol as PCM. During melting, the average domain temperature of
accelerated the solidification. The optimum eccentric case melted up to four tubes was higher than the single tube. A 3.5% temperature gradient
7.1 times faster than the concentric case. The study concluded that the was reported in the axial direction of MT-LHTESS. Johar et al. [45]
concentric case was optimum for the complete melting-solidification designed a waste heat recovery MT-LHTESS with 45 tubes, each with
cycle. four fins. The system was able to save 11.33% energy for a stationary
Few researchers have investigated the effect of using multiple HTF engine. Esapour et al. [46] investigated the melting of RT35 in MT-
tubes in LHTESS. A summary of the literature on multi-tube latent heat LHTESS with multiple HTF tubes and an outer HTF annulus. The
thermal energy storage system MT-LHTESS is enlisted in Table 1. melting performance of two, three, and four HTF tubes were compared
Agyenim et al. [44] conducted an experimental study to compare the with a single HTF tube case based on numerically simulated results. The

3
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 1. (a) Schematic of MT-LHTESS; (b) 2-D computational domain; (c) Computational mesh.

vortices generated due to natural convection currents in the molten PCM increased by using multiple tubes. A recommendation to stick to the
and the increased HTF surface area by using multiple HTF tubes reduced downward placement of HTF tubes in PCM was given by the authors. Shi
the melting time to a great extent. Esapour et al. [47] extended the et al. [51] investigated the performance of nanoparticle enhanced PCM
numerical investigation to more cases and studied the distribution of based vertical MT-LHTESS under the influence of a magnetic field
multiple tubes. The study concluded that the downward movement of numerically. The magnetic field enhanced both the melting and solidi­
inner HTF tubes reduced the melting time regardless of the number of fication rate. Qaiser et al. [52] numerically investigated the multiple
tubes. Anish et al. [48] investigated the performance of a finned tube HTF tubes with Y-fins for melting stearic acid in a shell and tube MT-
MT-LHTESS experimentally. HTF temperature was found more influ­ LHTESS. A base case with the single finned tube was compared with
ential as compared to the HTF flow rate. Sodhi et al. [13] optimized the multiple finned tube arrays containing two, three, four, and five HTF
number of tubes in an MT-LHTESS. It was found that the increasing tubes case keeping the constant PCM volume in all cases. Initially, for
number of tubes increased the natural convection, which promoted the first 5 min, all the cases performed almost identically due to con­
melting and reduced the conduction resistance, which promoted the duction dominance. Then the generation of small vortices occurred,
solidification of PCM. Leng et al. [49] investigated the charging and which enhanced the convection heat transfer in the PCM domain. Then
discharging of two different configurations of MT-LHTESS, the one with in the later stages, the dominance of vortices got reduced, and the
single and the other with four HTF tubes for the same quantity of PCM convection effects weakened. The researchers also modified the shell
and HTF flow rate. By using four tubes, the charging and discharging shape to elliptical and triangular and found that the melting dynamics
time was reduced greatly. The axial temperature gradient for four tubes improved. Mahdi et al. [53] investigated the enhancement potential of
was less as compared to the single tube. The theoretical efficiency using different geometrical arrangements of three HTF tubes in a shell
increased to 90.34% for four tubes case as compared to 75.88% for and tube MT-LHTESS. The different arrays of HTF tubes such as an
single tube case. Park et al. [50] investigated the effect of the number of inline-central array, upward and downward triangular arrays, and one
multiple HTF tubes and their arrangements in an MT-LHTESS numeri­ with all three tubes arranged as a circular arc at the bottom of the shell
cally. The melting dynamics of n-eicosane as PCM were observed with were compared for melting lauric acid. The downward shifting of HTF
one, two, three, four, and eight HTF tubes. A new parameter was tubes reduced the melting time significantly. Sodhi et al. [54] investi­
introduced to correlate the melting time with geometric parameters. The gated a shell and tube type MT-LHTESS experimentally for energy
study concluded that the heat transfer surface area and energy density storage in high temperature (400˚C) applications using sodium nitrate as

4
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 2. Cross-sections of various MT-LHTESS designs proposed in this study.

PCM. An array of 25 HTF tubes with hot air as HTF was used. More this study represents thirteen different cases of a horizontal shell and
number of HTF tubes helped in reducing the size of LHTESS and tube type MT-LHTESS. The melting dynamics are analyzed by
improved the melting and storage dynamics due to increased heat comparing the contours of liquid fraction, temperature, and flow
transfer surface area and increased natural convection effects. Mateo streamlines. The melting performance is evaluated based on the tran­
et al. [55] tested natural convection’s effect on melting and solidifica­ sient variation of liquid fraction and average domain velocity, melting
tion of a PCM in a finned tube MT-LHTESS. They concluded that the time, heat storage capacity, and heat storage rate. The results presented
effect of natural convection could be neglected while modelling the in this study are intended to provide a theoretical guide for designing an
solidification to save computation efforts. But natural convection MT-LHTESS for practical solar thermal applications. The explicit and
couldn’t be neglected while modelling the melting of PCM. Mahad et al. detailed investigation of melting dynamics in multiple tube arrange­
[56] investigated the performance of a solar salt based MT-LHTESS ments (arrays), namely the staggered and inline for different eccen­
numerically. The melting performance of two, three, and four HTF tricities and fins, is the novelty of this work.
tubes were compared with a single HTF tube case based on numerically
simulated results. The increasing number of HTF tubes and the HTF inlet 2. System description and computational cases
temperature reduced the melting time. The melting time reduced to 165
min by using four-tubes as compared to 240 min for the single tube case. Fig. 1(a) shows the schematic of the proposed MT-LHTESS. The unit
The study only considered an inline arrangement for four-tube case. length of the system is proposed in this study. The annular space be­
Song et al. [57] investigated the effect of incorporating innovative tween the HTF tubes and the shell is filled with paraffin wax as PCM. The
rectangular and tree type fins on the melting performance of three PCMs HTF is assumed to flow through the HTF tubes for melting the PCM
in an MT-LHTESS with four tubes. In finned tube cases, the melting time while charging the MT-LHTESS. Fig. 2 shows the thirteen different cases
reduced considerably due to the increased size and influence area of proposed in this study. The base case, i.e. Case 0 with a shell of 100 mm
vortices in the liquid PCM. This study also considered only the inline diameter and a single HTF tube of 25 mm diameter. Then the single HTF
arrangement of four tubes. tube is split into four HTF tubes of 12.5 mm diameter, keeping the same
Most of the research conducted on the shell and tube LHTESS quantity of PCM and HTF tube cross-section area. Two different ar­
considered concentric double pipe, concentric triplex tube type config­ rangements of four HTF tubes are designed. A rotated square type
urations, and some geometrical modifications such as eccentric and staggered array with 25 mm and 30 mm eccentricity for Case 1 and Case
finned HTF tube. To the best of the author’s knowledge, a comprehen­ 2, respectively. A simple inline array with 25 mm and 30 mm eccen­
sive study comparing the melting dynamics of a single tube with tricity for Case 3 and Case 4, respectively. Further, four fins of 1 mm
equivalent multiple tubes, the effect of different tube arrangements, thickness and 5 mm height are employed on the surface of HTF tubes in
eccentricity, and different fins configurations has not been done. Thus two different arrangements. Four plus (+ type) fins and cross (× type)

5
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Table 2 1. The PCM was assumed to be homogeneous, isotropic, and with

Geometrical details of various MT- LHTESS designs proposed in this study. constant thermo-physical properties [50].
Case Shell Tube Eccentricity Array Fin 2. The flow rate of HTF was assumed to be sufficiently high to
diameter diameter (e, mm) type type maintain a constant temperature boundary condition at the HTF tube
(D, mm) (d, mm) surface as shown in Fig. 1(c).
Case 0 100 25 – – – 3. The initial temperature of the PCM domain was considered
(Base 328.15 K.
Case) 4. The bulk changes in volume were neglected [27,61].
Case 1 100 12.5 25 Staggered
5. The flow of molten PCM was assumed to be incompressible,
–
Case 1(+) 100 12.5 25 Staggered + type
Case 1(×) 100 12.5 25 Staggered × type laminar, and transient [27,61].
Case 2 100 12.5 30 Staggered – 6. A no-slip boundary condition at the surface of the tube and shell
Case 2(+) 100 12.5 30 Staggered + type was assumed [61].
Case 2(×) 100 12.5 30 Staggered × type
The 2-D transient pressure-based computational fluid dynamics
Case 3 100 12.5 25 Inline –
Case 3(+) 100 12.5 25 Inline + type (CFD) model was designed using ANSYS Fluent 19.0 for melting. Flu­
Case 3(×) 100 12.5 25 Inline × type ent’s melting and solidification model is based on the fixed-grid based
Case 4 100 12.5 30 Inline – enthalpy formulation in which no need is there to track the phase
Case 4(+) 100 12.5 30 Inline + type transition front explicitly [62,63]. The phase transition front is repre­
Case 4(×) 100 12.5 30 Inline × type
sented by assuming it as a mushy zone. This zone is modelled as a pseudo
porous zone with the liquid fraction (porosity) varying from 0 (pure
solid) to 1 (pure liquid). The natural convection effects were taken into
Table 3 consideration by assuming a Boussinesq approximation for density
The thermo-physical properties of paraffin wax [59]. calculation [31,50]. The equations that govern the melting dynamics
Material Property Value (fluid flow and energy transfer) are given as follows [53,64,65]:
Density, ρ, (kg/m3) 870 780 Continuity equation:
(300 K) (340 K)
Specific heat, Cp, (J/kg.K) 2900
∂u ∂v
+ =0 (1)
Thermal Conductivity, k, (W/m.K) 0.22 ∂x ∂y
Dynamic Viscosity, μ, (kg/m.s) 0.002634
Thermal Expansion Coefficient, β, (1/K) 0.00011 Momentum equation (x-direction):
Latent Heat, L, (J/kg) 190,000 ( ) ( 2 )
Solidus Temperature, Ts, (K) 331 ∂u ∂u ∂u ∂P ∂ u ∂2 u
ρ +u +v = − +μ + + Sx (2)
Liquidus Temperature, Tl, (K) 331.8 ∂t ∂x ∂y ∂x ∂x2 ∂y2
Momentum equation (y-direction):
fins are employed on the HTF tubes of MT-LHTESS. The geometrical ( ) ( 2 )
∂v ∂v ∂v ∂P ∂ v ∂2 v
details of all the cases are enlisted in Table 2. Due to the incorporation of ρ +u +v = − +μ + + Sy + ρgβ(T − T0 ) (3)
∂t ∂x ∂y ∂y ∂x2 ∂y2
fins, there is a slight reduction of about 1.08% PCM quantity in finned
cases as compared with the unfinned cases. Multiple HTF tubes and fins Energy equation:
increase the heat transfer surface of MT-LHTESS. There is about a 100% ( ) ( ) ( )
∂H ∂ ∂ ∂ ∂T ∂ ∂T
increase in four tubes and a 304% increase in finned tubes as compared ρ + (uH) + (vH) = k + k (4)
to Case 0. The effects of using multiple tubes and employing fins on ∂t ∂x ∂y ∂x ∂x ∂y ∂y
melting dynamics are compared by taking Case 0 as the base case.
where ρ, β, μ, P, T0, g, t, H, and k are the mass density, coefficient of
The paraffin wax is used as a PCM for the study as it is non-toxic, non-
thermal expansion, dynamic viscosity, pressure, reference temperature,
corrosive, and is thermally stable. Paraffin waxes in the melting range of
acceleration due to gravity, time, total enthalpy, and thermal conduc­
50-60˚C are the most relied on PCM for SDWH [16]. The thermo-physical
tivity respectively. The total enthalpy H is composed of sensible and
properties of paraffin wax are given in Table 3.
latent heat and can be represented as follows [65]:
3. Numerical model H = h + ΔH (5)

A simplified two-dimensional (2-D) computational domain was where h represents the sensible part of enthalpy content can be repre­
considered for the numerical simulations to save computational time sented as follows:
and resources as LHTESS has a very high length to diameter ratio [28]. ∫T
Even this consideration was supported by the observations made by an h = h0 + Cp dT (6)
experimental study [60] and numerical studies [11,12]. Anish et al. [60] T0

reported that due to the horizontal orientation, the axial temperature The latent part of the total energy content can be represented as
difference was insignificant at four different axial locations in 1 m length follows:
of the LHTESS. Agyemin et al. [44] reported a 3.5% temperature
ΔH = λL (7)
gradient in PCM along the axial direction with four HTF tubes, sup­
porting the 2-D consideration in a horizontal MT-LHTESS. Being sym­
where Cp, h0, and L represent the specific heat capacity, enthalpy at the
metric about the vertical central line, only one half of the LHTESS cross-
reference temperature, and latent heat, respectively. The contribution of
section as shown in Fig. 1(b) was considered for saving computation
latent heat in the total heat content depends on the liquid fraction.
efforts as treated in [12,50]. The geometry and meshing of the compu­
As soon as the liquid fraction reaches 0, the velocity of the cell in the
tational domain were created using DesignModeler and Meshing tool of
mushy zone must turn to zero. The Sx and Sy in the momentum equations
ANSYS. A quad-dominated mesh was created for the PCM domain. A
are the terms to represent momentum sink which help in reducing the
body-fitted mesh feature was imposed on HTF tubes using an inflation
porosity of the mushy zone, are calculated as follows [66]:
layer as shown in Fig. 1(c) to capture the near-wall features accurately.
The following assumptions were taken for the numerical model:

6
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 3. (a) Grid independence test; (b) Time step independence test.

Fig. 4. Validation; (a) Contours of Liquid Fraction experimental results by Cao et al. [41] (left half) and numerical results of the current model (right half); (b)
Comparison of liquid fraction variation of the current model and that of Mahdi et al. [66].

7
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 5. Contours showing the melting progress of four HTF tube cases, left half (contours of liquid fraction and flow streamlines), right half (contours of
temperature).

(1 − λ)2 2.30 GHz, 128 GB RAM, 8 GB NVIDIA QUADRO RTX 4000 GPU).
Sx = Amushy 3
u (8)
λ + 0.001

(1 − λ)2 3.1. Grid and time independence test

Sy = Amushy 3
v (9)
λ + 0.001
For reliable computational results, before the actual analysis, the
where λ represents the variation in a liquid fraction which varies from designed model must be independent of the grid size and the time step.
0 to 1 for melting. It depends on the temperature of PCM as [37,67]: Finer grid size gives accurate results but consumes more computational
⎧ ⎫ time. Thus a compromise must be made between these two. So a grid
⎪
⎪ 0 T < Ts ⎪
⎪ independence test was performed by comparing the variation of liquid
⎪
⎨ ⎪
⎬
λ=
T − Ts
Ts < T < Tl (10) fraction for four different grid sizes as shown in Fig. 3(a). A test case of
⎪
⎪
⎪
Tl − Ts ⎪
⎪
⎪ the designed model for concentric tube, i.e. Case 0, was simulated with
⎩ ⎭
1 T > Tl an HTF temperature of 353.15 K (80˚C). The coarser grid with 5382 el­
ements showed about 5% relative deviation as compared to other grids.
Amushy is the mushy zone parameter whose value was set at 105
Grid with 8260 elements showed about 1% relative deviation as
[2,53]. A second-order upwind scheme was selected to discretize mo­
compared to 12,003 elements. The use of a finer grid with 14,070 ele­
mentum and energy equations [31,50]. The PRESTO technique was
ments showed only 0.78% maximum relative error as compared to
employed for the discretization of pressure terms with reference to the
12,003 elements but consumed more computation time and resources.
studies by [27,31,68]. The under-relaxation factors for the momentum,
Finally, mesh size according to 12,003 elements was found sufficient for
pressure, liquid fraction, and thermal energy were set to be 0.7, 0.3, 0.9,
computations in this study.
and 1 respectively [68]. The values of the single time-step convergence
Similarly, the time step was selected by comparing the results of
criterion for continuity, momentum, and energy equation residuals were
using different time steps of 0.05 s, 0.1 s, and 0.2 s on a grid size having
used as 10-3, 10-3, and 10-6 respectively. The computations were per­
12,003 elements. The results are compared as shown in Fig. 3(b). The
formed using HP Z6 Workstation (Intel(R) Xeon(R) Gold 5118 CPU @
liquid fraction is found to be independent and stable for the three chosen

8
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 6. Plots showing melting performance of unfinned MT-LHTESS (a) The variation of liquid fraction; (b) average domain velocity.

time step sizes. Thus 0.1 s time step size is chosen in this study for 4.1. Melting behavior analysis
solving the governing equations.
To assess the melting dynamics of PCM in MT-LHTESS, the melting
behavior of PCM around the HTF tubes is to be studied carefully. The
3.2. Model validation melting behavior of PCM in various tubes configurations is shown in
Fig. 5. The colored depiction of contours of liquid fraction, temperature,
The model designed in this study was validated by comparing the and flow streamlines for the first five cases Case 0 to Case 4 are shown in
results of a test case using the current model with the experimental a single image. The left half depicts both the melting contours and the
investigation by Cao et al. [41]. In Cao’s study, a concentric tube type flow streamlines. The right half depicts the contours of domain tem­
LHTESS with lauric acid as PCM and a PID controlled constant tem­ perature. As it can be seen from the contours, initially a thin circular
perature tube as HTF tube was used. The shell and tubes were 120 mm layer of molten PCM is formed around the HTF tube in all cases. The heat
and 40 mm in diameter. The initial temperature of PCM and the HTF is transferred from the HTF to the immediate surrounding solid PCM
tube temperature was taken as 25˚C and 80˚C, respectively. A qualitative during the initial stage by conduction. Also, the temperature difference
comparison of the melting contours of Cao’s study and the current model between the HTF tube surface and the solid PCM is very high and the
is shown in Fig. 4(a). The melting contours of the current model are in natural convection effects are negligible at the initial stage. After this,
good agreement with the contours of the experimental investigation. A the thin molten layer increases its size in the upward direction above the
slight difference in melting profile can be observed at 120 min. This is HTF tube. This happens because of the temperature difference driven
because the starting of melting during experiments was delayed due to buoyancy effects in the molten PCM around the HTF tube. The hot and
the setting-up time taken by the experimental setup and the thermal lighter PCM ascends upward and vice versa. This upward and downward
contact resistance between the HTF tube wall and the solid PCM. The flow of PCM in the molten pool creates two large vortices one in the left
current numerical model neglected these effects. The same reason was half and the other in the right half around the HTF tubes, which im­
also reported by Cao et al. as they also noticed the same difference at proves the natural convection and the melting dynamics. It can also be
120 min in their numerical contours and experimental contours. The observed that the molten PCM in the upper half is always greater than in
deviation was within the permissible limits. The developed model was the lower half. The melting front for all the cases moves very slowly in
also validated using the numerical results of melting in a rectangular the lower stages of melting and there is a large chunk of solid PCM in the
domain by Mahdi et al. [66]. In Mahdi’s study lauric acid in an 80 mm × lower half even if all the PCM in the upper half has been melted. The
160 mm rectangular cavity with a constant temperature heating wall at solid PCM far from the HTF tubes in the lower half forms a poor melting
65˚C was melted. With an average 5.4% deviation in liquid fraction as zone. The velocity streamlines are closer to each other along the vertical
shown in Fig. 4(b), the numerical results by the current model showed line just above the HTF tube, which indicates a high velocity vertical
good agreement with the results obtained by Mahdi et al. Thus the ascend of PCM.
current model guarantees an accurate and reliable solution based on the
validation with two different studies. 4.2. Effect of using multiple HTF tubes

4. Results and discussion This section discusses the effect of splitting a single HTF tube of 25
mm diameter into four HTF tubes of 12.5 mm diameter. Fig. 5 shows the
To compare the charging performance of the thirteen different cases, colored depiction of contours of liquid fraction, temperature, and flow
the melting behavior of PCM is observed in this section. The 2-D streamlines for comparing the effects of dividing the single HTF tube of
computational cases of MT-LHTESS designed in the previous section Case 0 into four HTF tubes in Case 1 - Case 4. As it can be seen from the
were simulated by employing a constant temperature 353.15 K (80˚C) contours, a thin circular molten PCM layer is formed around each tube of
boundary condition to the HTF surface. The HTF temperature was the MT-LHTESS initially. Each HTF tube seems to have its localized
chosen this value to make the MT-LHTESS fit for employment with the molten pool in the initial stage. This is the effect of conduction heat
solar thermal collectors for low and medium temperature applications transfer from the HTF tube to the immediate contact solid PCM. After
[7,8]. The performance is evaluated based on the transient observations this, the buoyancy-driven convection effect comes into play, and the size
of melting contours, temperature contours, flow streamlines, liquid of each localized molten pool increases in the vertically upward direc­
fraction, average domain velocity, melting time, heat storage capacity, tion above the HTF tubes. As time progresses, the localized molten pool
and heat storage rate in different cases. of individual HTF tubes comes in contact with each other, and the

9
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 7. Contours showing the melting progress of finned four HTF tube cases, left half (contours of liquid fraction and flow streamlines), right half (contours of
temperature).

coalescence of individual molten pools occurs. The coalescence phe­ is important for comparing different configurations of HTF tubes. At 25
nomenon results in larger vortices in the molten pool, strong convection mm eccentricity, this occurs at 25 min and 20 min for Case 1 and Case 3,
currents, and the mixing of different temperature layers of individual respectively. At 30 mm eccentricity, this occurs at 35 min and 30 min for
tubes which accelerate the melting in the upper half, and the solid PCM Case 2 and Case 4, respectively. It is observed that in the inline array
at the top of the domain gets melted quickly. The flow patterns and the cases (Case 3 and Case 4), the coalescence occurs earlier because there
thermal performance is greatly affected by the arrangement of tubes, as are two pairs of HTF tubes having tubes arranged on a common vertical
shown in Fig. 5. After looking at the contours of temperature for Case 1 line. In contrast to it in staggered array cases (Case 1 and Case 2), there is
to Case 4 at 120 min, it is evident that the temperature of PCM in the only one pair, and even the vertical distance between these tubes of such
upper half reached the same value as that of the HTF tube surface. Thus pair is more prominent than that in an inline case. The quantitative
the thermal gradients in the domain get reduced, and so does the melting assessment of different cases is done by comparing the transient plots of
rate. Due to this thermal stratification, the tubes in the upper half of the liquid fraction and average domain velocity of the PCM domain which
PCM domain become inactive for charging the MT-LHTESS in the later are shown in Fig. 6. It is quite clear that the four HTF tube cases
stages. outperform the single tube case considerably. The reason for this is the
100% increase in the HTF surface area. The effect can be seen in the
4.3. Effect of tube configurations initial stage as at 10 min, the liquid fraction increases to double from
0.129 for Case 0 to 0.258 for all four tube cases, as shown in Fig. 6(a). Up
This section compares the performance of four different arrange­ to this stage, the melting is governed by conduction which is a function
ments of HTF tubes in Case 1, Case 2, Case 3, and Case 4. The perfor­ of HTF surface area. Thus the liquid fraction gets doubled in the four-
mance is assessed based on the contours shown in Fig. 5. The previous tube cases. The staggered arrangement outperforms the inline arrange­
section discussed the acceleration of melting by the coalescence of in­ ment after the coalescence of molten pools. But the liquid fraction plot
dividual molten pools. Thus the time of occurrence of this phenomenon for inline cases Case 3 and Case 4 touches the plot of staggered cases

10
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 8. Flow characteristics for (a) unfinned tube; (b) (+ type) fins; (c) (× type) fins at 10 min.

Case 1 and Case 2 at 60 min and 50 min respectively and outperforms for effects dampen the vortices and thus the velocity magnitude decreases.
a few minutes. This happens because of more molecular travel activity in After 90 min the velocity reduces to very low magnitudes and the con­
the molten domain in the inline case due to two vortices near the melting vection effects start to vanish. Some fluctuations in velocity can also be
front as shown in streamlines at 60 min for Case 3 and Case 4 as seen which are caused by the fluctuations in the local velocity because of
compared to the single vortice in Case 1 and Case 2. After this, the the coalescence and dampening of local vortices.
staggered cases outperform the inline cases because of a similar reason,
new vortices get generated near the melting interface in Case 1 and Case 4.4. Effect of using fins on multiple HTF tubes
2, and the two vortices in inline cases merge into one. Thus for complete
melting, the staggered tubes outperformed the inline tubes array. To further reduce the melting time of MT-LHTESS, four longitudinal
The effect of eccentricity is also important while designing the tube fins of 5 mm height in two different configurations, namely plus (+ type)
configurations for MT-LHTESS. Up to 30 min, the melting in all the four- and cross (× type), are employed. A previous study [69] suggested using
tube cases is almost the same as flow behavior, and heat transfer dy­ four fins as no significant improvement was shown by increasing to more
namics around individual tubes were independent of the neighboring number of fins. Fig. 7 shows the colored depiction of contours of liquid
tubes. Thus it can be inferred that the melting dynamics of PCM for the fraction, temperature, and flow streamlines for comparing the effects of
same number of HTF tubes are independent of the tube arrangement using the two types of fins on the HTF tubes of Case 1 and Case 3. Case 1
before the coalescence of their molten pools in the domain. From Fig. 6 (+) and Case 1(×) have staggered arranged tubes with (+ type) and (×
(a), the cases with 30 mm eccentricity outperformed the cases with 25 type) fins, respectively, and Case 3(+) and Case 3(×) have inline ar­
mm eccentricity. This is because by providing the downward eccen­ ranged tubes with (+ type) and (× type) respectively all with 25 mm
tricity, the scope for the buoyancy-driven ascends of hot PCM increases, eccentricity. Like in unfinned cases, during the initial stages, the
which generates faster and hotter PCM currents in PCM domain, which conduction-dominated melting occurs. The heat transfer takes place
melts more solid PCM at the melting interface. This was also reported in from the HTF tube surface to the immediate surrounding solid PCM. For
many studies [27,43] that by providing the downward eccentricity, the the first 10 min, an almost equal amount of PCM gets melted as the HTF
melting performance improves. Thus if we consider only the hard to surface area is the same for these four cases. After that, the buoyancy-
melt half, i.e. the lower half of the PCM domain, the more we provide the driven convection effects come into play and the liquid fraction in­
eccentricity more will be the melting performance. The thermal strati­ creases in the upward direction, and the generation of small vortices
fication near the melting interface is also reduced by a downward ec­ takes place. These vortices accelerate the melting process. The flow
centricity which favors melting. This dominance of downward characteristics around the unfinned tube, (+ type) finned tube, and (×
eccentricity in the lower half of the domain is reflected in the overall type) finned tube is shown in Fig. 8. It is interesting to verify the upward
melting performance of the cases with 30 mm eccentricity. For complete motion of hot particles and the downward motion of cold particles as the
melting, it took 162 min for Case 1, 130 min for Case 2, 210 min for Case velocity vectors close to the hot surface (HTF tube) are in an upward
3, and 160 min for Case 4, respectively. Thus, the melting performance is direction and downward direction near the cold surface (solid PCM). It
improved by using multiple tubes in a staggered arrangement with a can also be seen that at 10 min there is more violent movement in the
larger eccentricity. The average velocity trend for unfinned tube MT- case of finned tubes than that of the unfinned tube. In the unfinned case,
LHTESS is shown in Fig. 6(b). It shows the peak domain velocity and there are only convection currents in the form of two big vortices on
the instance of maximum convection effects in the molten PCM in Case either side of the tube, but in the case of finned tubes, Rayleigh-Benard
0 to Case 4. Even more violent motion of molten PCM is evident for convection currents are generated due to the combined effect of buoy­
multi-tube cases as the velocity magnitude is more for Case 1 to Case 4 as ancy and restrictions created due to the fins in the liquid PCM. These
compared to Case 0. In multi-tube cases, the velocity is incredibly close Benard cells are shown in Fig. 8(b),(c). Depending on the intensity of
at the start of the melting and it increases in the same manner as the velocity and strength of vortices, the area around the tubes is divided
melting is almost identical around the individual tubes. Then the coa­ into two zones the upper zone as convection dominated zone and the
lescence of molten pools occurs and the magnitude of velocity increases lower zone as conduction dominated zone. Interestingly the convection
and reaches a maximum value. After this, the thermal stratification currents around the (+ type) fins are stronger than the (× type) fins. It is

11
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 9. Plots showing melting performance of finned MT-LHTESS (a) The variation of the liquid fraction of staggered array; (b) Average domain velocity of the
staggered array; (c) The variation of the liquid fraction of inline array; (d) Average domain velocity.

because the orientation (45◦ ) of (× type) fins restricts the upward mo­ peak is shifted towards the left as the quantity of molten PCM is more in
tion of the molten PCM and prevents the generation of larger and strong finned tube cases as compared to the unfinned cases. Just after coales­
vortices in the molten pool. Similar observations on flow characteristics cence i.e. 10 min more violent motion of molten PCM is evident for
were made in a previous study [70]. The number of fins and their ori­ finned tube cases as the velocity magnitude is more than unfinned tube
entations were studied in [69], and four fins (+ type) performed better cases. As the peak is reached the thermal stratification effects dampen
than (× type) fins. This can be verified from the results shown in Fig. 9 the vortices and the velocity magnitude decreases. After 60 min the
(a),(c) as the melting is faster in (+ type) fins as compared to (× type) velocity reduces to very low magnitudes and the convection effects start
fins for the same eccentricity and configuration of tubes. The average to vanish. Some fluctuations in velocity can also be seen which are
velocity trend for finned MT-LHTESS is shown in Fig. 9(b),(d). It shows caused by the fluctuations in the local velocity because of the coales­
the peak domain velocity and the instance of maximum convection ef­ cence and dampening of local vortices. A similar trend of average
fects in the molten PCM in finned tube cases. In finned tube cases, the domain velocity for finned tubes was reported by Li et al. [70]. Initially,

Fig. 10. Melting time taken for complete melting of PCM in all the cases.

12
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Fig. 11. The sensible, latent, and total heat stored in all cases.

the velocity magnitude was incredibly close for all cases, then it reached decrease in melting time was achieved by increasing eccentricity in both
a maximum value as a result of Rayleigh-Benard convection currents, inline and staggered arrangements of finned tubes. The (+ type) fins
and later it decreased as a combined effect of thermal stratification and outperformed the (× type) fins by saving about 10 min for the same
dampening of convection currents. eccentricity and tube arrangement. Interestingly there is not much dif­
ference (4%) in melting time for Case 2 with Case 1(×) and Case 4 with
4.5. Melting time and heat stored in all the cases Case 3(×). Thus instead of applying fins (× type) on Case 1 and Case 3, a
similar melting performance can be achieved by increasing the eccen­
Fig. 10 shows the melting time for all the designs of LHTESS inves­ tricity from 25 mm to 30 mm.
tigated in this study. The base case Case 0 with a single HTF tube took To assess the charging performance of LHTESS designs proposed in
the maximum melting time of 450 min. As the single HTF tube is divided this study it is important to compare their heat storage capacity (q). The
into four HTF tubes, the melting time is reduced considerably. The total energy stored by the LHTESS is the sum of both latent heat and
minimum melting time among the four unfinned tube cases is for Case 2, sensible heat. The latent heat stored is a function of the liquid fraction of
i.e. the case with a staggered arrangement of the four tubes with 30 mm the PCM domain. The sensible heat is stored in the PCM in solid and
eccentricity, which took 130 min (71.1% less than Case 0) for complete liquid states. The sensible heat is a function of specific heat capacity and
melting. The maximum melting time is for Case 3, i.e. the case with an the temperature difference of the PCM domain with its initial temper­
inline arrangement of the four tubes with 25 mm eccentricity, which ature. The initial temperature in this study is taken equal to the refer­
took 210 min (53.3% less than Case 0) for complete melting. By ence temperature at 328.15 K. The sensible heat, latent heat, and the
increasing the eccentricity of four tubes from the shell’s center, the total heat stored are calculated per meter length (kJ/m) using the
melting performance improves for both inline and staggered arrange­ equation (11), (12).
ments. About a 20% decrease in melting time was achieved by [ ]
ρAc cp (T − Ti ) + λinst .L
increasing eccentricity in both inline and staggered arrangements of qinst = (11)
1000
unfinned tubes.
By employing fins on four HTF tubes, the melting time was reduced ∫ tc
significantly. The minimum melting time among the four finned tube qtotal = qinst dt (12)
cases is for Case 2(+), i.e. the case with a staggered arrangement of the
0

four (+ type) finned tubes with 30 mm eccentricity, which took 75 min qtotal
q̇ = × 1000 (13)
(83.33% less than Case 0) for complete melting. The maximum melting tc
time is for Case 3(×), i.e. the case with an inline arrangement of the four
(× type) finned tubes with 25 mm eccentricity, which took 154 min where ρ is the density of PCM, Ac is the area of a cross-section of the
(65.78% less than Case 0) for complete melting. Increasing the eccen­ PCM, Tinst is the average PCM domain temperature at an instant, Ti is the
tricity of four finned tubes from the shell’s center improved the melting initial temperature, λinst is the liquid fraction at an instant, tc is the time
performance for both inline and staggered arrangements. About a 33% taken for complete melting.

Fig. 12. The average heat storage rate in all cases.

13
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

Table 4
Total energy stored and theoretical efficiency.
Case 0 Case 1 Case 1(+) Case 1(×) Case 2 Case 2(+) Case 2(×) Case 3 Case 3(+) Case 3(×) Case 4 Case 4(+) Case 4(×)

qtotal 1433.0 1463.1 1456.9 1443.9 1461.9 1449.0 1449.4 1466.6 1444.1 1443.0 1463.6 1450.4 1448.5
qmax 1507.6 1507.6 1491.2 1491.2 1507.6 1491.2 1491.2 1507.6 1491.2 1491.2 1507.6 1491.2 1491.2
ηth 95.1 97.0 97.7 96.8 97.0 97.2 97.2 97.3 96.8 96.8 97.1 97.3 97.1

As the total melting time was different for all the cases, for quanti­ compared in this study. At first, the performance of four cases with four
tative assessment, the total heat storage for 60 min (1 h) melting of PCM HTF tubes has been compared with a single HTF tube case. Then four
is compared and is shown in Fig. 11. It can be seen that the energy stored longitudinal fins of two types are incorporated to further improve the
in the single tube case Case 0 is the least as the heat transfer surface area melting. After analyzing the numerical results, the following conclusions
is least and the melting and heat diffusion is weak. Splitting a single HTF are drawn;
tube into four tubes comparatively resulted in an average 100% Splitting the single tube into four tubes improves the melting dy­
improvement in the total heat storage capacity. In the four unfinned namics. Initially, the conduction effects increase due to a 100% increase
tube cases, Case 1 to Case 4, the sensible content of heat stored is almost in heating surface area. After that, the coalescence of the localized
the same, i.e. approximately 295 kJ/m. The latent content makes the molten pool of each tube creates a more violent mixing of the different
difference in the storage capacity, which is clearly dependent on the temperature layers in the molten pool. A 53.3–71.1% reduction in
eccentricity as the higher eccentricity cases Case 2 and Case 4 stored melting time takes place using four tubes as compared to single tube
more heat than Case 1 and Case 3, respectively. Thus the total heat Case 0.
stored varies in accordance with the liquid fraction in the unfinned The tube arrangement significantly affects the melting dynamics.
cases. In multiple tubes and finned tube cases, the sensible heat content The staggered array of four tubes outperforms the inline array. The
increases as there are more natural convection effects. The (+ type) larger eccentricity of multiple tubes improves the melting dynamics.
cases stored more heat than the (× type) cases for the same eccentricity Almost 33% faster melting is achieved by increasing eccentricity from
and arrangement. Relatively this difference is more significant for the 25 mm to 30 mm for both type of arrays. As a result, Case 2 i.e. staggered
staggered arrangement in Case 1 and Case 2. The maximum heat storage array with 30 mm eccentricity took the least melting time i.e. 130 min
capacity (1411.1 kJ/m) was in Case 2(+) for 60 min of charging. among all the unfinned tube cases.
To get a clearer picture of the performance of MT-LHTESS, a quali­ The incorporation of fins accelerates the melting. Initially, the con­
tative assessment of heat storage is also required. To do so, the average duction effects increase due to a 304% increase in heating surface area.
heat storage rate (q̇) in W/m for complete melting is calculated using the After that, more violent movement of molten PCM as a combined effect
equation (13) and is shown in Fig. 12. The lowest q̇ is for Case 0, the coalescence of localized molten pools and Rayleigh-Benard cells takes
reason being the lowest heat transfer surface. For the same eccentricity, place. A 65.78%-83.33% reduction in melting time takes place using
30% more q̇ is observed in staggered Case 1 and Case 2 compared to four finned tubes compared to Case 0. The (+ type) fins outperform the
inline Case 3 and Case 4 arrangement respectively for unfinned tubes. (× type) fins for the same eccentricity and tube arrangement.
For the same arrangement, 30% more q̇ is observed in 30 mm eccen­ For quantitative assessment, the heat storage capacity of different
tricity Case 2 and Case 4 compared to 25 mm Case 1 and Case 3, cases is compared which depends more on the latent heat which is a
respectively. In finned MT-LHTESS, the staggered cases for the same function of liquid fraction. As a result, staggered array, with larger ec­
eccentricity and larger eccentricity for the same arrangements have centricity, and (+ type) fin on Case 2(+) has a maximum heat storage
more q̇. The (+ type) fins cases have more q̇ as compared to the (× type) capacity of 1411.1 kJ/m for 60 min of charging of MT-LHTESS.
cases. The maximum q̇ is found for Case 2(+), i.e. charged at a rate of For qualitative assessment, the heat storage rate (q̇) of different cases
322 W/m. is compared. About 30% more q̇ is observed in the staggered array and
To assess the effectiveness of heat storage, the theoretical efficiency 30 mm eccentricity cases as compared to their counterparts for both
ηth was also calculated using the equation (14). unfinned and finned tubes. The (+ type) fin cases have more q̇ as
qtotal compared to the (× type) cases. The maximum q̇ is reported for Case 2
ηth = (14) (+), i.e. charged at a rate of 322 W/m, which is five times more than 53
qmax
W/m for Case 0.
where qmax is the maximum heat storage capacity of the different ar­ The proposed designs have a theoretical heat storage efficiency of
rangements used in this study. It is calculated using equation (15). It is 95.1% and 97% for single and multiple tube cases, respectively.
taken as the total heat stored in the MT-LHTESS when it is completely The findings of this study can be used as a theoretical guide for
melted (λinst = 1) and the whole PCM domain is in thermal equilibrium designing an MT-LHTESS for practical solar thermal applications such as
with the HTF (T = THTF ). pre-heating, drying, washing, cooking, pasteurization, bleaching,
[ ] SDWH, etc. The findings on tube arrangement can be used for designing
qmax =
ρAc cp (THTF − Ti ) + L
(15) the upscaled and modular TES systems. The parallel and series ar­
1000 rangements for modular TES systems are discussed in [12].
The calculated values of qtotal and ηth are given in Table 4. It can be
seen that in the proposed designs the ηth for single tube is 95.1% and it Declaration of Competing Interest
increased to approximately 97% for both unfinned and finned four-tube
cases. The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
5. Conclusions the work reported in this paper.

This numerical study investigates the melting dynamics of paraffin Acknowledgements

wax as PCM in a multi-tube latent heat thermal energy storage system
(MT-LHTESS). A two-dimensional enthalpy-porosity model was The authors gratefully acknowledge the support provided by The
designed and simulated using ANSYS-Fluent. The model was validated, Department of Mechanical Engineering, Punjab Engineering College,
and using that thirteen different cases with different configurations are Chandigarh, India.

14
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

References [26] J. Guo, X. Wang, B. Yang, X. Yang, and M. J. Li, Thermal assessment on solid-liquid
energy storage tube packed with non-uniform angled fins, Sol. Energy Mater. Sol.
Cells, vol. 236, no. November 2021, p. 111526, 2022, doi: 10.1016/j.
[1] M. Zhao, Y. Tian, M. Hu, F. Zhang, M. Yang, Topology optimization of fins for
solmat.2021.111526.
energy storage tank with phase change material, Numer. Heat Transf. Part A Appl.
[27] V. Safari, H. Abolghasemi, L. Darvishvand, B. Kamkari, Thermal performance
77 (3) (2020) 284–301, https://doi.org/10.1080/10407782.2019.1690338.
investigation of concentric and eccentric shell and tube heat exchangers with
[2] J. Guo, Z. Liu, B. Yang, X. Yang, J. Yan, Melting assessment on the angled fin design
different fin configurations containing phase change material, J. Energy Storage
for a novel latent heat thermal energy storage tube, Renew. Energy 183 (2022)
vol. 37, no. March (2021), 102458, https://doi.org/10.1016/j.est.2021.102458.
406–422, https://doi.org/10.1016/j.renene.2021.11.007.
[28] Z.J. Zheng, Y. Xu, M.J. Li, Eccentricity optimization of a horizontal shell-and-tube
[3] H. Naik, P. Baredar, A. Kumar, Medium temperature application of concentrated
latent-heat thermal energy storage unit based on melting and melting-solidifying
solar thermal technology: Indian perspective, Renew. Sustain. Energy Rev. 76 (Sep.
performance, Appl. Energy, vol. 220, no. August 2017, pp. 447–454, 2018, doi:
2017) 369–378, https://doi.org/10.1016/j.rser.2017.03.014.
10.1016/j.apenergy.2018.03.126.
[4] S. Kalogirou, The potential of solar industrial process heat applications, Appl.
[29] M. Yusuf Yazici, M. Avci, O. Aydin, M. Akgun, Effect of eccentricity on melting
Energy 76 (4) (2003) 337–361, https://doi.org/10.1016/S0306-2619(02)00176-9.
behavior of paraffin in a horizontal tube-in-shell storage unit: An experimental
[5] A.K. Sharma, C. Sharma, S.C. Mullick, T.C. Kandpal, Solar industrial process
study, Sol. Energy 101 (2014) 291–298, https://doi.org/10.1016/j.
heating: A review, Renew. Sustain. Energy Rev. 78 (January) (2017) 124–137,
solener.2014.01.007.
https://doi.org/10.1016/j.rser.2017.04.079.
[30] S. Ebadi, K. Venkateshwar, S.H. Tasnim, S. Mahmud, A. Dutta, Charging
[6] C.A. Schoeneberger, C.A. McMillan, P. Kurup, S. Akar, R. Margolis, E. Masanet,
nanoparticle enhanced bio-based PCM in open cell metallic foams: An
Solar for industrial process heat: A review of technologies, analysis approaches,
experimental investigation, Appl. Therm. Eng., vol. 148, no. July 2018, pp.
and potential applications in the United States, Energy 206 (2020), 118083,
1029–1042, Feb. 2019, doi: 10.1016/j.applthermaleng.2018.11.121.
https://doi.org/10.1016/j.energy.2020.118083.
[31] Y.H. Diao, L. Liang, Y.H. Zhao, Z.Y. Wang, F.W. Bai, Numerical investigation of the
[7] L. Kumar, M. Hasanuzzaman, N.A. Rahim, Global advancement of solar thermal
thermal performance enhancement of latent heat thermal energy storage using
energy technologies for industrial process heat and its future prospects: A review,
longitudinal rectangular fins and flat micro-heat pipe arrays, Appl. Energy
Energy Convers. Manag. 195 (February) (2019) 885–908, https://doi.org/
233–234 (Jan. 2019) 894–905, https://doi.org/10.1016/j.apenergy.2018.10.024.
10.1016/j.enconman.2019.05.081.
[32] S. Rauf, S.K. Saha, Thermal performance of multitube latent heat storage using a
[8] Z. Wang, W. Yang, F. Qiu, X. Zhang, X. Zhao, Solar water heating: From theory,
metal matrix for solar applications: Numerical study, Heat Transf. Res. 50 (6)
application, marketing and research, Renew. Sustain. Energy Rev. 41 (2015)
(2019) 545–564, https://doi.org/10.1615/HeatTransRes.2018024568.
68–84, https://doi.org/10.1016/j.rser.2014.08.026.
[33] G. Wang, G. Wei, C. Xu, X. Ju, Y. Yang, X. Du, Numerical simulation of effective
[9] R. Saxena, K. Biplab, D. Rakshit, Quantitative Assessment of Phase Change Material
thermal conductivity and pore-scale melting process of PCMs in foam metals, Appl.
Utilization for Building Cooling Load Abatement in Composite Climatic Condition,
Therm. Eng. 147 (Jan. 2019) 464–472, https://doi.org/10.1016/j.
J. Sol. Energy Eng. 140 (1) (2017) Oct, https://doi.org/10.1115/1.4038047.
applthermaleng.2018.10.106.
[10] M. Rahimi, S.S. Ardahaie, M.J. Hosseini, M. Gorzin, Energy and exergy analysis of
[34] X. Yang, F. Xu, X. Wang, J. Guo, M.J. Li, Solidification in a shell-and-tube thermal
an experimentally examined latent heat thermal energy storage system, Renew.
energy storage unit filled with longitude fins and metal foam: A numerical study,
Energy 147 (2020) 1845–1860, https://doi.org/10.1016/j.renene.2019.09.121.
Energy Built Environ., no. May, 2021, doi: 10.1016/j.enbenv.2021.08.002.
[11] X. Huang, S. Yao, X. Yang, R. Zhou, Melting performance assessments on a triplex-
[35] J. Guo, Z. Du, G. Liu, X. Yang, M.J. Li, Compression effect of metal foam on melting
tube thermal energy storage system: Optimization based on response surface
phase change in a shell-and-tube unit, Appl. Therm. Eng., vol. 206, no. December
method with natural convection, Renew. Energy 188 (2022) 890–910, https://doi.
2021, p. 118124, 2022, doi: 10.1016/j.applthermaleng.2022.118124.
org/10.1016/j.renene.2022.02.035.
[36] M.J. Hosseini, A.A. Ranjbar, K. Sedighi, M. Rahimi, A combined experimental and
[12] X. Yang, X. Wang, Z. Liu, X. Luo, and J. Yan, Effect of fin number on the melting
computational study on the melting behavior of a medium temperature phase
phase change in a horizontal finned shell-and-tube thermal energy storage unit,
change storage material inside shell and tube heat exchanger, Int. Commun. Heat
Sol. Energy Mater. Sol. Cells, vol. 236, no. November 2021, p. 111527, 2022, doi:
Mass Transf. 39 (9) (2012) 1416–1424, https://doi.org/10.1016/j.
10.1016/j.solmat.2021.111527.
icheatmasstransfer.2012.07.028.
[13] G.S. Sodhi, K. Vigneshwaran, A.K. Jaiswal, P. Muthukumar, Assessment of Heat
[37] M.S. Mahdi, H.B. Mahood, J.M. Mahdi, A.A. Khadom, A.N. Campbell, Improved
Transfer Characteristics of a Latent Heat Thermal Energy Storage System: Multi
PCM melting in a thermal energy storage system of double-pipe helical-coil tube,
Tube Design, Energy Procedia 158 (March) (Feb. 2019) 4677–4683, https://doi.
Energy Convers. Manag. 203 (Jan. 2020), 112238, https://doi.org/10.1016/j.
org/10.1016/j.egypro.2019.01.737.
enconman.2019.112238.
[14] S.M. Shalaby, A.E. Kabeel, B.M. Moharram, A.H. Fleafl, Experimental study of the
[38] M. Gorzin, M.J. Hosseini, M. Rahimi, R. Bahrampoury, Nano-enhancement of
solar water heater integrated with shell and finned tube latent heat storage system,
phase change material in a shell and multi-PCM-tube heat exchanger, J. Energy
J. Storage Mater. 31 (2020) 101628.
Storage, vol. 22, no. December 2018, pp. 88–97, 2019, doi: 10.1016/j.
[15] H. Xu, J.Y. Sze, A. Romagnoli, X. Py, Selection of Phase Change Material for
est.2018.12.023.
Thermal Energy Storage in Solar Air Conditioning Systems, Energy Procedia 105
[39] N. Kousha, M.J. Hosseini, M.R. Aligoodarz, R. Pakrouh, R. Bahrampoury, Effect of
(2017) 4281–4288, https://doi.org/10.1016/j.egypro.2017.03.898.
inclination angle on the performance of a shell and tube heat storage unit – An
[16] M. H. Abokersh, “Review of the phase change material (PCM) usage for solar
experimental study, Appl. Therm. Eng. 112 (2017) 1497–1509, https://doi.org/
domestic water heating systems (SDWHS),” 2017, doi: 10.1002/er.
10.1016/j.applthermaleng.2016.10.203.
[17] A. Sharma, V.V. Tyagi, C.R. Chen, D. Buddhi, Review on thermal energy storage
[40] I. Al Siyabi, S. Khanna, T. Mallick, S. Sundaram, An experimental and numerical
with phase change materials and applications, Renew. Sustain. Energy Rev. 13 (2)
study on the effect of inclination angle of phase change materials thermal energy
(Feb. 2009) 318–345, https://doi.org/10.1016/j.rser.2007.10.005.
storage system, J. Storage Mater. 23 (2019) 57–68.
[18] K.A. Aly, A.R. El-Lathy, M.A. Fouad, Enhancement of solidification rate of latent
[41] X. Cao, Y. Yuan, B. Xiang, F. Haghighat, Effect of natural convection on melting
heat thermal energy storage using corrugated fins, J. Energy Storage 24 (Aug.
performance of eccentric horizontal shell and tube latent heat storage unit, Sustain.
2019), 100785, https://doi.org/10.1016/j.est.2019.100785.
Cities Soc. 38 (Apr. 2018) 571–581, https://doi.org/10.1016/j.scs.2018.01.025.
[19] K. Yang, N. Zhu, C. Chang, D. Wang, S. Yang, S. Ma, A methodological concept for
[42] Y. Pahamli, M.J. Hosseini, A.A. Ranjbar, R. Bahrampoury, Analysis of the effect of
phase change material selection based on multi-criteria decision making (MCDM):
eccentricity and operational parameters in PCM-filled single-pass shell and tube
A case study, Energy 165 (2018) 1085–1096, https://doi.org/10.1016/j.
heat exchangers, Renew. Energy 97 (2016) 344–357, https://doi.org/10.1016/j.
energy.2018.10.022.
renene.2016.05.090.
[20] S.K. Sahoo, P. Rath, M.K. Das, Solidification of phase change material
[43] M.R. Kadivar, M.A. Moghimi, P. Sapin, C.N. Markides, Annulus eccentricity
nanocomposite inside a finned heat sink: A macro scale model of nanoparticles
optimisation of a phase-change material (PCM) horizontal double-pipe thermal
distribution, J. Therm. Sci. Eng. Appl. 11 (4) (2019) 1–11, https://doi.org/
energy store, J. Energy Storage vol. 26, no. July (2019), 101030, https://doi.org/
10.1115/1.4043596.
10.1016/j.est.2019.101030.
[21] F. Li, M. Sheikholeslami, R.N. Dara, M. Jafaryar, A. Shafee, T. Nguyen-Thoi, Z. Li,
[44] F. Agyenim, P. Eames, M. Smyth, Heat transfer enhancement in medium
Numerical study for nanofluid behavior inside a storage finned enclosure involving
temperature thermal energy storage system using a multitube heat transfer array,
melting process, J. Mol. Liq. 297 (2020), https://doi.org/10.1016/j.
Renew. Energy 35 (1) (2010) 198–207, https://doi.org/10.1016/j.
molliq.2019.111939.
renene.2009.03.010.
[22] R.P. Singh, H. Xu, S.C. Kaushik, D. Rakshit, A. Romagnoli, Charging performance
[45] D.K. Johar, D. Sharma, S.L. Soni, P.K. Gupta, R. Goyal, Experimental investigation
evaluation of finned conical thermal storage system encapsulated with nano-
on latent heat thermal energy storage system for stationary C.I. engine exhaust,
enhanced phase change material, Appl. Therm. Eng., vol. 151, no. September 2018,
Appl. Therm. Eng. 104 (2016) 64–73, https://doi.org/10.1016/j.
pp. 176–190, 2019, doi: 10.1016/j.applthermaleng.2019.01.072.
applthermaleng.2016.05.060.
[23] M.S. Mahdi, A.F. Hasan, H.B. Mahood, A.N. Campbell, A.A. Khadom, A.M.
[46] M. Esapour, M.J. Hosseini, A.A. Ranjbar, Y. Pahamli, R. Bahrampoury, Phase
A. Karim, A.O. Sharif, Numerical study and experimental validation of the effects
change in multi-tube heat exchangers, Renew. Energy 85 (2016) 1017–1025,
of orientation and configuration on melting in a latent heat thermal storage unit,
https://doi.org/10.1016/j.renene.2015.07.063.
J. Energy Storage 23 (2019) 456–468.
[47] M. Esapour, M.J. Hosseini, A.A. Ranjbar, R. Bahrampoury, Numerical study on
[24] R. Karami, B. Kamkari, Experimental investigation of the effect of perforated fins
geometrical specifications and operational parameters of multi-tube heat storage
on thermal performance enhancement of vertical shell and tube latent heat energy
systems, Appl. Therm. Eng. 109 (2016) 351–363, https://doi.org/10.1016/j.
storage systems, Energy Convers. Manag. 210 (Apr. 2020), 112679, https://doi.
applthermaleng.2016.08.083.
org/10.1016/j.enconman.2020.112679.
[48] R. Anish, V. Mariappan, M. Mastani Joybari, Experimental investigation on the
[25] R.V. Seeniraj, N.L. Narasimhan, Performance enhancement of a solar dynamic
melting and solidification behavior of erythritol in a horizontal shell and multi-
LHTS module having both fins and multiple PCMs, Sol. Energy 82 (6) (2008)
535–542, https://doi.org/10.1016/j.solener.2007.11.001.

15
Vikas et al. Applied Thermal Engineering 214 (2022) 118803

finned tube latent heat storage unit, Appl. Therm. Eng. vol. 161, no. July (2019), J. Energy Storage 21 (Feb. 2019) 405–417, https://doi.org/10.1016/j.
114194, https://doi.org/10.1016/j.applthermaleng.2019.114194. est.2018.12.005.
[49] Z. Leng, Y. Yuan, X. Cao, J. Wang, F. Haghighat, Melting and solidification [60] A. Agarwal, R.M. Sarviya, An experimental investigation of shell and tube latent
performance in two horizontal shell-and-tube heat exchangers with different heat storage for solar dryer using paraffin wax as heat storage material, Eng. Sci.
structures, Int. J. Energy Res. 44 (14) (2020) 11288–11301, https://doi.org/ Technol. an Int. Journa 19 (1) (2016) 619–631.
10.1002/er.5743. [61] A. Kumar, S.K. Saha, Performance study of a novel funnel shaped shell and tube
[50] S.H. Park, Y.G. Park, M.Y. Ha, A numerical study on the effect of the number and latent heat thermal energy storage system, Renew. Energy 165 (2021) 731–747,
arrangement of tubes on the melting performance of phase change material in a https://doi.org/10.1016/j.renene.2020.11.023.
multi-tube latent thermal energy storage system, J. Energy Storage vol. 32, no. [62] V.R. Voller, C.R. Swaminathan, B.G. Thomas, Fixed grid techniques for phase
June (2020), 101780, https://doi.org/10.1016/j.est.2020.101780. change problems: A review, Int. J. Numer. Methods Eng. 30 (4) (1990) 875–898.
[51] E. Shi, S. Ahangar Zonouzi, H. Aminfar, M. Mohammadpourfard, Enhancement of [63] V.R. Voller, C. Prakash, fixed grid numerical modelling methodology for
the performance of a NEPCM filled shell-and-multi tube thermal energy storage convection-diffusion mushy region phase-change problems, Int. J. Heat Mass
system using magnetic field: A numerical study, Appl. Therm. Eng., vol. 178, no. Transf. 30 (8) (1987) 1709–1719.
December 2019, p. 115604, 2020, doi: 10.1016/j.applthermaleng.2020.115604. [64] H.B. Mahood, M.S. Mahdi, A.A. Monjezi, A.A. Khadom, A.N. Campbell, Numerical
[52] R. Qaiser, M.M. Khan, L.A. Khan, M. Irfan, Melting performance enhancement of investigation on the effect of fin design on the melting of phase change material in
PCM based thermal energy storage system using multiple tubes and modified shell a horizontal shell and tube thermal energy storage, J. Energy Storage, vol. 29, no.
designs, J. Energy Storage, vol. 33, no. November 2020, p. 102161, 2021, doi: November 2019, p. 101331, 2020, doi: 10.1016/j.est.2020.101331.
10.1016/j.est.2020.102161. [65] M.S. Mahdi, H.B. Mahood, A.F. Hasan, A.A. Khadom, A.N. Campbell, Numerical
[53] M.S. Mahdi, H.B. Mahood, A.A. Alammar, A.A. Khadom, Numerical investigation study on the effect of the location of the phase change material in a concentric
of PCM melting using different tube configurations in a shell and tube latent heat double pipe latent heat thermal energy storage unit, Therm. Sci. Eng. Prog. 11
thermal storage unit, Therm. Sci. Eng. Prog. vol. 25, no. August (2021), 101030, (Jun. 2019) 40–49, https://doi.org/10.1016/j.tsep.2019.03.007.
https://doi.org/10.1016/j.tsep.2021.101030. [66] M.S. Mahdi, H.B. Mahood, A.N. Campbell, A.A. Khadom, Natural convection
[54] G.S. Sodhi, K. Vigneshwaran, P. Muthukumar, Experimental investigations of high- improvement of PCM melting in partition latent heat energy storage: Numerical
temperature shell and multi-tube latent heat storage system, Appl. Therm. Eng. vol. study with experimental validation, Int. Commun. Heat Mass Transf. 126 (2021),
198, no. August (2021), 117491, https://doi.org/10.1016/j. 105463, https://doi.org/10.1016/j.icheatmasstransfer.2021.105463.
applthermaleng.2021.117491. [67] X. Yang, J. Guo, B. Yang, H. Cheng, P. Wei, Y.L. He, Design of non-uniformly
[55] M. Kirincic, A. Trp, K. Lenic, Influence of natural convection during melting and distributed annular fins for a shell-and-tube thermal energy storage unit, Appl.
solidification of paraffin in a longitudinally finned shell-and-tube latent thermal Energy vol. 279, no. July (2020), 115772, https://doi.org/10.1016/j.
energy storage on the applicability of developed numerical models, Renew. Energy apenergy.2020.115772.
179 (2021) 1329–1344, https://doi.org/10.1016/j.renene.2021.07.083. [68] M. Sriram, A. Bhattacharya, Analysis and optimization of triple tube phase change
[56] M. Shaikh, M. Uzair, U. Allauddin, Effect of geometric configurations on charging material based energy storage system, J. Energy Storage vol. 36, no. February
time of latent-heat storage for solar applications, Renew. Energy 179 (2021) (2021), 102350, https://doi.org/10.1016/j.est.2021.102350.
262–271, https://doi.org/10.1016/j.renene.2021.07.062. [69] R. Anish, M.M. Joybari, S. Seddegh, V. Mariappan, F. Haghighat, Y. Yuan,
[57] L. Song, S. Wu, C. Yu, W. Gao, Thermal performance analysis and enhancement of Sensitivity analysis of design parameters for erythritol melting in a horizontal shell
the multi-tube latent heat storage (MTLHS) unit, vol. 46, no. November 2021, and multi-finned tube system: Numerical investigation, Renew. Energy 163 (2021)
2022. 423–436, https://doi.org/10.1016/j.renene.2020.08.153.
[58] M.M. Joybari, S. Seddegh, X. Wang, F. Haghighat, Experimental investigation of [70] H. Li, C. Hu, Y. He, D. Tang, K. Wang, Influence of fin parameters on the melting
multiple tube heat transfer enhancement in a vertical cylindrical latent heat behavior in a horizontal shell-and-tube latent heat storage unit with longitudinal
thermal energy storage system, Renew. Energy 140 (2019) 234–244, https://doi. fins, J. Energy Storage, vol. 34, no. October 2020, p. 102230, 2021, doi: 10.1016/j.
org/10.1016/j.renene.2019.03.037. est.2020.102230.
[59] A. Yadav, S. Samir, Experimental and numerical investigation of spatiotemporal
characteristics of thermal energy storage system in a rectangular enclosure,

16