energies

Article
Investigation of Stratified Thermal Storage Tank
Performance for Heating and Cooling Applications
Azharul Karim *, Ashley Burnett and Sabrina Fawzia
Science and Engineering Faculty, Queensland University of Technology, Brisbane CBD, QLD 4001, Australia;
ashley.burnett@student.qut.edu.au (A.B.); sabrina.fawzia@qut.edu.au (S.F.)
* Correspondence: azharul.karim@qut.edu.au; Tel.: +61-7-3138-6879; Fax: +61-7-3138-1529




Received: 28 March 2018; Accepted: 18 April 2018; Published: 25 April 2018


Abstract: A large amount of energy is consumed by heating and cooling systems to provide comfort
conditions for commercial building occupants, which generally contribute to peak electricity demands.
Thermal storage tanks in HVAC systems, which store heating/cooling energy in the off-peak
period for use in the peak period, can be used to offset peak time energy demand. In this study,
a theoretical investigation on stratified thermal storage systems is performed to determine the factors
that significantly influence the thermal performance of these systems for both heating and cooling
applications. Five fully-insulated storage tank geometries, using water as the storage medium, were
simulated to determine the effects of water inlet velocity, tank aspect ratio and temperature difference
between charging (inlet) and the tank water on mixing and thermocline formation. Results indicate
that thermal stratification enhances with increased temperature difference, lower inlet velocities
and higher aspect ratios. It was also found that mixing increased by 303% when the temperature
difference between the tank and inlet water was reduced from 80 ◦ C to 10 ◦ C, while decreasing the
aspect ratio from 3.8 to 1.0 increased mixing by 143%. On the other hand, increasing the inlet water
velocity significantly increased the storage mixing. A new theoretical relationship between the inlet
water velocity and thermocline formation has been developed. It was also found that inlet flow rates
can be increased, without increasing the mixing, after the formation of the thermocline.

Keywords: thermal energy storage; simulation; thermal stratification; thermocline; mixing coefficient;
performance evaluation

1. Introduction
Heating and cooling systems are required to maintain comfortable thermal conditions within built
environments for the occupants of commercial and office buildings. Traditional heating, ventilating,
air-conditioning and cooling (HVAC) systems are subject to high costs and contribute to peak demand.
Most HVAC systems operate in the daytime and therefore contribute to the demand peak. The daytime
peak electricity load is in great excess compared to the daily average load [1,2]. Utility companies
must have larger capacities to meet this demand, that go unused most of the time. The requirement to
meet peak demand eventually is met either by constructing new generation facilities or by purchasing
power from competitors. Both options results in higher customer rates. Thermal energy storage (TES)
is an advanced method to generate the heating and cooling energy in the off-peak hours and deliver
to buildings during peak-hours and thus shifts the electricity demand to off-peak hours. With TES,
heating and cooling systems are run at off-peak times (mostly at night) to store the required energy
in the storage tanks. During the peak times, heating/cooling energy is supplied from the storage
tanks without running heaters and chillers. On the other hand, concern about climate change and
greenhouse emissions, as well as the increase in power costs and pressures on current infrastructure
forces utilities and governments to use more renewable energy, particularly solar energy [3,4]. Solar

Energies 2018, 11, 1049; doi:10.3390/en11051049 www.mdpi.com/journal/energies

Energies 2018, 11, x FOR PEER REVIEW 2 of 15
Energies 2018, 11, 1049 2 of 15

on current infrastructure forces utilities and governments to use more renewable energy, particularly
solar
thermalenergy
energy[3,4].
hasSolar
many thermal energyincluding
applications, has manyspaceapplications, including
heating [5], powerspace heating
generation [6][5],
andpower
solar
generation
drying [6] and solar
of agricultural drying of
products agricultural
[7–9]. However, products
one of [7–9]. However,
the major one ofofthe
drawbacks major
solar drawbacks
energy is that
ofcan
it solar
onlyenergy is thatinitthe
be utilised candaytime
only beand utilised in the
on clear daytime
days. and onthe
To maximise clear days. Toofmaximise
utilization the
solar energy,
utilization
proper of solar
energy energy,
storage mustproper energy storage
be developed. mustitbe
Therefore, developed.
can be arguedTherefore,
that thermalit can be argued
energy storagethat
is
thermal
most energyfor
essential storage
the bestis most essential
utilization for the bestasutilization
of traditional of traditional
well as solar energy. as well as solar energy.
Thermal energy
Thermal energystorage
storage(TES)
(TES) systems
systems areare
usedused
as a as a temporary
temporary energy energy
storagestorage that energy
that allows allows
energy
to to be
be held andheld
usedandforused
laterfor later
use. Inuse.
the In the literature,
literature, severalseveral
concepts concepts of thermal
of thermal storagestorage have
have been
developed [10,11]. Based on the medium of storage, water, oil and phase change materials have have
been developed [10,11]. Based on the medium of storage, water, oil and phase change materials been
been utilized
utilized for thermal
for thermal storage storage applications.
applications. However,
However, due todue
itsto its availability,
availability, highhigh specific
specific heat heat
and and
low
low cost,
cost, waterwater
has thehashighest
the highest potential
potential for useforasuse as storage
storage medium medium [2]. Moreover,
[2]. Moreover, storage storage
systemssystems
using
using have
water watersome
havesignificant
some significant advantages
advantages overmediums
over other other mediums including:
including:
•• Controls are
Controls are simpler
simpler than
than phase
phase change
change storage
storage systems.
systems.
•• Water can also serve as a water reservoir for fire protection.
Water can also serve as a water reservoir for fire protection.
•• Stored water
Stored water can
can serve
serve as
as aa standby
standby source
source for
for heating/cooling
heating/cooling in
in the
the case
case of
of power
power outage.
outage.
• Less costly than other mediums.
• Less costly than other mediums.
The separation of the cold/hot water stored in the storage tank and the cold/warm water
The separation of the cold/hot water stored in the storage tank and the cold/warm water
returning from the building HVAC system is the key factor that contribute to the performance of
returning from the building HVAC system is the key factor that contribute to the performance of water
water storage systems. Various methods such as the use of two tanks (one for cold and one for warm
storage systems. Various methods such as the use of two tanks (one for cold and one for warm water),
water), a membrane or diaphragm to separate cold and warm water in the same tank and natural
a membrane or diaphragm to separate cold and warm water in the same tank and natural stratification
stratification have been used. In the stratification method, cold and warm water are stored in the
have been used. In the stratification method, cold and warm water are stored in the same tank and are
same tank and are separated by natural stratification and reported to have better or similar
separated by natural stratification and reported to have better or similar performance compared to
performance compared to other methods [2].
other methods [2].
The principle of higher temperature water being less dense and therefore lighter than lower
The principle of higher temperature water being less dense and therefore lighter than lower
temperature water (up to 4 °C), being dense and hence heavier, allows for separation to occur.
temperature water (up to 4 ◦ C), being dense and hence heavier, allows for separation to occur.
Between these low and high temperature zones, a transitional zone known as the thermocline exists.
Between these low and high temperature zones, a transitional zone known as the thermocline
The thermocline consists of a mixture of hot and cold storage medium and consequently has
exists. The thermocline consists of a mixture of hot and cold storage medium and consequently has
intermediate density. Figure 1 shows the temperature profiles of a typical stratified storage tank [12].
intermediate density. Figure 1 shows the temperature profiles of a typical stratified storage tank [12].
It can be seen that a temperature gradient exists between the hot and cold zones, which makes the
It can be seen that a temperature gradient exists between the hot and cold zones, which makes the
thermocline. Thermocline thickness is one of the measures of the degree of stratification occurring in
thermocline. Thermocline thickness is one of the measures of the degree of stratification occurring
the storage tank. Highly stratified systems have a narrow thermocline with a low temperature
in the storage tank. Highly stratified systems have a narrow thermocline with a low temperature
gradient, whereas lowly stratified systems have thick thermocline with a high temperature gradient.
gradient, whereas lowly stratified systems have thick thermocline with a high temperature gradient.

Figure 1. Typical temperature profile in a stratified thermal storage tank [12].

Figure 1. Typical temperature profile in a stratified thermal storage tank [12].
Energies 2018, 11, 1049 3 of 15

Thermocline thickness determines the extent of mixing between warm and cold water. A small
stable thermocline also reduces the amount of heat transferred between warm and cold water inside
the tank during charging and discharging providing better stratification. Mechanisms that destroy
stratification of TES systems include [13]:
• Natural convection from the tank to the surrounding environment.
• Mixing of hot and cold water due to the kinetic force of water entering from the top or bottom of
the tank.
• Thermal diffusion and conduction within the water inside the tank and with storage tank wall,
piping and other materials inside the tank.
• Aspect ratio of the storage tanks.
Mixing is the most important cause of capacity loss in stratified tanks [14]. Heat transfer between
hot and cold water through the thermocline causes loss of thermal storage capacity and is also a major
cause of stratification degradation in TES systems [15]. Careful analysis of the operating variables and
storage tank design is required to minimise fluid mixing. Highly stratified TES systems can provide
enhanced capacity to store quality thermal energy for heating and cooling systems through minimising
heat losses. The first author Karim has developed a stratified chilled water thermal storage system and
demonstrated that a stratified storage tank for an air conditioning system consistently stratified in the
absence of any physical barrier [2]. He also investigated the effect of flow rates on storage efficiency
and showed that storage efficiency decreased at higher flow rate due to increased mixing. However,
the effect of temperature range and tank design were not investigated in that study. The present study
is the continuation of the previous study and aims to investigate the effect of the design and operating
parameters on the mixing and degradation of thermal stratification. The investigation was conducted
for both hot and cold water storage and hence applicable to any HVAC system.

2. Analysis
Better stratification can be ensured with the appropriate design of the tank, proper control of
temperature differences, maintaining the right inlet velocity and the introduction of cold water from
the bottom of the tank. If these are ensured, mechanical disturbances will be minimum, and the
density differences should create a region of vertical temperature gradient, called the thermocline
that separates warm and cold waters. The higher temperature water is introduced from the top of the
tank during discharge cycle and the cold water from the bottom of the tank during the charging cycle.
In this study, the mixing and thermocline behaviours of TES were investigated for changes in inlet
velocity, the temperature difference between the inlet water and stored water and the tank aspect ratio.
As mentioned earlier, five different tank geometries were examined in this study. Table 1 shows the
details of the tank geometries investigated in this study.

Table 1. Geometrical dimensions of the five different tanks investigated.

Tank Number Diameter, d (m) Height, h (m) Area, a (m2 ) Volume, V (m3 ) Aspect Ratio, AR (h/d)
TES-1 0.180 0.680 0.025 0.0173 3.778
TES-2 0.200 0.550 0.031 0.0173 2.750
TES-3 0.210 0.500 0.035 0.0173 2.381
TES-4 0.255 0.340 0.051 0.0173 1.333
TES-5 0.280 0.280 0.062 0.0173 1.000

Tank diameter and height varied to provide different aspect ratios determined by:

tank height
AR = (1)
tank diameter
Inlet water temperature was set at 283 K. The thermal properties of the inlet water at that
temperature are:
Energies 2018, 11, 1049 4 of 15

• Density: 999.7 kg/m3

• Dynamic viscosity: 1.307 × 10−3 kg/ms
• Thermal expansion coefficient: 0.0733 × 10−3 /K
The three initial tank water temperatures used for analysis are: 293, 323, and 363 K.
Temperature-dependent properties for water were considered for the simulations. Inlet flow velocities
ranged between 0.5 cm/s and 6 cm/s.
The effects of different parameters such as inlet water velocity were examined using the
dimensionless parameters of the Richardson and Reynolds numbers. Nelson et al. [16] found that
the ‘mixing coefficient’ is closely related to the thermocline decay and can be expressed as the ratio of
the Reynolds number and the Richardson number. They suggested that mixing coefficient Z can be
expressed as shown in Equation (2):
 0.67
Re
Z = 1.688 × 104 (2)
Ri

where the Reynolds number, Re, and Richardson number, Ri, are expressed as:

ρvd
Re = (3)
µ
gβ∆Th
Ri = (4)
v2
A mixing coefficient of one represents a perfectly stratified system with no mixing [16]. As the
ratio of Re
Ri decreases, the mixing is reduced. The ratio of Re and Ri can be written as:

Re ρv3 d
= (5)
Ri µgβ∆Th

As can be seen in Equation (5), a strong influence of flow velocity exists on the storage mixing, as
shown in Equation (5).
The parameters of water density, viscosity and thermal expansion are temperature dependent.
Therefore, reducing the mixing results in:
• Reduced inflow velocity.
• Increased temperature difference.
• Increased distance between ports (increasing the aspect ratio).

3. Simulation
Modelling and simulation of the thermal storage tank were conducted using ANSYS-CFX (ANSYS
CFX-Solver Modeling, Release 12.1, ANSYS Inc., Canonsburg, PA, USA, 2009) [17]. The governing
transport equations for a mass, momentum and energy conservation for a fluid are:
• Continuity:
∂ρ
+ ∇·(ρU) = 0 (6)
∂t
• Momentum:
∂(ρU)
+ ∇·(ρU ⊗ U) = −∇ρ + ∇·τ + S M (7)
∂t

Thermal energy is being transferred in the storage system. It only considers the transfer
of enthalpy.

∂(ρh)
+ ∇·(ρUh) = ∇·(λ∇ T ) + τ : ∇U + S M (8)
∂t
Energies 2018, 11, 1049 5 of 15

Due to the heat being transferred based on the temperature differences of the stored water,
the buoyancy effects of the water must be considered as a result of the density differences:
 
S M,buoy = ρ − ρre f g (9)

Buoyancy differences are governed by gravity. The “Boussinesq” model is used to account for
these small density changes through the use of varying thermal expansion β, based on the temperature
differences:  
ρre f = ρβ T − Tre f (10)

The CFX model will be one-dimensional in nature as the degree of stratification and thermocline
thickness is only dependent on one dimension. The energy equation is described in scalar components:
!
NC
∂ ∂P ∂ ∂ ∂T ∂Y µt ∂h
+ ∑ Γi hi i +


(ρH) − + ρU j H = λ + SE (11)
∂t ∂t ∂x j ∂x j ∂x j i
∂x j Pr t ∂x j

Boundary conditions that influence the system and hence model simulation are the inlet, outlet
and tank wall conditions and the conditions at the interface between the liquid and solid at the inlet,
outlet and wall. Total enthalpy is determined from the static enthalpy and the boundary conditions of
energy and pressure and enthalpy as a function of pressure and temperature [17]. Inlet and outlet flows
are considered to be subsonic and are specified to have static pressure. The simulation is considered to
be transient due to the changing water characteristics and heat transfer with time.
Tank geometries were constructed, using the CFX modeller, for the five TES systems according
to the dimensions provided in Table 1. As the tank geometries studied are simple and cylindrical in
shape, uniformity throughout the volume of tank was considered to provide consistent representation
of heat transfer during charging. For greater accuracy, a smaller element size was implemented around
the inlet/outlet connections with the tank. For initial simulations, the default ANSYS meshing plan
was adopted. However, with default meshing, results were not very accurate as temperatures in the
thermoclines were not uniform. It was assumed that to achieve better results, the number of meshing
elements needs to be increased. Therefore, an additional meshing configurations was implemented,
through refinement to the degree of three. This meshing scheme provided much better results, and the
results were validated with published results [2]. Figure 2 shows the meshing for the five TES systems.
In the model, a convergence criterion was selected as 10−4 for pressure and momentum and 10−6 for
the rest of the values.

Figure 2. Meshed tanks used in TES simulations.

Energies 2018, 11, 1049 6 of 15
Energies 2018, 11, x FOR PEER REVIEW 6 of 15

4. Results and Discussions

4. Results and Discussions

4.1. Analytical
4.1. Analytical
The mixing
The mixingcoefficient
coefficient waswas calculated
calculated for different
for different designdesign and variables
and process process variables using
using Equation
Equation (2), which was solved in the MS Excel programme. The results presented
(2), which was solved in the MS Excel programme. The results presented in Figure 3 show the effect in Figure 3 showof
the effecttank
different of different tank water(20
water temperatures temperatures
◦ ◦
C, 50 C and (2090°C,C)50and
◦ °C inlet
and velocities
90 °C) and oninlet velocities
the mixing on the
coefficient.
mixing
As can becoefficient.
seen, mixing As can be seen,when
is reduced mixingtheistemperature
reduced when the temperature
difference between the difference between
cool inlet water andthe
cool inlet water and warm stored water is increased. Mixing is more volatile
warm stored water is increased. Mixing is more volatile at lower temperature differences with greater at lower temperature
differences
mixing. with greater
At higher mixing.differences,
temperature At higher temperature
mixing is lower. differences,
Overall mixing is lower. Overall
mixing increases by 303%mixing
when
increases by 303% when the temperature differential
◦ is
◦ reduced from
the temperature differential is reduced from 80 C to 10 C. These results clearly imply that 80 °C to 10 °C. These results
a very low
clearly implydifferential
temperature that a verybetween
low temperature
the inlet and differential
tank water between
must bethe inlet and
avoided tank to
in order water mustthe
improve be
avoided in order to improve the performance of the stratified tanks. In practice,
performance of the stratified tanks. In practice, the temperature difference will be low in the cooling the temperature
difference
system, will will
as this be low in the cooling
be limited system,water
by the chilled as this will be limited
temperature by the chilled
and building returnwater temperature
temperature. This
and building
range is usuallyreturn temperature.
between 10 and 20This ◦ C. range
However,is usually
a higher between 10 anddifference
temperature 20 °C. However, a higher
is possible for a
temperature
heating system,difference
and theisabove
possible for a heating
limitation does system,
not apply andtothe above limitation
a heating does not apply
system. Therefore, it cantobea
heating system. Therefore, it can be concluded that stratified TES will have
concluded that stratified TES will have better performance in a heating system compared to a cooling better performance in a
heating The
system. system compared
results presentedto aincooling
Figure system.
3 will helpThedesigners
results presented in Figure
in determining the 3suitable
will help designers
temperature
in determining the suitable
difference for their system. temperature difference for their system.

2,500,000

2,000,000
Mixing Coefficient

1,500,000

1,000,000

500,000

0
0 1 2 3 4 5 6 7
Inlet Velocity cm/sec

90 deg 50 deg 20 deg

Figure 3.
Figure 3. Graph
Graph showing
showing the
the mixing
mixing effect
effect for
for different
different tank
tank water
water temperatures
temperatures and
and inlet
inlet velocities
velocities
◦ C).
(inlet water temperature is 10 °C).
(inlet water temperature is 10

The results
The results in
in Figure
Figure 44 show
show the
the effect
effect of
of the
the aspect
aspect ratio
ratio and
and inlet
inlet velocities
velocities on
on mixing
mixing in
in aa
TES system.
TES system.
Energies 2018, 11, 1049 7 of 15
Energies 2018, 11, x FOR PEER REVIEW 7 of 15

1,600,000

1,400,000

1,200,000
Mixing Coefficient

1,000,000
TES 1
800,000 TES 2
TES 3
600,000
TES 4
400,000 TES 5

200,000

0
0 1 2 3 4 5 6 7
Inlet Velocity cm/sec

Figure 4.
Figure 4. Graph
Graph showing
showing the mixing effect
the mixing effect for
for aspect
aspect ratios
ratios and
and inlet
inlet velocities.
velocities.

As can be seen in the figure, mixing is gradually reduced when the aspect ratio is increased.
As can be seen in the figure, mixing is gradually reduced when the aspect ratio is increased.
Moreover, mixing in high aspect ratio systems (TES-1) is less affected by inlet velocity changes, while
Moreover, mixing in high aspect ratio systems (TES-1) is less affected by inlet velocity changes, while
the effects of velocity increase in low aspect ratio tanks is significant. As can be seen, TES-2 (aspect
the effects of velocity increase in low aspect ratio tanks is significant. As can be seen, TES-2 (aspect
ratio of 2.75) has a 24% higher mixing coefficient compared to TES-1 (AR of 3.8) at the same inlet
ratio of 2.75) has a 24% higher mixing coefficient compared to TES-1 (AR of 3.8) at the same inlet
velocity. Similarly, replacing a TES-2 system with a TES-3 system (AR = 2.38) increased the mixing
velocity. Similarly, replacing a TES-2 system with a TES-3 system (AR = 2.38) increased the mixing
coefficient by 10%; replacing a TES-3 system with a TES-4 system (AR = 1.33) increased the mixing
coefficient by 10%; replacing a TES-3 system with a TES-4 system (AR = 1.33) increased the mixing
coefficient by 48%; and replacing a TES-4 system with a TES-5 system (AR = 1.00) increased the mixing
coefficient by 48%; and replacing a TES-4 system with a TES-5 system (AR = 1.00) increased the mixing
coefficient by 21%. Mixing increase for all the tanks was compared at the same inlet velocity (6 cm/s).
coefficient by 21%. Mixing increase for all the tanks was compared at the same inlet velocity (6 cm/s).
Overall mixing increases by 143% when comparing a TES-5 system with a TES-1 system.
Overall mixing increases by 143% when comparing a TES-5 system with a TES-1 system.
Figures 3 and 4 show that faster inlet flow velocities, a lower aspect ratio and a lower
Figures 3 and 4 show that faster inlet flow velocities, a lower aspect ratio and a lower temperature
temperature difference between the inlet and tank water result in increased water mixing. The
difference between the inlet and tank water result in increased water mixing. The designers need to
designers need to optimize these parameters for the best results based on particular applications
optimize these parameters for the best results based on particular applications of TES.
of TES.
4.2. Simulation
4.2. Simulation
4.2.1. Model Validation
4.2.1. Model Validation
In order to validate the simulation model developed, predicted temperature profiles in a chilled
waterInstorage
order totank
validate the simulation
are compared withmodel developed,
the authors’ predicted
previously temperature
published profiles in adata
experimental chilled
[2].
water storage tank are compared with the authors’ previously published experimental
The comparison of the experimental results to the simulation results is shown in Figure 5. The tankdata [2]. The
comparison water
dimensions, of theflow
experimental
rate, inlet results
and tanktowater
the simulation
temperatureresults is shown
and other in Figure
properties used5.inThe tank
[2] were
dimensions,
also water
used in this flow rate,
simulation forinlet and tankpurposes.
comparison water temperature and in
It can be seen other properties
Figure used in [2]results
5 that simulation were
also used in this simulation for comparison purposes. It can be seen in
closely agreed with experimental results and hence validated the model proposed.Figure 5 that simulation results
closely agreed with experimental results and hence validated the model proposed.
 Energies 2018, 11, 1049 8 of 15
Energies 2018, 11, x FOR PEER REVIEW 8 of 15
Energies 2018, 11, x FOR PEER REVIEW 8 of 15

Figure
Figure 5.5. Variation
Variationof of
thethe temperature
temperature profile
profile in a chilled
in a chilled water tank:
water storage storage tank: predicted
predicted and
and measured
Figure 5. Variation of the temperature profile in a chilled water storage tank: predicted and
measured
values [2]. values [2].
measured values [2].

4.2.2.
4.2.2. Simulation
Simulation Results
Results
4.2.2. Simulation Results
Simulation
Simulation results
results of
of thermocline
thermocline formation
formation in in tanks
tanks of
of different
different aspect
aspect ratios
ratios at
at inlet
inlet water
water
Simulation results of thermocline formation in tanks of different aspect ratios at inlet water
velocity
velocity of
of 1.5 cm/s
1.5 cm/s are
are presented
presented in
in Figure
Figure 6.
6. In
In these
these simulations,
simulations, the
the tank
tank water
water temperature
temperature was
was
velocity of 1.5 cm/s are presented in Figure 6.◦In these simulations, the tank water temperature was
50
50 °C,
◦ C, and
and the
the inlet
inlet water
water temperature
temperature was
was 66 °C.
C.
50 °C, and the inlet water temperature was 6 °C.

(a) (b) (c) (d) (e)

(a) (b) (c) (d) (e)
Figure 6. Simulation results of different TES systems after 1600 s: (a) TES-1; (b) TES-2; (c) TES-3;
Figure 6. Simulation results of different TES systems after 1600 s: (a) TES-1; (b) TES-2; (c) TES-3;
(d) TES-4;
Figure 6. (e) TES-5. results of different TES systems after 1600 s: (a) TES-1; (b) TES-2; (c) TES-3;
Simulation
(d) TES-4; (e) TES-5.
(d) TES-4; (e) TES-5.
It can be seen that mixing of hot and cold water increases with the reduction of the tank aspect
It can be seen that mixing of hot and cold water increases with the reduction of the tank aspect
rations. In Tank
It can 5, although
be seen that mixinga thermocline
of hot and is formed,
cold watermuch mixing
increases withcan
thebereduction
observedofbelow the
the tank
rations. In Tank 5, although a thermocline is formed, much mixing can be observed below the
thermocline.
aspect The In
rations. above
Tankresults
5, support athethermocline
although theoretical analysis
is presented
formed, much in Figurecan
mixing 4, showing
be that
observed
thermocline. The above results support the theoretical analysis presented in Figure 4, showing that
TES
below systems with higher
the thermocline. aspect results
The aspect
above ratios are characterised by reducedpresented
mixing and more
TES systems with higher ratios support the theoretical
are characterised by analysis
reduced mixing in andFigure
more4,
developed thermoclines.
developed thermoclines.
 x FOR PEER REVIEW
Energies 2018, 11, 1049 9 of 15

Figure 7 shows the non-dimensional thermocline thickness for different TES designs. The
showing that
thermocline
Energies TES11, xsystems
with
2018, the with
steepest
FOR PEER higher aspectgradient
temperature
REVIEW ratios areischaracterised
observed in by reduced
TES-1 mixing
(highest and
aspect more
ratio).
9 of 15
developed thermoclines.
TES-2 and 3 have a slightly thicker and less developed thermocline than TES-1. TES-4 has a larger
Figure Figure 7 showsthe
7whilst
shows thenon-dimensional
non-dimensional thermocline thickness for different TES designs.
TES The
thermocline, TES-5 (smallest aspect ratio)thermocline thickness
has the thickest for different
thermocline. designs.
thermocline with the steepest temperature gradient is observed in TES-1
The thermocline with the steepest temperature gradient is observed in TES-1 (highest aspect (highest aspect ratio).ratio).
TES-2TES-2
and 3and have3 have a slightly
a slightly thicker
thicker andless
and lessdeveloped
developed thermocline
thermocline than
thanTES-1. TES-4
TES-1. has has
TES-4 a larger
a larger
thermocline, 325 whilst TES-5 (smallest aspect ratio) has the thickest thermocline.
thermocline, whilst TES-5 (smallest aspect ratio) has the thickest thermocline.
TES 1
320 325 TES 2
TES 3 1
TES
Difference (K)

315 320 TES 4 2

TES
TES
TES 5 3
Temperature Difference(K)
Difference (K)

310 315 TES 4

TES 5
Temperature Difference(K)

310
305
Temperature

305
300
Temperature

300
295
295
290
290

285
285

280
280
0.5 0.55 0.6 0.65
0.5 0.55 0.6 0.65 0.70.7 0.750.75
Non-Dimensional Tank Height (mm)
Non-Dimensional Tank Height (mm)

Figure 7. Temperature profile and thermocline in different tank designs (water velocity of 1.5 cm/s,
Figure 7. Temperature
Figure profile
7. Temperature and
profile andthermocline
thermoclinein
in different tankdesigns
different tank designs (water
(water velocity
velocity of cm/s,
of 1.5 1.5 cm/s,
tans water temperature 50 ◦ C, inlet water temperature 6 ◦ C).
tans water
tans water temperature
temperature 50 °C,
50 °C, inlet
inlet watertemperature
water temperature 66 °C).
°C).

In In order
In order
order to to to investigate
investigate
investigate thetheimpact
the impact impact
of ofof temperature
temperaturetemperature differential
betweenbetween
differential
differential andtank
between
tank and
tank
inlet inlet inlet
and
temperatures,
TES-1 temperatures,
was simulated TES-1 was simulated
for three for
temperature three temperature differences
◦ ◦ (84 °C, 44◦ °C and 14 °C) at a
temperatures, TES-1 was simulated for threedifferences
temperature(84differences
C, 44 C and 14 44
(84 °C, C) °C
at aand
constant
14 °C)inlet
at a
constant
velocity of 1.5inlet
cm/s.velocity
It of ofbe1.5assumed
can1.5 cm/s. It can
thatbe assumed
these that
ranges ofthese ranges ofdifference
temperature temperature (14difference
◦ C to 84 ◦ C)
constant inlet velocity cm/s. It can be assumed that these ranges of temperature
(14 °C to 84 °C) cover both hot and cold water storage systems. The results are presented in Figure 8.
difference
(14 °Cboth
cover to 84hot
°C)and
cover both
cold hot storage
water and coldsystems.
water storage systems.
The results The results
are presented inare presented
Figure 8. in Figure 8.

(a) (b) (c)

Figure 8. Temperature profiles in TES-1 at temperature differences of (a) 84 °C, (b) 44 °C and
(c) 14 °C.
(a) (b) (c)

Figure 8. Temperature profiles in TES-1 at temperature differences of (a) 84 °C, (b) 44 °C and
Figure 8. Temperature profiles in TES-1 at temperature differences of (a) 84 ◦ C, (b) 44 ◦ C and (c) 14 ◦ C.
(c) 14 °C.
Energies 2018, 11, 1049 10 of 15
Energies2018,
Energies 11,x xFOR
2018,11, FORPEER
PEERREVIEW
REVIEW 1010ofof1515

Theresults
The
The resultspresented
results presentedin
presented ininFigures
Figures888and
Figures and999show
and showthat
show thatthe
that thetemperature
the temperaturegradient
temperature gradientis
gradient isisthin
thinat
thin atathigher
higher
higher
temperature differences and gradually becomes thicker with the reduction in temperature
temperature differences and gradually becomes thicker with the reduction in temperature difference.
temperature differences and gradually becomes thicker with the reduction in temperature difference.
difference.
Thesesimulation
These
These simulationresults
simulation resultscompared
results comparedwell
compared wellwith
well withthe
with thetheoretical
the theoreticalresults
theoretical resultspresented
results presentedin
presented ininFigure
Figure3,
Figure 3,3,showing
showing
showing
highermixing
higher
higher mixingfor
mixing forTES
for TESsystems
TES systemswhen
systems whenthe
when thetemperature
the temperaturedifference
temperature differenceis
difference isislow.
low.
low.

1.4
1.4
90deg
90 deg
Difference(K)(K)

50deg
50 deg
Non-Dimensional Temperature Difference(K)

1.2
Temperature Difference(K)

1.2 20deg
deg
20
Difference

11
Temperature
Temperature

0.8
0.8

0.6
0.6
Non-Dimensional
Non-dimensional
Non-dimensional

0.4
0.4

0.2
0.2

00350 400 450 500

350 400 450 500
TankHeight
Tank Height(mm)
(mm)

Figure9.
Figure
Figure 9.9.Temperature
Temperatureprofiles
Temperature profileswith
profiles withtank
with tankheight
tank heightin
height ininTES-1
TES-1at
TES-1 atatdifferent
differenttemperature
different temperaturedifferences.
temperature differences.
differences.

Theeffects
The
The effects
effects
of ofof varying
varying
varying inlet velocities
inletinlet velocities
velocities onon thermocline
on thermocline thermocline formation
formationformation and stratification
and
and stratification stratification were
were
were investigated
investigated
investigated
by varying inlet byvarying
by varyinginlet
velocity inlet
for velocity
velocity
TES-1 for
at afor TES-1at
TES-1
constant ataaconstant
tank constant tankof
tank
temperature temperature
temperature
◦
90 C. ofof90
90°C.
°C.
The
The results
The results
results inin Figures
inFigures 10 and
Figures1010and 11
and1111 show
show
show that
that as
as as
that the
thethe inlet
inlet velocity
velocity
inlet decreases,
decreases,
velocity less mixing
less less
decreases, mixing occurs.
occurs.
mixing The
The
occurs.
best
best thermocline
thermocline
The best thermocline formation
formation occurred
occurred
formation when
whenwhen
occurred the temperature
the temperature
the temperature gradient
gradient separating
separating
gradient the hot
the hot
separating and
theand the
hotthe cold
andcold
the
zones
zones
cold ofofthe
zones the water
ofwater storage
storage
the water tanktank
tank
storage wasperfectly
was perfectly
was linear.
linear.
perfectly AAsummary
linear. summary
A summary ofofthe
the simulation
ofsimulation
the results
results
simulation when
when
results full
full
when
thermocline
thermocline
full thermocline formation
formation occurred
occurred
formation are shown
are shown
occurred below
below
are shown below in Table
in Table 2.
2. 2.
in Table

(a)
(a) (b)
(b) (c)
(c) (d)
(d) (e)
(e) (f)
(f) (g)
(g)
Figure10.
Figure 10.Temperature
Temperatureprofiles
profilesininTES-1
TES-1atatdifferent
differentinlet
inletvelocities:
velocities:(a)
(a)88cm/s;
cm/s;(b)
(b)77cm/s;
cm/s;(c)
(c)66cm/s;
cm/s;
Figure 10. Temperature profiles in TES-1 at different inlet velocities: (a) 8 cm/s; (b) 7 cm/s; (c) 6 cm/s;
(d) 5 cm/s; (e) 4 cm/s; (f) 3 cm/s; (g) 2 cm/s.
(d) 5 cm/s; (e) 4 cm/s; (f) 3 cm/s; (g) 2 cm/s.
(d) 5 cm/s; (e) 4 cm/s; (f) 3 cm/s; (g) 2 cm/s.
Energies 2018, 11, 1049 11 of 15
Energies 2018, 11, x FOR PEER REVIEW 11 of 15

370

360

350
Temperature Difference (K)
Temperature Difference(K)
340
2 cm/ s 3 cm/ s 4 cm/ s 5 cm/ s 6 cm/ s 7 cm/ s 8 cm/ s
330

320

310

300

290

280
0 100 200 300 400 500 600 700
Tank Height (mm)

Figure11.
Figure 11.Temperature
Temperatureprofiles
profilesininTES-1
TES-1atatdifferent
differentinlet
inletvelocities:
velocities:non-dimensional
non-dimensionaltemperature
temperature
difference versus tank height.
difference versus tank height.

These results show a Table 2. Summary

general of the results:
trend of increased thermocline
mixing formation. difference is reduced.
as the temperature
In Figure 12Stored
below,Water
a general equation
Inlet for the relationship between the mixing coefficient and inlet
Reynolds
Flow Rate Richardson’s Mixing
velocity
TES for Temperature
fully-developed thermoclines
Velocity is presented. Number
(𝐋𝐋/𝐡𝐡) Number Ri Coefficient Z
(°C) (cm/s) Re
90 Summary of the
Table 2.0.90 results: thermocline
3.97 1239 formation.
482 31,757
1 50 0.85 3.75 1170 270 45,045
Stored Water Inlet Velocity Flow Rate Reynolds Richardson’s Mixing
TES 20 ◦
Temperature ( C)
0.70
(cm/s)
3.09
(L/h)
964
Number Re
100
Number Ri
77,186
Coefficient Z
90 0.85 3.75 1300 437 35,022
90 0.90 3.97 1239 482 31,757
1 2 50 50 0.80
0.85 3.53
3.75 1123
1170 246270 49,331
45,045
20 20 0.65
0.70 2.87
3.09 994964 94100 82,270
77,186
90 90 0.85
0.85 3.75
3.75 1365
1300 398437 38,572
35,022
50 0.80 3.53 1123
2 3 50 0.80 3.53 1285 224246 49,331
54,330
20 0.65 2.87 994 94 82,270
20 0.65 2.87 1044 85 90,609
90 0.85 3.75 1365 398 38,572
3 50 90 0.80
0.80 3.53
3.53 1560
1285 305224 50,358
54,330
4 20 50 0.75
0.65 3.32
2.87 1463
1044 17485 70,276
90,609
90 20 0.60
0.80 2.65
3.53 1170
1560 68305 113,767
50,358
4 50 90 0.80
0.75 3.53
3.32 1713
1463 251174 61,063
70,276
20 0.60 2.65 1170 68 113,767
5 50 0.70 3.09 1499 164 74,286
90 0.80 3.53 1713 251 61,063
5 50 20 0.60
0.70 2.42
3.09 1178
1499 66164 115,816
74,286
20 0.60 2.42 1178 66 115,816
These results show a general trend of increased mixing as the temperature difference is reduced.
In Figure 12 below, a general equation for the relationship between the mixing coefficient and inlet
The relationship
velocity for fully-developed × 107 v + 254,is526
Z = −3 thermoclines shown above represents the expected mixing when a
presented.
fully-developed thermocline is formed. Therefore, for any system, the inlet flow velocity required to
fully form a thermocline can be determined.
Once the thermocline is formed for the TES systems, increases in inlet flow velocity allow the
tank to charge more quickly without degrading the temperature distribution. The results below in
Figure 13 show that significant mixing is occurring (light blue colour of water) at 2 cm/s.
 120000
y = -3E+07x + 254526
100000

Mixing Coefficient
Energies 2018, 11, 1049 12 of 15
Energies 2018, 11, x FOR PEER REVIEW 12 of 15
80000
140000
60000
120000
40000 y = -3E+07x + 254526
100000
Mixing Coefficient

20000
80000

0
60000
0.005 0.0055 0.006 0.0065 0.007 0.0075 0.008 0.0085 0.009 0.0095 0.01
Inlet Velocity (m/s)
40000

Figure
2000012. Development of an equation for the mixing coefficient based on inlet velocity.

The relationship
0
𝑍𝑍 = −3 × 107 𝑣𝑣 + 254,526 shown above represents the expected mixing when
a fully-developed0.005
thermocline
0.0055 is formed.
0.006 Therefore,
0.0065 for any0.008
0.007 0.0075 system, the inlet
0.0085 0.009flow velocity
0.0095 0.01required
to fully form a thermocline can be determined. Inlet Velocity (m/s)
Once the thermocline is formed for the TES systems, increases in inlet flow velocity allow the
tank to charge more quickly without degrading the temperature distribution. The results below in
Figure 12. Development of
of an
an equation for
for the mixing coefficient based on
on inlet
inlet velocity.
velocity.
Figure 13 show that 12.
Figure significant mixing
Development is occurring
equation the(light
mixingblue colour
coefficient of water)
based at 2 cm/s.
The relationship 𝑍𝑍 = −3 × 107 𝑣𝑣 + 254,526 shown above represents the expected mixing when
a fully-developed thermocline is formed. Therefore, for any system, the inlet flow velocity required
to fully form a thermocline can be determined.
Once the thermocline is formed for the TES systems, increases in inlet flow velocity allow the
tank to charge more quickly without degrading the temperature distribution. The results below in
Figure 13 show that significant mixing is occurring (light blue colour of water) at 2 cm/s.

(a) (b) (c)

Figure 13.13.TES-1
Figure TES-1atat90
90°C,
◦ C, 2.0
2.0 cm/s:
cm/s: (a) 400s;s;(b)
(a) 400 (b)800
800s;s;(c)(c) 1200
1200 s. s.

However,
However, when
whenthe fully-developed
the fully-developed thermocline
thermocline is formed
is formed in Figurein14Figure 14 (inlet
(inlet velocity of 0.9velocity
cm/s), of
0.9 cm/s), thevelocity
the inlet inlet velocity
was then was then increased
increased (a) toat2 414
to 2 cm/s cm/s at
A 414
s.(b) smalls. effect
A small effect is with
(c)is observed observed with a
a larger
larger thermocline
thermocline ats500
at 500 a sresult
asFigureas 13.
aofresult of the velocity
the velocity
TES-1 at 90 °C,increase;
increase;
2.0 cm/s: however,
however,
as800
(a) 400 s; (b) charging ass. charging
continues,
s; (c) 1200
continues, the
the thermocline
thermocline
stabilises stabilises
and does not and does significantly.
degrade not degrade Thissignificantly. This with
result is consistent resultthe is consistentresults
experimental with the
experimental
reported results
However,
by reported
when
the first by the[2].
first author
the fully-developed
author Karim Karim [2].
thermocline is formed in Figure 14 (inlet velocity of
0.9 cm/s), the inlet velocity was then increased to 2 cm/s at 414 s. A small effect is observed with a
larger thermocline at 500 s as a result of the velocity increase; however, as charging continues, the
thermocline stabilises and does not degrade significantly. This result is consistent with the
experimental results reported by the first author Karim [2].
Energies 2018, 11, 1049 13 of 15
Energies 2018, 11, x FOR PEER REVIEW 13 of 15

(a) (b) (c) (d) (e)

Figure 14. TES-1

Figure 14. TES-1 at
at 90
90 °C:
◦ C: (a)
(a) 400
400 s;
s; (b)
(b) 500
500 s;
s; (c)
(c) 800
800 s;
s; (d)
(d) 1200
1200 s;
s; (e)
(e) 1600
1600 s.
s.

5. Discussion
5. Discussion
The stream of water entering the thermal storage tank causes mixing at the entry point and thus
The stream of water entering the thermal storage tank causes mixing at the entry point and
affects the formation of the thermocline. Therefore, introducing the water at the right velocity is a
thus affects the formation of the thermocline. Therefore, introducing the water at the right velocity
prime concern in TES. The work in [16] reports that mixing causes the loss of heating/cooling capacity
is a prime concern in TES. The work in [16] reports that mixing causes the loss of heating/cooling
during charging and discharging cycles, which depends on the inlet velocity, the temperature
capacity during charging and discharging cycles, which depends on the inlet velocity, the temperature
difference between incoming and stored water and the distance between the outlet and inlet ports.
difference between incoming and stored water and the distance between the outlet and inlet ports.
Analytical and simulated results in this study confirm that mixing is influenced by inlet velocity,
Analytical and simulated results in this study confirm that mixing is influenced by inlet velocity,
tank aspect ratio and the temperature differential between stored and inlet water. Both analyses
tank aspect ratio and the temperature differential between stored and inlet water. Both analyses
showed that mixing was reduced by increasing the tank aspect ratio, increasing the water
showed that mixing was reduced by increasing the tank aspect ratio, increasing the water temperature
temperature differential and reducing inlet water velocity. The temperature difference is shown to
differential and reducing inlet water velocity. The temperature difference is shown to greatly influence
greatly influence mixing and thermocline formation. For example, at a temperature difference of
mixing and thermocline formation. For example, at a temperature difference of 84 ◦ C, the effect of the
84 °C, the effect of the tank aspect ratio in mixing was minimum. However, as the temperature
tank aspect ratio in mixing was minimum. However, as the temperature difference is reduced,
difference is reduced, the effects of the tank aspect ratio resulted in a greater mixing. Simulation
the effects of the tank aspect ratio resulted in a greater mixing. Simulation results show that
results show that differences in mixing and the thermocline between TES-1 2, and 3 systems are
differences in mixing and the thermocline between TES-1 2, and 3 systems are minimal, as supported
minimal, as supported by the theoretical analysis. TES-4 and 5 show significantly higher mixing than
by the theoretical analysis. TES-4 and 5 show significantly higher mixing than the other systems.
the other systems. The mathematical relationship proposed in Equation (13) for a fully-developed
The mathematical relationship proposed in Equation (13) for a fully-developed thermocline will
thermocline will provide designers with general guidelines for developing thermally-efficient
provide designers with general guidelines for developing thermally-efficient systems by choosing the
systems by choosing the right operating and design parameters.
right operating and design parameters.
Inlet flow velocity can be measured using dimensionless parameters of Re and Ri. Mixing
Inlet flow velocity can be measured using dimensionless parameters of Re and Ri. Mixing reduces
reduces when Re is lowered and Ri is increased. Re values ranged between 7000 and 2000 for
when Re is lowered and Ri is increased. Re values ranged between 7000 and 2000 for thermocline
thermocline formation in this analysis. It can be mentioned here that for the same value of the
formation in this analysis. It can be mentioned here that for the same value of the Reynolds number,
Reynolds number, it is possible to have several values of the velocity as the Reynolds number
it is possible to have several values of the velocity as the Reynolds number depends on other variables,
depends on other variables, as shown in Equation (3). However, Re is mainly dominated by the
as shown in Equation (3). However, Re is mainly dominated by the velocity, as other changes are
velocity, as other changes are negligible for the same tank (characteristic length constant). For
negligible for the same tank (characteristic length constant). For example, if temperature is increased,
example, if temperature is increased, both the viscosity and density of water will decrease, and the
both the viscosity and density of water will decrease, and the combined effect of these two variables
combined effect of these two variables will be minimal. CFD modelling done by Walmsley et al.
will be minimal. CFD modelling done by Walmsley et al. showed that increasing inlet flow rate caused
showed that increasing inlet flow rate caused the thickness of the thermocline to be larger, indicating
the thickness of the thermocline to be larger, indicating more mixing [18]. Richardson’s number has
more mixing [18]. Richardson’s number has been used to determine the effect of mixing under
been used to determine the effect of mixing under different conditions. Bouhdjar and Hardad state
different conditions. Bouhdjar and Hardad state that stratification performance is improved with
that stratification performance is improved with higher Richardson numbers [19].
higher Richardson numbers [19].
The findings of this study should be of tremendous help to designers. They should make an
The findings of this study should be of tremendous help to designers. They should make an
optimum balance between the flow rate (inlet velocity), temperature difference and tank aspect ratio
optimum balance between the flow rate (inlet velocity), temperature difference and tank aspect ratio
in designing TES systems. A lower flow rate will require a longer time to fill the tank, but will have
in designing TES systems. A lower flow rate will require a longer time to fill the tank, but will have
better efficiency for TES. Moreover, at a lower flow rate, a smaller chiller or heater will be required
for the same heating/cooling capacity. Therefore, a lower flow rate in TES systems outweighs the
Energies 2018, 11, 1049 14 of 15

better efficiency for TES. Moreover, at a lower flow rate, a smaller chiller or heater will be required
for the same heating/cooling capacity. Therefore, a lower flow rate in TES systems outweighs the
disadvantage of longer charging cycles. Similarly, tanks with a higher aspect ratio (i.e., taller tanks)
will take up a much smaller space compared to tanks with a lower aspect ratio.

6. Conclusions
The analysis of stratified thermal storage systems allows several conclusions to be drawn:

• Higher inlet velocities result in increased mixing of water.

• TES tanks with higher aspect ratios result in reduced mixing and heat loss due to smaller relative
contact area between water temperature layers.
• Higher temperature differences between inlet water and stored tank water reduce mixing because
of greater density differences between layers.
• Thermoclines can form at higher inlet velocities when the aspect ratio and temperature difference
are greater.

Other aspects affecting mixing and stratification such as placement of inlet and outlet ports,
diffuser design and charging and recharging duration have not been investigated in this study. Future
studies considering these factors will bring better insight to stratification and storage performance.

Author Contributions: Azharul Karim in the leader of the project and developed and progressed the concept.
Ashley Burnett conducted the simulation work and prepared the report of the results. Azharul Karim and Sabrina
Fawzia drafted the manuscript and managed the submission and review process.
Conflicts of Interest: The authors declare no conflict of interest.

Nomenclature
d tank diameter
g acceleration due to gravity
h vertical distance between ports (tank height)
Re Reynolds Number
Ri Richardson Number
∆T temperature difference of inlet and tank water
v inlet velocity
Z mixing coefficient
Greek Letters
β thermal expansion coefficient
ρ fluid density
µ dynamic viscosity
Abbreviations
AR Aspect Ratio
CFD Computational Fluid Dynamics
TES Thermal Energy Storage

References
1. Shin, M.S.; Kim, H.S.; Jang, D.S.; Lee, S.N.; Lee, H.S.; Yoon, H.G. Numerical and experimental study on the
design of a stratified thermal storage system. Appl. Therm. Eng. 2004, 24, 17–27. [CrossRef]
2. Karim, M.A. Experimental investigation of a stratified chilled-water thermal storage system.
Appl. Therm. Eng. 2010, 31, 1853–1860. [CrossRef]
3. Karim, M.A.; Hawlader, M.N.A. Mathematical modelling and experimental investigation of tropical fruits
drying. Int. J. Heat Mass Transf. 2005, 48, 4914–4925. [CrossRef]
4. Karim, M.A.; Hawlader, M.N.A. Development of solar air collectors for drying applications.
Energy Convers. Manag. 2004, 45, 329–344. [CrossRef]
Energies 2018, 11, 1049 15 of 15

5. Karim, M.A.; Perez, E.; Amin, Z.M. Mathematical modelling of counter flow v-grove solar air collector.
Renew. Energy J. 2014, 74, 116–123. [CrossRef]
6. Islam, M.; Miller, S.; Yarlagadd, P.K.D.V.; Karim, M.A. Investigation of the effect of physical and optical
factors on the optical performance of a parabolic trough collector. Energies 2017, 10, 1907. [CrossRef]
7. Khan, M.I.H.; Nagy, S.A.; Karim, M.A. Transport of cellular water during drying: An understanding of cell
rupturing mechanism in apple tissue. Food Res. Int. 2018, 105, 772–781. [CrossRef] [PubMed]
8. Kumar, C.; Karim, M.A. Microwave-convective drying of food materials: A critical review. Crit. Rev. Food
Sci. Nutr. 2017, 1–16. [CrossRef] [PubMed]
9. Kumar, C.; Joardder, M.U.H.; Farrell, T.W.; Karim, M.A. Multiphase porous media model for intermittent
microwave convective drying (IMCD) of food. Int. J. Therm. Sci. 2016, 104, 304–314. [CrossRef]
10. Mawire, M. McPherson, Experimental and simulated temperature distri-bution of an oil-pebble bed thermal
energy storage system with a variable heat source. Appl. Therm. Eng. 2009, 2, 1086–1095. [CrossRef]
11. Guyer, E.C.; Brownell, D.L. Review of Heat Storage Materials; Final Report (No. EPRI-EM-3353); Dynatech
R/D Co.: Cambridge, MA, USA, 1983.
12. Njoku, H.O.; Ekechukwu, O.V.; Onyegegbu, S.O. Analysis of stratified thermal storage systems: An overview.
Heat Mass Transf. 2014, 50, 1017–1030. [CrossRef]
13. Haller, M.Y.; Cruickshank, C.A.; Streicher, W.; Harrison, S.J.; Anderson, E.; Furbo, S. Methods to
determine stratification efficiency of thermal energy storage processes—Review and theoretical comparison.
J. Sol. Energy 2009, 83, 1847–1860. [CrossRef]
14. Osman, K.; Khaireed, S.M.N.A.; Ariffin, M.K.; Senawi, M.Y. Dynamic modelling of stratification for chilled
water storage. J. Energy Conserv. Manag. 2008, 49, 3270–3273. [CrossRef]
15. Dincer, I.; Rosen, M. Thermal Energy Storage: Systems and Applications, 2nd ed.; Wiley: Hoboken, NJ, USA,
2011.
16. Nelson, J.E.B.; Balakrishnan, A.R.; Murthy, S.S. Experiments of stratified chilled-water tanks. Int. J. Refrig.
1999, 22, 216–234. [CrossRef]
17. ANSYS. (2009). ANSYS CFX—Solver Theory Guide. Available online: http://www1.ansys.com/customer/
content/documentation/120/cfx/xthry.pdf (accessed on 1 March 2017).
18. Walmsley, M.; Atkins, M.; Linder, J.; Neale, J. Thermocline Movement Dynamics and Thermocline Growth in
Stratified Tanks for Heat Storage. Chem. Eng. Trans. 2010, 21, 991–996.
19. Bouhdjar, A.; Hardad, A. Numerical analysis of transient mixed convection flow in storage tank: influence
of fluid properties and aspect ratios on stratification. J. Renew. Energy 2002, 25, 555–567. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).