BUILD SIMUL (2019) 12: 869–879

[Link]

Numerical investigation on the melting of nanoparticle-enhanced PCM

in latent heat energy storage unit with spiral coil heat exchanger

Research Article
Ruiqing Du1,2, Wenxin Li1,2, Teng Xiong3, Xun Yang1,2, Yong Wang1,2 (), Kwok Wei Shah3

1. National Centre for International Research of Low-Carbon and Green Buildings, Ministry of Science & Technology, Chongqing University,
Chongqing 400045, China
2. Joint International Research Laboratory of Green Buildings and Built Environments, Ministry of Education, Chongqing University, Chongqing
400045, China
3. Department of Building, School of Design and Environment, National University of Singapore, Singapore 117566, Singapore

Abstract Keywords
Latent heat storage units are widely used in building heating systems due to its high energy nanoparticle-enhanced,
storage density, whereas the practical performances of them are limited by the low thermal latent heat storage unit,
conductivities of phase change materials. In this paper, copper nanoparticles were added into spiral coil heat exchanger,
paraffin to enhance the heat transfer rate of a latent heat storage unit using a coil heat exchanger. heat transfer fluid,
A three-dimensional numerical model was built to simulate the melting process of phase change phase change material
material, and it was well validated against the experimental data. The simulation results showed
that the nanoparticle-enhanced phase change material saved 19.6% of the total melting time
Article History
Received: 14 September 2018
consumed by the pure phase change material. In addition, the dispersion of nanoparticles
Revised: 2 January 2019
significantly alleviated the temperature non-uniformity in the unit. Moreover, for the unit using
Accepted: 13 February 2019
nanoparticle-enhanced phase change material, the flow rate of heat transfer fluid was not
recommended higher than 0.75 m3/h. The dispersion of nanoparticles could enlarge the optimum
© Tsinghua University Press and
heat transfer fluid temperature range to 60–70 °C compared with that of pure phase change
Springer-Verlag GmbH Germany,
material (60–65 °C). Therefore, the application of nanoparticle-enhanced phase change material in
part of Springer Nature 2019
the latent heat storage unit can significantly enhance heat transfer, and the proposed optimum
inlet heat transfer fluid temperature range could contribute to higher energy efficiency.

1 Introduction industries such as in buildings (Alizadeh and Sadrameli

2016).

Building Systems and

The global energy crisis motivates the world to find The heat transfer efficiency of the PCM is crucial to the
environmental-friendly renewable energy such as in solar performance of the LHS unit. Compared to units with
Components
energy. Solar energy could avoid harmful emissions and straight pipes, a spiral coil heat exchanger achieves a greater
provide an infinite amount of energy (Thomas and André transfer coefficient due to the secondary flow of fluid (Guo
2012). However, the collection of solar energy is only available and Huai 2016). Bezyan et al. (2015) numerically studied
during the daytime and could be affected by weather the thermal performance of a vertical spiral shaped heat
conditions. To continuously heat the buildings throughout exchanger. The results showed that spiral shaped heat
the year, the latent heat storage (LHS) unit filled with phase exchanger performed better than the 1-U shaped and 1-W
change material (PCM) could be installed in heating systems shaped heat exchanger. Korti and Tlemsani (2016) numerically
(Delcroix et al. 2015). Due to the characteristic of chemical studied the melting process of the LHS unit with a vertical
stability, the constant temperature during phase change coil, and found that a 4 °C increase of the inlet temperature
and high energy storage density of PCM (Soares et al. 2017), of heat transfer fluid (HTF) could save the melting time of
LHS units are widely used in energy storage for various PCMS by up to 54.5%. Meanwhile, the HTF flow rate has a

E-mail: cqwangyong@[Link]
870 Du et al. / Building Simulation / Vol. 12, No. 5

List of symbols

C mushy zone constant V volume of the nano-PCM (m3)

Cp specific heat (J/(kg·K)) α thermal expansion coefficient (K−1)
d diameter (m) β liquid fraction
E/E0 cumulative energy transfer through the hot wall μ dynamic viscosity (kg/(m·s))
during the melting process inside tall C-TES system ρ density (kg/m3)

g gravity acceleration vector (m/s2) φ volume fraction of nanoparticles (%)
H total PCM enthalpy (J/kg)
Subscripts
ΔH latent heat content (J/kg)
h sensible enthalpy (J/kg) 0 initial
hsf latent heat of fusion (kJ/kg) c charging
k thermal conductivity (W/(m·°C)) h hot
L latent heat (kJ/kg) l liquidus
l length of the tube of coil heat exchanger m melting
P pressure (Pa) n nanoparticles
r radius of the tube of coil heat exchanger (m) nf liquid NePCM
rin inner radius of the tube of coil heat exchanger (m) ns solid NePCM
S source term p phase change material
T temperature (°C) ref reference
ΔT temperature difference (°C) s solidus
t time (s) w water

u velocity vector (m/s)

significant effect on the charging process if it is in 15–20 L/h, They found a reduction of 25.73% in evaporator temperature
while a negligible impact once it exceeded 20 L/h. Tay et al. with 1% volume fraction of Al2O3 and the maximum
(2012) investigated latent thermal energy storage (LTES) energy storage with 2% volume fraction of Al2O3. Das et al.
system with four serial coiled tubes and a cylindrical PCM (2017b) added different carbon-based nanocomposites
tank. It was observed to be efficient only when there was a (nano-diamond, SWCNT and graphene nanoplatelets) to
lower temperature difference between the heat exchanger n-eicosane and applied the NePCM in a vertical cylindrical
and PCM. Chen et al. (2016) found that the inlet temperature storage unit with an isothermal inner tube. The result
of HTF significantly influenced the melting process of presented an increase of 15% in melting rate with 1% volume
paraffin/graphite composite PCM, while the effect on the fraction of nano-diamond, an increase of 15% in melting
Reynolds number was not great. Therefore, the operational rate with 1% volume fraction of SWCNT and an increase of
temperature and flow rate of the HTF should be best selected 15% melting rate with 1% volume fraction of graphene
in the process of designing an LTES system. In this process nanoplatelets (GnP). They also conducted similar research
of design, the Reynolds number was not recommended to in a horizontal shell with a tube heat exchanger (Das et al.
be higher than 8700. 2017a). The results showed an increase of 2% in melting
The performance of the LHS unit can be significantly rate with 1% volume fraction of nano-diamond, an increase
enhanced by the better heat transfer efficiency of the PCM, of 27% in melting rate with 1% volume fraction of SWCNT
which is determined by its thermal conductivities. To enhance and an increase of 40% in melting rate with 1% volume
the heat transfer, nanoparticles (Li and Rogovchenko 2014), fraction of GnP. Alomair et al. (2018) investigated the
metal foam (Chen et al. 2014), multiple PCMs (Tian and melting process of a nano-PCM (copper oxide/coconut oil
Zhao 2013) and porous matrix (Elbahjaoui and El Qarnia bio-based PCM) in a concentric cylindrical thermal energy
2016) were applied. Among these methods, the dispersion storage system numerically and experimentally. They
of high thermal conductive nanoparticles in PCM is addressed compared the melting fraction, temperature distribution,
in this paper. Numerous studies have investigated the energy storage rate and heat transfer of pure PCM and
characteristics of the tube-in-tank LHS units with NePCM. nano-PCM. Both numerical and experimental results
Krishna et al. (2017) added Al2O3 nanoparticles in Tricosane exhibited that the melting rate was faster in nano-CPM
as NePCM to fill the cylindrical storage tank with a heat pipe. compared to the base-PCM.
Du et al. / Building Simulation / Vol. 12, No. 5 871

According to the literature reviewed, there were a limited

number of articles focusing on heat transfer enhancement
of LHS unit system using coil heat exchangers, and few
have considered to combining spiral heat exchangers and
NePCM. Therefore, the focus of this study is to investigate
the melting process of NePCM in an LHS unit with coil heat
exchangers numerically. Copper nanoparticles in NePCM
were synthesized by a controlled disproportionation of CuCl
in oleylamine reaction medium without the involvement of
strong reducing agents adopted in conventional approaches.
The particular NePCM used in this study shows excellent
stability and small sedimentation for a long time (Sreethawong
et al. 2014). The effects of NePCM on melting time of
the unit and the temperature distribution were studied. In
addition, the effects of inlet temperature and flow rate of HTF
on melting time and average charging power of NePCM were
investigated. The optimum ranges of inlet HTF temperature
and flow rate were proposed. The current outcome will
contribute to the design of the solar heating system with
LHS unit.

2 Numerical study
Fig. 1 3D geometry of LHS unit and mesh results of the latent
heat storage unit: (a) 3D geometry of LHS unit; (b) mesh results
2.1 Characterization of paraffin/nanoparticle-enhanced of the latent heat storage unit
PCM
and a cylindrical shell. The shell made of polyethylene and
A differential scanning calorimeter (DSC, Q20, TA was shaped as a vertical cylinder with an inner radius of
Instrument Inc., USA) with an average error of 1.0% was 200 mm, and a height of 1600 mm, acted as the container
used to measure the enthalpy and the heat capacity of PCM. of PCM. HTF (water) flowed inside the spiral copper coil
A thermal resistance tester (TIT, QuickLine-10-C, Anter tubes, which extended along the vertical axis of the PCM
Corp., USA) with a maximum error of 3.0% was used to tank, to provide heat for the thermal storage system. The
measure the thermal conductivity of PCM. Thermophysical outer diameter of the tube was 15mm and the thickness was
properties of pure PCM and NePCM are listed in Table 1 1 mm. The pitch of the spiral coil tubes was 80 mm and the
and imputed to the numerical simulation. spiral coil diameter was 140 mm. The HTF outlet was situated
at the center of the cylinder, and the distance between
2.2 Numerical simulation the inlet and outlet was 70 mm. In this LHSU, paraffin was
used as PCM.
Figure 1(a) presents the configuration of the cylindrical LHS In this study, the commercial software Gambit was used
unit used in this study, including a spiral coil heat exchanger for mesh generation and all governing equations together
with the boundary conditions over the whole domain were
Table 1 Thermophysical properties of PCM and NePCM solved by the commercial CFD code FLUENT 15.0. As shown
Physical NePCM NePCM Pure in Fig. 1(b), Tetra mesh and unstructured mesh were used.
parameters (φ = 5%) (φ = 2.5%) paraffin The Semi-Implicit Method for Pressure-Linked Equations
Density ρ (kg/m3) 932.2 920.3 890.5 (SIMPLE) algorithm was applied for the pressure–velocity
2457 (solid) 2531 (solid) 3110 (solid) coupling. The Pressure Staggering Option (PRESTO) scheme
Specific heat Cp (J/(kg·K))
3750 (liquid) 3820 (liquid) 4810 (liquid) was applied for the pressure interpolation. The standard k
Thermal conductivity λ
0.226 0.189 0.174 model was used to describe the turbulent flow of HTF. The
(W/(m·K))
turbulent intensity and turbulent viscosity ratio for the HTF
Dynamic viscosity μ (kg/(m·s)) 0.033 0.035 0.036
inlet were set as 5% and 10%, respectively. A second-order
Latent heat L (J/kg) 1.60×105 1.64×105 1.66×105 upwind scheme was adopted for the discretization of the
Melting point Tm (K) 332.72 329.72 327.53 momentum equation, energy equation, turbulent kinetic
Energy density (kJ/m3) 131538.46 134528.93 138461.54 energy equation and dissipation rate equation. Solution
872 Du et al. / Building Simulation / Vol. 12, No. 5

convergence target was set as 10−7 for the energy equation According to the subject and content of the study,
and 10−3 for other variables. To achieve the optimum value three-dimensional transient governing equations were
for grid number and time step, the precise investigation established based on the enthalpy-porosity method (Archibold
had been conducted before the main simulation. The liquid et al. 2014):
fraction (β) (see Eq. (9)) was selected to test the mesh Continuity equation:
dependency. As listed in Fig. 2, once the mesh elements 
exceeded 2,640,722, the liquid fraction became stable. ⋅ u = 0 (1)
Considering the computational efficiency and accuracy,
Momentum equation:
mesh with elements numbers of 1,113,043 were used for
further simulation in this study. In addition, once the time 
¶u   
step is smaller than 0.2 s, the liquid fraction become stable. ρ + ρ(  ⋅ u )u = -P + μ2 u
¶t 
Therefore, the simulation was performed with a fixed time 
+ ρg [ 1 - α ⋅ ( T - Tref ) ] + S (2)
step of 0.2 s.
In addition to this, several assumptions were made to
Energy equation:
simplify the CFD model as follow:
(1) The initial temperature of the TES unit is constant and ¶ 
uniform; ( ρH ) +  ⋅ ( ρuH ) =  ⋅ ( kT ) (3)
¶t
(2) Thermal resistance between nanoparticle-enhanced PCM
and heat exchanger tube, the HTF and heat exchanger where H is the total PCM enthalpy and it could be written as:
tube are negligible;
(3) The boundaries outside the PCM and straight tubes are H = h + ΔH (4)
adiabatic. The outer surface of the LHS unit is adiabatic.
where h is the sensible enthalpy and can be calculated as
(Seddegh et al. 2016):
Tc
h = href + ò Cp dT (5)
Tref

where href is the reference enthalpy under the reference

temperature Tref.
ΔH is the latent heat content that varies from 0 (solid
state) to L (liquid state) and can be written as (Seddegh et al.
2016):

ΔΗ = βL (6)

where L is the latent heat, β is the liquid fraction and it can

be expressed as (Seddegh et al. 2016):

ìï 0 if Tp £ Ts
ïï
β = ïí1 if Tl £ Tp (7)
ïï
ïïî ( Tp - Ts ) / ( Tl - Ts ) if Ts < Tp < Tl

The source term S in Eq. (2) is formulated as below
(Seddegh et al. 2016):

 
2 u
S = C (1 - β ) 3 (8)
β

where C is the mushy zone constant and ranging from 104

Fig. 2 Grid dependency and time step tests for numerical simulation: to 107. Several studies investigated the influence of the mushy
(a) grid dependency for numerical simulation; (b) time step tests zone constant on numerical results and found that 105–106
for numerical simulation was suitable for most cases (Shmueli et al. 2010). According
Du et al. / Building Simulation / Vol. 12, No. 5 873

to previous studies, 105 was applied. For HTF and PCM, the
initial temperature (30 °C) was equal to the temperature
that the data acquisition instrument read at the start of the
experimental test. Table 2 shows the different HTF inlet
temperatures and flow rates of different numerical simulation
cases. The melting peak of the phase change material was
about 52.44 °C. The inlet water temperature of 60 °C and
the flow rate of 0.25 m3/h were chosen as a base. The HTF
flow in all cases is turbulent flow because the Reynolds
number range for all cases is 17986.11–58235.5. According
to the research methods found in published articles (Fan
et al. 2016), a variety of values of HTF inlet temperature
and flow rate were chosen for different cases.
Fig. 3 Test-bed schematic diagram
Table 2 Operating conditions of HTF
Operating Inlet temperature Flow rate
condition (°C) (m3/h) PCM
Case 1 60 0.25 Pure
Case 2 60 0.25 NePCM (φ = 2.5%)
Case 3 60 0.25 NePCM (φ = 5%)
Case 4 60 0.50 NePCM (φ = 5%)
Case 5 60 0.75 NePCM (φ = 5%)
Case 6 60 1.00 NePCM (φ = 5%)
Case 7 65 0.25 NePCM (φ = 5%)
Case 8 70 0.25 NePCM (φ = 5%)
Case 9 75 0.25 NePCM (φ = 5%)

2.3 Validation of simulation

An experimental system was set up in Chongqing University,

China to validate the numerical model (Wang et al. 2017).
The schematic diagram and photos of the experimental
system are presented in Figs. 3 and 4. The system mainly
consisted of an electric heater, a cylindrical LHS unit, a
Fig. 4 Photos of the test-bed: (a) overall view; (b) electric heater;
pump, a flow meter, an expansion tank, and connecting (c) latent heat storage unit with insulation; (d) flow meter
pipes. In this system, water was used as HTF. Water was
heated by the electric heater with a temperature controller The experimental uncertainty analysis was conducted.
and maintained at 60 °C to ensure a stable HTF inlet Five same T-type thermocouples were pre-calibrated and
temperature. The HTF inlet flow rate was regulated at used to measure temperature, and the uncertainty of each
0.25 m3/h by a valve and a turbine flow meter was applied thermocouple is ±0.5 °C. A turbine flow meter with a
to measure the flow rate. The HTF was pumped by the measuring range of 0.2–1.0 m3/h and uncertainty of ± 0.2%
centrifugal pump flowed through the spiral coil tube to was used to measure flow rate. The temperature controller
provide heat to PCM. with an uncertainty of ±0.5 °C was used to keep the inlet
As shown in Fig. 3, three T-type thermocouples (T1–T3) temperature of HTF constant. In addition, for the DSC
were set in the tank to measure the temperature of PCM. used for thermal properties, it’s measuring range was −180
Another two T-type thermocouples (T4–T5) were set at to 725 °C, and the uncertainty was ±1%. The measuring
the inlet and outlet of the coil to measure the temperature range of TIT was 0.01 to 20 W/(m·K) and the uncertainty
of HTF. These five thermocouples (T1–T5) were linked was ± 3.0%.
to a data acquisition instrument (34972A, Agilent, Santa Figure 5(a) shows the numerical and experimental results
Clara, CA, USA). of temperature values of T1–T3. There was no constant
874 Du et al. / Building Simulation / Vol. 12, No. 5

3 Result and discussion

3.1 Effect of the dispersion of nanoparticles

The study in this section investigated the effect of nano-

particles on the performance of the LHS unit. Figure 6 shows
the liquid fraction (β) curves of the PCM for φ = 0% , 2.5%
and 5% when the constant inlet HTF temperature was 60 °C
and flow rate was 0.25 m3/h. The β can be calculated from
the following equation (Ebadi et al. 2018):

Vliquid PCM
β= (9)
Vliquid PCM + Vsolid PCM

where V represents the volume of the nano-PCM in solid

or liquid form as indicated in the subscript. It took 10000 s
for pure PCM (φ = 0%) to finish the phase change process,
and 9120 s for φ = 2.5%, 8080 s for φ = 5%. Dispersing
nanoparticles at φ = 2.5% shows a gentle impact on melting
time (8.8% less), while the total melting time was shortened
by nearly 19.6% for φ = 5%. Therefore, the volume fraction
Fig. 5 Experimental and numerical temperature result comparison: at 5% was suitable as part of the practical engineering.
(a) experimental and numerical temperature profiles of ther- To present energy stored rate inside the LHS system,
mocouples inside the LHSU over time for Case 1; (b) discrepancy the cumulative energy storage (E) through the heater surface
between the numerical and experimental results during the to the PCM up to a specified time during the melting
charging process process can be calculated using the time integration given
in Eq. (10). The Qh (t) in Eq. (10) is the instantaneous heat
temperature region due to the motion of molten PCM transfer, which can be estimated from Eq. (11) (Ebadi et al.
induced by natural convection. In addition, the PCM was 2018). E can be put into its dimensionless form by using E0
heated continuously, and the temperature of these three (Eq. (12)), which represent maximum possible energy that
points increased continuously. The trends of temperature can be stored inside the LHS system during the melting
variation curves in experimental and numerical results were process (Alomair et al. 2018).
very similar. All three measuring points finally reaching t
333 K, were different from the experimental results. The E =ò Qh (t )dt (10)
0
differences can be explained by the heat loss from the PCM
ltube ¶T
tank in the experiment. Moreover, the heat resistance between Qh (t ) = (πdtube )ò (-k ¶r )r =r dl (11)
0 in
nanoparticle-enhanced PCM and heat exchanger tube were
neglected in the simulation. Furthermore, higher absolute
errors were observed before the second phase change process,
although the temperature difference in the second process
was small. That meant that the errors were higher when the
nanoparticle-enhanced PCM was in the solid state compared
to when it was in the liquid state. That could be explained
by the heat resistance of nanoparticle-enhanced PCM in its
solid state is much higher than that in the liquid state. That
was due to the bad contact of solid paraffin and heat
exchanger. As shown in Fig. 5(b), above all, the relative errors
of numerical results were all lower than 7.5%, which meant
that the model was fit enough to predict the behavior of
LHSU system. Therefore, the following study was based on Fig. 6 Volume-average liquid fraction profiles of pure PCM and
the simulation data. NePCM in latent heat storage units
Du et al. / Building Simulation / Vol. 12, No. 5 875

where dtube is the diameter of the tube of coil heat exchanger,

ltube is the total length of the tube of coil heat exchanger.
2
πdinner H unit - πdtube
2
ltube é
E0 = ê Cp,ns ρns ( Tm - T0 )
4 ë
ρns + ρnf
+(
2
)hsf,n + Cp,nf ρnf ( Th - Tm ) ùúû (12)

where dinner is the inner diameter of the latent heat storage

unit, Hunit is the height of the unit, and hsf,n is the latent heat
of fusion of NePCM. The E/E0 for φ = 0%, 2.5% and 5% is
shown in Fig. 7. E/E0 increases rapidly in the initial period
of time due to larger temperature differences between heater
surface and PCM. Then E/E0 increases linearly before
becoming non-linear and approaching to 1 once thermal
equilibrium is established. The latent heat capacity decreases
and sensible heat capacity increases after the dispersion of
nanoparticles to base-PCM. The energy storage rate of LHS
system is higher with the increasing volume fraction of
nanoparticles.
Figure 8 presents temperature and liquid fraction
contours of a vertical plane cutting the unit through its
longitudinal axis. The bottom areas of units are circled in
Fig. 8. In the LHS unit filled with pure PCM, it took 1920 s
for the portion of PCM in the bottom area of the unit to be
completely melted even though the melting process of PCM
in the other part had ended. The volume of the bottom part
occupied less than 5% of the entire unit, while its melting
time took nearly 24.3% of the total melting time, which
means the temperature non-uniformity lead to a waste of
time Figs. 8(a)–(b). The reason was that the main way of
heat transfer in this area was heat conduction rather than
natural convection or the combination of these two. However,
the dispersion of copper nanoparticles solved this problem.
The high thermal conductivity of the NePCM enabled heat
to be timely transferred within the unit. By 8080 s, the PCM

Fig. 8 Volume-average temperature and volume-average liquid

Fig. 7 Dimensionless cumulative energy transfer profiles of pure fraction contour of melting processes of pure PCM and NePCM
PCM and NePCM in latent heat storage units in latent heat storage units
876 Du et al. / Building Simulation / Vol. 12, No. 5

Fig. 9 Volume-average liquid fraction profiles for NePCM in a

latent heat storage unit at different HTF inlet flow rates

Fig. 8 Volume-average temperature and volume-average liquid

fraction contour of melting processes of pure PCM and NePCM
in latent heat storage units (Continued)

in the bottom area of the unit had been completely phase

changed almost synchronously with the PCM in other parts
Figs. 8(c)–(d).

3.2 Effect of HTF inlet parameters

The melting process of NePCM in LHS unit is affected by

HTF inlet temperature and flow rate, which will be discussed
in this section.
Fig. 10 Temperature difference between the inlet and outlet
3.2.1 Effect of HTF inlet flow rate
water at different HTF inlet flow rates
The nanoparticle-enhanced PCM liquid fraction curves with
the HTF inlet temperature of 60 °C are shown in Fig. 9. It curves saw a similar trend and the temperature differences
took 8080, 7811, 7498, and 7329 s for nanoparticle-enhanced were 1.5 °C higher on average (Yang et al. 2017).
PCM to finish the phase change process when HTF inlet The average charging power can be calculated from the
flow rate was 0.25, 0.50, 0.75, and 1.00 m3/h, respectively. following equation:
The HTF inlet flow rate can give rise to shorter thermal t melt
storage time, but the effect was insignificant, especially when
P=
ò 0
Cp,w f ρw ΔTdt
(13)
HTF inlet flow rate was higher than 0.75 m3/h. Therefore, t melt
accelerating the energy storage process by applying an
inlet flow rate higher than 0.75 m3/h is not recommended, where P is the average charging power, tmelt is the total
considering the additional energy consumption of the pump. melting time, f is the flow rate of HTF, ΔT is the
As shown in Fig. 10, with the inlet water temperature temperature difference between the inlet and out HTF. The
at constant (60 °C), the effect of HTF inlet flow rates average charging power of the thermal storage unit in
(0.25–1.00 m3/h) on the temperature difference between the Cases 2 to 5 is shown in Fig. 11. In the case where the inlet
inlet and outlet water was investigated. The form NePCM water temperature was constant, the effect of the increasing
stored energy was sensible energy in the initial period of the inlet flow rate (from 0.25 to 1.00 m3/h) on the average
charging process, and there observed a sharp drop in the charging power was small (from 0.756 to 0.789 kW). The
temperature difference between inlet and outlet water. When reason was that the heat transfer area dominated the heat
form NePCM stored energy became latent energy, the results exchange capacity of the heat exchanger, although the
reached a stable value. Comparing the results of PCM, the increasing inlet flow rate could increase the forced convection
Du et al. / Building Simulation / Vol. 12, No. 5 877

Fig. 11 Average charging power of NePCM at different HTF inlet Fig. 12 Volume-average liquid fraction profiles for NePCM in a
flow rates latent heat storage unit at different HTF inlet temperatures

strength in the spiral heat exchanger and reduce the heat

transfer resistance of the heat exchanger. In practical
engineering applications, a smaller flow rate should be chosen
to reduce the energy consumption of the circulating water
pump for energy saving.

3.2.2 Effect of HTF inlet temperature

Figure 12 shows the liquid fraction curves when HTF inlet

flow rate was 0.25 m3/h. It took 8080, 6771, 5870 and 5278
s for NePCM finishing phase change process with the HTF
inlet temperature of 60, 65, 70, and 75 °C, respectively. A
higher HTF inlet temperature gave rise to the shorter thermal
storage time, and the effect was more significant than that
of the inlet flow rate shown in Fig. 9. However, the difference
of total melting time got less apparent with an increasing Fig. 13 Temperature difference between the inlet and outlet
inlet temperature. The reason was that the heat transfer water at different HTF inlet temperature
efficiency was dominated by the heat transfer area rather
than HTF inlet temperature. To improve the energy efficiency
of the heat source, a reasonable HTF inlet temperature range
should be chosen for energy savings considering the additional
energy of the heat source.
As shown in Fig. 13, with a constant inlet flow rate
(0.25 m3/h), the effect of HTF inlet water temperature
(60–75 °C) on the temperature difference between the inlet
and outlet water was investigated. The result of Case 3 and
Cases 6 to 9 reached a stable value after a rapid decrease.
This trend was similar to that shown in Fig. 10.
Average charging power in Cases 3 and 6 to 9 of the LHS
unit is shown in Fig. 14. The influence of the inlet water
temperature on the average charging power was significant
when the inlet flow rate was constant. The average charging
power was 0.645, 0.9326, 1.350 and 1.490 kW when the HTF
inlet temperature was 60, 65, 70, and 75 °C, respectively. It Fig. 14 Average charging power of NePCM at different HTF inlet
indicated that the increase of the average charging power got temperature
878 Du et al. / Building Simulation / Vol. 12, No. 5

less obvious with an increasing HTF temperature. When (4) For the LHS unit using NePCM, increasing HTF inlet
the HTF inlet temperature ranged from 60 to 70 °C, the temperature (from 60 to 75 °C) can shorten the melting
average charging power increase with the relatively higher time by 34.68% (from 8080 to 5278 s) and increase the
gradient, which indicated that 60–70 °C was the optimum average charging power by 2.3 times, while temperature
range of HTF inlet temperature. It is noted this range was being too high was not advisable due to the additional
larger than that of pure PCM (60–65 °C) (Yang et al. 2017). energy consumption of the heat source. Therefore, an
In the unit filled with pure PCM, when the inlet temperature optimum range (60–70 °C) of HTF inlet temperature
increased from 65 to 70 °C, the average charging power should be chosen when designing a similar LTES system.
grew with the smaller gradient. That means the effect of
increasing inlet temperature (65–70 °C) on heat transfer Acknowledgements
enhancement was insignificant. Meanwhile higher inlet
temperature means higher energy consumption and poorer This research is financially supported by the Science and
energy efficiency. Besides, in the unit with the same structure, Technology Ministry of China (No. 2016YFC0700406),
for NePCM, when the inlet temperature increased from the National Natural Science Foundation of China (No.
65 to 70 °C, the average charging power grew with high 51576023), the 111 Project (No. B13041) and the Fundamental
gradient because NePCM has higher thermal conductivity Research Funds for the Central Universities (No.
than pure PCM. Therefore, increasing inlet temperature 106112016CDJCR211221).
(65–70 °C) is energy-efficient and effectively enhanced heat
transfer for NePCM. References

4 Conclusions Alizadeh M, Sadrameli SM (2016). Development of free cooling based

ventilation technology for buildings: Thermal energy storage
A three-dimensional model of an LHS unit using a spiral (TES) unit, performance enhancement techniques and design
coil exchanger was constructed. Based on this model, considerations—A review. Renewable and Sustainable Energy
the effects of dispersion of copper nanoparticles, HTF inlet Reviews, 58: 619–645.
Alomair M, Alomair Y, Tasnim S, Mahmud S, Abdullah H (2018).
temperature and HTF inlet flow rate on the melting process
Analyses of bio-based nano-PCM filled concentric cylindrical
of NePCM (paraffin/Cu) were investigated. The conclusions
energy storage system in vertical orientation. Journal of Energy
arising from this study are summarized as follows:
Storage, 20: 380–394.
(1) The NePCM in the LHS unit using a spiral coil exchanger Archibold AR, Gonzalez-Aguilar J, Rahman MM, Goswami DY,
had a significant reduction of nearly 19.6% in melting Romero M, Stefanakos EK (2014). The melting process of storage
time and alleviated the temperature non-uniformity. materials with relatively high phase change temperatures in
Even if the bottom area occupied less than 5% of the partially filled spherical shells. Applied Energy, 116: 243–252.
entire volume, it took nearly 24.3% of the total melting Bezyan B, Porkhial S, Mehrizi AA (2015). 3-D simulation of heat
time to melt the pure PCM in this part, and this figure transfer rate in geothermal pile-foundation heat exchangers
was 1920 s longer than that spent by NePCM. In addition, with spiral pipe configuration. Applied Thermal Engineering, 87:
the energy storage rate of LHS system increased with 655–668.
the growing volume fraction of nanoparticles. Chen J, Yang D, Jiang J, Ma A, Song D (2014). Research progress of
(2) Comparing to the pure PCM, the higher temperature phase change materials (PCMs) embedded with metal foam (a
conductivity of NePCM can overcome the disadvantage review). Procedia Materials Science, 4: 389–394.
brought by the limited heat exchange area. The optimal Chen C, Zhang H, Gao X, Xu T, Fang Y, Zhang Z (2016). Numerical
temperature range (60–70 °C) of HTF in unit with and experimental investigation on latent thermal energy storage
system with spiral coil tube and paraffin/expanded graphite
NePCM was larger than that in unit with pure PCM
composite PCM. Energy Conversion and Management, 126:
(60–65 °C).
889–897.
(3) For the LHS unit using NePCM, increasing HTF inlet
Das N, Kohno M, Takata Y, Patil DV, Harish S (2017a). Enhanced
flow rate (from 0.25 to 1 m3/h) would shorten the melting
melting behavior of carbon based phase change nanocomposites
time by 9.29% (from 8080 to 7329 s) and increase the in horizontally oriented latent heat thermal energy storage system.
average charging power by 4.37% (from 0.756 to Applied Thermal Engineering, 125: 880–890.
0.789 kW), while the effects of flow rate were negligible. Das N, Takata Y, Kohno M, Harish S (2017b). Effect of carbon nano
The flow rate higher than 0.75 m3/h was not recommended inclusion dimensionality on the melting of phase change
when designing a similar LTES system due to the nanocomposites in vertical shell-tube thermal energy storage unit.
additional energy consumption of the pump. International Journal of Heat and Mass Transfer, 113: 423–431.
Du et al. / Building Simulation / Vol. 12, No. 5 879

Delcroix B, Kummert M, Daoud A, Bouchard J (2015). Influence of Shmueli H, Ziskind G, Letan R (2010). Melting in a vertical cylindrical
experimental conditions on measured thermal properties used to tube: Numerical investigation and comparison with experiments.
model phase change materials. Building Simulation, 8: 637–650. International Journal of Heat and Mass Transfer, 53: 4082–4091.
Ebadi S, Tasnim SH, Aliabadi AA,Mahmud S (2018). Melting of Soares N, Reinhart CF, Hajiah A (2017). Simulation-based analysis of
nano-PCM inside a cylindrical thermal energy storage system: the use of PCM-wallboards to reduce cooling energy demand and
Numerical study with experimental verification. Energy Conversion peak-loads in low-rise residential heavyweight buildings in Kuwait.
and Management, 166: 241–259. Building Simulation, 10: 481–495.
Elbahjaoui R, El Qarnia H (2016). Transient behavior analysis of the Sreethawong T, Shah KW, Zhang S-Y, Ye E, Lim SH, Maheswaran U,
melting of nanoparticle-enhanced phase change material inside a Mao WY, Han M-Y (2014). Optimized production of copper
rectangular latent heat storage unit. Applied Thermal Engineering, nanostructures with high yields for efficient use as thermal
112: 720–738. conductivity-enhancing PCM dopant. Journal of Materials Chemistry
Fan LW, Zhu ZQ, Xiao SL, Liu MJ, Lu H, Zeng Y, Yu ZT,Cen KF A, 2: 3417–3423.
(2016). An experimental and numerical investigation of constrained Tay NHS, Bruno F, Belusko M (2012). Experimental validation of a
melting heat transfer of a phase change material in a circumferentially CFD and an ε-NTU model for a large tube-in-tank PCM system.
finned spherical capsule for thermal energy storage. Applied International Journal of Heat and Mass Transfer, 55: 5931–5940.
Thermal Engineering, 100: 1063–1075. Thomas S, André P (2012). Numerical simulation and performance
Guo J, Huai X (2016). Numerical investigation of helically coiled tube assessment of an absorption solar air-conditioning system coupled
from the viewpoint of field synergy principle. Applied Thermal with an office building. Building Simulation, 5: 243–255.
Engineering, 98: 137–143. Tian Y, Zhao CY (2013). Thermal and exergetic analysis of metal
Korti AIN, Tlemsani FZ (2016). Experimental investigation of latent foam-enhanced cascaded thermal energy storage (MF-CTES).
heat storage in a coil in PCM storage unit. Journal of Energy International Journal of Heat and Mass Transfer, 58: 86–96.
Storage, 5: 177–186. Wang Y, Yang X, Xiong T, Li W, Shah KW (2017). Performance
Krishna J, Kishore PS, Solomon AB (2017). Heat pipe with nano evaluation approach for solar heat storage systems using phase
enhanced-PCM for electronic cooling application. Experimental change material. Energy and Buildings, 155: 115–127.
Thermal and Fluid Science, 81: 84–92. Yang X, Xiong T, Dong JL, Li WX, Wang Y (2017). Investigation of
Li T, Rogovchenko YV (2014). Asymptotic behavior of higher-order the dynamic melting process in a thermal energy storage unit
quasilinear neutral differential equations. Abstract and Applied using a helical coil heat exchanger. Energies, 10: 1129.
Analysis, 2014: 395368.
Seddegh S, Wang X, Henderson AD (2016). A comparative study of
thermal behaviour of a horizontal and vertical shell-and-tube
energy storage using phase change materials. Applied Thermal
Engineering, 93: 348–358.