HEFAT2014

10th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics

14 – 16 July 2014
Orlando, Florida

HEAT TRANSFER WITH PHASE CHANGE IN A SHELL AND TUBE LATENT HEAT
STORAGE UNIT

García-Alonso J. M., Aguilar F. and Montero E.*

*Author for correspondence
Department of Electromechanical Engineering, Escuela Politécnica Superior
University of Burgos,
Burgos, 09006,
Spain,
E-mail: [email protected]

ABSTRACT
Concerning thermal energy storage, latent heat thermal INTRODUCTION
energy storage is particularly attractive technique because it One of the present issues to improve energy efficiency is the
provides a high energy storage density. The development of a need to store excess energy that would otherwise be wasted and
latent heat thermal energy storage system therefore involves also to bridge the gap between energy generation and
first the understanding of heat transfer in the phase change consumption. Latent heat thermal energy storage is particularly
materials (PCM) when they undergo solid-to-liquid phase attractive technique because it provides a high energy storage
transition in the required operating temperature range, and density. When compared to a conventional sensible heat energy
second, the design of the container for holding the PCM and the storage system, latent heat energy storage system requires a
formulation of the phase change problem. smaller weight and volume of material for a given amount of
The paper presents a study on the heat transfer mode of a energy. In addition, latent heat storage has the capacity to store
PCM stored in a horizontal cylindrical shell and tube heat heat of fusion at a constant or near constant temperature which
exchanger, being the PCM placed inside the tubes. For this corresponds to the phase transition temperature of the phase
purpose, an experimental bench has been developed to test the change material (PCM). Reference [1] presents a general
heat transfer process, coupled with a Ground-Source Heat review of thermal energy storage systems, including PCMs,
Pump (GSHP), controlled by a computer program developed to while reference [2] presents a review on the PCM materials
manage the coupled GSHP+PCM system. The aim is to used.
compute the heat transfer of the installation during the loading, Once the PCM has been selected, based primarily on the
storage and downloading energy processes. More precisely, the temperature range of application, the next most important
interest is to check experimentally to what extent some factors to consider are: (i) the geometry of the PCM container
common hypothesis used for heat transfer calculations are valid and (ii) the thermal and geometrical parameters of the container
or not. In particular: (i) Comparison between the radial and the required for a given amount of PCM. These factors have a
axial heat transfer through the wall of the PCM tube; (ii) The direct influence on the heat transfer characteristics in the PCM
temperature distribution of the heat transfer fluid in the central and ultimately affect the melt time and the performance of the
and peripheral locations; (iii) The heat transfer rate between the PCM storage unit.
container and the ambient temperature. Results indicate that PCMs are mostly placed in cylindrical or rectangular
density gradients could occur during energy exchange between containers. A survey of previously published papers dealing
the heat transfer fluid and the PCM when melting or with latent heat thermal energy storage systems shows that the
solidification are taking place, leading to modification of most intensely analyzed unit is the shell and tube system,
central flow with respect to peripheral flow. Also, the accounting for more than 70% [3]. This is probably due to the
temperature gradients recorded in the axial direction during the fact that most engineering systems use cylindrical pipes and
change of phase were very small compared with those of that of also because heat loss from the shell and tube system is
the radial direction, indicating essentially a two-dimensional minimal.
heat transfer mode. Fig. 1 shows the classification of common PCM containers
in terms of the geometry.

924
 Three modes of cylindrical PCM container configurations For each pair, the upper arrow represents the direction of
are distinguished. The first is where the PCM fills the shell and HTF flow during charging and the lower arrow represents
the heat transfer fluid flows through a single tube, designated as discharge direction of the HTF.
the pipe model. In the second model the PCM fills the tube and Gong and Mujumdar [7] investigated the effect of the
the heat transfer fluid (HTF) flows parallel to the tube, said the parallel and counter-current flow modes and showed that
cylinder model. The third cylinder model is the shell and tube parallel flow increases the energy charge/discharge rate by 5%
system. Several authors [4-6] recommend shell and tube more than counter-current flow.
configuration as it performs heat transfer better than the pipe Apart from the container geometry and configuration, some
model. other thermal and geometric parameters are known to affect the
thermal performance of latent heat thermal energy storage
systems. For example, shell and tube containers often place
baffles to support the tubes, and then parallel, counter-current
and angular flows could take place inside the cylinder
container. Then, tests and experimental data should be
performed to test mathematical modelling and simulation of
heat transfer modes.
This paper presents a study on the heat transfer mode of a
PCM stored in a horizontal cylindrical shell and tube heat
exchanger, being the PCM placed inside the tubes. For this
purpose, an experimental bench has been developed to test the
heat transfer process, coupled with a Ground-Source Heat
Pump (GSHP), controlled by a computer program developed to
manage the coupled GSHP+PCM system. The aim is to
compute the heat transfer of the installation during the loading,
storage and downloading energy processes. More precisely, the
interest is to check experimentally to what extent some
common hypothesis used for heat transfer calculations are valid
or not. In particular: (i) Comparison between the radial and the
axial heat transfer through the wall of the PCM tube; (ii) The
temperature distribution of the heat transfer fluid in the central
and peripheral locations; (iii) The heat transfer rate between the
Figure 1. Classification of common PCM containers in
container and the ambient temperature.
terms of the geometry.
In a cylindrical container there exist two possibilities for the NOMENCLATURE
flow direction of the heat transfer fluid during charging and A [m2] Area
discharging of the PCM energy. The two modes are the parallel AD [ºC] Average difference of temperature
e [m] Thickness
flow (either the hot and cold fluids are introduced into the heat E [J] Energy
exchanger from the same end) and the counter-current flow (the k [W/mK] Thermal conductivity
hot and cold fluids are introduced from the opposite ends). Fig. MRD [%] Maximum relative difference
2 illustrates the schematic diagram of the parallel and counter- Q [W] Power (heat)
r [m] Radius
current flow principles. Rth [K/W] Thermal resistance
RAD [%] Relative average difference
t [s] time
T [ºC] Temperature

Special characters
∆ [-] Difference

EXPERIMENTAL APPARATUS
To test the heat transfer mode of a horizontal cylindrical
shell and tube container for PCM energy storage, an
experimental bench has been designed and built, as shown in
Fig. 3.
The PCM used is a hydrated salt with a melting temperature
of 41ºC. It was selected because it is an adequate temperature
Figure 2. The physical model of parallel and counter- for domestic hot water production and radiant floor heating.
current HTF flow in a shell and tube heat exchanger. Table 1 presents the thermo-physical properties of the PCM,
obtained from reference [8].

925
 nozzles. The tank is externally insulated with a blanket of
thickness 50 mm made of a commercial elastomer whose
thermal conductivity is 0.04 W/m·K.
Five baffles are placed inside the container, to stand the
PCM tubes in horizontal position, then the container could be
considered as a shell and tube one. A maximum of 24 PCM
tubes can be allocated inside the container. The holes of the
baffles are made of higher diameter than those of the PCM
tubes, in order to allow the water-flow. Fig. 4 shows the
geometrical distribution of PCM tubes.
Figure 3. Diagram of the experimental bench for testing the
heat transfer rate of PCM placed in a cylindrical shell and tube
container.

The PCM used is a hydrated salt with a melting temperature

of 41ºC. It was selected because it is an adequate temperature
for domestic hot water production and radiant floor heating.
Table 1 presents the thermo-physical properties of the PCM,
obtained from reference [8].
The storage temperature and the size of the container has
been chosen to meet the energy needs for the heating and the
domestic hot water supply of a single family house of about
150 m2. A Ground-Source Heat Pump (GSHP) supply the
energy needed to charge/discharge the PCM energy storage.

Table 1. Thermo-physical properties of the PCM placed

inside the cylindrical container [8].
Phase Change Temperature Latent Heat Capacity Density
(ºC) (kJ/kg) (kg/m3)
41 210 1587
Specific Heat Capacity Thermal Conductivity
(kJ/kg·K) (W/m·K)
Solid Liquid Solid Liquid
1,68 2,59 0.450 0.245

The experimental bench consists of (A) two water tanks,

from Domusa™, model SANIT 150, each one of 150 l of
volume, equipped with temperature control to set the
temperature of the cold-source of the heat pump between 10ªC Figure 4. Configuration of PCM tubes and baffles inside
and 40ºC; (B) the GSHP from Giordano™, model SUNEO N5 the circular section of the container.
(open loop heat pump, which uses the two tanks A as if they
were aquifer energy sources), nominal heating power 5.35 kW, A set of 10 Pt100 temperature probes have been used to
fluid R-407C; (C) one horizontal storage tank of 210 l, filled obtain the temperature distribution inside the shell and tube
with the PCM modules; (D) one fan-coil Saunier Duval™, tank, as shown in Fig. 5. T103 to T107 probes measure the
model 3-020 AF, to dissipate the energy stored in the PCM outside temperature of the tank, T102 measures the inner water
water tank, simulating the energy use in a single family house. temperature, T115 and T116 register the temperature of the
For energy calculation purposes, the experimental stand was internal surface of the PCM tubes and T117 and T118 measure
equipped with thermal energy meters Kundo™, model the temperature of the external surface of the same. Ambient
G20/G21, class B, which includes two temperature probes temperature, and inlet and outlet temperatures of water are also
Pt1000 and a flow meter with a maximum relative error of measured.
±4%. All the measurement devices are computer controlled by Temperature measurement was performed by means of the
means of the Agilent VEE 7.0 software. Pt100 probes and the multi-meter Agilent 34970A. After
The PCM is encapsulated in cylindrical tubes, each of 1000 calibration of the equipment, uncertainty of temperature
mm long and 50 mm external diameter, 2 mm thickness. The measurement has been estimated to be less than 0.05 K.
tubes are made of high density polyethylene, with a thermal
conductivity of 0.2 W/m·K.
The stainless steel cylindrical tank is 1020 mm long and 510
mm internal diameter, with a capacity of 210 l. It is placed in
horizontal position, with 1” diameter of input and output

926
 The PCM liquid phase transfers sensible heat between the
temperature T2 and the liquid-solid boundary temperature Tm,
which is the melting temperature. The solid phase is expected
to be at Tm, and new energy transfer produces melting and the
boundary moves to an inner radio. The heat balance across the
interface, known as Stefan condition [9], means that the latent
heat released due to the interface displacement equals the net
amount of heat delivered to (or from) the interface per unit area
and unit time (flux normal to the moving surface).

Figure 5. Distribution of Pt100 temperature probes in the

shell and tube cylindrical container C. PCMs are placed inside
the tubes (C2) while water flows by the external surface (C3).

Fig. 6 shows a typical energy charging/discharging mode of

the PCM storage tank on a daily basis. The conditions during
the experiment are that GSHP operates 14 h (from 22.00h to Figure 7. Heat transfer during the melting process. T1 =
12.00h, when the electricity cost is cheaper) while the fan-coil external temperature of the polyethylene tube at r1; T2 = internal
operates 24.00h (all day house demand). The hot water as HTF temperature of the polyethylene tube and external temperature
is supplied by the GSHP, and the set point is fixed at 50ºC. of the PCM liquid phase at r2; Tm = boundary melting
During the charging period, the GSHP starts heating the HTF, temperature at the moving boundary r.
and the PCM in solid phase increases its temperature during the
sensible heat transfer stage, before melting, showing a quasi- Once the set-point of 50ºC is reached by the HTF, typical
linear slope in the increasing temperature. When the melting temperature evolution of the temperature in a saw-tooth shape
temperature of PCM is reached (41ºC) at the external surface of around the temperature set point occurs, as the GSHP is on/off
the PCM, it starts changing to the latent heat transfer mode and controlled. This mode is kept repeatedly along the rest of the
melting process takes place. The rate of temperature increase is charging period. Along this period, the PCM ends its melting
smaller, and the slope of the curve then decreases compared to and all the PCM is in the liquid phase. Moreover, superheating
the sensible mode. Along this period, heat from the HTF is of the liquid PCM should take place if the GSHP still continues
transferred to the PCM through the thermal resistance of the working. It should be noticed that, at a fixed time, the radius of
polyethylene tube by heat conduction. The external annulus of the moving boundary of the melting PCM will be different
the PCM is in the liquid phase, while the internal circular along the axial direction of the tube, and also the temperature of
section is still solid, as shown in Fig. 7. the liquid phase in contact with the local temperature of the
HTF.
When the GSHP ends its heat supply, only the fan-coil is in
operation, and the discharging period starts, the PCM being in
the liquid phase. Initially, the PCM temperature decreases as
sensible heat exchanges, and super-cooling of the PCM (under
its nominal solidification temperature) could sometimes appear.
When solidification of the PCM begins, the temperature profile
shows a nearly flat shape, enlarging the period of time of
energy availability at a useful temperature, which constitutes
one of the advantages of the PCM. After the solidification ends,
then the discharge uses the sensible heat exchange of the solid
PCM.

RESULTS AND DISCUSSION

Several tests were conducted to study the heat transfer
Figure 6. Distribution of the storage tank temperature vs. phenomena in the cylindrical container. The temperature
time during the charge and discharge periods of the PCM control of the two water tanks (A), acting as low temperature
container. heat reservoirs, was set to 20ºC. The temperature of the water
supply by the GSHP was set to 50ºC, equipped with an off/on
control regulated by the return line of HTF. The charging cycle

927
of the GSHP was programmed to operate along 14 h, from relative terms. Maximum differences between inner/outer
22.00h to 12.00h. Three discharging modes were studied, using temperatures of PCM tube occur always at the beginning of the
three air flow rates at the fan-coil (D), 160 m3/h, 250 m3/h and charging period, frequently during the first 20 minutes after the
390 m3/h, named Test 1, Test 2 and Test 3 respectively. The fan- GSHP started increasing the HTF temperature. Even in this
coil was switched-on along 24 h, and then the discharging cycle case, the differences are always less than 4.7 %.
duration was 10 h, from 12.00h to 22.00h. Temperature and The second concern is the thermal resistance along the
other measurements were registered every 300 seconds, which axial direction of the tube. At the inner side of the tube the
means 288 records in a 24 h cycle. temperatures are measured by probes T115 and T116, and the
Some energy transfer processes were studied through the outer side temperatures are measured by probes T117 and T118.
three tests carried on. The interest is to check experimentally to Results are shown in Table 3.
what extent some common hypothesis used for heat transfer
calculations are valid or not. In particular: Table 3. Differences of axial temperatures of the PCM
a) Comparison between the radial and the axial heat transfer tubes. Average and maximum values.
through the wall of the PCM tube, made of high density Parameter ∆T115-116 ∆T117-118 Test
polyethylene (HDPE). Relative Average Difference (%) 2.822 2.530 1
𝐹
b) The temperature distribution of the HTF (water) in the 100 ∆𝑇 2.658 2.243 2
𝑅𝑅𝑅 = �� �
central and peripheral locations. 𝑁 𝑇 2.641 2.288 3
𝑖=1
c) The heat transfer rate between the container and the 11.641 9.488 1
Maximum Relative Difference (%)
ambient temperature ∆𝑇 8.996 9.389 2
𝑀𝑀𝑀 = 𝑀𝑀𝑀 �100 � ��
𝑇 10.334 9.013 3
Comparison between the radial and the axial heat transfer Average Difference (ºC) 1.219 1.105 1
through the wall of the PCM tube. 𝐹
100 1.113 0.965 2
As stated previously, PCM is encapsulated in cylindrical 𝐴𝐴 = �|∆𝑇|
𝑁 1.129 0.998 3
tubes, each of 1000 mm long and 50 mm external diameter, 2 𝑖=1

mm thickness. The tubes are made of HDPE, with a thermal

conductivity of 0.2 W/m·K (similar to the liquid PCM and It can be considered that the inner temperature difference
slightly smaller than that of the solid PCM). The thermal ∆T115-116 is representative of the PCM side, while ∆T117-118 is
resistance Rth is defined as closer to the HTF temperature (water side). It can be observed
that average differences of temperature along the axial direction
𝑅𝑡ℎ = 𝑒�𝑘 · 𝐴 (1) are less than 1.22ºC in any test, which means less than 2.83 %
in relative terms. As the area of the front annular surface is very
where e is the thickness of the material, k the thermal small, and the considered length is 1000 mm, the thermal
conductivity and A is the front area of the heat flux. In this case, resistance in this case is 16556 K·W-1. That means that heat
as the external area of the cylinder is 0.16 m2, this leads to a transfer along the axial direction is almost negligible compared
thermal resistance of 0.063 K·W-1, which can be similar to to the radial direction.
some other encapsulating materials. Maximum differences of temperature along the axial
Moreover, the temperature differences measured at both direction (up to 11.65 %) take place during the initial rise of
sides of tube wall are small, as shown in Table 2. At the left temperature of the charging mode.
side of the container, near the HTF outlet, the temperatures are
measured by probes T115 and T117. At the right side of the
The temperature distribution of the HTF (water) in the
container, near the HTF inlet, the temperatures are measured by
central and peripheral locations.
probes T116 and T118. Distribution of HTF temperature along the PCM
cylindrical container is measured in three sections. Related to
Table 2. Differences of internal/external temperatures of the Fig. 5, probe T117 measures the temperature of HTF just onto
PCM tubes. Average and maximum values. the external surface of the outlet side; probe T118 does the
Parameter ∆T115-117 ∆T116-118 Test same at the inlet side; and probe T102 measures the
Relative Average Difference (%) 0.730 0.426 1
𝐹 temperature of HTF in an intermediate point of the cylinder,
100 ∆𝑇 0.650 0.370 2
𝑅𝑅𝑅 = �� � placed in the upper side of the central baffle that supports the
𝑁 𝑇 0.695 0.424 3
𝑖=1 tubes. Probes T117 and T118 will be more influenced by the
Maximum Relative Difference (%) 4.688 3.322 1 tube temperatures, while T102 will be dominated by the HTF
∆𝑇 3.033 2.641 2 bulk temperature. In different sections of the tank, parallel,
𝑀𝑀𝑀 = 𝑀𝑀𝑀 �100 � ��
𝑇 3.280 2.887 3 counter-current and angular flows could take place. Results of
Average Difference (ºC) 0.314 0.190 1 temperature measurement during a full charging/discharging
𝐹
100 0.267 0.162 2 cycle are presented in Fig. 8. When the GSHP is switched-on,
𝐴𝐴 = �|∆𝑇|
𝑁 0.288 0.187 3 the storage tank presents a positive energy balance (more
𝑖=1
energy enters than exits), as well as when the GSHP is
Average differences of temperature through the thickness switched-off the energy balance is negative (more energy exits
of HDPE are less than 0.32ºC in any test, or less than 0.73 % in than enters).

928
 For the discussion purpose of this section, only data for the net amount of energy entering the tank is the difference
Test 2 air flow rate at the fan-coil (D), 250 m3/h, are presented. between them. Considering the local discharging period, only
Similar conclusions could be obtained from the analysis of Test the energy exit due to the fan-coil is operating, then the
1 and 3. absolute energy that leaves the tank is higher than the one of the
It can be observed that the increase of temperature during charging period.
the initial phase of the charging period is almost the same for
the three temperature probes, showing a uniform behavior of
the HTF. It must be recalled that under the melting temperature,
41ºC, the PCM is solid and exchange sensible heat with the
HTF. When the HTF is over 41ºC, the solid PCM is surrounded
by an annulus of liquid PCM and its temperature rises over
41ºC, depending on the exchange of sensible heat with the
HTF.
Once the HTF reaches the set-point temperature of 50ºC,
the temperature behaves in a saw-tooth shape, due to the on/off
control of the GSHP, though with some delay of the
intermediate temperature (T102) with respect to both side
temperatures (T117 and T118). Temperature of the outlet side
(T117) also presents a small delay compared with the inlet side
(T118). Fig. 9 shows a closer description of this phenomenon. Figure 9. Detailed distribution of the HTF temperature vs.
time while keeping the set-point temperature (51ºC) of the
PCM container.

More complex is the behavior of the HTF measured at the

intermediate position, probe T102. As this probe is not placed
onto the tube surface but immersed into the HTF flow, its
measure is the result of the weighted average between the
energy supplied by the HTF and the heat exchange with the
PCM tubes, being always more influenced by the HTF flow.
For the local charging mode, when the energy balance is
positive, the energy exchange is fully dominated by the HTF
energy supply, all temperatures increase and, as the heat
transfer from the HTF to the PCM takes place, the relative
values of temperature follows the sequence
Figure 8. Distribution of the HTF temperature vs. time
𝑇118 > 𝑇102 > 𝑇117
along the charge and discharge periods of the PCM container.
For the local discharging mode, the energy balance of the
The on/off control of the GSHP is regulated by the HTF
storage tank is negative. Temperatures T117 and T118 decrease
return line temperature, which is related to the outlet
rapidly but temperature T102 does it slowly. The measured
temperature T117 of the cylindrical container. As seen in Fig. 9,
sequence, Fig. 9, is now
during the charging period (GSHP on) the inlet temperature
T118 presents a higher value than the outlet temperature T117,
𝑇102 > 𝑇117 > 𝑇118
as corresponds to a positive energy balance of the tank. At the
opposite, while the discharging period (GSHP off), the inlet
The T102 behavior is now conditioned by heat transfer
temperature is lower than outlet temperature, the energy
from the PCM tubes to the HTF, which produces internal
balance of the tank being negative. The sequence is repeated on
gradients of density inside the HTF. The sudden decrease of
and on till the end of the charging period.
inlet temperature means the heat is transferred from the external
It can be observed that the average temperature of the local
surface of the tubes to the HTF, provoking a local increase of
charging/discharging modes increases along this period, for
HTF density. This phenomenon around the whole set of tubes
example from 51.5ºC to 52.5ºC for probe T117, as the melting
lead to a preferential HTF flow through the central part of the
of PCM progresses and ends, and the annular liquid phase
tank, with respect to the low density HTF placed at the
increase its external temperature. The decreasing branch shows
peripheral part, where probe T102 is placed. This experimental
a high slope (only 15 minutes to descent from 53ºC to 49ºC),
behavior is coherent with theoretical assumptions for the
while at the increasing branch the slope is lower (45 minutes to
mathematical modelling of heat transfer in similar cases [9].
rise from 49ºC to 53ºC), due to the different result of the energy
Then, temperature profile of T102 is delayed with respect T117
balance. During the local charging period, both the energy
and T118 till the charging period starts again.
supply (GSHP) and the energy exit (fan-coil) are operating and

929
 The same explanation serves for the discharging period of uniform temperature distribution is supposed. Temperature
the cylindrical tank, when the energy balance is negative. As probes T104, T105 and T106 are used to measure the external
long as the HTF evacuates energy from the tank, temperatures temperature of the insulation wall at the cylindrical surface. An
T117 and T118 decrease initially while the liquid PCM average of these three temperatures is used as the reference
transfers heat to the HTF. When solidification of PCM begins, a temperature of this surface. As surface temperatures are
plate shape of temperature profile occurs, keeping the measured no ambient temperature and external convection is
temperature almost constant while the latent heat is transferred. needed.
This process means a longer period of energy transfer between Concerning the internal convection, only probe T102 is
the storage system and the facility system (fan-coil), which is representative of the bulk temperature of the HTF. As stated
one of the advantages of the energy storage approach. After previously, parallel, counter-current and angular flows could
solidification ends, a new period of decreasing temperature take place in the HTF, then, it is difficult to estimate convection
appears, the PCM in solid phase. Probe T102 shows the same coefficients at different locations of the HTF flow inside the
kind of displacement with respect to the inlet and outlet bulk tank. As turbulent flow is supposed, the HTF bulk temperature
temperatures as previously stated, due to the peripheral and the internal surface temperature of the container should be
placement of the probe. Finally, at the end of the discharging very close, then, we can consider the hypothesis of negligible
period, temperatures become closer because of the thermal resistance of the conductive layer with respect to the
homogenization of the HTF flow. thermal resistance of the solid conduction (insulation).
Comparison of the experimental estimation of QEXT under
The heat transfer rate between the container and the these heat transfer hypothesis with the corresponding value
ambient temperature obtained from the energy equation agree quite well, being the
Heat transfer between the storage tank and the average deviation of 2% for the full cycle charge/discharge.
surroundings affect the energy balance of the system. At any Table 3 shows the results of the evaluation of QEXT and the
time, the energy conservation in the storage tank can be estimation of EHTF for the full cycle of the system, as well as for
expressed as the three partial processes that can de distinguished.
∆𝐸𝐻𝑇𝐹
𝑄𝐺𝐺𝐺𝐺 + 𝑄𝑃𝑃𝑃 + 𝑄𝐸𝐸𝐸 + 𝑄𝐹𝐹𝐹 = (2) Table 3. Estimation of the energy accumulated by the HTF
∆𝑡
inside the storage tank, EHTF, and measured values of the heat
where QGSHP is the power supplied by the GSHP; QPCM is the
transfer between the storage tank and the surroundings QEXT.
power absorbed/supplied by the PCM; QEXT is the heat transfer
EHTF QEXT QEXT/EHTF
between the inside of the storage tank and the surroundings; Heat transfer period
(kJ) (kJ) (%)
QFAN s the power extracted by the fan-coil system; and ∆EHTF is Total Cycle (average) 26639 10.8 0.034
the variation of the energy accumulated by the HTF inside the Partial processes (average)
tank along a period of time ∆t. The aim is to estimate the Charging period (<Tset point) 30117 10.9 0.035
impact of the heat transfer with the surroundings with respect Charging period (=Tset point) 35753 14.8 0.041
the accumulation of energy in every functioning period. The Discharging period 22967 6.6 0.026
energy equation allows the estimation of QEXT if the rest of the
terms of eq. 2 are known. With respect to the total cycle of the PCM insulated tank,
QGSHP and QFAN are determined by its respective energy- results show that the heat transfer to the surroundings is less
meters. QPCM is estimated by the heat transfer between the HTF than 0.034 % compared to the total amount of energy stored by
and the PCM, once transport properties, flow rates and the HTF, which means that QEXT is almost negligible in the
temperatures are determined. And the ∆EHTF is computed by energy equation.
means of temperature T102 and the heat capacity of the HTF If we take into consideration the partial processes taking
mass control in the container. Then QEXT can be obtained. A part during the charge/discharge cycle, it can be observed that
comprehensive computer program, using Engineering Equation (i) the average energy content of the HTF inside the tank
Solver EES software (©F-Chart Software, LLC), has been increases during the charging process before the set point
developed to compute the energy balance in the system. temperature of 50ºC is reached, (ii) energy keeps in the higher
Detailed description of these calculations is outside the scope of value during the saw-tooth period of fixed set point, and (iii)
this paper, and will be provided in future publications, though energy decreases to a lower value during the discharging
some preliminary results were presented in reference [10]. period. The same profile is showed by the heat transfer to the
For the purpose of checking experimentally the heat surroundings, increasing as long as the internal temperature
transfer with the surroundings QEXT, heat transfer across the two (and the energy content) increases, and decreasing when the
circular plates at both sides of the tank, and across the internal temperature do it, being coherent with the heat transfer
cylindrical surface of its length, is evaluated. As the thermal dependence on temperature differences. Nevertheless, the ratio
resistance of the insulation is much greater when compared to QEXT/EHTF is always very small, less than 0.041%.
that of the stainless steel, heat conduction through the stainless
steel wall is not considered. Temperature probes T103 and T107 CONCLUSION
are placed at the respective center of the external surface of the A case study of heat transfer study of a low-temperature
insulated circular plate walls, following Fig. 5. In these surfaces PCM energy storage system has been presented. The study has

930
shown data on the characterization of the heat transfer mode of [8] Plus Ice©, Phase Change Materials (PCM) Thermal Energy Storage
a PCM stored in a horizontal cylindrical shell and tube heat (TES) Design Guide, Phase Change Material Products Ltd, Yaxley,
exchanger, being the PCM placed inside the tubes. Some Cambridgeshire, UK, 2011.
experimental data have been obtained from a bench has been [9] Alexiades, V., Solomon, A. D., Mathematical modeling of melting
and freezing processes, Hemisphere Publ. Co. Washington, 1993.
developed to test the heat transfer process. [10] García-Alonso, J. M., Aguilar, F., Montero, E., Energy simulation
Some energy transfer processes were studied through the and feasibility of a Ground-Source Heat Pump coupled with a Phase
three tests carried on. Concerning the comparison between the Change Material energy storage system for heat supply, Renewable
radial and the axial heat transfer through the wall of the PCM Energy & Power Quality Journal, Vol. 11, 2013, paper 358.
tube, made of HDPE, experimental data lead to the conclusion
that heat transfer along the axial direction is almost negligible
compared to the one in the radial direction.
Another issue of interest is the temperature distribution of
the HTF along the container, which influences the heat transfer
rate to and from the PCM. It has been show that, during periods
of positive/negative energy balance and sensible heat exchange
with the PCM, the HTF temperatures are dominated by the bulk
temperature of HTF, but when melting/solidification processes
take place in the PCM, some blockage of peripheral HTF flow
can occur due to density gradient.
Finally, in relation with the heat transfer rate between the
container and the ambient temperature, experimental data have
shown that the heat transfer to the surroundings is negligible
compared to the total amount of energy stored by the HTF in
the insulated tank, which means that the hypothesis of adiabatic
performance is valid.

AKCNOWLEDGEMENT
We acknowledge support for this research to the firm
ENERFUTURE S.L.L., Spain, project on “Integration of low-
temperature storage and renewable energy systems in air-
conditioning of small scale buildings”, ENERFUTURE-
Universidad de Burgos, 2010.
This contribution is part of the Thesis Doctoral of J. M.
García-Alonso

REFERENCES

[1] Dinçer, I., Rosen, M. A., Thermal Energy Storage Systems and
Applications, Wiley, London, 2011.
[2] Abhat, A. Low temperature latent heat thermal energy storage: heat
storage materials. Solar Energy, Vol. 30, No. 4, 1983, pp. 313–32.
[3] Agyenim, F., Hewitt, N., Eames, P., Smyth, M., A review of
materials, heat transfer and phase change problems formulation for
latent heat thermal energy storage systems (LHTESS), Renewable
and Sustainable Reviews, Vol. 14, 2010, pp. 615-628.
[4] Ghoneim, A. A., Comparison of theoretical models of phase-
change and sensible heat storage for air and water-based solar
heating systems, Solar Energy, Vol. 42, No. 3, 1989, pp. 209–20.
[5] Hendra, R, Hamdani, Mahlia, T. M. I., Masjuki, H. H., Thermal
and melting heat transfer characteristics in a latent heat storage
system using mikro, Applied Thermal Engineering, Vol. 25, 2005,
pp. 1503–15.
[6] Agyenim, F, Eames, P, Smyth, M., Heat transfer enhancement in
medium temperature thermal energy storage system using a
multitube heat transfer array. Renewable Energy, Vol. 35, No. 1,
2010, pp. 198–207.
[7] Gong, Z, Mujumdar A. S., Finite-element analysis of cyclic heat
transfer in a shell and tube latent heat energy storage exchanger,
Applied Thermal Engineering, Vol. 17, No. 4, 1997, pp. 583–91.

931