A numerical study of external building walls containing phase change materials (PCM)

Applied Thermal Engineering 47 (2012) 73e85

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

A numerical study of external building walls containing phase change materials (PCM) M.A. Izquierdo-Barrientos a, *, J.F. Belmonte b, c, D. Rodríguez-Sánchez b, A.E. Molina b, c, J.A. Almendros-Ibáñez b, c a b c

Thermal and Fluid Engineering Department, Universidad Carlos III de Madrid, Avda. de la Universidad 30, 28911 Leganés, Madrid, Spain Renewable Energy Research Institute, Section of Solar and Energy Efficiency, C/de la Investigación s/n, 02071 Albacete, Spain Escuela de Ingenieros Industriales, Dpto. de Mecánica Aplicada e Ingeniería de Proyectos, Castilla La Mancha University, Campus Universitario s/n, 02071 Albacete, Spain

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 July 2011 Accepted 19 February 2012 Available online 20 March 2012

Phase Change Materials (PCMs) have been receiving increased attention, due to their capacity to store large amounts of thermal energy in narrow temperature ranges. This property makes them ideal for passive heat storage in the envelopes of buildings. To study the influence of PCMs in external building walls, a one-dimensional transient heat transfer model has been developed and solved numerically using a finite difference technique. Different external building wall configurations were analyzed for a typical building wall by varying the location of the PCM layer, the orientation of the wall, the ambient conditions and the phase transition temperature of the PCM. The integration of a PCM layer into a building wall diminished the amplitude of the instantaneous heat flux through the wall when the melting temperature of the PCM was properly selected according to the season and wall orientation. Conversely, the results of the work show that there is no significant reduction in the total heat lost during winter regardless of the wall orientation or PCM transition temperature. Higher differences were observed in the heat gained during the summer period, due to the elevated solar radiation fluxes. The high thermal inertia of the wall implies that the inclusion of a PCM layer increases the thermal load during the day while decreasing the thermal load during the night. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Phase change materials Building simulation

1. Introduction The idea of integrating Phase Change Materials (PCMs hereafter) into building walls has been studied since the early 1980s, although their use has been increasing in the last decade, due to the need to reduce energy consumption and the cost of Heating, Ventilating and Air Conditioning (HVAC) systems [1]. Currently, people spend most of their time in enclosed spaces and demand narrow temperature ranges to live comfortably. Consequently, there has been an increase in both our energy demand, as well as the release of polluting agents into the environment [2]. Traditional building materials store energy in sensible forms, and by varying their temperature, these materials contribute to a decrease in the magnitude of internal air temperature swings. The main drawback of these materials is the heavy masonry walls that are needed to stabilize the temperature swings, due to their limited thermal capacity [3]. An alternative solution is the use of PCMs,

* Corresponding author. Tel.: þ34 967 599 200x8217; fax: þ34 967 555 321. E-mail address: [email protected] (M.A. Izquierdo-Barrientos). 1359-4311/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2012.02.038

which store part of the energy in a latent form (constant temperature) by melting or solidifying, provided that a suitable material is selected [4]. In general, a PCM to be used for thermal energy storage should have a high heat of fusion, high thermal conductivity, high specific heat and density, long-term reliability during repeated cycling and low volume change during phase transition, should be noncorrosive, non-toxic and non-flammable and should exhibit little or no supercooling [5]. Two main PCM groups, organic and inorganic, are differentiated. Inorganic PCMs (salt hydrates and metallics) exhibit supercooling and phase segregation during transitional processes. Organic PCMs, such as paraffins, fatty acids and polyethylene glycol, show little supercooling or segregation, are available over a large temperature range and are compatible with conventional construction materials. However, these materials are flammable and have low thermal conductivity [6]. For this study, the material chosen is the granulate PCM Rubitherm GR, which is commercially available at different melting temperatures. The composition of the granule is 65% ceramic and 35% paraffin wax. This PCM offers the advantage of maintaining its macroscopic solid form during a phase change. The PCM is bound

74

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

within a secondary supporting structure that prevents leakage of the PCM when it is in liquid form [7]. This type of PCM was also employed by Huang et al. [8] in their numerical study. One of the purposes of this study is to determine the optimum phase change temperature of the PCM at which the maximum amount of heat is stored or released. This temperature has also been studied by different researchers [3,9e11]. Peippo et al. [3] concluded that the phase change temperature that maximizes the heat absorbed during the diurnal period should be between 1 and 3  C higher than the indoor temperature. Neeper [9] analyzed a PCM wallboard with diurnal variations in the indoor temperature but without direct solar radiation. This researcher observed that the optimum transition temperature of the PCM should be close to the indoor temperature to minimize the thermal load of the building. Heim and Clarke [10] selected a phase change temperature of 22  C, which is 2  C higher than the indoor comfort temperature. Finally, Zhang et al. [11], in their review, concluded that the phase change temperature should be near the indoor temperature. Considering these results, a more general conclusion would be that the optimal value of the melting temperature depends on the average room temperature, which varies from building to building and from season to season. This paper presents a numerical simulation of the transient heat transfer through a typical building exterior wall with a PCM layer for two different periods of time: 6 days in the winter and 6 days in the summer. In both cases, the orientation of the wall, the position of the PCM in the wall and the phase change temperature have been varied to find the optimal parameters that minimize the energy demand of the building due to heat transmission through the walls. In the following sections, the building exterior walls and the heat transfer model used in the simulations are described. Next, the results achieved for the different configurations studied are presented and discussed, and finally, the main conclusions of the work are summarized. 2. Description of the cases studied To evaluate the performance of PCMs in building exterior walls, simulation results are compared with results obtained for a base composite wall without PCMs. This wall represents standard Spanish construction and consists of a first layer of cement that is

Cement

Exterior

I

T

Convection he

Tamb_ext

Perforated brick wall

Table 1 Thermal properties of the masonry wall materials selected, where k is the thermal conductivity, r is the density and c is the specific heat capacity. Constitute material

k [W/(m K)]

r[kg/m3]

c [J/(kg K)]

Concrete grout Perforated brick wall Insulator Brick wall II Plaster

1.3 0.5 0.038 0.4 0.57

1900 900 32 920 1100

1000 1000 840 1000 1000

15 mm thick followed by two layers of brick wall that are 115 and 40 mm thick with a 40 mm layer of insulation between them and, finally, a 15 mm plaster layer on the interior of the building [12]. Fig. 1 shows a schematic of the wall. The thermal properties of the masonry wall materials, used as input data in the simulation, are summarized in Table 1. The transient heat transfer through the base composite wall shown in Fig. 1 is studied for a north, south, east and west orientation over 6 days in the winter and over another 6 days in the summer. January and July are the months selected to represent these extreme seasons. The weather data correspond to the city of Madrid, and the irradiation and external dry air temperature were measured every 10 min [13]. The indoor temperature is fixed at 20  C for winter and 26  C for summer. A second wall (PCM composite wall) is simulated under the same conditions. This wall is identical to the first one, which is shown in Fig. 1, but incorporates a PCM laminated into a single 20 mm-thick layer in the masonry. The position of the PCM influences the thermal behavior of the wall. Due to this property, three different configurations, which are shown in Fig. 2, are proposed:  Configuration 1: the PCM is placed next to the interior face of the insulator, as shown in Fig. 2(a).  Configuration 2: the PCM is placed next to the exterior face of the insulator, as shown in Fig. 2(b).  Configuration 3: the PCM is substituted for the insulator, as shown in Fig. 2(c). Under transient conditions, with the PCM storing and releasing heat, configuration 1 (Fig. 2(a)) seems to be most appropriate for a building with the highest thermal load during the winter. This

Insulation

Brick wall II

Plaster

Interior

Convection hi

Tamb_int

Fig. 1. A schematic diagram showing the wall layers of the typical external wall (base composite wall) used in the simulations.

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

75

Table 2 Thermophysical properties of the PCM GR, obtained from [7].

a Exterior

Interior

P C

Properties

GR

Thermal conductivity k[W/(mK)] Specific heat capacity c[J/(kgK)] Density r[kg/m3] Latent heat of fusion hfg[J/kg]

0.15 1500 750 63,000

3. Heat transfer analysis and numerical resolution

M

3.1. Numerical scheme and resolution

b Exterior

Interior

P C M

To study the transient heat transfer through the different walls, as previously described, for typical winter and summer days, a 1D transient conduction model has been developed. The equations of the model have been solved numerically using an implicit finite difference scheme. The wall is discretised with a uniform spatial step of Dx ¼ 1 mm, resulting in 226 nodes (of which 20 correspond to the PCM layer). The time step used in the simulations is Dt ¼ 10 min, which corresponds to the measurement frequency of the meteorological data. The time step has been tested to assure the convergence and consistency of the numerical scheme. A control volume is defined around each node and is analyzed using energy balances to develop the set of equations that must be solved to obtain the temperature at each node and time step (see Fig. 3). The energy balance for the first node, which is situated on the exterior face of the wall, is




Dx vT

rc
2

c Exterior

Interior

P C M

vt

x¼0


  vT
¼ Ia  h Tx¼0  TN;ext þ k

vt x¼ Dx

where r is the density of the wall material and c is the specific heat capacity. The first term on the right side of equation [1] corresponds to the fraction of energy incident on the wall surface that is absorbed (a value of a ¼ 0.6 is assumed [14]). The second term represents the heat transfer by convection and radiation from the wall surface to the environment. On a typical summer day, the temperature of the exterior environment is higher than the temperature at the external surface; therefore, this term is positive. In contrast, on a typical winter day (or during a summer night), this term could be negative. The coefficient h is the sum of the convective heat transfer coefficient and the linearized radiation coefficient [2].

h ¼ hc þ hrad Fig. 2. (a) Configuration 1: PCM to the right of the insulator. (b) Configuration 2: PCM to the left of the insulator. (c) Configuration 3: PCM substituted for the insulator.

configuration permits the PCM to retain more energy in the interior of the building when the latent energy stored in the PCM is released because of the high resistance to heat transfer caused by the insulation. Following similar reasoning, configuration 2 (Fig. 2(b)) may be more appropriate for a building with the highest thermal load during the summer. Nevertheless, the selected configuration must be the same for each building, regardless of the period of the year. For this reason, the performance of both configurations is analyzed for both the summer and winter. The type of PCM selected is the GR from Rubitherm, which is available over a wide temperature range. This material fills the cavity in the same manner as in the study by Huang et al. [8]. The thermal properties of the PCM are listed in Table 2 [7]. The same property values for the solid and liquid phase have been used.

(1)

2

(2)

Defraeye et al. [15] studied the convective coefficient for the exterior of a building surface for different correlations and for CFD (Computational Fluid Dynamics) modeling. These researchers concluded that there are significant differences between correlations related to the specific conditions under which each correlation has been derived, and these differences limit, to some extent, the correlations’ applicabilities for other building configurations. Nevertheless, for this study, the following expression [14] has been used

 h  i 6 þ 4V hc W= m2 K ¼ 7:4V 0:78

if V  5 m=s if V>5 m=s

(3)

where V is the air velocity in m/s. The linearized radiation term is calculated using the equation below

   2 2 Tx¼0 þ TN;ext hrad ¼ 3 s Tx¼0 þ TN;ext

(4)

76

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

Fig. 3. Example of three control volumes and their energy balance. The red arrows indicate the positive direction of the heat flow. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

where 3 is the emissivity with a value of 0.9 [14], and s is the StefaneBoltzmann constant. The third term of Eq. (1) takes into account the heat transferred by conduction through the masonry wall. This term can also be positive or negative depending on the sign of the temperature gradient through the right surface of the control volume. Similarly, the energy balance in a control volume associated with an interior node reduces to





vT Dxrc

vt

x¼Dxði1Þ



vT
vT
¼ k

þ k

vt x¼Dxði1Þ Dx vt x¼Dxði1Þþ Dx 2

(5)





vt

x¼Dxði1Þ



vT
vT
¼ k

þ k

vx x¼Dxði1Þ Dx vx x¼Dxði1Þþ Dx 2

(6)

2

where hfg is the change in enthalpy associated with the phase change, and f is the liquid fraction in the PCM. The value of f is zero if the material is in a solid state, and one if the material is in a liquid state. For every time step in which there is a change in temperature from a value lower than the phase change temperature to a value higher than the phase change temperature (or vice versa), the simulation detects that a phase change is occurring and the f value is actualized using the following expression

"

jþ1 fi

  jþ1 c Ti  Tpcm

2

vt

x¼L


  vT
¼ hint Tx¼L  TN;int  k

vx x¼L Dx

(8)

2

Many discrepancies in the literature can be found regarding the value of the internal heat convection coefficient hint. Hongim et al. [16] concluded that the convective heat transfer coefficient calculated for the PCM wall surface using the equations for an ordinary wall usually underestimates the true value. Nevertheless, for this study, the equation presented by Awbi and Hatton [17], which is based on experimental data, has been used to calculate hint.

2

where the two terms on the right side of Eq. (5) represent the heat conduction through the left and right faces of the control volume of the node, respectively. Eq. (5) is valid when there is no phase change in the material, i.e., the stored energy is in a sensible form and undergoes a temperature change. In contrast, when a phase change is present, the left side of equation (5) should be modified to take into account the storage of energy in a latent form at a constant temperature. Therefore, the energy balance equation for an interior node i when there is a phase change can be expressed by the following

vf Dxrhfg






Dx vT

rc


#

8 > j > > ;1 > < max fi þ hfg   ¼ " # jþ1 > > c Ti  Tpcm > > min f j þ : ; 0 i hfg

jþ1

if Ti

>Tpcm (7)

if

Tijþ1

< Tpcm

In Eq. (7), the superscript indicates the time instant, the subscript indicates the node position, and Tpcm is the constant temperature at which the phase transition occurs. Finally, the energy balance in the control volume associated with the node closest to the room is

hint ¼

 1:823  Tx¼L  TN;int 0:121 D

(9)

where D is the hydraulic diameter of the surface, which has been calculated assuming a 3 m square wall. In order to numerically solve the set of equations for all nodes, the spatial derivatives in Equations (1)e(6) are discretised in time j þ 1 using first-order differences, and the temporal derivatives are approximated using a forward difference scheme. To obtain the N temperatures at time j þ 1 (with N number of nodes), it is necessary to solve a set of N equations because the method is fully implicit. The initial condition needed to solve the system of equations is a constant temperature for the whole wall that is equal to the interior temperature (20  C for winter and 26  C for summer). To avoid the influence of the initial condition on the results, the calculations were initialized eight days before, which has been found to provide sufficient time. With more than five days of previous calculations, no differences were observed in the final results.

3.2. Numerical outputs The main numerical outputs of the simulation are the temperature evolution with time for each node (Ti (t)) and the liquid fraction evolution (fi (t)) in the PCM. From these data, other numerical outputs have been computed and are presented in the following section. These parameters are  The instantaneous heat flux lost/gained during the winter/ summer period Q_ ½W=m2 , which is defined as the heat flux through the interior face of the wall for each instant of time

Q_ ðtÞ ¼



  hint ðtÞ Tx¼L ðtÞ  TN;int   hint ðtÞ TN;int  Tx¼L ðtÞ

for summer period for winter period

(10)

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

 The total heat lost/gained during the winter/summer period Q [J/m2], which is defined as the instantaneous heat flux integrated over the time period tZ¼ tf

Q ¼

Q_ dt

77

Where tf ¼ 6 days is the total time computed. The total heat can be separated into two components: the heat transferred between the sunrise and sunset of each day (sunlight time period) Qs and the heat transferred between the sunset and sunrise of each day (nosunlight time period) Qns, i.e.,

(11) Q ¼ Qs þ Qns

t¼0

Dry Air Temperature (ºC)

a

(12)

15 No Sunlight time periods Sunlight time periods

10

5

0

−5

0

1

2

3

4

5

6

Day

100

800

2

2

Radiaton (W/m )

c 1000

Radiaton (W/m )

b 120

80 60 40

600 400 200

20 0 0

2

4

0 0

6

2

Day

d

e

600

6

4

6

600 500

2

2

Radiaton (W/m )

500

Radiaton (W/m )

4 Day

400 300 200

400 300 200 100

100 0

0

2

4 Day

6

0

0

2 Day

Fig. 4. (a) External dry air temperature and (b)e(e) radiation on the wall for different orientations for the 6 days simulated in January.

78

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

 the averaged heat flux lost/gained during the winter/summer period Q_ ½W=m2 , which can be calculated by dividing the total heat by the total computed time

Q Q_ ¼ tf

where ts and tns are the portions of the total time computed during the sunlight and no-sunlight time periods, respectively. These time periods are different for each season of the year. For summer, there are ts ¼ 84 h (14 h/day) and tns ¼ 60 h (10 h/day), and for winter, there are ts ¼ 55 h (9.17 h/day) and tns ¼ 89 h (14.83 h/day).

(13)

 In the same manner, we can define the average heat flux for the sunlight and no-sunlight time periods, as follows

Qs _ Qns ; Q ns ¼ ¼ Q_ s ¼ ¼ ts tns

a

4. Results One objective of the simulation is to compare, under the same external solar radiation and temperature conditions, the wall temperature profiles obtained with and without a PCM layer in the masonry. Additionally, the influence of the phase change

(14)

b

7.5 NORTH SOUTH EAST WEST

7

c

8 NORTH SOUTH EAST WEST

7.5

7.5 NORTH SOUTH EAST WEST

7

6.5 7 6.5

6 6.5

5.5

6

6

5 5.5

5.5

4.5 5

5

4.5

4 0

4.5

5

10

15

7.5 NORTH SOUTH EAST WEST

7

3.5

4 0

20

5

10

15

3 0

20

e

8 NORTH SOUTH EAST WEST

5

10

15

20

Phase Change Temperature (ºC)

Phase Change Temperature (ºC)

Phase Change Temperature (ºC)

d

4

f

7.5 NORTH SOUTH EAST WEST

7

7.5

6.5 7

6.5

6

6

6.5

5.5

6

5

5.5

5.5

5

4.5

4 5

4.5

3.5

4 0

5

10

15

20

Phase Change Temperature (ºC)

4.5

0

5

10

15

20

Phase Change Temperature (ºC)

3

0

5

10

15

20

Phase Change Temperature (ºC)

Fig. 5. Heat loss over 6 days in winter for a wall facing different directions as a function of the phase change temperature for configuration 1 (a)e(c) and configuration 2 (d)e(f). The filled, isolated points represent the data for the base composite wall, and the empty, isolated points represent the heat loss of the PSM composite wall.

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

temperature on the amount of heat exchanged with the room is studied. As explained in the previous section, this heat is differentiated between two time periods. The first time period is the time between sunrise and sunset (sunlight time period), and the second time period is defined as the time between sunset and sunrise (nosunlight time period). In this way, the effects of the PCM are fully revealed, rather than the global effect in which certain results would go unnoticed because of the balance between day and night cycles (energy released/stored). In the numerical calculations, the external convection heat transfer coefficient hc has been evaluated from Eq. (3) with the assumption of natural convection, i.e., V ¼ 0 and hc ¼ 6 W/(m2K). This coefficient was varied up to values of hc ¼ 30 W/(m2K) (which corresponds to air velocities of V ¼ 6 m/s), but no noticeable differences were observed in the results. The total amount of heat transferred decreased slightly, but the optimum temperature remained constant. Therefore, only the results for natural convection conditions are presented in this article. 4.1. Results of the winter period for configurations 1 and 2 Fig. 4(a) shows the external dry air temperature for 6 days in January. In Fig. 4(b)e(e), the radiation is represented for the four different wall orientations. It is observed that the first, second and sixth days are cloudy, as indicated by the lower radiation and ambient temperature in comparison to the rest of the days. Fig. 5(a)e(c) show the results obtained for configuration 1, with the PCM layer next to the interior face of the insulator (see Fig. 2(a)). Fig. 5(a) shows the heat loss through the wall over the 6

a

days, and Fig. 5(b) and (c) show the heat loss during the sunlight and no-sunlight periods, respectively. In these figures, the averaged heat flux ðQ_ ½W=m2 Þ is shown, rather than the total heat lost [Q[J/m2]], due to the different durations of the two time periods. The heat flux only provides information regarding the wall configuration, and the heat flux is not influenced by the total number of hours of the time period analyzed, which differ depending on the time period and the season. The points joined by a line correspond to the heat lost through the wall for different phase change temperatures. The filled and isolated points correspond to the heat loss for the base composite wall (scheme of Fig. 1) for each orientation. However, the results of the base composite wall and the PCM composite wall are not directly comparable because the PCM layer adds additional conduction resistance, and the two configurations have different global heat transfer coefficients. Illustrating the influence of the phase change in the PCM layer only, the isolated and empty symbols represent the averaged heat flux through a wall with an additional layer that has the same properties as the PCM layer in a solid state. This wall is known as the PSM (Phase Stabilised Material) composite wall. Therefore, the differences in the heat between the two isolated points (filled and empty points) are attributed to the additional conduction resistance that is included in the wall with the PCM composite layer but without the phase change process. The results of the heat loss (Fig. 5(a)) clearly show that the heat losses are at a maximum when the wall is facing north and at a minimum when it is facing south; this phenomenon is due to the influence of solar radiation. It can be seen that there are phase changes for 16  C in the west and east, 15  C in the north and 18  C

b

9

79

10

8.5 8 8 7.5

6

7 4 6.5 6

2

5.5 0 5 4.5

0

1

2

3

4

5

−2

6

0

1

2

Day

c

3

4

5

6

Day

d

9

9 8

8

7

7 6

6 5

5 4

4

3 0

3 2

1

2

3

Day

4

5

6

0

1

2

3

4

5

6

Day

 Fig. 6. Comparison between the instantaneous heat flux Q_ loss during winter for a PCM composite wall (configuration 1 and Tpcm ¼ 16 C) and a PSM composite wall over 6 days in January for the different orientations.

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

in the south. However, no reduction in the heat loss is obtained compared to the base composite wall. Fig. 5(c) shows how the phase change process during the nosunlight time period increases the thermal losses as compared to the base composite wall. In contrast, during the sunlight time period (Fig. 5(b)), there is a beneficial reduction in heat loss for all of the orientations (most pronounced for the south orientation). Fig. 5(d)e(f) show the same results for wall configuration 2, with the PCM layer being next to the exterior face of the insulator (Fig. 2(a)). No noticeable differences are observed in the maximum and minimum averaged heat fluxes compared to the results obtained for configuration 1, although the temperatures at which the PCM stores and releases heat are different. For configuration 2, this temperature is between 5  C and 10  C, depending on the time period and orientation and is lower than the temperatures observed for configuration 1. The results shown in Fig. 5 may lead to the conclusion that integrating a PCM layer into a standard wall would not be worthwhile, but it does help to diminish the amplitude of the instantaneous heat flux Q_ released or absorbed (depending on the time of day). Fig. 6 shows the variation of Q_ for the PCM composite wall  with configuration 1 at Tpcm ¼ 16 C (solid line) and for the PSM composite wall (dashed line). Regardless of the orientation, the amplitude and the maximum of the instantaneous heat flux are reduced. Fig. 7(a) shows the temperature profile of the PCM composite wall with configuration 1 from the exterior to the interior for different solar hours during a typical winter day. The simulated wall faces the west and has a phase change temperature of 16  C. It can be seen that the profile temperature at the PCM layer (between x ¼ 0.17 and x ¼ 0.19 m) is flat most of the time. The PCM remains at the phase change temperature, which diminishes the amplitude of the temperature swing. Fig. 7(b) shows the temperature profiles obtained for the base composite wall (without the additional PCM layer) under the same external conditions. Fig. 8 shows the liquid fraction over the 6 days along the thickness of the PCM layer for a west-facing PCM composite wall (configuration 1) with a transition temperature of 16  C. The black color indicates that the material is solid, and the gray color shows that part of the material has changed into liquid. It can be seen that the nodes closer to the interior of the building are in the liquid state for the whole period, and the nodes closer to the exterior remain in solid form.

Temperature along the wall (ºC)

a

With PCM

25 20 15 10

0 0

8:00h 12:00h 16:00h 18:00h 20:00h

0.05

0.1

0.15

Wall position (m)

4.2. Results of the summer period for configurations 1 and 2 The external dry air temperature and radiation at the vertical wall for the four different orientations for 6 days in July are presented in Fig. 9. The results are illustrated in the graphs shown in Fig. 10, which are the equivalents to the graphs in Fig. 5 for summer. Fig. 10(a)e(c) show the results obtained with configuration 1 (the PCM layer located before the insulator), whereas Fig. 10(d)e(f) show the data obtained with configuration 2 (the PCM layer located after the insulator). During this period of the year the aim is to gain the least possible heat. The points joined by a line correspond to the heat transferred to the interior of the building for different phase change temperatures. The isolated symbols correspond to the averaged heat flux gained with the PSM composite wall, and the filled and isolated ones correspond to the heat gain with the base composite wall for each orientation and period. In this case, taking a look on Fig. 10(a) and (d), there is not an optimum point in the range of phase change temperatures tested. For each orientation the total energy saved, with the inclusion of

b

30

5

Fig. 8. Liquid fraction along the thickness of the PCM layer for a west-facing PCM composite wall (configuration 1) with a transition temperature of 16  C for 6 days in January. 0 cm corresponds to the side of the PCM closest to the insulator.

0.2

Temperature along the wall (ºC)

80

Without PCM 30 25 20 15 10 8:00h 12:00h 16:00h 18:00h 20:00h

5 0 0

0.05

0.1

0.15

0.2

Wall position (m)

Fig. 7. (a) Temperature profile for configuration 1 of the PCM composite wall facing west for a day in January. (b) Temperature profiles for the same day for the base composite wall.

Dry Air Temperature (ºC)

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

81

40 No Sunlight time periods Sunlight time periods

35 30 25 20 15 10 0

1

2

3

4

5

6

Day 500 400

150

2

Radiaton (W/m )

2

Radiaton (W/m )

200

100

50

0 0

2

4

300 200 100 0 0

6

2

800

600

600

2

Radiaton (W/m )

2

Radiaton (W/m )

800

400

200

0 0

2

4

6

4

6

Day

Day

4

6

Day

400

200

0 0

2

Day

Fig. 9. (a) External dry air temperature and (b)e(e) radiation on the wall for different orientations for 6 simulated days in July.

PCM in the wall, does not vary appreciably with the melting temperature. However, the wall performance is different if we focus on sunlight and no-sunlight time ranges. Fig. 10(e) shows how through all the orientations the heat transferred to the interior of the building is much higher than for the base composite wall when the melting temperature is between 30  C and 33  C, depending on the orientation. For the configuration 1 of the PCM composite wall this temperatures range is between 28  C and 29  C. Regardless of the wall configuration, this increase in the heat transferred to the interior is not desired for a summer day. On the other hand, during the no-sunlight time ranges the situation is the opposite and the averaged heat flux

gained through the wall presents much lower levels around the same range of temperatures. For higher and lower temperatures there is no phase change in the PCM material and this only introduces a simple conduction resistance through the wall. As stated previously, these results may not show any benefit of integrating PCMs into building walls, but the PCM does help to diminish the amplitude and maximum of the instantaneous heat flux Q_ . In Fig. 11, the instantaneous heat flux for configuration 2 is plotted for both a PCM composite wall that changes its phase at  Tpcm ¼ 32 C and the PSM composite wall. In Fig. 12, the temperature profiles of the PCM composite wall and the base composite wall are represented for different solar

82

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

a

b

4.5

4 NORTH SOUTH EAST WEST

NORTH SOUTH EAST WEST

8 NORTH SOUTH EAST WEST

3.5

7

4

3.5

3

6

3

2.5

5

2.5

2

4

2

1.5

3

1.5

20

25

30

35

1

40

e

4.5

20

25

30

35

2

40

4 NORTH SOUTH EAST WEST

NORTH SOUTH EAST WEST

f

7

3.5

3

6

3

2.5

5

2.5

2

4

2

1.5

3

20

25

30

35

40

1

20

25

30

35

40

Phase Change Temperature (ºC)

30

35

40

NORTH SOUTH EAST WEST

3.5

Phase Change Temperature (ºC)

25

8

4

1.5

20

Phase Change Temperature (ºC)

Phase Change Temperature (ºC)

Phase Change Temperature (ºC)

d

c

2

20

25

30

35

40

Phase Change Temperature (ºC)

Fig. 10. Heat gained over 6 days of summer for a wall facing different orientations as a function of the phase change temperature for configuration 1 (a)e(c) and configuration 2 (d)e(f). The filled, isolated points represent the data for the base composite wall, and the empty, isolated points represent the heat gained by the PSM composite wall.

hours. The wall is facing west, and the phase change temperature in the PCM composite wall is fixed at 32  C. The temperature profiles shown in Fig. 12 demonstrate how the inclusion of a PCM layer reduces around 2  C the maximum temperatures reached in the wall. When the temperature profile goes flat between x ¼ 0.13 and x ¼ 0.15 m in Fig. 12(a) it means that the PCM material remains at the phase change temperature storing/releasing heat. In Fig. 13 the phase changes in the PCM layer, for the same case represented in Fig. 12(a), are displayed. It is appreciated that the cycle solid (black) and liquid (white) repeats everyday. 4.3. Results for the winter and summer periods for configuration 3 In the last configuration, the insulator is replaced by the PCM layer. The same procedure as in the other cases has been followed,

but based on Fig. 14, we can conclude that the results are not promising. There is no benefit from eliminating the insulator and replacing it with a PCM layer. For the PCM used in this work, there is no benefit in the summer or the winter. A PCM with a higher capacity of latent heat storage is needed. 5. Discussion Several authors have concluded that the phase change temperature should be close to the indoor comfort temperature [11]. Peippo et al. [3] analyzed a PCM wall in a passive solar house and indicated that the optimal diurnal heat storage occurs for a melting temperature 1e3  C above the average room temperature. Heim and Clark [10] performed a numerical analysis on a PCM-gypsum composite for the heating season and showed that the optimal PCM solidification temperature was 22  C, which was

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

a

b

5

83

7 6

4

5 3

4 2

3 1

2 0

1

−1

0

1

2

3

4

5

0

6

0

1

2

c

3

4

5

6

4

5

6

Day

Day

d 10

8

9

7

8

6 7

5

6

4

5 4

3

3

2 2

1 0

1

0

1

2

3

4

5

0

6

0

1

2

Day

3

Day

Fig. 11. Comparison between the instantaneous heat flux Q_ gained during the summer period for a PCM composite wall (configuration 2 and Tpcm ¼ 32 C) and a PSM composite wall for 6 days in July for different orientations. 

2  C higher than the heating set point for the room. Similar conclusions can be found in other references [9,18e20]. However, these studies were performed under different conditions, for example, using PCM-impregnated gypsum wallboards instead of a PCM layer or modeling PCMs like fatty acids. Obviously, the climate in which the building is located is also a determining

With PCM

b

55 50 45 40 35 30 25

8:00h 12:00h 16:00h 18:00h 20:00h

20 15 0

0.05

0.1

0.15

Wall position (m)

0.2

Temperature along the wall (ºC)

Temperature along the wall (ºC)

a

factor of the PCM characteristics. The same requirement applies to the orientation of the wall or the assumed values for some parameters in numerical simulations. The results obtained in this study do not show any clear optimum phase change temperature (a temperature that minimises the heat gained through the wall in summer and lost in

Without PCM 55 8:00h 12:00h 16:00h 18:00h 20:00h

50 45 40 35 30 25 20 15 0

0.05

0.1

0.15

0.2

Wall position (m)

Fig. 12. (a) Temperature profile for configuration 2 of the PCM composite wall facing west during one day of July, (b) Temperature profiles during the same day for the base composite wall.

Width of the PCM layer(cm)

84

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

20

0.9

18

0.8

16

0.7

14

0.6

12

0.5

10

0.4

8

0.3

6

0.2

4 0.1 2 1

2

3

4

5

6

0

Day Fig. 13. Liquid fraction along the thickness of the PCM layer for a west-facing PCM composite wall (configuration 2) with a transition temperature of 32  C over 6 days in July. 20 cm correspond to the side of the PCM closest to the insulator.

winter) that is valid for all periods of the year. To select the phase change temperature of a PCM, one must know beforehand the orientation of the external building walls, the location of the PCM layer in the wall and the period of the year during which the PCM is

a

expected to reduce the thermal load of the building. Depending on these factors, the selected phase change temperature can vary between approximately 5  C and 35  C for the wall configurations and the PCM studied in this work. In addition, the results could be quite different if only day or night periods are considered, due to the influence of solar radiation. During the winter and sunlight time periods, the results achieved for a south-facing wall show that the inclusion of a PCM layer could increase the thermal load of the building. In this situation, the PCM could prevent the transfer of energy (coming from the sun’s radiation) absorbed by the external surface of the wall to the interior of the building. For other orientations in which the radiation intensity is lower, this effect is less important but is still present. On typical summer days, the wall’s behavior is different due to the importance of solar radiation during this period of the year. In general, and for nearly any phase change temperature and orientation, the PCM increases the thermal load during the sunlight time periods and reduces the thermal load during the no-sunlight time periods. This means that the heat stored during the night is released during the day. This surprising result is attributed to the high thermal inertia of the wall studied because the base composite wall also presented a higher thermal load during the day for the same period of the year. Therefore, selection of the melting temperature of the PCM and its location in the wall strongly depends on the season during which the thermal load of the building needs to be reduced, the wall orientation and the activity carried out in the building. Focusing on the activity of the building, the average heat flux for no-sunlight periods ðQ_ ns Þ is more or less significant, as it is strongly

b

16

16

14

14 12

12 10

8

10

6

8 4

6 NORTH SOUTH EAST WEST

4

10

15

20

Phase Change Temperature (ºC)

2

0

NORTH SOUTH EAST WEST

20

25

30

35

40

Phase Change Temperature (ºC)

Fig. 14. (a) Average heat loss for 6 days in January. (b) Average heat gain for 6 days in July. Both plots are for the configuration in which the insulating layer is replaced by the PCM.

M.A. Izquierdo-Barrientos et al. / Applied Thermal Engineering 47 (2012) 73e85

related to the number of hours the HVAC works. For instance, in a hospital (24 h on service) the HVAC system works the whole day. As a result, sun and no-sunlight periods have to be taken into account, while in other buildings is not necessary to include the nosunlight period on the conditioning schedule. Another example is the office buildings, where the main activities are carried out during the daytime, the inclusion of a PCM with an appropriate melting temperature would help to maintain a comfortable internal temperature, thereby reducing the building thermal loss during the winter season.

6. Conclusions A comparative simulation of a building wall with and without PCMs has been conducted by subjecting the wall to different conditions (orientation, position of the PCM layer, phase change temperature and weather conditions). The results of this work show that there is not a clear optimum temperature that minimises the thermal load through the wall. Under the conditions used in this work, this temperature is between 5  C and 35  C, depending on the season, the wall orientation and the location of the PCM layer. When Tpcm is properly selected, the PCM helps to diminish the maximum and amplitude of the instantaneous heat flux Q_ . As a consequence, the power needed for the HVAC system to overcome the thermal load is reduced. In contrast, the total heat Q lost/ gained during the winter/summer is not always reduced. For the winter, the total heat lost during the sunlight (Qs) and no-sunlight (Qns) time periods is both reduced and increased, respectively. The opposite result is observed for the summer. These surprising results are due to the high thermal inertia of the standard wall selected.

Acknowledgements This work was partially founded by the Spanish Government (Project ENE2010-15403), the regional Government of Castilla-La Mancha (Project PPIC10-0055-4054) and Castilla-La Mancha University (Project GE20101662).

85

References [1] L.V. Shilei, Guohui Feng, Neng Zhu, Li Dongyan, Experimental study and evaluation of latent heat storage in phase change materials wallboards, Energy and Buildings 39 (2007) 1088e1091. [2] F. Kuznik, J. Virgone, J. Noel, Optimization of a phase change material wallboard for building use, Applied Thermal Engineering 28 (2008) 1291e1298. [3] K. Peippo, P. Kauranen, P.D. Lund, A multicomponent PCM wall optimized for passive solar heating, Energy and Buildings 17 (1991) 259e270. [4] L.F. Cabeza, H. Mehling, Heat and cold storage with PCM, An Up to Date Introduction Into Basics and Applications (2010) ISBN: 978-3-540-68556-2. [5] A.A. Ghoneim, S.A. Klein, J.A. Duffie, Analysis of collector storage building walls using phase-change materials, Solar Energy 47 (3) (1991) 237e242. [6] Zhang Yinping, Zhou Guobing, Lin Kunping, Qunli Zhang, Hongfa Di, Applicaton of latent heat thermal energy storage in buildings: state of the art and outlook, Building and Environment 42 (6) (2007) 2197e2209. [7] M. Rady, Granular phase change materials for thermal energy storage: experiments and numerical simulations, Applied Thermal Engineering 29 (2009) 3149e3159. [8] M.J. Huang, P.C. Eames, N.J. Hewitt, The application of a validated numerical model to predict the energy conservation potential of using phase change materials in the fabric of a building, Solar Energy Materials and Solar Cells 90 (2006) 1951e1960. [9] D.A. Neeper, Thermal dynamics of wallboard with latent heat storage, Solar Energy 68 (5) (2000) 393e403. [10] D. Heim, J.A. Clarke, Numerical modelling and thermal simulation of PCMgypsum composites with ESP-r, Energy and Buildings 36 (8) (2004) 795e805. [11] Y. Zhang, G. Zhou, K. Lin, Q. Zhang, H. Di, Application of latent heat thermal energy storage in buildings: state-of-the-art and outlook, Building and Environment 42 (2007) 2197e2209. [12] Documentos técnicos de instalaciones en la edificación (DTIE 7.03) Atecyr (2008) ISBN 978-84-95010-26-1. [13] Meteonorm. Global Meteorological Database for Engineers, Planners and Education. [14] O. Larsson, S. Thelandersson, Estimating extreme values of thermal gradients in concrete structures, Materials and Structures 44 (2011) 1491e1500. [15] T. Defraeye, B. Blocken, J. Carmeliet, Convective heat transfer coefficients for exterior building surfaces: Existing correlations and CFD modelling, Energy Conversion and Management 52 (2011) 512e522. [16] H. Liu, H.B. Awbi, Performance of phase change material boards under natural convection, Building and Environment 44 (2009) 1788e1793. [17] H.B. Awbi, A. Hatton, Natural convection from heated room surfaces, Energy and Buildings 30 (1999) 233e244. [18] M. Marinovic-Cincovic, S. Gadzuric, I.J. Zsigrai, New materials for solar thermal storage solid-liquid transitions in fatty acid esters, Solar Energy Materials Ans Solar Cells 79 (3) (2003) 285e292. [19] U. Stritih, P. Novak, Solar heat storage wall for building ventilation, Renewable Energy 8 (1e4) (1996) 268e271. [20] U. Onishi, H. Soeda, M. Mizuno, Numerical study on a low energy architecture based upon distributed heat storage system, Renewable Energy 22 (1) (2001) 61e66.