Energy and Buildings 39 (2007) 364373 www.elsevier.

com/locate/enbuild

The most efcient position of shading devices in a double-skin facade

Elisabeth Gratia *, Andre De Herde
Universite Catholique de Louvain, Architecture et Climat, Place du Levant 1, B-1348 Louvain-La-Neuve, Belgium Received 16 August 2006; received in revised form 26 August 2006; accepted 1 September 2006

Abstract In recent years, there has been a great deal of interest in double-skin facades due to the advantages claimed for this technology in terms of energy saving in the cold season, protection from external noise and wind loads and their high-tech image. The advent of computers and other ofce equipment has increased the internal heat gains in most ofces. Highly glazed facades, together with the extra heat gains from the electric lighting made necessary by deep oor plans and the wider use of false ceilings, have increased the risk of overheating. To preserve comfort and reduce cooling loads, it is important to apply natural cooling strategies, including solar protections use. External solar protections are more effective than internal shading devices. In the case of the double-skin facade, the blinds can be integrated in the cavity. It is thus protected from the bad weather and pollution. Solar protection can remain in place even in the event of important wind, which represents an undeniable advantage for the buildings with great height. This article examines the inuence of the position and the colour of the blinds on the cooling consumption of an ofce building with a doubleskin facade. # 2006 Elsevier B.V. All rights reserved.
Keywords: Double-skin facade; Thermal modelling; Ofce building; Shading devices; Cooling strategies

1. Introduction The double-skin facade is an architectural phenomenon driven by an aesthetic desire for an all-glass facade. Transparency is often seen as the main architectural reason for a double-skin facade, because it allows close contact with the surroundings. From the clients point of view, physical transparency may appear to indicate a transparent organisation with a large degree of openness [1]. This emerging technology of heavily glazed facades is also often associated with buildings whose design goals include energy efciency, sustainability, and a green image. So there has been an increase in the numbers of this type of building. The success of these facades also lies in the fact that they admit a large amount of daylight, exhibit a uniform exterior, and have attractive aesthetics. The costs of double-skin facades are higher than those of normal facades, but claims of energy and productivity savings are used to justify some of these increased costs [2].

* Corresponding author. Tel.: +32 10 47 22 23; fax: +32 10 47 21 50. E-mail addresses: elisabeth.gratia@uclouvain.be (E. Gratia), deherde@arch.ucl.ac.be (A. De Herde). 0378-7788/$ see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2006.09.001

The advent of computers and other electric ofce equipment has increased the internal heat gain in most ofces. Highly glazed facades, often with poor shading, have become very common. This, together with the extra heat gain from the electric lighting made necessary by deep oor plans, and the widespread use of false ceilings, has increased the risk of overheating [3,4]. In the 1990s, concern about global warming resulted in a resurgence of interest in natural cooling strategies, including solar protections use [57]. There is also an increasing demand for high-quality ofce buildings. The occupants and developers of ofce buildings require healthy and stimulating working environments [8]. This is usually provided by an air conditioning system. But in many cases, with some effort to reduce internal heat gain (well chosen equipment), solar protection and natural ventilation may be sufcient to ensure good comfort levels for the buildings occupants. In that case, air conditioning system will not be necessary, which will result in considerable energy and cost savings. It will also indirectly reduce the burden on the environment, since the use of energy is always associated with the production of waste materials [9]. Double-skin facades are assuming an ever-greater importance in modern building practice. They are already a common

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

365

feature in architectural competitions in Europe; but there are still relatively few buildings in which they have actually been used, and there is too little information on their behaviour in operation [10,11]. There are many unknowns: optical and thermal modeling of these systems is not routine and coupling heat transfer and air ows from an isolated facade system to the whole building is complex. A variety of thermal coupling strategies must be simulated. Moreover, although subjective claims abound in the architectural literature, it is extremely difcult to nd any objective data on the actual performance of buildings with double-skin facades. Results of simulations show that heating loads are decreased in an ofce building with a double-skin facade. Indeed, the temperature of the air layer in the double-skin is more important than the outside temperature and so the cavity protects the building from the cold. Moreover, double-skin hot air can be recovered to heat the coldest zones of the building [12]. The addition of a double-skin decreases the heating loads of 11.313%. The greatest reduction is observed for the northern double-skin, indeed, in this case, the southern zones benet fully from the solar prots which are not ltered by the doubleskin and the zones north are protected by a buffer space [13]. On the other hand, results of simulations show that cooling loads are increased in an ofce building with a double-skin

facade. Indeed, the hot air layer becomes an obstacle with the cooling of the building. Application of natural cooling strategies becomes still more important in the building with double-skin than in building without double-skin. If no natural strategy is implemented to try to decrease cooling consumption, the addition of a southern double-skin increases the cooling loads of 19.7%. Indeed, since no strategy is applied, the temperature in the double-skin is quickly very important. There is an important transfer of direct solar radiation and heat by transmission towards the ofces through the windows. The addition of a double-skin in an east-west oriented building increases the cooling loads of 16.818.4%. If no natural strategy is implemented to try to decrease cooling consumption (the double-skin remains closed, solar protections are not used, the strategies of day and night natural ventilation are not used, etc.), cooling loads in an insulated building are more important than heating loads. Cooling loads are 2.5 to three times more signicant in the case of a building well insulated, two times more signicant in the case of a mean insulated building. On the other hand, if everything is implemented to reduce cooling loads by natural cooling strategies use, cooling loads are divided by six comparatively to the case without natural cooling strategies use. If double-skin is south oriented, for buildings with various insulation levels, the repartition of the loads is as follow:

Fig. 1. Temperature evolution in a closed double-skin according to orientation and use of mean coloured solar shading devices.

366 Well insulated building Heating loads = 67% of the energy demand Cooling loads = 33% of the energy demand Mean insulated building Heating loads = 76% of the energy demand Cooling loads = 24% of the energy demand Not insulated building Heating loads = 90% of the energy demand Cooling loads = 10% of the energy demand

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

Many studies have already simulated the behaviour of double-skin facades [1619]. This research is only one step in the search for a better understanding, in a qualitative way, of the thermal behaviour of one particular type of double-skin. Architectural design guidelines would help architects and owners achieve a better understanding of the applicability of various concepts to their specic building projects. 2. Methods

So, application of natural cooling strategies becomes extremely important in a building with double-skin [14]. One of the most efcient natural cooling strategies is the use of solar blinds. The temperature of the air layer in the double-skin is inuenced by many factors (solar radiation, outside temperature, wind speed, windows openings, type of glazing, etc.) but also by the presence of shading systems [15]. Fig. 1 gives, by clear sky conditions, temperature evolution in a double-skin facade according to the orientation of this one and according to the use or not of mean coloured solar shading devices in the double-skin facade. Shadings devices are then placed just along the windows of the internal skin in the cavity. All the openings of the double-skin are closed. When the solar protection devices are shut down, an important temperature increase is observed in the double-skin. The greenhouse effect increases since an additional part of the sun rays is absorbed by the solar protection devices rather than transmitted in the ofces. The position of the blind within the air cavity affects the rate of the heat transfer to the interior and amount of thermal stress on the glazing layers. Placed too close to the interior facade, inadequate air ow around the blind may occur and conductive and radiative heat transfer to the interior are increased. The blind should be placed toward the exterior pane with adequate room for air circulation on both sides. With wind-induced ventilation or high velocity thermal-driven ventilation, the bottom edge of the blind should be secured to prevent uttering and noise. To undertake the study, we chose an ofce building with a high level of thermal insulation. With the thermal program TAS we simulated various features of the double-skin. The research will help us to understand how the double-skin operates.

2.1. TAS program TAS is a software package for the thermal analysis of buildings. It includes a 3D modeller, a thermal/energy analysis module, a systems/controls simulator and a 2D CFD package. There are also CAD links into the 3D modeller as well as report generation facilities. It is a complete solution for the thermal simulation of a building, and a powerful design tool in the optimisation of a buildings environmental, energy and comfort performance [20]. 2.2. The building The simulations where undertaken using the building proposed in Subtask A of Task 27 (performance of solar facade components) of the International Energy Agency, Solar Heating and Cooling Program. Some modications were made to adapt this to Belgian practices. This is a medium-sized ofce building with ofce modules on two facades separated by a central corridor and staircase/service spaces at both ends of the building. It comprises 150 ofces, distributed over ve oors and two orientations: 15 ofces on either side of the building per oor. Fig. 2 presents the geometrical data for the ofce building. A vertical cross section of an ofce with its main measurements, is shown in Fig. 3. The internal wall between the ofce module and corridor has an openable window above the door to facilitate the air ow between northern and southern spaces (the false oor is not included in the drawing). Each ofce has four windows (two top and two bottom) to allow natural day or night ventilation.

Fig. 2. View of the ofce building studied.

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

367

Fig. 3. Geometrical data of the ofce building.

Each oor is divided into 5 zones; the building thus comprises 25 zones plus the double-skin space.
Thermal characteristics Building envelope Roof: U = 0.3 W m2 K1 Ground oor: U = 0.379 W m2 K1 Opaque part of facade: U = 0.373 W m2 K1 Low-e double glazing: U = 1.8 W m2 K1, direct solar transmission: 0.62, total solar transmission: 0.708 Double skin Clear single glass: U-value = 5.33 W m2 K1, shading factor = 0.76 Width of the air cavity: 1.2 m H (double-skin facade) = H (building) + 1 m Use of the building (internal heat gains in the ofces: 29.37 W m2) HVAC system For heating MondayFriday 0 a.m. ! 7 a.m. (15 8C) 7 a.m. ! 6 p.m. (21 8C) 6 p.m. ! 12 p.m. (15 8C) Weekend 0 a.m. ! 12 p.m. (15 8C) For cooling MondayFriday 0 a.m. ! 7 a.m. (nothing) 7 a.m. ! 6 p.m. (24 8C) 6 p.m. ! 12 p.m. (nothing) 0 a.m. ! 12 p.m. (nothing)

The factitious walls have a very small thermal resistance and a solar transmission of 1. Vertically and horizontally subdivisions are possible. The horizontal subdivisions are tracked oors. The Track label has a special function which allows the oor to be modelled as an air ow aperture. The solar radiation which strikes the tracked oor is not stopped by the oor and enters in the adjacent zone. The vertical subdivisions are modelled by permanently open window. 2.5. Shading devices Shading devices (roller blind, venetian blind with orientable slats, etc.) are modelled only as transparent layers; so optical properties are only angular dependent. There is thus no difference between roller blind and venetian blind if the solar transmittance and the external and internal reectance are the same. The convection coefcient for gas layer adjacent to blind is automatically increased by a factor 4/3 to allow for the effects of convection currents in the larger gas space enclosing the blind. In a double facade, the shading device can be placed in various positions. 2.5.1. Shading device against a facade If the shading device is against a facade it is modelled as a device associated to an external window. TAs allows having substitute building element. The construction and feature

Week end

Each stage was divided into 5 zones; the building thus counts 25 zones + the double-skin space. 2.3. Climatic data assumptions The simulations were performed with climatic data from UCCLE (Belgium). For this study, we chose a sunny summers day (24 July). Fig. 4 shows the climatic data for this day. Ten days preceding the chosen day were also modelled to take account of the effect of inertia. 2.4. Subdivision of the double-skin In Tas, only one temperature per zone is calculated. To try to understand more precisely the various phenomena which occur in the double-skin, we subdivided this one in several zones. Separations are factitious walls which can be completely open.

Fig. 4. Climatic data of the sunny summer day.

368

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

shading properties of the building element are replaced by those of a Substitute Building Element at certain times. These times are specied by the Substitution Schedule. In the most common application of this feature, the main building element represents a window and the substitute building element the same window with a blind. The substitution process then represents the drawing of the blind. 2.5.2. Shading device far from the facade If the shading device is far from the facade, the double facade can be divided into several zones and air movement can be studied. The impact of the size, the colour and the localization of those can be studied. 3. Study of the most efcient position of shading devices in a double-skin facade 3.1. Efciency of the solar protections placed in a double-skin facade The most effective solar protections are external shading devices. In the case of the double-skin facade, the blinds can be integrated in the cavity. They are thus protected from the bad weather and pollution. Solar protection can remain in place even in the event of important wind, which represents an undeniable advantage for the buildings with great height. However, the heat retained by the blinds involves a rise in the temperature in the air layer. If it is possible to crate an opening in the bottom and in the top of the double-skin facade, by the stack effect, hot air goes up and escapes outside. If no strategy is used to reduce cooling consumption in a well insulated building with southern double-skin, cooling consumption is 1123 kWh/day during a sunny summer day. The addition of blinds directly in front of the glazings of the interior skin makes it possible to reduce this consumption of 17.5% if the double-skin remains closed and of 25.7% if this one is open.

3.2. Mean coloured blindsdouble-skin closed The characteristics of the mean coloured blinds are  coefcient of solar absorptance0.42,  coefcient of solar reexion0.40. 3.2.1. Mean coloured blinds placed against the windows of the inside skin The blinds strongly warm up and reach a temperature of 69.8 8C at 1 p.m. They communicate part of their heat to the glazing although this one is quite insulating. The glazing temperature on the side of the ofces is of 37.6 8C. The phenomenon would be marked more if the glazing were normal double glazing. The window sill warms up strongly too; but as it is good insulated, it communicates less heat in the ofces. Fig. 5 compares cooling loads and windows temperature of the inside skin (side ofces) at 1 p.m. for the three positions When the double-skin is sunny and closed, a movement of thermocirculation is established. Indeed, the inside skin of the double-skin and the blinds warm up: the air in contact with these walls warms up and acquires an upswing. Quite to the contrary, along the external skin in single glass, the air cools and acquires a downward movement. The thermocirculation of the air is present day and night. It is accompanied by exchanges of air between the zones close to the external skin and the zones close to the interior skin. Air movements are illustrated in Fig. 6. Air ows are expressed in kg/s. 3.2.2. Mean coloured blinds placed against the windows of the outside skin The cooling consumption is lower than in the preceding case, whereas the air temperature inside the double-skin is higher. The air temperature in the double-skin is more important because cooling along the external skin in single glass is removed. On the contrary the single glazing becomes a

Fig. 5. Cooling loads and windows temperature of the inside skin (side ofces) at 1 p.m. for the three positions (mean coloured blindsdouble-skin closed).

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

369

Fig. 6. Air movement for the three positions (mean coloured blindsdouble-skin closed).

very hot surface since it is covered by the sunscreens. The temperature of the interior blind reaches 65.6 8C at 1 p.m. It is also for this reason that the movement of thermocirculation is reversed. It is to be noticed that during the night, the movement of air takes again its normal direction (Fig. 6). On the other hand, we note that the cooling need in the ofces is weaker. If we observe the temperatures of the inside skin, we note, that being safe from any solar radiation and not being overheated by the blinds, the glazing and the window sills have a temperature lower than in the preceding cases. Temperature of the windows on the side of the ofces is of 33.5 8C (Fig. 5). 3.2.3. Mean coloured blinds placed in the middle of the cavity The blinds protect completely the inside skin from the sun ray. They are placed over all the width and the height of the facade, only a space of 5 cm is left free in the higher part of each stage to simulate the mechanism of operation. Moreover one space of 50 cm is left free in bottom to allow the movement of thermocirculation. The part of the double-skin exceeding the building is also not occulted. The movement of thermocirculation is well established, with small passages of air by the slits of 5 cm. The air of the zones close to the interior skin rises due to the heating of the blinds (and not either due to the heating of the inside skin and the blinds); and it goes down in the zones close to the external skin due to the low temperature of the single glazing of the external skin (Fig. 6). The temperature of the interior facade is denitely lower than in the preceding cases. For that reason, cooling loads are

still reduced. Indeed, the interior facade is completely protected from the solar radiation, and the blinds cannot transmit their heat directly to the inside skin (except by radiation). Moreover, the air of the double-skin can cool partially by licking the single external glazing (Fig. 5). 3.3. Light coloured blindsdouble-skin closed The characteristics of the light coloured blinds are  coefcient of solar absorptance0.17,  coefcient of solar reexion0.65. 3.3.1. Light coloured blinds placed against the windows of the inside skin Compared to the case of mean coloured blinds, the effect of greenhouse is less important since the colour of the blinds is lighter. The temperature in the double-skin is decreases about 2.5 8C. The movements of air have the same directions but are slightly weaker. The solar gains in the double-skin decrease since the solar radiation reected by the blinds arises mainly by the external glazing. On the other hand, the solar gains very slightly increase in the ofces, that is due to the increase in the coefcient of reexion to the back of the blinds (Fig. 7). Compared to the case of mean coloured blinds, blinds temperature is decreased of 12 8C. This one passes from 69.8 to 57.8 8C. That has as a consequence that the temperature of the glazing side ofces decreases from 37.6 to 34.9 8C. The temperature of the window sills side double-skin decreased from 75.2 to 73.3 8C; that is due only to the fact that the

370

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

Fig. 7. Cooling loads and windows temperature of the inside skin (side ofces) at 1 p.m. for the three positions (light coloured blindsdouble-skin closed).

temperature in the double-skin is lower. The window sills being well insulated, that makes it possible to gain only 0.2 8C side ofces. Thanks to the reduction in surface temperature of the glazings and in spite of the light increase in solar prots in the ofces, the cooling consumption passes from 926 to 894 kWh/ day (either 3.5% less). 3.3.2. Light coloured blinds placed against the windows of the outside skin In this case, most of the solar gains is directly reected towards outside. Compared to the case of mean coloured blinds, the temperature in the double-skin is denitely lower (14 8C of less). The movements of air have the same directions but are slightly weaker. The air of the double-skin cannot cool any more by licking the single glazing, but the solar gains are denitely weaker. The temperature reached in the blind is also lower, this one passes from 65.6 to 47.4 8C; so a difference of 18 8C is observed (Fig. 7). The inside skin being safe from any solar radiation and not being overheated by the blinds, the glazings and the window sills have a temperature lower than in the preceding cases. Thanks to the reduction in surface temperature of the inside facade and in spite of the fact that the air cannot cool by the single glazing, the cooling consumption passes from 870 to 830 kWh/day (either 4.5% less). 3.3.3. Light coloured blinds placed in the middle of the cavity The blinds protect completely the inside skin from the solar radiations. They are placed over all the width and the height of the facade. The movement of thermocirculation is well established, with small passages of air by the slits of 5 cm. The air of the zones close to the interior skin rises due to the heating of the blinds (and not either due to the heating of the inside skin and the blinds); and it goes down in the zones close to the external skin due to the low temperature of the single glazing of the external skin. Compared to the case of mean coloured blinds, blinds temperature is decreased of 8.5 8C. This one passes from 62.5

to 48.3 8C. That has as a consequence that the temperature of the glazings side ofces decreases from 33.1 to 31.1 8C (Fig. 7). The cooling consumption is decreased because the temperature of the interior facade and the temperature of the air of the double-skin which can cool partially by licking the single external glazing are lower. The cooling consumption passes from 801 to 768 kWh/day (either 4.2% less). 3.4. Double-skin opened When the double-skin is sunny, the opening of the lower and higher windows in the double-skin causes an upswing of the air in the majority of the cases. Indeed, the stack effect is often more important than the effect due to the wind when the double-skin is sunny. The ventilation of the double-skin by the outside air causes a considerable reduction in the temperature of this one. The temperature is about 24 8C in the bottom of the double-skin and about  29 8C in the top for mean coloured blinds,  28 8C in the top for light coloured blinds. Note. When the double-skin is closed, the temperature is about  between 51 and 57 8C according to the location of the mean coloured blinds,  between 43 and 48 8C according to the location of the light coloured blinds. The temperature in the opened double-skin is much lower than in the closed double-skin. However, the temperature of the blinds and the opaque walls of the inside skin remains relatively important due to the direct solar radiation. In the case of mean coloured blinds placed against the windows of the inside skin, the temperature of the blinds is  69.8 8C when the double-skin is closed,  51.4 8C when the double-skin is open.

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

371

blinds or the opening of the double-skin would have been much more important if the inside facade had been less insulated. 4.1. If double-skin is closed The inuence of the position and the size of the blinds is important. Indeed, the judicious choice of the location and the size of the blinds makes it possible to save up to 14.1% of the cooling consumption of the building during a sunny summer day. Table 1 shows the inuence of the blinds position on the cooling consumption. The colour of the blinds makes it possible to save up to 4.6% of the cooling consumption of the building during a sunny summer day (Table 2). 4.2. If double-skin is opened The inuence of the position and the size of the blinds is important. Indeed, the judicious choice of the location and the size of the blinds makes it possible to save up to 13.9% of the cooling consumption of the building during a sunny summer day. Table 3 shows the inuence of the blinds position on the cooling consumption. The colour of the blinds makes it possible to save up to 3.5% of the cooling consumption of the building during a sunny summer day (Table 4). When the blind is located interior side of the external skin, the cooling consumption of the building is even slightly lower for mean coloured blinds than for light coloured blinds (see Section 3.4).
Table 1 Inuence of the blinds position on the cooling consumption when the doubleskin is closed Double-skin closed Mean coloured blinds Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 926 kWh/day 6.0% 13.5% Light coloured blinds 894 kWh/day 7.2% 14.1%

Fig. 8. Cooling loads and windows temperature of the inside skin (side ofces) at 1 p.m. for the two blind positions and two blind colours (double-skin opened).

A temperature decreasing of 18.4 8C is reached thanks to the cooling caused by the air circulation in the double-skin. That has as a consequence that the temperature of the glazings side ofces decreases from 37.6 to 33.1 8C (Fig. 8). The cooling consumption of the building is 834 kWh/day. When the blind is placed in the middle of the cavity, the blinds are cooled on the two sides and reach a temperature of 36.1 8C (light coloured blind, open double-skin). When the double-skin was closed, the blinds temperature was 48.3 8C. That has as a consequence that the temperature of the glazing side ofces decreases from 31.1 to 28.1 8C (Fig. 8). The cooling consumption of the building is then 711 kWh/ day. Fig. 8 presents the two situations. When the double-skin is opened, the colour impact is denitely less important. When the blind is located interior side of the external skin, the cooling consumption of the building is even slightly lower for mean coloured blinds than for light coloured blinds. Indeed, the double-skin being ventilated, the difference in temperature between the mean coloured and the light coloured blind decreases. More, the light coloured blind tends to reect more light towards the inside of the ofces. That explains the small difference on the level of cooling consumption. 4. Discussion Simulations were performed for a double-skin ofce building whose inside facade is well insulated and whose windows are performing. The inuence of the choice of the

Table 2 Inuence of the blinds colour on the cooling consumption when the double-skin is closed Double-skin closed Mean coloured blinds (kWh/day) Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 926 870 801 Light coloured blinds (%) 3.5 4.6 4.1

372

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373

Table 3 Inuence of the blinds position on the cooling consumption when the doubleskin is opened Double-skin opened Mean coloured blinds Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 834 kWh/day 8.9% 13.9% Light coloured blinds 812 kWh/day 6.2% 12.4%

effect of cold surface close to the windows is decreased, and that the radiators should not any more be obligatorily along sills. Let us note that the current insulating glazing already mainly make it possible to limit the problems of cold surface of the glazing. This argument can be turned over against the defenders of the double-skin. Indeed, in summer period, the glazing of the interior facade can reach high temperatures due to the important temperature in the cavity of the double-skin and due to the hot radiation coming from the blinds. So, the study of the optimal position of the blinds makes it possible to reduce cooling consumption of the building but also implies an increasing of the occupants comfort. 5. Conclusions

Table 4 Inuence of the blinds colour on the cooling consumption when the double-skin is opened Double-skin opened Mean coloured blinds (kWh/day) Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 834 760 718 Light coloured blinds (%) 3.5 +0.1 0.8

Table 5 compares all the studied cases. We see that a cooling consumption decreasing until 23.2% can be reached by paying attention to  the location of the blinds,  the blinds colour,  the opening of the double-skin. The impact of the opening of the double-skin is obviously important; the reduction of consumption varies from 7.4 to 12.6% (Table 6). The inuence of the position of the blinds is even larger than that due to the opening of the double-skin. The judicious choice of the location and the size of the blinds makes it possible to save up to 14.1% of the cooling consumption of all the building during this sunny summer day (Table 7).

4.3. Comfort impact The defenders of the double-skin advance the following argument: the usable space is increased due to the fact that the

Table 5 Cooling loads comparison between all the congurations compared with the case where the mean coloured blinds are placed against the windows of the inside skin in a closed double-facade Double-skin closed Mean coloured blinds Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 926 kWh/day 6.0% 13.5% Light coloured blinds (%) 3.5 10.4 17.1 Double-skin opened Mean coloured blinds (%) 9.9 17.9 22.5 Light coloured blinds (%) 12.3 17.7 23.2

Table 6 Impact of the opening of the double-skin on the cooling consumption Double-skin closed Mean coloured blinds (kWh/day) Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 926 870 801 Light coloured blinds (kWh/day) 894 830 768 Double-skin opened Mean coloured blinds (%) 9.9 12.6 10.4 Light coloured blinds (%) 9.2 8.2 7.4

E. Gratia, A. De Herde / Energy and Buildings 39 (2007) 364373 Table 7 Impact of the blinds location on the cooling consumption Double-skin closed Mean coloured blinds Blinds placed against the windows of the inside skin Blinds placed against the windows of the outside skin Blinds placed in the middle of the cavity 926 kWh/day 6.0% 13.5% Light coloured blinds 894 kWh/day 7.2% 14.1% Double-skin opened Mean coloured blinds 834 kWh/day 8.9% 13.9%

373

Light coloured blinds 812 kWh/day 6.2% 12.4%

This study has showed the great inuence that the position and the colour of the blinds have on the cooling consumption in an ofce building with a double-skin. It also highlights the importance of the opening of the double-skin. Other interesting factor is the impact of the blinds characteristics on the human comfort. The position and the colour of the blinds have an inuence on the temperature of the windows of the inside skin and so on the hot radiation coming from the windows to the occupants. Acknowledgements The authors wish to thank the Walloon Regional Government of Belgium for its support in funding this research. We would also thank Jose Flemal (draftsman) who helped us to illustrate the article. References
[1] O.J. Hendriksen, H. Sorensen, A. Svenson, P. Aaqvist, Double skin facadesfashion or a step towards sustainable buildings, in: Proceedings of ISES, Eurosun, 2000. [2] H. Poirazis, Double skin facades for ofce buildingsliterature review, Report EBD-R-04/3. Division of Energy and Building Design, Lund University, 2004. [3] E. Gratia, A. De Herde, Guidelines to improve efciency of a double-skin facade in an ofce building, in: Proceedings de la 26th AIVC Conference, 2005. [4] E. Gratia, A. De Herde, Is day natural ventilation still possible in the ofce buildings with a double-skin facade? Building and Environment 39 (4) (2004) 399409. [5] F. Allard, Natural Ventilation in Buildings, a Design Handbook, Altener Project, James and James, London, 1998. [6] M. Liddament, A Guide to Energy Efcient Ventilation, Air Inltration and Ventilation Centre, University of WarWick Science Park, Great Britain, 1996.

[7] D. Dickson, Ventilation Technology in Large Non-Domestic Buildings. Air Inltration and Ventilation Centre, University of WarWick Science Park, Great Britain, 1998. [8] D. Rennie, F. Parand, Environmental design guide for naturally ventilated and daylit ofces. Building Research Establishment Ltd., Construction Research Communications Ltd., PO Box 202, Watford, Herts, WD2 7QG, 1998. [9] N. Wong, A. Mahdavi, J. Boonyakiat, K. Lam, Detailed multi-one air ow analysis in the early building design phase, Building and Environment 38 (2003) 110. [10] A. Zollner, E.R.F. Winter, R. Viskanta, Experimental studies of combined heat transfer in turbulent mixed convection uid ows in double-skinfacades, International Journal of Heat and Mass Transfer 45 (22) (2002) 44014408. [11] L. Zalewski, S. Lassue, B. Duthoit, M. Butez, Study of solar walls validating a simulation model, Building and Environment 37 (1) (2002) 109121. [12] E. Gratia, A. De Herde, Optimal operation of a south double-skin facade, Energy and Buildings 36 (1) (2004) 4160. [13] E. Gratia, A. De Herde, Thermique des immeubles de bureaux, UCL Presses Universitaires de Louvain, 2006 http://www.i6doc.com. [14] E. Gratia, A. De Herde, Natural cooling strategies efciency in an ofce building with a double-skin facade, Energy and Buildings 36 (11) (2004) 11391152. [15] E. Gratia, A. De Herde, Greenhouse effect in double-skin facade, Energy and Buildings 39 (2) (2007) 199211. [16] C. Balocco, A simple model to study ventilated facades energy perfor mance, Energy and Buildings 34 (5) (2002) 469475. [17] T. Pasquay, Natural ventilation in high-rise buildings with double facades, saving or waste of energy, in: Proceedings of the 18th International Conference on Passive and Low Energy Architecture, Brazil, 2001. [18] H. Manz, Numerical simulation of heat transfer by natural convection in cavities of facade elements, Energy and Buildings 35 (3) (2003) 305311. [19] J. von Grabe, A prediction tool for the temperature eld of double facades, Energy and Buildings 34 (9) (2002) 891899. [20] A. M. Jones, EDSL Ltd, TAS, Software package for the thermal analysis of buildings, 13/14 Cofferidge Close, Stony Stratford, Milton Keynes, Mk11 1BY, United Kingdom.