Case study of a double skin façade : focus on the gap between predicted and measured

Full text

VII International Congress on Architectural Envelopes
May 27, 28, 29 2015, San Sebastian-Donostia, Spain
VII International Congress on Architectural Envelopes
Case study of a double skin façade: focus on the gap between
predicted and measured
Energy efficiency – Thermal storage
M. Comminges, A. Dugué
Research Technical Organization NOBATEK
e-mail: mcomminges@nobatek.com ; adugue@nobatek.com
Key words: Dynamic facades, Analyze and feedback, Technical and Technological improvements, gap
between predicted design and real design
Abstract
Double skin façade are particularly interesting for their esthetic values and their high energy
performances for both new design and renovated buildings. The vast majority of the studies focus on the
simulation aspects as complex CFD calculations but, conversely, few studies pay attention to the most
practical aspects which are, to provide useful pragmatic feedbacks for future design of such complex
double skin façade.
This paper tackles the particular issue of the complexity of the design and the piloting of double skin
façade. The headquarter of the Office64 in Bayonne, France is used as a case study. This building
integrates a 50m wide and 4 storey high double skin façade orientated toward south. The double skin
façade was designed to limit the thermal losses in winter and to preheat the fresh air entering in an air
handling unit connected to the double skin via a by-pass.
A monitoring campaign has been carried out during the year 2014. Sensors for temperature, humidity
and solar irradiation were placed into the double skins. Sensors of energy consumption were placed into
HVAC systems. The thermal behavior of the double skin has been studied and qualified. The focus was
not so much on the explanation of thermal transfers but more on the identification and explanation of
the gap between what was designed and predicted and what is now measured.
Results show that the origins of differences are often due to bad communication during the numerous
actors of the design phase.
This feedback allowed us to highlight a number of recommendations and identify some technical and
technological improvements in design.

VII International Congress on Architectural Envelopes
May 27, 28, 29 2015, San Sebastian-Donostia, Spain
1 Context and objectives
Facades are a central item in contemporary architecture because it is, firstly, the main way to manifest
the aesthetic of the project and, secondly, it is an essential component of the energy efficiency of
buildings (e.g. solar gains and protections, natural light benefits, etc.). It has also a great impact on the
occupant comfort as it is the shelter that provides him light, protects him from noise and rain.
The technicality of the facades is continuously growing with a strong development of both passive and
active solutions. This improvement of the façades focuses on the optimal design either by using new
materials (glass or thermal break for example), or by an optimization approach of the design in a given
context. But it does not take as much into account feedbacks of existing buildings where such technics
and systems were applied.
Indeed the complexity of façades implies a great interaction with others systems, but the relation to the
users are also strong and hard to anticipate. Therefore the analysis of the functioning of façades, of the
behaviours of users can also provide great knowledge in order to improve the design of façade.
Double skin façades (DSF) are a good example of such complex façade which have both a high
architectural value and a great energy impact. In addition to the aesthetics DSF are often used for
increasing the acoustic attenuation, for reducing the thermal thermal losses through the envelope. They
can also be used as solar collectors as the air between the two skins can heat due to greenhouse effect.
The approach of most of the studies on DSF is theoretical. It can be on the thermal modelling, as done
by [1, 6], or often on how to optimally design those in a given context [2, 4]. Even when the positive
impact of the DSF is questioned, the answer comes from simulation as did Radhi [5] who used CFD to
analyze the behavior and impact of the DSF of a real building in summer. [3] used TAS software to
analyze if the added DSF lowers the energy consumption of a fictional building.
But we know that due to the complexity of the installation, piloting, interaction with users, there can be
a gap between modeled and real behavior.
This paper tackles the particular issue of the complexity of the design and the piloting of double skin
façade. This feedback presents results from one year measurement campaign of the headquarter of the
Office64 in Bayonne, France This building integrates a 50 m wide and 4 storey-high double skin façade
orientated toward south. The double skin façade was designed to limit the thermal losses in winter and
to preheat the fresh air entering in an air handling unit connected to the double skin via a by-pass.
2 General presentation of the case study building
2.1 Presentation of the building and its double skin façade
The studied building is the headquarter of the Office64, based in Bayonne, France with an area about
4300m².
The envelope can be divided into three parts:
 South façade: The south façade is a double skin with:
o Interior façade: Double glazing associated with an insulated concrete wall of 100mm +
a zinc cladding
o Exterior façade: Single glazing curtain wall fixed on a primary metal frame and
secondary on a Douglas pine frame.
 North façade: Triple glazing with an insulated concrete wall of 100mm + a zinc cladding.
East and West façade: insulated concrete wall of 100mm

VII International Congress on Architectural Envelopes. May 27, 28, 29 - 2015, Donostia-San Sebastián, Spain
Fig 1 :South façade from East
Fig 2: South facade
Fig 3: North facade
At the early stage of the design phase, the DSF was contemplated in order to increase the acoustic
attenuation from a close street and in order to reduce the thermal energy loss through the envelope.
A quick energy analysis focused on the relevance of a DSF on the north façade but it was finally
discarded. Later on, it was decided that in order to take advantage from the greenhouse effect that takes
place in the DSF, the preheated air in the cavity would be used for ventilation purposes. Next HVAC
systems have been designed accordingly.
2.2 Focus on the systems
2.2.1 Vents in the DSF
Fig 4 : Air cavity
Fig 5 : Motorized
openings
Fig 6 : Motorized openings distribution
The double skin is equipped with motorized vents controlled by the Building Management System
(BMS). The opening of the vents is done accordingly to the following rule that depends on the air
temperature and the relative humidity in the DSF. It does not depend on the external air temperature.
 OPEN either if TDSF ≥ 25 ° C OR RHDSF ≥ 80%
The largest number of vents is implemented at the bottom and the top part to increase the air circulation.
2.2.2 Connection of the double skin façade with HVAC system
About HVAC systems, two independent air/water reversible heat pumps allow heating and cooling. And
three different Air Handling Units (AHU) allow the air renewal. Two AHU have a double flux exchanger
and are connected to the DSF: the fresh air can be taken either from the outside or from the DSF, a valve
allowing to connect it to the DSF or the outside. The grid for the air intake in the DSF is located at the
top of it. The functioning in winter is represented in the figure 7.

M. Comminges, A. Dugué
The objective is to take advantage of the greenhouse effect in the DSF. When there are thermal losses
associated to air renewal (e.g external temperature is lower than internal temperature), the new air can
be taken in the DSF instead of the outside, this way the loss can even be turned in a heat gain if the air
temperature in the DSF is higher than the internal air. All elements are controlled by a BMS and allow
an accurate tracking of overall performance with measurements made by sensors (temperature, humidity,
operating status, etc.).
The figure 7 illustrates the general functioning of the system in winter. In summer, the new air is always
taken from the outside.
Fig 7 : Winter functioning of the HVAC systems connected do the DSF
3 The monitoring protocol
3.1 Objectives of the monitoring
This building is a good example of an active DSF integrated in a building. Thus it can provide an
interesting feedback to disseminate to designers of future buildings with DSF.
The first objective of this monitoring campaign is to assess the DSF behavior compared to external
solicitations and the building behavior. Next, the idea is to assess the gap between what was designed
and predicted and what is now measured. Lastly, from the data analysis some pieces of information will
be provided to the building owner in order to improve the functioning of the building and the DSF; and
more general feedbacks will be compiled in order to be included in a general methodology for the design
of active façades.
It has to be noted that this is a quick monitoring campaign that aims more at focusing on some precise
practical elements such as the placement of the sensors than one global one that would aim at measuring
precisely the energy reduction brought by the DSF.

VII International Congress on Architectural Envelopes. May 27, 28, 29 - 2015, Donostia-San Sebastián, Spain
3.2 Implementation and protocol
A BMS allows the visualization of data and the possible modifications of some key parameters. But in
order to complete missing data, and to obtain more reliable measurements, sensors were added for the
one year long campaign.
Fig 8 : South facade sensors
Fig 9 : AHU sensors
All the temperature sensors located in the DSF are illustrated in the figure 9. Two pyranometers
operating on the spectral range 300 to 2800 nm are installed facing south, one in front of the external
glazing of the DSF and one just behind it. A sensor for the external temperature is also added. Others
sensors have been installed in the AHU systems in order to complete missing data from the BEM, as can
be seen on figure 10.
4 Exploitation of data
4.1 Behavior of the double skin façade
Thermal analysis
As seen in figure 9, 9 different temperatures sensors were placed in the DSF. Vertical and horizontal
gradient have been studied and it shows quite well the heterogeneity of the temperature field in the DSF.
Here we present two different two-day long sequences. First one is in summer, the 16th and 17th of July
2014 and the second shows two days in winter, 18th and 19th of January 2014.
Fig 10 : Temperatures and Solar irradiation in
the DSF during 2 hot days in summer
Fig 11 : Temperatures and Solar irradiation in
the DSF during one cloudy then one sunny day in
winter

M. Comminges, A. Dugué
In summer, we can see based on incident solar radiation that the first day is sunny and second one is
cloudy. In winter, the first one is cloudy and the second one is sunny. On a sunny day the incident solar
radiation is way higher in winter as it is over 800W/m² while it is under only 400W/m² in summer.
Moreover, the solar transmittance of the external glazing of the DSF does not appear to be the same: 0,2
in summer and 0,6 in winter. This is due to the fact that the incident angle of the solar radiation is high
in summer, thus the transmittance of both the glazing and the sensor is lowered. This measure cannot
served as an evaluation of the solar transmittance but reveals the fact that solar gains are higher on a
sunny winter day than on a sunny summer day.
On both sequences, a strong correlation between the incident solar radiation and the temperature in the
DSF can be seen. The maximum difference between outside air and DSF air is 10°C in summer and
20°C in winter. The vertical gradient associated to the chimney effect can be seen, and the difference of
temperature between the highest point (level 3) and the lowest point (level 0) is constantly around 2-3°C
during daylight. In summer, the vents are always open, which favor natural ventilation and limit the
vertical gradient. In winter, the vents also open on the sunny day as the DSF air temperature rises above
25°C.
One key point is that at night in summer, the DSF air temperature is the same than outside air, while in
winter it stays 3°C above outside air. The DSF acts as a buffer zone and limits the thermal loss of the
associated building.
Vents functioning analysis
The operating of the vents has been checked. The opening law is respected. In summer and winter, when
it is sunny, the 25°C temperature - the upper limit for the opening - is often reached. This limit was set
in order to prevent overheats and thus the deterioration of some elements of the façades due to high
temperatures such as the silicone, or motors… But this temperature limits the energy potential from the
DSF in winter and also makes the opening movement of the vents very frequent. Therefore while
wanting to preserve those elements, it may be the contrary. As an example, during a sunny day in March
with external air temperature beneath 20°C, at peak hour, the vents open and close every 5 minutes.
As a way to improve the efficiency of the DSF, we advise to have different temperatures limits
depending on the season and, secondly, to use a hysteris with a delay time.
4.2 Are the BEM sensors relevant?
The BMS temperature sensor for the DSF is located at the level 3, on the west side on the double skin,
behind the cladding elements and the temperature it gives is used to manage the opening-closing of the
double skin vents and the control of the mixture of outdoor / greenhouse air entering into the AHU. It
is therefore essential to have a representative measure in order to use the energy systems efficiently.
The temperature measured by this sensor is compared with the temperatures at the lowest and highest
point of the DSF in figure 12 for the same 2 days than in figure 10. Firstly as it is located on the 3rd
floor, it is at the warmest point of the vertical temperature field of the double skin and, secondly, it is
behind the cladding elements so with a measured temperature higher than those measured by our HOBO
sensors. It was actually placed behind the cladding in order to prevent direct solar radiation on it.
For the functioning of the vents, a maximal temperature in the DSF is useful, but as said above the upper
limit should be greater. And for the air intake in the DSF, the BMS needs the average temperature in the
DSF which is not given by the sensor.
Thus to improve the AHU and the vents functioning the value measured by the BMS sensor should be
an average value of the DSF air temperature and not only a maximum. This would require two sensors
located at the levels 0 and 3.

VII International Congress on Architectural Envelopes. May 27, 28, 29 - 2015, Donostia-San Sebastián, Spain
Fig 12: Comparison of the DSF temperature given by the BMS with temperatures measured by our
monitoring campaign
4.3 Connection with HVAC system and identification of the cause of the gap
The precise connection with the HVAC system and the global energy consumption of the building is yet
to be analyzed. Though as explained about the opening of the vents, those open at 25°C and make
potential heat unused. At the moment of thermal simulation of the buildings those practical elements
have not been taken into account.
It is also important to note that the occupants of the building can open their windows that are connected
to the DSF. In this situation, thanks to a contactor, the air renewal of the room is stopped. It was not
possible to model this behavior.
Also, during winter the air temperature rises above the internal air temperature, in this case, the heat
exchanger of the double flux should be avoided. But the installation of an automatic bypass is also
expensive.
Therefore there are many elements that are different between what was designed and predicted and what
has been built and how it is now used. Those are mainly due to practical reasonings that came after the
first design phase. Indeed, at the moment of the modeling, we can say the approach is optimistic and
does not take into account all the possible situations, occupants’ behaviours, mechanical and financial
aspects.
5 Analysis, toward a methodology for an efficient design of double skin
façade
Elements to be included in a design methodology for double skin façade
For the future conception of a building with an integrated double skin façade, this feedback has enabled
us to put forward a number of recommendations.
 Implementation and selection of sensors for the double skin control: As seen above, the temperature
field in the DSF is not homogeneous. Thus, according to the implementation of the sensor it can
measure a maximum temperature, an average temperature or even none of them. However, these
sensors are used to set a precise piloting. For instance for the heat recovery in the BMS, the average
temperature is required and conversely for the vents opening the maximum temperature is preferable.

M. Comminges, A. Dugué
The parameters to be measured must be well identified, and the sensors and their location must be
chosen accordingly.
 Location of the vents openings in the double skin facade: During the summer the temperature of the
double skin constantly remains above the outdoor temperature (from few degrees to 10 ° C) while
the vents are always open. Then, natural ventilation with horizontal openings is not enough and could
be improved by implementing vents in the upper part of the double skin facade.
 Management of the vents by the BMS: To improve the benefits from the preheated air from the
double skin façade we should have different setpoints of relative humidity and temperature according
to the seasons. The setpoints change will allow the vents to not be opened in winter and then take the
maximum advantage from the preheated air in the double skin facade and thus reduce the system
consumption. In order to reduce the mechanical stress of vents and get a system more robust we
should have a range of setpoints values rather than a single guidance value.
 Traceability of the information from the conception: For the BMS part, we have had many unclear
information, reflecting poor communication between actors and an excessive number of stakeholders
during the design phase. Feedbacks allow us to realize the differences between the conception and
the reality and then to correct them but here the system rigidity prevented from doing it so simply.
Therefore for BMS, flexibility and scalability must be considered. There must be a commitment in
the Special Technical Specifications (STS) which must ask for a better traceability of the information
about the building system.
 The taking into account of misuse of the building at the modeling phase: The thermal modeling are
often optimistic. For a building which behavior and thus its energy consumption relies deeply on the
systems, a difference in the use can correspond to great differences in the energy consumption. It is
then needed to model potential misuse and see their impacts.
 An adaptive BMS and the commissioning phase: The importance of the commissioning phase of a
building is now well accepted. It will be even more important for a building that relies deeply on its
system, occupants’ behaviors and piloting. Thus at the design phase, it must be included in the
contracts that the BMS must be adaptive in the following years of the completion of the building.
Acknowledgements
We acknowledge the Office 64 for letting us run this monitoring campaign of their building and the
architect Patrick Arotcharen for providing all the necessary elements.
Reference
[1] Ghadamian, H., Ghadimi, M., Shakouri, M., Moghadasi, M., & Moghadasi, M. (2012). Analytical
solution for energy modeling of double skin façades building. Energy and Buildings, 50, 158–165.
[2] Gratia, E., & De Herde, A. (2004). Optimal operation of a south double-skin facade. Energy and
Buildings, 36(1), 41–60.
[3] Gratia, E., & De Herde, A. (2007). Are energy consumptions decreased with the addition of a
double-skin? Energy and Buildings, 39(5), 605–619.
[4] Joe, J., Choi, W., Kwak, Y., & Huh, J.-H. (2014). Optimal design of a multi-story double skin facade.
Energy and Buildings, 76, 143–150.
[5] Radhi, H., Sharples, S., & Fikiry, F. (2013). Will multi-facade systems reduce cooling energy in
fully glazed buildings? A scoping study of UAE buildings. Energy and Buildings, 56, 179–188.
[6] Stec, W. J., van Paassen, a. H. C., & Maziarz, a. (2005). Modelling the double skin façade with
plants. Energy and Buildings, 37(5), 419–427.