Glass in Buildings II

University of Bath

April 7/8, 2005

THE CASE FOR VENTILATED FACADES LATEST DEVELOPMENTS TO PREVENT SOLAR OVERHEATING OF HIGHLY GLAZED BUILDINGS.
Achim Geissler Senior Faade Engineer of Schmidlin AG, Aesch, Switzerland.

ABSTRACT

Architects love to design highly glazed facades, providing continuity from the outside world into the building interior, and providing light, airy spaces in which the occupants can work and live. However, energy considerations particularly in relation to use of air conditioning have provided significant challenges for the design of such buildings. Put in a nutshell, building occupants want thermally and visually comfortable buildings. As one possible approach, ventilated or so called double facades have been developed as an approach to provide energy efficient faades with both high thermal insulation values and good solar control, and allowing large areas of glazing to the faade. Different configurations can be employed, and each has a range of features and benefits that must be evaluated to ensure the appropriate solution is selected for a given project. In addition to the usual concerns in regard to the faade itself such as design aesthetics, buildability and weather tightness, the thermal faade performance must be determined together with the building as a system. This paper discusses some basic questions of double faade design and presents some recently built double faade constructions. Levels of (thermal) performance that are achieved by each are outlined.

INTRODUCTION Highly glazed faades provide visual continuity from the outside world into the building interior. They provide light, airy spaces in which the occupants can work and live. Notwithstanding their current popularity with architects, such highly glazed facades have been much discussed, e.g. [Gertis99] and are often damned [Schulz04]. Energy considerations particularly in relation to necessary air conditioning provide significant challenges for the design of buildings with highly glazed faades [Saelens03]. Of course, such faades can and should not be considered by themselves at any time in the design process when crucial design decisions in respect to thermal building performance are made. The whole building including its proposed usage in terms of internal loads, its construction in terms of thermal masses available and its control in terms of HVAC system must be viewed in an integrated way. Such detailed performance evaluation may often lead to the result that energy considerations are not in favour of a static double skin faade (that is to say, a double skin faade e.g. with unchanging ventilation openings over the seasons). Of course, there are other reasons to choose a double skin faade, for example the protection of the solar shading device from wind forces in high rise buildings or the possibility of night time natural ventilation. In order to correctly evaluate the energy efficiency of a given faade design, an annual energy simulation focussing on both heating and cooling load is necessary [Saelens03]. Apart from the above mentioned factors, the simulation model used should possibly take the building site especially any shading by adjacent buildings into account. A detailed model of a building will also take the environmental situations for different faade orientations into account (open country, adjacent building, concrete plaza). Oversimplified models, e.g. models that disregard

Proceedings

Contribution Page 1/8

Glass in Buildings II

University of Bath

April 7/8, 2005

the performance of solar shading devices for diffuse radiation will lead to wrong possibly quite wrong results for heating and cooling loads. As faade contractor that is generally involved in a stage of the overall building process when most basic parameters have already been fixed, we must assume that the specifications given for the faade design have been drawn out taking all the above mentioned parameters into consideration. However, specification documents often seem to tell differently. More often than not, the approach to faade modelling is very simplistic, centre of U-values and area percentages (that need not correspond to the architectural design, as design evolves and corresponding simulations are not necessarily updated) are used to describe thermal properties. Shading schedules and/or control strategies are often rough. Sometimes it may be a good idea to involve a faade contractor at an earlier stage of the design process to avoid unnecessary mistakes.

FAADE TYPES Basically, two double faade configurations are being used most often: a) External curtain glass gap internal IGU and b) External IGU gap internal curtain glass. In this paper, configuration a) is called external natural ventilated faade and configuration b) is called extract mechanical ventilation faade. Of course, a wide range of variations in regard to macroscopic geometric details of these two basic configurations is used e.g. box type window, band type faade, the gap spanning more than one storey, various gap widths etc. Furthermore, the ventilation rate may vary in a wide range. For the external natural ventilated faade, the ventilation rate depends mainly on the size of openings available. Figure 1 shows above mentioned basic faade configurations with their main elements. Transom

low- coating curtain glass Solar blind

Extract fan

curtain glass Insulting glazing unit

thermal break

Figure 1: Typical basic double faade configurations. External natural ventilation (left) and extract mechanical (right).

Proceedings

Contribution Page 2/8

Glass in Buildings II

University of Bath

April 7/8, 2005

Often, the architect does not wish to exhibit functional openings such as the ventilation openings of an external natural ventilated faade. Then, only the necessary (or possibly slightly enlarged) movement joints between faade elements remain for ventilation opening cross section. Subtract from this any insect mesh or the like which is inserted into the openings and the resulting free ventilation opening area is quite small. To achieve low g-values, high quality coatings in the IGU will be necessary, here. Taking the next step in this procedure and omitting air outlets at the top of the faade gap leads to a pressure equalised system. That is, a system in which there is no ventilation, but simply a pressure equalisation between the faade gap and the ambient. The frequency and duration of interstitial condensation mainly on the outside and inside face of the curtain glass will generally increase with the reduction of ventilation. However, if (or when) this condensation poses a problem is a question of taste, mainly. One interesting aspect of the extract mechanical ventilation faade is the internal surface temperature (the temperature of the internal face of the curtain glass). This surface temperature will always be somewhat higher than for an external natural ventilation system due to the highest system temperature generally being the closed blind.

PERFORMANCE PARAMETERS General For overall faade design, at least the following general parameters should be considered: Architectural requirements / restrictions Thermal performance to be achieved (U-value, g-value, layer temperatures) Flexibility (adjustable performance) Interaction strategy with HVAC systems (extract rate, natural ventilation).

Apart from these parameters of a more general nature, the following more specific parameters may prove to have significant impact on possible design and, therefore, thermal faade performance: Loads Maintenance (inside or outside?) Elements module sizes (e.g. 1.5 vs. 3.0 m) Availability of local goods Investment vs. running costs (integrated view)

In the following, however, we will focus on the two main thermal performance parameters Uvalue and g-value from a faade contractors point of view. Typically, these values are given in the project specifications. However, in the case of double faades we are dealing with here, there are quite a few uncertainties in regard to compliance confirmation due to a lack of clear definitions in relevant standards.

U-value In general, U-value calculation is based on detailed two-dimensional thermal simulation for all frames, panels and glazing edges involved in a given faade system. The calculation procedure is described in [ISO10077-1, ISO10077-2] for framing elements and in [prEN13947] for faade systems. Still, uncertainties in various definitions / procedures for specific constructions abound.

Proceedings

Contribution Page 3/8

Glass in Buildings II

University of Bath

April 7/8, 2005

E.g., how to treat faades with projections incidentally, experience shows that this is another general area of uncertainty in early stage thermal modelling, or how to treat the faade air gap has been a matter of lengthy discussions. According to [EN6946], any ventilated air gaps are disregarded for U-value calculation. This is the approach to take for ventilated double faades. However, for external natural ventilated systems, the external heat transfer coefficient hse can be reduced due to the additional radiation barrier and reduced convection on the external surface of the internal IGU (and frames and panels, of course). In more than one project featuring an internal extract mechanical ventilated system, the specification clearly assumed that the heat balance of the air going through the faade gap should be taken into account. To our knowledge, there is no standardised procedure to do this, though. Yet, clarification of calculation methods is usually quite difficult in context of a project. One is asked to confirm performance, end of story. It must be kept in mind, however, that at this point the faade contractor is being asked to mess around with parts of the HVAC of a building.

g-value For practical purposes, the calculation of g-values for double faade systems is even less defined, than U-value calculation. Just to name a few possible points of discussion, calculating the g-value according to [EN410] does not allow for ventilated air gaps or solar shading devices. Moreover, the g-value is defined for normal incidence of solar radiation only and is evaluated with temperature equilibrium between inside and outside. This standardised g-value gives a value for direct-diffuse radiation (direct radiation as input, diffuse radiation as output). All these assumptions and restrictions lead to a value which is not very realistic for the assessment of the performance of a faade in regard to solar radiation. The international draft standard [ISO15099] gives algorithms for a much more detailed evaluation of the thermal behaviour of windows including shading devices and ventilated gaps. However, this standard does not define a g-value as such. It does define a total solar energy transmittance which could be viewed as an effective g-value, that is to say, a g-value that takes ventilated layers and temperature differences between inside and outside into account as well as non-normal incidence angles of the solar radiation. For double faades, we generally use this approach. Typically, in addition to the specification of the g-value of the IGU, a g-value for blind down and closed is specified for double faades. However, in the early stages of a given project, more often than not no detailed information on the shading (colour, surface finish, possible perforation) is available. Additionally, detailed information about the glazing especially any coatings used is difficult to come by. Therefore, a similar product must be used for the assessment. With luck, such a similar product can be found e.g. in [IGDB]. If this is the case, the calculation results themselves can be viewed as quite accurate, as comparison with measurement results show. However, as for example the blind colour does have an impact on resulting g-values, the question remains how the building performance was evaluated without defining the shading system in detail. Measurement of system g-values for double faades including shading devices usually does not take ventilation into account. Such measurements can be viewed as a worst case scenario for ventilated systems. As measurements show, insulating glazing units of modern build allow the achievement of very low g-values even for unventilated systems.

Proceedings

Contribution Page 4/8

Glass in Buildings II

University of Bath

April 7/8, 2005

Miscellaneous For external natural ventilated faades, condensation on the curtain glass (both on the internal and external surface) is often an issue. In general, such condensation can never be completely avoided. However, given a reasonable ventilation of the faade gap the frequency of occurrence during daylight hours will usually be acceptable. Of course, design variations can be checked in this respect, as well. A further question being brought up in more and more cases is that of thermal breakage. As faade contractor, we usually put this question through to the glass supplier. However, agreed upon procedures (what is the worst case scenario?) and decision criteria (Which local temperature gradient is too high? What time scale is to be regarded?) would be very welcome for improved planning possibilities.

EXAMPLES Recently built double faades give an impression of the wide variety of possible visual and performance solutions. Figures 2 through 5 each show a building overview and a detailed view of the faade of a recent Schmidlin project. The achieved performance for each of the faades is given below the figures. Hereby, U-value is the system (winter) U-value and the g-value is for blind down and closed (without temperature difference). Table 1 summarises the performance parameters of these faades and for comparison the performance data of a double faade for which the g-value was measured as well as a simple box window type faade.

Figure 2: ADIA, Abu Dhabi. Internal extract mechanical ventilation, gap width 2 approx. 0.2 m. U-value 1.7 W/(m K), g-value 0.18 0.02 (air extract rate 8.5 l/(s m)). The vision glass area is 90% of the faade area.

Proceedings

Contribution Page 5/8

Glass in Buildings II

University of Bath

April 7/8, 2005

Figure 3: Swiss Re, London. Extract mechanical ventilation, gap width approx. 1.0 2 1.5 m. U-value 1.3 W/(m K), g-value 0.17 0.02 . The vision glass area is 85 % of the faade area.

Figure 4: RZVK, Kln. Externally naturally ventilated faade, gap width approx. 0.7 2 m. U-value 1.5 W/(m K), g-value 0.09 0.02. The vision glass area is 90 % of the faade area.

Proceedings

Contribution Page 6/8

Glass in Buildings II

University of Bath

April 7/8, 2005

Figure 5: Gemeindezentrum Reinach. Externally naturally ventilated faade, gap 2 width approx. 0.15 m. U-value 1.2 W/(m K), g-value 0.11 0.02. The vision glass area is 90 % of the faade area.

Table 1:

Typical U-values and g-values for some double faade systems. Values given are taken from Schmidlin projects. IGU g-values correspond to [EN410]. System g-values given are effective g-values according to [ISO15099] calculated without temperature difference between inside and outside, however. IGU U IGU g [-] 0.55 0.56 0.48 0.40 0.51 0.54 low- [-] 0.02 0.02 0.08 0.03 0.16 0.04 Venetian, light colour Venetian, light beige Venetian, light colour Venetian, light colour Venetian, light grey Venetian, silver grey Blind type System UCW
[W/(m K)]
2

System type

System geff, [-] 0.09 0.02 0.11 0.02

Vision area [%] 90 90 90 85 42 80

[W/(m K)]

ext. nat. ext. nat. extract mech. extract mech. pressure equalised unventilated
1) 2)

1.1 1.0 1.4 1.4 1.91) 1.2

1.5 1.2 1.7

0.18 0.02 0.17 0.02 0.16 0.02 0.124) 0.02

1.33) 1.5 1.1

Box type window with two single panes, low- hard coating on pos. 3. * 2 Design extract flow rate 8.5 l/(s m), taking extract air into account for U-value calculation gives U CW 0.89 W/(m K). 3) Design extract flow rate 14 l/(s m). 4) Measurement result.

Proceedings

Contribution Page 7/8

Glass in Buildings II

University of Bath

April 7/8, 2005

RESULTS AND DISCUSSION System U-values of 1.1 W/(m K) and g -values of below 0.1 can be achieved using IGUs with modern glazing coatings. Highly glazed thermally comfortable buildings without excessive CO 2 emission values due to heating/cooling loads should be feasible. However, realisation of such buildings will usually require close cooperation between HVAC, building automation and faade contractor; preferably from a very early phase of the building design process. It is shown, that even unventilated double faades can achieve very good performance in regard to solar control. This has become possible with modern high quality coatings for glass products. A literature review shows that deciding on a static double faade in most cases can not be based on energy efficiency considerations. Other design aspects will be paramount. As faade contractor, we must assume that the early design process has taken all aspects of the building as a system into account. As long as we are not involved at an earlier stage of the design process, we can do no more than design an actual faade that complies with specifications given. In actual fact, though, we could do a lot more, given the chance.
2

REFERENCES [Gertis99] Gertis, K.: Sind neuere Fassadenentwicklungen bauphysikalisch sinnvoll? Teil 2: Glas-Doppelfassaden (GDF), Bauphysik 21 (1999), Heft 2, p. 54-66. [Schulz04] Schulz, M.: Leben im Schwitzkasten, DER SPIEGEL 47/2004. [Saelens03] Saelens, D. and Carmeliet, J. and Hens, H.; Energy Performance Assessment of Multiple Skin Facades; International Journal of HVAC&R Research, vol. 9, nr. 2, pp. 167-186. [ISO10077-1] ISO/DIS 10077-1:2004-05; Thermal performance of windows, doors and shutters Calculation of thermal transmittance Part 1: General. [ISO10077-2] ISO 10077-2:2003-10; Thermal performance of windows, doors and shutters Calculation of thermal transmittance Part 2: Numerical method for frames. [prEN13947] prEN 13947:2004-11; Thermal performance of curtain walling Calculation of thermal transmittance Simplified method. [EN6946] EN ISO 6946:1996; Building components and building elements Thermal resistance and thermal transmittance Calculation method. [EN410] EN 410:1998-12; Glass in building Determination of luminous and solar characteristics of glazing. [ISO15099] ISO/FDIS 15099:2003; Thermal performance of windows, doors and shading devices Detailed calculations. [IGDB] International Glazing Database (IGDB); Version 14.0 release date: 23 October 2004.

Proceedings

Contribution Page 8/8