These experiments are evaluated by using a thermodynamic model and the comfort criteria according to DIN 1946 [5]. Based on this evaluation, differences between the summer 2002 and the very hot summer 2003 can be evaluated for every building in order to characterise the building physics and weak points in the specific building concept.

The thermodynamic model is based on the specification in EN 832 [6] and focuses on the essential building parameters according to Keller’s approach [7]. The underlying theory can be taken from Baehr’s textbook [8]. The indoor room temperature Ti oscillates periodic around its mean value T, m with the amplitude ATi. Mean temperature and amplitude can be derived from the heat loss coefficient H, the specific heat gain Gm and its amplitude AG and the time constant т or the heat storage C of the room, respectively.

T(t) = Tm +AT’sin(®-1) with ® = 2n/t0 with t0 = cycle period

with Tm = Tam + — and AT = to -[at, +— | with C/H

l, m am H 1 2-n-z I a H )

Applications

The authors [9] have shown that this simplified model corresponds to a sophisticated building simulation and have applied this method successfully to experiments (with well — known boundary conditions) and field measurements which are associated with uncertainties.

As the thermal performance of a specific building can be discerned accurately by this method, the data model can be practically used

— for data analysis of field measurements (here: passively cooled buildings).

— for quality assurance of the building simulation during the design phase since the complex interactions calculated by the building simulation can be verified with a few concise parameters.

— for measurements during the implementation phase or during the operation to determine the building parameters for advanced control strategies, i. e. predictive controllers.

with an acceptable time effort.

The Summer of 2003

In Germany, the three summer months June, July and August 2003 were significantly too warm. The daily mean temperature was 19.6 °C, 3.4 K above the reference temperature. This extreme summer weather can be predicted by the numerical climate models; summer conditions like those experienced in 2003 should occur statistically every 1,000 years [10], even if the anthropogenic global warming is taken into account. In particular regions (e. g. Freiburg), the mean ambient air temperature was 5 K higher than the reference year.

If the summer 2003 was used as design weather for passive cooling concepts, (almost) every building would need air-conditioning to meet the comfort criteria. Though the summer 2003 is not suitable for design studies, its impact on the design of mechanical and passive cooling systems is discussed intensely.