Fig. 1 compares the room and the ambient air temperature in / at the four low-energy office buildings. The cumulative duration curves show how often the temperature exceeds a given limit. However, these curves cannot be used to evaluate the comfort since the comfort depends on the ambient air temperature. The indoor/outdoor temperature graphs, which are independent of the actual weather, evaluate the comfort.

— The room temperature in the Fraunhofer ISE building is too high. While the room temperature was slightly too high in 2002, the room temperature did not meet the comfort range during the summer 2003 (temporarily 3 — 4 K too warm).

— The Pollmeier (low heat gains) and the Lamparter building (earth-to-air heat exchanger) provided comfortable room temperatures in summer 2002. In 2003, the room temperature was 1 — 2 K too high.

The graphs take only working hours into account. The following statements s1 — 4 can be derived from an extensive data analysis based on these measurements and are discussed in this paper:

(s1) Self evident, if the weather (and the other boundary conditions) do not change, the thermal performance of a building remains also unchanged.

(s2) Obviously, if the weather changes, the thermal behaviour of a building (here: degree hours over 25°C) will change.

(s3) If the chronology of climate situations (here: periods of warm weather) changes, the thermal performance of a building will change.

(s4) In both years, the buildings did not meet the comfort criteria strictly. The increased failure to meet the comfort standard in 2003 can be explained by using smaller time constants t than in 2002. That means that the heat storage capacity was completely utilised due to the longer cycle periods of warm weather.

Fig. 1: Cumulative duration curves of room and ambient air temperatures and indoor- versus-outdoor graphs with comfort criteria according to DIN 1946.

A "summer day” is defined by a daily maximum temperature of 25 °C or higher. Furthermore, the frequency of room temperatures above 25 °C is often used as a simplified comfort criteria for passively cooled buildings. For these reasons, this temperature limit is used as a consistent standard for comparison. The graphs for Fraunhofer ISE, Pollmeier and Lamparter show generally the same behaviour: Each building exceeded 25 °C more often during the warm summer 2003 than during the typical summer 2002, which is characterised by the degree hours Dh in Table 1.

The difference of the degree hours approximately corresponds to the area between the 2002-line and the 2003-line for the ambient and the room air temperature in Fig. 1. The ratio of the two areas indicates how the building can compensate for high ambient air temperatures: The smaller the ratio, the less the ambient air temperature affects the room temperature. The Fraunhofer ISE building (98 %) has coped with the summer 2003 worse than the Pollmeier (52 %) or the Lamparter building (63 %). This characteristic building behaviour is discussed — together with the energy balance of the buildings — in detail in the results analysis.