The simulation model has been used to evaluate how the building performs compared to a reference building close to the regional common practice and under identical operational and environmental conditions.

The reference building considered is identical to the real building except for the following passive techniques: shading devices are removed, double glazed windows are replaced for single glazed windows, the ceiling polystyrene vault is removed, walls insulation are reduced 1cm. The reference building accomplishes the obligated normative, Spanish Standard for Thermal Conditions in Buildings, 1979. The design is close to the

26 24

22 20 18 16

72 172 272 372 472 572

Time (hours)

Figure 11: Simulated indoor air temperature for the reference ad the real building.

common construction practice on the region.

The predicted indoor air temperature for the reference and the real building is showed in Figure 11. The mean day maximum temperature is for the reference building 22.5 °C while for the real building this is 20.8 °C.

2

1,5 1

0,5 0

72 172 272 372 472 572

Time (hour)

Figure 12: Values of the difference between the indoor air temperature simulated for the real building and the indoor air temperature simulated for the reference building.

The difference mean is 1.05 °C and its standard deviation is 0.45 °C. It present a clear 24 h periodicity, differences are higher during the day and lower during nights. In addition differences are lower during the colder days than during warmer ones. It indicates that the implemented techniques have stronger influence during warm periods. This is reasonably as the implemented techniques were conceived for and arid Mediterranean climate, with extreme summer conditions and longer summer periods than the winter ones.

Conclusions

The thermal performance of an energy-efficient building in South Spain has been analysed by means of measurements and simulation, during three weeks of Spring.

Experimental results show that the indoor air temperature standard deviation is reduced from the outdoor 5.1 °C to the indoor 1.6 °C, while the mean air temperature augments 1°C the 18.8 °C outdoor air temperature. The air relative humidity standard deviation is reduced from the outdoor 21.3 % to the indoor 6.3 % The spectrum analysis of the available data shows that the system is mainly excited over the frequency range [0,1/10 h" 1] and that the building acts as a low pass filter.

The model has been created using the TRNSYS simulation code and the whether data recorded. Usual modelling hypothesis have been adopted. Differences between the measured and simulated indoor air temperature (residuals) mean is 0.37 °C and standard deviation is 0.45 °C. Residuals are not stationary and present trends which follows outside temperature trends, residuals are lower during the colder days. Besides residual spectrum analysis shows that main disagreements between measurements and simulations are observed at the frequencies in which the system is mainly excited.

The simulated thermal behaviour of the reference building in comparison to that one of the real building shows that the techniques under analysis present stronger influence during the day than during the night. The day maximum temperature mean is for the reference building 22.5 °C while for the real building this is 20.8 °C. As well they present stronger influence during warm days than during cold days, this is reasonable since the strategies were conceived for a sub arid Mediterranean climate.

This preliminary results will be used in further analysis on the building energy performance.