Experimental data analysis

The measured variables over the course of the trial are presented in Figure 5 and 6 The estimates of mean and standard deviation of the recorded data are summarised іг Table 3. The measured data show that indoor air temperature is not homogeneous and present moderate vertical and north-south gradient. The differences observed betweei them are always less than 1 °C. It has to be notice that the indoor air temperature dati could be influenced by the wall surface temperatures, because the air temperature data are close to a ‘resultant temperature’, which takes values in between the actual mean radiant and indoor air temperatures. The indoor air temperature considered for analysis i the average of the air temperature measured by the sensors located near the north and the south walls at 1.5 m height.

SHAPE * MERGEFORMAT

5 4

3 2 1 0 -1 -2 -3

0 100 200 300 400 500

Time (Hour)

Figure 6: Measured data: Indoor air Ta. Heat fluxes through different walls.

The air temperature standard deviation is reduced from the outdoor 5 °C to the indoor

1.6 °C, outdoor the mean air temperature is 18.8 °C while indoor it increases around 1 °C. Mean air relativity humidity decreases indoor 5 % and it standard deviation is reduced from the outdoor 21.3 % to the indoor 6.3 %. Wall flux data show that ceiling heat flux is smaller comparing to the others walls, this is due to the ceiling sensor location on the vertical of the polystyrene beam. Also it has been observed that main wind direction is E-W.

Table 3: Mean and standard deviation of the recorded data.

Mean

Standard deviation

Outdoor air temperature (°C)

18,8

5,1

Outdoor air relativity humidity (%)

55,8

21,3

Wind velocity (m/s)

3,0

1,9

Global horizontal solar flux (W/m2)

284,3

352,5

Diffuse solar flux (W/m2)

80,4

107,6

Air Ta near southwall 50 cm height (°C)

20,0

1,7

Air Ta near southwall 150 cm height (°C)

20,1

1,6

Air Ta near southwall 250 cm height (°C)

20,0

1,6

Air Ta near northwall 150 cm height (°C)

20,6

1,6

Indoor air relativity humidity (%)

50,4

6,3

Heat flux south wall (W/m2K)

0,5

1,3

Heat flux north wall (W/m2K)

0,5

1,2

Heat flux west wall (W/m2K)

0,5

1,5

Heat flux ceiling (W/m2K)

-0,1

0,5

The spectral analysis of the main outdoor variables allows to identify the frequency ranges over which the building are mainly excited.

The normalised cumulative spectra of two main variables affecting the system thermal performance are presented in Figure 7: the outdoor air temperature and the solar global horizontal radiation. The conclusions from their analysis are:

Outdoor temperature: 97% of the variance is concentrated over the frequency range [0, 1/10 h-1]. It exhibits a clear 24 h periodicity, as well as secondary spectral peaks (variance concentration) at 1/12 h-1 and 1/6 h-1 frequencies.

Solar radiation: 94% of the variance is concentrated over the frequency range [0, 1/11 h — 1]. As in the previous case, it exhibits a clear 24 h periodicity and a secondary spectral peak at 1/12 h-1 frequency.

The normalised cumulative spectra of indoor variables describing the building response: air temperature and heat fluxes through different surfaces are presented in Figure 7. This spectral analysis shows that the building acts as a low-pass filter. 97% of the variance of the indoor temperatures is distributed over the frequency range [0, 1/24 h1], it is 24 h harmonic and presents a secondary spectral peak at 1/12 h-1 frequency.

All the heat flux time series present a clear 24 h periodicity as well as spectral peaks at 1/12 h"1 and 1/6 h"1 frequencies. South, north and west walls heat flux present similar spectral density. 97% of their variance is concentrated over the frequency range [0,1/7 h-1]. While ceiling heat flux presents significant spectral power over the whole spectrum, 95% of the variance is concentrated over the frequency range [0, 1 h-1].