Three reference wind speeds of 6.4 m/s, 9.0 m/s and 11.1 m/s were applied. The reference height for wind speed is the building height. For those measurements, the building model was not present. The spires located upwind in the test section of the wind tunnel and the roughness elements in the test section were positioned to achieve the target average velocity and turbulence vertical profiles at the planned building location. These profiles correspond to an Atmospheric Boundary Layer (ABL) flow with a slightly rough terrain and the Reynolds number based on the building size ranges from 1.9x105 to 3.3x105.

Details about the ABL flow characterization can also be found here:

http://dx.doi.org/10.1016/j.jweia.2023.105442

To verify ventilation airflow rates prior to testing in the wind tunnel, the indoor-outdoor pressure difference was measured at the centre of each building facade. For those measurements, the building model was outside the wind tunnel. The indoor pressure was calibrated at -40 Pa compared to outdoor and the extracted flow rate was set at 100 m3/h to achieve the design total ventilation airflow rate.

This experiment aims at analysing the effect of wind conditions on the re-introduced pollutant concentration. Results are expressed as the percentage of pollutant concentration measured in the pipe upwind the exhaust fan normalized by the pollutant concentration released at the stack. The pollutant release flow rate at the stack is controlled by mass flow controllers and a calibration procedure was carried out before each experiment to verify the accuracy of the tracer gas measurement by means of a known ethylene source. The background pollutant concentration in the approach flow was measured simultaneously to be subtracted from the measurements in the pipe upwind the exhaust fan. After reaching the quasi-steady state, the concentrations were averaged over 180 s. The air temperature and the absolute atmospheric pressure were also tracked in order to compute the real air density.

This experiment was conducted for each of the 13 reference wind directions and each of the 3 reference wind speeds. A non-dimensional velocity (M) was introduced as the ratio between the release velocity at the stack and the reference wind velocity, leading to M values ranging from 1.0 to 1.7. Four measurements have been replicated 3 times to estimate the reproducibility error.

Based on Experiment 3 data, the two most critical scenarios were chosen for detailed measurements of the internal pollutant concentration, namely for wind directions of 0° and 45° at M=1.7. Measurements were conducted on a uniform 3x3x3 grid, with an averaging of 180 s after reaching the quasi-steady state. The background pollutant concentration measurement was maintained and the internal mapping of the pollutant concentration was realized by means of a probe mounted to a three-axis traverse system installed inside the wind tunnel. The air temperature and the absolute atmospheric pressure were also tracked in order to compute the real air density.

Based on Experiment 3 data, the most critical scenario was chosen for detailed measurements of the pollutant concentration at ventilation inlets, namely for the 45° wind direction at M=1.7. Measurements were conducted at the centre of each of the 40 air inlets, with an averaging of 180 s after reaching the quasi-steady state. The background pollutant concentration measurement was maintained and the pollutant concentration measurement was realized at each air inlet by means of a probe mounted to a three-axis traverse system installed inside the wind tunnel. The air temperature and the absolute atmospheric pressure were also tracked in order to compute the real air density.