A probabilistic approach to the energy-saving potential of natural ventilation: Effect of approximation method for approaching wind velocity

The approaching wind velocity is an important input parameter in wind-driven ventilation studies. The approaching wind velocity at a certain height must be approximated from the measured meteorological wind velocity because the wind data in weather data files are usually measured at a meteorological station at only one given height. The most commonly used approximation method is the power law, in which the power law index (PLI) is only one parameter. Its typical values are provided in some data sources, the use of which generally involves several assumptions. An important assumption is that it is constant. We conducted observations of the wind velocity profiles above a high-density area in Tokyo, Japan, using a Doppler LIDAR system. Our observations revealed that the PLI value is not a constant but rather a time-dependent variable. In ventilation studies, neglecting this phenomenon possibly limits the usefulness of the results because the ventilation airflows depend on the PLI value in given building characteristics and given weather data. This paper presents a probabilistic simulation model for evaluating the performance of natural ventilation, which takes the uncertainties in the PLI values into account. The simulation results provide a set of possible values for savings in cooling energy that are possible with natural ventilation. Different choices of PLI values can lead to errors of up to 45% in the estimation of potential savings in cooling energy when using natural ventilation. This study can support building designers and engineers in the reasonable design of naturally ventilated buildings.