An Accurate Prediction of Daylight Pipe Harvesting of Interior Space

Abstract

The performance of a light guide under arbitrary meteorological conditions is difficult to predict numerically because none of the present methods can account for the random configuration of clouds. Although, partly cloudy skies typically occur very frequently in many regions over the globe, the effects of cloud sizes, distributions, or altitudes on the luminous effectiveness of a light guide remain virtually unknown. The physical installation of light guides does not allow for a controlled experiment in which a single parameter can be varied within a defined range, while holding all other parameters constant. Numerical modeling is the only practical way to provide deep insights on the role of a specific property, such as cloud coverage, and its importance relative to all others (e.g., atmospheric turbidity). We have developed a unique solution through linking the UniSky simulator and HOLIGILM (hollow light guide interior illumination method) tool, while providing for the first time a powerful method that can provide accurate predictions and understanding of the fundamental differences of light guide behaviors under homogeneous and inhomogeneous sky states. The use of homogeneous skies in low-accuracy empirical models often results in overestimated/underestimated effectiveness of some light-guide systems. The model we have developed can predict complex optical signatures that are normally impossible to reproduce using up-to date empirical models (e.g., the illuminance patterns due to isolated clouds or clouds traversing across the sky). The optical efficiency, average cosine and working-place illuminance are analyzed and compared in order to demonstrate the model capabilities.