Prediction of discomfort glare from windows under tropical skies

Abstract

Discomfort glare is common problem in modern buildings that employ highly-glazed façades. Glare not only has negative impacts on occupant comfort but can also increase energy use in buildings. Yet a robust discomfort glare metric, particularly with regards to daylight, has proven elusive. The Unified Glare Probability (UGP), developed using 493 surveys under clear skies in Brisbane, Australia, shows promise as useful tool for estimating glare from windows. This investigation attempts to validate the UGP using a similarly large number of surveys under different sky conditions, different building types and demographics in the tropical climate of Kuala Lumpur, Malaysia.The study uses a similar methodology to the initial investigation of the UGP, using a combination of luminance mapping with a post-occupancy evaluation (POE) questionnaire on discomfort glare. A total of 341 surveys were collected from six buildings; three green-rated and three regular office buildings. Hypothesis testing of the luminance data collected in these buildings indicate that the ratio of window to background luminances are a more sensitive measure of occupant discomfort than ratios of window to task luminances.The results for the UGP show agreement in predicting discomfort from windows between both data sets using regression analysis. A simple hypothesis test showed discomfort was accurately predicted by the old UGP model for 69% of new surveys. The agreement between the two data sets for window glare enabled them to be combined into one massive data set of 813 surveys to update the UGP. The updated metric uses a logistic curve in place of the original linear transformation of the Unified Glare Rating (UGR) to prevent illogical values of probability being obtained from the metric.