Validation of dynamic RADIANCE-based daylight simulations for a test office with external blinds

The study encompasses the validation of the dynamic, RADIANCE-based daylight simulation method DAYSIM, which uses the concept of daylight coefficients and the Perez sky model to predict the short-time-step development of indoor illuminances. Measured and simulated illuminances have been compared under 10,097 sky conditions in a full-scale test office with a double glazing and external venetian blinds. The additional planning effort for the designer compared to a conventional daylight simulation is addressed. It has been found that the treatment of direct sunlight strongly influences the accuracy of the daylight coefficient method. Three different simulation modes for the direct sunlight are investigated. The simulation results prove that indoor illuminances can be modeled with comparable accuracy for various blind settings under arbitrary sky conditions. Daylight autonomies are predicted with an accuracy below 2% points, where simulation errors stem with roughly equal parts from the raytracing and the sky model.