Development of a fast and accurate annual daylight simulation approach for complex window systems
Open Access
- Author:
- Yoon, Younju
- Graduate Program:
- Architectural Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 11, 2006
- Committee Members:
- Martin Moeck, Committee Chair/Co-Chair
Richard George Mistrick, Committee Member
William P Bahnfleth, Committee Member
John Michael Cimbala, Committee Member - Keywords:
- Annual photosensor performance modeling
RADIANCE
Annual daylight simulation
Daylight Coefficient Approach
Annual daylight simulation for blind and complex w - Abstract:
- The introduction of daylight through window systems to building interiors has the potential to save electric lighting energy and heating and cooling loads, but improper selection/design of window systems can negate the benefits of electric lighting energy reduction by increasing space conditioning requirements. Therefore, combined lighting and building thermal energy simulations based on hourly local weather data must be performed to estimate accurate total building energy use. However, computation of annual daylight availability, which provides annual electric lighting energy use, requires long computation time, and there are no annual daylight simulation tools capable of simulating annual performance of complex daylight systems with reasonably short computation time. The main hypothesis of this research is that a fast and accurate annual daylight simulation approach for spaces with daylight systems and exterior obstructions can be performed. To compute the illuminance contribution from the sky, a daylight coefficient approach and sky matching methods using the exterior vertical to horizontal illuminance ratio, illuminance vectors, and illuminance ratios were studied. The daylight coefficient approach provides more accurate simulation results and requires less computation time than the sky matching methods. To compute the illuminance contribution from the sun, a daylight coefficient approach, a sun matching method, and a modified sun matching method were investigated. The modified sun matching method provides the most accurate sun illuminance results, but is the slowest method. The daylight coefficient approach provides slightly less accurate results than the modified sun matching method, but requires the least computation time among the three approaches studied. Annual daylight simulation results are ultimately used to compute annual electric lighting energy use. Therefore, the accuracy in the prediction of electric lighting energy use for the fastest and the most accurate annual simulation approaches were investigated. Both methods provide comparable accuracy to the standard RADIANCE hour by hour approach. In this thesis, the daylight coefficient approaches for the sky and the sun were also successfully applied to simulate the performance of the photosensor that exhibits a directional sensitivity distribution. The simulation of the hourly photosensor performance for an entire year permits an evaluation of realistically accurate hourly electric lighting energy use for a given lighting control system, type, location, and orientation of the photosensor. The proposed annual daylight simulation method consumes 0.0001~0.0002 times the computation time required for the standard RADIANCE rtrace method for illuminance sensor modeling and 0.004~0.006 times the computation time required for the RADIANCE psens method for photosensor modeling.