Refining building energy modeling through aggregate analysis and probabilistic methods associated with occupant presence

The building sector represents the largest energy consumer among the United States' end use sectors. As a result, the public and private sector will continue to place great emphasis on designing energy efficient buildings that minimize operating costs while maintaining a healthy environment for its occupants. Creating design-phase building energy models can facilitate the process of selecting life-cycle appropriate design strategies aimed at maximizing building energy efficiency. The primary objective of this research study is to gain greater insight into likely causes of variation between energy predictions derived from building energy models and building energy performance during post-occupancy. Identifying sources of error can be used to improve future modeling efforts that can potentially lead to greater accuracy and better decisions made during the building's design phase. My research approach is to develop a method for conducting retrospective analysis of building energy models in the areas that affect the building's predicted and actual energy consumption. This entails collecting pre-construction and post-occupancy related data from various entities that exhibit influence on the building's energy performance. The method is then applied to recently-constructed military dormitory buildings that utilized building energy modeling and now have actual, metered building energy consumption data. The study also examines how building occupancy impacts energy performance. The value of this work will provide additional insight to future building energy modeling efforts.