A Method for Developing Energy Budgets and Energy Design Guidelines for Institutional Buildings.

This study is directed at a more complete underst and ing of energy consumption in buildings and , by extension, further possibilities for energy conservation. The approach is to search for major factors, relating to buildings as they are now constructed, occupied and operated, which appear to explain the variations in annual energy consumption among them. These factors may differ in kind, order of importance, magnitude or interpretation from those conventionally employed when designing buildings. Thus new or enhanced insight will be gained into reduction of the operating energy. The present study applied the method of multiple regression analysis to historical data about existing institutional buildings in the State of Michigan. Factors which have some impact on building energy consumption were isolated and examined in reasonable detail. Study indicated some complex interactions between these factors. However, a final regression model was constructed from a set of factors possessing both strong impact and relatively low intercorrelations. This model, consisting of 10 major factors, has an ability to explain up to 94 percent of the energy consumption variation among buildings, with the st and ard error for the prediction of about (+OR-)14 percent of the mean energy consumption. The factors which are included in the model are variables that related to: exhaust air quantity, opaque wall area, percent of air conditioning and cooling degree day, type of fuel used, hours of air equipment used, type of refrigeration, type of air h and ling system, shading of buildings in winter, shading of buildings in summer, and CFM per horse power of supply fan. Tests were conducted on the performance of the predictive model and the validity of the selected factors. The results of these tests support the choices embodied in the final model. Finally, the approach used in this study has merit for the underst and ing of energy consumption in buildings, for prediction, evaluation and budgeting of energy consumption and probably for other energy investigations of different focus with respect to building. However, a larger, more diverse data base as well as further developmental studies are needed to fully realize the demonstrated potential of the regression method.PhDArchitectureUniversity of Michiganhttp://deepblue.lib.umich.edu/bitstream/2027.42/158780/1/8214946.pd