A Framework for Benchmarking the Total Energy Efficiency of Net Zero Energy Buildings

Abstract

The need for net zero energy buildings (NZEB) is a result of global depletion of natural fuel resources and the effects of the usage of these fuel resources on climate. For Instance, the building sector in the US alone contributes to 40% of energy use and greenhouse gas emissions as opposed to 32% in the industrial sector and 28% in the transportation sector. This large percentage reflects the need for reducing the energy consumed by buildings and using renewable energy resources. In essence, NZEB utilize renewable energy resources to offset the amount of energy used in a building over the course of a year. In the literature, there are various definitions for NZEB that relate to the energy used and the costs and emissions from that energy. Also, there is no unified benchmarking approach that can be used to compare NZEB. In this thesis, a framework for benchmarking the total energy efficiency of NZEB is introduced which indicates an overall reduction in the energy costs and emissions and utility energy use of NZEB. The total energy efficiency was computed using a network of two data envelopment analysis (DEA) models. The first DEA model benchmarks the energy efficiency of the building based on uncontrollable weather and functional factors. The second DEA model benchmarks the renewable energy system of the buildings based on the energy costs and emissions and utility energy used. Through combing both efficiencies in one benchmark, the total energy efficiency was able to indicate reductions in energy, costs and emissions as it indicated buildings with highest ratios of outputs to inputs that were included in the models. It is important to note that the models were applied on commercial office buildings and from the application; a framework was developed to benchmark all types of NZEB. It is recommended that the framework is applied to the benchmarked NZEB rather than the results of the models as the data used contained simulated data.