Analysis Methodology for Large Organizations' Investments in Energy Retrofit of Buildings

Abstract

This paper presents a formal methodology that supports large organizations' investments in energy retrofit of buildings. The methodology is a scalable modeling approach based on normative models and Bayesian calibration. Normative models are a light- weight quasi-steady state energy models, which makes them scalable to large sets of buildings due to highly enhanced modeling efficiency. Then, Bayesian approach calibrates normative models such that calibrated models quantify uncertainty in the model while representing a building as operated. Calibrated models can further incorporate additional uncertainty from ECMs, and provide information about underperforming risks of ECMs. This paper illustrates the proposed retrofit analysis process through a case study, and demonstrates its feasibility to support large-scale retrofit decisions under uncertainty in the context of the ESCO industry.