Measurement uncertainty in energy monitoring : present state of the art

Abstract

Measurement uncertainty is a key component in the overall uncertainty calculation for Measurement and Verification (M&V) projects. However, in some cases, it is reduced to outlier detection or basic uncertainty propagation calculations. In other cases, funds are spent on determining uncertainties that have little effect on project decisions. Therefore a need exists for a fuller treatment of the subject in the light of literature from M&V and other fields. This paper surveys general M&V literature, as well as relevant research from metrology, electrical engineering, economics, decision analysis, and statistics. Electrical metering and sub-metering uncertainty is investigated, as well as often-overlooked considerations such as power quality and the cost of calibration. The effect of mismeasurement on energy models and practical techniques for mitigating such effects are assessed. Last, research on building simulation and project decisions in the light of measurement error is surveyed. Bayesian methods are found to be a recurring theme in much of the research being conducted on all of these aspects. Power quality and mismeasurement effects have also been found to make a material difference in project evaluation. The survey is concluded with recommendations for further research in the light of current trends in data analysis and energy evaluation.