Low-cost energy meter calibration method for measurement and verification

Abstract

Energy meters need to be calibrated for use in Measurement and Verification (M&V) projects. However, calibration can be prohibitively expensive and affect project feasibility negatively. This study presents a novel low-cost in-situ meter data calibration technique using a relatively low accuracy commercial energy meter as a calibrator. Calibration is achieved by combining two machine learning tools: the SIMulation EXtrapolation (SIMEX) Measurement Error Model and Bayesian regression. The model is trained or calibrated on half-hourly building energy data for 24 hours. Measurements are then compared to the true values over the following months to verify the method. Results show that the hybrid method significantly improves parameter estimates and goodness of fit when compared to Ordinary Least Squares regression or standard SIMEX. This study also addresses the effect of mismeasurement in energy monitoring, and implements a powerful technique for mitigating the bias that arises because of it. Meters calibrated by the technique presented have adequate accuracy for most M&V applications, at a significantly lower cost.

Bayesian calibration of building energy models for uncertainty analysis through test cells monitoring

“…Again, this tend to be for energy consumption [73] and utility data analysis [89] rather than thermal conditions. Moreover, most of the mentioned studies assess Bayesian calibration in commercial buildings such as offices [88] and university buildings [90], while a small number of them present residential buildings as calibration targets [89], again putting great efforts into analysing energy use data.…”

“…Thirdly, in this research a 24-hour training period and a 120-hour testing period were considered, which meant a larger amount of computational running time was required compared to monthly calibration (Bayesian hourly calibration for this case study took on average 12-hours processing time for each protocol). To tackle this issue and in comparison to other studies where only Bayesian approaches were implemented[90], this technique was complemented with a sensitivity analysis for identifying the most influential variables on the output results. The combination of both methods has affordably reduced computation time and has led to a considerable improvement in prediction, in contrast to uncalibrated results.…”

Deep Highway Networks and Tree-Based Ensemble for Predicting Short-Term Building Energy Consumption

“…A violin plot is similar to box-and-whisker plot, a box plot indicates variability outside the upper and lower quartiles with a box and whiskers. In a violin plot, the full probability density function in a mirrored form is presented on a vertical axis [55]. The white circle in the middle of the plot indicates median, and the upper and lower ends of the box inside the violin plot indicates the quartiles.…”

Efficient longitudinal population survival survey sampling for the measurement and verification of lighting retrofit projects

“…There are three kinds of uncertainty that need to be mitigated for an M&V savings estimate to be accurate. These are measurement uncertainty [4,5], sampling uncertainty, and modelling uncertainty [6][7][8]. Of these, sampling uncertainty is often the dominant component [9,10].…”