Climate adaptive energy efficiency modeling using a generalized additive approach to optimize building performance across Chinese climate zones

Abstract

Abstract Accurate measurement and verification (M&V) of energy efficiency measures (EEM) in commercial buildings is a key requirement to improve energy performance and meet sustainability goals. Research suggests a new method to M&V EEM using generalized additive models (GAM) to provide a way to measure how EEMs perform across different commercial buildings (i.e., offices, mixed-use developments, and healthcare). Comparisons suggest GAM is a preferred method of predicting energy savings from previous years and provides good estimates on a new dataset (comparable to previous years). The CV(RMSE) value is acceptably low. Lighting upgrades and HVAC improvements are areas of best practice for energy savings, and all sectors studied achieved significant energy savings with reasonable return times on investment compared to all other studies conducted to date (examples include offices and healthcare). We also focus on and show climate-related factors affecting energy consumption and had some success differentiating results based primarily on temperature/RH relative humidity-triggered variables and indicated the primary “thresholds” that appeared to alter energy demand behavior. Particular high humidity and temperatures carry serious energy penalties, and future climate change calls for climate-responsive energy policies. Furthermore, Monte Carlo simulations were used to measure uncertainty and backlog of data readings, not all prompted by climate factors alone, to confirm our results were sound.

Keywords

• Generalized additive models
• Measurement and verification
• Energy efficiency measures
• Building energy performance
• Monte Carlo simulation