PNNL

Abstract

Understanding the impact of residential building mechanical systems on indoor air quality requires monitoring air handler unit (AHU) function, which can be expensive and time consuming. This paper considers the ability of low-cost acoustic sensors to predict home HVAC on-off state and runtimes and compares its logged power level measurements on a dedicated AHU circuit. This exploration is demonstrated in a full-scale laboratory home designed for energy performance testing. A classification algorithm applied to both the power and noise data saw 96.7% agreement, suggesting that overall level can predict AHU state for two-state systems with careful sensor placement.