Uncertainty Quantification of Occupancy Sensor Errors on Energy Savings and Thermal Comfort in Residential Homes

Document Type

Conference Proceeding

Publication Date

1-1-2026

Abstract

Smart thermostats have gained widespread popularity due to their affordable cost, easy installation and maintenance, and significant potential for energy savings. However, prior studies have identified significant accuracy limitations in the presence sensing of commercially available smart thermostats, leading to thermal discomfort from false negative (FN) errors and diminished energy savings from false positive (FP) errors. This study quantifies the impacts of occupancy sensing accuracy in smart thermostats on energy and comfort outcomes through large-scale EnergyPlus simulations, incorporating Monte Carlo Markov Chain (MCMC)-based models of occupant behavior and sensor error, both informed by field data. The simulation results indicate an HVAC energy-saving ratio of 18.27% with a standard deviation of 0.74% when occupancy sensing is adopted. In addition, the FP ratio is found to be inversely related to energy saving, especially during high thermal demand. Besides, the FN errors did not show immediate impact on thermal comfort; however, its frequency is identified to be a critical factor for discomfort intensity, with unmet temperatures exceeding 3° C (5.4°F) being observed in extended durations. This study provides a design reference for occupancy sensing systems in smart building applications, supporting the optimization of trade-offs between energy efficiency and occupant comfort.

Publication Source (Journal or Book title)

ASHRAE Transactions

First Page

230

Last Page

238

This document is currently not available here.

Share

COinS