A scalable framework for impact analysis of faults on indoor CO2 in buildings
Tian, Jinzhao Bie, Haipei Keene, Kevin Loftness, Vivian Zhang, Kun Gilligan, Brian King, David
Tian, Jinzhao
Bie, Haipei
Keene, Kevin
Loftness, Vivian
Zhang, Kun
Gilligan, Brian
King, David
Supervisor
Department
Machine Learning
Embargo End Date
Type
Journal article
Date
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Fault impact analysis is essential for translating fault detection metrics into actionable maintenance priorities by identifying which faults most strongly degrade HVAC performance. Prior studies have assessed energy penalties through simulation; however, empirical evidence linking operational faults with indoor air quality remains limited. This study addresses this gap by analyzing more than 90 million rows of data from building automation systems (BASs) across 299 air handling units (AHUs) in 23 U.S. commercial buildings and 335 zones in 4 buildings to estimate the impacts of operational faults on indoor CO2 levels. The analysis quantifies the effects of 23 fault types on indoor CO2, including 19 AHU-level faults and 4 terminal unit faults. The statistically significant findings reveal that “low-static-pressure” faults and “unstable outdoor-air damper” faults yield the largest increases in AHU return air (RA) CO2. On the other hand, economizer and simultaneous cooling-or-heating faults that hold the “outdoor air (OA) damper excessively open” reduce CO2 due to unintended high outdoor air intake. Among terminal unit faults, “airflow lower than setpoint” produces the greatest increases in zone CO2. These results provide large-scale empirical evidence of the impacts of operational HVAC faults on indoor air quality and offer clear guidance for prioritizing fault diagnostics and maintenance actions in commercial buildings.
Citation
J. Tian, H. Bie, K. Keene, V. Loftness, K. Zhang, B. Gilligan , et al., "A scalable framework for impact analysis of faults on indoor CO2 in buildings," Building and Environment, vol. 295, pp. 114443-114443, 2026, https://doi.org/10.1016/j.buildenv.2026.114443.
Source
Building and Environment
Conference
Keywords
33 Built Environment and Design, 3301 Architecture
Subjects
Source
Publisher
Elsevier