Paper Conference

Proceedings of eSim 2022: 12th Conference of IBPSA-Canada


AHU Doctor: An inverse model-based software platform for commissioning controls hardware and sequences in VAV AHU systems

Burak Gunay 1, Darwish Darwazeh 1, Narges Torabi 1, Scott Shillinglaw 2
1 Carleton University
2 National Research Council

Abstract: Faults in variable air volume (VAV) air handling unit (AHU) system control hardware and software detrimentally affect indoor environmental quality and energy performance. Existing fault detection and diagnostics (FDD) platforms employ simplistic expert rules, offer limited visualization capabilities to their end-users, and do not utilize state-of-the-art FDD approaches. This paper presents an inverse model-based FDD tool for commissioning controls hardware and sequences in VAV AHU systems. The software tool inputs widely available building automation system trend data types to train inverse models characterizing the heat and air mass balance of a VAV AHU system. The models are then used to detect nine common controls hardware and sequence logic faults. The platform is demonstrated with data from a 69-zone VAV AHU system in Ottawa, Canada. Five of the nine fault categories are detected and confirmed to be present in the case study building.
Keywords: Fault detection, VAV AHU systems, Building commissioning, HVAC systems