Characterization of a building's operation using automation data: A review and case study

- Methods to characterize building operation from automation data were reviewed.
- Common building faults and diagnostic approaches were identified from the literature.
- Based on the survey findings, an inverse greybox model-based method was put forward.
- Strengths and weaknesses of this approach were demonstrated through a case study.
- The case study was conducted by using the archived automation data from an office building.

This paper presents a critical review of the automated on-going commissioning (AOGC) methods for air-handling units (AHU) and variable air volume terminal (VAV) units in commercial buildings. The common faults studied in the literature were identified. The diagnostic approaches taken and the characteristics of the fault-symptom datasets utilized were categorized. It was found that the diagnostics methods were vastly fragmented, and most of them employed pure-statistical approaches. Only a few studies attempted to assimilate the automation data within the underlying physical processes. In addition, a large fraction of the reviewed literature has been devoted to physical faults in AHUs. Only a few studies were conducted to diagnose faults-related with controls programming and faults at the zone level. Upon the literature survey findings, an inverse greybox modelling-based AOGC approach was put forward. Its strengths and weaknesses were demonstrated through a case study conducted using the archived building automation system (BAS) data of an office building in Ottawa, Canada. The results of this case study indicate that inverse greybox modelling-based AOGC is a promising method to diagnose both physical and controls programming related faults at AHUs and VAVs.