“Investigation of the Prevalence of Faults in the Heating, Ventilation, and Air-Conditioning Systems of Commercial Buildings”
Architectural Engineering PhD Defense by Amir Ebrahimifakhar
9:30 am –
10:30 am
Peter Kiewit Institute Room: 250
Contact:
Kelly Johnson, (402) 554-5935, kelly.johnson@unl.edu
Under the supervision of Dr. David Yuill
Abstract: This dissertation presents the study for a large-scale investigation of heating, ventilation, and air-conditioning (HVAC) fault prevalence in commercial buildings in the United States. A multi-year dataset with thousands of HVAC equipment including air handling units, air terminal units, and packaged rooftop units is analyzed to determine a range of HVAC fault prevalence metrics. The primary source of data for this study comes from three commercial fault detection and diagnostics (FDD) providers. Fault data from each provider are converted to a standard format, which is called binary daily fault data. Since each FDD provider uses different fault names to refer to the same fault in an HVAC system, mapping functions are created for each FDD provider to convert their fault reports to a unifying taxonomy.
Since the commercial FDD software outputs are inherently subject to some level of error, i.e., they could have false negatives and false positives, a field study is conducted to verify the commercial FDD software results. Two buildings from among the buildings of one of the FDD providers are selected. Packaged rooftop units of these two buildings are monitored for about two weeks using our installed data loggers. Using our FDD methods, the actual faults in these buildings are identified. The results of our field study are compared with the FDD provider fault reports to find the false negatives and false positives.
This study also proposes and demonstrates a data-driven FDD strategy for packaged rooftop units using statistical machine learning classification methods. The FDD task is formulated as a multi-class classification problem. Seven typical rooftop unit faults are discriminated against one another as well as the normal condition. Nine well-known classification methods are applied to our dataset, and their performance is compared. The performance of the classification methods for individual faults is also characterized using true positive rate and false positive rate statistical measures. The relative importance of input variables is also discussed.
Abstract: This dissertation presents the study for a large-scale investigation of heating, ventilation, and air-conditioning (HVAC) fault prevalence in commercial buildings in the United States. A multi-year dataset with thousands of HVAC equipment including air handling units, air terminal units, and packaged rooftop units is analyzed to determine a range of HVAC fault prevalence metrics. The primary source of data for this study comes from three commercial fault detection and diagnostics (FDD) providers. Fault data from each provider are converted to a standard format, which is called binary daily fault data. Since each FDD provider uses different fault names to refer to the same fault in an HVAC system, mapping functions are created for each FDD provider to convert their fault reports to a unifying taxonomy.
Since the commercial FDD software outputs are inherently subject to some level of error, i.e., they could have false negatives and false positives, a field study is conducted to verify the commercial FDD software results. Two buildings from among the buildings of one of the FDD providers are selected. Packaged rooftop units of these two buildings are monitored for about two weeks using our installed data loggers. Using our FDD methods, the actual faults in these buildings are identified. The results of our field study are compared with the FDD provider fault reports to find the false negatives and false positives.
This study also proposes and demonstrates a data-driven FDD strategy for packaged rooftop units using statistical machine learning classification methods. The FDD task is formulated as a multi-class classification problem. Seven typical rooftop unit faults are discriminated against one another as well as the normal condition. Nine well-known classification methods are applied to our dataset, and their performance is compared. The performance of the classification methods for individual faults is also characterized using true positive rate and false positive rate statistical measures. The relative importance of input variables is also discussed.