The purpose of this research is to develop effective methods of noncapital improvements in the operation of existing large scale data-rich heating, ventilation, and air conditioning (HVAC) systems, through the use of system data commonly routinely logged by building automation systems. Using the Iowa State University campus system as a test bed, we develop a systematic hierarchical method of HVAC system analysis and the setting of monitoring schemes. Our approach takes account of commonalities of location, use, subsystem identity, equipment type, etc. provided by basic engineering knowledge in order to monitor for signals of abnormal/degraded system behavior. It takes account of systematic temporal effects like those of season, use patterns that vary with day of the week, diurnal effects related to weather and sunlight, etc. In the end, our methodology becomes a battery of multivariate control charts, carefully thought out and matched to the application.
This work has huge potential economic and environmental impacts. Short term savings are potentially available by reducing energy consumption through utilizing smarter control. Long term savings come through monitoring, which helps identify newly developing areas of energy waste and system problems, allowing for quick correction. Preventative maintenance opportunities arise from detection of evolving system inconsistencies. The reductions in energy consumption and waste potentially resulting from this methodology support a sustainable future.
Kisch, Wendy, "Logic and methodology for hierarchical monitoring as a means of fault detection in existing large scale, data rich HVAC systems" (2014). Mechanical Engineering White Papers. 1.