Case Study: The Effective Use of an Extensive Logical rule Based Data Analytics Approach in Establishing Root Cause of Performance Issues in Widespread Deployments of Unitary Space Air Conditioning Units

Abstract

Today a significant percentage of office spaces are air conditioned using widely deployed unitary systems, either Fan Coil Units (FCU) or Variable Air Volume (VAV) boxes, to achieve high degrees of air conditioned zonal control. However establishing a near realtime overall control monitoring regime to ensure proper control adherence across such large estates can be difficult, or possibly unattainable, where the number of units could effectively run into the hundreds, if not the thousands. And, as is outlined in the problem definition section of this paper, compounding normal operational control issues such as system tuning and hardware failures, with inbuilt design limitations and poor building fabric issues, coupled with embedded deep rooted and undetected commissioning problems, and all functioning within a dynamic operational environment, one can see that the task of effective and timely root cause analysis can be an extremely difficult one. Having being challenged with such a problem environment, the author attempts, through presentation of a series of actual real life working examples, to offer the reader the case for use of an effective and efficient aggregated data driven analytical approach, which includes temporal and geo spatial visualisation techniques, that makes the identification of the fundamental root cause problems, and system interactions (positive and negative) within this complex operational environment analysis, possible. The presented approach therefore offers the SME community a means to detect the current underlying problems without the need for deployment of costly disruptive diagnostic procedures or preventative maintenance regimes. Such an approach also demonstrates the flexibility to cater for future problem diagnosis scenarios where the underlying logical rules built within the underlying architecture can be written, tested and deployed in a timely fashion. Finally it is contended that this targeted aggregated data driven fault diagnosis approach is equally applicable in any other situation which entails wide spread deployment of energy assets, for example, high volume deployments of store refrigeration units used within the Retail sector.