In-situ fault detection for chillers

Abstract

Faults in chillers causes economic loss, inconvenience to building occupants and higher energy consumption.Recent interests in the field of fault detection and diagnostics (FDD) model have resulted in the development of several models. This project introduces the use of one of the many models for actual implementation in the operation of chillers with the objective of minimizing the building cost. The Gordon-Ng Universal Chiller model uses the minimum numbers of readily measurable variables. It provides insights to the physical conditions of the chiller, diagnoses the possible faults and provides a direction to the fault rectification effort.The success of implementing fault detection and diagnostics program depends on its acceptance by building managers whose works of work includes maintaining the chiller in optimum condition. Towards this end, a Microsoft Excel program was developed to simplify the fault detection and diagnostics processes and reduces the tasks of transferring the measured data and predicting the changes in chiller operation profiles.