Validating BMS sensors for chiller condition monitoring

A method based on energy balance is developed for diagnosing and validating the health states of the building management system (BMS) sensors (faults and biases) for monitoring chiller condition. A normalized energy balance residual is defined and used as the indices for fault detection. A characteristic quantity is derived for diagnosing flow meter faults without needing to know the temperature sensor biases. A robust algorithm is developed for estimating the magnitude of the flow meter biases. The sensor fault validation method, verification of the method using dynamic simulation programs and application to an existing chilling plant are presented.