A Model-Based Fault Detection and Diagnosis Methodology for Cooling Tower

This paper presents a model-based method for detecting and diagnosing some faults in the cooling tower of healing, ventilating, and air-conditioning systems. A simple model for the cooling tower is employed. Faults in cooling tower operation are detected through the deviations in the values of system characteristic parameters such as the heat transfer coefficient-area product, the tower approach, the tower effectiveness, and fan power. Three distinct faults are considered: cooling tower inlet water temperature sensor fault, cooling tower pump fault, and cooling tower fan fault. As a result, most values of the system characteristics parameter variations due to a fault are much higher or lower than the values without faults. This allows the faults in a cooling tower to be detected easily using above methods. The diagnostic rules for the faults were also developed through investigating the changes in the different parameter due to each faults.