Principal components analysis was performed using process data from a waste-to-energy plant (WtE), together with Hotelling’s T2 statistics and Q statistics, to evaluate the effect of maintenance routines on the performance of different units. The results showed that shower cleaning of evaporator units improves their efficiency while decreasing the heat exchanged in superheater and economizer. Q statistic detected the effects of maintenance earlier than Hotelling’s T2 statistics. The analysis could indicate which of the cleanings were less effective. Therefore, the proposed method can contribute to the design of a more effective cleaning strategy to prevent failures due to particle deposition. All in all, the fault detection and diagnosis proposed in this work can provide operational improvements, increased efficiency and reliability in WtE plants.
Waste-to-EnergyPrincipal component analysisfault detectionfault diagnosisplant maintenance