Intelligent monitoring and selective cleaning control of deposits in pulverised coal boilers (Clean selective) : final report

Within the EU-funded project 'Clean selective', software and hardware sensors have been developed and used to monitor the ash deposition in the Teruel plant. As hardware sensors in the convective section, strain gauges were applied to the hanger rods of a superheater for the first time. Heat flux sensors were used in the furnace area to observe deposition evolution. Additionally, 3D-CFD calculations were performed for the boiler and results were used as input for the Ash Deposition Predictor developed, which monitors the ash deposition on the surfaces as maps. The sensor signals are used to decide when and where soot-blowers should be activated to effectively clean the heat exchangers. In the furnace section, heat flux sensor signals do not indicate the effectiveness of soot-blowing on deposition removal. Therefore, neural network models have been developed to decide on the activation of a soot-blower. These models are trained on historical data of soot-blowing in the furnace. In addition, the heat flux signals are compared to the depositions maps. In the convective section, the hardware sensor indicates deposition evolution without the location on the observed superheater. Therefore, historical data have been used to identify areas of heavy deposition, which are compared to the maps from the ash deposition calculations. Thus a predictive tool for the detection of ash deposition in a boiler has been developed combining hardware and software sensors. This system has been successfully tested for a two month period, where an efficiency rise could be predicted.