Research On Combustion State Monitoring And Optimization Strategy For Power Station Boiler

With the increasing competition from the power market, the shortage of energy, and the severe requirement on environment protection, it is urgent for the thermal power units to improve the operation level of the boiler combustion system, and to meet the requirements on energy saving and emission reduction. This dissertation studies the state monitoring and the performance assessment of the boiler combustion in thermal power units via different data-driven methods, and the best combination mode and load dispatch of the air and fuel for the boiler from the viewpoints of stability, economy and environment are obtained based on the proposed methods. The result can be used as guidance for optimal operation.The main contributions of this dissertation are summarized as follows:1) By extensive selection and comparison, the flame intensity, furnace pressure and steam heat-parameters are used to realize the state monitoring of the combustion stability. It is found that the flame intensity and furnace pressure are sensitive to the change of combustion state, and they can be used as real-time monitor of the combustion stability. The steam heat-parameters can reflect the quality of stability, which accords with the economy performance at the same operating point.2) To remedy the deficiency of the previous methods, a comprehensive index to evaluate the combustion stability performance is proposed as the evaluation index of combustion stability, which is computed from the steam heat quality using the PCA (Principal Components Analysis) method. The result can not only be helpful to search the best combustion process for optimal operation, but also be used as one assess-index for improving the operation level of the engineers.3) The excess-air coefficient is selected as the main monitoring state of combustion economy, and the boiler efficiency is selected as the assess index. To compute the excess-air coefficient and boiler efficiency accurately, the grey-box modeling method is used in the soft sensing of oxygen content in flue gases. The results show that the grey-box model has clearer physical meaning and more exact forecast precision than other identification models such as neural network.4) The NOx (nitrogen oxides) emission is used as the monitor state and evaluation index of combustion environment. The stage combustion methods are discussed for the NOx emission reduction, which is air-stage and fuel-stage. Some rules of air supply and mill combination mode is found by analyzing the history data in a 600MW boiler.5) Taking the NO.3 boiler of Datang Panshan Power Plant as an example, this dissertation obtain the best mode of air supply and mill combination by statistical analysis, and mine the best rules of the load dispatch, with the best combustion performance of stability, economy and environment as separate objective, The result shows that this scheme for combustion optimization is significant in real application.