Research On Combustion Optimization Of CFB Boiler Blended With Petroleum Coke And The Prediction Of The Wear Of Heating Surface In The Boiler

Petroleum coke is a by-product of the oil refining process.Compared with coal,it has a higher carbon content,heating value and lower ash content.CFB boilers have been widely used due to their wide fuel adaptability and high combustion efficiency.The mixed combustion of coal and petroleum coke in CFB boilers can not only rationally utilize petroleum coke,but also achieve the effect of waste treatment.At present,there is a lack of research on the combustion optimization and wear prediction of CFB boiler blended petroleum coke.This paper takes the blended combustion of petroleum coke in a 410t/h CFB boiler of a Sinopec plant as the research object.Through BP neural network and genetic algorithm,a GA-BP prediction model is established for the combustion system of the CFB boiler,and the conditions of different coal and petroleum coke blending ratios are established.The fly ash carbon content and NOx emissions were predicted.The optimization goal was the linear combination of fly ash carbon content and NOx emissions.The genetic algorithm was used to optimize the results.The research results were carried out in this CFB boiler.The test results show that the GA-BP combustion prediction model can accurately predict the results.In the actual operating conditions of the optimized boiler,the fly ash carbon content and NOx emissions under different blending ratios have been reduced.FLUENT software was used to numerically simulate the combustion process of the CFB boiler with different primary and secondary air ratios.The results show that increasing the primary air and reducing the secondary air will help the stable combustion of the CFB boiler.The increase of the mixing ratio and the excessively high temperature in some areas of the boiler will affect the safe operation of the CFB boiler.In actual operation,a large proportion(>40%)of mixed burning should be avoided.Finally,based on the results of numerical simulation in FLUENT and the wear mathematical model,the boiler is divided into three parts: upper,middle and lower parts.GA-BP neural network is used in Matlab to establish wear prediction for different areas of the CFB boiler Model.The results show that the model has a good generalization ability and can better predict boiler wear.Then,according to the established wear prediction model and control variable method,the influence of air distribution plate air volume,primary air volume,secondary air volume,fuel volume,and mixing ratio in the dense phase area on the wear of the CFB boiler boiler is analyzed.The results show that: The increase of the air volume of the air plate and the fuel volume will increase the overall wear of the boiler,while the increase of the primary air volume and the secondary air volume in the dense phase area will cause the "cutting" effect on the fluidized air and the wear is reduced,and the increase in the mixing ratio leads to a decrease in the amount of wear in the boiler.