Uncertainty Analysis And Stochastic Optimization Of Operating Parameters Of Coal Water Slurry Gasifier

The global optimization of a gasifier is difficult to be achieved due to the operation uncertainties.Therefore,a reduced-order model for SE coal water slurry(CWS)gasifier of Sinopec Zhenhai Refining&Chemical Co.,Ltd.was developed.The uncertainty of the gasifier operating parameters was analyzed and the optimization method of operating parameters was explored.According to the gasifier flow field characteristics and chemical reaction characteristics,the gasifier was divided into zones,and the reduced-order model was developed in Unisim,which was verified to be reliable using industrial data.On this basis,the single-factor sensitivity of operating parameters was carried out analyze.The results showed that the gasifier temperature was more sensitive to the changes of coal slurry flow rate and oxygen flow rate,and the effective gas production was more sensitive to the changes of coal slurry flow rate and coal slurry concentration.The uncertainty of the gasifier operating parameters was analyzed by Monte Carlo method,and the multi-factor sensitivity of the gasifier operating parameters was tested using Spearman rank correlation coefficient method.Taking the effective gas production as the optimization goal,the control parameter optimization was carried out based on the stochastic optimization method.Our results showed that the uncertainty of CWS flowrate,oxygen flowrate,CWS concentration was 0.55%,0.06%,and 0.67%,respectively.The consequent uncertainty of effective syngas production and operation temperature was 1.33%and 1.20%.The multi-factor sensitivity analysis showed that the output parameters were strongly influenced by the uncertainty of the CWS flowrate,while slightly influenced by the uncertainty of the oxygen flowrate.Within 95%confidence interval,the optimized effective syngas production was obtained at the oxygen coal ratio of 449.43～457.65 Nm3/m3 in the conditions that the temperature is in 1190.8～1237.5℃.In order to solve the problem of high computational cost of operating parameter optimization based on the reactor network model,a gasifier operating parameter optimization algorithm based on Kriging surrogate model combined with particle swarm optimization was proposed.The results showed that under certain conditions,the optimal solution based on the surrogate model optimization can converge to the optimal solution of the reactor network model.After considering the uncertainty,the maximum relative error of the optimization result was only 0.21%,and the calculation time is greatly reduced,which made it possible for the optimization process considering the uncertainty to meet the requirements of real-time optimization calculation speed.The research results in this paper provide theoretical support for the optimality of gasifier operating parameters.