Ptimization of process parameters based on data mining for urea plant

Zhang Yuntao,Zhang Mingrong,Zheng Huaijiang,Fu Weizhong
DOI: https://doi.org/10.3969/j.issn.1001-4160.2009.08.012
2009-01-01
Abstract:A prediction equation of the urea plant production was set up about the relationship between the response value(conversion percentage of carbon dioxide(η%)) and the variables selected from the main factors by stepwise regression.The selected variables in the equations,including the temperature TR_0102 at the bottom of reactor,ammonia to carbon dioxide ratio N/C and water to carbon dioxide ratio H/C at the inlet of urea synthesis tower,CO_2%and Ur%in tower export.Regression equation and selected variables all passed the significant test.The correlation coefficient R of the equation to predict the test set is 0.9991.Based on the double variable analysis and the statistic classification of raw data,the range of variables of optimized production conditions are given.After the Simulation and test,taking the program of optimized process parameters,the N/C and H/C was controlled in the 3.8~4.0 and the 0.70~0.85,to implement the production test,respectively.With the test results device Ur%(outlet) increasing by 1.56,NH_3%,CO_2%and H_2O%in tower export decreasing by 3.35.The optimized target of improving the synthesize percentage and the energy-saving emission reduction was achieved.
What problem does this paper attempt to address?