On The Optimal Design And Control Of Dividing Wall Column

In this paper,we summarize the research meaning,the research principle and the research progress of energy saving of distillation process.We investigate the steady state design and dynamic control Petlyuk column,Kaibel column and reactive dividing wall column.At first,we study the most simple dividing wall column configuration,three-product Petlyuk column.Petlyuk column contains prefractionator and the main column.There are two pairs of thermally coupled vapor and liquid streams connecting prefractionator and the main column.Usually,Petlyuk column is used for three-component feed mixture.Petlyuk column is a fully thermally coupled distillation column.A fully thermally coupled distillation column is thermodynamically equivalent to a dividing wall column with the dividing wall in the middle of the column.,if the heat transfer between the dividing wall is ignored.We design and optimize the steady state configuration of Petlyuk column;then we study the stochastic optimization of Petlyuk column;then we study the stablizing contol(pure temperature control)of Petlyuk column;at last,we compare the temperature control and the temperature –composition cascade control of Petlyuk column.Results show that traditional PID control,pure temperature control(three-point temperature control)without composition control,is able to handle well the feed disturbances added to the system,and the deviations of product compositions are very small.We then study the more complex Kaibel column configuration.Usually,Kaibel column is used for four-component feed mixture.We design and optimize the steady state configuration of Kaibel column;then we several suitable control structures for Kaibel column.Results show that traditional PID control,pure temperature control(four-point temperature control)without composition control,is able to handle well the feed disturbances added to the system,and the deviations of product compositions are very small.These results are good for the industrialization of DWC.For reactive dividing wall column(RDWC),we select an industrial example.We combines the selective hydrogenation reaction with the propylene distillation column,and integrates two distillation columns into one thermally coupled reactive distillation column.At first,we study the design and optimization of RDWC(using particle swarm optimization);then we investigate the traditional PID control of RDWC;at last,we study the advanced intelligent control(model predictive control)of RDWC.In this paper,the RDWC is a partially thermally coupled distillation column.The dividing wall is at the bottom of the column.We not only propose a new process RDWC with 27.88% total annual cost saving,but also propose several suitable control structures for the new process(including temperature control,composition control and model predictive control).In this paper,we summarize the main research results of this paper and we make prospects of the design and control of DWC at last.The innovation points of this paper are as follows:(1).For three-component Petlyuk column and four-component Kaibel column,the results of rigorous simulation prove that traditional PID control can handle the feed disturbances inserted into the columns.(2).For three-component Petlyuk column,the stochastic optimization algorithm contains both support vector machine(SVM)and particle swarm optimization(PSO).For reactive dividing wall column(RDWC),the stochastic optimization algorithm is particle swarm optimization.(3).For RDWC for the selective hydrogenation and separation of a C3 Stream in an ethylene plant,model predictive control(MPC)and traditional PID control are studied.Results show that MPC is more suitable for multi-component,complex and interactive RDWC.