Dynamic Simulation Of The Fully Thermal Coupling Distillation

Distillation system is a multi-variable coupled systems, which have multipleinput variables and multiple output variables.Its problems about controlling is verycomplicated. For full thermal coupling distillation, due to the inverse circulation flow,making the rectification process is nonlinear and coupling between the operationvariables is more serious. In order to achieve saving energy, distillation system mustbe reasonable control.Otherwise,it will affect the normal production. In order toresolve the control problem in the process of distillation, we should have sometheoretical research of its dynamic characteristics, which has a certain practicalsignificance.Taking water, ethylene glycol and diethylene glycol ternary mixtures separationprocess as an example, discussing the dynamic characteristics of the fully thermalcoupling column separation process, the main research results are as follows:1.Adopting strict dynamic mechanism modeling method, and using Aspen Dynamics,the dynamic model of thermal coupling distillation is evaluated. When the system hasthe positive and negative disturbances in two directions, overhead and bottomproducts are sensitive to negative disturbance,which offset is larger;while sideproduct is sensitive to disturbance changes in positive direction, which offset islarge.According to the dynamic response, the system is asymmetric and severenonlinear. The fluctuation of the overhead product and the intermediate products inconcentration for10hours, while the concentration of the bottom products for5hours,system recovery from one steady state to another steady state for a long time, thesystem has strong hysteresis.2.Determining the control scheme of "three concentrations", namely the reflux flowcontrols overhead product concentration (L-x_D), and side flow controls side productconcentration (S-x_S), and using reboiler heat duty controlling product concentration(Q_R-x_B).Selecting the proportional integral (PI) control, and determining the positive and negative role of controller and controller model equation.To obtain the optimalproportional gain K_c and integral time T_i,we use Ziegler Nichols method to adjustingthe control parameters. Adding±10%of the step disturbance to the system, thechange of the top, side and bottom products concentration is larger,and the amplitudefluctuation time is10hours.Returning tank liquid level and the top pressure ofdynamic response time need5hours, and the bottom level takes15hours to returnstability. When having±1%feed composition disturbance of EG, the dynamicsimulation of the overhead product concentration fluctuation amplitude was0.5%; theside product concentration fluctuation amplitude is0.1%;the amplitude of bottomproducts concentration is0.35%, and the recovery time is6hours.Although thesystem can restore stability, the recovery time is very long.3.Using relative gain matrix, we put forward a new control scheme, namely the topreflux flow controls top level (L-L_D), and the flow at the top of the column cancontrol product concentration (D-x_D) and side flow can control side productconcentration (S-x_S), and reboiler heat duty controls bottom products concentration(Q_R-x_B), and bottom flow controls the bottom level (B-x_B), and condenser heatduty controls the top pressure of condenser (Q_C-P). When adding load±10%of thestep disturbance,overhead, the recovery time of the side and bottom productconcentration reduces significantly,which only needs3hours. The changes of the toplevel and the bottom level is more flat and the curve is less fluctuation with2hours,;after3hours,the dynamic response of the top pressure returns stability. When having±1%feed composition disturbance of EG, the biggest swings of the overheadproduct is0.2%, and the biggest swings of side product is0.175%, at the sametime,the change curve of the swings are more ideal, the time shortens to3hours.Theindicators in the time domain are more excellent, satisfying the requirement ofindustrial production.