Nonlinearity Analysis And Control Of Solvent-based Post-combustion CO₂ Capture System

Greenhouse gas emission has a severe impact on global climate change.Among all the greenhouse gases,CO₂ contributes to the major effects.Power generation from coal-fired power plant is currently the largest source for CO₂ emission.In the context of energy conservation and emission reduction,carbon capture technologies have grown increasing public concern.Post-combustion carbon capture?PCC?through monoethanolamide?MEA?-based absorption is regarded as the most mature technology exist due to its low cost and rich availability.With the worldwide application and deployment of PCC technology,scientific researches in terms of dynamic modelling and control have been widely conducted.From these studies,PCC process exhibits strong nonlinearity,large time constant and strong interactions between different variables.Meanwhile,PCC process must be operated flexibly to reduce the energy penalty in solvent regeneration process.In this regard,it is of great importance to gain a thorough understanding of transient behaviors for PCC process,concurrently to propose an advanced control structure to operate PCC process in a flexible manner.In this work,there are 3 major contributions:?1?Dynamic models of MEA-based PCC process are implemented in gCCS modelling environment.In this platform,open-loop step response tests have been realized under several typical working conditions.Dynamic characteristics of key output variables to key input variables are presented and discussed.A critical nonlinearity measurement using gap metric is carried out to acquire the nonlinear distribution of PCC process in a quantitative manner.The nonlinear analysis is able to reveal the working conditions of PCC process which may exhibit significant variations.?2?Based on the nonlinear analysis,linear model predictive controller?MPC?is obtained to improve the controllability of PCC process in the presence of weak nonlinear working conditions.Flue gas flowrate is viewed as a measurable disturbance.In order to eliminate the effect of varied flue gas flowrate,feed-forward control is therefore adopted with the addition of disturbance model.In consideration of unmeasurable disturbances,this thesis proposes a disturbance rejection predictive controller based on extended state observer.Disturbance signals can be estimated and then rejected with the modification of manipulated variables.?3?In the face of strong nonlinear working conditions,a multi-model predictive controller is presented.In order to operate the PCC process in an economically optimal trace,improvement has been employed with the utilization of economic model predictive controller with a terminal constraint.The cost penalty in terms of CO₂ emission and steam extraction are considered for the construction of economic index.With the minimization of mentioned index,the optimal economic profits can be achieved.