|Ph.D Thesis||Department of Chemical Engineering|
|Supervisors:||Prof. Sheintuch Moshe|
|Prof. Lewin Daniel|
|Full Thesis text|
Fixed bed catalytic reactors play a key role in many chemical processes. These units should be optimal in their design. Moreover, such reactors may exhibit extremely complex spatio-temporal dynamics including patterned states like thermal fronts and periodic or aperiodic temperature oscillations. Thus, such reactors are of significant practical and academic interests.
The first part of the thesis deals with loop reactor design and control for exothermic reversible reaction. A loop reactor (LR) is an N-units system composed as a loop with gradually shifted inlet/outlet ports. Loop reactors may compete with other heat recuperating technologies, like reverse flow reactors, for catalytic abatement of low-concentration volatile organic compounds (VOC). The need for recuperating energy in VOC combustion stems from the low adiabatic temperature rise in combustion of lean wastes and consequently the inability of sustaining a reaction even in a catalytic adiabatic process.
The aim of control part of this work is the design of a robust algorithm that stabilizes the performance of the Loop Reactor. The design utilises mathematical
models of reaction front propagation in a fixed catalytic bed and rotating pulse in Loop Reactor. The design of a control algorithm for Loop Reactor is divided into a number of steps. First, a homogeneous one-dimensional model of fixed bed was adopted from Sheintuch and Nekhamkina (2005). The model was extended to reversible reactions and ported to a Matlab/Simulink environment.
At the second step the dynamics of fixed bed was studied for both reversible and irreversible reactions. The control design uses thermal front location as the
measured variable and the switch velocity as the control variable. To implement the control system an observer algorithm was created to track the location of the thermal front. As a result a new reliable approach to Loop Reactor control method was developed in the presented research. This part was published in “Industrial &Engineering Chemistry Research” (Sheinman and Sheintuch, 2009).
This research proved possibility of controller synthesis for reactor networks using accurate analysis of underlying mathematical models. The research achieved robust control over Loop Reactor. Extensive and comprehensive mathematical modeling made possible to understand and fine tune control algorithm. This method of control design provides simpler algorithm in comparison to existing control methods.
Moreover, it provides analytical tools for quantifying robustness of the designed controller and controller parameter estimation using physical properties of the system.