MODELLING AND CONTROL OF A PACKED BED REACTOR (DISTRIBUTED PARAMETER, STATE ESTIMATION, METHANOL OXIDATION, FORMALDEHYDE, OPTIMIZATION)

A detailed mathematical model of a packed bed reactor is developed and utilized in devising algorithms for state estimation, parameter estimation, control, and optimization. The specific case considered is an exothermic wall-cooled reactor for the gas-phase partial oxidation of methanol to formaldehyde over an Fe-Mo oxide catalyst. Reactor selectivity is important due to an undesirable consecutive reaction producing carbon monoxide. The ideas presented here are tested by closed-loop simulation of the model-based control system, and by experiments with a computer-controlled laboratory reactor.;Several levels of mathematical models involving one and two dimensions as well as pseudohomogeneous and heterogeneous formulations are used to predict the steady state and dynamic behavior of the reactor. A detailed comparison is made between different models. Orthogonal collocation enables solving the equations faster than real time using realistic parameters. Parametric sensitivity is examined for several heat and mass transfer parameters, kinetic parameters, and operating conditions. Model parameters, including the catalyst activity profile, are optimized to the fit data over a wide range of experimental conditions.;A two-dimensional, nonlinear, distributed parameter optimal state estimation algorithm is developed in order to estimate the temperature and composition profiles from a limited number of measurements. The performance sensitivity to designer-selected parameters is investigated. State estimation during start-up of the reactor, changes in the inlet composition or flowrate, and unmeasured disturbances in the wall or inlet temperature is accomplished in the present of poor initial conditions, erroneous parameters, and/or model mismatch with the experimental system.;The estimated temperature profiles and a 2-D heterogeneous reactor model are the basis for estimation of the catalyst activity profile, control of the maximum temperature in the reactor, and optimization of the outlet stream composition. Manipulation of the wall temperature gives excellent control over the estimated maximum temperature. Efficient reactor optimization is accomplished at various flowrates, in the presence of erroneous modelling parameters and changes in the objective function, and with flow as an additional manipulated variable.