DETECTING AND AVOIDING UNSTABLE OPERATION OF AUTOTHERMAL REACTORS

A control system is developed that avoids unstable operation of autothermal chemical reactors. The exothermal reaction between hydrogen and oxygen occurs on catalyst in two fixed-bed chemical reactors coupled to an external feed-product heat exchanger. An optimization searches for values of process setpoints that maintain an acceptable, specified degree of stability. Two setpoints are adjusted to track feed changes.;Process stability is inferred from the eigenvalues of a physically-based mathematical model of the reactor-exchanger system. The search for setpoint values is performed by an optimization procedure with a steady-state objective function. In this work, the objective function was fashioned to drive the setpoints toward unstable conditions. The search is constrained by stability constraints and process constraints. The stability constraints require the eigenvalues to lie to the left of a specified piecewise-linear curve in the complex plane.;The mathematical models of the reactor beds resulted in state-space models that were unwieldy for real-time computations in their original form. Novel application of the Taylor-dispersion approximation to the original partial differential equations results in a reduction of the number of state variables from 14 to 3-6.;Tests on a laboratory autothermal reactor show that changing the process setpoints influences the stability of the process. The model is able to give information about stability provided the model closely represents the steady-state temperature profiles of the reactor system. A standard nonlinear optimization algorithm is able to find setpoint values that satisfy the stability and process constraints under changing conditions.