Identification of Heat Transfer Parameters in Fixed-Beds

Abstract

In a fixed-bed chemical reactor, the heat transfer parameters dictate the behaviour of the reactor and thus their accurate determination is vital to the successful design and effective control of this complex, highly nonlinear distributed parameter system (DPS) which will ensure stable operation while mainaining high performance. Optimal performance depends on the quality of information regarding the state of the system which requires estimation of conditions other than those where the measurement probes are placed. The best sensor location is not fixed but will in general vary spatially in time because of the nature of external disturbances A simple but effective scheme is developed to achieve a robust system in terms of a fixed finite number of sensor locations for parameter identification. This requires the development of an effective model reduction technique based on moving finite elements and orthogonal collocation which approximate the DPS. Using this, a low order model is obtained which can be used as a flexible and versatile parameter estimator to identify the critical thermal parameters of the bed and guarantee that stable conditions are always maintained and temperature runaway does not occur