AN ADAPTIVE INFERENTIAL MEASUREMENT STRATEGY FOR ONLINE MONITORING OFCONVERSION IN POLYMERIZATION PROCESSES

Much research effort has been devoted to the development of sensors ca pable of providing reliable on-line information about reaction advance ment in polymerization processes. In particular, calorimetric techniqu es which have been widely investigated in the past, still present real difficulties in multipurpose applications. This article describes a n ew strategy for the accurate determination of conversion during batch polymerizations. In addition to the usual calorimetric data, infrequen tly-available gravimetric measurements are used to track variations of key parameters, such as the overall heat transfer coefficient. It is shown that accurate estimation of conversion can be obtained through t he design of an adaptive state-observer, even if unpredictable convers ion-dependent and/or batch-to-batch variations of the system an encoun tered. Batch and semi-batch experiments were performed in a 7-litre be nch-scale well-mixed reactor to evaluate the technique, The method has been successfully applied both to solution polymerization system, i.e . vinyl acetate/butyl acrylate in ethyl acetate, and to emulsion copol ymerization operations, i.e. styrene/butyl acrylate in water.