Monitoring slurry-loop reactors for early detection of hydrodynamic instabilities

Abstract Various regimes can be encountered in solid–liquid flows. In slurry-loop systems, such as an industrial polyethylene reactor, instabilities can occur at high solids concentration. An early warning method based on high-frequency pressure measurements is presented that can detect these instabilities in an early stage, using a monitoring method based on attractor comparison. With this method, deviations from the preferred process conditions are detected using a statistical quantifier of the changes of the system's attractor. In this way, process control actions can be taken to prevent unwanted instabilities to occur, which offers opportunities to operate closer to an undesirable flow regime. An early warning procedure is developed and validated, using both a scaled version of an industrial polyethylene reactor, and full-scale operational data.