R. A.Stephenson, R.A., S. J. Healy, S.J.

Bayer Process Precipitation circuits can undergo periodic variations in key operating and process parameters. Product size distribution in particular, can undergo significant change, and has implications for product quality and productivity. Causes of instability can be external to the circuit’s process, such as changes in incoming pregnant liquor flow and quality, equipment performance changes and failures, mistuned plant controllers or non-ideal operator intervention, or from internal changes to process conditions. One example of an internal cause of circuit instability is a change in the oxalate related nucleation rate, destabilising circuit and product sizing. Over a period which is characteristic of the circuit, changes in the particle size distribution can result in gross and undesirable changes to seed and product size, demanding operational control responses. These control responses themselves may propagate instability. This work investigates the drivers of these variations using a detailed dynamic simulation circuit model. Simulations indicate that “unattended” circuits are surprisingly stable and will tend to a steady state (albeit not necessarily a desirable one). External control of the circuit using PID to achieve a particular target (such as pump-off solids concentration or overall yield) by controlling fine and coarse seed rates can lead to gain-driven instability. Interventions must necessarily account for the process sensitivities and timescales involved, if this is to be avoided. Overall, Dynamic Circuit Simulation provides key insights into the behaviour of a complex interconnected system and can be a powerful tool for optimizing process control strategies and responses, plant performance, assisting in upset recovery, and operator training.