Refinery expansions or a need to process lower grade bauxites will lead to larger mud loads through Clarification circuits. This can be handled either by building more process equipment or increasing the loads on existing equipment.
This work was centred on the latter alternative assessing the equipment capability and possible process risk of planned increases in average slurry flowrates and mud loadings of some forty percent. This was done before the event using proprietary statistical multivariate analysis software to model future performance based on targeted historical data correlating multiple key performance parameters.
Focusing primarily on modelling for future increased equipment loadings, this paper reviews the purpose of the work, the software methodology, its capability to produce correlated results quickly and reliably from extensive data, the analysis outcomes and also possible future work. Based on Key Performance Indicators that define Settler performance it was shown that conditions for higher rates should not differ from historical standard Settler operating conditions, and therefore they are unlikely to cause poor Settler performance.
With the methodology and philosophy of the work established, there is a brief review of trending work comparing performance of different flocculents in full-scale flocculent field trials undertaken over two years by Worsley Alumina to improve Settler performance at existing flow conditions. Multivariate statistical analysis modelling of existing operations demonstrated changes in underlying processing modes over the two years that are not evident from the data using simpler trending. Further data analysis is being undertaken to better establish the business case.
As more data becomes available from the new operating regimes it will be used with the Knowledge Extraction and Benefit Estimation components of the software to optimise overall operation of the units. It is also proposed to replicate the models developed in other areas of Clarification and the Refinery