Predicting and interpreting uncertainty propagation in separation circuits using functional unit evaluation

Abstract

The estimation and analysis of uncertainty propagation in separation circuits is an important, but challenging, aspect of a comprehensive circuit design strategy. Owing to the sophisticated and complex modeling requirements, many of the current separation circuit design tools rely on deterministic models, despite the ubiquity of uncertainty in the model input parameters. Additionally, many of the uncertainty assessment protocols, which rely on mathematical solutions, provide few insights on how particular circuit modifications will influence uncertainty propagation. As a supplement to these methods, the current paper describes a method of assessing and evaluating separation circuit uncertainty by integrating the fundamentals of linear circuit analysis, the law of propagation of error, and the separation functional unit evaluation technique. This analysis shows how the addition or modification of units within circuits, based on separation functional unit definitions, can be used to predict how uncertainty propagation will change, even with limited a priori data. The capabilities of this innovative approach are illustrated and critically evaluated through a systematic separation experimental study using the roll-type corona electrostatic concentrator for separating titanium from silica within a heavy mineral sand feed.