A mineral liberation distribution estimator for monitoring and process control applications

Particle size control in mineral processing is typically achieved using an index characterising the whole distribution, e.g. the sieve opening through which a given mass fraction of the population would pass. Although mineral liberation is the goal of size reduction, its use for monitoring and process control applications is not reported because its online measuring is currently not possible. This paper proposes to address this issue using a mineral liberation distribution estimator consisting in two models. The first one uses a polynomial expression to relate the size-by-size mineralogical distribution of every mineral species to that of the ore. The second model relies on a beta distribution function to sort the masses in every size fraction into liberation classes. The characteristic constants of the beta function are parameterised with the properties of the ore. Results demonstrate that the liberation estimator can represent liberation distributions of both fresh and processed comminution products. [Display omitted] •The estimator can model liberation distributions of fresh and processed comminution products.•The grade of the analysed minerals has an impact on the precision of the model calibration.•The calibration methods allow to estimate parameter uncertainty and thus to confer reliability.•The realistic distributions are encouraging for future monitoring and control applications.