A Rational Approach To Computerized Optimization Of Mineral Processing Plants

Steady-state flowsheet simulators for mineral processing plants have evolved to a current high level of utility and accuracy. Reliable, general-purpose, semi-empirical models now exist in nearly all unit operations categories particularly in the areas of comminution and classification. .These models, installed as user-selectable functions within the integrated frame- work of flowsheet balance-solver software systems, can be strung together to represent any plant configuration. In addition, CAD-like graphics, data reconciliation and model parameterization functions have become standard features of the current generation of integrated plant simulators. These advances have given engineers the capability to investigate many different plant scenarios while requiring a minimum of actual plant testwork. This paper describes a methodology and an integrated software package which are being successfully applied in a number of operating plants worldwide to predict plant performance under changing conditions. Those conditions which can be changed include effect-of ore type, feed size, water addition rates, equipment geometries and flowsheet modifications. By using data from actual plant optimization exercises in real mills, the paper illustrates and out- lines a technique which includes plant data collection, ore breakage testing, flowsheet simulation, model-fitting (parameterization), data adjustment and model experimentation. Brief descriptions of some unit operations models, such as autogenous mills and cyclones, are provided as examples of the current state of such models. Ore breakage testing and its relation to certain models is also discussed. Computer- generated size-distribution graphs, along with tabular data generated by the simulator, round out the presentation and serve to demonstrate the real- ism and accuracy of the models. The paper concludes with a partial list of cases where this technology has been successfully applied.