Development of a grinding simulator and calibration by the reconciled data of Zarshuran gold processing pilot plant

Document Type : research - paper

Authors

Abstract

Steady-State simulation of grinding circuit is an efficient tool for achieving reasonable estimates of the process and improving its performance. In this study, along with coding the grinding simulator for a closed circuit and providing applicable sub-routins for calculation of breakage and selection functions and etc in MATLAB and creating a user friendly interface, it was attempted to calibrate the models of this software using sampling data of Zarshuran gold processing pilot plant. Models’ Calibration was done using raw, average and also reconciled data using bilinear method and in each stage, simulation results and measured data were compared. Results of simulations showed that selecting valid models for the units and using reconciled data in model calibration process, leads to calculating true parameters for selected models and consequently, resulting in better simulations. Maximum error reduced from 153.16 percent in water flowrate of hydrocyclone underflow calculated using raw data down to 4.57 percent while reconciled data were used. In addition, it was shown that repetitive sampling of every sapmling point and averaging the data could reduce the errors to some extent.

Keywords

Main Subjects