Analysis of Wet High-Intensity Magnetic Separation of Low-Grade Indian Iron Ore using Statistical Technique

Abstract

The particle separation in a wet high-intensity magnetic separator depends on a number of variables. Applied magnetic field, particle size, and wash water rate play a vital role among them. Influences of these important variables were studied in detail following 3(3) full factorial designs of experiments using a laboratory/pilot-scale wet high-intensity magnetic separator (Gaustec G-340 Minimag). Statistical analysis of the results indicated that the influence of all these variables are significant (95% confidence level) on the recovery of magnetics and the order of significance follows, particle size > magnetic field > wash water rate. The experimental results used to develop the regression models to predict yield and grade at unknown operating conditions in the study range. Investigations carried out on a wet high-intensity magnetic separator using a low-grade iron ore sample containing 49.27% Fe shows that it is possible to upgrade to 62% Fe in the concentrate with poor yield values. A "tree" procedure (generally used to evaluate the flotation performance) was followed to evaluate maximum possible yield and grade by the magnetic separator. The "tree" procedure results provide useful information about the magnetic separation competence.