A prediction of blend grinding result based on individual ore size measurements and grinding tests. (January 2023)

• Record Type: Journal Article
• Authors: de Jong, Tako.P.R.
• Is Part Of: Minerals engineering. Volume 191(2023)
• Publication Date: 2023-01
• Subjects: Ball mill -- Grinding -- Iron ore -- Ore blend -- Pelletizing -- Population balance; Mines and mineral resources -- Periodicals; Ressources minérales -- Périodiques; Mines and mineral resources; Periodicals; Electronic journals; 622
• DOI: 10.1016/j.mineng.2022.107950
• Languages: English
• ISSNs: 0892-6875

Abstract

Graphical abstract: Highlights: Laboratory grinding tests based on laser diffraction analysis. Grinding parameters of iron ores for population mass balance calculations. Individual ore data predict mill test results of binary ore blends. Simulated test results predict plant ground ore quality from input blends. Abstract: Dry ball mills of Tata Steel IJmuiden's pelletizing plant grind a blend of iron ores, coke breeze and mineral additives. Blend composition can be adjusted depending on raw material supply and product demand. A grinding prediction method on blends has been developed in order to aid improvement of blend selection. A batch testing procedure on individual ores was applied. Parameters describing grinding kinetics were determined by using simulation and fitting to the results of the tests. The grinding of certain ore mixtures was simulated using the obtained parameters of the individual ores. The calculated particle size distributions were compared to laboratory grinding results of the mixtures. Grinding parameters of each ore were subsequently summarised in a single parameter, being the time required in a laboratory mill to achieve a specified target average particle diameter. The incorporation of the grinding time into linear models improved the prediction of product quality from the industrial grinding system. Results and mean average error of this prediction were determined by means of cross validation.