Experimental investigation and three-dimensional computational fluid-dynamics modeling of the flash-converting furnace shaft: Part I. Experimental observation of copper converting reactions in terms of converting rate, converting quality, changes in particle size, morphology, and mineralogy

Abstract

An experimental investigation was conducted to elucidate the main features of the processes taking place in the shaft of a continuous flash-converting furnace for solid copper mattes. The experiments were conducted in a large laboratory furnace. The test variables included the matte grade, oxygen content in the process gas, particle size of the feed material, and oxygen-to-matte ratio. The observed variables included the fractional completion of the oxidation reactions, fraction of sulfur remaining in the particles, copper-to-iron atomic ratio, particle-size distribution, morphology, and mineralogy of the reacted particles. The experiments showed substantial differences in the oxidation behavior of high-grade (72 pct Cu) and low-grade (58 pct Cu) matte particles. Low-grade matte particles reacted evenly throughout the furnace, increased in size, and experienced no substantial fragmentation during oxidation. High-grade matte particles tended to be oxidized unevenly and experienced severe fragmentation leading to generation of dust. The order of the effects of the test variables on the observed variables was found to be (1) the oxygen-to-matte ratio, (2) the particle size of the feed material, and (3) the oxygen content in the process gas. Microscopic examination revealed that the oxides of copper and iron were the main oxidation products, with little elemental copper present in the reacted particles. Potential implications of the experimental findings on the operation of an industrial flash-converting furnace are discussed.

Evolution of Size and Chemical Composition of Copper Concentrate Particles Oxidized Under Simulated Flash Smelting Conditions

“…Jokilaakso et al [11] made similar observations regarding the fraction 28 to 38 lm of chalcopyrite concentrate. During oxidation tests with high-grade copper matte particles, Perez-Tello et al [21] observed that particles of fraction <37 lm tended to increase their size upon oxidation, whereas large particles decreased their size with substantial generation of dust. In a previous study with La Caridad copper concentrate particles, Thomas et al [16] noted that particles of fraction 25 to 38 lm were less reactive than those of fraction 45 to 53 lm.…”

“…[2] The elucidation of the fundamental aspects governing the behavior of the flash smelting reactor may help optimize its operation in terms of energy requirements, the quality of the molten products obtained, and its potential impact on the environment, among other issues. Over the last decades, extensive experimentation has been conducted on the flash smelting [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17] and flash converting [18][19][20][21] processes by means of stagnant-and laminar-flow reactors. As a result, the reaction path originally proposed by Kim and Themelis [10] and further modified by Jokilaakso et al [11] and Yli-Pentila et al [20] has been widely accepted as a descriptive model which explains most of the experimental observations made under controlled laboratory conditions.…”

“…As a result, the reaction path originally proposed by Kim and Themelis [10] and further modified by Jokilaakso et al [11] and Yli-Pentila et al [20] has been widely accepted as a descriptive model which explains most of the experimental observations made under controlled laboratory conditions. The model is shown schematically in Figure 1 and is discussed in detail in the literature, [11,20,21] thus only a brief description is provided here. Upon entering the reactor, the initial particles are first heated up by the gas phase by interphase heat transfer and the reactor walls by radiation.…”

“…The morphology of most reacted particles was spherical. Pe´rez-Tello et al [21] reported on oxidation experiments with high-grade copper matte particles (72 pct Cu) in a large laboratory furnace. The authors found that the finest particles in the feed (< 37 lm) tended to increase their mean size upon oxidation and decrease the amount of particles with sizes less than 20 lm.…”

“…Adapted from Refs. [11] and [21]. collision followed by agglomeration of particles may occur depending on the local values of particle number density and turbulent intensity in the industrial reactor.…”