Products & Services

A model for the bacterial leaching of copper sulfide ores in pilot-scale columns

A model for the bacterial leaching of copper sulfide ores in pilot-scale columns
of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  International Journal of Mineral Processing, 31 ( 1991 ) 247-264 247 Elsevier Science Publishers B.V., Amsterdam A model for the bacterial leaching of copper sulfide ores in pilot-scale columns Heinz J. Neuburg, Jorge A. Castillo, Miguel N. Herrera, Jacques V. Wiertz, Tom~s Vargas and Ricardo Badilla-Ohlbaum t Department of Chemical Engineering, Faculty of Physical and Mathematical Sciences, University of Chile, Casilla 2777, Santiago, Chile (Received February 5, 1990; accepted after revision October 19, 1990) ABSTRACT Neuburg, H.J., Castillo, J.A., Herrera, M.N., Wiertz, J.V., Vargas, T. and Badilla-Ohlbaum, R., 1991. A model for the bacterial leaching of copper sulfide ores in pilot-scale columns. Int. J. Miner. Pro- cess., 31:247-264. The kinetics of the bioleaching of a low-grade copper ore was studied in a pilot-scale leaching col- umn over a period of 360 days using an open-circuit leaching regime with a fresh sulfuric acid solution at pH = 2 at the inlet of the column. The effluent solutions were periodically analyzed for Cu, total Fe and Fe(II) concentrations as well as Eh and pH values. Bacterial population was measured both in the solution (free cells) and at the surface of the ore particles. Bacterial growth parameters and the true kinetic constant for the dissolution of the sulfide minerals present in the ore were determined in independent experiments. A generalized phenomenological model for the bioleaching of sulfide ores is presented, for which practically all the required parameters can be independently measured. The model is capable of reproducing well measured values of the process variables, and the results are extended to the simulation of the process under different operating conditions. INTRODUCTION Hydrometallurgical processes based on bacterial leaching/solvent extrac- tion/electrowinning are increasingly applied worldwide, particularly for the recovery of copper from low-grade sulfide ores which are otherwise wasted. Depending upon the characteristics of the deposit and the ore, the technology is applied in different ways: heap leaching, in situ leaching and in place min- eral processing operations (e.g., Murr, 1980 ). This technology is also applied to the recovery of other metals such as gold (by bacterial oxidation of refrac- tory ores) and uranium. Process design and operation of these complex me- tallurgical operations can be better approached with the aid of mathematical For all correspondence. 0301-7516/91/$03.50 © 1991- Elsevier Science Publishers B.V.  248 H.J. NEUBURG ET AL. models which provide a useful tool to understand the interaction of the most important operating variables in the process behaviour. Modelling of the bacterial leaching of copper sulfide ores has been devel- oped mainly with reference to the metallurgical aspects of the process, being the bacterial role included in an implicit form. Madsen et al. (1975) mo- delled the leaching of sulfide ores by ferric iron in a pilot column, assuming a constant ferric ion concentration profile along the column, a condition ful- filled in practice by the action of bacteria on the ore, which catalyzes the reox- idation of ferrous iron. These models are based on the shrinking core concept of the mineral particles and present equations with parameters to be opti- mized from experimental measurements. In general they consider two reac- tion stages, the first taking place at the beginning of the process where the conversion rate is controlled by surface reactions and the second occurring at later stages, where the process rate is controlled by diffusion through the re- acted product zone. Among these models are the ones developed by Bartlett (1973 ), Braun et al. (1974), Madsen et al. (1975 ) and Madsen and Wads- worth ( 1981 ). Energy transport equations have also been included by Cathles and Apps ( 1975 ) and Gao et al. ( 1983 ). Cathles et al. ( 1975 ) introduced the concept of the bacterial controlling effect in the modelling of the bacterial leaching in heaps. A temperature range was defined (between 10 ° and 50 °C) in which bacteria are active and, therefore, an instant conversion of ferrous iron into ferric iron can be assumed. Outside this temperature range, bacterial activity was considered negligible, thus limiting the supply of ferric iron for the leaching process. A mathematical model that would jointly consider biological and metal- lurgical aspects of bacterial leaching operations of copper ores is still lacking. The present work describes the modelling work for the bacterial leaching of a copper sulfide ore in a pilot column. Bacterial activity is related with the ox- idation of ferrous to ferric iron and described by means of a modified Mi- chaelis-Menten mechanism, distinguishing between the catalytic action of the populations attached to the ore and free in solution. The model, using param- eters obtained from the literature, describes well the behaviour of the metal- lurgical and biological phenomena in a column experiment. The results are used to analyze the interaction between metallurgical and biological factors in the leaching process. MATERIALS AND METHODS Sulfide ore The column was fed with a -4.699 + 0.833 mm particle size fraction pre- pared from an ore sample from the Mine of El Teniente. This narrow particle size range was selected in order to minimize solution chanelling and air sup- ply limitations throughout the reaction zone, usually found in materials con-  BACTERIAL LEACHING OF COPPER SULFIDE ORES 249 taining high proportions of fines (Madsen and Groves, 1976). Copper and iron contents in the initial ore, determined by chemical analysis, were 1.0 wt% Cu and 5.8 wt% Fe; soluble copper and iron were 0.085 wt% and 0.44 wt%, respectively, determined from a 24-h leaching test of a -0.147 mm par- ticle size sample in a 5 wt% sulfuric acid solution. Specific acid consumption was determined to be around 4.2 kg H2SO 4 per kg of Cu. Mineralogical reflec- tion microscope analysis indicated that copper was mainly present as chalco- pyrite, with much smaller amounts of chalcocite and covellite. Pyrite was the main iron-containing species, with some amounts of magnetite and hematite also present. The gangue was in nature mainly siliceous with the presence of small amounts of gypsum. Leaching column and procedures A 180-cm high PVC column with a 28-cm diameter was used and loaded with around 185 kg of the ore. Temperature measurements and gaseous and solid samples were taken at 36-cm intervals starting from the top of the col- umn. The leaching solution was circulated by means of a Cole Palmer Model 7554-20 peristaltic pump and sprinkled onto the ore through a perforated spiral shaped 1/4 tygon pipe attached to the lead of the column. The column was sealed at the top with a gas trap connected to the head and a water seal at the bottom. A flow of 24 I h-~ air (saturated in water at room temperature) was fed countercurrently to the solution throughout the experimental run. The ore bed was leached with a fresh sulfuric acid solution at pH = 2 in an open circuit (no recirculation) with a flow of 1.91 I h -~. The temperature of the column varied in the range of 14 ° to 28°C, following room conditions and no significant variations along the column were observed at any time during the 360 days run. Analysis of the leaching solutions The leaching process was monitored by means of daily analysis of free-bac- teria counts, Cu 2÷, Fe 2+ and total Fe concentrations as well as Eh and Ph determinations in the effluent leaching solutions. Cu concentration was de- termined by atomic absorption spectrophotometry whereas ferrous and total iron concentrations were determined by the o-phenanthrolin spectrophoto- metric method (Muir and Anderson, 1977). Eh and pH values were mea- sured in a Cole Palmer Model 5982-10 pH-Eh meter. Bacterial activity The iron oxidation activity arises from free bacteria in the solution and from cells attached to the surface of the ore particles. Free-bacteria numbers were evaluated by determining viable bacteria concentrations in the solution by microscopic counts of colonies formed in solid plate cultures (Harrison, 1984; Espejo and Ruiz, 1987). The population of iron-oxidizing microorga-  250 H.J. NEUBURG ET AL. nisms attached to the ore along the column was estimated through the mea- surement of the initial ferrous iron oxidation rate of a sample of ore. Two grams of ore from each sampling point were contacted with 100 ml of MS-9B incubation medium, using HCI instead of H2SO 4 (Espejo and Romero, 1987 ) and the Fe(II) oxidation rate during the first 2-3 h was measured. The at- tached bacteria population was then estimated by comparing the observed Fe (II) oxidation rate with the specific ferrous iron oxidation rate of a strain of T. ferrooxidans (ATCC 19859) iron oxidizing bacteria (Espejo et al., 1989). Intrinsic oxidation kinetics of the ore The true mineral oxidation rate coefficient of the sulfide minerals present in the ore was determined through the measurement of the copper dissolution kinetics of a flotation concentrate, obtained from a - 0.417 + 0.104 mm par- ticle size fraction of the ore. A 2-g sample was leached in an acidic ferric so- lution in a shake flask at 30°C, from which the value of 4.63× 10 -7 cm s -~ was calculated for the kinetic coefficient (referred as fl in eq. 11 ) for the chal- copyrite oxidation. MODEL FORMULATION Stoichiometry of the leaching process Copper ores nearly always contain a mixture of oxide and sulfide minerals (Dana and Ford, 1971). Oxide minerals are easily solubilized by just acid solutions, whereas sulfide minerals are solubilized only under strong oxidiz- ing conditions and, therefore, the ore mineral composition varies continu- ously during leaching. Oxide species are completely dissolved during the early stages of the process, whereas sulfide species are gradually leached, with the secondary sulfide minerals being depleted first. In this study the main sulfide species present in the ore were chalcopyrite and pyrite, and therefore, from a practical point of view, only these two species are considered for modelling purposes. Although many nonstoichiometric compounds are formed on the surface of these minerals during leaching, the overall stoichiometry for the reactions that take place under the pH and Eh conditions in the column can be represented as: chalcopyrite (Dutrizac et al., 1969 ) CuFeS2 +4Fe 3+ ,5Fe 2+ +Cu 2+ +2S ( 1 ) pyrite (Garrels and Thompson, 1960) FeS2 +8H20+ 14Fe 3+ ,15Fe2+ +2SO42- + 16H + (2) It has been demonstrated elsewhere (Herrera et al., 1989 ) for the bioleaching  BACTERIAL LEACHING OF COPPER SULFIDE ORES 2 51 of low-grade chalcopyritic ores, that the microbial effect on the overall leach- ing rate is mainly related to the bacterial oxidation of ferrous ions, by both bacteria free in solution and attached onto the surface of the ore particles (Almendras et al., 1987; Espejo et al., 1989 ). Thus, the resulting ferrous ions are metabolically oxidized by bacteria according to: bacteria 2Fe2+ +O2 +2H + ,2Fe3++H20 (3) Mass balances in the flowing phases The mass balance for the chemical species i present in the system is repre- sented by the continuity-equation in the axial direction of the column: O ¢~C~) ~_vz00~ 0 Da OC~ Ot Oz\ --~-z = Y.R, 4) where es is the liquid fraction (volume of solution per unit volume of ore bed), C~ concentration of species i in solution, vz superficial liquid velocity and Oai the axial dispersion coefficient of species i. ~R~ is the net rate of mass change of the species i by production or consumption through chemical and biochemical reactions and/or transport from or into another phase per unit volume of packing, Since the process is slow and the ore particle size distri- bution used in the present work is very narrow, a small value for the Da~ dis- persion coefficient will prevail (Satterfield, 1975 ), and therefore, a plug-flow behaviour for both liquid and gaseous phases and pseudo-steady-state condi- tions can be assumed throughout the column. Thus the mass balance eq. 4 can be reduced for each species i to: v~ dC Eq. 5 is applied to the oxygen balance in the gas phase and to oxygen, ferrous ion, ferric on, cupric ion and hydrogen ion mass balances in the liquid phase. Changes in bed porosity with liquid-flow rate were estimated through stan- dard correlations (Coulson and Richardson, 1979 ), whereas unit mass trans- fer surface area per unit bed volume was estimated as suggested by Onda et al. (1968 ). Overall gas-liquid mass transfer coefficients were estimated by the correlations of Bolles and Fair ( 1979 ). Liquid holdup was estimated us- ing the correlations presented by Kafarov et al. (1969). In the present study, mainly isothermal conditions throughout the column were observed, so that the heat balances were not considered in the model.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks