Studies On Bio-Oxidation A Refractory Gold Containing Sulphidic Concentrate With Respect To Optimization And Modeling

Abstract

Although bacterial leaching of sulphidic minerals is a well-known phenomenon, it is only in the last ten years that full-scale bacterial leaching plants have been commissioned for gold processing. In order for bacterial leaching to compete successfully with other pretreatment processes for refractory ores, particularly with established technologies such as roasting and pressure leaching, it needs to be efficient. This requires the optimization of the parameters affecting the leaching reaction and the growth of bacteria.

The entire biotreatment process is agitation leaching, carried out in stirred reactors or Pachuca type reactors. The bacterial oxidation is a complex reaction involving gaseous, liquid and solid phases. The interactions are highly complex, and analysis is complicated by the presence of solids in the leaching medium. Inspite of the amount of research that has been performed, kinetic and process models are underdeveloped. Since kinetic data varies widely with the type and source of concentrate, experimental data should be generated before doing the full-scale reactor design. In sizing reactors for a commercial scale process, it would be useful to have a mathematical model that one could use to predict the amount and rate of release of metal, as a function of the various operating parameters of the system.

G.R.Halli arsenical gold sulphide concentrate obtained from Hutti Gold Mines Ltd., Karnataka, was chosen for our study, because of its high refractoriness. An indegenous strain of Thiobacillus ferrooxidans was used for biooxidation. The experiments were conducted in a well-agitated stirred tank reactor under controlled conditions. Sparged air was supplemented with carbon-dioxide for optimized growth. In this work, more than 90% gold and 95% silver could be recovered from the sulphidic gold concentrate when bioleaching was used ahead of cyanidation, compared to 40% and 50% by direct cyanidation.

A generalized model, which accounts for both direct bacterial attack and indirect chemical leaching, has been proposed for the biooxidation of refractory gold concentrates. The bacterial balance, therefore, accounts for its growth both on solid substrate and in solution, and for the attachment to and detachment from the surface. The overall process is considered to consist of several sub-processes, each of which can be described in terms of a mechanism and related rate expressions. These sub-processes were studied seperately under kinetically controlled conditions. The key parameters appearing in the rate equations were evaluated using the experimental data. Since the refractory concentrate contains pyrite and arsenopyrite as the major leachable entities, leaching studies have been done on pure pyrite and arsenopyrite as test minerals and the key parameters in the rate equations are evaluated using this data. The model so developed is tested with the leaching kinetics of the concentrate.

The growth of bacteria is dependent on the availability of the substrate, ferrous iron, and the dependence is modelled by the widely accepted Monod equation. The effect of carbon dioxide supplementation on the bacterial activity was studied and the optimal concentration for growth was found to be l%(v/v). Studies on indirect chemical leaching showed that the rate is sensitive to surface area of concentrate. Indirect rate constant of arsenopyrite was found to be greater than that of pyrite, since pyrite is more nobler than arsenopyrite. Conditions of direct leaching alone was obtained at high pulp density and using substrate adapted bacteria. The rate constant of arsenopyrite was found to be greater than that of pyrite. The parameters obtained were tested with the overall batch leaching data of the concentrate and favourable comparision was obtained.

Thus, it has been possible to isolate the various simultaneous sub-processes occurring during the leaching and propose useful models to describe these processes in some detail. The model has been extended successfully to predict the continuous leaching behaviour using the parameters obtained from the batch data. Studies on the effect of residence time and pulp density on steady state behaviour showed that there is a critical residence time and pulp density below which washout conditions occur. The critical residence time at 10% pulp density was found to be 11 hrs. Operation at pulp densities lower than 5% and residence times lower than 72 hrs is not favourable for efficient leaching. Studies on the effect of initial ferric iron concentration showed that there exists an optimum concentration of ferric iron at which the time required to reach steady state is minimum.