This paper presents a case study of the modelling of the natural oxidation of pyrite in a refractory gold ore stockpile at a mine on the island of Aniolam in Papua New Guinea to predict the gold recovery through direct cyanide leaching. The stockpile comprises low-grade material that was mined twenty years ago to access the orebody. The value of this low-grade resource could be increased if the processing cost could be reduced. The natural oxidation of the gold-bearing pyrite in the stockpile may reduce the processing cost by reducing or eliminating the need for pre-treatment of the material and enable gold recovery via direct cyanide leaching. Depending on the oxidation level and the gold recovery that could be achieved by direct cyanide leaching, reclamation of the stockpile may be more profitable if the pyrite oxidation stage could be fully or partially bypassed. In this paper, we describe the modelling of the pyrite oxidation level in the stockpile using the numerical model reported in our previous work. The characteristics of the stockpile used in the model are the stockpile geometry, the geothermal heat underneath the stockpile, and the properties of the different types of material, including the acid neutralization capacity, rock particle size distribution and pyrite grain size distribution. Limited onsite test results were used to calibrate the model and the final oxidation level over the stockpile was estimated using the calibrated model. Based on the simulation results, it was found that the accessibility to oxygen is the limiting factor for the level of pyrite oxidation when diffusion is the dominant mechanism for oxygen transport within the stockpile. However, experimental tests on samples showed that sufficient oxygen resupply should have been available in the stockpile and therefore there may be air movement mechanisms other than diffusion and thermal-induced air convection onsite. This study also demonstrates that the numerical model can be used to estimate the gold recovery via direct cyanide leaching from the simulated oxidation profile together with the gold distribution in pyrite grains and gold fractions in pyrite of different sizes.

本文介绍了对巴布亚新几内亚 Aniolam 岛上一个矿山的难熔金矿石库存中黄铁矿自然氧化建模的案例研究，以预测通过直接氰化物浸出的黄金回收率。该库存包括 20 年前为获取矿体而开采的低品位材料。如果可以降低处理成本，则可以增加这种低品位资源的价值。库存中含金黄铁矿的自然氧化可以通过减少或消除材料预处理的需要来降低加工成本，并通过直接氰化物浸出实现黄金回收。根据直接氰化物浸出可以实现的氧化水平和金回收率，如果可以完全或部分绕过黄铁矿氧化阶段，回收库存可能会更有利可图。在本文中，我们使用我们之前工作中报告的数值模型描述了库存中黄铁矿氧化水平的建模。模型中使用的料堆的特征是料堆的几何形状、料堆下方的地热以及不同类型材料的特性，包括酸中和能力、岩石粒度分布和黄铁矿粒度分布。有限的现场测试结果用于校准模型，并使用校准模型估计库存的最终氧化水平。根据仿真结果，研究发现，当扩散是库存内氧气传输的主要机制时，氧气的可及性是黄铁矿氧化程度的限制因素。然而，对样品的实验测试表明，库存中应该有足够的氧气补给，因此现场可能存在扩散和热诱导空气对流以外的空气流动机制。该研究还表明，该数值模型可用于通过从模拟氧化剖面中直接氰化物浸出以及黄铁矿颗粒中的金分布和不同尺寸黄铁矿中的金分数来估计金回收率。对样品的实验测试表明，库存中应该有足够的氧气补给，因此现场可能存在扩散和热诱导空气对流以外的空气流动机制。该研究还表明，该数值模型可用于通过从模拟氧化剖面中直接氰化物浸出以及黄铁矿颗粒中的金分布和不同尺寸黄铁矿中的金分数来估计金回收率。对样品的实验测试表明，库存中应该有足够的氧气补给，因此现场可能存在扩散和热诱导空气对流以外的空气流动机制。该研究还表明，该数值模型可用于通过从模拟氧化剖面中直接氰化物浸出以及黄铁矿颗粒中的金分布和不同尺寸黄铁矿中的金分数来估计金回收率。