Quantitative methods for sustainability analysis and integrated management of mining systems

矿山开发项目是资本密集和有风险的，因此必须分阶段设计。早期阶段的预算相对较少，并决定是否追求需要更大预算的后续阶段。这种多相方法对可持续工程至关重要，被称为前端加载（FEL）。它要求对经济、社会和环境影响进行全系统的定量分析，在每一阶段内加以应用，并须达到多目标最优化；它决定最终设计是否应该实现。此外，如果设计得到实际实施，定量分析将成为对所产生的采矿系统进行综合和可持续管理的基础。目前的建议是制定定量框架，以评价应用地质统计和地质冶金建模的现代控制和自动化战略，以便在全系统范围内协调矿物价值链。在项目的早期，系统可能仅用主要的物质和能量流表示，就像在标准生命周期评估中一样，由静态蒙特卡罗模拟支持。更发达的模型需要动态表示，由离散事件模拟支持；它们可以描述计划外的事件，和/或材料和能源投入的激增，以及环境释放。每当接近或超过系统容限时，最终呈现环境风险的是动态涌动。矿体、设备和市场条件的变化可能引发操作模式的变化。作业做法的标准化确保了根据事件和趋势向整个矿山的各个作业单位发出连贯的指示。此外，操作数据不断积累，通过机器学习支持流程优化。例如，岩石类型的变化可能需要改变爆破模式，以及相应的（机器学习）下游粉碎参数的变化。交替操作模式因此是对工艺变化的综合响应。然而，可用的模式受到所选设备的限制，因此必须在设计过程中尽早考虑。考虑到采矿业的竞争性质，矿山的可持续性可能确实取决于自动化技术的智能协调。

Mine development projects are capital intensive and risky, and must therefore be engineered in phases. Early phases have relatively small budgets, and determine whether or not to pursue the subsequent phases that require larger budgets. This multiphase approach is essential to sustainable engineering, and is known as Front End Loading (FEL). It requires system-wide quantitative analysis of economic, social and environmental impacts, applied within every phase, and is subject to multiobjective optimization; it determines if the final design should be implemented. Furthermore, if the design is actually implemented, the quantitative analysis becomes the basis for integrated and sustainable management of the resulting mining system. The current proposal is to develop the quantitative frameworks to evaluate modern control and automation strategies, which apply geostatistical and geometallurgical modeling, for system-wide coordination of the mineral value chain. Early within a project, the system may be represented with only the principle material and energy flows, as in standard Life-Cycle Assessment, supported by static Monte Carlo simulation. The more developed models require dynamic representations, supported by discrete event simulation; they can describe unplanned events, and/or surges in material and energy inputs, as well as environmental releases. It is ultimately the dynamic surging that presents environmental risks, whenever system tolerances are approached or exceeded. Variations in the orebody, equipment, and market conditions can trigger changes in the mode of operation. The standardization of operational practices ensures that coherent instructions are sent to the various unit operations throughout the mine, in response to events and trends. Moreover, operational data is continually accumulated, which supports process optimization through machine learning. For example, a change in the rock type may require a change in the blasting pattern, as well as corresponding (machine learnt) changes in the downstream comminution parameters. Alternating operational modes are thus an integrated response to process variation. However, the available modes are limited by the equipment that is selected, and must therefore be considered early within the design process. Considering the competitive nature of the mining industry, the sustainability of mines may indeed depend on the intelligent coordination of automated technology.