Ahead of the February issue of International Mining and its annual examination of the latest in mining software, Peter Mokos, Principal Mining Engineer with AMC Consultants offers some great insights. Writing in AMC’s November 2015 Digging Deeper newsletter he notes that “most software vendors providing geological modelling and mine planning software now include their own activity sequencers and Gantt-chart-style schedulers, or use an integrated third-party provider’s scheduling package. Examples of these are CAE 5DP, MineRP, Deswik, Gemcom Surpac, Maptek Vulcan, RungePincockMinarco XPAC, and Dassault Systemes GEOVIA MineSched.
“The largest step forward in recent times for mine-specific scheduling packages has been the ability to:
- Link activities using mining logic (for example, a stope cannot be mined until the required access development has been completed)
- Apply mining rules (such as geotechnical sequence), priorities and resource pools (for example, equipment fleet), to practically constrain capacity
- Provide customisable reporting (and customisable time-period reporting)
- Apply solver logic to optimise schedules
- Provide Gantt-chart-style visualisation.
“Most of the mining-specific schedulers are now integrated with their mine design structure files, and their scheduled progress can be visualised in 3D. The visualisation feature has been a powerful development, helping a vast range of schedule stakeholders (for example, the mine scheduler, mine operators, mine management, mine boards, and community forums) to better understand the proposed mine schedule. Future schedule visualisation developments are heading towards immersive technology to enable mine engineers to ‘test-drive’ the proposed mine designs and schedules. These schedulers are excellent for identifying project bottlenecks and resource inadequacies.
“The current and future drive for the mine-specific schedulers (for example, Deswik Scheduler and Enhanced Production Scheduler, or EPS) is to have functions that optimise key KPIs such as project cash flow or net present value, head grade, tonnage throughput, capping metres and developed or backfill volume placed. For the current generation of schedulers, the ability to optimise is directly related to the duration of the mining time periods being evaluated (for example, weekly, monthly, quarterly, yearly, etc.) and the available activities in that time period—and hence, the size of the problem to be solved and the limitations of computing. As a consequence, mine-specific schedulers will only look one time period ahead as opposed to true schedule optimisation, which would look at all possible outcomes for the life-of-mine schedule and then select the best outcome. These schedule optimisers typically use linear and/or mixed integer solver programming, but can also use other optimisation techniques.
“There are schedulers (for example Mirarco Schedule Optimisation Tool, or SOT) that use heuristic-type generic algorithms that can examine multiple possible schedule paths, thus learning which paths are more optimal and generating a substantial number of schedule cases from which a user can select their preferred optimum. This approach can be very useful for rapidly appraising conflicting or competing schedule priorities or strategies.
“A common feature of the mine-specific scheduling packages is the ability to integrate with third-party modules, such as HardDollar and Candy, for project management control. These are add-on modules that use the design and scheduled bill-of-quantities to build a detailed mine-specific cost model for tracking earned value and producing detailed project performance.
“Generic Gantt-type scheduling packages, such as Oracle’s Primavera and MicrosoftTM Project 2010, can be adapted for mine scheduling, however at the cost of substantially simplifying the considerations involved. These are really PERT/Critical Path Method programs and do not deal practically with mass-weighted average grades or the interaction delays between development, stoping and filling activities. These are more specifically suited to project management, and because the form of output is not suited to mining, it is best downloaded into a spreadsheet for final analysis and presentation. These packages are most successful if used for an annual budget, where the possible variations are not great, and are really designed for construction projects.
“Mine scheduling at the conceptual through to feasibility-study level can be done using ordinary spreadsheets. These are very flexible and easily understood by others. In addition, the presentation of outcomes using graphs, bar charts and time bars is excellent. Spreadsheets can handle the grade and tonnage computations while leaving the engineer free to consider the problem holistically. However, like the generic scheduling packages, they will over-simplify the activity interactions and thus have a strong tendency to overstate project capacity.
“Scheduling systems that are more complex than generic types, like Microsoft Project, require a dedicated operator. As with geological modelling and mine-planning systems, regular involvement is required to maintain skill levels after an initial learning period. These are not tools that will reduce the workload of a general mine-planning engineer, but they have the ability to produce, assess and analyse schedules more rigorously. Care must be taken to ensure that the scheduling function is well-documented and sufficiently simple such that staff transfers can be accommodated regularly without total disruption of the planning-scheduling function.
“AMC uses a variety of planning-scheduling packages to accommodate client requirements and has developed methodologies to use some of these packages to maximise project value. It has expertise in contemporary specialist packages such as CAE 5DP, MineRP, Deswik, Gemcom Surpac and EPS, as well as scheduling expertise using generic packages such as Microsoft Project and Excel.”
