Development quarterly mine plans is generally time-consuming and requires consideration for many operational constraints and competing objectives. With lengthy turnaround times, there is often limited opportunity to adjust the quarterly plan details to account for differences between actual state of the mine operation at the time of execution and expected state at the time that the plan was developed. Actual mining task completion times can vary significantly from the expected task completion times, due to many factors. In the absence of revision to account for differences between the planned and actual state, resource allocations and mining sequences, embodied in the quarterly mine plan may no longer be the best approach at the time of execution to achieve an efficient mining operation.
Simulation and optimisation tools can be used to support rapid refinements for development and execution of weekly mine plans. When initialised with an operations data feed to establish the actual state of the mine operation, optimisation models can be used to facilitate both automation and revision of the quarterly plan to produce a weekly or daily plan for execution. This approach can remove much of the manual burden of revision if the existing quarterly plan no longer represents the best approach to meeting production objectives. Following initialisation with mining actuals, simulation models can provide a representation of the range of likely mine operations outcomes. The models can be used to accelerate the planning process by providing visibility of future performance and identifying likely operational issues. They can also be used to test the robustness of the plan with respect to performance variability and response to disturbances.
Following identification of issues, simulation models can also be used for evaluating mitigation measures, through automated testing of alternative plans. Integration of mine data feeds with simulation and optimisation tools enables automated optimal responses to variations between the quarterly mine plan and actuals to be tested prior to release of and execution plan and implementation. The use of simulation and optimisation in weekly mine planning will be discussed in this paper with reference to surface mining examples.
The opportunity cost of potential improvements in converting quarterly or longer-term mine plans directly into weekly execution plans is likely to amount to billions of dollars every year. Quarterly plans often remain relatively unchanged into execution for most mining operations, particularly for cut-off grades and prescribed material destination (e.g., waste dump location) details that may be no-longer ideal at execution. Over time, variations in actual task completion times can affect or alter the critical path for the mining operations. Variable task completion times can also change the instantaneous ore quality profile and ore to waste ratios from intended outputs if mine production follows quarterly plan without revisions to account for changes. Without optimisation of the execution plan to realign production with planning objectives, value can be lost as production falls behind schedule and ore quality and ore/waste ratios vary from planned targets. Retaining some of the value in execution can be achieved through the use of support tools for the weekly planning process as well as operations data feeds and associated process changes
Mine Planning process
Mine planning is complex and requires consideration of many practical constraints and dependencies between tasks. Some of these include mine block sequencing constraints, equipment resourcing constraints, relocation costs, balancing of waste and ore production, management of varying ore grades, consideration for maintenance requirements and upstream and downstream process constraints. To develop a good mine plan, consideration for all these factors takes considerable time and effort and many planning iterations. The time taken to develop a feasible plan limits the ability of the mine plan execution team to replan in response to recent changes, such as an unplanned shovel failure, or slower than expected completion of critical tasks.
Mine schedule creep caused by variable task completion times
The actual time required to complete a mine plan task is likely to vary relative to the time allowed for completion in the mine plan. Variations can be due to one of many factors including unplanned outages, availability of interfacing equipment, upstream or downstream constraints and external factors such as weather.
A mining operation is a set of concurrent tasks, with both task dependencies (e.g., drill and blast before loading) and resource dependencies (e.g., shovel completion of one bench before moving to the next), which constrain the earliest possible starting time for a task. Variable task completion times can have knock-on effects for execution of the mine plan due to task and resource dependencies. For a task to start early, all the precedent tasks need to be completed early with sufficient resource availability. However, for a task to be delayed only one of the precedent tasks needs to be delayed or insufficient resources need to be available. This effect compounds through a mine schedule with both task and resource dependencies, such that any variability in task completion time is significantly more likely to result in mine operations progress running behind rather than in advance of the schedule over time.
An example of schedule creep is shown in Figure 1. White arrows indicate either a lead or lag for shovels 1,2,3 and 5. The dashed black line shows overall schedule creep due to a dependency for a shovel 3 task. The bottom of the figure highlights the effect on ore to waste ratios because of task completion time variations when compared with the plan.
If a mining operation follows a quarterly mine plan, schedule creep over time will be realised as reduced throughput as dependent tasks and resources are delayed by precedent tasks. Task delays may also cause equipment idle time, waiting for planned tasks to become available if the equipment can’t be effectively redeployed to an alternative task. Schedule creep has other effects including misalignment between tasks that were intended to be executed concurrently. This can affect the ore quality profile or ore to waste proportion of tonnes hauled, which can starve or exceed downstream processing capability.
Two mechanisms are generally employed to manage variability of task completion time and effect on maintaining the mine schedule.
A third mechanism which can also be employed is to implement changes to the plan. There is often optionality in the mining operation that can be exploited to maintain productivity when the system state is not what it was estimated to be at the time that the quarterly plan was developed. Changes to the quarterly plan can take the form of:
For mining systems with downstream processing constraints, these mechanisms may benefit from active management of RoM or feeder stockpile levels and use of tactical stockpiles to manage varying ore profiles. While net improvements in mine performance can be achieved with a dynamic execution plan, the mechanisms can be difficult to implement effectively. Thorough analysis is required to evaluate and identify the best alternatives and operations data feeds and reporting systems are required to support operational execution.
Often used to inform long-term strategic decisions, detailed discrete event simulation models of the mining operation can also be used to provide forward visibility of expected performance for the weekly execution plan. Simulation can provide value to the weekly planning process whenever delays due to equipment interactions (e.g., truck queuing, shovel hangtime, unplanned outages) or other variations are likely to affect the time required to complete mine plan tasks.
Simulation techniques can be used to support the mine planning process across the planning spectrum from long-term LoM strategic planning to execution. Aydiner et. al. (2006), Eustace et. al. (2020) and Nguyen et. al (2021) provide examples of simulation models used to evaluate production operations to inform the mine design and long-term mine plan. Greberg et. al. (2011) discusses the use of simulation in the mine planning process. The appropriate approach and level of simulation model detail will depend on the planning horizon, operational complexity, performance data availability and required insights. For strategic planning support, value can be derived from considering variability in completion time for a mine plan task (e.g. completion of loading for a mine block) with task dependencies. For some applications, this may be sufficient to develop a predicted distribution of completion times, without necessarily considering the detail of all the mining activities and interactions that comprise the task. At the execution end of the planning spectrum, a detailed representation is typically required to provide an accurate evaluation of the likely operations performance and completion times for each task.
To provide an accurate estimate of short-term operational performance, suitable for rapid evaluation of a weekly plan, the simulation model requires initialisation to represent the current system state. This typically requires direct integration with site operations data systems. Details that describe the current state of the mine operation include progression through each mine plan task, equipment availability and equipment location. Once initialised, the simulation model can run from the current state, following a weekly or quarterly plan to provide an assessment of likely operations performance over the period of the plan.
Many replications of the simulation of the weekly plan can be run with stochastic inputs to understand the range and distribution of potential completion times for each task in the plan, potential for knock-on task delays and metrics such as throughput and ore quality profile. This provides a significantly more robust view of the likely performance of the execution plan, provided inputs used to describe variability for fundamental parameters (activity rates, likelihood of outages etc.) align with actual performance.
In addition to providing a robust assessment of the performance range for an execution plan, a simulation model initialised to represent the current state of the mine operations, also provides a means for developing a comparison of the performance of two or more alternative weekly plans. Simulation assessment of alternative execution plans, enables comparison of distributions of performance metrics for each alternative.
Optimisation can benefit mine planning across the planning spectrum. Franco-Sepúlveda et. Al (2019) discusses the use of optimisation for long-term mine planning, including consideration for uncertainties to minimise risk and maximise profit. While often employed to improve the Net Present Value for a mine by Optimising the Life of Mine (LoM) planning, optimisation is equally important at the execution end of the planning spectrum. Optimisation provides a mechanism for considering all possible planning options and identifying the best alternative. It also enables automation of the development of alternative execution plans quickly and with minimal manual effort. This makes effective replanning for execution possible.
At execution, an optimisation tool provides agility to respond in an optimal way to variations between the mine plan and actuals. Significant departures from the mine plan can occur in any shift (such as an unplanned equipment outage) or accrue slowly over time. With the amount of time and effort that goes into developing quarterly mine plans including consideration for interdependencies between tasks, an iterative process of manually revaluating the mine plan in response to changes, such as an unplanned shovel outage is typically not feasible.
When used for execution, optimisation tools can be configured to automatically read in the current state of the mining operation including: task completion relative to the plan; the current state of RoM/feeder stockpiles; equipment availability and any downstream processing conditions to reoptimise the plan. Optimisation objectives and constraints can be configured to preserve the strategic/ tactical value embedded in the plan, while exploiting any operational flexibility to maximise output.
Optimisation models typically provide an abstracted representation of mining operations to achieve tractable solve times. Simulation models can include the additional operational details required to ensure that an optimised plan is suitable for execution. Detailed operations simulation models can also enable testing with stochastic inputs for critical variables to ensure that the plan is sufficiently robust. Integration of both simulation and optimisation tools into the weekly planning process provides both the ability to:
Upadhyay (2015), discusses the use of shovel allocation optimisation within a simulation model to determine the optimal short-term schedule for improved mine plan compliance. Embedding weekly planning optimisation in a simulation model representation of the mining operation can also be used to quantify potential long term operations performance improvements. Figure 2 provides an illustration of this process, where mathematical optimisation is integrated with the mine operations simulation model to test the performance of the optimisation model and estimate performance improvements. The same mathematical optimisation tool can then be used for actual mining operations as part of the weekly planning process for execution.
The following examples describe the application of optimisation and simulation support for weekly planning. Each example application also relies on optimisation of longer-term plans to establish strategy and performance targets for weekly planning optimisation.
An existing open pit mining operation uses drill and blast at multiple pit locations with transfer of ore to Run of Mine (RoM) heap leach pads to extract a gold-bearing solution. Haulage of ore and waste from open pits to leach pads and waste dumps is carried out by a mixed fleet of shovels and 300t haul trucks. A complex haulage system to transport ore and waste with multiple pits, leach pads and waste dump locations changes over time as pits, leach pads and waste dumps are opened and closed. The mining sequence is governed by a mine plan with a block model that describes mining locations and allocated shovels over time. Tipping locations for ore carrying trucks to heap leach locations or waste carrying trucks to waste dumps are also prescribed by the mine plan based on a shortest path calculation.
Over the life of the mine, multiple waste dump options are available and the average distance from pit locations to waste dump locations varies significantly over time. Always choosing the closest waste dump location results in significant variations in average haul cycle time over the life of the mine and creates short-term throughput bottlenecks with truck haulage capacity shortfalls when the average haul cycle is long and surplus haulage capacity when the average haulage cycle is short.
Life-of-mine strategic optimisation considers the material haulage requirements (location and volume) and capacity of waste dump locations over the life of the mine to reallocate dump locations for waste haulage tasks in the mine plan. The optimised strategic plan minimises variation in average haulage cycle times for each mine plan period. Reallocated waste dump locations are then used as targets for each mine plan period (e.g., 40% long, 60% short) to inform the weekly plan. Optimisation of the weekly plan considers the current state relative to haulage targets, shovel fleet and truck fleet availability to determine weekly targets that are used by dispatch algorithms to determine the best waste dump location for each truck. The application of weekly targets and dispatch algorithms was tested using a simulation model of the system prior to implementation.
The approach is used to align truck fleet capability with shovel fleet capability and significantly reduce variation in average haulage cycle time over the life of mine, reducing haulage bottlenecks and improving production performance.
In a second example, a large fleet of ultra-class haul trucks is used to haul ore from several different shovels located in different sectors of a large open pit mine. The mining sequence is controlled by a mine plan which describes the number of concurrent shovels, locations, mining rates and is linked with a block model that describes ore/ waste classification and ore attributes for material loaded by each shovel over time. Haulage operations follow the quarterly mine plan and the specified ore and waste destinations for each mining block. Ore is fed through a primary crusher to a downstream processing facility with a large feeder stockpile used to buffer downstream processing from variable the haulage operational performance. With variable progress through each mine plan task, the set of concurrent blocks mined and therefore ore/waste ratios and ore grade profiles vary relative the mine plan.
Based on the block sequence provided by the quarterly plan, implementation of weekly plan optimisation considers actual progress through each mine block, feeder stockpile levels and current equipment availability via automated operations data feeds. The optimisation tool uses available shovel and haulage capacity to improve average ore ranking and consistency for ore sent to the crusher. This is achieved through:
Optimised weekly plans are tested using a simulation model of the haulage operation to quantify the range of likely system performance given variability in downstream processing rates and delays to the haulage operation. A comparison of daily haulage totals over a quarterly period is provided in figure 3. Daily tonnes hauled to the crusher follows the quarterly mine plan in the upper chart, with the haulage system ultimately constrained by downstream processing rates. The lower chart shows the results of a weekly planning optimisation process. For the optimised operation, sufficient ore is sent to the crusher to feed downstream processing, but ore is redirected to remote ore stockpiles when there is sufficient shovel and haulage capacity. Significantly higher (~30%) overall daily haulage targets are achieved, enabling a corresponding increase in the cut-off grade for ore fed to downstream processing.
Enabled by automated data feeds, simulation and optimisation models can be used effectively to support the weekly execution planning process and to drive operations performance improvements through plan execution. Optimisation models are used in execution planning to identify where changes to quarterly or longer-term plans are likely to lead to improved operations performance, while remaining aligned with the overall production objectives. Simulation models can be used to evaluate the likely range of performance outcomes for the existing plan, identify potential issues and compare performance outcomes for alternative plans.
Aydiner, Kerim & Çelebi, Neşe & Paşamehmetoǧlu, A.G.. (2006). A simulation model for mine production sequences, Proceedings of the 2006 Applied Simulation and Modelling Conference, Rhodes, Greece,pp 290-295.
Franco-Sepúlveda, Giovanni & Branch, John & A, Patricia. (2019). Stochastic Optimization in Mine Planning Scheduling. Computers & Operations Research. 115. 10.1016/j.cor.2019.104823.
Greberg, J and Sundqvist, F, 2011. Simulation as a tool for mine planning, in Proceedings Second International Future Mining Conference 2011, pp 273-278 (The Australasian Institute of Mining and Metallurgy: Melbourne).
Eustace, C, Lagace, D & Bobbermen, L 2020, ‘Simulation Capacity Analysis for the Carrapateena Block Cave’, Edited by K-H Bae, B Feng, S Kim, S LazarovaMolnar, Z Zheng, T Roeder, and R Thiesing, Proceedings of the 2020 Winter Simulation Conference, http://www.wintersim.org, Orlando, Florida
Nguyen, K, & Hegarty-Cremer, S, 2021, ‘Haulage Simulation with Complex Routing and Tactical Stockpiling’ Edited by S. Kim, B. Feng, K. Smith, S. Masoud, Z. Zheng, C. Szabo and M. Loper, Proceedings of the 2020 Winter Simulation Conference, http://www.wintersim.org, Phoenix, Arizona
Upadhyay, Shiv & Askari Nasab, Hooman & Tabesh, Mohammad & Badiozamani, Mohammad. (2015). Simulation and Optimization in Open Pit Mines, Proceedings of the 2015 SME Annual Meeting, Denver, Colorado.