A New, More Strategic, Way to Mine ?
Mining, like many industries, can be slow to change. We often stick to traditional processes far longer than we should because we are comfortable with them and don’t want to take the risk of trying out new ones. But that means we may also miss significant opportunities both to improve profits and increase sustainability.
A common refrain from mine planners is that they do not have enough time to look at all the possible options for a solution space, which leaves them with sleepless nights worrying about such questions as: How much value was left in the last untested cut-off grade or mining capacity limit? Which direction and sequence would have created the best schedule? Have we followed the pit optimisation angles closely enough?
The fact is, however, that this situation can be solved — and these important questions can be answered — if we adopt a different approach to strategic mine planning.
Strategic mine planning ?
Mine planning concentrates on long-range production planning aimed at maximising the value derived from exploiting an ore deposit. However, by its very nature, because it is long-term, a mine plan can be affected by a variety of internal and external forces including, for example, increased knowledge of the orebody, unexpected staffing issues, technical advancements ,and changes in legislation, economy, and market.
Strategic mine planning attempts to de-risk a mine plan, to make it flexible enough to adapt to changes as and when they rise.
Traditional approach ?
The traditional approach to developing a mine plan is to assess a mine project based on the net present value (NPV). That means the NPV, which is calculated by applying a rate to progressively discount cash flows based both on how much profit the mine project must make and on its risks, becomes the primary KPI for the mine plan and drives decisions about where to start the extraction and how to orient the sequence.
For open pit mines, mine planners traditionally define reserves using the Lerchs-Grossmann (LG) algorithm, which identifies an economic envelope (pit shell), constrained to maximum slope angles, that will maximise the total undiscounted cash flow. With that final pit identified, the planner builds a sequence to reach the final pit often by creating nested pit shells using the same algorithm but constraining the volume of the output envelopes or adjusting the block model valuation using revenue factors (RFs). To select a subset of the nested pits to serve as pushback expansions toward the final pit, the mine planner then calculates the preliminary schedules.
Issue with this approach ?
The issue with this traditional approach is that most of the time, the nested shells available for the planner to select as pushbacks are not operationally feasible, and that may in turn require:
Often, the planner will try to override these issues by building some feasible pushback designs loosely based on a set of nested pit shells and by splitting and merging different envelopes. However, this often seals the decision to use a pushback sequence based on the RF-limited pit shells instead of looking for other possible sequences towards the same final envelope. Plus, as a side effect, because the traditional approach is based on maximising undiscounted cash flow for simulated price-levels through different RFs, there is no guarantee that the sequence obtained will maximise NPV and could be out of alignment with other feasibility-focused KPIs.
A more flexible approach ?
Using process-automation tools with mine planning software allows mine planners to better appraise the optimisation solution space, delivering a workflow such as this: ?
2. Run thousands of possible scenarios based on mining rate and production capacity, their corresponding CAPEX and OPEX (making both the mine and the processing plant, and their corresponding costs, the right size), and cut-off grade, with each scenario producing its own mine plan and production schedule.
3. Optimise the schedule to maximise NPV by identifying what material to mine from each pushback and when, since the “what and when” will affect the mine’s order of revenues and costs (aka cashflow). ?
For example, the traditional approach has been to optimise the material send to the processing plant based purely on the mining and processing capacity. This can lead to low grade material taking up vital plant capacity in the early periods and reducing NPV among other KPIs. ?
A better approach would be to stockpile lower grade ore in the early years of production in order to prioritise higher-grade processing early on, and then use the remainder of the viable ore later, increasing NPV over the life of the mine. Even better still would be to not only optimise the processing capacity, cut-off grades, and stockpile usage, but to do this at the same time as choosing the sequence and the pit shells. This would free the optimisation to look at a wider solution space and not lock it in to decisions that were made in the previous step.
Strategic mine design ?
It is important to remember that optimisation and scheduling is only one side of the coin that is mine planning and that it takes a design to make a schedule actionable. In a process where we are creating large numbers of scenarios for optimisations and schedules, it is critical to establish a living design model with an?intelligent workflow?that updates as objects and inputs change. ?
Traditional CAD-based mine design works well but because it requires a designer to modify the entire shape of a design in response to a change, it is often slow, manual work that is prone to mistakes. This slowness can mean that a designer is able to produce just one or maybe two design options by deadline, with no time left for engineers to evaluate the integrity of the design. ?
Adding an automated parametric capability to the traditional design process not only ensures faster execution, it does so with improved accuracy and flexibility over traditional mine design. ?
领英推荐
Parametric design fundamentals
Parametric design?does not produce a solution as much as generate a family of possible outcomes through dynamic automation.
Parametric modeling can use either a:
Propagation-based systems often include ‘form-finding’ processes that optimise specific design goals against a set of design constraints, so that the final form of the designed object is ‘found’ based on these constraints.
Both types of parametric modeling have been used for years in other industries, such as civil construction, aviation, and manufacturing, as a replacement for traditional 3D CAD-based design. ?
Creating a living model
The parametric?model-based approach?incorporates traditional CAD functions but differs by adding links between objects and parameters.
This associativity preserves the connection between reference data — such as terrains and geology or resource models — and existing infrastructure models. This in turn allows the mine designer to update designs automatically every time there is new input data because, while the input data may have changed, the parameters of the design may not. The designer can also create templates by searching a series of functions?and parameters, to speed up the time needed to design repetitive tasks, and deploy them manually or automatically through scripting.
The result is a “living” model where design changes made in a localised area will update the global mine design, and designs are ready for review days or even weeks faster than traditional practice allows. ?
Running limitless simulations
Mining projects are complicated, expensive, and extremely risky ventures. Being able to simulate everything from the?mine design?to the material movement in advance is critical to de-risking a project.
By?automating the manual and iterative work done by the mine designer, parametric simulation enables the designer to compare the original design configuration with a larger spectrum of data. It works like this: regression models?are first trained on simulation data and then progressively calibrated on measured data during a set monitoring period in order to (1) evaluate the?robustness of design-phase performance and detect potentially critical assumptions, and (2) maintain a continuity with operation-phase performance with feed-back from measured data. ?
Applying simulations in real life
Parametric simulations can be used to design mining phases that consider unexpected variations and uncertainties, such as?the metal content available in a mineral deposit and shifting commodity prices.
In the illustration below, we used a Design of Experiments (DoE) to perform a wide range of input modifications to a pit optimisation run. This allowed us to calculate tens of thousands of ?scenarios and explore the entire solution space, with the output being a dynamic set of pit shells linked to and associated with a set of pit design parameters. This associativity, coupled with parametric design, created the design shown below.
The design now maintains a constant link with the optimisation results. As alternative scenarios are selected, new designs are automatically created and stored with their own revision and life cycle. We can also choose to link and associate them with other restriction criteria not made available to optimisation, such as pit crusher locations that require their own areas for infrastructure, flat areas in the ramp for regulatory purposes, or sump locations for pumping requirements. And we can assign a template to each of these criteria that is associated with the design and will be used to automatically update it. ?
Finally, each design can be used again within the life-of-mine scheduling, closing the planning loop and confirming the assumptions taken previously in the optimisation step.
Conclusion
Strategic mine planning and parametric design are critical innovations at a time when mining companies are looking to reduce time to market and address marginal economic deposits, social, and ESG challenges.
If we can reduce our reliance on traditional mine planning tools and embrace new and innovative technologies, we will find the opportunities we need to move forward into a secure and responsible future.
*****
Glenn Barlow and Ross Milne , Dassault Systèmes, consider how companies can adopt strategic mine planning and parametric design to improve flexibility and revenue, for Global Mining Review
Associate Professor at University of Nevada, Reno
9 个月liked the innovation.
Chief Mining Engineer (Senior Manager & Head of Mine Engineering | Adjunct Professor - AI & Machine Learning)
9 个月Parametric designs, Design of Experiments: I like this shift in thinking ????.
Intriguing insights on the evolution of mining strategies—embracing innovation in planning and design could indeed be a game-changer for the industry's future.