How integrated mining simulation reveals hidden production capacity, challenges capital assumptions, and strengthens investment decisions.
By Bryan Monk, P.Eng. | Founder, MISIM Modelling
How much production capacity might be hidden within a mining operation that already appears to be adequately designed?
At BHP's Jansen Potash Project, integrated simulation helped answer that question, with significant implications for the investment case.
During prefeasibility, Jansen's estimated Stage 1 annual production capacity increased by approximately 15–20%, with roughly two-thirds of that improvement credited to the modelling. The potential additional production was projected to add $500 million to Stage 1 net present value.
These were improvements to the project's production and investment basis, identified before the mine began operating.
The work was documented in a peer-reviewed INFORMS publication that I co-authored in 2018.
What made the modelling valuable was its ability to represent the entire mine-to-market system, including the interactions between underground mining, hoisting, processing, rail transportation, port operations, and customer demand.
After more than 30 years developing simulation models for mining and complex industrial operations, I've seen the same principle emerge repeatedly.
Some of the greatest opportunities don't lie within individual assets. They lie in understanding how those assets work together.
And the most valuable findings are often those that challenge an assumption that initially seemed entirely reasonable.
1. Why do mines miss production targets when every operating area has sufficient capacity?
Because individual equipment capacity does not necessarily translate into achievable system production.
Consider a mining operation designed to produce eight million tonnes annually.
The mine plan supports the required extraction rate. The processing plant has adequate nameplate capacity. Storage, rail, and export infrastructure have all been designed around the same production target.
On paper, the pieces fit.
But mines don't operate continuously at average design conditions.
Equipment fails. Maintenance takes assets offline. Material accumulates in storage. Production rates fluctuate. Operating schedules change.
More importantly, these events interact.
A processing outage can cause upstream inventory to accumulate until mining is restricted. A shortage of feed can starve downstream equipment, even though that equipment is fully operational. An outbound logistics constraint can cause finished-product inventory to build, eventually affecting plant production.
Storage may absorb these disruptions temporarily, but it can also delay their consequences until another part of the system is affected.
Engineers naturally concentrate on their respective areas of responsibility. Mining, processing, maintenance, and logistics teams each develop sophisticated analyses of their systems. Those analyses are necessary, but the assumptions made at the interfaces between areas can be difficult to reconcile.
We've encountered this repeatedly in our work. Each discipline may have a sound understanding of its own equipment and operating requirements, while the combined system behaves differently from what the individual analyses suggest.
The result is that every operating area may appear adequately sized, yet the integrated operation still falls short.
The critical question is whether all those systems, operating together under realistic conditions, can deliver the required production.
Discrete-event simulation (DES) provides a way to test this. It represents equipment behaviour, material flows, storage, resources, operating rules, and their interactions over time.
We can then examine the production losses that emerge from the system, rather than relying solely on the capacities of its individual components.
2. Where is the real bottleneck, and why does it keep moving?
Most engineers are familiar with the concept of a bottleneck: the constraint limiting the output of a system.
The challenge in mining is that the bottleneck isn't necessarily fixed.
The governing constraint can move between operating areas over time, even when no deliberate change has been made to the operation.
Ore characteristics may change, affecting mining productivity, processing rates, or recovery. A temporary equipment failure can create a short-lived constraint. Crew availability, shift coverage, and operating practices can also alter the system's capacity.
Some effects last hours. Others persist for months.
As a mine develops, ore characteristics and production requirements evolve. New equipment, automation, or changes in operating strategy may introduce further changes to the constraint structure.
A plant might limit production when processing difficult ore. Mining may become the governing constraint during periods of reduced equipment availability. Logistics may constrain output when production is strong and outbound capacity is fully utilized.
The operation hasn't necessarily changed its design, yet the system constraint has moved.
This creates a problem for conventional debottlenecking studies.
A snapshot of operating performance might identify a genuine bottleneck, but will eliminating it improve annual production? How often does that asset actually govern system output? What happens during the periods when something else is limiting production?
An investment based on a temporary constraint may provide considerably less value than anticipated.
An example from a copper-gold project
On a copper-gold project we modelled, the processing plant was identified as the life-of-asset design bottleneck.
However, the mining operation needed to support several mine plans, each with different requirements for the truck fleet.
The integrated simulation revealed periods when the available truck fleet could not deliver sufficient ore to keep the processing plant fully supplied. During those periods, mining became the governing constraint, even though the plant remained the nominal design bottleneck over the life of the operation.
The analysis allowed us to evaluate fleet requirements across the mine plans and identify circumstances where insufficient mining capacity could result in lost plant production.
The governing constraint changed with operating conditions, without any change to the physical design of the system.
Identifying the nominal bottleneck was only part of the analysis. Understanding when other constraints emerged, how long they persisted, and their consequences for production provided a much stronger basis for equipment sizing and investment decisions.
Understanding how constraints move over time can fundamentally change where capital and operating improvement efforts should be directed.
3. When does additional storage actually improve production?
Storage is one of the most familiar ways to manage interruptions between operating areas.
If upstream equipment stops, downstream operations can continue drawing from inventory. If downstream equipment becomes unavailable, upstream production can continue until storage reaches capacity.
The engineering logic is sound.
But how much storage is enough? And at what point does adding more stop delivering meaningful value?
On one project, a client was considering investing in an additional storage buffer, expecting greater inventory capacity to improve operating performance.
We modelled the proposed configuration and evaluated its effect on the wider system.
The results showed no significant improvement in overall performance.
The modelling gave the client reason to reconsider whether the proposed investment was justified.
The broader engineering question is how much storage is actually needed to protect production, and at what point additional capacity stops delivering meaningful value.
How much storage is enough?
Published research provides an interesting illustration.
A 2021 study in Minerals used discrete-event simulation to examine how target ore stockpile levels affected a conceptual oil-sands processing system, using data derived from Canadian oil-sands operations.
The researchers investigated how inventory targets and operating strategies influenced production under varying geological conditions.
The results were revealing.
With a target stockpile inventory of approximately 4.37 million tonnes, simulated throughput averaged 26.5 thousand tonnes per hour.
Doubling the target inventory to 8.75 million tonnes increased throughput to approximately 28.7 thousand tonnes per hour.
But increasing the target fivefold, to nearly 21.9 million tonnes, produced only a marginal further increase, to 28.8 thousand tonnes per hour.
Most of the simulated throughput benefit had already been achieved at the smaller inventory target.
These were changes to target inventory levels in a conceptual model, not comparisons of physical storage expansion projects. Nevertheless, the results illustrate how quickly the production benefits of additional buffering can diminish.
The study also examined the risk of stockouts and interruptions to normal operating modes. Inventory targets and control thresholds could materially influence operating resilience, even when average throughput changed very little.
That introduces another investment consideration.
Is additional inventory justified by higher expected production, improved resilience, or both?
Why can storage have diminishing returns?
The value of a buffer depends on the capacities and reliability of the operations it connects, the duration of interruptions, and the ability to recover production once normal operations resume.
More storage may allow an upstream process to run longer during a downstream interruption. But if downstream equipment cannot subsequently process the accumulated inventory, the production benefit may be limited.
Conversely, a modest buffer can sometimes protect a critical process against frequent short interruptions and produce a substantial improvement.
Storage also has costs, including construction, equipment, inventory carrying costs, and additional material handling.
The economically justified capacity depends on the operating system and the value of the benefits being provided.
The opportunity is to establish how much storage the system needs, rather than assuming that more must be better.
4. Can mine-to-market simulation help defer major capital expenditure?
Mining projects routinely consider investments in equipment capacity, redundancy, storage, transportation infrastructure, and operating flexibility.
These investments often have valid technical justifications within their respective engineering disciplines.
But do they create sufficient value across the complete system?
This was one of the important questions addressed during the Jansen studies.
The project's Detailed Integrated Capacity Estimate (DICE) model represented the complete production and logistics chain, from underground mining through customer delivery.
It captured equipment behaviour, reliability, material flows, storage, maintenance, and operating rules, allowing different capital configurations to be evaluated against achievable production.
One particularly significant decision involved the second hoisting shaft.
The integrated analysis provided the justification to defer approximately $300 million in capital expenditure associated with equipping that shaft during the initial stage of development.
The investment wasn't necessarily eliminated permanently. The model supported a different decision about when the additional capacity would be needed.
Capital invested before it is required imposes an opportunity cost. Deferring expenditure, while preserving the ability to expand when justified, can materially improve the investment case.
The same reasoning applies to processing capacity, storage, redundancy, and logistics infrastructure.
Can the planned production be achieved with a different equipment configuration? Would an operating change reduce the need for additional infrastructure? Could an expansion be staged rather than built entirely at the outset?
The Jansen experience demonstrates what can be achieved when integrated production modelling becomes part of the capital allocation process.
The goal is to invest in the right capacity, in the right location, at the right time.
5. Can operating changes improve mine performance without additional equipment?
Operating decisions can have consequences that rival those of major equipment investments.
Mining companies routinely make decisions about maintenance, operating schedules, crew assignments, equipment priorities, production sequencing, and resource allocation.
These choices influence how effectively installed capacity is used.
Consider planned maintenance.
Taking equipment offline doesn't necessarily reduce final output if sufficient inventory is available or another part of the operation is already constraining production.
Conversely, a relatively short interruption at a critical interface can create significant losses elsewhere.
How much downtime can an operation absorb before production is affected? Could maintenance be coordinated with other operating constraints? Would changing equipment priorities or production sequencing deliver more value than improving equipment availability?
Underground mine development is a useful example. The sequence in which development headings are advanced affects equipment access, shared-resource availability, and the progress achievable over time.
A published simulation study of Codelco's New Mine Level development examined these interactions, evaluating alternative development sequences, equipment maintenance strategies, and orepass availability.
Over a modelled six-year period, revising the initial development sequence increased projected advance by 1,598 metres relative to the base case. Changes to maintenance strategy and orepass availability produced smaller improvements of 457 metres and 194 metres, respectively.
These were simulated scenario outcomes rather than subsequently measured operating gains. The comparison is nevertheless instructive.
In that study, changing the sequence of work offered substantially greater potential than the individual equipment-related interventions.
The finding illustrates how operating logic and resource coordination can influence performance just as materially as physical equipment capacity.
From a focused bottleneck to broader operating support
We've encountered related questions in our own work.
At Shell's Albian oil sands operation, our modelling began with a maintenance-intensive bottleneck associated with the conditioning lines.
What started as a focused model expanded into broader mine and plant decision support. Related work across Shell's Athabasca operations involved extensive interconnected processing facilities, storage, blending, and material flows.
The models were calibrated against actual operating performance and used in long-term planning.
Across these assignments, we've found that improving an equipment-level performance indicator doesn't necessarily improve the output of the wider system.
An equipment reliability improvement may be valuable. So might changing maintenance timing, resource priorities, or production sequencing.
The potential benefit needs to be assessed against its effect on the complete operating system.
6. Can integrated simulation help define rail service agreements and logistics requirements?
For bulk mining operations, rail service and terminal capacity can be critical to the production and delivery plan.
Rail agreements commonly specify volume commitments, service frequency, and performance measures. These establish an important planning basis.
But how should the required service terms be determined?
Mine-to-market simulation can help define the rail service levels, terminal performance, and storage requirements needed to support planned production.
Consider a mine with a defined annual export target.
What train frequency is required? What service timing and variability can the system accommodate? What unloading turnaround times are needed? How much storage is economically justified?
And if the mine is planning an expansion, which service commitments must change to support the higher production rate?
These questions can inform commercial negotiations before agreements are finalized.
They can also be used to evaluate existing service arrangements against new production requirements.
Simulation has proven useful in detailed mining freight planning. For example, research by Jones and Gun combined simulation and scheduling methods to generate feasible train timetables and destination assignments across a mining freight network.
The research illustrates how operating constraints can be incorporated into detailed transportation planning rather than relying solely on aggregate annual capacity.
Our work with bulk-export terminals extends this type of analysis across the interfaces between rail delivery, unloading, storage, and marine operations.
From logistics capacity to commercial commitments
In one shared bulk-export terminal application, our modelling represents rail unloading, stockyard inventory, shared operating resources, berth operations, and vessel movements.
The analysis supports decisions involving storage requirements, third-party volume commitments, operating priorities, infrastructure capacity, and demurrage exposure.
These are interconnected engineering and commercial decisions.
If third-party volumes increase, what commitments can the terminal reliably support? What infrastructure or operating changes would be needed? What rail service arrangements protect the production objectives of individual users?
We've found that these questions are particularly important when critical infrastructure is owned or operated by third parties. The engineering assessment needs to inform the service arrangements the operation depends upon.
A model can evaluate proposed terms, compare alternatives, and establish the performance requirements needed to support production and delivery commitments.
The analysis can influence both what infrastructure is needed and what service commitments should be negotiated.
7. How reliable is the production forecast in your mining feasibility study?
This is an area where I believe many mining projects miss a significant opportunity.
Feasibility studies bring together mine plans, process designs, equipment capacities, reliability assumptions, and production schedules to establish the project's production and financial basis.
Sensitivity analyses typically examine the financial consequences of changes in production, costs, prices, and other project assumptions.
But how well has the underlying production range actually been established?
An integrated simulation can represent operating behaviour in considerable detail, including equipment failures, maintenance, material flows, storage, and interactions across operating areas.
Yet even a sophisticated model can produce a misleadingly narrow range of outcomes if the underlying project assumptions are treated as known.
Variability and uncertainty are different problems
Consider equipment reliability in a proposed mining operation.
A simulation might represent equipment failures and repairs using assumed reliability and maintenance parameters. Each simulation run produces different failure events, interruption durations, and production outcomes.
That's operating variability. We can quantify how the system performs under those assumptions.
But how confident are we in the reliability parameters themselves?
What if the equipment proves less reliable than expected? What if maintenance takes longer, or operating conditions are more demanding than anticipated?
Those questions concern uncertainty in the project basis.
Repeatedly simulating the same assumptions won't tell us whether they're correct. We need to evaluate credible alternative assumptions and understand how they affect achievable production.
In our experience, the resulting production range can be substantially wider than the variability observed within any single project basis.
This is particularly important for greenfield mining projects, where some of the most consequential assumptions have yet to be validated against actual operating performance.
Detailed simulation is entirely capable of supporting uncertainty analysis when uncertain inputs are explicitly represented. The concern is when repeated simulation runs using one assumed project basis are treated as sufficient evidence of the full range of production outcomes.
What does research show about geological uncertainty?
The consequences of uncertainty have been studied extensively in mine planning.
Research by Dimitrakopoulos, Farrelly, and Godoy examined how geological uncertainty affects open-pit mine design and project economics.
Rather than relying on one representation of an orebody, the researchers considered multiple geological realizations and evaluated the consequences for production, costs, and project value.
The resulting range of outcomes differed from the projections obtained using a single orebody model.
Geological uncertainty is one part of a broader problem.
Mining productivity, equipment reliability, process performance, operating strategy, and logistics assumptions can also change the project's achievable production.
The relative importance of these uncertainties depends on the project, but their combined effects can be considerably greater than the operating variability observed under a single design basis.
For greenfield developments, this is particularly important because many fundamental production assumptions have yet to be tested against actual operations.
Why isn't conventional sensitivity analysis enough?
A financial sensitivity might calculate project NPV at production rates five or ten percent above and below the base case.
That tells us how sensitive project value is to production.
It doesn't establish whether those production outcomes are plausible, how likely they are, or which assumptions would produce them.
Nor does it necessarily capture how the governing system constraints change under different project conditions.
A lower mining productivity assumption might reduce production directly. It could also change equipment utilization, inventory behaviour, and the importance of downstream capacity.
These interactions can alter both expected production and the distribution of possible outcomes.
Detailed simulation is valuable for understanding those relationships, but exploring a large number of uncertain project bases can require additional analytical methods.
These may include structured scenario analysis, probabilistic treatment of uncertain inputs, faster production-ranging models, and independent engineering checks.
The appropriate approach depends on the decision, the available evidence, and the level of confidence required. Probability distributions should not be assigned arbitrarily simply to produce a statistical result.
The objective is to develop a defensible assessment of expected production, downside exposure, and the conditions under which different outcomes arise.
It also allows capital alternatives to be evaluated against more than one assumed future.
Does additional redundancy improve expected production? Does it materially reduce downside risk? Would a lower-capital design still perform satisfactorily across plausible project conditions?
A financial model tells us what production is worth. Integrated simulation and uncertainty analysis help establish what production is achievable and which decisions can improve the outcome.
For major mining investments, both are needed.
8. When is integrated mining simulation worth the investment?
Not every mining problem requires a sophisticated simulation model.
If a conventional engineering calculation can reliably answer the decision, it may be the most appropriate approach.
Simulation becomes valuable when significant financial consequences depend on system interactions, operating variability, or uncertainty that simpler methods cannot adequately represent.
For a greenfield project, this could involve validating production capacity, evaluating capital configurations, assessing ramp-up, or strengthening the investment case.
For an existing operation, it might involve a persistent production shortfall, changing bottlenecks, operating strategies, maintenance, or expansion decisions.
The most useful starting point is usually a specific question:
- What decision needs to be made, and what is its financial significance?
- What remains uncertain about the system's response?
- What existing engineering and operating information is available?
- What evidence would change the decision?
A focused storage or equipment problem may require a relatively small model. A major sanction decision might warrant a comprehensive integrated simulation and extensive uncertainty analysis.
We generally favour starting with the smallest credible model capable of supporting the decision. The modelling scope should reflect the complexity of the system and the consequences of getting the answer wrong.
For high-consequence decisions, credibility requires transparent assumptions, verification, validation, reproducible results, and appropriate treatment of uncertainty.
The analysis should be capable of withstanding scrutiny from engineering, operations, finance, and executive stakeholders.
Better decisions across the mine value chain
Some of the most valuable findings from our modelling work have challenged the assumptions behind proposed investments and operating strategies.
A governing bottleneck may move as ore characteristics and operating conditions change. Additional storage may offer almost no increase in throughput. A different operating strategy may create greater value than an equipment upgrade. Commercial service requirements can be just as consequential as infrastructure capacity.
And integrated analysis may reveal opportunities to defer substantial capital expenditure while maintaining production objectives.
These outcomes are rarely obvious from individual equipment specifications or static capacity assessments.
Understanding the complete operating system gives project teams a stronger basis for challenging assumptions, testing alternatives, and allocating capital.
Transforming the mine value chain means identifying the production, operating, and investment decisions that create the greatest value across the entire system.
Sometimes that involves building more capacity. Sometimes it involves changing how existing assets operate.
And sometimes the most valuable result is discovering what doesn't need to be built, or doesn't need to be built yet.
About the author
Bryan Monk, P.Eng., is the founder of MISIM Modelling and has more than 30 years of experience developing simulation models for mining, processing, logistics, and complex industrial operations. He is an INFORMS Franz Edelman Laureate and co-author of the published research on BHP's Jansen Potash Project.
References and further reading
- Bouffard, S. C., Boggis, P., Monk, B., Pereira, M., Quan, K., and Fleming, S. (2018). Discrete-Event Simulation Modeling Unlocks Value for the Jansen Potash Project. Interfaces, 48(1), 45–56.
- Bouffard, S. C. (2020). Benefits Realized and Lessons Learned from Modeling the Production Capacity of the Jansen Potash Project. Mineral Processing and Extractive Metallurgy Review, 41(1), 50–58. Published online in 2018.
- Wilson, R., Mercier, P. H. J., Patarachao, B., and Navarra, A. (2021). Partial Least Squares Regression of Oil Sands Processing Variables within Discrete Event Simulation Digital Twin. Minerals, 11(7), 689.
- Soto, I., Anani, A., and Córdova, E. (2022). A Discrete Event Simulation Approach for Mine Development Planning at Codelco's New Mine Level. Journal of the Southern African Institute of Mining and Metallurgy, 122(10), 549–559.
- Jones, W., and Gun, P. (2024). Train Timetabling and Destination Selection in Mining Freight Rail Networks: A Hybrid Simulation Methodology Incorporating Heuristics. Journal of Simulation, 18(1), 1–14. Published online in 2022.
- Dimitrakopoulos, R., Farrelly, C. T., and Godoy, M. (2002). Moving Forward from Traditional Optimization: Grade Uncertainty and Risk Effects in Open-Pit Design. Mining Technology, 111(1), 82–88.
