9) The KPI Trap-How Mine Simulation and Flow-Based Thinking Turn Mine Plans into Bankable Reality

In the previous post, we looked at why so many mine plans break by Wednesday, and how Flow (TOC) changes the physics of the system so plans actually have a chance. We showed that when you stop “balancing capacity” and instead:

• Put the true constraint at the centre,

• Protect it with buffers and protective capacity, and

• Align KPIs to Flow rather than local utilisation,

You move from brittle, KPI-driven chaos to Superflow – higher tonnes, at lower cost/tonne, with less firefighting.

We also touched on stochastic planning. But the reality is that most mines still live on standard planning tools – Gantt charts, critical path, Excel, and a lot of pressure.

That doesn’t mean you have to stay stuck in the KPI trap – green dashboards, red tonnes. What you do need, though, is more than another piece of software.

We’ve found two things make the difference::

1. Mine digital twins (like Mine-Twin) act as a digital wind tunnel for Flow-based plans, and

2. Flow-based training and thinking give planners, supervisors, managers, and executives the mental model to use the wind tunnel effectively.

Simulation on its own won’t move behaviour. Training on its own won’t move tonnes. Together, they give you a way to change the system and prove the ROI across the whole business.

Why Planners Need More Than a Better Gantt Chart

If you sit in Technical Services, your life probably looks something like this:

• Targets arrive from above.

• You build a plan that looks feasible on paper.

• Variability and interdependence smash it in Week 1.

• Operations blame Planning, Planning points to gear, geology or labour.

Underneath sits the KPI trap:

• Each department is measured on its own numbers – metres, hours, utilisation, cost per unit.

• Anyone can be “green” while the constraint is starved or blocked, and the mine misses shipment.

When you suggest a Flow-based change – new or bigger buffer, different cut sequence, a bit of slack at a key upstream unit – you run into the next wall:

“We’re already under cost pressure. Prove that this buffer or extra unit will pay for itself.”

That’s not a spreadsheet argument. It’s a physics argument. And physics is where digital twins earn their keep.

What a Mine Digital Twin Really Is (for Non-Modellers)

A mine digital twin is not there to replace your planning system. Think of it as a dynamic model of how your mine actually behaves when the plan hits reality.

Typically, it captures:

• Layout – pits/benches or stopes/headings, ore passes, haul roads, stockpiles, plant.

• Fleets – trucks, loaders, LHDs, shovels, hoists, conveyors, key plant units.

• Rules – shift patterns, hot seats, blasting windows, maintenance logic, priorities.

• Variability – breakdown distributions, travel times, loading variability, and weather impacts.

You feed your existing plan into this model and let it play out over time. The twin shows where queues form, where the constraint is starved, and what happens to tonnes and cost as you tweak the design.

If Flow is the rulebook, the digital twin is the wind tunnel where you test those rules on your own mine before you start cutting steel.

Three Jobs the Twin Does for Planners and the Business

1. It Puts the “Where’s the Real Constraint?” Argument to Bed

On most sites, people disagree about the bottleneck:

• Ops says it’s the plant.

• Maintenance says it’s the hoist.

• Planners suspect it’s actually stope or bench availability.

With a twin, you don't need a religious debate; you just:

• Run the current plan,

• Track queues and starvation, and

• See which resource actually governs sustained Flow over weeks and months.

That clarity:

• Gives planners a solid basis for a Flow design,

• Helps operations understand why certain areas get priority, and

• Stops capital from being thrown at the loudest problem instead of the real one.

2. It Sizes Buffers and Protective Capacity with Numbers, Not Hope

Flow tells you to:

• Put a buffer in front of the constraint, and

• Give upstream/downstream units enough protective capacity to catch up after a bad patch.

The twin lets you test:

• Different buffer sizes and locations,

• Different fleet mixes and maintenance strategies, and

• The impact of modest extra capacity at key links.

Instead of “trust me, this works elsewhere”, you can say:

“For an extra $X in buffer and $Y in protective capacity, we reduce plant starvation events by about Z% and unlock A kt per year.”

Finance gets a real business case, not a war story. Maintenance sees how their choices play into Flow. Operations sees why some of their favourite local KPIs may have to change.

3. It Ranks Improvement Ideas by Impact on Stable Tonnes

Most mines have a long list of “good ideas” – new ore passes, more trucks, re-levelling, new layouts – but limited bandwidth and capital.

With a twin, planners can put those options against the same digital mine and see:

• Which ones genuinely move stable weekly tonnes at the constraint?

• Which ones are nice-to-have but low impact, and

• Which ones look good in isolation but actually create more problems by shifting the bottleneck.

That means Engineering effort, capital and management attention go to the few changes that really matter, which is precisely what your GM, CFO and crews want.

Two Short MineTwin Case Studies

Here’s what this looked like on real sites using MineTwin:

1. Strip coal – proving the value of a ROM buffer

2. At a large strip coal operation in South Africa, planners used MineTwin to answer a simple question: “Do we really need an ex-pit stockpile in front of the main conveyor, and if so, how big?” The twin replayed full cycles with realistic distances, cycle times and availabilities. Scenarios compared “no buffer” to different stockpile sizes and fleet mixes. Result: a buffer of roughly 10 kt ROM almost eliminated plan shortfalls – around a 90% drop in the risk of missing monthly tonnage – and required only a modest increase in fleet size to support the expansion.

3. Nickel portfolio – stopping unnecessary fleet purchases

4. For a portfolio of six underground nickel mines, MineTwin was used at the head office to test equipment requests. The twin showed, year by year, how many machines each site actually needed to deliver the approved plans. By redistributing existing gear and cancelling over-optimistic purchase requests, the team identified about US$262 million in avoided fleet spend while still hitting production targets.

In both cases, the digital wind tunnel didn’t replace planning – it de-risked it, and gave executives a clear ROI story.

Why Simulation Alone Is Not Enough

Here’s the catch: seeing the physics is not the same as thinking in Flow.

Most supervisors and managers have spent their careers in the KPI trap:

• Keep “my” utilisation high.

• Keep “my” cost per unit down.

• Hit “my” department’s targets.

The idea of deliberately unbalancing the system – giving one area slack, building buffers “that don’t produce tonnes” – will feel wrong until people understand the logic of Flow and how Flow KPIs replace the old measures.

Which is exactly why we start with Flow-based training:

• Planners need a solid grasp of constraint, buffer and protective capacity thinking so they know what to test in the twin.

• Supervisors and control room operators need to see why some of their old KPIs are a trap, and how Flow KPIs actually make their lives easier.

• Executives need a simple story that links Flow, simulation results and business outcomes – tonnes, cost, safety, and risk.

Simulation sessions become part of the training:

• You teach the Flow concepts in plain language.

• Then you show their own mine in the twin, breaking under the old rules and stabilising under Flow rules.

• People see their mental model change in real time.

Without that conceptual shift, a digital twin risks becoming just another niche technical tool. With it, the twin becomes a shared “single source of physics truth” across Planning, Ops, Maintenance and Finance.

A Practical Roll-Out: Training + Flow + Twin

If I were starting on a new site tomorrow, I’d do it like this::

1. Start with Flow training Short, targeted workshops for planners, supervisors and key managers. Focus on constraint, buffers, protective capacity, and the KPI trap. Introduce Flow KPIs and the idea of a Flow Room.

2. Build a focused digital twin. Choose one value chain and bottleneck (e.g. “Pit → ROM → CHPP” or “Stopes → Hoist → Plant”). Work with a vendor to build a simplified MineTwin model and calibrate it with recent data.

3. Run Flow scenarios in the twin Current “balanced” state. One or two Flow-aligned designs (buffer size/location, protective capacity changes). A couple of popular “good ideas” you want to test.

4. Use the twin in cross-functional sessions Bring Ops, Maintenance, Planning and Finance into the same room. Show how each scenario changes constraint stability, tonnes, cost and risk. Agree on which design to implement and which KPIs need to change.

5. Implement and stabilise via a Flow Room Use Flow KPIs and simple daily control to keep the chosen design on track. Use the twin periodically to test new ideas or major changes before touching the real mine.

   
   

Closing the Loop: From Intractable Problems to Testable Designs

The mines I see aren’t short of intelligence or effort. They’re trapped in systems where KPIs and physics are misaligned:

• Departments are optimised.

• The system is brittle.

• Plans have the shelf life of milk.

Part 1 of this series was about fixing the physics with Flow, so the mine can absorb shocks and still deliver.

This second step is about making that change understandable, sustainable and bankable:

• Flow training gives people a new mental model – Flow KPIs instead of local-efficiency KPIs.

• Flow implementation adds the buffers and protective capacity that turn brittle chains into robust ones.

• Digital twins like MineTwin act as a wind tunnel, proving in your mine that the changes will lift tonnes, improve safety, reduce the cost per tonne, and avoid pointless capex.

• 
  

Our experience over two decades is consistent: once resilience is built in, Flow becomes predictable, plan shelf life increases dramatically, output, cost and safety improve together, and management attention is freed from constant firefighting to real improvement work.

For planners, it means stepping up from “spreadsheet victim” to architect of how the whole system runs.

For everyone else, it means fewer nasty surprises, fewer heroic recoveries, and more weeks where the mine quietly does what the plan says and still beats the numbers.

In the end, that’s what the digital wind tunnel plus Flow-based training is for: taking the ‘intractable problems’ we all complain about and turning them into designs we can test, argue about, and repeat on purpose.