7) The KPI Trap: How Traditional KPIs Shrink the Safe Production Envelope

1.      Introduction

Over the last two decades, there has been a substantial decrease in mining fatalities. This trend, however, has levelled out and stubbornly resists further improvement. Automation and removing humans from dangerous situations remain viable levers, but justifying the cost can be a constraint for smaller organisations. Figure 1 illustrates the decline and levelling out of fatalities as well as a reduction in mining productivity over the same time period.

Figure 1: Productivity vs mining deaths- safetydifferently.com

In over 20 years of flow-based consulting and more than 90 productivity improvement projects, we have noticed a recurring pattern. As soon as the production flow increased and stabilised, safety performance improved visibly (as measured by safety prizes and safety statistics).

This article will argue that the core problem in improving safety is not the commitment of leaders and workers (or the quality of safety programs), but the way we set up and use our KPIs (planning and measuring performance). Traditional KPIs create flow instability, shrinking the safe production envelope. In contrast, Flow KPIs, built around the fundamental constraint, stabilise the system and expand that envelope – allowing safety and production to improve together.

 Acceptance is growing (personal discussions with safety professionals and mining executives) that further improvement cannot be achieved without adjustments to the way we plan and incentivise production. The days of running “safety programs” on one track and “production initiatives” on another are coming to an end. If the system that delivers tonnes is unstable, overloads attention, and requires constant replanning, then (in the absence of full automation and the removal of all human interactions) no amount of rules, campaigns, or investigations can make it safer.

The lived experience on many sites looks like this (Figure 2):

• ROM production sits 20–25% below budget.

• Daily variation is high.

• Supervisors are constantly re-planning work (often resorting to WhatsApp groups).

• Everyone knows month-end will mean a frantic rush to “make the numbers”.

Figure 2: Production output versus budget

2.      Complications arising from the safety productivity tradeoff

At a Queensland parliamentary enquiry, senior leaders spent days reviewing their safety initiatives and efforts to improve safety. These laudable efforts were then contradicted within 30 minutes by a Site Safety and Health Representative who presented examples of production being prioritised over safety. These included supervisors knowingly ignoring procedures near the end of the month to protect output. Sentis' State of Safety Culture in Mining (2022) report shows a similar pattern: high production pressure, under-reporting of incidents, and a negative safety culture across many sites.

So despite genuine intent and significant investment, many mines still live the trade-off between safety and tonnes.

3.      How traditional KPIs move Rasmussen’s boundary the wrong way

Rasmussen’s model visualises operations as taking place inside an envelope bounded by three pressures:

• Economic failure (insufficient production, excessive costs)

• Unacceptable workload (people simply cannot cope)

• Unacceptable safety risk

 In an ideal world, your mine would operate comfortably inside these boundaries. In reality, the system is constantly pushed and pulled by cost pressure, production targets, and day-to-day variability. The “safe zone” is only as large as your system’s stability allows.

Figure 3: Rasmussen and Cook on acceptable performance boundaries

 Most mines are still planned and operated using a balanced capacity approach:

• Each department is expected to run at high utilisation.

• KPIs reward local efficiency – hours, utilisation, cost per tonne – in each silo.

• The underlying assumption is: if every part is efficient, the whole mine will be efficient.

 But a mine is a highly variable, interdependent system. When you drive every area hard with efficiency KPIs:

• Any disturbance (breakdown, quality issue, contractor delay, weather) ripples through the chain.

• There is no protective capacity to absorb it.

• The result is instability – frequent plan changes, excessive queueing, starvation at critical resources, and constant firefighting.

 In Figure 2, this appears as a jagged, sawtooth actual output line well below the budget line. Even if you work harder, you will still not get the tonnes you planned.

In Rasmussen’s terms, this instability moves the productivity–safety boundary to the right:

• The system becomes harder to control.

• The safe operating space shrinks.

• Leadership is forced to cut back production or add more controls to stay on the safe side. (see Figure 1)

 The paradox is brutal: by chasing local efficiency through traditional KPIs in a balanced-capacity system, you end up with lower average production, higher costs, greater pressure on safety, and demotivated workers. You are closer to the cliff edge than you need to be.

4.      When Flow KPIs stabilise the system

Flow KPIs tune and refocus your existing MOS/BOS/CI system:

• The tiered meetings and dashboards stay.

• What changes is the “first page” – the vital few metrics now tell the story of constraint flow, buffer health, and stability, not silo efficiency.

• The MOS stops amplifying firefighting and starts reinforcing stable, safe Flow.

By applying Flow KPIs we achieve

• A higher average output

• Much lower day-to-day variation 

Figure 4: The effect of Flow principles on production output and variability

 Once Flow KPIs are established and the constraint is properly protected:

• Plans have a longer shelf life; operators are not shuffling work on the fly.

• Supervisors and contractors are no longer required to perform end-of-month heroics.

• Leaders regain the time and attention to spot weak signals and run effective learning teams.

• Workers experience “doing it properly” as the simplest way to get the job done.

From Rasmussen's perspective, smoother Flow reduces the attention and coordination load on leaders and crews. The productivity–safety boundary shifts outward – meaning you can operate at a higher average output while staying in a more comfortable safety margin.

4.      What this unlocks for Safety-1 and Safety-2

Many mines now blend elements of:

• Safety-1 – controls, procedures, investigations, audits

• Safety-2 – learning from everyday work, simplifying tasks, enabling adaptation

 Both are essential – and both rely on a system that is not in constant crisis.

When Flow is unstable

Under a balanced-capacity Standard KPI-driven regime:

• Safety-1 suffers Pre-starts and risk assessments are compressed into box-ticking. Procedures are seen as obstacles when the plan changes for the third time that shift. Investigations become shallow paperwork exercises to keep up with the queue.

• Safety-2 is strangled Learning teams and “work as done” conversations are cancelled because “we’re in recovery mode”. Improvement ideas from crews are parked; there is no bandwidth to experiment. Supervisors spend their day replanning and reprioritising instead of listening and coaching.

The result: workers are asked to be both safe and more productive in a system designed for neither.

 Safety-1 practices become realistic rather than aspirational. Safety-2 stops being a side project and becomes the normal way of improving work. The mine produces more tonnes with less pressure, and people feel less compelled to “bend the system” just to keep up.

6. What mine executives can do next

If you suspect your site is caught in this KPI trap, explore three lines of enquiry:

• How are we planning capacity? Are we still managing to a balanced-capacity model? Are we expecting every department to operate at high utilisation? How often does a disturbance in one area trigger site-wide chaos?

•  What are our “vital few” KPIs really optimising? Do our key metrics describe Flow at the constraint, or local efficiency everywhere? How many KPIs are directly linked to constraint tonnes, stability, and buffer health? How many drive departmental self-protection and conflict?

•  Where are we in Rasmussen’s model today? Are we cutting back production, adding layers of control, or relying on worker heroics to stay inside the safety boundary? Or are we operating in a calm, predictable regime where higher production and better safety genuinely go hand in hand?

 Redesigning your KPI set around Flow is a high-impact, “consultant-lite”, low-interference intervention. Done well, it transforms both safety and productivity without a massive program or new technology layer.

Conclusion: Expanding the safe production envelope

If your site is living with constant re-planning, end-of-month pressure spikes, and you worry about how close you are to the safety boundary, the root cause is unlikely to be a lack of leadership commitment or careless workers (these are outcomes of system design). It is more likely that a balanced-capacity, local-efficiency KPI regime is driving instability into the system and pushing you closer to Rasmussen’s productivity–safety boundary than you need to be.

Flow KPIs, anchored on the true constraint, reverse this effect. By deliberately building protective capacity away from the constraint and managing a small set of flow-based indicators, you:

• Increase average output – because the constraint runs more consistently.

• Reduce operational pressure – because plans hold and “heroics” are no longer required.

• Improve safety – because leaders and crews regain the time and attention to coordinate the job properly and learn from normal work.

 In this more stable environment, both Safety-1 and Safety-2 become more effective. Safety is no longer something you “trade” against tonnes; it becomes a property of a well-designed flow system.

As a practical first step, ask your team to run a 90-day experiment:

1. Agree on the one constraint for the next three months.

2. Define 3–5 Flow KPIs that describe its throughput and stability (tonnes, buffer health, disruptions, plan changes).

3. Make these the first page of your daily and weekly performance story.

You don’t have to choose between tonnes and safety. By shifting the productivity–safety boundary outward with Flow KPIs, you create the space for both to improve – together, and on purpose.