Blog / Metrics
Metrics · Guide

OEE in Mining: Fleets, KPIs and Where It Fits

OL By OEE Lab Editorial |Updated 2 November 2026 |6 min read
Short answer

OEE can be calculated for mining equipment, but mobile fleets already run on a Time Usage Model with KPIs such as physical availability and use of availability. Use those as your main fleet numbers. OEE adds value as a combined view that includes cycle speed and payload, and it fits crushing and processing plants more naturally than trucks.

Mining had its own language first

A haul truck has no line and no scrap bin, and its rate changes with the route. Mines track fleets through a time model that splits the whole calendar into categories, and fleet systems report against it.

So the honest answer is that OEE is a second lens in mining, not a replacement. The Global Mining Guidelines Group framework (2020) lists a standard OEE definition for surface mining as future work, on printed page 11, so we found no agreed version.

The measures mines already use

GMG's guideline defines these KPIs for mobile surface equipment, on printed page 10. It offers them as recommendations, not a formal standard.

GMG KPIFormulaClosest OEE idea
Physical availabilityAvailable time / scheduled timeAvailability, but counting only downtime
Mechanical availabilityOperating time / (operating time + downtime)A maintenance view with no direct match
Use of availabilityOperating time / available timeStandby, which OEE would log as stops
Operating utilizationOperating time / scheduled timeClose to OEE availability
Asset utilizationOperating time / calendar timeClose to the time part of TEEP
Effective utilizationWorking time / scheduled timeAvailability after operating delays
All of these are time ratios. None of them looks at speed, payload or quality.

That last line is the gap. A truck can score well on every row while hauling light and running slow cycles.

Downtime, standby and delays

In GMG terms, downtime means the machine is not in a condition to work. Standby means it could work but isn't, and GMG's external standby examples include a shovel out of digging and a workforce shortage.

Operating delays are temporary stops inherent to the work. A factory OEE would treat all three as losses, and the mine's model keeps them apart because they have different owners.

Weather, blasting clearances and shift change need a site rule. Write down which category each goes in, the same way a plant builds reason codes.

The ideal rate: route and payload

For a truck or shovel, the ideal is not a machine rating alone. It is the planned cycle time on that route and the planned payload.

A longer haul as the pit deepens changes the ideal rate, so reset it by route or period. Otherwise a performance fall may just be geometry.

What stands in for quality

A loader moves rock, and it does not make scrap in the usual sense. Mohammadi, Rai and Gupta (2017) replace the quality factor with bucket capacity utilization for bucket-based excavating, loading and transport equipment, and they tested this on a dragline at a large surface coal mine.

Elevli and Elevli (2010) do much the same for shovels, setting quality equal to the bucket filling factor. GMG also notes that some operations classify actual productivity against forecast as "quality loss".

True ore quality, such as dilution or grade, depends on the deposit and mine planning as well as the machine. We would keep it out of a fleet OEE and report it separately.

Worked example: one haul truck, one month

Example: a truck in a 720-hour month, with a plan of 220 tonnes per load and a 30-minute cycle, or 440 tonnes per hour. It has 24 hours unscheduled, 96 hours of downtime, 60 hours of standby and 30 hours of operating delays.

Calendar 720 h Scheduled 696 h (24 unscheduled) Available 600 h (96 downtime) Operating 540 h (60 standby) Working 510 h (30 delays)
Each step down removes one category of time. Hypothetical numbers.

The fleet report would say physical availability is 600 / 696 = 86.2%, use of availability is 540 / 600 = 90.0% and asset utilization is 540 / 720 = 75.0%. All three look respectable.

Now add the two things those numbers skip. Actual cycles took 33 minutes, giving 927 loads in the 510 working hours, and the average payload was 207 tonnes, so the truck hauled 191,889 tonnes.

Availability: 600 / 696 = 86.2%
Utilization: 510 / 600 = 85.0%
Speed: 30 / 33 min = 90.9%
Payload: 207 / 220 t = 94.1%
Product of the four = 62.7%
Check: 191,889 t / (440 t/h x 696 h) = 62.7%
Four factors in the Mohammadi style, on scheduled time. The direct check agrees within rounding.

The same tonnes against the calendar give 191,889 / (440 x 720) = 60.6%. Always state which time base you used, because it changes the result.

Calendar time or scheduled time?

Elevli and Elevli favor a calendar-time basis for shovels and trucks, because mining is so capital-intensive. Mohammadi and co-authors note that a loading-time basis gives higher values than a calendar basis.

In the Elevli shovel example, both bases give the same production, and only the percentage moves. A calendar-based figure is closer to TEEP than to factory OEE, so never compare one with the other.

Trucks, shovels or fixed plant: where it earns its place

  • Poor fit: a fleet KPI to replace the time model. Keep the KPIs your fleet system already reports, and add a speed and payload view on top.
  • Poor fit: one truck judged alone. A truck queues because of the shovel, the road or dispatch, so read shovels and trucks together.
  • Poor fit: comparing mines. Haul distance, rock and climate differ, and we found no measured OEE benchmark for mining equipment across mines.
  • Better fit: fixed plant. Crushing, screening and conveying lines have a rate in tonnes per hour and a clear run state, so the continuous process method applies almost directly.

GMG's framework covers mobile surface equipment, and it names underground and fixed plant as future work. Check which definitions your site uses before you compare anything across areas.

How to start

Keep your time model as it is, because rebuilding it would break years of reporting. Pick one fleet or one crushing line, and add just the two missing factors.

For fixed plant, run state and tonnage from a belt weigher are usually enough to begin. Crusher troubleshooting helps once stops start showing a pattern, and the MTBF and MTTR calculator sizes repeat failures.

See it live in Fabrico

Fabrico is a light MES for OEE on lines, so it suits fixed processing lines better than haul fleets. It takes stops, speed and counts live from PLCs or IoT sensors, and operators tag the cause at the line. See how Fabrico measures OEE on lines.

What people ask next

What is OEE in the mining industry?

It is availability x performance x quality adapted to mining, often with payload or bucket fill in place of quality. Some papers use a calendar-time basis, so the result reads closer to TEEP.

What is the difference between physical availability and mechanical availability?

GMG defines physical availability as available time over scheduled time. Mechanical availability is operating time over operating time plus downtime, which isolates the effect of maintenance. Use the uptime calculator for the basic time ratios.

What is a Time Usage Model?

It is a framework that sorts all calendar time into categories such as scheduled time, downtime, standby and operating delay. It gives a mine consistent inputs for its KPIs, much as a loss tree does for a plant.

Is there a good OEE figure for haul trucks?

We found no measured benchmark for trucks. Elevli and Elevli propose over 77% for scheduled shovels, from 90% availability, 90% performance and 95% quality, but that is their suggested target. Build your own baseline first, as described in how to set OEE targets.

Can OEE work for a crusher or processing plant?

Yes, this is where it fits best. Use tonnes per hour as the rate and on-spec tonnes as good output, as shown in OEE for continuous processes.

See OEE per line and shift on your own machines

Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.

Book a Fabrico demo

← More OEE Lab guides · OEE vs TEEP →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES