Tiegel home Talk to a metallurgist

Plant availability, utilisation and OEE: accounting for lost hours and tonnes

Definitions and calculation of availability, utilisation, MTBF, MTTR and OEE for a processing plant, with a time model and worked example.

· 4 minute read · By

A plant generates revenue only while it operates at full rate producing on-specification product. Every hour lost to breakdowns, planned maintenance, waiting for ore or reduced rate is lost production, and these losses must be measured consistently before they can be reduced.

Availability measures the proportion of time the plant could operate, utilisation the proportion of time it did operate, and performance its rate while operating. Together they show where production was lost. Availability and utilisation appear in every monthly report, but they are calculated differently from site to site, which makes comparisons and targets meaningless unless the definitions are clear. This article sets out a simple time model, the standard calculations, and their use in identifying lost production.

The time model

Time model for one month720 calendar hours split into 36 planned maintenance, 96 unplanned downtime, 24 standby and 564 running.Running 564 hRunning 564 hStandby 24 hUnplanned 96 hPlanned 36 hStandby 24 hUnplanned 96 hPlanned 36 hCalendar time: 720 h
Where a month's hours went in the ramp-up example. Availability counts standby as available time; utilisation counts only running time. Illustrative figures.
CategoryMeaningExample hours
Calendar timeAll hours in the period720
Planned maintenanceScheduled shutdowns for maintenance36
Unplanned downtimeBreakdowns and trips96
StandbyPlant could run but did not: no ore, no power from the grid, downstream full24
RunningTreating ore564

Availability, utilisation, MTBF and MTTR

Availability=calendar hours−maintenance downtimecalendar hoursUtilisation=run hourscalendar hours% of nameplate=tonnes treatednameplate rate×calendar hours×100

where

maintenance downtime
planned plus unplanned hours when the plant could not run because of equipment (h)
run hours
hours the plant actually treated ore; also excludes standby, such as no feed or no power from the grid (h)
nameplate rate
design throughput (t/h)

Availability = (720 − 36 − 96) ÷ 720 = 81.7%, and utilisation = 564 ÷ 720 = 78.3%. Standby is counted as available time because the plant itself was able to operate.

Availability=MTBFMTBF+MTTROEE=Utilisation×Performance×Quality

where

MTBF
mean time between failures (h)
MTTR
mean time to repair, including waiting for parts and people (h)
Performance
average rate while running ÷ design rate
Quality
fraction of production that meets specification; for a concentrator, often the fraction of time the grind or product grade is on target

If the plant failed 8 times in the month with 564 running hours, the MTBF is about 70 hours. If each failure took on average 4 hours to rectify, availability considering failures alone would be 70 ÷ (70 + 4) = 95%. Raising MTBF (fewer failures, through better maintenance and design) and reducing MTTR (faster repairs, through spares, procedures and access) both increase availability.

Overall equipment effectiveness

Overall equipment effectiveness (OEE) originated in manufacturing. Applied to a concentrator for the same month, utilisation is 78.3%; performance, at an average of 260 t/h while running against a 300 t/h design, is 86.7%; and quality, the proportion of production within specification (for example, the fraction of running time with the grind on target), is 98%. OEE = 0.783 × 0.867 × 0.98 = 66.5%.

Time and rate measures for the monthAvailability 81.7%, utilisation 78.3%, performance 86.7%, quality 98%, overall equipment effectiveness 66.5%.0255075100Availability: 81.7%81.7%AvailabilityUtilisation: 78.3%78.3%UtilisationPerformance (rate): 86.7%86.7%Performance (rate)Quality: 98.0%98.0%QualityOEE: 66.5%66.5%OEEPer cent
OEE multiplies utilisation, rate performance and quality, so modest losses in each compound to a large overall loss. Illustrative figures.

Three measures each in the 80s and 90s multiply to two-thirds. The plant treated 146,640 t against a possible 216,000 t at design rate around the clock. OEE shows the size of the opportunity and identifies the largest lever.

Loss analysis

Every stoppage should be logged with its start, end, equipment, cause and category. Most control systems record trips automatically, and operators add the cause. Each month the losses are ranked in a Pareto chart, in hours and in tonnes; the top two or three causes usually account for half the loss, as shown in the example in the commissioning guide.

Rate losses should be separated from time losses. A plant that runs 95% of the time at 85% of design rate has a throughput problem, often related to ore hardness, grind or a bottleneck, rather than a maintenance problem.

Definitions should be agreed with finance, maintenance and operations and kept stable, so that trends are meaningful. Performance should also be tracked by circuit, because crushing, grinding and leach have different availabilities, and buffers such as stockpiles decouple them.

Glossary

Availability
The share of calendar time the plant was able to run.
Utilisation
The share of calendar time the plant actually ran.
Standby
Time the plant could run but did not, for external reasons.
MTBF
Mean time between failures.
MTTR
Mean time to repair, including waiting for parts and people.
OEE
Overall equipment effectiveness: utilisation × performance × quality.
Performance
Average rate while running divided by design rate.
Pareto chart
A bar chart of causes sorted from largest to smallest.

Frequently asked questions

What is the difference between availability and utilisation?

Availability counts the time the plant could run, including standby. Utilisation counts only the time it did run, so utilisation is always equal to or lower than availability.

How are MTBF and MTTR calculated?

MTBF is running time divided by the number of failures. MTTR is total repair time divided by the number of failures.

Is OEE useful for a processing plant?

Yes, as a means of combining time, rate and quality losses into one figure and identifying the largest. Quality should be defined appropriately for the plant, for example as time with the product on specification.

Why do availability figures differ between sites?

Sites define categories such as standby, planned downtime and external delays differently. Figures should be compared only when the definitions match.

Discuss your plant with a metallurgist

Tiegel is a Perth metallurgy company. We help with testwork, plant optimisation, metal accounting and investigations, and we build the Tiegel Suite software.

Talk to a metallurgist

References and further reading

  1. Nakajima, S. (1988). Introduction to TPM: Total Productive Maintenance. Productivity Press.
  2. Applying the AMIRA P754 Code of Practice for Metal Accounting. AusIMM MetPlant 2008.

Worked examples and charts marked illustrative use invented figures, not data from any operation.