An evidence-led look at extended drain intervals in industrial and fleet lubrication: why the practice is more common than the manuals admit, what a wrong call costs per hour, and how to extend safely.
Ask any lubricant buyer whether they have ever pushed an oil beyond the drain interval printed in the OEM manual, and you will get an honest answer in private and a careful one in public. So we asked them anonymously instead.
Of the 114 plant engineers, fleet managers and procurement leads who voted in our poll, 61% said they run oils past the OEM drain interval as a matter of routine, while 39% stick rigidly to spec.
That split is something you will not find on Google, in a product datasheet or in any AI summary, because nobody publishes what they actually do behind the workshop door. So here is the honest breakdown: why it happens, what it costs when the maths is wrong, and how to extend a drain interval without betting the plant on it.
What "running past spec" actually means
An OEM drain interval is a conservative default. It is set for the worst duty the manufacturer expects the equipment to face, with a safety margin layered on top to cover customers who never sample their oil. For a plant running clean, well-monitored machinery, that default is often pessimistic by design.
That creates two camps. The first changes oil on the calendar or the hour-meter, no questions asked. The second extends on condition, meaning they sample the oil, confirm it is still fit for service and keep running until the data says otherwise.
Done with discipline, the second approach is recognised as best practice. Done on hope, it is a gamble dressed up as efficiency. The gap between those two outcomes is the entire subject of this article.
Why it happens
The pressure to extend is real, and it is rational.
Lubricant is a recurring cost with a recurring disposal bill attached, so every litre not drained early drops to the bottom line twice. Waste-reduction and sustainability targets push in the same direction, because used oil is a tracked waste stream. Modern synthetics genuinely outlast the intervals written for the mineral oils of a generation ago, so the manual frequently lags the fluid in the sump. Suppliers reinforce this with long-life performance claims. And there is a quieter driver: everyone assumes a competitor is already running leaner, so nobody wants to be the plant changing oil on a fixed schedule while the industry moves on.
Here is the part the datasheets leave out. The practice stays invisible because the wins are silent and the failures get quietly relabelled. A successful extension produces nothing to report. A failed one is written up as a "bearing failure" or a "gearbox failure," not as "we ran the oil too long." That mislabelling is precisely why there is no public data on how common extension really is, and why the true rate is almost certainly higher than anyone admits.
What it costs when it goes wrong
Start with the asymmetry, because it is the whole argument. The oil is the cheapest fluid in the plant. The failure it guards against is one of the most expensive events the plant can have.
When an oil is run too far, it degrades in predictable ways. Anti-wear and extreme-pressure additives deplete, the base number falls as acids build, viscosity drifts out of grade, and oxidation produces varnish and deposits. The lubricating film thins, metal begins to touch metal, and wear accelerates toward failure.
That failure lands in the most costly category of all. Bearing-failure data needs to be quoted with its scope intact, or a reliability engineer will rightly correct you. SKF's widely cited breakdown attributes 36% of premature bearing failures to lubrication, alongside fatigue at 34%, mounting and handling at 16% and contamination at 14%. Using a broader definition that folds in lubricant selection, application, contamination and degradation, industry estimates commonly place 60 to 80% of bearing failures in the lubrication-related bucket. Both framings are defensible. Only the unscoped version is wrong.
Then comes the downtime, and this is where the numbers turn serious. The figures below are converted at approximately £0.76 and €0.86 to the US dollar (June 2026) and rounded.
| Benchmark | Per hour (USD) | Per hour (GBP) | Per hour (EUR) |
|---|---|---|---|
| Manufacturing average (Aberdeen) | ~$260,000 | ~£198,000 | ~€224,000 |
| Cross-sector median (ABB Value of Reliability survey) | ~$125,000 | ~£95,000 | ~€108,000 |
| Automotive assembly (Siemens True Cost of Downtime) | ~$2,300,000 | ~£1.75 million | ~€1.98 million |
Set that against the savings. A drum of oil kept in service a little longer saves a few hundred. The stoppage can trigger costs six figures an hour in any currency on that table. Extending without data is not a cost-saving measure. It is an uninsured bet where the downside dwarfs the upside.
How to extend safely
None of this is an argument against extending. It is an argument for extending the evidence. The plants that do this profitably share the same habits.
They run an oil-analysis programme rather than the occasional sample. The panel that matters is viscosity, acid and base number, wear metals, particle count, additive elements and water content. One sample tells you almost nothing. They trend the results over time, because a value heading the wrong way matters more than any single reading, and they set condemning limits in advance, agreed with both the OEM and the lubricant supplier, so the decision to drain is made by the data and not by a feeling.
They take sampling seriously. Same point, same operating conditions, clean technique, every time. A contaminated sample produces a confident wrong answer, which is worse than no answer.
They control contamination at the source. Better filtration, sealed breathers and clean top-up practice extend oil life more reliably than the choice of oil itself, and they keep the film doing its job. And they document the decision, so an extended interval is an evidence-led, defensible call rather than an unrecorded gamble that only surfaces at the failure investigation.
Crucially, they treat their supplier's technical team as part of the programme. An extension backed by analysis and a clear set of limits is a service the supplier can stand behind. An extension based on hope is one nobody will defend when the bearing lets go.
The real story in the data
The split in our poll, 61% extending against 39% holding to spec, is not the headline on its own. The headline is what sits behind it. The industry is already running leaner than the manuals admit, but the discipline behind that decision varies enormously.
The operations that extend on trended analysis save money quarter after quarter. The ones that extend on optimism eventually pay for it at around $125,000 (£95,000 / €108,000) an hour and upward, before the cost of replacement parts, expedited freight and a missed delivery is counted. The difference between the two is never the oil in the drum. It is the data behind the decision to leave it there.
Want to know where your own drain intervals sit against the rest of the field, and whether your analysis programme can support a safe extension? Contact us
Sources
- SKF, premature bearing-failure cause breakdown (lubrication, fatigue, mounting and handling, contamination).
- Aberdeen, benchmark on the average cost of unplanned manufacturing downtime.
- ABB, Value of Reliability survey of more than 3,200 plant maintenance decision-makers (cross-sector median downtime cost).
- Siemens, True Cost of Downtime research (automotive assembly downtime cost).
- Currency conversions calculated at approximately £0.76 and €0.86 to the US dollar, June 2026.
