Long service intervals are supposed to be the thing quietly destroying modern engines. We went looking for the measurements behind that argument. What the record shows is narrower and more interesting: the interval stopped being a margin and became a specification, and the head-to-head wear comparison the popular claim depends on does not appear to have been published.
- Applies to
- Cars on 15,000 km / 12-month service schedules
- Best evidence
- Used-oil analysis over single intervals
- Not measured
- Engine wear at short vs long intervals, same engine
- Australian change
- 91 RON sulfur limit fell 150 to 10 mg/kg, 15 Dec 2025
- Usually binds first
- The 12-month limit, not the kilometres
The oil-life number is a model, not a measurement
In US Patent 4,742,476, "Automatic engine oil change indicator system" (filed 27 January 1986, issued 3 May 1988), General Motors set out an oil-life indicator in which claim 1 computes an "effective usage amount" from a measure of actual engine usage and a penalty factor, accumulates those amounts, and signals an oil change once the running total hits a preset value; claim 3 specifies that usage measure as engine revolutions multiplied by the penalty factor. The patent states that the penalty factor is "determined as a sole function of the engine oil temperature as it relates to empirically determined data without regard to the engine load".
Read that carefully, because the useful point is a narrow one. The system does take an oil measurement - temperature, from an oil temperature sensor. What it does not sense is any property describing the oil's actual condition: viscosity, total base number, contamination, dielectric constant. The remaining life it displays is a modelled estimate built from empirically determined data, not a reading off the sump. The accompanying validation work was published as SAE 870403 in 1987.
That is the pattern in one component. The tolerance came out; a specification went in. And a specification holds until the thing it assumed stops being true. We could not reach any primary validation data showing how closely monitors of this family track measured oil condition in service, so we are not going to tell you they work well or badly - we can only tell you what they measure.
What the oil looks like when an interval ends
In a single conference-proceedings study, Tashev and colleagues sampled one car's SAE 0W-20 at its first scheduled oil change, at about 13,000 km on a new petrol engine. The oil had thinned roughly 17.5 per cent (8.5 to 7.01 mm2/s at 100 C), which the authors attribute to shear of the viscosity modifiers rather than dilution, since no fuel was detected. Its total base number measured 3.7 mgKOH/g, about half the 7 to 9 mgKOH/g the authors cite as typical for fresh oil of that class - though no fresh-oil TBN was actually measured, so the loss is inferred against a literature figure, not weighed before and after. Their conclusion was that the oil still "maintains adequate lubrication and protective properties", with copper, iron and aluminium at levels characteristic of engine running-in rather than abnormal wear.
Two samples, one vehicle, one normal-length interval, on a brand-new engine. The authors say themselves the results are indicative of general trends rather than statistically generalisable. It is not evidence about extended drains or long-term engine life, and we are citing it as what it is.
The chemistry underneath is better established. In a 2019 bench study by Nagy, Knaup and Zsoldos (Szechenyi Istvan University and Audi Hungaria), three light-duty diesel engines on a test bed running one boron-containing SAE 0W-30 over 250-hour drains showed total base number falling and total acid number rising consistently with oil hours, at correlation coefficients around 0.8 and independent of duty cycle. Both also correlated with friction and wear measured on a reciprocating tribometer, but with lower significance - friction tracked soot and boron content more closely than TAN or TBN, and the wear model was weak, with prediction error above the test's own repeatability. The authors assumed linear relationships without testing for non-linearity, and state their models apply to boron-containing oils in these bench-tested engines, not to arbitrary oils or real-world driving.
The comparison we could not find
Here is the hole at the centre of the whole argument. We could not locate, at any primary source, a study measuring engine wear head-to-head at a short versus a long drain interval in the same engine family under the same duty cycle. Every wear result we reached is either a bench tribometer test on used oil, or wear metals from a single interval with no comparison arm.
That means any sentence of the form "engines on 15,000 km intervals wear X per cent more than engines on 5,000 km intervals" is unsupported at source - in either direction. It also means several famous claims should be treated as unproven rather than merely disputed. We found no primary source defining severe-service qualifying conditions and counting how many owners meet them without realising. We found no manufacturer statement saying European long-life intervals are shortened for Australia because of our fuel or heat. And we found no validated accuracy figure for any oil-life monitor against measured oil condition.
One document is worth naming precisely, because it is routinely dragged into this argument and does not belong there. A December 2018 American Honda dealer message, filed by the US regulator as manufacturer communication MC-10152402-0001, told service managers about a product update campaign covering certain 2017-18 CR-V vehicles with the 1.5-litre turbo engine sold in certain cold-weather US states, addressing engine oil dilution attributed to "various control unit software/threshold settings". The documented repair was to reset the maintenance minder, "replace only the engine oil and the A/C control unit (if applicable)", clear fault codes, update the PGM-FI and TCU software, reset throttle position learned values and run the PCM idle learn procedure. The document states no warranty terms, and says nothing about oil change intervals or engine life. The mechanism it concerns - fuel entering the sump during cold short trips - is separately documented in the engineering literature, where one SAE study measured dilution reaching roughly 9 per cent of crankcase oil under simulated driving cycles, with a matching loss of viscosity.
The evidence that cuts against this article
The strongest single result runs the other way. In road tests of three civilian cars on four oils over 5,000 km of urban stop-start driving - the duty cycle usually named as the most damaging - the oil retained more than half its alkaline reserve, kinematic viscosity at 100 C changed by less than 16 per cent, oxidation onset temperature stayed above 210 C in every used sample, and four-ball wear scar diameters and friction coefficients of the used oils "did not increase significantly". If 5,000 km leaves that much headroom, the implication is that 5,000 km was conservative and the room to extend it was real.
Three more facts sit on that side of the ledger. Oil specifications tightened rather than loosened: API SP with ILSAC GF-6 became licensable on 1 May 2020, and the American Petroleum Institute describes it as introducing "seven new tests, including a chain wear test and a new test for very low-viscosity engine oils", with one test specifically addressing low-speed pre-ignition; Infineum records that four existing engine tests were replaced and three new ones introduced, with the Sequence VIII carried over unchanged. The fuel-sulfur difference that gets raised against Australian conditions is shrinking rather than growing, since regular 91 RON dropped from a 150 mg/kg sulfur limit to 10 mg/kg for all grades after 14 December 2025. And the Australian fleet is not dying young in the extended-interval era: the ABS Motor Vehicle Census counted 20.1 million vehicles at an average age of 10.6 years at 31 January 2021.
Even the industry's own warning was not the popular one. Mercedes-Benz argued at SAE in 1995 that further lengthening of fixed intervals "could seriously compromise the durability of engines in the overall vehicle population", particularly for drivers whose usage differs from the average, and that the answer was condition-based scheduling (SAE 951035). It then shipped a flexible system reported to run 15,000 to 30,000 km, or one to two years, conditioned explicitly on premium oil quality (SAE 981443, 1998). The complaint was about fixed numbers, not about long ones.
What genuinely varies in Australia
Fuel is the documented difference, and it is historical. Under the Fuel Quality Standards (Petrol) Determination 2024, 91 RON was legally permitted up to 150 mg/kg sulfur and 95 RON up to 50 mg/kg until 14 December 2025, against a 10 mg/kg limit now applying to every grade. So a European-derived interval was run here for years on regular unleaded permitted to carry fifteen times the sulfur the same standard now allows. Nobody has published what that difference cost an engine, and the factor is now shrinking - this is a story about the last fifteen years, not the next fifteen.
Second, the kilometre figure mostly does not apply. The ABS Survey of Motor Vehicle Use put average passenger-vehicle travel at 11.1 thousand kilometres over the 12 months to 30 June 2020, a COVID-affected period. At that rate a 15,000 km allowance takes about sixteen months, so the 12-month limit arrives first for a typical private owner. The average hides the split that matters: high-kilometre regional and towing users hit the distance limit, while low-kilometre urban owners sit on the time limit with the harshest trip pattern.
Third, what Australians buy. Isuzu UTE Australia publishes a free 3-month or 3,000 km inspection plus five scheduled services to 5 years or 75,000 km for 24MY and later D-MAX and MU-X - a 15,000 km or 12-month interval on a diesel 4x4 ute and SUV sold for towing and unsealed-road work. The same page notes that items such as fuel filter replacement are "not specified by a particular distance or time as each vehicle will reach the appropriate timing differently", and refers owners to the Warranty and Service Booklet. So the severe-duty adjustments exist, but they sit in a booklet we could not find published online, while the schedule on the public servicing page is the light-duty one.
The kilometre figure is a ceiling, not a target - servicing to the odometer alone can quietly push a short-trip car well past its 12-month limit, which is the duty cycle the interval handles worst.