An engine oil analysis programme is not a test, it is a series. One sample tells you almost nothing useful; five samples from the same engine, drawn from the same port on a fixed interval, tell you whether the wear rate is changing and whether the drain interval is right. The two numbers that decide the interval are TBN, the base reserve available to neutralise the acids combustion produces, and TAN, the concentration of acidic constituents already in the oil. Their relationship is predictable: TAN just begins to increase when TBN depletion reaches 50 percent, and TBN depletion can reach about 65 percent before it becomes necessary to change the oil. There is one trap that catches fleets before they start — TBN can be measured by two different ASTM methods, and the field method can yield a result as much as 2 mgKOH/g lower than the method oil makers use on their data sheets. Same oil, two numbers. This guide covers the crossover, the two test methods, what each result on the report actually means, how to take a sample worth paying for, and what to do with each finding. Book a 30-minute demo to see sample results trended per unit, or start free with 3 vehicles.
One Oil Sample Is a Data Point. Five Is a Decision.
For fleet maintenance managers, shop leads and technicians: how to read the TBN and TAN curves, why two labs give two different TBN numbers, and how to build a programme that shortens some intervals and safely extends others.
No credit card required. Attaches to the PM records you already keep.
The Curve That Sets the Drain Interval
Everything in an oil analysis programme eventually comes back to two lines on one chart. TBN falls as the additive package is consumed neutralising acids; TAN sits flat while there is base reserve to spare, then climbs once there is not. Where they meet, the oil is finished for that duty cycle — and that point is measurable rather than guessed. Trend it per unit free on 3 vehicles.
Curve shape and figures follow the worked example in Machinery Lubrication's guidance on optimising drain intervals with TBN and TAN: for an oil starting at TBN 9.0, TAN will hold steady at around 1.75 to 2.0 until TBN depletes to around 5.5, at which point TAN begins to increase, and the two will meet at around 3.15 to 3.5. Note that 65 percent depletion of a 9.0 starting TBN is 3.15 — the crossover and the practical change point land in the same place. Your own numbers depend on the oil, the engine and the duty cycle, which is exactly why the programme has to be built on your samples rather than someone else's chart.
TBN has two test methods, and they do not agree
ASTM D2896 uses perchloric acid, measures both the hard and soft base components, and is what oil manufacturers use for technical data sheets. ASTM D4739 uses hydrochloric acid, is safer to handle, and is what field labs generally use — and it can yield a result as much as 2 mgKOH/g lower than D2896 on the same oil. A fresh oil with an initial TBN of 11.00 could be measured at 9.0 before it is even put into service, depending on the method. D4739 is the one to use on used, in-service oils. The practical rule is simple: one lab, one method, for the life of the programme.
TBN and TAN definitions, the two ASTM methods and the 2 mgKOH/g gap are per Chevron Lubricants' technical explainer on used engine oil analysis; the depletion thresholds and the worked crossover example are per Machinery Lubrication. New heavy-duty oils typically start in a 6.0 to 13.0 mgKOH/g TBN range.
What Each Result on the Report Is Telling You
A used oil analysis report is six separate conversations printed on one page. Read them as groups, because the interesting findings are always a combination — silicon with iron, fuel with viscosity, oxidation with TBN. We'll read one of your own reports with you in a demo.
- Iron from liners, rings, valve train and gears
- Copper, tin, lead and aluminium from bearings and bushings
- Chromium from hardened surfaces
- What matters is the rate of change across samples, not one value
- Dirt getting past the air intake
- Elevated silicon usually means faulty intake filtration
- Abrasive, so it drives every other wear metal up with it
- Fix the leak before you fix the wear
- Sodium, potassium and boron from coolant inhibitors
- Glycol confirms it directly
- A head gasket, liner seal or cooler is leaking
- This is a stop-and-investigate result, not a trend to watch
- Thins the oil and drops film strength
- Low viscosity may indicate improperly functioning injectors
- Common on heavy idle and short-cycle duty
- Read it with the viscosity result, never alone
- Soot loading thickens the oil and abrades
- Oxidation raises viscosity and acid content together
- Nitration rises with high-idle and gas-heavy combustion
- Monitor these alongside TBN rather than in isolation
- High viscosity can plug oil filters as the oil breaks down
- Low viscosity points at fuel or the wrong grade going in
- TBN is the base reserve left to neutralise acids
- TAN is the acid already there — the two together set the drain
Wear-metal to component mapping, the silicon and coolant-inhibitor indicators and the viscosity interpretations are per Edward Covington of WIX Filters, quoted in Fleet Equipment's guide to interpreting heavy-duty used oil analysis. Cummins notes there that most engine manufacturers will define acceptable TBN limits, so the condemning figure for your fleet comes from your engine maker rather than from a general rule.
Why the Shape of the Line Beats Any Single Number
This is the part that separates a programme from a spend. The same iron figure means three different things depending on what came before it, and no condemning limit can tell them apart. Hold the history against each unit, free.
Illustrative trends, not measured fleet data. The point is structural: a flat line at a moderate level is a healthy engine, a curve bending upward is a developing problem you still have time to plan around, and a step change after a stable history is either a real event or a sampling error — and the cheapest way to tell which is to take another sample before spending anything.
Do Your Sample Results Live in Email, or Against the Truck?
A programme only pays when the last four results are visible at the moment someone decides whether to extend an interval or open the engine. In 30 minutes we'll set up oil analysis as a record attached to each unit, so the trend is in front of the person making the call rather than in a PDF somebody forwarded in March.
Taking a Sample That Is Worth Paying For
Most of the money wasted on oil analysis is wasted at the bottle, not at the lab. Six rules, and the first two account for most of the noise in a fleet's data. Put them on the inspection report and into preventive maintenance so the sample is taken the same way by everyone.
A sample from the drain stream and a sample from the dipstick tube are not comparable. Pick one location per engine and never change it, because a change of port looks exactly like a change of condition.
Take it at operating temperature with the oil circulating, and take it from the middle of the flow — not the first or last of it, where the debris concentrates and misrepresents the sump.
A result with no service life against it cannot be trended. Record miles or hours since the last change and since the last top-up, on the bottle, not from memory afterwards.
TBN can be measured two ways, and the two do not agree. Switching labs mid-programme resets your trend line whether you meant it to or not.
The sample answers whether the interval was right. Drawn after fresh oil goes in, it answers nothing and costs the same.
A programme is a series. Sampling only when something sounds wrong gives you single data points at exactly the moments they are hardest to interpret.
What to Do With Each Finding
A report that does not change a decision was not worth the postage. Here is the action for each of the patterns that actually come back. Set the actions up as job templates with our team. Swipe the table on mobile.
| What the report shows | What it usually means | What to do |
|---|---|---|
| Iron climbing steadily across three or more samples | Normal wear accelerating, or an abrasive getting in | Check silicon on the same reports; if silicon is up too, it is dirt, not the engine |
| Silicon elevated, wear metals following it up | Dirt ingress through intake filtration | Find the air leak or failed filter seal first — replacing the oil changes nothing |
| Sodium, potassium or boron present, or glycol found | Coolant entering the oil | Stop and find it. This is not a drain-interval question |
| Viscosity below grade, fuel detected | Fuel dilution, possibly injectors | Inspect injectors and check duty cycle; extended idle makes it worse |
| Viscosity above grade, oxidation rising | Oil oxidising, soot loading, or overheating | Shorten the interval and check for a cooling or combustion problem behind it |
| TBN falling but everything else stable | Normal additive depletion | Low TBN alone is not condemning; read it with viscosity, oxidation and nitration |
| TAN rising to meet TBN | Base reserve nearly exhausted, acids winning | The interval is over for this duty cycle. Shorten it and re-baseline |
| One high reading with a normal history | Sampling error, or a real step change | Re-sample before spending. A single number with no trend behind it is a question, not an answer |
What the Programme Costs Against What It Catches
A sample bottle and a lab fee against a liner kit, a bearing failure or an engine out of the frame. Industry benchmarking puts unplanned downtime at roughly $448 to $760 per truck per day before the repair, and ATRI puts 2025 repair and maintenance cost at $0.215 per mile, up 8.6%. The programme also works in the other direction — it is the only defensible basis for extending a drain interval rather than guessing at one. Work out your own numbers with our team.
Relative cost by where the problem is caught, illustrative — not prices. Downtime benchmark: FleetNet America and ATA's Technology & Maintenance Council via industry reporting; per-mile cost: ATRI.
Oil Analysis Questions Fleets Ask
What does TBN tell me about my oil?
How much base reserve additive is left to neutralise the strong acids combustion produces. It falls through the drain interval as the additive is consumed — but a low TBN on its own is not a condemning result. Lower TBN values seen in the field today are not necessarily a reflection of the overall health of the oil, which is why it is read alongside viscosity, oxidation, nitration and TAN. Have us read a report with you.
Why did two labs give me different TBN numbers for the same oil?
Because there are two test methods. ASTM D2896 measures both hard and soft base components and is what appears on manufacturer data sheets; ASTM D4739 is what field labs typically run, and it can read as much as 2 mgKOH/g lower on the same oil. Neither is wrong. Pick one lab and stay with it. Record the lab and method per sample, free.
When is the oil actually finished?
TAN begins to increase once TBN depletion reaches about 50 percent, and depletion can reach about 65 percent before a change becomes necessary. Once TAN has risen to meet TBN, the oil's ability to neutralise acids has dropped significantly and the interval is over for that duty cycle. Your engine maker's stated limits take precedence over any general rule. Set condemning limits per engine in a demo.
Can oil analysis let me extend drain intervals?
It is the only sound basis for doing so. An extension defended by three or four consecutive samples showing stable wear rates and healthy base reserve is an engineering decision; the same extension made because the oil "looked fine" is a gamble with an engine attached. Confirm any extension against the engine maker's position and your warranty terms. Hold the evidence per unit, free.
What does high silicon mean on a diesel oil report?
Dirt. Elevated silicon in used oil indicates dirt contamination, usually caused by faulty intake filtration — and because dirt is abrasive, it drives iron and the other wear metals up alongside it. Find the air path before touching the engine. Map intake checks to the PM in a demo.
What does it mean if viscosity is low?
Usually fuel dilution, and low viscosity may indicate improperly functioning injectors. High viscosity is the opposite problem and can lead to premature plugging of oil filters as the oil breaks down. Read viscosity with the fuel and soot results rather than on its own. Track viscosity trends free.
How often should we sample?
On a fixed interval tied to the drain, every unit, whether or not anything sounds wrong — because the value is in the series. Sampling only when a problem is suspected produces isolated numbers at the worst possible moment to interpret them. Set the sampling schedule in a demo.
Put the Last Four Samples in Front of the Person Deciding
FleetRabbit holds every sample result, the lab, the method and the service life against the unit it came from, so the trend is visible at the moment someone extends an interval or opens an engine — and any finding becomes a work order with that history already attached.
No credit card required. Start with your next drain.
