The Gearbox Is Hotter, but the Viscosity Report Looks Normal
Two oil reports agree at 40 °C. The gearbox sump is 14 °C hotter. Learn what the laboratory number can tell you, and what to check next.
- Training exercise
- Two records
- Next-check table
Read both records
The sump temperature has climbed from 68 °C to 82 °C. The new oil report comes back at 219 cSt, against 221 cSt at the last check. Read on its own, that small change looks reassuring.
The gearbox, Gear Oil A and these readings were created for this exercise. Read the laboratory report alongside the machine log before deciding what the stable result says about the extra heat.
Laboratory report
Same laboratory and test method for both samples.
| Item | Earlier check | Later check |
|---|---|---|
| Product on record | Gear Oil A, ISO VG 220 | Same product |
| Test method | ASTM D445 | ASTM D445 |
| Test temperature | 40 °C | 40 °C |
| Kinematic viscosity | 221 cSt | 219 cSt |
Machine log
Same installed sump sensor, read when each sample was taken.
| Item | Earlier check | Later check |
|---|---|---|
| Sump oil temperature | 68 °C | 82 °C |
| Input shaft speed | 1,480 rpm | 1,480 rpm |
| Operating note | Steady production | Steady production |
| Transmitted load | Not recorded | Not recorded |
Viscosity of the oil at the sump temperature
- Earlier, 68 °C
- Not given
- Later, 82 °C
- Not given
Left blank in both records.
In this article
The answers
Check your answers
What stayed close
The reported viscosity fell by about 0.9%. Both results sit inside the ISO VG 220 band of 198 to 242 cSt, so there is little movement at the test temperature. To judge how much that movement means, ask the laboratory about the method’s precision. A difference of 2 cSt should not become a trend simply because the spreadsheet draws a line between the points.
The rest of the report still matters. Water, wear metals and acid number may tell a different part of the story. A stable viscosity result alone cannot rule out water, wear debris, oxidation or a mixed product.
What changed
The sump sensor reads 14 °C hotter. That deserves attention, but first confirm the measurement: the sensor’s condition and position, and when each reading was taken. A comparison is useful only if the readings describe comparable conditions.
What is missing
Neither report gives the viscosity at 68 °C or 82 °C. There is no product curve to estimate it, and no gearbox requirement against which to judge an estimate.
The load is missing too. A shaft can turn at the same speed while transmitting a different load; “steady production” does not settle that comparison.
What we can say is narrower: the two samples behaved similarly at 40 °C. If the oil itself is unchanged, heating it from 68 °C to 82 °C will reduce its viscosity. The two 40 °C results do not establish its viscosity at either sump temperature, or prove that its behaviour across the temperature range is unchanged.
The test method
Why the laboratory tests at 40 °C
On the report, read “kinematic viscosity at 40 °C” as one complete result. Leave out the temperature and you lose the condition under which the number was measured.
In ASTM D445, the capillary method named on these reports, the laboratory holds the sample at a controlled temperature and times its flow under gravity through a calibrated tube. Multiplying that time by the tube’s calibration constant gives the kinematic viscosity. ISO 3104 also describes capillary kinematic-viscosity measurement. Noria’s explanation makes the practical point: a viscosity figure needs its test temperature to be meaningful.
The ISO viscosity grades use 40 °C, which is why it is a familiar test temperature for industrial oils. Holding the temperature constant lets you compare a used-oil sample with the previous sample or with new oil, subject to the test method and its precision. The laboratory has deliberately removed temperature as a variable. The gearbox has not.
Referenced here Noria: Oil Viscosity, How It’s Measured and Reported
ISO VG 220 is a label at 40 °C
Under ISO 3448, ISO VG 220 has a midpoint of 220 cSt and a band of 198 to 242 cSt, all at 40 °C. Finding both results inside that band answers a grading question. It does not give you a viscosity at 82 °C.
Nor is the band a used-oil acceptance limit. A result can remain within the grade while moving away from that product’s new-oil value or its established trend. Use the product baseline and the gearbox’s condition-monitoring limits to decide whether the change deserves attention.
The same care with specifications appears in Reliability Engine’s coolant article, Inside PG25: The Invisible Chemistry Protecting Every Cold Plate. The useful connection is the habit of checking the product, method and sample location before interpreting a limit. The coolant tests and limits themselves do not transfer to gear oil.
Referenced here ISO 3448 record; Shell Marine Pocketbook 2025, p. 41
Test temperature and machine temperature
Hotter oil is thinner oil
Take the same, unchanged lubricating oil from 68 °C to 82 °C and its viscosity falls. That temperature effect is the principle behind this exercise. To put a number on it for Gear Oil A, we need more than the two results measured at 40 °C.
Shell’s 2025 marine pocketbook shows how large the difference can be. It lists typical viscosities for Shell Omala S2 GX 220, a mineral ISO VG 220 gear oil, of 220 mm²/s at 40 °C and 19.0 mm²/s at 100 °C. The second value is less than one tenth of the first.
Those are values for a different oil, at temperatures different from our sump readings. They cannot be used to fill in the missing results for Gear Oil A. Their value here is to show how much a temperature can change the meaning of a viscosity number.
Viscosity at running temperature is called operating viscosity. It matters to the gears and to the bearings inside the gearbox. SKF’s handbook on bearings in twin screw compressors identifies operating viscosity as the main parameter for choosing a bearing lubricant.
Referenced here Mobil: High viscosity index oils; Shell Marine Pocketbook 2025, p. 22; SKF compressor bearing handbook, p. 27
Ask for the curve that belongs to this oil
Different oils respond differently to heat. Viscosity index, or VI, describes that temperature dependence: a higher VI means less change in viscosity with temperature. It tells you nothing by itself about oxidation, contamination or additive condition.
Calculating VI requires viscosities at 40 °C and 100 °C. Our two samples were both tested at 40 °C. Repeating the measurement at one temperature does not provide the second temperature needed for VI, or a value at 68 °C or 82 °C.
Ask the supplier for the exact product’s viscosity-temperature curve, or a validated calculation, and match it to the formulation and data-sheet revision. That curve describes new oil. Confirming the oil’s identity and a 40 °C result close to the new-oil value makes it a starting estimate for the used oil, not proof of how that used oil behaves at another temperature.
Keep the routine 40 °C test, and keep estimates distinct from measurements. Without a suitable curve or calculation, leave the operating-viscosity boxes blank.
Check the assumptions on generic charts as well. The SKF handbook’s viscosity-temperature chart for ISO grades assumes mineral oils with a VI of 95. Its next page discusses a synthetic oil with a VI of 150 that retains more viscosity when hot, while noting that this alone does not mean a thicker bearing film. These examples explain the behaviour of oils; neither supplies the missing curve for Gear Oil A.
Even the correct curve gives bulk-oil viscosity at a stated temperature. The sump sensor does not measure every gear mesh or bearing, and the curve does not say whether the viscosity is adequate there. For that judgement, you still need the gearbox manufacturer’s lubrication requirement.
Referenced here Shell Marine Pocketbook 2025, glossary p. 57; Noria on viscosity index; SKF compressor bearing handbook, pp. 27 and 28
Next checks
Decide what to check next
If the gearbox has reached a site alarm or trip limit, follow that procedure before anything else. Follow the site’s safe-work rules for every reading, and never open a running gearbox to measure its temperature.
Once any immediate alarm has been dealt with, use the records to decide where to look. Each row below pairs a finding with the evidence needed to make it useful.
| What you have | What it supports | Next check |
|---|---|---|
| Viscosity of 221 and 219 cSt at 40 °C | The oil’s measured viscosity has changed very little | Ask the laboratory for the method’s precision, compare both results with the new-oil value for Gear Oil A, and read the other results on the same report |
| Sump oil at 68 and 82 °C | The oil at the sensor is hotter | Confirm the sensor’s position, calibration status and reading times, and record the ambient temperature |
| Same speed and “steady production” | That operation looked similar, not that the load was the same | Record the load, or a clearly labelled stand-in such as motor current, for both conditions |
| Gear Oil A in the lubrication record | The product that should be in the gearbox | Check top-ups, oil level by the manufacturer’s method and recent work, and ask the laboratory whether the results still match the new-oil baseline |
| If a cooler is fitted: no cooler readings in these records | Nothing yet about heat removal | Record cooler inlet and outlet temperatures and flow, look for any logged cooler history, and compare it with the timeline of the heating |
| No product curve or gearbox requirement on file | No judgement of operating viscosity yet | Request the exact-product curve and the gearbox manufacturer’s viscosity requirement |
Choose the investigation steps to suit the gearbox and its operating conditions.
The cause may emerge as these records are checked together. If it does not, widen the investigation through the machine’s approved inspection and condition-monitoring procedures.
A hotter gearbox alone is not a reason to buy thicker oil
Buying a heavier grade would be premature. The cause of the heating is still unknown, and additional viscosity can create its own problem: SKF warns that excessive viscosity in bearings causes friction and heat. Any grade change must also suit start-up temperatures, oil delivery to the gears and bearings, and the manufacturer’s approved list.
Keep the approved oil unless the people responsible for the equipment and its lubrication decide, on evidence, to change it. These records do not select a replacement grade, diagnose gear damage or establish an alarm temperature.
A useful review note separates what was measured from what still needs an answer.
The next person can now see why the oil grade stayed the same and which checks are still open. That is a more useful handover than a single word beside the viscosity result: “normal”.
References
- Shell Marine, Pocketbook 2025, October 2025: ISO viscosity classification, p. 41; Shell Omala S2 GX gear oils, p. 22; glossary entry for viscosity index, p. 57.
- SKF, Bearings in twin screw compressors: application handbook, PUB BU/P7 14510 EN, March 2018: Table 1 and Diagram 1, p. 27; oil viscosity notes and Diagram 2, p. 28.
- Noria Corporation, Oil Viscosity: How It’s Measured and Reported, Machinery Lubrication, first published in Practicing Oil Analysis, November 2002.
- Rick Russo, ExxonMobil, High viscosity index oils can help improve productivity, Mobil, 23 February 2016.
- ISO, ISO 3448:1992 Industrial liquid lubricants: ISO viscosity classification, public catalogue record.
- Reliability Engine, Inside PG25: The Invisible Chemistry Protecting Every Cold Plate, 12 August 2026.
