Grease Analysis and Advice for Wind Turbine Bearings
A grease analysis of a wind-turbine bearing measures wear and water in grease sampled representatively from the bearing in service: ferrous wear debris by magnetometry, wear particles per gram in size bands and, for main bearings, elements. We know of no standardised limits for in-service grease, so each bearing is judged against its own trend and the fleet.
We interpret grease analyses from main bearings and blade (pitch) bearings.
TriboTech is a consulting engineering company, not a laboratory. The analyses are carried out by our partner laboratories. Ask us for an analysis programme.
Why grease analysis, and why advice
A blade or pitch bearing does not rotate continuously. It oscillates through a few degrees, so according to Møller & Esbensen (2022, 2025) vibration monitoring does not apply. Here the grease can tell you how the bearing is doing.
The main bearing rotates, and there grease analysis adds to the other monitoring. ASTM D7718 states that grease analysis can be combined with infrared imaging, vibration analysis and ultrasonic vibration analysis.
The method decides what you see. Iron measured by RDE spectroscopy misses large particles. In a validation study on blade bearings that had a defect, iron by RDE spectroscopy stayed below 50 ppm while magnetometric FdM-Fe on the same samples reached 1,400 ppm (Møller et al. 2016). ASTM D8120 describes magnetometry as insensitive to particle size. So our programme determines iron in grease by magnetometry.
The laboratory measures. What you buy from us is the judgement on the bearing and the recommendation on what to do.
The bearing types
Main bearing
The main bearing rotates under heavy load. We trend fine wear particles in size bands, together with ferrous wear, water and elements, on every sample.
We choose the sampling point by bearing design. Grease that has left the bearing past the seals is not used to judge the bearing: it holds debris from the exit path, not from the load zone. The principle is in DS 3077: a grab sample is representative only in the rarest of instances.
Blade (pitch) bearings
Used grease is pressed out into grease cups that can be reached in service. Two randomly selected cups are sampled per bearing (Møller & Esbensen 2022, 2025). In our programme both samples are analysed for ferrous wear and water, and the sample with the higher FdM-Fe is then counted for coarse particles. According to Møller & Esbensen (2025), only wear particles larger than ~100 µm are relevant for a condition check of blade bearings.
Agreement between the two cups supports that the samples represent the bearing. We report a wide pair as variation along the bearing.
Sampling decides the result
A sample is representative only when every part of the lot has the same non-zero chance of ending up in it. This is the Fundamental Sampling Principle in DS 3077.
The large errors arise at the turbine, not in the laboratory: according to Møller et al. (2016), non-representative primary sampling typically carries at least 10 times the error of the subsequent laboratory sampling.
ASTM D7718-11 is the standard practice for sampling in-service grease, and its goal is a sample you can trend. For a large bearing that does not fully rotate, such as a slew bearing, the standard notes that more than one sample may be needed. The sampler must know the component's internal layout and its sampling points.
Grease-cup sampling has been validated against dismantled blade bearings (Møller et al. 2016), and in one published case the finding was confirmed when the bearing was subsequently opened (Møller & Esbensen 2022). In the validation study, two blade bearings were dismantled and their grease sampled all the way round the active zone (360°) as a reference. Grease-cup averages agreed with that reference for ferrous wear and water, while outlet-hole samples were much more variable (Møller et al. 2016). The bearings were deep-groove ball bearings, and according to the authors the result may likely also apply to other turbine sizes and types with a similar type of blade bearing.
The studies were published by Hans Møller and co-authors (see References). Hans Møller is today a partner in TriboTech.
An empty grease cup is not a sample. We only use cups with enough grease in them.
The analysis programme
The programme is agreed per fleet:
| Parameter | Main bearing | Blade/pitch bearing |
|---|---|---|
| Ferrous wear (FdM-Fe) by magnetometry, mg/kg | every sample | both samples |
| Water, Karl Fischer with oven | every sample | both samples |
| Wear particles per gram in size bands | >5, >15, >25, >50 and >100 µm | >100, >200, >300, >400 and >500 µm on the sample with the higher FdM-Fe |
| Elements (XRF) | every sample | – |
| Microscopy of selected particles | when the count warrants it | when the coarse count warrants it |
Ferrous wear is measured by magnetometry (ASTM D8120, which also covers bearing greases). The ferrous wear particles are separated and counted per gram. The counting and the size bands follow our own programme, not an ISO or ASTM method. Background in Counting particles.
How we interpret: the bearing's history and the fleet
We do not pass or fail the bearing. You get the counts per band and where each bearing stands against the fleet and its own history. Where the report uses assessment levels, they are TriboTech's own and are labelled as such, not as a standard.
- The bearing's own trend. ASTM D8120 frames magnetometric iron as a way to pinpoint when the build-up of ferrous debris departs from normal. We trend water the same way.
- Healthy blade bearings stay low and steady. In five years of offshore monitoring, well-functioning blade bearings showed a low, constant content of ferromagnetic wear particles, with no rise as they aged. One bearing with a rising trend was, in the authors' words, clearly predicting premature failure (Møller & Esbensen 2022).
- On blade bearings the coarse particles weigh most. They are the primary wear parameter, and the iron level alone does not decide the recommendation.
- The fleet as reference. We place each bearing in its fleet by percentile ranking. ASTM D7720 describes the technique as a cumulative distribution of sorted values that does not depend on a presumed statistical profile. The standard is written for in-service oil; we apply the same statistical principle to grease.
- The manufacturer's criteria come first. ASTM D7720 stresses that its techniques do not supersede limits established and validated by the equipment manufacturer. Where the turbine manufacturer or the grease supplier has validated criteria, we read your data against them once you give them to us.
How it works
- Agreement. Before we deliver the first analysis or interpretation, you and we sign a framework agreement, and each task is agreed separately.
- Planning. We agree bearing types, sampling points per turbine type and the purpose: routine monitoring, end of warranty or fault investigation.
- Sampling. We send a sampling instruction. Sampling on site can be agreed.
- Analysis. Our partner laboratories analyse the samples.
- Interpretation and report. You receive an assessment per bearing and a fleet overview, with the recommended action and the date of the next sampling.
- Follow-up. The campaign can be repeated so that each bearing is followed over time, and it can be timed to the end of warranty.
Who it is for
We work for clients in offshore and onshore wind in Northern Europe.
Independent interpretation
We sell neither grease, lubrication systems nor bearings. Our recommendation rests on your data, the standards, the published literature and our professional judgement.
References
- Møller, H.S., Esbensen, K.H., Wurzbach, R. (2016). Grease sampling and analysis for in-service Condition Monitoring (CM) of wind turbine blade bearings. LUBMAT 2016, Bilbao.
- Møller, H., Esbensen, K.H. (2022). Representative Sampling for condition monitoring of in-service wind turbine bearings: challenges and solutions over 10 years. TOS Forum 11, 207–215. doi:10.1255/tosf.150
- Møller, H.S., Esbensen, K.H. (2025). A Manual Process Sampling Solution for In-Service PAT Inspection of Wind Turbine Blade Bearing Lubricants. Sampling Science & Technology 4, 95–100. doi:10.62178/sst.004.009
Frequently asked questions
What does grease analysis show that vibration monitoring does not?
On blade (pitch) bearings vibration monitoring does not apply, because the bearing oscillates a few degrees instead of rotating. There, a grease sample can show the bearing's condition (Møller & Esbensen 2022, 2025). On main bearings, grease analysis adds the wear debris itself, its amount, size and shape, alongside vibration. ASTM D7718 states that grease analysis can be combined with vibration analysis.
Is a grease-cup sample representative of the blade bearing?
Yes, for the bearing type it was validated on. Two dismantled blade bearings were sampled all the way round the active zone and compared with grease-cup samples. The averages agreed for ferrous wear and water, and the cups held the same particle types and sizes. Outlet-hole samples varied much more for ferrous wear and water (Møller et al. 2016). The study covered deep-groove ball bearings, and according to the authors the result may likely also apply to other turbine sizes and types with a similar type of blade bearing.
How often should blade and main bearing grease be analysed?
We set the interval per fleet in the programme, from the bearings' trend. In the validation study the grease cups were changed at the yearly service (Møller et al. 2016), and in the published method samples are taken during inspection visits (Møller & Esbensen 2025).
Why can iron measured by spectroscopy be low while the bearing is wearing?
Because iron measured by RDE spectroscopy does not see the large particles. In a validation study on blade bearings that had a defect, RDE iron stayed below 50 ppm while magnetometric iron ranged up to 1,400 ppm on the same samples (Møller et al. 2016). ASTM D8120 describes magnetometry as insensitive to particle size, while noting that very large debris particles can make the reading depart from the linear calibration. More on what ICP sees and misses in ICP elemental analysis.
Are there limits for in-service grease?
We know of no standardised limits for in-service grease. We rank each bearing against its own history and the fleet, using the statistical principle ASTM D7720 describes for in-service oil. Where the report uses assessment levels, they are TriboTech's own. Where the turbine manufacturer or grease supplier has validated criteria and you give them to us, those apply. A grease count is never converted to an ISO 4406 code: the microscope route to that code is counting to ISO 4407, which counts per volume of fluid, not per gram of grease.
Can you interpret a grease analysis from another laboratory?
Yes. Send the report and tell us which method measured iron and particles: ICP/RDE, PQ index or magnetometry. We will come back with a proposal for interpreting it. The method decides what the number can bear, and we cannot re-check the other laboratory's sampling.
Can grease analysis be used at end of warranty?
Yes. End of warranty is a common reason for a grease campaign on a fleet. The report describes wear and water per bearing from the samples at the time of sampling. Where the manufacturer or grease supplier has its own criteria and you give them to us, the data are read against those as well. The report does not assess whether damage is covered by the warranty; any claim is raised by the owner.
Talk to an adviser
Tell us about your fleet, the bearing types and the purpose, and we will propose an analysis programme. What you get from us is advice, not only analysis results.