From the field · Case 02

Gearbox Temperature Rising With Wind Speed

A Vestas V126. The gearbox oil was running at 78–82 °C where it used to sit at 60–65 °C. Still below the alarm threshold, so nothing had stopped — and that is precisely why it was worth looking at. A temperature that has moved without tripping anything is a machine telling you something quietly.

Symptom

• Oil temperature: 78–82 °C, against a normal band of 60–65 °C for the same conditions
• Alarm: none yet; the threshold is 80 °C with a short averaging window, so it was only brushing it
• Pattern: the gap widened as wind speed rose — at low wind the temperature was almost normal
• Oil: level below the mark, colour dark, a burnt smell, metallic particles present

First assessment

The detail that mattered was that the rise tracked wind speed. A blocked radiator or a failed cooling fan raises temperature fairly uniformly. A rise that grows with load points at the oil itself: either there is not enough of it, or it has lost its ability to do the job.

The state of the oil settled it quickly — low level, dark colour, burnt smell. But a low level is a symptom, not a cause. Oil does not evaporate. The real question was where it was going.

Step-by-step diagnosis

1. Read temperature against wind speed rather than against time. Plotting the two together made the relationship obvious and ruled out a cooling-side fault, which would not have scaled with load this way.
2. Checked the cooling circuit anyway. Radiator clean, fan running, coolant at level. Eliminating the obvious alternative is worth the twenty minutes.
3. Sampled and inspected the oil. Dark, burnt smell, metallic particles. The oil had been running hot long enough to begin oxidising, and wear had started.
4. Followed the oil trace to find the loss. There was oil on the housing below the high-speed shaft. The seal there had hardened and was passing oil.
5. Estimated the quantity lost. Against the fill volume and the level drop, something over 50 litres had gone — consistent with a slow leak rather than a sudden failure.
6. Pulled the eight-month trend. The pitch of the curve told the story: temperature had crept up month by month, and in parallel the drive motor current had risen from about 11 A to 18 A. Two independent numbers describing the same decline.

Root cause

A degraded shaft seal had been leaking oil for months. As the level fell, the remaining oil had to carry more heat, its temperature rose, and above roughly 75 °C viscosity drops and oxidation accelerates — which degrades the oil further and raises the temperature again. A self-reinforcing loop, running slowly enough that no alarm ever fired.

Temperature against wind speed — reference

The shape of this table is the diagnosis. A fault that grows with load is an oil problem; a fault that is flat across load is a cooling problem.

Fix

1. Seal replacement (4 hours): the failed shaft seal was replaced and the sealing face inspected for scoring.
2. Full oil change (3 hours): the entire charge was drained rather than topped up — the old oil was already oxidised and carrying wear metal.
3. Filter replacement: done at the same time, since the existing filter had been loaded by the contaminated oil.
4. Follow-up sampling (3 and 6 months): metal concentration was tracked to confirm the wear had stopped progressing.

Preventive maintenance

• Record oil temperature together with wind speed. A temperature without a load reference cannot be interpreted. The same 70 °C is normal at 12 m/s and alarming at 5 m/s.
• Follow the trend at a fixed wind speed. Pick a band, say 9–10 m/s, and plot the temperature in that band month to month. The slope is the early warning.
• Treat a falling oil level as a leak until proven otherwise. Topping up without finding the path only buys time and hides the trend.
• Take oil samples seriously. Colour, smell and metal content together tell you more than the level gauge.
• Watch drive current alongside temperature. Two unrelated instruments moving in the same direction is strong confirmation.