Wind Turbine Gearbox Intelligence
Evaluates gearbox condition using five input parameters and produces a structured maintenance report.
When to Use
Load this skill when the user wants to:
- Assess gearbox health from on-site inspection or sensor data
- Classify damage severity on a 1-5 scale
- Determine whether a turbine should be shut down or kept running under monitoring
- Generate a structured maintenance or escalation plan
Quick Reference
| Input Parameter |
What to Collect |
| Visual inspection |
Surface cracks, pitting, spalling, discoloration, debris |
| Oil iron (Fe ppm) |
Iron particle concentration in gear oil (ppm) |
| Temperature (C) |
Bearing/gear temperature, normalized to baseline |
| Vibration |
RMS or peak-to-peak acceleration (g), frequency anomalies |
| Acoustic / Sound |
Noise type: grinding, knocking, whining, clicking |
Fault Thresholds (Reference)
| Parameter |
Normal |
Warning |
Critical |
| Oil Fe (ppm) |
< 100 |
100 - 300 |
> 300 |
| Temp above baseline |
< 5 C |
5 - 15 C |
> 15 C |
| Vibration RMS (g) |
< 0.5 |
0.5 - 1.5 |
> 1.5 |
| Acoustic |
No anomaly |
Intermittent |
Continuous |
| Visual |
Clean surface |
Minor pitting |
Spalling/crack |
Common Failure Modes
| Failure Mode |
Typical Indicators |
| Micropitting |
High Fe ppm, slight vibration increase, no visible cracks |
| Spalling |
High Fe ppm, elevated vibration, visible surface damage |
| Fatigue crack |
Knocking sound, vibration spike at gear mesh frequency |
| Bearing wear |
Whining noise, high temperature, broadband vibration increase |
| Oil contamination |
Very high Fe ppm, discolored oil, possible foaming |
Procedure
- Collect inputs across all five parameters. If any are unavailable, note as "not measured" and proceed.
- Evaluate each parameter against the thresholds table. Flag Warning or Critical zones.
- Cross-correlate symptoms:
- Fe ppm + vibration increase → wear / spalling progression
- Knocking sound + vibration spike → fatigue crack
- Temperature + whining sound → bearing failure
- Multiple Critical flags → Severity 5
- Assign severity:
- 1 Healthy: All parameters normal. No action required.
- 2 Early wear: 1-2 parameters in warning zone. Increase monitoring frequency.
- 3 Moderate damage: 2-3 parameters in warning/critical. Inspect within 2 weeks.
- 4 Significant damage: Multiple critical flags. Plan shutdown within 48-72 hours.
- 5 Critical: Imminent failure risk. Immediate shutdown required.
- Determine root cause from the failure modes table.
- Generate the output report using the format below.
Output Format
=== GEARBOX HEALTH REPORT ===
ROOT CAUSE : [e.g., Progressive spalling on intermediate shaft gear]
SEVERITY : [1-5] - [Healthy / Early Wear / Moderate / Significant / Critical]
SHUTDOWN : [Yes / No / Conditional]
MONITORING STRATEGY:
- [e.g., Repeat oil sample in 72 hours]
- [e.g., Daily vibration trend monitoring for 1 week]
ESCALATION TRIGGERS:
- [e.g., Fe ppm exceeds 400 - immediate shutdown]
- [e.g., Vibration RMS exceeds 2.0 g - immediate shutdown]
Pitfalls
- Never assign Severity 5 based on one parameter alone. Cross-validate with at least two sources.
- Temperature readings can be misleading in extreme ambient conditions. Ask for baseline-normalized values.
- Acoustic descriptions are subjective. Ask for noise type and whether continuous or intermittent.
- If sensor data is unavailable, rely on visual and oil condition as primary indicators.
- Do not conflate oil change interval with oil health. New oil can still show high Fe ppm.
Verification
After generating the report, confirm with the user:
- Does the severity match their on-site observations?
- Are escalation thresholds feasible for their monitoring setup?
- Are there additional data points (CMS trending, historical Fe ppm) that could refine the assessment?
1---2name: wind-turbine-gearbox3description: Assesses wind turbine gearbox health from multi-sensor and inspection data. Classifies damage severity (1-5), identifies root cause, and recommends shutdown or monitoring actions.4---56# Wind Turbine Gearbox Intelligence78Evaluates gearbox condition using five input parameters and produces a structured maintenance report.910## When to Use1112Load this skill when the user wants to:13- Assess gearbox health from on-site inspection or sensor data14- Classify damage severity on a 1-5 scale15- Determine whether a turbine should be shut down or kept running under monitoring16- Generate a structured maintenance or escalation plan1718## Quick Reference1920| Input Parameter | What to Collect |21|-------------------|------------------------------------------------------------|22| Visual inspection | Surface cracks, pitting, spalling, discoloration, debris |23| Oil iron (Fe ppm) | Iron particle concentration in gear oil (ppm) |24| Temperature (C) | Bearing/gear temperature, normalized to baseline |25| Vibration | RMS or peak-to-peak acceleration (g), frequency anomalies |26| Acoustic / Sound | Noise type: grinding, knocking, whining, clicking |2728## Fault Thresholds (Reference)2930| Parameter | Normal | Warning | Critical |31|---------------------|---------------|----------------|----------------|32| Oil Fe (ppm) | < 100 | 100 - 300 | > 300 |33| Temp above baseline | < 5 C | 5 - 15 C | > 15 C |34| Vibration RMS (g) | < 0.5 | 0.5 - 1.5 | > 1.5 |35| Acoustic | No anomaly | Intermittent | Continuous |36| Visual | Clean surface | Minor pitting | Spalling/crack |3738## Common Failure Modes3940| Failure Mode | Typical Indicators |41|------------------|---------------------------------------------------------------|42| Micropitting | High Fe ppm, slight vibration increase, no visible cracks |43| Spalling | High Fe ppm, elevated vibration, visible surface damage |44| Fatigue crack | Knocking sound, vibration spike at gear mesh frequency |45| Bearing wear | Whining noise, high temperature, broadband vibration increase |46| Oil contamination| Very high Fe ppm, discolored oil, possible foaming |4748## Procedure49501. Collect inputs across all five parameters. If any are unavailable, note as "not measured" and proceed.512. Evaluate each parameter against the thresholds table. Flag Warning or Critical zones.523. Cross-correlate symptoms:53 - Fe ppm + vibration increase → wear / spalling progression54 - Knocking sound + vibration spike → fatigue crack55 - Temperature + whining sound → bearing failure56 - Multiple Critical flags → Severity 5574. Assign severity:58 - 1 Healthy: All parameters normal. No action required.59 - 2 Early wear: 1-2 parameters in warning zone. Increase monitoring frequency.60 - 3 Moderate damage: 2-3 parameters in warning/critical. Inspect within 2 weeks.61 - 4 Significant damage: Multiple critical flags. Plan shutdown within 48-72 hours.62 - 5 Critical: Imminent failure risk. Immediate shutdown required.635. Determine root cause from the failure modes table.646. Generate the output report using the format below.6566## Output Format6768=== GEARBOX HEALTH REPORT ===6970ROOT CAUSE : [e.g., Progressive spalling on intermediate shaft gear]71SEVERITY : [1-5] - [Healthy / Early Wear / Moderate / Significant / Critical]72SHUTDOWN : [Yes / No / Conditional]7374MONITORING STRATEGY:75 - [e.g., Repeat oil sample in 72 hours]76 - [e.g., Daily vibration trend monitoring for 1 week]7778ESCALATION TRIGGERS:79 - [e.g., Fe ppm exceeds 400 - immediate shutdown]80 - [e.g., Vibration RMS exceeds 2.0 g - immediate shutdown]8182## Pitfalls8384- Never assign Severity 5 based on one parameter alone. Cross-validate with at least two sources.85- Temperature readings can be misleading in extreme ambient conditions. Ask for baseline-normalized values.86- Acoustic descriptions are subjective. Ask for noise type and whether continuous or intermittent.87- If sensor data is unavailable, rely on visual and oil condition as primary indicators.88- Do not conflate oil change interval with oil health. New oil can still show high Fe ppm.8990## Verification9192After generating the report, confirm with the user:93- Does the severity match their on-site observations?94- Are escalation thresholds feasible for their monitoring setup?95- Are there additional data points (CMS trending, historical Fe ppm) that could refine the assessment?
dvcrn