Fault Prevention and Energy Optimization with a Predictive Yaw Model

Context

Delfos was informed of failures in the yaw gearboxes of two Alstom ECO 122 wind turbines at a wind farm. The root cause of the problem was an imbalance between the yaw motors, leading to unexpected failures and impacting the availability of the wind turbines. These failures resulted in energy loss and increased corrective maintenance costs.

‍

Implemented Solution

Delfos has developed a specific predictive model for wind turbine yaw, with the aim of anticipating these failures and enabling preventive action. Using machine learning algorithms, the model was able to identify patterns that indicated imminent failures and generate predictive alerts 15 days in advance.

Case 1:

Energy loss:
348.11 MWh
(Yaw Motor Overload)

‍

Case 2:

Energy loss:
137.84 MWh
(Corrective Maintenance)

‍

Achieved Results

The impact of the solution was significant:

• Failures predicted 15 days in advance, allowing preventive maintenance and reducing costs.
• Avoidance of failures in 4 additional wind turbines, ensuring greater operational availability in the first month of intelligent monitoring.
• Reduction in energy losses, as demonstrated in the case:
  
  • Case 1: Fault reported on 15/11, predictive alert issued on 02/11, loss avoided of 348.11 MWh.‍
  • Case 2: Fault reported on 16/11, predictive alert issued on 01/11, loss avoided of 137.84 MWh.

‍

Conclusion

The implementation of Delfos' predictive model has brought significant gains to the wind farm, avoiding critical failures and significantly reducing energy losses. The model's ability to anticipate and its accuracy make Delfos an essential partner in the optimization of assets in the renewable energy sector.

- 4 failures avoided
- 15 days advance warning
- 485.95 MWh reduction in energy losses

‍