Wind Farm Optimization — GreenWind Corp
Visual: turbine twin outputs, SCADA overlays and blade health diagnostics.
Client Overview
GreenWind Corp operates mid-sized wind farms across coastal regions in the USA. They aimed to boost Annual Energy Production (AEP), reduce blade and drivetrain failures, and optimize curtailment decisions through a digital twin and advanced analytics platform.
Challenge
Frequent gearbox and blade maintenance, unpredictable wake losses between turbines, and conservative curtailment strategies limited revenue. GreenWind needed a physics-informed digital twin combined with data-driven models to tune control settings, predict component degradation, and optimize farm-level power capture.
Solution — Digital Twin & Performance Suite
We developed a digital twin that replicates turbine aerodynamics, wake interactions, and drivetrain thermal behavior. Coupled with SCADA analytics and ML for anomaly detection, the twin enabled scenario testing for yaw, pitch, and curtailment strategies to maximize AEP while protecting equipment.
Core modules
Approach
- Calibrate twin per turbine using historical SCADA, met mast, and CMS data.
- Run farm-level simulations to test wake-steering and curtailment strategies under measured conditions.
- Deploy anomaly detectors for drivetrain & blade signals and integrate maintenance workflows.
- Run A/B tests for control changes with limited rollouts before full adoption.
Technology stack
Implementation — Phases
Phase 1 — Twin Calibration (Weeks 1–12)
Calibrate physics-based twin using SCADA & met mast data and validate predictive performance.
Phase 2 — Pilot Wake Steering Tests (Weeks 13–22)
Run controlled wake steering experiments on a subset of turbines to measure AEP impact and loads.
Phase 3 — Predictive Maintenance & Fleet Rollout (Weeks 23–34)
Deploy anomaly detection, schedule preventative tasks, and supply-chain planning for spares.
Phase 4 — Optimization Governance (Weeks 35–44)
Embed governance for control changes, monitoring, and continuous twin recalibration.
Impact & Results
3–5%
Net AEP uplift from wake-steering & optimized controls
30%
Reduction in unexpected drivetrain failures
15%
Lower scheduled maintenance costs via better planning
12 months
To measurable farm-level improvements
Qualitative outcomes
- Confidence in gradual control changes enabled revenue improvements with accepted loads.
- Predictive maintenance reduced emergency interventions and improved availability.
- Supply-chain optimization for spares minimized downtime during repairs.
Client Testimonial
Key Highlights & Learnings
- Start with a calibrated twin for a representative turbine—the rest can be scaled by parameterization.
- Controlled field experiments are essential to verify twin predictions and ensure safe adoption.
- Integrate predictive maintenance with spare-parts logistics to fully realize uptime gains.