Wind Turbine Power Curve Data Cleansing Checklist

Role: Principal Renewable Asset Data Custodian with extensive expertise in IEC 61400-12 wind turbine performance analytics.

Context

• Wind Generation Facility: {{wind_farm_name}}
• Turbine Fleet Model: {{turbine_model_series}}
• Active Curtailment Flag Field: {{curtailment_event_flag}}
• Anemometer Drift Cutoff: {{anemometer_drift_threshold}}
• Status Code Reference Matrix: {{scada_status_code_dictionary}}
• Analysis Target Period: {{cleaning_window_period}}

Task

Formulate a rigorous operational data cleaning checklist to filter turbine sensor logs, isolate environmental and mechanical anomalies, and prepare an IEC-compliant empirical power curve dataset for {{wind_farm_name}}.

Method

1. Filter turbine status logs using {{scada_status_code_dictionary}} to separate normal operational state records from grid curtailment, planned maintenance, and fault halts.
2. Isolate derated power production records indicated by {{curtailment_event_flag}} so market-driven curtailments do not distort aerodynamic efficiency baselines.
3. Detect anemometer icing or physical sensor blockage by cross-checking wind speed variance against neighboring turbines in {{turbine_model_series}}.
4. Apply the {{anemometer_drift_threshold}} rule to identify nacelle wind speed calibration shifts versus met mast reference values.
5. Flag and eliminate unrealistic operating points, such as negative active power during high wind speeds or above-rated generation during low winds.
6. Normalize power measurements by ambient air density derived from temperature and barometric pressure sensor readings over {{cleaning_window_period}}.
7. Verify bin completeness across wind speed bins from cut-in to cut-out speeds to ensure statistical significance.

Constraints

• Checklists MUST explicitly require density normalization before power bin aggregation.
• Cleaned datasets MUST NOT contain records logged during turbine pitch testing or emergency stop states.
• Include specific IEC 61400-12 data filtering categories for every checklist item.
• Do not include financial revenue modeling or PPA billing steps.

Output format

1. Operational State & Curtailment Filtering (4 markdown checklist items)

2. Environmental & Sensor Integrity Cleansing (4 markdown checklist items)

3. Aerodynamic Sanity Bounds & Binning (3 markdown checklist items)

4. Dataset Sign-off & Completeness Metrics (3 markdown checklist items with pass criteria)

Self-review

• Are curtailment states cleanly separated from genuine aerodynamic performance issues?
• Did I incorporate all six variables including {{scada_status_code_dictionary}}?
• Is the checklist aligned to industry-standard wind performance testing procedures?