Three Hidden Things SCADA Can Tell You About Turbine Performance

Operators don’t lack data. They lack clarity.

SCADA is available everywhere, but raw signals are noisy, unstable and heavily shaped by turbine operation. Standard analysis often misses real changes or reacts to noise. It becomes either too sensitive, detecting issues that are not there, or not sensitive enough, missing issues that matter.

If SCADA is cleaned, calibrated and interpreted with the right methods, it reveals far more than production figures or standard KPIs suggest. It shows what the wind conditions actually were, how turbines reacted and what caused the performance changes.

Here are three things SCADA can reveal when the data is treated correctly.

1. Wind speed can be far more reliable than the raw signal

Nacelle wind speed is affected by the rotor, yaw movements, averaging and turbulence. In raw form it is not precise enough for performance analysis, and it introduces uncertainty in almost every metric operators care about.

Unreliable does not mean unusable.

By applying multi-dimensional stability models, cross-turbine consistency checks and bin-dependent calibration functions, we can detect any change in the wind-speed measurement and correct it on a relative scale. This creates a stable, calibrated wind-speed signal that is consistent over time. This greatly reduces uncertainty in wind measurement and provides a reliable basis for tracking true performance behaviour. Corrected wind speed is one of the foundations for understanding how the turbine is really operating.

Fig. 1: Monthly wind-speed corrections for a wind farm. DEM13 shows a clear drift of almost 0.3 m/s caused by the installation of the vortex generators. If not corrected, the power curve appears 11.7% better than reality.

2. Performance becomes clear when several independent indicators align

Assessing performance on wind turbines is quite complex and a single indicator is rarely enough.

SCADA becomes reliable when several independent signals confirm the same pattern. Corrected power curves, power-prediction residuals, turbine-to-turbine consistency, pitch and rotor regimes, each reveal a part of the behaviour. When they move together, the trend is real.

This approach reduces both false positives and false negatives caused by external effects, temporary conditions and unstable signals. Instead of reacting to a single noisy curve, operators can rely on performance conclusions supported by multiple clean, calibrated indicators.

Fig. 2: Power curves and measured vs predicted power. Even with the corrected power curve lower than the raw one, measured power stays above the prediction, confirming the real performance gain from the vortex generators.

3. SCADA reveals how turbines are actually controlled

Even if SCADA was not designed for performance analysis, it contains enough information to understand how turbines behave in real operation. By analysing the relationships between wind speed, pitch, rotor speed, yaw activity and torque, we can see:

• how each turbine responds to the same wind conditions
• how and when it begins to limit power
• how it ramps, restarts and transitions between modes
• how its control strategy changes over time
• why two identical turbines react differently to identical conditions

These control signatures explain why performance evolves and what causes the change.

Fig. 3: Control curves vs power, showing how the turbine is actually operated. Key control parameters can be extracted from control curves, revealing how the unit reacts, limits and transitions under different conditions.

Why this matters

Standard SCADA analytics rely on raw or lightly processed signals, which leads to missed issues or wrong conclusions. These approaches struggle with noisy wind, inconsistent sensors and turbine-specific behaviour.

Once SCADA is stabilised, calibrated and made comparable:

• performance conclusions become precise instead of approximate
• wind effects can be separated from turbine behaviour
• behaviour differences between turbines become clear
• real issues can be detected earlier and with higher confidence

Much of SCADA’s value is hidden and needs the right tools to be extracted. Once treated correctly, the data shows how turbines behave, why performance changed and where concrete improvements can be made.

---

Want to learn more about ExpertWind? Feel free to Schedule a Demo for a live demonstration of our services and see how your wind farms could benefit from it.