What 10-Minute SCADA Data Hides—and What Seconds-Level Data Reveals

Ten-minute SCADA data has served the wind industry well for decades. It is compact, widely available and convenient for baseline reporting, availability tracking and long-term yield trends.

However, wind turbine does not operate in 10-minute increments.

Wind conditions change second to second. In response, turbine controllers continuously adjust blade pitch, generator torque, yaw alignment, and rotor speed. A massive amount of dynamic behavior occurs within every 10-minute window (600-second) and much of that operational reality disappears the moment the data is flattened into an average.

For routine high-level reporting, averaging is fine. For diagnosing recurring or subtle or transient performance problems, it may not be.

An Average Is a Summary, Not a Sequence

Consider two identical turbines operating side-by-side, both reporting an average output of 2.0 MW over the exact same 10-minute period:

• Turbine A operated steadily at 2.0 MW with smooth controller responses and optimal aerodynamic efficiency.

• Turbine B oscillated wildly between 1.2 MW and 2.8 MW due to unstable control behavior, excessive pitch hunting, or repeated momentary recovery events.

On the standard SCADA dashboard, both turbines look identical. In reality, Turbine B suffered severe dynamic fatigue, wasted kinetic wind energy, and exposed hardware to unnecessary stress.

This is the fundamental flaw of 10-minute data: it records the destination but completely hides the journey.

Transient Events Can Disappear Into the Average

Some turbine performance problems last only a few seconds or minutes. Individually, they may appear insignificant. If they recur frequently, however, their cumulative energy impact can become substantial.

Examples include:

Unoptimized Power-mode transitions

When a turbine toggles between standard operation, noise-reduced modes, or grid curtailment limits, controller logic dictates a precise sequence. If power dips lower than necessary during transitions, or if the controller lags when ramping back up, a recurring energy deficit develops.

A 10-minute record may show only a modest reduction in average power. Seconds-level data can reveal whether pitch, rotor speed and active power followed the expected trajectory.

Pitch-control behavior

Rapid wind gusts can cause control software to overshoot pitch adjustments. If the blades take 30 to 40 seconds to settle back to optimum angle, the turbine operates under-pitched repeatedly—a loss completely smoothed out by 10-minute averaging.

High-resolution data allows pitch behavior to be examined alongside wind speed, rotor speed and power.

Yaw misalignment detection

Yaw misalignment changes continuously as wind direction shifts and the turbine’s yaw controller responds. Ten-minute averages can smooth out short but repeated misalignment events and hide their duration and direction. High-frequency data preserves the time sequence of how the turbine responds to changes in wind direction. The added resolution supports more accurate diagnosis, energy-loss estimation and verification after corrective action.

Short-duration shutdowns and resets

A turbine may experience brief stops, resets or controller events that have little effect on monthly availability but recur often enough to reduce production. Averaged records can obscure the frequency and duration of these events.

High-Resolution Data Does Not Replace 10-Minute Data

The choice is not between conventional SCADA and seconds-level data. Each serves a different purpose.

A practical analytics strategy uses lower-resolution data to screen and prioritize the fleet, then applies higher-resolution data where additional detail can support diagnosis and corrective action.

More Data Is Not Automatically Better

Collecting higher-frequency data creates value only when it can be converted into an operational decision.

A useful analysis requires:

• Consistent timestamps and time-zone handling

• Clear signal definitions and units

• Correct mapping across turbine models and OEMs

• Data-quality checks

• A method for quantifying recurrence and energy impact

• A process for working with the operator or OEM on corrective action

Without this foundation, high-resolution data can become a large archive rather than a useful performance tool.

The objective is not to produce more plots. It is to create sufficient evidence to answer three questions:

1. What is the turbine doing?

2. Why is it doing it?

3. What action should be taken?

From Hidden Behavior to Actionable Evidence

When an owner approaches an OEM with a general statement that a turbine appears to be underperforming, the cause may be difficult to establish.

A stronger case combines:

• The operating condition in which the issue occurs

• The relevant high-resolution signals

• The sequence of turbine responses

• Comparisons with unaffected turbines

• The frequency of the behavior

• An estimate of the energy impact

• Evidence showing whether the issue persists after correction

This converts a broad performance concern into a specific, technically supported discussion.

Start by Understanding What You Already Have

Many wind plant owners are already generating or storing data at intervals shorter than 10 minutes. The immediate opportunity may not require installing new hardware. It may begin with understanding which signals are available, how frequently they are recorded and whether they are retained long enough for analysis.

Useful questions include:

• What is the native sampling interval?

• Are raw records stored, or only 10-minute statistics?

• Are setpoints, curtailment states and event codes retained?

• How long is high-resolution data archived?

• Can the owner export it without relying on manual downloads?

The answers determine which performance questions can be investigated today—and which data should be captured for the future.

See Beyond the 10-Minute Blindspot with Wind PulseSense

Catching short-duration anomalies requires looking at high-frequency (seconds-level) data—without overwhelming your O&M team with noise.

Wind PulseSense ingests both high-frequency telemetry and standard SCADA data to reconstruct the exact sequential events within every 10-minute window. By analyzing sub-interval dynamics, the solution allows you to unlock hidden value.

Explore how Wind PulseSense works or contact our team to run a high-resolution data audit on your asset portfolio.