Energetic availability of wind turbines

The KPI “availability” is often mentioned in the wind industry – but rarely truly understood.Anyone who wants to capture the actual technical performance of a wind turbine (WT) should focus on energetic availability (EV).

Thanks to modern artificial intelligence (AI), this metric can now be determined more realistically and accurately than ever before, since AI can account for effects such as turbulence, temperature, air density, and wind farm layout.

What Is Energetic Availability?

Energetic availability (EV) measures how much of the potential energy production of a wind turbine was actually available during a given period – that is, without technical or operational restrictions.

Unlike classical temporal, contractual, or technical availability, EV does not just consider whether a turbine is running, but how much energy is lost when it isn’t.

A downtime during strong wind is energetically much more relevant than one during weak wind.

Why Energetic Availability Is a Strong KPI for Comparing Turbine Performance

1. Realistic Performance Assessment

A short outage during turbulent, high-wind conditions can cause significant energy loss.EV weights such losses by wind conditions, showing how much energy is actually left on the table.

2. More Precise Root Cause Analysis

Energetic availability helps identify critical downtimes with high energy impact.This enables targeted prioritization of maintenance and service efforts.

3. Objective Comparability

EV normalizes for wind availability.This means turbines, models, and sites can be compared fairly, independent of wind frequency or seasonal effects.

How AI Makes the Calculation Far More Accurate

Historically, energetic availability was calculated using simple wind speed models or power curves.These approaches oversimplify reality and ignore many influences that significantly affect actual energy yield.

With AI-based models, this has fundamentally changed.Modern machine learning approaches (like those used by Turbit for performance monitoring) learn each turbine’s individual behavior under real site conditions.

These models consider parameters such as:

• Turbulence and wind direction: influence the aerodynamic efficiency of the turbine

• Temperature and air density: affect torque and energy conversion

By incorporating these factors, AI can predict potential power output far more accurately than traditional power-curve-based models – and therefore calculate energetic availability much more precisely.

Example

As a technical operator, I want to compare two turbines in a fictional two-turbine wind farm and answer the question: Which one operates more efficiently?

Of course, I could simply compare total energy production over a given period.However, there’s a problem: Turbine 2 is partially shaded by Turbine 1, and vice versa. Depending on wind direction frequency in different years, one or the other will appear more performant.

Since both turbines are at the same site, they experience identical temperature and air density, so those factors cancel out here – but they would matter if the turbines were at different locations.

Finally, terrain elevation (height above sea level) also affects the power curve. None of these are true technical performance factors, they are environmental or physical influences, not issues like damaged rotor blades or component faults.