Availability

Availability measures how much of the time a machine was ready to produce. Time-based availability counts hours; energy-based availability weights them by the energy that was on offer, which matters because an outage during a windy week costs far more than the same outage in calm conditions. When a contract quotes a single figure without saying which, it is almost always the time-based one, because it is the flattering one.

The distinction is where availability claims get slippery. An availability warranty may exclude grid outages, curtailment, extreme weather, force majeure, and scheduled work, so two sites quoting the same figure can mean quite different things. When comparing numbers, the exclusions matter more than the percentage.

High availability also does not mean high output. A turbine can be available every hour and still underproduce for the whole period through a control fault or a degraded blade, because it was running the entire time - just not well. That kind of loss only appears against a measured expectation such as a learned power curve, which is exactly the gap availability reporting leaves open.

In practice this makes availability a necessary and insufficient metric. It is the right way to talk about unplanned downtime and the wrong way to talk about production, and a fleet managed to availability alone will hit its target while quietly losing energy on machines that never stopped.