Fault code

Every turbine controller maintains a set of status and fault codes describing what it is doing and why it stopped. They arrive in SCADA data alongside the operating channels, and they are the first thing an operations team looks at when a machine goes down.

They are also unstandardized. Each manufacturer defines its own list, the same physical condition carries different identifiers on different platforms, and code definitions change between software releases. Mapping a mixed fleet onto one taxonomy is real work, and it is the main reason a portfolio assembled by acquisition ends up with several separate reporting systems instead of one.

The deeper limitation is that a fault code is a lagging indicator. It fires when a measured value crosses a threshold the controller was configured with, which by design happens once something is already wrong. The degradation that led there was visible in temperature and vibration trends for weeks beforehand, and none of it produced a code, which is exactly the gap condition monitoring fills.

Read as a history rather than as individual events, though, the code log is genuinely informative. Rising frequency of intermittent trips on one machine, or a pattern that only appears in particular wind conditions, tells you more about the state of a turbine than any single alarm does - and it is one of the strongest inputs to ranking which unplanned downtime risk to address first.