Markov Modeling of the Availability of a Wind Turbine Utilizing Failures and Real Weather Data

Abstract

The maintenance and reliability issues of the wind farms are a matter of high importance both for the offshore and onshore constructions. The technology applied in the wind turbines is fast growing and the wind farms expand rapidly all over the world. European Union countries have set high targets for electricity production through wind till 2020. The maintenance is very important part for a wind farm because we can prevent very important faults even disasters in vital parts of the Wind Turbines (WTs) that may cost hundred thousand euros and lead to long downtimes. Condition monitoring methods are applied in order to prevent high cost damages, save money from maintenance and increase the availability of the system. Moreover, the weather phenomena play a very important role for both the operation and maintenance of the system. For example, in the case of wind or bad weather we cannot have access for maintenance and repair, as a result the downtime increases. The weather parameter is more important in offshore wind farm because the accessibility is more difficult and costly. Using Markov chains we develop a model describing the availability of a wind turbine considering the wind intensity and the operational condition or the downtime of a wind turbine.