Field-Experience Based Root-Cause Analysis of Power-Converter Failure in Wind Turbines
Journal article, 2015

The frequent power-converter failure experienced in wind turbines has a strong economic impact through both the related turbine unavailability and the maintenance cost. Up to now, the prevailing mechanisms and causes underlying the converter failure in wind turbines are mostly unknown. Their identification is, however, a prerequisite for the development of effective solutions. This paper describes a multitrack empirical approach to failure analysis including systematic field-data evaluation, exploration of the real converter operating environment, and postoperational laboratory investigation of converter hardware. The analysis is carried out for two widely used multi-MW wind turbines with low-voltage, insulated-gate bipolar transistor-based converters (topology 1: doubly fed induction generator with partially rated converter, topology 2: induction generator with fully rated converter). The findings suggest that the principle failure mechanisms of power electronics found in other applications, namely solder degradation and bond-wire damage, play a minor role in the investigated types of wind turbines. Instead, the analysis reveals indications of insufficient protection of the converter hardware against the environment (salt, condensation, and insects) as well as indications of electrical overstress.

failure

power electronics

root-cause analysis

reliability

wind turbine

Converter

Author

Katharina Fischer

Chalmers, Energy and Environment, Electric Power Engineering

T. Stalin

Vattenfall

H. Ramberg

Volvo Cars

J. Wenske

European Wind Energy Association

G. Wetter

Swerea

Robert Karlsson

Chalmers, Energy and Environment, Electric Power Engineering

Torbjörn Thiringer

Chalmers, Energy and Environment, Electric Power Engineering

IEEE Transactions on Power Electronics

0885-8993 (ISSN) 19410107 (eISSN)

Vol. 30 5 2481-2492 6917013

Subject Categories

Other Electrical Engineering, Electronic Engineering, Information Engineering

DOI

10.1109/tpel.2014.2361733

More information

Latest update

4/5/2022 6