Wind Turbine Converter Reliability Analysis Under Realistic Loading

Abstract

Electricity generated by wind energy is getting increasing attention as a successful alternative to fossil fuel electricity generators.
One of the challenges of wind energy is the reliability of the wind turbine power converter (WTPC) as its failure rate impacts the levelised cost of wind energy.
The semiconductor devices have been found to contribute to the WTPC failure rate due to the thermomechanical stress generated by the cycling of their junction temperatures and differences in the thermal expansion coefficients of their internal layers.
WTPC lifetime estimation models based on its semiconductors junction temperatures have become widely accepted by industry and academia.

The published articles in this field use wind turbine modelling to analyse the WTPC lifetime considering operating conditions and different wind turbine technologies.
However, important modelling details are required to ensure results accuracy alongside a proper analysis method to develop valid conclusions.
To emulate the actual WTPC loading, variable wind speed profiles covering a good range of wind speeds and turbulence intensities are essential for WTPC lifetime estimations.
Also, the complexity of wind speed changes requires statistical analyses in order to assess WTPC reliability related to variable wind turbine operating conditions.
Furthermore, the simulated parameters of modelled wind turbines have to be validated to ensure accurate outputs related to input wind speeds.

In this research, a new WTPC reliability analysis method is presented.
It is based on testing hundreds of variable wind speed profiles and uses statistical tools to assess WTPC reliability against wind speeds and turbulence intensity.
The method considers the simulated outputs of a validated wind turbine model for WTPC lifetime estimation.
The model is demonstrated and tested for converter topologies and control systems that are widely deployed in wind turbine applications.
In a contribution to this field, the research found that wind turbulence intensity shows a statistically significant impact on WTPC lifetime at a 95 confidence level in all tested models.
In comparing the impact of wind turbine design factors on WTPC lifetime, the research found that WTPC with direct torque control is more impacted by increasing wind speeds and turbulence intensity, resulting in 91 of the lifetime of WTPCs controlled by field-oriented control.
WTPC has been significantly developed in their design by implementing the three-level topology in the last ten years. In comparing converter topologies, the research found that 3L-NPC WTPCs lifetime achieves 2.7 to 3.9 times the lifetime of 2L-VSC depending on wind speed and turbulence intensity.
This method can be utilised to select the converter topology and control system for a longer WTPC lifetime based on the specific wind speed profile of the proposed site.
This method contributes a valid approach to assessing the reliability of future WTPC considering minor changes in the wind turbine model while the overall framework stays unchanged.