Reliability is becoming more and more important as the size and number of installed Wind Turbines (WTs) increases. Very high reliability is especially important for offshore WTs because the maintenance and repair of such WTs in case of failures can be very expensive. WT manufacturers need to consider the reliability aspect when they design new power converters. By designing the power converter considering the reliability aspect the manufacturer can guarantee that the end product will ensure high availability. This paper represents an overview of the various aspects of reliability prediction of high power Insulated Gate Bipolar Transistors (IGBTs) in the context of wind power applications. At first the latest developments and future predictions about wind energy are briefly discussed. Next the dominant failure mechanisms of high power IGBTs are described and the most commonly used lifetime prediction models are reviewed. Also the concept of Accelerated Life Testing (ALT) is briefly reviewed.

An increasing number of power-electronic-based distributed generation systems and loads generate not only characteristic harmonics but also unexpected harmonics. Several methods, such as impedance-based analysis, which are derived from the conventional average model, are introduced to perform research about the harmonic interaction. However, it is found that the linear-timeinvariant-based model analysis makes it difficult to analyze these phenomena because of the time-varying properties of the powerelectronic-based systems. This paper investigates a grid-connected converter by using the harmonic state-space (HSS) small-signal model, which is based on a linear time-varying periodically theory. The proposed model can include the switching behavior of the model, where it makes the model possible to analyze how harmonics are transferred into both the ac-side and dc-side circuits. Furthermore, a harmonic matrix of the grid-connected converter is developed to analyze the harmonic interaction at the steady-state behavior. Besides, the frequency-domain results are compared with time-domain simulation results by using the HSS modeling to verify the theoretical analysis. Experimental results are finally discussed to verify the proposed model and study. Index Terms-Harmonic coupling analysis, harmonic state-space (HSS) modeling, three-phase grid-connected converter. I. INTRODUCTION W ITH the increased use of power-electronic-based distributed generation (DG) Systems, the stability and the dynamic performance of the system are important issues today [2]-[6]. Particularly, various DG systems are gathered into the same grid network, where complex connections, active control, and bidirectional current flow can make it difficult to analyze the dynamics and interaction of the power-electronic-based systems [3], [7]-[11]. Besides that, even if "N" identical power converters, which operate simultaneously or independently, are connected to the same bus, the cancellation, generation,

No abstract

With the continuously increasing demand for energy and the limited supply of fossil fuels, renewable power sources are becoming ever more important. Knowing that future energy demand will grow, manufacturers are increasing the size of new wind turbines (WTs) in order to reduce the cost of energy production. The reliability of the components has a large impact on the overall cost of a WT, and press-pack (PP) insulated gate bipolar transistors (IGBTs) could be a good solution for future multi-megawatt WTs because of advantages like high power density and reliability. When used in power converters, PP IGBTs are stacked together with other components in a clamping mechanism in order to ensure electrical and thermal contact. Incorrect mechanical clamping of PP IGBTs has a negative impact on their reliability and consequently on the reliability of the WT. In this study the impact of mechanical clamping conditions on the static thermal distribution among chips in PP IGBTs is investigated.

The wind power industry is continuously developing bringing to the market larger and larger wind turbines. Nowadays reliability is more of a concern than in the past especially for the offshore wind turbines since the access to offshore wind turbines in case of failures is both costly and difficult. Lifetime modeling of future large wind turbines is needed in order to make reliability predictions about these new wind turbines early in the design phase. By doing reliability prediction in the design phase the manufacturer can ensure that the new wind turbines will live long enough. This paper represents an overview of the different aspects of lifetime modeling of high power IGBTs in wind power applications. In the beginning, wind turbine reliability survey results are briefly reviewed in order to gain an insight into wind turbine subassembly failure rates and associated downtimes. After that the most common high power IGBT failure mechanisms and lifetime prediction models are reviewed in more detail.