A Guideline for Reliability Prediction in Power Electronic Converters

Abstract

Reliability prediction in power electronic converters is of paramount importance for converter manufacturers and operators. Conventional approaches employ generic data provided in handbooks for random chance failure probability prediction within useful lifetime. However, the wearout failures affect the long-term performance of the converters. Therefore, this paper proposes a comprehensive approach for estimating the converter reliability within useful lifetime and wear-out period. Moreover, this paper proposes a wear-out failure prediction approach based on a structural reliability concept. The proposed approach can quickly predict the converter wear-out behavior unlike conventional Monte Carlo based techniques. Hence, it facilitates reliability modeling and evaluation in large-scale power electronic based power systems with huge number of components. The proposed comprehensive failure function over the useful lifetime and wear-out phase can be used for optimal design and manufacturing by identifying the failure prone components and end-of-life prediction. Moreover, the proposed reliability model can be used for optimal decisionmaking in design, planning, operation and maintenance of modern power electronic based power systems. The proposed methodology is exemplified for a photovoltaic inverter by predicting its failure characteristics.

System-Level Design for Reliability and Maintenance Scheduling in Modern Power Electronic-Based Power Systems

“…The random chance failures are associated with overstressing of the components triggered by sudden single event such as overvoltage and overcurrent. Furthermore, the aging failures are associated with the wear-out of power modules, capacitors and Printed Circuit Boards (PCB) solder joints [3], [14], [23], [37], [38].…”

“…In order to predict the wear-out failure probability of these components, the concept of structural reliability has been adopted [11], [23], [50]. Based on this approach, the components resistances are compared to the applied stress and the corresponding lifetime consumption is obtained by using the linear Miner's rule as:…”

“…In order to identify the distribution function of LC, Monte Carlo simulations can be used for modeling the impact of uncertainties. This procedure has been explained in detail in [22], [23]. The wear-out failure probability of each device can be presented by a Weibull distribution as:…”

“…This is required to find the converter loading due to the fact that the converter reliability depends on its operating conditions. For instance, the impact of converter loading on the stress of semiconductor devices and capacitors are demonstrated by experiments in [22], [23], [31], [53], [63] [37]. Thus, the converter loading is determined by PF analysis and its functionality in the system with respect to the energy management strategy.…”

“…Furthermore, the converter-level activities are associated with reliability modeling and enhancement in power converters using the lifetime models of its components. The converter reliability is predicted based on a stressstrength analysis comparing applied stresses induced by a mission profile to its components lifetime [23]. Therefore, the converter reliability depends on its components lifetime, climate and operating conditions, converter topology [24]- [26], control algorithm [22], [27]- [31], and cooling system etc.…”