A unified MMC reliability evaluation based on physics-of-failure and SM lifetime correlation. (March 2019)

• Record Type: Journal Article
• Authors: Xu, Jiayou; Wang, Longjun; Li, Yan; Zhang, Zheng; Wang, Gang; Hong, Chao
• Is Part Of: International journal of electrical power & energy systems. Volume 106(2019)
• Page Start: 158
• Page End: 168
• Publication Date: 2019-03
• Subjects: Modular multilevel converters -- Reliability -- Lifetime estimation -- Physics-of-failure -- Copula function; Electrical engineering -- Periodicals; Electric power systems -- Periodicals; Électrotechnique -- Périodiques; Réseaux électriques (Énergie) -- Périodiques; Electric power systems; Electrical engineering; Periodicals; 621.3
• DOI: 10.1016/j.ijepes.2018.09.044
• Languages: English
• ISSNs: 0142-0615

Abstract

Highlights: A unified physics-of-failure-based lifetime estimation approach for IGBT modules and capacitors. Redundant MMC system reliability modeling considering SM lifetime correlation. Explaining the reasons for the change in the reliability of the MMC system after considering the correlation. Obtaining component lifetime Weibull distributions using Monte Carlo method at a specific mission profile. Abstract: The reliability of a modular multilevel converter (MMC) is determined both by the component lifetime distribution and the MMC structure. In this paper, a unified submodule (SM) component lifetime is estimated based on the physics- of- failure, and the overall reliability of the MMC system is evaluated considering the SM lifetime correlation. First, the component lifetime estimation is performed in three sub-steps. (a) The current stress of each component, i.e., the insulated gate bipolar transistor modules (IGBTs) and capacitors, is computed according to the working principle of the MMC. (b) The core temperatures of the components are calculated using thermal equivalent networks. For the IGBTs, the thermal cycles are extracted to compute the cumulative damage. (c) The lifetime of each device is estimated in a static state using the lifetime model and damage model. Second, the lifetime Weibull distribution of each component is obtained using the Monte Carlo method. This distribution links the component lifetime estimation and MMC reliability analysis. Finally, considering the SM lifetime correlation, the overall reliability of an MMC system with redundancy is evaluated using a Copula function. The reasons for the impact of SM lifetime correlation on MMC reliability are revealed in a case study.