Reliability and Quality assurance methods for power electronics converters

The strong acceleration of energy system electrification and the growing requirements for a resilient power grid, place power electronics as a fundamental enabling technology. The grid is evolving to become a digital system of systems where the critical function of transmission, control and protection is shifting from passive components to active Electronic Components and Systems (ECSs). Prominent examples in power transmission are the massive introduction of HVDC interconnections and STATCOMs, based on the Modular Multilevel Converter topology. These power electronics designs are built from a large quantity of electrically interconnected identical power electronics modules. ECSs in such equipment are critical parts of the power grid infrastructure and expected to satisfy the strict reliability and availability requirements of grid utilities. This is challenged by the complex supply chain of microelectronic components used in ECSs, where short product lifecycles and material changes for cost or regulatory motivation frequently occur. We present the foundational work for the implementation of Quality and Reliability assurance methods for the building blocks of power electronics converters, using a combination of design-for-reliability, reliability growth, validation and lifetime estimation by (highly) accelerated testing, stress screening inspection and digital product tracking designed to enable advanced data analytics. This contribution is intended to outline quality and reliability process implementation for mission critical power electronics components, where novel methods for physics-informed deep learning and advanced data analytics can fundamentally enhance the knowledge creation out of continuous product testing and lead to optimized product lifecycle management. The discussion will also cover the challenges of quantitative knowledge generation out of highly accelerated testing, conceptually designed to only be a qualitative source of product validation.

Keywords: Accelerated Testing; Reliability Assurance; Quality Assurance; Physics-Informed Deep Learning