Reliability Modelling of Power Electronics with Mission Profile Forecasting for Long-Term Planning

Abstract

The power electronics system reliability is strongly dependent on the environmental and operating conditions at the installation site. During the reliability modelling, a mission profile-based procedure is traditionally employed to account for this dependency. To accurately estimate the reliability, mission profiles need to include the low resolution time intervals (e.g., several minutes per sample). The main forecasting challenge is the accurate prediction of such profiles for long-term prediction horizons, such as several years ahead. Therefore, in this paper, a mission profile forecasting method suitable for long-term reliability-oriented planning is proposed. The benefit of this method includes multi-year mission profile prediction with high time resolution (i.e., 1 minute per sample) suitable for power electronics reliability studies. A case study reveals that the proposed forecasting method enables power electronics modelling with less than 5% relative error in reliability estimation over different time horizons.

Multi-Year PV Generation Planning Incorporating Power Electronics Impacts in Sizing Decisions

“…4. The relevant stress information are then extracted from the thermal loading profile to determine the number of cycles to failure N f [14], [15]:…”

“…PV systems for residential applications is studied, where PV arrays and power electronics are modeled as in [15]. The cost information are based on [16].…”

“…The available PV inverter and PV array values are P load (T 0 )×{0.33, 0.67, ...1.66} and P nom P E ×{0.5, 0.6, ..., 1.5}, respectively. The input high time resolution profiles of the environmental conditions used for the estimation of PV inverters LC are taken from [15]. It is assumed that the yearly profiles do not change over T P H .…”