Electrical Insulation Challenges in Future Smart Grids: Power Electronics Meets High-Voltage Technology

The need to interface a large number of electrical sources and loads having different voltage and/or frequency levels (including DC) is ubiquitous. This function is typically provided by a power electronic converter, which uses semiconductor switches to achieve full control of terminal voltages and currents. In recent years, much research has focused on extending these “electronic transformers” to operate at higher voltages (e.g. 10 kV) and corresponding high power ratings (>10kW). They would allow interfacing medium-voltage AC or DC distribution grids with, e.g., renewable energy sources, battery storage, fast electric vehicle chargers, etc.

Galvanic isolation in electronic transformers is achieved through a “conventional” magnetic transformer, which is however operated at frequencies up to a thousand times higher than low-frequency (50 Hz) transformers. This decreases the required magnetic core volume and therefore substantially increases the conversion power density (up to 30 kW/dm³) and/or efficiency (up to 99.7%).

The voltage stress on the medium-frequency transformer insulation (and other inverter-fed medium-voltage loads in general) thus consists of a succession of voltage pulses (up to 100 thousand per second) of several kilovolts in amplitude. The transitions between voltage levels (two in the simplest case) occurs with speeds up to several tens of kilovolts per microsecond. Moreover, the high conversion power density and the unavoidable conversion losses lead to high insulation temperatures (>100°C) even in high-efficiency transformer designs

The research carried out within this project aims at quantifying the degradation of polymeric insulating materials (main constituent in dry-type transformer insulation) under these operating conditions. To do so, a “converter stress simulator” has been designed (see image). Experiments are currently carried out to quantify the influence various parameters such as pulse amplitude, DC offset, fundamental frequency, pulse rise time, duty cycle, electrode material and foil impregnation (air/oil) on sample lifetime. This data set is used as an empirical basis for inferring the main degradation mechanisms leading to failure, as well as to identify the most critical influencing parameters. In addition, diagnostic tools have been developed to help identify the main degradation mechanisms. These include a high-precision dielectric spectrometer for sub-Picofarad samples as well as optical partial discharge monitoring.