With the massive deployment of direct current (DC) technologies on the grid, particularly photovoltaics and grid-connected battery energy storage systems (BESS), a growing share of electrical energy now flows through static power converters. Unlike classical grids dominated by rotating machines, which benefit from high natural inertia, power-electronics-dominated networks exhibit very limited inertia and may therefore experience highly dynamic voltage spikes, voltage drops, or even complete collapse. Some research focuses on synthetic inertia, emulated through specific control strategies implemented in static converters, but these approaches depend on equipment manufacturers and do not rely on established standardization. Another approach consists in designing dedicated equipment specifically intended for the active stabilization of low-inertia power systems, which is the direction explored in this PhD project.
A particularly demanding case concerns MVDC grids, which by construction rely entirely on static power converters, therefore exhibiting extremely low natural inertia, and requiring the use of converters based on specific technologies. Within the framework of this PhD, we propose the study and proof of concept of a converter connected to an MVDC electrical network operating between 6 and 12 kV, capable of injecting or absorbing very high levels of power in a transient manner, on the order of ten megawatts for durations ranging from 10 µs to 100 ms. The system will rely on an isolated Dual Active Bridge (DAB) topology, with a medium voltage capacitive DC bus at its primary.
This power electronics topic presents several technological bottlenecks. Synthetic switches (series-connected SiC devices, as investigated in a previous PhD in the laboratory) will have to be implemented in a real DAB converter. A highly isolated power supply for the gate drivers of these synthetic switches will need to be designed. The medium-frequency DAB transformer must be designed to transfer very high transient power while minimizing volume. Particular attention will therefore be paid to transient-oriented design, with the objective of identifying the key parameters that maximize, within a complex structure, the ratio between the converter rated power and its peak power.
Potential extensions toward other pulsed-power applications that could benefit from such a converter will be explored, taking into account their specific constraints.