Accelerating detailed simulations of an HVDC system based on modular multilevel converters in a multi-core environment

Recently, a modular multilevel converter (MMC) technology was introduced suitable to extend the application of voltage source converters to high power HVDC systems. The MMC technology is based on a modular topology which builds on elemental constituent parts called submodules (SMs). This modularity enables reduced switching losses and increased power quality. Concomitantly with the development and adoption of the MMC technology, there is an increasing need for simulation models capturing the behavior of the technology at different levels of granularity. Detailed models able to reproduce the dynamics of the SMs are essential to study voltage balancing algorithms which ensure stable operation of the converters. However, the large number of semiconductor switches utilized in this technology makes detailed models computationally burdensome. In this paper, the potential of parallel computing is explored by making use of a multi-core model implementation of an MMC-based HVDC system in the PSCAD/EMTDC environment. As demonstrated through different test cases, this approach achieves reduced computation times without resorting to approximated models.