Derisking of subsynchronous torsional oscillations in power systems with conventional and inverter-based generation
Nicolas Bonafé, Julian Freytes, Hani Saad
TL;DR
The paper tackles SSTI risks in power systems with conventional and inverter-based generation by applying a derisking methodology that combines UIF screening and the complex torque coefficients method within EMT simulations. Using a realistic Aramon oil plant adjacent to a 1 GW HVDC link, it demonstrates how multi-mass shaft dynamics and frequency-dependent electrical damping can reveal instability in the first torsional mode after network perturbations, even when higher modes remain damped. It then presents mitigation strategies, including blocking filters and a narrow-band SSDC design, and discusses a protection scheme to prevent damaging interactions while maintaining grid reliability. A key contribution is a practical, step-by-step industrial workflow that emphasizes data and resource requirements, collaboration among asset owners, and the feasibility of using black-box models to derisk SSTI in real-world settings.
Abstract
This article proposes an application of a derisking methodology of subsynchronous torsional oscillations considering a realistic use case. The main objective is to summarize and draft a synthetic paper clarifying the complete methodology highlighting the main information needed step-by-step. For exemplification, a real model from a decommissioned oil power plant is adopted, where a fictitious high voltage direct current power link is connected. In this article, stress is laid on details of the application of the derisking methods: the unit interaction factor and the complex torque coefficients method. Then, the different steps to obtain results are explicitly explained. Moreover, the design and tuning process of supplementary subsynchronous damping controller is discussed. This mitigation section uses minimal information to correctly damp the unstable oscillations, as one would expect from industrial projects where the data sharing may be limited. Finally, the resources needed to perform each step of the study were summarized.
