An Effective Real-World Load-Side Operational Strategy to Mitigate Sub-Synchronous Oscillations in Inverter-Dominated Power Systems
The rapid growth of inverter-based resource (IBR) integration and large-scale data centers has introduced a new class of sub-synchronous oscillation (SSO) risks in weak grids. This paper presents a novel, effective, real-world load-side operational strategy to mitigate SSO in high-density industrial loads. The authors analyze the case of a Texas-based data center (hereinafter referred to in this article as "the company"), a hyperscale facility with a capacity of 400 MW, located in a region with a weak grid and a short-circuit ratio (SCR) below 3. During the second half of 2024, persistent SSO events in the 20–24 Hz range were detected and validated via spectral analysis and time-domain data, which is consistent with recent reports from the Electric Reliability Council of Texas (ERCOT) and the International Council on Large Electric Systems (CIGRÉ in French) on increasing SSO frequency in power electronics-dominated systems. The proposed mitigation strategy comprises three coordinated actions: (1) disabling Fast Reboot in miners' motherboards to avoid synchronized transient load responses; (2) scheduling load ramp-up operations during periods of lower IBR generation variability, thereby reducing simultaneous interactions between generation-side inverter controllers and load-side PSU/PFC controllers; and (3) tuning PSU firmware to increase the phase margin from 25° to 67°, improving damping and small-signal stability. The company's engineering team validated the proposed strategy through a controlled field test at the Corsicana facility, in which the operating load was increased from approximately 300 MW to 330 MW in staged 2 MW increments under real-time power-quality monitoring. This work supports the feasibility of load-side operations as complementary or alternative solutions to generation-side SSO mitigation strategies, thereby improving operational reliability in IBR-penetrated industrial parks.
Authors
- D Flores (ORCID: https://orcid.org/0000-0001-7358-0980)
- Manuel Briceño (ORCID: https://orcid.org/0000-0003-1433-5074)
- Leandro Delgado (ORCID: https://orcid.org/0000-0001-9485-766X)
- Luis Díaz
- Jose Canelon
- Dewei Kong
- Ricardo S. Carrillo
- Ricardo Carrillo
- Frank Durant
- Steve Yates
Institutions
- Universidad del Zulia (VE)
- Doctors Company (United States) (US)
- Pontificia Universidad Católica Argentina (AR)
- Shenzhen Technology University (CN)
- Institute of Automation (DE)
- Massachusetts Institute of Technology (US)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-06-29
- DOI
- https://doi.org/10.5281/zenodo.21041617
- Primary Topic
- Microgrid Control and Optimization
- Type
- article
- Field-Weighted Citation Impact
- 0.00