Fault Response of Inverter-Based Resources and Their Impact on Grid Protection Systems

Abstract

Inverter-based resources (IBRs) now supply a substantial portion of modern power systems, yet their short-circuit behavior during grid faults differs fundamentally from that of synchronous machines. These differences challenge traditional protection schemes and motivate the need for models that capture the essential physics of IBR fault response while remaining tractable for large-scale studies. This dissertation develops a physics-informed, data-driven reduced-order model (ROM) of IBR fault current tailored for protection applications. The work begins with requirements analysis that translates interconnection standards, fault ride-through (FRT) requirements, and protection needs into modeling targets. These requirements specify the necessary outputs (sequence-domain currents and voltages), behaviors (current limiting, negative-sequence control), interfaces (compatibility with short-circuit and coordination tools), and quality metrics (relay-relevant timing and magnitude accuracy). Guided by these requirements, detailed electromagnetic transient (EMT) models are developed in Simulink and PSCAD to generate IBR fault responses under diverse control modes, grid strengths, and fault conditions. A theoretical analysis of IBR behavior during grid faults is then performed to characterize the inverter fault response. This includes temporal decomposition into subtransient, transient, and steady-state regimes; dq- and sequence-frame current trajectory analysis; and harmonic and interharmonic studies under PWM saturation, current limiting, PLL and sequence-extractor imperfections, and grid-code reference-current policies. These investigations show that IBR fault currents follow a structured sequence of event-driven dynamic regimes. Based on these insights, a parameterized ROM architecture is formulated in which the total current is represented as the sum of subtransient, transient, and steady-state components with physically interpretable parameters. A data-driven generalizability layer allows these parameters to vary systematically with pre-fault operating point and sequence-voltage depression using empirically fitted sensitivity coefficients. The ROM is implemented across multiple platforms (Python, Simulink, PSCAD) to support interoperability with EMT tools and protection-oriented short-circuit programs. A validation framework is developed using experimental data from a commercial off-the-shelf (COTS) inverter, detailed EMT simulations, and relay-based assessments. Quantitative metrics such as cumulative error, RMSE, phase-shift alignment, and a composite performance metric demonstrate that the ROM reproduces EMT and experimental waveforms with acceptable accuracy, especially beyond the first cycle after fault inception. Relay studies indicate that the ROM captures key protection behaviors, including distance and overcurrent element operation, negative-sequence directional response, and voltage supervision, with timing differences generally within one cycle. The ROM of the IBR presented in this dissertation was developed using a systems-engineering approach, beginning with requirements-based problem formulation, followed by a theoretical analysis of inverter fault dynamics, and concluding with the construction of a reduced-order model intended for use in protection studies and grid-level applications.