Root Cause Analysis of Frequency Oscillations Observed in Ontario's Distribution System

Recurrent frequency oscillations have been observed in Ontario's distribution system, initially misinterpreted as mains loss by the rate of change of frequency relay, resulting in 16 false trips of a biomass power plant over 1.5 years. These oscillations, ranging from 5.6 to 6.45 Hz, persisted across varying load levels without coinciding faults or sudden system changes. Traditional simulation-based investigations failed to replicate these conditions due to high uncertainties in system parameters, prompting temporary adjustments to relay settings to prevent further nuisance trips. This paper details root-cause analysis approach that contrasts with conventional methods. First, the system is treated as a black box, applying signal processing tools to field-recorded data to extract signal features and eliminate unlikely scenarios—providing early insight without model dependence. Next, sensitivity analysis identifies critical parameters, refining the model to isolate the underlying issue. Using the shooting method and small-signal analysis, Hopf bifurcation is pinpointed as the cause, generating limit cycles and local frequency oscillations. Analysis showed that systems in UPF mode are more susceptible to bifurcation than those in PV mode. Finally, control parameter modifications are proposed to broaden the system load margin and improve stability. Findings are validated through Real-Time Digital Simulator and Hardware-in-the-Loop testing.