Energy Digest

A frequency-domain framework has been introduced to assess oscillation propagation dynamics in power systems integrating Inverter-Based Resources. Two quantitative indicators, Frequency Sensitivity Index (FSI) and Voltage Sensitivity Index (VSI), have been proposed to quantify system sensitivity to oscillations of different types and frequencies. These indices are derived from linear analysis transfer functions and provide insights for oscillation detection and propagation analysis.

A deep reinforcement learning framework is proposed for joint risk and anomaly identification in distribution network operation, explicitly addressing pervasive uncertainty through second-order uncertainty quantification that decomposes total uncertainty into aleatoric and epistemic components. The framework integrates distributional and Bayesian deep reinforcement learning to characterize inherent operational risks and out-of-distribution anomalies. Simulation results demonstrate the performance of the proposed DRL agent.

A coupling-aware two-level aggregation framework is proposed to address uncertainty propagation in aggregating building HVAC loads for power systems, utilizing tailored Gaussian elimination and coordinate transformation techniques at the building level and a matrix-transformation technique at the aggregator level. The framework clarifies zone-level uncertainties, decouples intra-building multi-zone coupling, and provides full-dimensional flexibility sets for tractable reformulation. The proposed framework has been validated through case studies, demonstrating accuracy, reliability, and scalability.

Aggregating distributed energy resources aims to encode their collective flexibility into a single set for efficient grid dispatch, but existing methods are conservative due to dimensional and type heterogeneity. A novel preference-oriented aggregation method is proposed to address these challenges by reformulating the Minkowski sum as a polytope projection problem and incorporating reserve dispatch preferences into aggregation. The method effectively aggregates dimensionally heterogeneous DERs and significantly reduces flexibility loss when aggregating type-heterogeneous resources.

Interharmonics are used to locate power system oscillations by extracting beating waves from synchrophasor data, which may support similar applications with phasor data alone. For certain oscillation phenomena, phasor data can be used for interharmonic extraction and oscillation source location. A cloud-based software tool is developed to enable PMU-based source location based on this method.

A new operational adequacy metric called Dynamic Reserve Margin (DRM) is proposed to assess reserve capacity in a time-varying manner. DRM measures the additional generation capacity available within a 5-minute window relative to uncertainty and contingency reserve requirements, providing an interpretable characterization of reserve adequacy. The proposed metrics also reveal reserve accessibility and stress persistence not captured by conventional reserve margin metrics.

Large data centers pose system-level stability problems when their power-electronic load is suddenly disconnected or transferred during voltage disturbances, potentially causing a significant demand removal on the power system. A new co-planning framework assesses voltage ride-through capability in data centers to ensure IT service while limiting customer-initiated load reduction. The study finds that internal VRT capability becomes a binding constraint for data center capacity, with grid-forming mode providing greater ride-through margin than current-limited grid-following mode under similar network fault conditions.

Thyristor rectifiers used in renewable power-to-hydrogen systems may violate grid-code limits due to harmonics; a new operational harmonic coordination scheme proposes coordinating operation of multiple electrolyzers to mitigate these effects; the method can increase profit by 31% and reduce transformer investment costs.