Renewable energy generation plays a fundamental role in sustainable development. Hydropower plants provide highly reliable power generation, while synchronous generators contribute essential rotational inertia to power system stability. In this context, variable speed hydrogenerators have emerged as an important technology for the energy transition, particularly in pumped storage and conventional hydropower plants, owing to their potential to enhance reservoir management, energy storage, operational flexibility, and overall system performance. This paper investigates the efficiency characteristics of large variable speed hydrogenerators by analyzing electromagnetic and mechanical losses as functions of rotor speed. The proposed methodology enables the mapping of machine efficiency over a wide range of operating conditions, considering different load levels and rotational speeds. The methodology is applied to a 160 MVA hydrogenerator, providing a comprehensive assessment of its efficiency and performance under variable speed operation.
Modern power systems are experiencing a rapid increase in inverter-based resources (IBRs). Not only do IBRs operate at high switching frequencies, but also often lead to erosion of system inertia. As such, IBR-intensive power systems manifest complicated, fast, and transient behavior that challenge their electromagnetic transient (EMT) simulations. Although EMT simulation algorithms are, in essence, well-suited for capturing high-frequency phenomena, doing so in the context of complex power systems with a large number of high-frequency converters exceeds the capabilities of existing EMT algorithms. Repetitive factorizations of large conductance matrices is the core computational challenge that cripples the performance of EMT simulation of large, IBR-intensive systems. To reduce the computation time while retaining the established modeling capabilities of commercial EMT tools, this paper investigates a GPU-CPU co-simulation framework in which an IBR-intensive portion of the electrical network, including its device calculations and nodal network solution, is executed on the GPU, while the remaining portion is retained in a commercial CPU-based EMT simulator. The two portions are coupled through transmission-line models, requiring only boundary quantities to be exchanged between the CPU and GPU. A Schur complement-based partial re-factorization strategy is also evaluated for cases in which switching affects only a small portion of the GPU-modeled conductance matrix. Simulation results using a modified IEEE 39-bus system demonstrate close agreement with the commercial EMT simulator and speedups of up to 43.2 times for a system containing 1000 IBRs.
This paper presents a framework for integrating real-world operational data—including advanced metering infrastructure, Supervisory Control and Data Acquisition (SCADA), and other field measurements— into distribution network models to support realistic application development and operational controls. This methodology uses detailed metering data and supervisory sensor inputs from utility partners, systematically mapping them to OpenDSS models. It features a rigorous verification and validation process to ensure that the simulated models accurately represent both the physical assets and operational states of the grid, enabling realistic simulation of grid physics and supporting confident implementation of advanced control strategies. For demonstration, a power factor correction application was developed and integrated within this framework, achieving near-unity power factor at the source bus and effectively addressing voltage violations under high distributed resource activity. This methodology was applied and validated across two distinct distribution circuits in Vermont—South Hero and South Alburgh—demonstrating its practical effectiveness in real-world settings. By incorporating time-series metering and diverse sensor data streams, this approach enables greater situational awareness and operational flexibility across distribution system modeling, control, and advanced application deployment.
Nonlinear loads in industrial electrical systems cause harmonic distortion, reactive power losses, and reduced efficiency. While passive filters offer proven mitigation, accurately predicting their performance before installation remains challenging due to the cost and complexity of physical testing. Conventional analytical methods rely on simplified assumptions that poorly represent real facility conditions. This paper presents a digital replica-based simulation framework that predicts passive filter performance using measured voltage and current waveforms (≤80 μs resolution) from target facilities. Developed in MATLAB Simulink with Simscape Electrical, the framework incorporates detailed transformer modelling derived from systematic parameter extraction and simulates both filter-active and bypass modes to calculate active power, reactive power, total harmonic distortion, power factor, and energy consumption. Validation across five real-world installations (1400 A to 3200 A ratings) demonstrated average absolute prediction error of 0.99 percentage points and average relative error of 0.42. For the detailed validation case, the simulated results predicted reductions in active power, reactive power, and THD, alongside power-factor improvements across all phases. Unlike analytical approaches requiring idealized load models, this data-driven methodology directly incorporates facility electrical signatures, reducing the uncertainty associated with constructing an explicit equivalent model of the facility load. The framework provides facility managers and filter suppliers with a scalable, non-invasive tool for quantifying energy-saving potential before capital investment, reducing financial risk and supporting evidence-based procurement decisions.
Long-term capacity and transmission expansion planning develops the infrastructure needed to meet projected demand, using economic-dispatch models that streamline plant representation at the gigawatt scale. However, comprehensive planning also necessitates reliability assessments, which requires power flow models for dynamic simulations to ensure system reliability under contingencies. This paper proposes a systematic engineering workflow that unifies capacity-expansion results with reliability studies across multiple operating snapshots featuring high penetrations of inverter-based resources. By closing this modeling gap, the workflow enhances reliability analysis, supports compliance with planning standards, and improves system performance during disturbances. It also enables grid planners to evaluate future scenarios with substantial new generation by building study cases that determine whether system inertia and governor response are sufficient, grid voltages recover adequately after faults, and frequency load shedding events might occur. Finally, studies on the Western Electricity Coordinating Council 2030 planning system demonstrate that detailed, disaggregated modeling yield more accurate reliability assessments.
