Energy Digest
Summary
Summary
Summary
Summary
Summary
Summary
Summary
Summary
Summary
Technical Papers & Research
AI-curated academic research for power system engineers
Grid Operations & Resilience 3 papers
A new power-swing blocking scheme uses higher-order incremental quantities to improve performance in systems with inverter-based resources, addressing interoperability issues with faster power swings and unconventional impedance trajectories. The proposed scheme outperforms conventional dual-blinder methods in simulations, particularly in challenging swing scenarios. The scheme is designed to handle the limitations of traditional blinders, such as not operating when the measured impedance is within the blinders.
HVAC systems consume a significant portion of a building's energy, prompting the development of energy- and cost-efficient control strategies. Conventional strategies are limited to simple, reactive control, and this paper proposes two online algorithms for designing control inputs with a limited computational budget, achieving results comparable to those of a baseline algorithm with unlimited budget. The proposed algorithms are computationally efficient, allowing for effective control of HVAC systems.
A cost-effective submodule with fault-blocking capability for modular multilevel converters is proposed, featuring a reduced component count compared to existing SM topologies. The proposed submodule uses two fully-controlled switches to block DC-side fault-currents, with one variant employing just two switches. The functionality is validated through experiment and demonstrated through a test simulation for a high-voltage direct-current transmission system.
Energy Storage & Markets 1 papers
A new framework using polynomial diffusion models is proposed for real-time joint electricity price and load forecasting in data centers, enabling the shift of electricity purchases from high-price to low-price periods. The proposed approach achieves significant cost savings, up to 2.63 times more than a traditional rule-based strategy. It leverages closed-form conditional expectations of polynomial diffusion models to minimize electricity procurement costs.
Other 5 papers
A new methodology has been proposed to disaggregate county-level building electrification projections onto representative distribution feeders to evaluate their impact on distribution upgrade costs. The approach combines bulk system end-use electrification projections with representative feeder sets to generate building-level electrification measures. This allows for more realistic distribution system upgrade costs to be calculated, capturing localized load variations and spatial diversity of electrification demand.
A policy distillation framework using explainable boosting machines (EBMs) is proposed to obtain a transparent wind farm power control policy, reducing imbalance costs by 82% in real-world data, outperforming other control strategies. The framework incorporates a value-oriented perspective and expert knowledge to increase model value and justify the control strategy to stakeholders. The approach achieves this using data from the Belgian grid and an offshore wind farm.
EnergyBench is a live, gate-aligned benchmark for day-ahead forecasts of European power markets, evaluating 18 time-series foundation models and their performance without and with weather covariates. The benchmark shows that including weather covariates reduces the error of each model by 37-51% on matched races, with some models outperforming a LightGBM model. The grid operators' pre-gate forecasts remain ahead in wind and solar forecasts, but the models with weather covariates close the gap in load forecasts.
A unified framework for multi-terminal DC (MTDC) stability analysis is presented for bipolar high-voltage direct current (HVDC) systems, enabling sensitivity studies and providing a closed-form expression for stability improvement. The framework combines analytical network representation with black-box converter models and is valid for both symmetric and asymmetric conditions. It is demonstrated through several case studies, highlighting the importance of considering intra-station converter couplings and network parameter uncertainty.
The article proposes using data-driven methods to design current controllers for modular multilevel converter-based HVDC transmission systems, which can deliver less biased models that capture parameter uncertainties and control-loop interactions. The proposed method, integral Subspace Predictive Control, is synthesized directly from input-output data measured in a high-fidelity real-time digital simulator. The method is compared with a state-of-the-art solution and validated through simulation studies in MATLAB/Simulink.
Was this digest helpful?