Energy Digest

Large AI data center loads can introduce persistent sub-synchronous active-power oscillations that may impact nearby generators by exciting torsional modes and increasing shaft stress. A model-based framework is presented to evaluate the risk of hydro-generator shaft fatigue under oscillatory loading, considering factors such as generator-to-turbine inertia ratio and damping. The study suggests that Kaplan-type units are more susceptible to fatigue due to lower inertia ratios and reduced damping.

A distributed PI+Reset Scheme for Discrete-Time Economic Dispatch of a Grid-connected BESS Network investigates a novel approach to reduce costs in battery energy storage system networks, utilizing discrete-time multi-agent systems with a proportional integral (PI) controller. The scheme uses a marginal cost consensus controller and a power mismatch estimation algorithm to optimize energy routing. It achieves faster convergence speeds and improved accuracy compared to traditional P controllers.

Data-driven algorithms are developed to optimize block replacement scheduling for maintaining N identical machines, aiming to minimize costs when lifetime distribution is unknown. The proposed algorithms achieve regret rates of O(K log T) and O((K-k*) log T), outperforming the Lai-Robbins lower bound in some cases. Numerical experiments confirm the effectiveness of these algorithms and reveal structural cost gaps between optimal block and age-dependent replacement strategies.