Michael Murillo | Searcharxiv

arXiv2026

Modal Analysis of Spatial Load Correlation in AI Data Center-Dominated Power Systems

Hyperscale AI data centers induce spatially and temporally correlated load fluctuations that violate classical independence assumptions and are not captured by time-averaged spectral methods. These correlations are episodic and non-stationary, so they demand analysis that resolves transient structure. This paper applies Dynamic Mode Decomposition (DMD) to the temporal evolution of pairwise inter-bus correlation coefficients and forms a low-dimensional state representation that enables modal analysis without a stationarity assumption. The recovered modes distinguish sustained coherence, decaying transients, and intensifying events, and their oscillation timescales map to underlying physical coupling mechanisms. The method is evaluated on an IEEE 39-bus Real-Time Digital Simulator (RTDS) testbed with three converter-interfaced AI data center loads driven by synthetic workload profiles. A global analysis attributes the dominant correlation energy to a slow thermal band, and a sliding-window analysis identifies brief intensification events in a small fraction of windows that align with stochastic workload coincidences. Cross-validation with RTDS voltage coherence confirms elevated coupling during these intervals. The proposed modal growth indicator provides an early-warning signal of correlation intensification, with a lead of of about 4~s before pairwise coherence reaches its peak.

eess.SY↗

arXiv2024

Conservative dielectric functions and electrical conductivities from the multicomponent Bhatnagar-Gross-Krook equation

A considerable number of semi-empirical and first-principles models have been created to describe the dynamic response of a collisionally damped charged-particle system. However, known challenges persist for established dynamic structure factors (DSF), dielectric functions, and conductivities. For instance, the semi-empirical Drude-Smith conductivity [N.M. Smith, Phys. Rev. B 64, 155106 (2001)] lacks interpretability, and the first-principles Mermin dielectric function [N.D. Mermin, Phys. Rev. B, 1, 2362 (1970)] does not satisfy the frequency sum rule [G.S. Atwal and N.W. Ashcroft, Phys. Rev. B 65, 115109 (2002)]. In this work, starting from the multicomponent Bhatnagar-Gross-Krook (BGK) kinetic equation, we produce a multi-species susceptibility that conserves number and momentum, which we refer to as the ``completed Mermin'' susceptibility, and we explore its properties and uses. We show that the completed Mermin susceptibility satisfies the frequency sum (f-sum) rule. We compute the associated DSF and find that momentum conservation qualitatively impacts the DSF's shape for a carbon-contaminated deuterium and tritium plasma under NIF hot-spot conditions. In the appendices, we provide numerical implementations of the completed Mermin susceptibility, for the reader's convenience. Further, we produce a new non-Drude conductivity model, by taking the single-species limit and introducing free parameters in the terms that enforce number and momentum conservation. To illustrate how number and momentum conservation impact the dynamical conductivity shape, we apply our conductivity model to dynamical gold conductivity measurements [Z. Chen, et al., Nature communications, 12.1, 1638, (2021)]. Finally, comparing our model to the Drude-Smith conductivity model, we conclude that Smith's phenomenological parameter violates local number conservation.

physics.plasm-ph↗