Harnessing Flexible Spatial and Temporal Data Center Workloads for Grid Regulation Services

Data centers (DCs) are increasingly recognized as flexible loads that can support grid frequency regulation. Yet, most existing methods treat workload scheduling and regulation capacity bidding separately, overlooking how queueing dynamics and spatial-temporal dispatch decisions affect the ability to sustain real-time regulation. As a result, the committed regulation may become infeasible or short-lived. To address this issue, we propose a unified day-ahead co-optimization framework that jointly decides workload distribution across geographically distributed DCs and regulation capacity commitments. We construct a space-time network model to capture workload migration costs, latency requirements, and heterogeneous resource limits. To ensure that the committed regulation remains deliverable, we introduce chance constraints on instantaneous power flexibility based on interactive load forecasts, and apply Value-at-Risk queue-state constraints to maintain sustainable response under cumulative regulation signals. Case studies on a modified IEEE 68-bus system using real data center traces show that the proposed framework lowers system operating costs, enables more viable regulation capacity, and achieves better revenue-risk trade-offs compared to strategies that optimize scheduling and regulation independently.

💡 Research Summary

The paper addresses the emerging role of data centers (DCs) as flexible loads capable of providing frequency regulation services to the power grid. While prior work has examined either DC workload scheduling or regulation capacity bidding in isolation, it has largely ignored the interplay between queue dynamics, latency constraints, and spatial‑temporal dispatch decisions that determine whether a DC can sustain the committed regulation signal in real time. To bridge this gap, the authors propose a unified day‑ahead co‑optimization framework that simultaneously decides (i) how workloads are distributed across a set of geographically dispersed DCs and (ii) how much regulation capacity each DC should commit.

Modeling framework
The core of the methodology is a space‑time network representation. Each DC l at each discrete time slot t is mapped to a virtual node p(l,t) = (l‑1)·T + t, yielding ψ = N·T total nodes. Workloads are aggregated as execution‑rate variables x_i,t,l ∈