Co-optimizing Data Center Workloads for Grid Regulation Services

Yingrui Fan and Junbo Zhao have presented a new unified day-ahead co-optimization framework for scheduling geographically distributed data center workloads and committing regulation capacity (arXiv submission 2 Feb 2026).

• Main announcement: The paper proposes a unified day-ahead co-optimization framework that jointly decides workload distribution across geographically distributed DCs and regulation capacity commitments, using a space-time network model to capture workload migration costs, latency requirements, and heterogeneous resource limits; it also introduces chance constraints on instantaneous power flexibility and Value-at-Risk queue-state constraints to ensure sustainable regulation delivery.
• Background and validation: The authors evaluate the approach via case studies on a modified IEEE 68-bus system using real data center traces, reporting lower system operating costs, more viable regulation capacity, and better revenue-risk trade-offs versus independent scheduling and bidding strategies; the work is published on arXiv (submitted 2 Feb 2026) with PDF, HTML, TeX source and DOI links.