Grid Operational Benefit Analysis of Data Center Spatial Flexibility: Congestion Relief, Renewable Energy Curtailment Reduction, and Cost Saving

Haoxiang Wan Linhan Fang Xingpeng Li Department of Electrical and Department of Electrical and Department of Electrical and Computer Engineering Computer Engineering Computer Engineering University of Houston University of Houston University of Houston Houston, TX, USA Houston, TX, USA Houston, TX, · 2025

Read Grid Operational Benefit Analysis of Data Center Spatial Flexibility: Congestion Relief, Renewable Energy Curtailment Reduction, and Cost Saving by Haoxiang Wan Linhan Fang Xingpeng Li Department of Electrical and Department of Electrical and Department of Electrical and Computer Engineering Computer Engineering Computer Engineering University of Houston University of Houston University of Houston Houston, TX, USA Houston, TX, USA Houston, TX, (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-08759.txt.

Introduction

Haoxiang Wan, Linhan Fang, and Xingpeng Li ask whether data centers can help the power grid simply by moving their computing around. Data center demand is set to double globally by 2026 under AI growth. Big fixed loads congest transmission lines, while solar and wind farms lose output whenever the grid cannot absorb it.

An optimal power flow model coordinates generation, security reserves, and movable data center load together. On a modified IEEE seventy-three bus test system, rigid placement causes severe violations, with some lines overloaded by about thirty percent. Allowing workloads to migrate restores feasibility in the studied scenarios. Solar curtailment falls by up to sixty-one percent when load shifts toward sunny regions. The authors argue spatial flexibility could defer costly transmission upgrades.

Catalog record

CREATORHaoxiang Wan Linhan Fang Xingpeng Li Department of Electrical and Department of Electrical and Department of Electrical and Computer Engineering Computer Engineering Computer Engineering University of Houston University of Houston University of Houston Houston, TX, USA Houston, TX, USA Houston, TX,
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-08759

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Grid Operational Benefit Analysis of Data Center Spatial Flexibility: Congestion Relief, Renewable Energy Curtailment Reduction, and Cost Saving Haoxiang Wan Linhan Fang Xingpeng Li Department of Electrical and Department of Electrical and Department of Electrical and Computer Engineering Computer Engineering Computer Engineering University of Houston University of Houston University of Houston Houston, TX, USA Houston, TX, USA Houston, TX, USA [email protected] [email protected] [email protected] Abstract Data centers are facilities housing computing infrastructure for processing and storing digital information. The rapid expansion of artificial intelligence is driving unprecedented growth in data center capacity, with global electricity demand from data centers projected to double by 2026. This growth creates substantial challenges for power transmission networks, as large concentrated loads can cause congestion and threaten grid reliability. Meanwhile, the intermittent nature of solar and wind generation requires flexible resources to maintain grid reliabitliy and minimize curtailment. This paper assesses whether data center spatial flexibility—the ability to migrate computational workloads geographically—can serve as a grid resource to address these challenges. An optimal power flow model is developed to co-optimize generation dispatch, security reserves, and flexible data center loads. Case studies on a modified IEEE 73-bus system show that inflexible data center placement can lead to severe transmission violations, with line overloads reaching 30.1%. Enabling spatial flexibility mitigates these violations in the studied scenarios and restores system feasibility. This flexibility also reduces solar curtailment by up to 61.0% by strategically reallocating load to solar-rich areas. The results suggest that spatial flexibility offers a viable approach to defer transmission upgrades and enhance renewable utilization. I Introduction The proliferation of artificial intelligence and big data analytics has catalyzed substantial growth in data center (DC) infrastructure, both in number and scale [ 1 ] . This expansion has contributed to rising energy consumption, with global data centers currently accounting for approximately 3% of total electricity supply [ 2 ] . The integration of these large, geographically concentrated loads into power transmission networks presents challenges for operational stability and capacity management. Concurrently, efforts to mitigate carbon emissions and diversify energy portfolios have accelerated the grid integration of renewable energy sources [ 3 ] , introducing additional operational complexities. The increasing deployment of variable renewable energy sources has introduced new dimensions to power system operations, particularly in transmission congestion management and economic dispatch optimization [ 4 , 5 , 6 ] . Recent advances in machine learning–based optimal power flow (OPF) formulations have further expanded this field by enhancing computational efficiency and incorporating system-level constraints such as frequency stability and network congestion [ 7 , 8 ] . According to Lawrence Berkeley National Laboratory’s 2024 report, data centers consumed approximately 4.4% of total U.S. electricity in 2023, with projections indicating growth to 6.7–12% by 2028 [ 9 ] . This rapid expansion, driven by AI applications, manufacturing growth, and electrification initiatives, introduces geographically concentrated loads that may intensify transmission stress and contribute to locational marginal price (LMP) volatility [ 10 , 11 ] . Addressing these challenges may require approaches that extend beyond conventional transmission expansion, which often involves substantial capital expenditures and extended implementation timelines [ 12 ] . Traditional congestion mitigation relies primarily on physical infrastructure investments such as transmission upgrades and energy storage deployment, each involving considerable capital costs and multi-year development cycles [ 13 ] . Demand-side management offers a complementary approach by leveraging load flexibility to potentially defer or reduce infrastructure requirements [ 14 , 15 ] . However, existing frameworks have predominantly focused on temporal flexibility, with limited investigation of spatial flexibility—the strategic redistribution of computational workloads across geographically dispersed facilities. Data centers may present opportunities in this regard: their distributed architectures and delay-tolerant workloads can enable workload migration across locations with relatively minimal service impact, potentially functioning as spatially controllable demand resources [ 16 ] . Recent literature has examined data center participation in ancillary services and demand response programs [ 14 , 15 , 16 ] , generally treating data centers as localized flexible loads. However, the potential for these facilities to serve as network-aware assets capable of influencing transmission congestion through spatial workload reallocation remains relatively unexplored. While Fridgen et al. [ 17 ] demonstrated the economic viability of spatial load migration for balancing power provision across international markets, spatial flexibility not only mitigates localized congestion but can also reshape power flow distributions at the system level, exerting a more direct influence on renewable energy utilization and price stability. The concept of demand-side resources providing transmission-like services, where spatial load redistribution can substitute for or complement physical network capacity, has received limited systematic investigation. Existing studies have not sufficiently examined the technoeconomic role of data center spatial flexibility in mitigating renewable curtailment, alleviating transmission congestion, and reducing locational marginal price variability within integrated optimal power flow frameworks. This paper presents an optimal power flow framework that models spatially flexible data centers and evaluates their potential role in alleviating transmission congestion and renewable curtailment. Using a modified IEEE 73-bus test system with distributed data center capacity and high solar penetration, the analysis examines four operational scenarios to address several interrelated questions: how inflexible deployment influences transmission feasibility under varying load conditions, whether spatial flexibility can restore system operability without physical network expansion, what transfer capability thresholds are required for congestion mitigation, and to what extent demand-side flexibility may reduce solar curtailment in transmission-constrained systems. The proposed framework identifies critical feasibility thresholds, economic saturation points, and curtailment reduction potential through systematic case studies. The findings suggest that spatial reallocation of data center loads can offer operational benefits; however, practical implementation would necessitate consideration of market design and coordination challenges beyond the scope of this study. II Formulation of the Optimization Model This section details the integrated optimization model that co-optimizes conventional generation, security reserves, renewable resources, and flexible demand within an OPF framework. II-A Objective Function The objective minimizes total operational cost: \min\sum_{g\in\mathcal{G}}c_{g}P_{g}, (1) where P_{g} and c_{g} denote the active power dispatch and marginal cost for generator g\in\mathcal{G} . II-B Network and Generation Constraints Conventional generation is bounded by capacity limits: P_{g}^{\min}\leq P_{g}\leq P_{g}^{\max},\quad\forall g\in\mathcal{G}, (2) where P_{g} is the real power output of generator g , P_{g}^{\min} and P_{g}^{\max} are minimum and maximum capacity limits, and \mathcal{G} is the set of all generators. Under Direct Current power flow approximation, active power flow P_{k} on branch k\in\mathcal{K} is: P_{k}=\frac{\theta_{f(k)}-\theta_{t(k)}}{x_{k}},\quad\forall k\in\mathcal{K}, (3) where \mathcal{K} is the set of transmission branches, \theta_{f(k)} and \theta_{t(k)} are voltage angles at the from-bus and to-bus terminals, x_{k} is the series reactance, and f(k),t(k) map branches to buses in set \mathcal{N} . Branch flows are constrained by thermal ratings: -\text{RateA}_{k}\leq P_{k}\leq\text{RateA}_{k},\quad\forall k\in\mathcal{K}, (4) where \text{RateA}_{k} is the thermal capacity limit of branch k . Nodal power balance at bus n\in\mathcal{N} is: \sum_{g\in\mathcal{G}(n)}P_{g}+P^{\text{solar}}_{n}+\sum_{k\in\mathcal{K}^{-}(n)}P_{k}-\sum_{k\in\mathcal{K}^{+}(n)}P_{k}\\ =d_{n}+L_{n}^{\text{optimized}},\quad\forall n\in\mathcal{N}, (5) ·where \mathcal{G}(n) is the set of generators at bus n , P^{\text{solar}}_{n} is dispatched solar generation, \mathcal{K}^{-}(n) and \mathcal{K}^{+}(n) are branches flowing into and out of bus n , d_{n} represents the traditional inelastic electrical load, and L_{n}^{\text{optimized}} denotes the optimized data center computational workload at bus n after spatial redistribution. II-C Security and Reserve Constraints For N\!-\!1 reliability, spinning reserve r_{g} for generator g satisfies: \displaystyle P_{g}+r_{g}\leq P_{g}^{\max}, \displaystyle\quad\forall g\in\mathcal{G}, (6) \displaystyle r_{g}\leq

Read the complete text