pith. sign in

arxiv: 2605.18517 · v1 · pith:RQJUSNQXnew · submitted 2026-05-18 · 📡 eess.SY · cs.SY

Data Center Spatio-Temporal Load Flexibility in Security-Constrained Unit Commitment for Enhanced Grid Efficiency and Reliability

Haoxiang Wan , Xingpeng Li This is my paper

Pith reviewed 2026-05-20 09:13 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords data center flexibilitysecurity-constrained unit commitmentrenewable curtailmenttransmission congestiondemand responseMILP optimizationgrid reliabilityspatio-temporal scheduling
0
0 comments X

The pith

Integrating spatial and temporal flexibility from data centers into security-constrained unit commitment eliminates transmission violations and cuts renewable curtailment by up to 84 percent.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops three mixed-integer linear programming models within a security-constrained unit commitment framework to coordinate data center loads with grid scheduling. It demonstrates that enabling both spatial redistribution of workloads across sites and temporal shifting at each site, while keeping daily energy use constant, allows the system to resolve all base-case and post-contingency line violations at a 40 percent flexibility level. The same mechanism reduces renewable energy curtailment by as much as 84.4 percent at a 30 percent flexibility ratio compared with an inflexible baseline. A reader would care because data center demand is rising rapidly and already stresses transmission networks, yet this controllable demand offers a practical way to ease congestion and support more renewables without new infrastructure.

Core claim

The paper formulates a modular SCUC framework with three models: DC-S for instantaneous spatial redistribution of workloads across distributed data centers, DC-T for temporal shifting of deferrable load at each site while preserving daily energy balance, and DC-ST that activates both mechanisms to span the largest feasible region. Case studies on a modified IEEE 24-bus system show that the DC-ST model removes every base-case and post-contingency transmission violation once flexibility reaches 40 percent and reduces renewable curtailment by up to 84.4 percent at 30 percent flexibility relative to the inflexible case. Sensitivity results indicate that most benefits appear at moderate levels of

What carries the argument

The Data Center Spatio-Temporal (DC-ST) model, a MILP formulation that jointly optimizes spatial workload redistribution across sites and temporal load shifting within sites subject to daily energy balance and service constraints.

Load-bearing premise

Data center operators will permit spatial redistribution and temporal shifting of workloads while preserving daily energy balance and service quality under real operational constraints.

What would settle it

A field trial on an actual transmission system in which data centers do not shift or redistribute loads as assumed, after which the model-predicted elimination of violations and curtailment reductions either appear or fail to materialize.

Figures

Figures reproduced from arXiv: 2605.18517 by Haoxiang Wan, Xingpeng Li.

Figure 2
Figure 2. Figure 2: Number of lines with flow violations under contingency across [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 1
Figure 1. Figure 1: DC-ST load profiles at Bus 9 (left) and Bus 18 (right) under i [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Cost savings vs. Fixed DC ($M): DC-S (Spatial), DC-T (Temporal), [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: presents curtailment results at β = 0.30. The Fixed DC baseline curtails 1,781.9 MWh over 24 hours. DC-T reduces curtailment to 895.5 MWh (49.7%) by concentrating load during midday hours 9–16, when PV peaks and lines are otherwise overloaded. DC-S achieves 597.3 MWh (66.5%) by directing workload toward the site with greater local renew￾able availability, cutting inter-regional power transfer. DC-ST attain… view at source ↗
read the original abstract

Data center electricity consumption reached 4.4% of U.S. total in 2023 and is projected to grow to 6.7--12% by 2028, imposing increasing stress on transmission networks while representing a largely untapped source of controllable demand-side flexibility. This paper proposes a modular security-constrained unit commitment (SCUC) framework that coordinates flexible data center workloads with system-level scheduling to reduce renewable curtailment, alleviate congestion, and lower operating costs. Three mixed-integer linear programming (MILP) models are formulated: the Data Center Spatial model (DC-S), enabling instantaneous workload redistribution across geographically distributed sites; the Data Center Temporal model (DC-T), permitting each site to shift its deferrable load across time while preserving the daily energy balance; and the Data Center Spatio-Temporal model (DC-ST), jointly activating both mechanisms and spanning the largest feasible operating region. Case studies on a modified IEEE 24-bus reliability test system show that DC-ST eliminates all base-case and post-contingency transmission violations at a flexibility ratio of 40%, and reduces renewable curtailment by up to 84.4% at 30% relative to the inflexible baseline. Sensitivity analysis further reveals that moderate flexibility levels of 20%--30% already capture most of the achievable benefits, supporting practical deployment with limited operational burden on data center operators.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes a modular security-constrained unit commitment (SCUC) framework that integrates data center load flexibility via three MILP models: DC-S (spatial workload redistribution across sites), DC-T (temporal shifting at each site while preserving daily energy balance), and DC-ST (joint spatio-temporal activation). Case studies on a modified IEEE 24-bus reliability test system claim that DC-ST eliminates all base-case and post-contingency transmission violations at a 40% flexibility ratio and reduces renewable curtailment by up to 84.4% at 30% flexibility relative to an inflexible baseline, with sensitivity analysis indicating that 20-30% flexibility captures most benefits.

Significance. If the idealized flexibility assumptions hold under real SLAs, the framework could meaningfully improve grid efficiency and renewable integration by treating data centers as controllable demand, with the use of a public IEEE test system and standard MILP formulations aiding reproducibility. The modular structure (DC-S, DC-T, DC-ST) is a clear strength for isolating the value of each flexibility mechanism.

major comments (2)
  1. [Abstract and Model Formulation] Abstract and Model Formulation section: the central claims of zero violations at 40% flexibility and 84.4% curtailment reduction at 30% rest on the assumption that substantial workloads can be spatially redistributed or temporally shifted while exactly preserving daily energy balance and service quality, yet no explicit MILP constraints for latency bounds, data locality, or SLA penalties are provided; if these limits bind, the feasible region shrinks and the reported outcomes no longer hold.
  2. [Case Studies] Case Studies section (IEEE 24-bus results): the performance metrics are stated quantitatively but without accompanying equations for the flexibility ratio definition or the exact form of the added DC-ST constraints, preventing direct verification that the elimination of violations is not an artifact of the chosen test-system modifications.
minor comments (2)
  1. [Introduction] Notation for the three models (DC-S, DC-T, DC-ST) is introduced clearly in the abstract but could be reinforced with a summary table comparing their constraint sets and feasible regions.
  2. [Sensitivity Analysis] The sensitivity analysis paragraph would benefit from explicit statement of the range of flexibility ratios tested and the corresponding objective values or violation counts.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which help improve the clarity and transparency of our work. We address each major comment below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and Model Formulation] Abstract and Model Formulation section: the central claims of zero violations at 40% flexibility and 84.4% curtailment reduction at 30% rest on the assumption that substantial workloads can be spatially redistributed or temporally shifted while exactly preserving daily energy balance and service quality, yet no explicit MILP constraints for latency bounds, data locality, or SLA penalties are provided; if these limits bind, the feasible region shrinks and the reported outcomes no longer hold.

    Authors: We acknowledge that the current MILP models do not include explicit constraints on latency bounds, data locality, or SLA penalties. The flexibility ratio is instead used to limit the fraction of load that may be shifted or redistributed while enforcing exact daily energy balance at each site. This formulation is intended to quantify the maximum potential benefits under idealized but operationally plausible flexibility levels. In the revised manuscript we will add a new paragraph in the Model Formulation section that explicitly states these modeling assumptions, discusses their implications for service quality, and outlines how latency or SLA constraints could be incorporated as extensions. This will make clear that the reported results represent an upper-bound case. revision: yes

  2. Referee: [Case Studies] Case Studies section (IEEE 24-bus results): the performance metrics are stated quantitatively but without accompanying equations for the flexibility ratio definition or the exact form of the added DC-ST constraints, preventing direct verification that the elimination of violations is not an artifact of the chosen test-system modifications.

    Authors: We agree that the current text does not present the flexibility-ratio definition or the full DC-ST constraint set in the Case Studies section. In the revision we will insert the explicit mathematical definition of the flexibility ratio (maximum fraction of daily energy that may be spatially or temporally reallocated) together with the complete set of DC-ST constraints added to the base SCUC formulation. We will also expand the description of the IEEE 24-bus modifications, specifying the added data-center buses, their capacities, and the renewable-generation profiles. These additions will enable readers to reproduce the results and confirm that violation elimination arises from the flexibility mechanisms. revision: yes

Circularity Check

0 steps flagged

No circularity; standard MILP SCUC augmented with flexibility constraints and solved on public IEEE test system.

full rationale

The paper defines three MILP models (DC-S, DC-T, DC-ST) by extending conventional security-constrained unit commitment formulations with explicit spatial redistribution and temporal shifting constraints while enforcing daily energy balance. Results on the modified IEEE 24-bus system are generated by solving these optimization problems for given flexibility ratios (e.g., 40% and 30%), producing reported outcomes such as violation elimination and curtailment reduction. No step reduces by construction to a fitted parameter renamed as prediction, no self-citation chain is load-bearing for the central claims, and the derivation relies on standard external test data rather than self-referential inputs.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claims rest on the domain assumption that data center workloads can be shifted spatially and temporally without violating operational requirements, plus standard power-system constraints from the IEEE test system; no free parameters are fitted to data and no new entities are postulated.

free parameters (1)
  • flexibility ratio
    Scenario parameter representing the fraction of load available for shifting; tested at discrete values such as 20%, 30%, and 40%.
axioms (1)
  • domain assumption Workload redistribution across sites and time preserves daily energy balance and service quality
    Invoked in the definitions of DC-T and DC-ST models in the abstract.

pith-pipeline@v0.9.0 · 5782 in / 1409 out tokens · 51329 ms · 2026-05-20T09:13:06.856338+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

  1. [1]

    Energy and AI: Energy demand from AI,

    International Energy Agency, “Energy and AI: Energy demand from AI,” International Energy Agency, Paris, France, Tech. Rep., 2025, accessed: Jan. 15, 2026. [Online]. Available: https://www.iea.org/report s/energy-and-ai/energy-demand-from-ai

    work page 2025

  2. [2]

    Gartner says electricity demand for data centers to grow 16% in 2025 and double by 2030,

    Gartner, Inc., “Gartner says electricity demand for data centers to grow 16% in 2025 and double by 2030,” Press Release, Nov. 2025, accessed: Jan. 20, 2026. [Online]. Available: https://www.gartner.com/en/newsr oom/press-releases/2025-11-17-gartner-says-electricity-demand-for-d ata-centers-to-grow-16-percent-in-2025-and-double-by-2030

    work page 2025

  3. [3]

    2024 united states data center energy usage report,

    A. Shehabi, S. J. Smith, A. Hubbard, A. Newkirk, N. Lei, M. A. B. Siddik, B. Holecek, J. Koomey, E. Masanet, and D. Sartor, “2024 united states data center energy usage report,” Lawrence Berkeley National Laboratory, Berkeley, CA, USA, Tech. Rep. LBNL-2001637, Dec. 2024. [Online]. Available: https://eta-publications.lbl.gov/sites/default/files/2 024-12/lb...

    work page 2024

  4. [4]

    How data centers are reshaping PJM’s energy market,

    Powwr, “How data centers are reshaping PJM’s energy market,” Blog Article, 2025, accessed: Feb. 5, 2026. [Online]. Available: https://www. powwr.com/blog/how-data-centers-are-reshaping-pjms-energy-market

    work page 2025

  5. [5]

    Grid congestion is posing challenges for energy security and transitions,

    International Energy Agency, “Grid congestion is posing challenges for energy security and transitions,” IEA Commentary, 2024, accessed: Feb. 10, 2026. [Online]. Available: https://www.iea.org/commentaries/g rid-congestion-is-posing-challenges-for-energy-security-and-transitions

    work page 2024

  6. [6]

    Carbon impact of intra-regional transmis- sion congestion,

    S. E. Sofia and Y . Dvorkin, “Carbon impact of intra-regional transmis- sion congestion,”Cell Rep. Sustainability, vol. 2, no. 7, p. 100577, 2025

    work page 2025

  7. [7]

    Energy campus: Why data centers are embracing this model,

    Hanwha Data Centers, “Energy campus: Why data centers are embracing this model,” Blog Article, 2025, accessed: Feb. 3, 2026. [Online]. Available: https://www.hanwhadatacenters.com/blog/energy-c ampus-why-data-centers-are-embracing-this-model/

    work page 2025

  8. [8]

    Data centres as a source of flexibility for power systems,

    M. T. Takci, M. Qadrdan, J. Summers, and J. Gustafsson, “Data centres as a source of flexibility for power systems,”Energy Rep., vol. 13, pp. 3661–3671, 2025

    work page 2025

  9. [9]

    Grid operational benefit analysis of data center spatial flexibility: Congestion relief, renewable energy curtailment reduction, and cost saving,

    H. Wan, L. Fang, and X. Li, “Grid operational benefit analysis of data center spatial flexibility: Congestion relief, renewable energy curtailment reduction, and cost saving,” inProc. IEEE Power Energy Soc. Gen. Meeting (PESGM), Montr ´eal, QC, Canada, Jul. 2026, to appear

    work page 2026

  10. [10]

    DOE data center load flexibility workshop summary,

    T. W. Kirchstetter, S. J. Smith, A. Shehabi, and D. A. Sartor, “DOE data center load flexibility workshop summary,” Lawrence Berkeley National Laboratory, Berkeley, CA, USA, Workshop Summary Rep., Mar. 2025. [Online]. Available: https://eta-publications.lbl.gov/sites/de fault/files/2025-03/final doe data center load flexibility workshop s ummary.v0307.pdf

    work page 2025

  11. [11]

    Datacenter demand response for carbon mitigation: From concept to practicality,

    J. Xing and B. C. Lee, “Datacenter demand response for carbon mitigation: From concept to practicality,” inProc. IEEE 15th Int. Green Sustain. Comput. Conf. (IGSC), 2024, pp. 142–144

    work page 2024

  12. [12]

    Opportunities and challenges for data center demand response,

    A. Wierman, Z. Liu, I. Liu, and H. Mohsenian-Rad, “Opportunities and challenges for data center demand response,” inProc. Int. Green Comput. Conf. (IGCC), Dallas, TX, USA, Nov. 2014, pp. 1–10

    work page 2014

  13. [13]

    Greening geographical load balancing,

    Z. Liu, M. Lin, A. Wierman, S. H. Low, and L. L. H. Andrew, “Greening geographical load balancing,”IEEE/ACM Trans. Netw., vol. 23, no. 2, pp. 657–671, 2015

    work page 2015

  14. [14]

    Using geographic load shifting to reduce carbon emissions,

    J. Lindberg, B. C. Lesieutre, and L. A. Roald, “Using geographic load shifting to reduce carbon emissions,”Electr. Power Syst. Res., vol. 212, p. 108586, 2022

    work page 2022

  15. [15]

    Spatio-temporal deep learning-assisted re- duced security-constrained unit commitment,

    A. V . Ramesh and X. Li, “Spatio-temporal deep learning-assisted re- duced security-constrained unit commitment,”IEEE Trans. Power Syst., vol. 39, no. 2, pp. 4735–4746, 2024

    work page 2024

  16. [16]

    An accelerated-decomposition approach for security-constrained unit commitment with corrective net- work reconfiguration,

    A. V . Ramesh, X. Li, and K. W. Hedman, “An accelerated-decomposition approach for security-constrained unit commitment with corrective net- work reconfiguration,”IEEE Trans. Power Syst., vol. 37, no. 2, pp. 887– 900, 2022

    work page 2022

  17. [17]

    Security-constrained unit commitment considering locational frequency stability in low-inertia power grids,

    M. Tuo and X. Li, “Security-constrained unit commitment considering locational frequency stability in low-inertia power grids,”IEEE Trans. Power Syst., vol. 38, no. 5, pp. 4134–4147, 2023

    work page 2023

  18. [18]

    National solar radiation database (NSRDB) data viewer,

    National Renewable Energy Laboratory, “National solar radiation database (NSRDB) data viewer,” Online database, accessed: Feb. 13,

    work page

  19. [19]

    Available: https://nsrdb.nrel.gov/data-viewer

    [Online]. Available: https://nsrdb.nrel.gov/data-viewer

    work page