{"paper":{"title":"Grid Integration of Gigawatt-Scale AI Data Centers under Connect-and-Manage","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Gigawatt-scale AI data centers can connect to transmission grids without upgrades using a hierarchical coordination protocol that slashes curtailment while maintaining training workloads.","cross_cats":["cs.SY"],"primary_cat":"eess.SY","authors_text":"Qianwen Xu, Xin Lu","submitted_at":"2026-05-13T20:48:47Z","abstract_excerpt":"Emerging connect-and-manage interconnection practices allow gigawatt-scale artificial intelligence data centers (AIDCs) to connect to the transmission network without prior network upgrades, at the cost of real-time curtailment during grid stress. This paper formalizes the resulting AIDC-transmission system operator (TSO) coordination as a sequential request-acceptance protocol with an explicit curtailment variable and a strict information boundary between the two parties. Physical models are developed on both sides of the point of common coupling: the AIDC is decomposed into frontier training"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Case studies with a gigawatt-scale AIDC on the IEEE 39-bus system with Australian market data show that the framework reduces curtailment from 9.1% to 2.8% while preserving 98.1% frontier training workload, that batch training acts as the primary grid-elastic resource with the largest throughput swing during peak demand, and that the on-site battery provides curtailment buffering through active discharge and charge deferral.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The TSO's acceptance mapping is completely opaque to the AIDC and can be treated as a robust black-box mechanism whose worst-case behavior is known in advance; if real TSO decisions depend on private information or non-robust criteria not captured by the model, the hierarchical architecture loses its performance guarantees.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A hierarchical request-acceptance protocol with learning-based planning and robust TSO evaluation reduces curtailment for GW-scale AI data centers from 9.1% to 2.8% while preserving 98.1% of frontier training workload.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Gigawatt-scale AI data centers can connect to transmission grids without upgrades using a hierarchical coordination protocol that slashes curtailment while maintaining training workloads.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4f09a954e507b82fa6ceaea4d5fc9387a5502b1968534cbe8bccbc6c75a4f0e0"},"source":{"id":"2605.14109","kind":"arxiv","version":1},"verdict":{"id":"849b79e1-8a93-4c50-8897-ec2519a46c14","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:01:18.062233Z","strongest_claim":"Case studies with a gigawatt-scale AIDC on the IEEE 39-bus system with Australian market data show that the framework reduces curtailment from 9.1% to 2.8% while preserving 98.1% frontier training workload, that batch training acts as the primary grid-elastic resource with the largest throughput swing during peak demand, and that the on-site battery provides curtailment buffering through active discharge and charge deferral.","one_line_summary":"A hierarchical request-acceptance protocol with learning-based planning and robust TSO evaluation reduces curtailment for GW-scale AI data centers from 9.1% to 2.8% while preserving 98.1% of frontier training workload.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The TSO's acceptance mapping is completely opaque to the AIDC and can be treated as a robust black-box mechanism whose worst-case behavior is known in advance; if real TSO decisions depend on private information or non-robust criteria not captured by the model, the hierarchical architecture loses its performance guarantees.","pith_extraction_headline":"Gigawatt-scale AI data centers can connect to transmission grids without upgrades using a hierarchical coordination protocol that slashes curtailment while maintaining training workloads."},"references":{"count":27,"sample":[{"doi":"","year":2025,"title":"NVIDIA Launches Omniverse DSX Blueprint, Enabling Global AI Infrastructure Ecosystem to Build Gigawatt -Scale AI Factories,","work_id":"89e6ebbd-2ced-4a04-ae4e-6082ecbef965","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"The cost of compute: A $7 trillion race to scale data centers,","work_id":"83958039-1209-46b8-aa3f-2590193dae68","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"I. E. Agency, “Energy and AI,” World Energy Outlook Special Report, 2025","work_id":"ad624471-a350-4528-811a-adb2dfc11aab","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"Structural alignment for energy –computation co-design,","work_id":"03678f49-4690-43d1-bf5c-6543cd99198f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"PJM, stakeholders begin work on Board's plan to reliably integrate large loads,","work_id":"a89d0f68-8ced-4df1-b5dd-a32b9988c14d","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":27,"snapshot_sha256":"eddab22c6182f0d00d6c2605938aa95a6bf310949336147b7918c0ecea4c55bf","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"fb41be7f341bb32b7734722be51fcbdfa0e07ed594effc8e18bc9535810ce7f4"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}