--- 格式版本: 2 标题: "DOE Retreat on Transmission Corridors Tests Grid Planning" 原文链接: "https://www.datacenterknowledge.com/energy-power-supply/doe-retreat-on-transmission-corridors-tests-the-case-for-building-ahead" 发布日期: "2026-08-28" 发布时间校准状态: "found" 发布时间需复核: "否" 发布时间来源: "rule:configured_publication_date_rule" 发布时间证据: "datacenterknowledge-jsonld-date-published html:original: \"datePublished\":\"2026-08-28T13:12:06.000Z\"" 发布时间校准原因: "信源发布日期识别规则直接确认发布时间" 发布时间校准置信度: "high" 发布时间候选数量: 2 发布时间严格候选数量: 2 发布时间原页读取状态: "source template page reused from URL open" 发布时间未找到原因: "" 发布时间校准时间: "2026-08-30T09:21:23+08:00" 发布时间仲裁状态: "skipped" 发布时间仲裁尝试次数: 0 发布时间仲裁耗时毫秒: 0 发现时间: "2026-08-30T09:20:43+08:00" 入库时间: "2026-08-30T01:21:23.456Z" 来源平台: "Data Center Knowledge 搜索" 搜索渠道: "source_template" 搜索词: "https://www.datacenterknowledge.com/search?q=AI%20Rack" 匹配关键词: - "AI Rack" - "AI" - "delivery" 相关厂家: [] 相关专家: [] 内容类型: "网页" 抓取工具: "Free Fetch + Defuddle" 清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取" 原始附件: [] AI优质: "否" AI打分: 53 AI分档: "非优质" AI质检状态: "不通过" AI打分理由: "正文主线是DOE输电走廊决策及数据中心大负荷的电网规划，并非超节点、AI Rack或机架级关键部件。Data Center Knowledge属专业媒体，引用DOE、ESIG、ERCOT资料并包含专家采访，来源可追溯。固定知识库未见该输电决策的重复记录；新增事实包括DOE否决三个NIETC、ERCOT约474.7GW大负荷申请中约420.8GW来自数据中心，以及按证据门槛分阶段投入输电资本的建议。但内容没有新机架拓扑、供电产品、客户部署或可复用的机架级工程机制，商业信号主要是未承诺的并网申请，命中“主体为通用能源与电网规划、超节点相关性弱”的强否决项。" AI质检模型: "gpt-5.6-sol" AI质检时间: "2026-08-30T09:21:37+08:00" AI主题相关性: 6 AI来源权威性: 11 AI新颖性: 14 AI技术细节: 8 AI商业部署信号: 5 AI完整性: 9 AI评分提示词版本: "v17-精简生产版" AI评分提示词SHA256: "48fb9777f386026761b4873eaff30807694fb11e9b352d7c69bf2dfde750cc7d" AI评分知识库版本: "knowledge_base_v1-20260819+runtime.53" AI评分知识库SHA256: "4fcd6d53a16f56063efcfb4a47c89832a62c826f84d2abb3ce64c62a46d9fa61" AI评分知识库检索词: "[\"AI Rack\",\"RAS\",\"NPU\",\"DOE\",\"US\",\"NIETCs\",\"CREZ\",\"NIETC\",\"FERC\",\"CREZ-enabled\",\"ESIG\",\"REAP\"]" AI评分知识库命中: "[{\"id\":\"historical-jun-042\",\"title\":\"AI Rack/机柜级超节点架构，详细列出Vera Rubin/GB200 NVL72规格、NVLink 6互连、HBM3e、液冷与供电设计，具备明确的产品化与\",\"sourceType\":\"curated_item\",\"time\":\"2026-06\",\"matchedTerms\":[\"AI Rack\"],\"rank\":-7.435259563490922},{\"id\":\"july-correct-0088\",\"title\":\"StrataCL: Fabric-Native Communication Library for Production Supernodes - 智源社区论文\",\"sourceType\":\"labeled_article\",\"time\":\"2026-07\",\"matchedTerms\":[\"NPU\",\"US\",\"ESIG\"],\"rank\":-6.164150341681708},{\"id\":\"runtime-5581f72bc66fda8e6c6bf711\",\"title\":\"Advancing Standards-Based AI Fabric RAS Through COSMOS Integration\",\"sourceType\":\"ai_excellent_article\",\"time\":\"2026-08-10\",\"matchedTerms\":[\"RAS\",\"US\",\"ESIG\"],\"rank\":-5.948482066000196},{\"id\":\"runtime-b86df24f43e82ced83868dce\",\"title\":\"AWS, Anthropic Complete Project Rainier AI Cluster\",\"sourceType\":\"ai_excellent_article\",\"time\":\"2026-08-17\",\"matchedTerms\":[\"RAS\",\"DOE\",\"US\",\"ESIG\"],\"rank\":-5.8044159284953585},{\"id\":\"historical-jun-025\",\"title\":\"英伟达、谷歌与国产超节点的三种网络选择\",\"sourceType\":\"curated_item\",\"time\":\"2026-06\",\"matchedTerms\":[\"NPU\",\"US\"],\"rank\":-5.797545736487382}]" AI摘要: "美国能源部（DOE）拒绝指定三条拟议的国家利益输电走廊，而同一时期其2026年输电需求研究草案显示，数据中心等负荷增长急需新增输电能力。" AI摘要模型: "ali-deepseek-v4-flash" AI摘要时间: "2026-08-30T18:09:25.835Z" 采集批次: "2026年8月30日8点42分04秒" 采集批次ID: "20260830-084204-060" 去重键: "https://www.datacenterknowledge.com/energy-power-supply/doe-retreat-on-transmission-corridors-tests-the-case-for-building-ahead" --- ## DOE Retreat on Transmission Corridors Tests the Case for Building Ahead DOE declined three proposed transmission corridors as its own study cites data center load growth, raising questions about how far ahead utilities should build. Getty Images Data center developers can seek gigawatts of power years before a transmission line can be permitted and built. That mismatch is putting new pressure on grid planners just as the US Department of Energy (DOE) [has decided](https://www.energy.gov/oe/national-interest-electric-transmission-corridor-designation-process) not to designate three proposed National Interest Electric Transmission Corridors (NIETCs) that had advanced to Phase 3 of federal review. DOE’s decision came after its draft [2026 National Transmission Needs Study](https://www.energy.gov/oe/articles/does-office-electricity-publishes-2026-draft-national-transmission-needs-study) identified accelerating electricity demand from data centers, domestic manufacturing, large industrial loads, and electrification as major drivers of future transmission needs. The draft finds a pressing need for additional transmission capacity to accommodate that growth. That leaves utilities and regulators with a familiar problem: build transmission early enough to support new demand, or risk making the grid the bottleneck after projects are committed. Building too far ahead, however, may leave utilities and customers paying for infrastructure that serves a load that never materializes. Texas offers an example of proactive planning that delivered results at scale. The state’s Competitive Renewable Energy Zones (CREZ) program was designed to move electricity from resource-rich West Texas to major load centers, with transmission planned ahead of the full generation buildout. Nearly two decades later, data center developers are gravitating toward parts of the same grid because of the transmission and power infrastructure that CREZ helped create. ## DOE’s NIETC Decisions: What It Means Former DOE Secretary Jennifer Granholm questioned the NIETC decision. In a [LinkedIn post](https://www.linkedin.com/feed/update/urn:li:activity:7498945116825374720/), Granholm pointed to the timing of the decisions, noting that DOE’s own transmission study identified a “pressing need” for new transmission to accommodate load growth and maintain reliability. “You can’t declare an energy emergency, demand that America pump out more electricity, and then purposefully make it harder to deliver that power where it’s needed,” Granholm wrote. The decision not to designate the three corridors does not necessarily conflict with the broader finding that the country needs more transmission, said Neil Osnato, founder of Persistence Analytics Group. “The Needs Study identifies a national transmission problem,” Osnato said. “The NIETC process asks a different question: whether specific geographic corridors satisfy the statutory and evidentiary basis for federal designation.” DOE concluded there was not a sufficient basis to designate any of the three potential NIETCs that advanced to Phase 3 – the Tribal Energy Access Corridor, Southwestern Grid Connector Corridor, and Lake Erie-Canada Corridor – citing the 2023 National Transmission Needs Study, public and governmental input, and other information gathered during the designation process. A NIETC is a geographic designation, not approval for a specific transmission project. Designation can unlock certain federal transmission tools and, in limited circumstances, trigger [Federal Energy Regulatory Commission](https://www.datacenterknowledge.com/build-design/ferc-targets-grid-rules-for-data-centers-and-large-loads) (FERC) authority to issue permits for eligible transmission facilities within the corridor. Osnato said the episode highlights a broader problem: the US has tools to identify transmission needs, but fewer to turn those needs into timely, coordinated interstate projects. Compounding that, transmission planning and large-load development operate on very different timelines. ## Lessons from Texas CREZ Planning As Data Center Knowledge has [previously reported](https://www.datacenterknowledge.com/build-design/texas-may-have-accidentally-built-the-perfect-grid-for-ai), Texas built CREZ to connect expected wind development in West Texas to population centers, planning the transmission ahead of the generation buildout. Nearly two decades later, major data center projects are locating in areas where CREZ-enabled infrastructure provides access to substantial power. [Galaxy Digital’s Helios campus](https://www.datacenterknowledge.com/data-center-construction/galaxy-bets-its-data-center-edge-on-power-not-bragawatts-) in Dickens County, for example, is being developed as a multi-gigawatt data center platform in an area with significant transmission capacity not built specifically for data centers. A [July 2026 report](https://www.esig.energy/reports-briefs/transmission-planning-large-loads/) from the Energy Systems Integration Group (ESIG) Large Loads Task Force, prepared by Brattle principal Johannes Pfeifenberger, Warren Lasher, and ESIG’s James Okullo, cites CREZ as an example of planning transmission around expected development rather than waiting for every project to materialize. The report recommends moving beyond project-by-project upgrades toward proactive, scenario-based, and multi-value planning. It also points to development-zone approaches such as Texas CREZ and [Illinois Renewable Energy Access Plan](https://www.illinois.gov/news/release.html?releaseid=30099) (REAP). Osnato noted that while the CREZ approach can inform data center planning, the evidence supporting the two types of development differs. “CREZ showed the value of identifying development zones, planning transmission at scale, and avoiding a sequence of one-off upgrades that become slower and more expensive over time,” Osnato said. “That logic absolutely applies to clustered data center growth.” But whereas CREZ was built around geographically concentrated generation development, data center planning often relies on future load that remains conditional on customer contracts, power delivery, financing, permitting, site readiness, and technology decisions. “I would borrow the CREZ concept of proactive zones and scale planning,” he said, “but pair it with a much stronger requirement to prove which loads have actually earned the right to drive that investment.” ## Build Ahead, Not Ahead of Evidence Transmission planning, permitting, and construction can take five to 10 years or longer, while large-load projects can seek interconnection within months or a few years, according to the ESIG report. Waiting until every customer is fully energized risks making transmission the bottleneck, but building billions of dollars in infrastructure around every announced gigawatt creates stranded capital and cost allocation risks. “In some places, yes,” Osnato said when asked whether today’s data center pace justifies building transmission ahead of firm commitments. “But ‘ahead of firm commitments’ should not mean ‘ahead of evidence.’” Osnato favors a staged approach: develop broad scenarios; preserve transmission corridors and development options early; then release increasingly irreversible capital as the underlying load passes defined evidence gates. For example, a utility might secure rights-of-way or design a scalable transmission solution before every megawatt is contracted, while deferring full capital investment until the evidence of durable load is stronger. He suggested separating the process into three questions: What is represented? What is independently supported today? What change would invalidate that conclusion? ## ERCOT’s Queue Tests the Model The ESIG report underscores how today’s large loads differ from historical demand growth: individual facilities can reach gigawatt scale; projects can cluster geographically; and development can move faster than major transmission projects. That creates a persistent mismatch between the time required to build transmission and the speed at which large loads can seek interconnection. ERCOT’s large-load pipeline shows the scale of the problem. As of June 2026, ERCOT [was tracking](https://www.datacenterknowledge.com/www.ercot.com/files/docs/2026/07/29/ERCOT-Senate-July-29-Panel-1-Assessing-The-Grid.pdf) about 474.7 GW of large-load interconnection requests, including roughly 420.8 GW – about 90.2% – associated with data centers. These requests are not equivalent to committed demand and span a range of development stages. The size of the queue has prompted Texas officials to scrutinize which projects are real enough to shape grid planning. On June 10, [Gov. Greg Abbott directed](https://www.datacenterknowledge.com/regulations/texas-pushes-ai-data-centers-to-pay-their-own-grid-costs) the Public Utility Commission of Texas (PUCT) and ERCOT to review data center projects and address development risks and costs that could otherwise fall on Texas customers. In August, Abbott [ordered a comprehensive audit](https://www.datacenterknowledge.com/energy-power-supply/texas-orders-statewide-audit-of-ai-data-center-projects-in-ercot-queue) of data centers advancing through ERCOT’s interconnection process. For transmission planners, the central question is how much of that pipeline should influence today’s investment decisions. The ESIG report recommends moving beyond isolated, project-by-project upgrades toward proactive, scenario-based, and multi-value planning that can scale as evidence of durable load accrues. ## Plan Early, Commit Capital in Stages Utilities may need to start planning years before a data center signs a final power contract. That raises the stakes for the evidence behind load forecasts and for how much capital is committed at each stage. Development zones, scenario-based planning, and expandable transmission designs can give planners more options as demand takes shape. Texas shows how early transmission can enable future development in ways planners might not fully anticipate. For Osnato, the lesson is not to wait for every megawatt to be certain before planning – it’s to distinguish between early planning decisions that preserve options and later irreversible capital commitments that require stronger evidence. “The real discipline is not choosing between ‘build early’ and ‘wait,’” Osnato said. “It is deciding which decisions are reversible, which are irreversible, and how much evidence should be required before crossing each one.”