Underwriting
Powered Land and Grid Advantage
California's September 2026 laws address rate design, grid-upgrade cost allocation, clean-energy supply, and water planning for data centers. Texas's September 21 executive directive pauses TCEQ data-center permits pending an ERCOT review and conditions other state approvals on information sought by PUCT, ERCOT, and TWDB. For powered-land diligence, these actions add jurisdiction-specific policy and timing checks; neither establishes site-level power availability, a completed permit denial, project cancellation, or a quantified capacity change. See Source: Texas and California Data Center Policy Actions September 2026.
Also known as: grid advantage, power-and-fiber advantage, powered land
Updated
Loudoun's pending application-pause proposal adds a separate entitlement-timing check for sites needing legislative approval: filings and Planning Commission review could continue, while final Board action may be paused if the October 20 resolution is approved. Treat approval path and timing as unresolved until the Board acts. See Source: Loudoun County Data Center and Substation Application Pause Proposal 2026.
Maryland now has an executive-order review gate for data-center projects of at least 25 MW seeking state action. Virginia's EO 22 restricts specified discretionary state site-readiness or expedited-review assistance at 25 MW or greater, while a broader local-approval rule is only proposed for 2027. In Prince William County, the September overlay change is reported as adopted, but final ordinance text and the acreage remain unresolved; check the parcel's overlay position and SUP path against the final county record. See Source: Mid-Atlantic Data Center Rules September 2026.
Property-Specific Entitlement and Power Evidence
For Data Center Underwriting and Powered Land and Industrial Logistics Underwriting, record separate evidence for site control, local entitlement, utility service, interconnection classification, deliverable MW and service dates. Keep requested, approved, contracted and energized capacity distinct. Price and financing remain conditional on the actual rights, obligations, capital required and tenant cash. Test delays, curtailment and failed approvals rather than treating queue position or nearby infrastructure as a guaranteed premium.
Phoenix provides a bounded example: a 5.43-acre site was reported sold for a data-center plan after the City adopted LKY's property-specific waiver item under its special-permit ordinance. The same reporting describes adjacent APS substation infrastructure, but no source here documents a service commitment, approved MW, or completed expansion. The screen therefore remains three separate gates: site control, a valid local entitlement path, and utility-confirmed deliverable power. See Source: GI Partners South Phoenix Data Center Site 2026.
Fort Worth provides a separate distinction between entitlement and energized capacity. Local reporting says an amended plan for a 187-acre, four-building Black Mountain data-center phase was approved in August 2026, and the plan includes an electrical substation. The captured city case page still displays an earlier continuation; neither the reported site-plan action nor a planned substation proves grid interconnection, deliverable MW, water service, or committed tenants. Keep those gates separate. See Source: Fort Worth Black Mountain Data Center Site Plan.
Americas Delivery and Precommitment (H1 2026)
Cushman & Wakefield reports 37.7 GW of Americas capacity under construction with 91.7% precommitted, alongside 50.3 GW operational and 285 GW planned. That combination points to strong booked demand but also a large uncommitted future pipeline. These are provider-defined regional figures, not a local utility commitment or parcel-level power verification. A site's grid position, interconnection schedule, transmission funding, and executed tenant commitments still require direct evidence. See Source: Cushman & Wakefield Americas Data Center Update H1 2026.
This concept captures a recurring pattern across the Texas geography and strategy materials: in data-center, semiconductor, logistics, and advanced-manufacturing corridors, land is only valuable if power, transmission capacity, fiber access, and entitlement readiness make it actually usable.
The Nvidia / Corning manufacturing announcement adds a supply-chain corollary: grid and fiber advantage do not only affect data-center campuses; they also affect the manufacturing locations that produce optical connectivity and transmission inputs for those campuses. See Source: Nvidia Corning US Data Center Manufacturing Plants 2026.
The industrial power-shortage source adds a demand-side corollary: powered land is no longer only a hyperscale data-center screen. Automation, robotics, EV charging, switchgear-heavy manufacturing, and high-throughput logistics facilities can all turn electricity availability into a site-selection constraint. See Source: Industrial Automation Power Shortages Data Center Competition 2026.
Batch 72 adds three extensions to the powered-land frame. DOE nuclear financing and the reported NextEra / Dominion merger show that generation ownership, utility balance sheets, and procurement lead times are becoming part of the data-center site screen. CIM's Permanent Power Co. launch shows QOZ-oriented capital trying to own the rural generation / storage / transmission layer. The "whale hunting" source adds a demand-side counterpoint: inference workloads may favor smaller, closer-in projects where 20 MW to 90 MW can clear faster than a remote multi-GW campus. See Source: DOE To Finance Parts for 10 Nuclear Reactors to Power Data Centers, Source: NextEra Dominion Utility Merger and Data Center Power Demand 2026, Source: CIM Group Launches Platform To Invest In Rural QOZ Power Projects, and Source: The Whale Hunting Era Of Data Center Development May Be Ending.
The 2027 OZ round makes rural tax eligibility a potential data-center site-selection variable, while the Hawley bill and Wyden staff proposal show emerging federal debate about whether data centers should retain those benefits. The enacted rural incentive remains law in the sources reviewed; proposed restrictions are not current eligibility rules. A tax overlay cannot substitute for confirmed utility capacity, tract designation, or project-level QOF qualification. See Source: Data Centers and Opportunity Zone Tax Policy 2026.
Gradiant's Series E financing adds the water-services corollary: powered-land underwriting cannot stop at megawatts and fiber where AI infrastructure, semiconductor manufacturing, and high-density cooling create water-treatment and reliability requirements. Treat the financing as a platform signal, not a site-level proof point. See Source: Gradiant Closes on Series E Financing to Drive Advanced Water Solutions.
Hut 8 reports securing key interconnection and site approvals for its 1,000-MW Beacon Point campus in Nueces County, then contracting 704 MW of IT capacity under two 15-year leases. This is a sponsor-reported example of power access preceding staged delivery and tenant commercialization; the 1,000-MW utility figure is not IT load, and the Q3 2027 delivery target remains prospective. See Beacon Point Data Center Campus and Source: Hut 8 Beacon Point Data Center Campus 2026.
The June 2026 RSS verification pass checked the May 18-19 powered-land / data-center batch and promoted the source notes to reviewed. Use the batch for source-scoped support around power procurement, utility structure, generation/storage, inference-scale site selection, water/cooling services, DFW project finance, and non-core powered-market watchlists. Do not treat policy commentary, utility M&A reporting, rankings, or announcement-stage projects as delivered megawatts, closed loan terms, final interconnection capacity, or structured market observations without primary utility, permit, lender, company, or regulatory records.
Source: NAIOP Electric Grid Reliability and Commercial Real Estate July 2026 adds an independent public diligence frame: rising demand from AI data centers, EV charging, and cold storage can combine with baseload retirements to make capacity, redundancy, utility reliability, and substation funding site-selection determinants. Its Texas/Southwest/corridor characterization is directional, not a capacity map. The underwriting consequence is to require documented queue position, interconnection, energization timing, utility commitments, and tenant need before crediting a powered-land premium or proposed rent uplift.
Source: Colliers AI Infrastructure Spending on Track to Exceed $800 Billion adds a capital-demand overlay, citing Morgan Stanley's projection of more than $800 billion of 2026 investment by major cloud and technology providers and Bloomberg analysis of more than $850 billion in future cloud-provider lease obligations. Treat those as corporate-level, source-attributed demand signals; they strengthen the need to test power, developable land, fiber, tenant credit, and delivery timing but do not prove local contracted NOI or delivered capacity.
Source: Cushman & Wakefield Turning Land into Value Data Centres June 2026 adds an APAC site-value control: infrastructure-ready land can command a premium only where utility access, power capacity, fiber, water, zoning, and delivery timing are demonstrably usable. Keep the source qualitative and do not capitalize a generic land premium without technical diligence, utility engagement, and an executable energization path.
Source: RCLCO Data Center Pipeline Boom or Bubble 2026 adds a scenario-control overlay to that power-first screen. RCLCO's 80-to-134-GW pipeline cases and associated value / equity estimates are scenario outputs rather than market benchmarks; the transferable instruction is to favor projects with a near-term road map to power and stabilization, then test construction funding and exit timing against a larger potential supply wave. A negotiated forward sale can improve residual certainty only when its buyer, price, conditions, and closing path are executed; it does not substitute for utility, tenant, or delivery evidence.
Source: CBRE North America Data Center Trends H1 2025 adds the first-half 2025 operating proof for powered-land scarcity. CBRE's applied rows show 5,242.5 MW under construction across primary markets with 74.3% already committed, while the public text says power availability and infrastructure delivery timelines remained decisive for site selection, leasing, and pricing. Pricing also separated by requirement scale: 250-to-500-kW requirements rose 2.5%, while larger 10 MW-plus deployments rose up to 19% where contiguous power blocks were scarce.
Source: CBRE North America Data Center Trends H2 2025 turns the same logic into a market-wide operating constraint. CBRE's applied rows show 1.4% primary-market vacancy, 2,497.6 MW of 2025 net absorption, 5,994.4 MW under construction, and Northern Virginia / Northeast site costs exceeding $8M per acre. That makes grid position and speed-to-power a pricing input rather than a background diligence item.
Source: JLL 2026 Global Data Center Outlook adds the global rent and delivery-risk version of the same point. JLL forecasts the Americas as the fastest-growing region at 17% supply CAGR through 2030 and 7% lease-rate CAGR, but it also reports multiyear grid-connection waits, 33-week equipment lead times, and 57% of projects delayed three months or more in 2025. Powered land therefore deserves premium treatment only when the power path, equipment procurement, cooling design, and local approval path are concrete, not merely claimed.
Source: JLL North America Data Center Report Year-end 2025 adds the regional pipeline proof behind that premium. JLL reports more than 35 GW under construction in North America, 92% of under-construction capacity precommitted, and 64% of that capacity in frontier markets. That makes powered land a scarce asset only when the grid and tenant path are real: West Texas, Tennessee, Wisconsin, and Ohio can matter more, but only where power, land, utility process, and capital structure convert into deliverable capacity.
Source: JLL Power Progress in Your Global Data Center Expansion adds the portfolio-execution version of the same constraint. JLL's August 2025 article reports U.S. power-delivery waits ranging from roughly 2 to 3 years in the Mountain West and New York metro areas to 8 to 10 years in the Pacific Northwest, with Northern Virginia at 3 to 5 years unless sites sit close to existing power lines. That reinforces powered land as an energization-timing asset: proximity to power, utility process, and energy-provider partnership structure determine whether global data-center demand becomes usable capacity.
Source: Green Street Launches Green Street Infrastructure and Source: Green Street Infrastructure Research Now Available add the lender / capital-market coverage layer. Green Street's public launch release describes infrastructure transaction, financing, and asset database coverage, while the later research release says its infrastructure research includes data centers plus infrastructure debt capital markets, lending volumes, and pricing-trend analytics. For powered-land underwriting, that supports linking power and utility feasibility to capital availability, but the public releases do not expose spreads, LTVs, loan terms, or market-level observations.
Source: CBRE Global Data Center Trends 2026 turns the powered-land thesis into a current global market-statistics screen. CBRE reports that emerging markets such as Queretaro, Johor, Batam, Tennessee, and West Texas are gaining traction because scalable land and power availability can support large deployments, while established hubs like Northern Virginia, Chicago, London, and Frankfurt face grid and development-timeline constraints. The implication is that powered land is valuable only when speed-to-power is underwritten alongside zoning, utility delivery, community opposition, and construction-cost exposure.
Source: CBRE Atlanta Data Center Market H2 2025 turns that concept into a Georgia utility-capacity example. CBRE cites over 3 GW of power commitments and over 10 GW of future expansion enabled by approved Georgia Power generation capacity, but the same page warns that clustered mega-scale projects are creating transmission constraints and that state incentive uncertainty complicates hyperscaler planning. The powered-land read is therefore conditional: grid commitments create option value, but transmission upgrades, load geography, and incentive durability decide whether the option is financeable.
Source: CBRE Atlanta Data Center Market H1 2025 shows the earlier utility-screening layer behind that H2 option value. Georgia Power had secured over 8 GW of new generation capacity scheduled after 2030, but it was also requiring developers to show end-user commitments, zoning confirmation, and collateral for power procurement. The same H1 source says a major hyperscale user preleased over 300 MW and planned another 600 MW by year-end. Powered land in Atlanta should therefore be priced as customer-backed utility capacity, not raw land near a growth market.
Source: CBRE Chicago Data Center Market H2 2025 adds the credit-support version of grid advantage. CBRE reports Chicago data-center vacancy at 1.9% and no contiguous 5 MW-plus space, but also says ComEd-served projects face delivery delays until 2032 or later and require 10-year letters of credit for 50 MW-plus projects. Powered land here is not merely land near Chicago demand; it is a site with financeable utility timing, balance-sheet support, and transmission-revenue security.
Source: CBRE Chicago Data Center Market H1 2025 shows why that credit-support screen matters. The source reports 72 MW leased to an AI-related tenant and Yorkville projects seeking approval for 38 buildings and 3+ GW of power, while also saying ComEd delivery was already delayed to 2031 or later. Powered land in Chicago is valuable only where large-load demand can clear approval, transmission, letter-of-credit, and energization gates.
Source: CBRE Dallas-Ft. Worth Data Center Market H2 2025 adds the Texas scale-market version of grid advantage. CBRE reports a 1 GW colocation market with 2.4% vacancy, roughly 700 MW under construction at 94.5% preleased, and 3 GW of planned greenfield development, while also flagging throttled power delivery, queue-entry controls, rising site pricing, and behind-the-meter power solutions. Powered land here means a site whose utility queue position and preleasing path can turn broad greenfield demand into deliverable capacity.
Source: CBRE Dallas-Ft. Worth Data Center Market H1 2025 adds the earlier setup for that same DFW gate. The preserved H1 profile reports 869.5 MW of existing inventory and 425.1 MW under construction at 78% preleased, with Oncor transmission planning, rising data-center development-site pricing, unprecedented utility requests, and behind-the-meter power studies already in the source text. Source: CBRE North America Data Center Trends H1 2025 Market Profiles adds a collection-page snippet that DFW was expected to double in size by 2026 and had 89% of under-construction space preleased; keep that as source-scoped support for heavy preleasing, not a denominator-clean replacement for the dedicated profile row. That reinforces powered land as a queue-position and utility-execution asset rather than merely a cheap-land asset.
Source: CBRE Phoenix Data Center Market H1 2025 and Source: CBRE Phoenix Data Center Market H2 2025 add the utility-queue version of grid advantage. CBRE's H1 profile reports just over 75 MW of Phoenix space-and-power absorption and SRP study work for 26 citywide projects; the H2 profile reports 256 MW of colocation space-and-power absorption across four citywide projects, while also saying SRP and APS restrictions are limiting powered-land availability and SRP's first 25-applicant cluster study had mixed initial results. The powered-land lesson is that demand and near-term project completions are not enough; the site must survive utility study, power-delivery timing, and local infrastructure allocation.
Source: CBRE Hillsboro OR Data Center Market H2 2025 adds the policy-design version of grid advantage. CBRE says no Hillsboro facilities had 5 MW-plus contiguous space available, PGE plans $6.5 billion of clean-energy and grid-modernization investment over five years, and Oregon's POWER Act creates a data-center customer class starting at 20 MW. Grid advantage here is not only access to power; it is the interaction of scarce capacity, utility capex, and a tariff / customer-class framework that can support or constrain large-load development.
Source: CBRE New York Tri-State Data Center Market H2 2025 adds the power-price and low-latency metro version. CBRE says PJM pricing increases, high power costs, and limited tax incentives are pushing occupiers to consider adjacent markets or limiting relative growth, even as New Jersey powered-land interest and financial-services low-latency workloads keep demand alive. Grid advantage here is not cheap land; it is a site whose power economics, tax setting, and latency profile can overcome a tight 5 MW-plus contiguous-space market.
Source: CBRE New York Tri-State Data Center Market H1 2025 adds the earlier contiguous-capacity version of the same gate. CBRE said no capacity of more than 5 MW was currently available and AI occupiers were preleasing in 5-to-10-MW increments where available, while Equinix's NY3 project was coming to market as its 10th Tri-State data center. Powered land or powered building exposure in the Tri-State market has to be underwritten through facility-level vacancy, operator campus expansion, power certainty, and latency requirements rather than broad metro demand alone.
Source: CBRE Northern Virginia Data Center Market H2 2025 adds the mature-corridor preleasing version. CBRE says Northern Virginia delivered more than 1 GW of capacity in 2025 and still ended H2 at 0.5% colocation vacancy, while 15-to-20-MW contiguous blocks were difficult to secure and Dominion batching extended new-project delivery timelines. Grid advantage here is proven capacity plus a credible power path, not simply being inside the largest data-center market.
Source: CBRE Northern Virginia Data Center Market H1 2025 shows that the same gate was visible earlier in 2025. The H1 rows combine 0.72% vacancy, AWS's 97 Leesburg acres, BlackChamber's 65 Prince William acres, and EdgeCore's 114 MW Loudoun construction start. Powered land here means acreage with a credible Dominion, zoning, and preleasing path, not raw county exposure.
Source: CBRE Silicon Valley Data Center Market H2 2025 adds the power-cost premium-market version. CBRE says 10 MW-plus contiguous space commands a premium, Silicon Valley pricing surpasses other North American markets, and new construction remains limited by power challenges. Grid advantage here is scarce and expensive: innovation-cluster adjacency matters, but the site has to survive utility cost, power availability, and approval risk.
Source: CBRE Silicon Valley Data Center Market H1 2025 adds the H1 capacity-scarcity base. CBRE says power constraints limited new supply even while over 50 MW was immediately available or under construction, vacancy dropped to 4.5%, and some planned data-center sites reverted to industrial use because of power limitations. Grid advantage here means having power that can actually support the intended data-center use, not merely owning Bay Area industrial land.
Source: CBRE Minneapolis Data Center Market H1 2025 adds the secondary-market renewable / stability version. Minneapolis benefits from power stability and renewable-energy access, but CBRE also flags permitting and environmental regulation as cost and timing constraints. The Eagan 61,000 SF / 5 MW build-to-suit activity is useful as a site-specific powered-land signal, while the proposed Minnesota pipeline should remain context until project-level power, entitlement, and delivery evidence is preserved.
Source: CBRE Houston Data Center Market H1 2025 adds the behind-the-meter energy-market version. Houston's oil-and-gas occupier demand and multiple behind-the-meter power studies make it relevant to powered-land underwriting, but the 19.7% vacancy row after recent single-operator delivery keeps the signal disciplined: power adjacency does not automatically equal scarcity.
Source: CBRE Denver Data Center Market H1 2025 adds the power-block scarcity version. Denver has a modest market footprint at 20.1 MW across 27 facilities, with only two colocation facilities offering large amounts of power and space. Powered land here is valuable only where utility delivery, water / cooling, state incentive treatment, and facility-level power blocks are proven.
Source: CBRE Austin-San Antonio Data Center Market H1 2025 adds the Central Texas tight-market version. Austin-San Antonio shows 2.3% vacancy, 97% preleased under-construction capacity, and 462.0 MW under construction. Powered land here should be screened by Oncor, Lower Colorado River Authority, water / cooling, entitlement, community acceptance, and tenant commitments rather than by generic Austin or San Antonio growth.
Source: CBRE Seattle Data Center Market H1 2025 adds the dense-power / network-market version. Seattle's value is tied to AI, enterprise, and hyperscaler colocation demand, state tax-exemption support, fast-track occupancy, liquid cooling, and generator options; the clean structured row is Sabey's scheduled 6 MW SDC Building 5 expansion, not a broad Seattle powered-land capacity claim.
Source: CBRE Central Washington Data Center Market H1 2025 adds the rural utility-delay version. PUD transmission-line delays, private power systems, developer-funded substations, alternative power, microgrids, and interim generators all become part of the powered-land screen. The planned Tri-Cities campus scale ($4.8B, 16 data centers, 215,000 SF each) is meaningful only if the power-delivery path can clear.
Source: CBRE Charlotte-Raleigh Data Center Market H1 2025 adds the Carolinas utility-commitment version. Duke Energy's end-user commitment and capital-contribution requirements make powered land less about nominal site control and more about who can fund, contract, and wait for deliverable power. The structured rows are deliberately project-scale (1 MW delivered, 1,000,000 SF conversion, 1,600,000 SF announced campus, $10B announced buildout), not a claim that all Carolinas sites are powered.
Source: CBRE Southern California Data Center Market H1 2025 adds the infill carrier-hotel version of grid advantage. One Wilshire's over-10 MW available shell-capacity row is valuable because of connectivity and existing building position, but it still has to clear power cost, equipment, fit-out, and customer-demand underwriting. Powered land and powered buildings are different risk packages.
Source: CBRE Montreal Data Center Market H1 2025 adds the power-allocation version of grid advantage. Quebec hydro power supports AI/HPC demand, but CBRE says operators without pre-approved Hydro-Quebec service agreements or under the 5 MW threshold faced delays. Grid advantage here is an allocation and service-agreement asset, not simply a cheap-power geography.
Source: CBRE Toronto Data Center Market H1 2025 adds the GTA land-and-utility-study version. CBRE says data-center development land commands a premium over traditional industrial projects because of limited sites with sufficient grid capacity, while developers are paying for power studies and new applications with Toronto Hydro and Alectra. Powered land here means grid-capable, securable sites, not just industrial acreage.
Source: CBRE Hillsboro Data Center Market H1 2025 adds the Oregon capacity-expansion version. NTT's additional 216 MW toward 354 MW of total planned Hillsboro capacity is meaningful only because larger requirements struggle to find operator vacancy and Oregon's utility-cost allocation framework can affect developer economics. Powered land here means capacity that can clear operator, utility, and regulatory gates together.
Source: CBRE Data Centers - U.S. Real Estate Market Outlook 2026 sharpens that speed-to-power rule. CBRE's 2026 outlook says 300 MW-plus delivery inside 36 months can now matter more than pure connectivity, 500 MW-plus AI campuses push schedules into multi-year territory, and added high-voltage transmission or generation can push interconnection timelines to 24, 36, or 48-plus months. Treat powered land as premium only when those timelines are evidenced, not merely asserted.
Source: Avison Young Q1 2026 U.S. Data Center Market Overview adds a U.S. absorption signal to the powered-land screen. Austin recorded 1,909 MW of Q1 2026 colocation net absorption and Dallas-Fort Worth recorded 912 MW, inside a national market with 29.0 GW of inventory and only 1.2% vacancy. Use that as evidence that Texas absorption can be very large when capacity is deliverable, but still verify utility queue position, energization timing, water/cooling design, labor availability, preleasing, and local process before treating a site as truly powered.
Source: Colliers 2026 Data Center Marketplace Report turns powered land into an explicit capital and utility-deposit screen. Colliers says power infrastructure can account for 40%-50% of total project costs, utility deposits of $25M-$75M-plus are now standard, private credit supplies roughly 60%-75% of early-stage funding, and more than $64B of U.S. projects has been delayed or canceled since 2023. That makes powered land valuable only when power certainty, deposit funding, local approvals, environmental review, community acceptance, and phasing discipline are underwritten together.
Source: Newmark 2026 U.S. Data Center Market Outlook adds a Texas-specific powered-industries framing to that screen. Newmark's public page says the U.S. market has a 160 GW under-construction / announced pipeline and 2% vacancy, while also identifying Texas as a concentrated ecosystem where data-center growth, power infrastructure, advanced manufacturing, and cross-border supply chains overlap. Use that as support for Texas infrastructure convergence, not as proof that announced capacity will clear the grid or local approval process.
Source: Cushman & Wakefield Data Centres: EMEA Growth Is Robust But Being Reshaped adds a non-U.S. stress test for the same powered-land rule. C&W's EMEA interview says demand is not the issue; power availability and pricing, permitting timelines, supply chain, regulatory complexity, and off-taker certainty determine how fast capacity can come online. That makes "shovel ready" land valuable only when timely power delivery and planning are confirmed, not simply when a market has strong AI or cloud demand.
Source: Cushman & Wakefield From Megawatts to Meaning May 2026 narrows the powered-land premium even further. If AI data centres are increasingly evaluated by cost per token, a site with nominal megawatts is not automatically superior; the underwritten advantage has to convert power into workload output through the right cooling, hardware, utilization, latency, and customer economics. Use it as KPI framing, not as a market-data row.
Source: CBRE Q2 2026 U.S. Distributed Generation Opportunities adds the distributed-generation revenue-stack layer. CBRE's public article ties state battery programs, utility tariffs, and capacity-market structures to front-of-the-meter leases and behind-the-meter bill savings, including PJM data-center capacity contracts, Maryland / New Jersey battery solicitations, Illinois rebates, New York offsite-battery reforms, and Massachusetts long-term storage contracts. The underwriting implication is that powered land should be screened for monetizable flexibility, not just nameplate service.
Source: JLL Valuing the Rooftop Opportunity 2026 adds the warehouse-rooftop version of that flexibility screen. JLL frames industrial/logistics rooftops as distributed-generation infrastructure because warehouse energy intensity is low relative to roof area; BTM systems may use only about 20% of available roof area, while FTM leases can monetize nearly the full roof as contracted third-party income. The power-premium caveat is policy-specific: JLL names New Jersey, Maryland, Illinois, New York, and Massachusetts as the strongest current FTM markets, so rooftop optionality should not be capitalized uniformly across every warehouse state.
The June 15 digital-infrastructure tranche reinforces powered land as a combined power, process, and capital-stack asset. Goodyear and San Jose show Microsoft accumulating and activating very different site types; Bastrop County shows local process becoming part of the site-control screen; Rowan / Temple and Switch show debt and credit facilities tied to large powered pipelines; and DigitalBridge / ArcLight plus Helix show capital moving upstream into generation, transmission, and AI-infrastructure platforms. See Source: Microsoft Goodyear Data Center Land 2026, Source: Microsoft North San Jose Data Center Groundbreak 2026, Source: Bastrop County EdgeConneX Data Center Community Meeting 2026, Source: Rowan Temple Third Data Center Deal 2026, Source: Switch Credit Facilities Nearly 10B 2026, Source: DigitalBridge ArcLight Power Infrastructure Acquisition 2026, and Source: KKR NVIDIA Kuwait Helix AI Infrastructure Platform 2026.
Source-Derived Pattern
Across the current canonical branch, the legacy materials keep returning to the same strategic logic:
- data-center growth is constrained less by demand than by access to power and fiber
- semiconductor and manufacturing corridors depend on utility capacity and infrastructure timing, not only on labor or land price
- industrial corridors with strong intermodal access become even more valuable when they also offer grid headroom
- Texas metros increasingly compete on infrastructure readiness rather than just population growth
Why It Matters
This concept connects:
- Texas because the state thesis explicitly flags AI infrastructure and power constraints as a key theme
- Digital Infrastructure Real Estate because power, fiber, and readiness are the enabling layer beneath the real-estate side of AI and connectivity infrastructure
- Conviction Theme Investing because data centers and infrastructure-heavy sectors are part of the current high-conviction capital rotation
- Equinix because the listed data-center and interconnection layer provides a public-market benchmark for the same infrastructure logic
- American Tower because wireless and communications infrastructure extend the same digital-physical layer beyond data centers alone
- Alliance and North Fort Worth because powered land and intermodal infrastructure are central to the Alliance thesis
- Williamson County Semiconductor Corridor because semiconductor growth depends on utility and corridor capacity, not just employer announcements
- Plano Richardson Telecom Corridor because semiconductor, telecom, and data-center adjacency create a distinct infrastructure-led investment case
- Northern Virginia Digital Infrastructure Corridor because Dominion's contract pipeline and load forecast show how power delivery timing becomes the binding risk in a mature hyperscale corridor
Related Pages
-
Texas
-
Equinix
-
American Tower
-
Industrial Hub
-
Alliance and North Fort Worth
-
Williamson County Semiconductor Corridor
-
Plano Richardson Telecom Corridor
-
Northern Virginia and Washington DC
-
Northern Virginia Digital Infrastructure Corridor
Hyperscaler Validation of the Powered-Land Thesis (2025–2026)
The following announcements provide distinct project and procurement examples. They do not establish delivered capacity or a universal ranking of site constraints:
OpenAI Stargate — announcement scope (October 2025): OpenAI announced over 8 GW of planned program capacity and more than $450B of intended investment over three years. Its Michigan announcement describes DTE using existing excess transmission capacity and the project funding required upgrades. These are sponsor statements about plans and cost allocation, not proof of completed capacity, actual ratepayer outcomes, or power being the sole site-selection constraint. See Source: OpenAI Plans to Build Data Center Near Ann Arbor, Invest $450B Over Next Three Years.
The April 27 Related Digital report, captured May 18, adds Saline capital-stack announcement context. Source: Related Digital Secures Financing for $16B Oracle Data Center Project in Michigan keeps the developer’s closing conditions, subsequent construction account and conditional utility-contract evidence separate. Use Michigan Data Centers and Powered Land, CRE Capital Markets, Data Center Underwriting and Powered Land, Energy Infrastructure and CRE and Source: MPSC Saline Data Center Contract Admission 2025-26 for actual funding, load boundaries, approvals and service. A financing announcement does not prove a site premium, complete closing or energized capacity.
Google / DTE Michigan — 2.7 GW of new resources (Mar 2026): Google says a proposed Michigan data center will be served by 2.7 GW of new grid resources spanning solar, advanced storage, and demand flexibility under a Clean Capacity Acceleration Agreement with DTE. The article makes the cost-allocation logic explicit: Google says it will cover its electricity and infrastructure needs while supporting grid reliability and ratepayer protection. The Van Buren Township site was still under evaluation, so this is procurement-structure evidence—not disclosed IT load, final site control, completed generation, or delivered capacity. See Michigan Data Centers and Powered Land and Source: Google Michigan Clean Energy and Reliability Announcement 2026.
Google Texas — announcement scope (November 2025): Google announced $40B of Texas cloud and AI infrastructure investment through 2027, including Armstrong and Haskell County campuses. It identified one Haskell data center alongside solar and battery storage. This supports integrated infrastructure diligence; it does not show that every campus owns generation or that metro congestion caused the location choices. See Source: Google to Invest $40B in Texas Data Centers Through 2027. Use Energy Infrastructure and CRE to distinguish planned co-location, ownership, contracted resources, parcel service and operating continuity. Assess actual delivery and cost obligations through Data Center Underwriting and Powered Land before applying a site or investment advantage.
Microsoft Castroville — January 2026 reported plan: The SAT82 article describes a Medina County proposal. Keep its historical cost and schedule separate from the later-read accessibility registry and actual delivery. Review Complete does not verify construction completion or usable power. The report supplies no verified transmission headroom, utility commitment or comparative site-selection analysis. See Source: Reports: Microsoft Plans $400M Data Center in Castroville, Texas, Austin-San Antonio Data Centers and Powered Land Market, Data Center Underwriting and Powered Land, Energy Infrastructure and CRE and CRE Capital Markets for geography, phase/service and funding diligence.
Use these cases to ask for site-specific power, transmission, procurement, entitlement, capital, labor, and delivery evidence. Their announcements do not establish which constraint dominates any other project.
Penzance Bedington WV (March 4 report; May 18 capture): The reported 600 MW / 548-acre Berkeley County campus is a planned West Virginia watchlist case. Source: Penzance to Develop $4B Data Center Campus in Bedington, West Virginia separates the article’s unspecified MW label from the governor’s dated critical-IT description and designation announcement. Use Data Center Underwriting and Powered Land, Energy Infrastructure and CRE, CRE Tax and After-Tax Underwriting and CRE Capital Markets for phase delivery, operative terms, service and actual funding; designation is not proof of those conditions.
Amazon Ashburn campus acquisition (May 2026 source): Amazon's reported purchase of GW's Virginia Science and Technology Campus is a mature-corridor land-control example, where redevelopment optionality matters because Ashburn already sits inside the data-center / Dulles infrastructure ecosystem.
Amazon Bristow / Prince William acquisition (November 11, 2025 report; May 18, 2026 capture): Source: Amazon Acquires Data Center Site in Northern Virginia for $700M relays a $700 million land acquisition near Devlin Road and I-66. The report does not establish powered-land scarcity as the cause, a scarcity premium or delivered service. Use Northern Virginia Digital Infrastructure Corridor, Data Center Underwriting and Powered Land, Energy Infrastructure and CRE, CRE Capital Markets and CRE Valuation Methodologies for acquired rights, current legal scope, utility obligations, funding and value. Hold the relayed rezoning/court account separately from current clearance.
Amazon Indiana — Utility Partnership (2026): The retained November 26, 2025 report describes a planned $15B expansion and an Amazon/NIPSCO power arrangement. It does not establish Amazon ownership of GenCo, completed power delivery or measured ratepayer protection. Keep the reported investment and capacity plans separate from asset ownership and service performance. Use Energy Infrastructure and CRE to verify the actual counterparties, corporate structure, contracts, approvals, upgrade costs and service dates. The arrangement is a project-specific diligence case, not proof that all hyperscalers own utility infrastructure or that this structure matches Google's Texas model. See Source: Amazon Plans to Invest $15B to Build New Data Center Campuses in Northern Indiana.
Beale / De Soto Kansas (April 30 report; captured May 18, 2026): Project Pilot is a phased Kansas-side watchlist case. The investment plan, forecast first-phase start, unnamed tenant and relayed site/incentive accounts do not establish usable service, actual spending or an investment premium. Use Olathe and Western Johnson County Growth Corridor, Data Center Underwriting and Powered Land, Energy Infrastructure and CRE and CRE Tax and After-Tax Underwriting to keep phase, service and incentive conditions distinct. See Source: Beale Infrastructure Unveils Plans for $3B Data Center Campus in De Soto, Kansas.
Vantage VA4 / Stafford County (November 2025 report; captured May 2026): The announced investment, 192 MW campus plan and late-2027 first-building target are distinct scopes. Stafford / Fredericksburg is outside the Loudoun / Prince William core used here. Cooling design and a reported groundbreaking celebration do not establish usable power/fiber, operating water performance, committed tenants or completed delivery. See Source: Vantage Data Centers to Invest $2B for New Campus in Northern Virginia, Data Center Underwriting and Powered Land and Energy Infrastructure and CRE for phase/service and cooling boundaries.
PW Digital Gateway / Compass withdrawal (May 2026 sources): Compass's decision to walk away from the Prince William County project after the court ruling adds a local-consent gate to the powered-land screen. The reported scale was enormous, but the source stack shows that power, fiber, and land do not become usable capacity without durable rezoning, notice, and community-process footing. See Source: Compass Abandons PW Digital Gateway After Court Ruling 2026 and Source: Compass Datacenters CEO On Walking Away From Prince William 2026.
Powerhouse / Caldwell County TX (May 2026 source): The proposed Caldwell Valley Technology Park project near Uhland adds an Austin-exurban powered-land watchlist item. The reclaimed-water framing is relevant because water use can become a gating issue for Texas data centers, but the source remains announcement-stage and should be verified against local, utility, and developer records.
CBRE Austin-San Antonio H1 2025: CBRE's combined Central Texas profile adds hard market context to the Austin / San Antonio exurban watchlist. The source reports 2.3% vacancy, 97% preleased under-construction capacity, and 462.0 MW under construction, while pointing to Oncor-served northeast Austin suburbs and activity moving south into Lower Colorado River Authority territory. Treat this as a utility-service-territory screen, not a blanket premium for all Austin-San Antonio land. See Source: CBRE Austin-San Antonio Data Center Market H1 2025.
Prime SMF02 / Sacramento (May 8 report; captured May 18, 2026): The planned 150,000 SF / 18 MW critical IT load is a project watchlist input. Route McClellan evidence through Sacramento Data Centers and Powered Land Market, keeping county building-program areas, reported groundbreaking and SMF01 operation separate from SMF02 delivered capacity. Verify utility obligations and actual phases through Data Center Underwriting and Powered Land; company demand/location views do not establish an investment advantage.
St. Louis powered-infrastructure commentary (February 12, 2026 web report): JLL's legacy-site thesis is a watchlist input, not verified spare power or shorter delivery. Its 7.69 / 8.65 cents/kWh comparison has no explicit measurement period and cannot price site service. The primary-summary review separates the minimum service term from an optional additional ramp period and holds exact tariff/exit-fee terms for contract review. Use Energy Infrastructure and CRE, Data Center Underwriting and Powered Land and St Louis Industrial and Logistics Market for service, underwriting and Missouri / Illinois scope. See Source: Why Data Centers Are Emerging as a Defining Opportunity for the St. Louis Region.
Connect CRE's summary of IIF's PNNL-atlas analysis distinguishes the geography of facility counts from the geography of hyperscale campus capacity: dense hubs such as Data Center Alley and Silicon Valley may host more facilities, while larger campuses are more common in less-populated regions. The report summary leaves causality unresolved: available power may attract campuses, or generation may expand in response. Treat this as a qualitative screen and verify utility capacity and project scale directly. See Source: Connect CRE IIF Data Center Geography 2026.
Sources
-
Source: Fort Worth Black Mountain Data Center Site Plan
-
Legacy Texas Market Thesis
-
Legacy CRE Investment Strategy
-
Source: OpenAI Plans to Build Data Center Near Ann Arbor, Invest $450B Over Next Three Years
-
Source: Google to Invest $40B in Texas Data Centers Through 2027
-
Source: Reports: Microsoft Plans $400M Data Center in Castroville, Texas
-
Source: Amazon Plans to Invest $15B to Build New Data Center Campuses in Northern Indiana
-
Source: Penzance to Develop $4B Data Center Campus in Bedington, West Virginia
-
Source: George Washington University Sells Virginia Science and Technology Campus to Amazon for $427M
-
Source: Amazon Acquires Data Center Site in Northern Virginia for $700M
-
Source: Beale Infrastructure Unveils Plans for $3B Data Center Campus in De Soto, Kansas
-
Source: Related Digital Secures Financing for $16B Oracle Data Center Project in Michigan
-
Source: Vantage Data Centers to Invest $2B for New Campus in Northern Virginia
-
Source: Powerhouse Data Centers Caldwell County Data Center Proposal
-
Source: CBRE Austin-San Antonio Data Center Market H1 2025
-
Source: Prime Data Centers Breaks Ground on Second Sacramento Facility
-
Source: Why Data Centers Are Emerging as a Defining Opportunity for the St. Louis Region
-
Source: Compass Abandons PW Digital Gateway After Court Ruling 2026
-
Source: Compass Datacenters CEO On Walking Away From Prince William 2026
-
Source: DOE To Finance Parts for 10 Nuclear Reactors to Power Data Centers
-
Source: CIM Group Launches Platform To Invest In Rural QOZ Power Projects
-
Source: NextEra Dominion Utility Merger and Data Center Power Demand 2026
-
Source: The Whale Hunting Era Of Data Center Development May Be Ending
-
Source: Battery Storage in Data Centers
-
Source: Gradiant Closes on Series E Financing to Drive Advanced Water Solutions
-
Source: CBRE North America Data Center Trends H1 2025
-
Source: JLL Power Progress in Your Global Data Center Expansion
-
Source: CBRE Global Data Center Trends 2026
-
Source: CBRE Atlanta Data Center Market H2 2025
-
Source: CBRE Atlanta Data Center Market H1 2025
-
Source: CBRE Chicago Data Center Market H2 2025
-
Source: CBRE Chicago Data Center Market H1 2025
-
Source: CBRE North America Data Center Trends H1 2025 Market Profiles
-
Source: CBRE Dallas-Ft. Worth Data Center Market H1 2025
-
Source: CBRE Dallas-Ft. Worth Data Center Market H2 2025
-
Source: CBRE Phoenix Data Center Market H1 2025
-
Source: CBRE Phoenix Data Center Market H2 2025
-
Source: CBRE Seattle Data Center Market H1 2025
-
Source: CBRE Central Washington Data Center Market H1 2025
-
Source: CBRE Charlotte-Raleigh Data Center Market H1 2025
-
Source: CBRE Southern California Data Center Market H1 2025
-
Source: CBRE Montreal Data Center Market H1 2025
-
Source: CBRE Toronto Data Center Market H1 2025
-
Source: CBRE Hillsboro Data Center Market H1 2025
-
Source: CBRE Northern Virginia Data Center Market H1 2025
-
Source: CBRE Silicon Valley Data Center Market H1 2025
-
Source: CBRE Hillsboro OR Data Center Market H2 2025
-
Source: CBRE Minneapolis Data Center Market H1 2025
-
Source: CBRE Houston Data Center Market H1 2025
-
Source: CBRE Denver Data Center Market H1 2025
Loudoun Proposal — September 24 Evidence Boundary
Full retained county and reporting-source review confirms that September 15 advanced a proposed final-Board-action pause. The October 20 resolution was pending as of September 24. This record does not establish an enacted moratorium, an October 6 decision, project cancellations or loss of powered capacity. Source: Source: Loudoun County Data Center and Substation Application Pause Proposal 2026.
Complete held source review, October 11, 2026
The complete September 16 Connect article reports Cushman & Wakefield's Americas data-center construction at a record 37.7 GW. It does not provide a named facility, power deliverability, commissioning date, market allocation or dataset methodology. This is secondary coverage of the same primary research, not an independent normalized capacity series. John McWilliams, head of data-center insights at C&W, describes power availability, grid infrastructure and regulatory certainty as competitive advantages and demand as insatiable; these are attributed commentary, not measured unlimited demand or investor return. Established Northern Virginia, Atlanta, Chicago and Dallas are named for connectivity, customer concentration and mature ecosystems. West Texas, Pennsylvania, Alberta and Cheyenne are named as emerging areas with some large proposed developments. Proposed projects are not delivered, financed, permitted or energizable capacity; no individual allocation, ranking, price, available MW, utility commitment or parcel is supplied. Primary operational, precommitment, planned-pipeline and vacancy numbers discussed elsewhere in the source note are not additional values in this recap. All three complete faithful panels and linked rack illustration were inspected; panels are identical because the article fits in one viewport. The unlabeled rack image does not establish a real named facility, electrical load, cooling specification, rack count, physical condition or operating capacity. Portrait, biography, CTA and navigation are excluded from asset facts. Use held primary research for its own supported aggregates and keep missing embedded maps/datasets excluded. This article supports a source-specific siting hypothesis, not an investment conclusion without investor price, objectives, horizon, execution ability and power/site diligence.
- Source: Connect CRE Cushman & Wakefield Power and Data Center Map 2026.
Liberty application — readiness claim and service proof
Source: Metrobloks and Lincoln Property Co. to Develop 568,000 SF Data Center Campus in Metro Kansas City is an April 2026 Northland watchlist case. A described zoned/powered site and proximity to an exchange point do not establish actual service rights, usable IT capacity, latency or an investment premium. Preserve total utility-power versus critical-IT scope and reconcile developer/city plan versions through Data Center Underwriting and Powered Land and Energy Infrastructure and CRE. KCI Airport and Northland Logistics Corridor keeps the Liberty campus separate from downtown reuse and Kansas-side industrial projects.
GW Ashburn application — acquisition and usable service
Source: George Washington University Sells Virginia Science and Technology Campus to Amazon for $427M provides a reported institutional-campus acquisition, not a powered-site certification. Reconcile acquired parcels, continued university use, redevelopment rights, transmission/substation needs and utility commitments before assigning an energization advantage. Use Northern Virginia Digital Infrastructure Corridor for market context, Data Center Underwriting and Powered Land and Energy Infrastructure and CRE for service, Adaptive Reuse of Obsolete Office for reuse alternatives, and CRE Capital Markets / CRE Valuation Methodologies for funded costs and comparable rights. Keep price and acquisition evidence separate from demand or capacity measurement.
Complete official packet review, October 11, 2026
The full29-page September9 packet contains a blank resolution template, the adopted June9 initiation, separate board/staff drafts and an August13preliminary estimate. Page28 reports approximately237parcels/3654acres versus9698existingacres; the criterion counts1/0/78/3/155 sum237. Its pie labels use inconsistent rounding. Page29 depicts the existing overlay only and disclaims precise construction/property-boundary use. These figures and redlines do not certify the final ordinance, final post-deadline map, parcel eligibility, current supply or a scarcity premium. Keep the later county summary’s deadline and historical recommendations in their own dated sources.
- Source: Prince William County Overlay Complete Staff Packet Review September 9 2026.
Complete reproduced source review, October 11, 2026
As captured October11, the county project page states September22 adoption as Ordinance26-67, describes five eligibility paths and explicitly gives December21 as the fifth-criterion close. The county says the updated qualifying-parcel map will follow that deadline. These are county summary statements; the complete final ordinance and parcel-specific rights remain unheld. Older September9 recommendations for120days and500-foot setbacks, and a retained September22 hearing notice, are historical material rather than adopted deadline proof. Keep the listed SUP pathway distinct from a blanket development ban and from guaranteed approvals or power service.
- Source: Prince William County Overlay Complete Project Page Review October 11 2026.
Complete reproduced source review, October 11, 2026
Complete September22reporting distinguishes the8–0overlay vote from5–3rejection of zero-day grace. The3654-acre figure derives from an August13preliminary staff estimate; the packet map shows existing overlay only, and the county’s later summary says the updated qualifying map follows December21. Keep the3641-acre secondary figure unresolved. USDataCenterMap’s36operational/40pipeline sites include11under construction;4.5GW mixes reported and operational power and is not available energized capacity. Loudoun’s contemplated legacy change and up-to12monthpause belong to a different jurisdiction. QTS Digital Gateway100B/2100acres and DullesSouth1940acres are separate dated proposals/outcomes, not delivered supply or quantified losses. No final parcel rights or investor premium is certified.
- Source: Bisnow Prince William Overlay Complete Vote and Pipeline Review September 22 2026.