Core Scientific’s Texas Power Approvals Advance 1.4 GW of Data Center Capacity — but Final ERCOT Allocation Still Matters

Core Scientific has moved a large portion of its Texas power portfolio further through ERCOT’s interconnection process, a meaningful step for a company whose growth increasingly depends on converting secured electricity into revenue-producing AI and high-density colocation capacity.

In a September 10 filing, Core Scientific said its Denton, Pecos and Hunt sites received a mix of conditional approvals, prior study validation and exemptions under ERCOT’s new Batch Zero framework. BTI calculates that the capacities discussed across the three sites total 1,402 megawatts: 371 MW at Denton, 600 MW at Pecos and 431 MW at Hunt. That total should not be read as 1.4 GW of fully cleared, immediately deployable capacity because the regulatory status differs by site and some allocations remain conditional.

Denton has the clearest position. Core Scientific said 297 MW of existing electrical load was conditionally approved as “Base Load, Pathway (a),” reflecting load energized before March 25, 2022. That capacity is not subject to Batch Zero allocation. An additional 74 MW at Denton had already been incorporated into ERCOT’s 2025 Regional Transmission Plan study cases and is also outside the Batch Zero process. Together, those two blocks represent 371 MW with less Batch Zero uncertainty than the company’s newer Texas projects.

Pecos is more dependent on the new framework. Core Scientific said 300 MW of existing power was conditionally approved as “Base Load, Pathway (b),” covering load energized after March 25, 2022 and before July 10, 2026. A separate 300 MW block was conditionally approved as “Studied Load.” The company said it submitted the required documentation and posted collateral, but it also made an important qualification: final allocation and the load-ramp schedule will not be known until ERCOT completes the Batch Zero process.

That distinction is central for investors. ERCOT created Batch Zero as a transitional, system-wide process for large electricity users, including data centers, after the Public Utility Commission of Texas approved the framework in June. Projects of 75 MW or more can qualify through defined pathways, but ERCOT evaluates their combined impact on grid reliability and available transmission capacity. A conditional pathway designation therefore reduces one layer of development uncertainty without eliminating the risk that the final ramp, timing or usable allocation differs from a developer’s preferred schedule.

The Hunt County project is the most development-intensive of the three. Core Scientific said 431 MW of planned capacity received conditional approval as “Base Load, Advancing Large Load, Pathway (e).” The company and its utility partner have completed required interconnection studies, obtained ERCOT approval for a 431 MW stability study, posted financial collateral, ordered long-lead equipment and begun construction. Those steps indicate a more advanced project than a site that is merely seeking access to power, but Hunt still sits inside the Batch Zero process.

The Texas update matters because power access has become one of the most valuable constraints in AI infrastructure. Core Scientific is no longer primarily a bitcoin-mining story. In its second-quarter filing, the company said high-density colocation had become its primary business. Colocation generated $136.7 million of Q2 revenue versus $10.6 million a year earlier, and represented 77% of total revenue for the first half of 2026. As of June 30, Core Scientific controlled about 2.1 GW of gross utility power and roughly 1.3 GW of leasable customer capacity across 11 U.S. data centers.

The company has also been signing long-duration AI infrastructure agreements. Core Scientific disclosed in July that its strategic relationship with AMD could support up to 2.5 GW of leasable capacity, anchored by 15-year agreements for approximately 530 MW across five sites and more than $14 billion of potential base contracted revenue. It also said total leased customer power capacity had increased to about 1.1 GW, representing more than $24 billion of potential contracted revenue. Those are company-reported contract values, not BTI forecasts.

This is why the ERCOT process is financially important. Contract announcements create revenue potential, but power, construction, equipment delivery and commissioning determine when that potential becomes billable. Core Scientific’s own filings identify interconnection sequencing, equipment lead times, permitting, construction execution and customer deployment as factors that can delay the conversion of leased capacity into revenue.

For investors, the positive read-through is that Core Scientific has cleared additional regulatory and technical milestones across three strategically important Texas campuses. Denton appears comparatively de-risked from Batch Zero, while Pecos and Hunt have advanced far enough to receive conditional classifications and, in Hunt’s case, to proceed with construction and long-lead equipment commitments.

The risk is that “conditional approval” is not the same as final capacity allocation. The next material datapoint is ERCOT’s completion of Batch Zero and the resulting allocation and ramp schedules, particularly at Pecos and Hunt. If those outcomes align with Core Scientific’s development plans, the Texas portfolio could support a larger and more diversified high-density colocation business. If timing or allocation is constrained, revenue conversion could shift further out even if underlying AI demand remains strong.

BTI’s bottom line: the September 10 filing is a constructive infrastructure milestone, not a final all-clear. Core Scientific is building an increasingly power-rich AI data center platform, and the Texas approvals reduce execution risk at key sites. But investors should track the final ERCOT allocations and commissioning timelines rather than treating the full 1.4 GW discussed in the filing as immediately available capacity.