Texas’s 438GW queue: when does a power request become revenue?

ERCOT said on June 18 that it was tracking more than 438GW of large-load connection requests, with nearly 89% from data centers. The approved Batch Zero process groups qualified projects of at least 75MW to assess where and how much load the grid can reliably support. The queue measures development requests, not committed future consumption. ^[1]

For data-center investors, the question is which projects can obtain credible connection scope, cost and timing sooner. Batch studies may reduce repetitive analysis, but do not immediately supply transformers, lines or generation. Revenue still requires engineering, finance and customer contracts to align.

The queue cannot be multiplied by rent to size a market

Requests can span years, locations and development stages. Preserving a location option differs from a customer committing to long-term rent. The announcement does not establish contract quality, deduplication or delivery probability for the entire queue, so the total cannot imply firm equipment orders. ^[1]

The nearly 89% data-center share shows the sector's importance to the workload, not its eventual completion rate. Multiplying requests by that share still produces a request measure. Qualification, completed studies, service arrangements and construction each need their own denominator.

Equal capacity has different value at different locations. Network headroom, upgrades and stability requirements affect cost and timing. Cheap land requiring extensive transmission work may collect cash later than an expensive but better-connected site.

Pipeline valuation should distinguish early development options from capacity supported by customers and defined electrical work. Observable conditions explain the difference better than an invented uniform success probability.

Batch analysis improves comparison without completing construction

Sequential studies can require repeated calculations as more applications enter. Batch Zero groups qualified large loads to assess aggregate demand, allocate available capacity and identify upgrades under shared assumptions. ^[1]

Qualification makes a project eligible for analysis; it does not guarantee commercial energization. Design, service agreements, procurement, construction and tests can remain after a study. Entering or completing analysis should not be described as immediately usable power.

Load characteristics, ramp and behavior during disturbances affect reliability assessment. Repeated design changes can make earlier results less applicable. This is engineering reasoning; exact submission duties and deadlines require the governing document version.

Developers with complete designs, equipment coordination, customers and funding may benefit more than those controlling land alone. Better rules improve information, while execution capability remains a separate requirement.

A six-month delay affects more than reported rent

Assume, only for illustration, 100MW of IT capacity and monthly rent of $150 per kW. Full-year billing produces $180 million. At an average billable share of 60%, revenue is $108 million. Announced capacity cannot substitute for the billable fraction.

Without advance customer payment, a six-month shift in full-capacity lease commencement could defer $90 million of rent. It does not necessarily destroy that revenue: extensions and payment terms matter. Deferred cash, lost revenue and reduced present value are distinct effects.

If $1 billion is already invested at an 8% annual capital cost, an extra six months represents about $40 million of carrying cost. This should not mechanically be added to deferred rent as a net loss. Avoided operating expense, compensation and the timing of later receipts must also be considered.

Equity may face a cash call even when the long-term contract remains profitable. Construction and interest continue before billing. A temporary funding gap can force unattractive financing, making liquidity and draw conditions as important as the target energization date.

Onsite generation and curtailment involve an exchange

The June announcement provides paths for customers using onsite generation and for those accepting curtailment during local transmission constraints. Truly islanded facilities generally fall outside ERCOT interconnection, although PUCT registration may still apply. Flexibility does not remove all regulatory or operating duties. ^[1]

Onsite supply transfers some dependency from transmission to fuel, generating equipment, maintenance and environmental requirements. Continuous cloud service also requires redundancy and recovery. Nameplate capacity alone does not establish hourly availability under a customer SLA.

Curtailment may improve connection options while changing the service that can be sold. Deferrable batches differ from real-time work. Moving tasks elsewhere requires data, network, software and spare resources; interruption cannot be assumed commercially free.

As an illustration, losing 5% of saleable GPU-hours at unchanged prices would reduce the relevant volume revenue by 5% if every lost hour would otherwise have sold. Idle hours could make the effect smaller; compensation and lost customers could make it larger. Compare access benefits with lost receipts and added operating costs.

Upgrade responsibility determines who retains the return

A clearer study can expose previously uncertain infrastructure needs. Dedicated substations, shared-line allocation and payment timing affect capital per unit of capacity. The reform does not justify assuming a uniform cost allocation; project documents and applicable rules are needed.

A customer contract can help fund upgrades, but pricing must cover them. Fixing a low rent before costs are known can leave weak returns even after successful delivery. Reasonable adjustment terms can instead share the benefit of earlier connection.

Lenders will assess whether electrical conditions are sufficiently defined and who supplies cash if upgrades delay. Wider spreads, lower leverage and larger reserves can all reflect uncertainty. These conditions affect equity needs beyond the nominal interest rate.

Supplier receipts arrive in stages: design work, executable electrical orders and later server deployment. They cannot all be placed in the quarter of a procedural announcement, nor allocated to named suppliers from aggregate queue MW.

How much service flexibility is earlier access worth?

Connection options should be compared on saleable service. Two 100MW facilities with different curtailment and reliability conditions may command different prices. Earlier access reduces waiting, while customer discounts, backup and compensation can absorb the benefit.

In a purely illustrative case, a flexible arrangement earns $100 million of annual gross rent but incurs an assumed $10 million for interruption, substitute resources and discounts, leaving $90 million. An otherwise identical option begins a year later. Earlier net receipts have value, but extra generation or migration capital could reverse the comparison.

The $10 million is a scenario expense, not a forecast of curtailment frequency or probability. A real model requires contractual conditions, acceptable downtime, compensation limits and replacement-resource prices. Without them, reliability cannot be precisely valued.

Workload portability matters. Distributed data and cross-site scheduling can make flexibility relatively inexpensive; latency and data-residency constraints can make it difficult. Customer mix therefore shapes the power arrangement rather than simply following the lowest tariff.

Phased capacity can let customers start earlier and expand with grid work. Reserved interfaces and repeated construction add design considerations. It may reduce initial cash needs while extending recovery of the full investment.

Project advantage comes from matching power, engineering and workload. The longest application pipeline is not necessarily the best service inventory. Clearly priced constraints and reliable collections turn connection rights into economic value.

Better rules need operating outcomes

The favorable case is less repeated analysis, earlier clarity on upgrades and reliable power at a sensible cost. Construction, customer installation and payment must follow before process improvements become investment returns. Study completions and energized capacity should be tracked separately.

The adverse case is faster analysis but slow transformers, lines, permissions or construction. Joint analysis may also reveal greater shared-resource costs than developers budgeted. Transparency improves the ability to assess projects, not necessarily every project's result.

Compare developers through qualified phases, upgrade obligations, executable power schedules and billing. Companies that price engineering constraints into contracts and finance the timing gap have a stronger path from large demand requests to durable cash.