Prime West Texas and coastal sites are saturated, transmission corridors are congested, and curtailment risk is suppressing new investment

Decision Lens

The core tension is structural: data center operators need firm, always-on power, but the grid design that shaped Texas for 25 years rewarded cheap intermittent generation over dispatchable capacity. Gas projects in the ERCOT interconnection queue have grown more than 400% in three years, while ERCOT data registers roughly 360,000 MW of prospective data center demand — more than four times the grid’s record peak. If your Texas portfolio is anchored to wind PPAs or 24/7 CFE commitments, the generation mix you planned against is no longer the mix being built.

90-Second Brief

Now, gas generation has overtaken wind in the ERCOT interconnection queue for the first time since January 2016, driven primarily by data center demand for dispatchable power. Gas projects grew from 12,500 MW in March 2023 to nearly 64,000 MW by April 2026, a pace wind cannot match. ERCOT CEO Pablo Vegas has stated publicly that materializing data center loads will require substantially more gas generation to maintain grid reliability. Texas state policy is reinforcing this shift, with the Texas Energy Fund backstopping approximately 9,000 MW of dispatchable projects via low-interest loans.

What’s Actually Happening

The mechanism is a demand-pull effect, not a supply-side technology story. Data centers require continuous, schedulable power, which intermittent wind and solar cannot guarantee without storage. That operational requirement has made gas the preferred pairing for large load commitments — developers are actively speccing gas capacity to support co-location or behind-the-fence supply agreements.

Gas turbine supply chain constraints complicate delivery timelines. Mainstream turbine orders placed today reportedly face delivery dates extending to 2031, creating a hardware bottleneck that mirrors the transformer problem familiar to anyone managing large-scale grid interconnection. Some developers are bypassing this by aggregating smaller industrial generators — El Paso Electric’s plan to cluster 813 small gas units to produce 366 MW for a Meta data center campus is one confirmed example — while others, like xAI in Mississippi, have revived dormant plants entirely.

Wind, meanwhile, faces a structural ceiling. Prime West Texas and coastal sites are saturated, transmission corridors are congested, and curtailment risk is suppressing new investment. Solar costs dropped to roughly $70/MWh in 2024 from around $160/MWh in 2010, pulling clean-energy capital toward solar-plus-storage rather than wind. Federal permit uncertainty around turbine structures adds a further financing deterrent.

Why It Matters for Global Heads of Data Center Energy?

For portfolio managers with Texas exposure, this is a planning assumption that needs revision. The 360,000 MW of queued data center demand will not all materialize — LBNL data places ERCOT’s queue completion rate at 22%, the highest in the nation — but even a fraction of that load coming online will stress a grid that is not yet built to serve it. ERCOT CEO Pablo Vegas has told the Texas Senate directly that balancing that load requires more gas, not less.

That shapes three near-term decisions. First, any wind PPA under negotiation in Texas carries elevated basis risk given transmission saturation and curtailment exposure — structuring language around curtailment compensation deserves review. Second, 24/7 CFE commitments backed by Texas wind may not perform as modeled if new generation continues shifting toward gas; CFE matching strategies may require geographic diversification outside ERCOT. Third, operators planning new Texas campuses should model gas offtake or co-location scenarios as primary, not contingency — the queue signals that counterparties are pricing dispatchability at a premium.

The Forward View

The Texas Energy Fund’s approximately 9,000 MW of loan-backed projects represents the most predictable tranche of new gas capacity; those are the projects most likely to survive the 22% completion rate filter. Operators looking to lock in long-term firm power in Texas should track which TEF-supported projects are advancing through permitting and which data center anchor tenants they are connected to.

ERCOT is simultaneously advancing plans for three new 765 kV transmission lines, which would relieve wind curtailment and potentially reset wind economics by the late 2020s. If those lines are built on schedule, the current gas-dominant queue dynamic could partially reverse — but transmission infrastructure buildouts in Texas have historically slipped. The more durable forward signal is that gas will anchor Texas dispatchable capacity for the next planning cycle, and any decarbonization strategy for Texas operations will need to treat that as a structural baseline rather than a transitional state.

What We’re Uncertain About?

  • Actual data center demand realization rate. The 360,000 MW figure represents queued intentions, not committed load. What fraction converts to operating load — and on what timeline — is unknown. Resolution requires tracking actual interconnection agreements signed, not queue entries.

  • Gas turbine delivery timelines vs. TEF project milestones. If mainstream turbines don’t arrive until 2031, near-term TEF projects may face the same equipment bottleneck driving unconventional solutions like generator clustering. Whether this creates a gas capacity gap in the 2027–2029 window is unresolved.

  • Wind recovery potential post-transmission upgrade. ERCOT’s planned 765 kV lines could reopen wind economics, but neither a completion date nor a financing structure has been confirmed in available evidence. The impact on future queue composition depends on execution.

  • Federal permitting risk duration. FAA permit uncertainty for wind turbines above 200 feet is currently a financing deterrent, but the scope and longevity of the policy shift remain unclear. A change in administrative posture could accelerate or reverse wind investment sentiment quickly.

One Question to Bring to Your Team

Given that gas is now the dominant dispatchable capacity being built in Texas, what is our current exposure to wind curtailment in existing Texas PPAs, and have we stress-tested our 24/7 CFE matching assumptions against a scenario where new gas — not renewables — sets the marginal generation profile for the next decade?

Sources

  • Texastribune — Gas power making a comeback in Texas’ grid connection queue (Link)

Tags