Planned on-site gas generation for U.S. data centers nearly doubled in the first half of 2026, jumping from 97 GW to 189 GW, while overall U.S. gas development rose 50% to 378 GW – yet only 14% of those proposed projects have started construction, signaling a massive speculative pipeline that could lock in emissions and raise power costs if even a fraction materializes.
Data Center Demand Drives Unprecedented Gas Pipeline Expansion
Global Energy Monitor’s mid-2026 update reveals a staggering acceleration in gas-fired capacity proposed specifically for data center campuses. The 92 GW increase in just six months exceeds the total installed generating capacity of many mid-sized European economies. Texas alone accounts for 122 GW of the national 378 GW gas development queue – a figure larger than the entire gas fleet of any other country except the United States itself. Ohio, Pennsylvania, and West Virginia also posted significant gains, reflecting the geographic concentration of new hyperscale campuses in regions with abundant gas supply and permissive permitting.
The scale of this queue is without recent precedent. For context, total U.S. natural gas capacity additions averaged roughly 10-15 GW per year over the past decade. A 378 GW development pipeline represents roughly 25-30 years of historical build rates compressed into a single planning horizon. That compression is driven almost entirely by the projected electricity appetite of AI training and inference workloads, which hyperscalers have signaled could require gigawatt-scale, 24/7 power at individual sites within the next five years.
Yet the construction-start rate tells a different story. At 14%, the vast majority of this capacity exists only in interconnection queues, permit applications, and press releases. Jenny Martos of Global Energy Monitor emphasized the opacity: “It is nearly impossible nowadays to guess what is a pie in the sky proposal, and what has a real chance of getting built.” That uncertainty is not academic – it directly affects turbine procurement, gas supply contracting, and ratepayer exposure.
Cross-Cutting Dynamics: Turbine Supply Chains, Emissions Lock-In, and the Clean Energy Collision
The projects that do advance face a seller’s market for heavy-duty gas turbines. Global OEMs – GE Vernova, Siemens Energy, Mitsubishi Power – are already reporting multi-year backlogs for H-class and J-class units. If even 20% of the 378 GW queue (roughly 75 GW) reaches firm order status, it would absorb several years of global turbine production capacity. That points to sustained premium pricing for equipment and extended lead times, which in turn raises the levelized cost of electricity for any gas plant commissioned after 2028.
Emissions lock-in is the parallel risk. A combined-cycle plant built today operates for 30-40 years. At current U.S. grid average emission factors, each GW of new gas adds roughly 3-4 million metric tons of CO₂ annually if run at baseload. Multiply that by a plausible 50-75 GW of actual builds, and the incremental emissions are on the order of 150-300 million metric tons per year – comparable to the total annual emissions of a mid-sized industrialized nation. That trajectory collides directly with state-level clean electricity standards (California, New York, Washington) and federal 2035 power-sector decarbonization targets.
Electricity price impacts are already visible in forward curves. PJM and ERCOT capacity auction clearing prices have risen sharply over the past 18 months, driven partly by load growth expectations and partly by the retirement of coal and nuclear units without firm replacement. The gas queue, if realized, would eventually moderate energy prices by adding supply – but in the near term, the scramble for turbines and gas firm transport contracts is pushing up capital costs that regulators will ultimately recover from ratepayers.
Meanwhile, the storage and renewables industries face a crowded-out interconnection queue. In PJM, gas projects now compete with hundreds of GW of solar, wind, and battery projects for limited interconnection study slots. The “first-come, first-served” queue reform efforts are moving slowly, meaning speculative gas applications can block or delay cleaner resources that are further along in development.
Who This Affects
- Utility resource planners: Must model a wide range of gas-build scenarios (0-75 GW actualized) in integrated resource plans, stress-testing portfolio costs against turbine price escalation and potential carbon regulation.
- Storage and renewable developers: Face longer interconnection timelines and higher upgrade costs as gas projects consume queue capacity; should prioritize sites with existing interconnection rights or co-location agreements.
- State public utility commissions: Need to establish clear prudence standards for gas capex recovery before ratepayers bear the risk of stranded assets if data center load growth underperforms or clean alternatives underbid gas.
- Grid operators (ISOs/RTOs): Must update reliability assessments to reflect the low probability of most queued gas capacity materializing, while preparing for localized gas-heavy pockets that could create congestion and voltage stability issues.
- Institutional investors: Should scrutinize sponsor equity commitments and offtake structures – projects without long-term power purchase agreements from creditworthy hyperscalers carry high merchant risk in a decarbonizing market.
What to Watch Next
- Construction start rate quarterly tracking: A sustained rise above 20% would signal the queue is converting to steel; a drop below 10% would confirm most projects are speculative.
- Turbine OEM order backlog disclosures: Quarterly earnings calls from GE Vernova, Siemens Energy, and Mitsubishi Power will reveal firm order intake versus soft commitments.
- State permitting decisions in Texas, Ohio, and Pennsylvania: Air permit approvals and water withdrawal authorizations are the first hard gates that filter speculative proposals.
- Hyperscaler power procurement announcements: New long-term PPAs for gas-fired output (or gas-backed “firm” renewable products) would validate demand; silence suggests load forecasts are softening.
- FERC Order 2023 implementation and queue reform outcomes: Cluster study results due in late 2026 will show how many gas projects survive the new financial commitment requirements.
Bottom line: The gas development queue has ballooned into a speculative hedge against AI load uncertainty – not a committed build plan. The 14% construction-start rate is the single most important number: it means 86% of the 378 GW is still optional, and the industry’s next moves will be written in turbine contracts and air permits, not press releases.
Read the full report at Energy Central
Note: facts and figures attributed above to reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply