Natural Gas Capacity proposals in the U.S. have expanded quickly in 2026, with data centers now cited as a direct driver for new gas-fired generation and related pipeline work. The available figures point to a sharp increase in planned capacity, but they also show why the regulatory question is not only whether projects can be built faster. The harder issue is whether power plant, pipeline, air-permit, and community-impact reviews can keep pace with load growth claims that remain project-specific and, in some cases, dependent on data-center buildout schedules.
Several estimates from energy and environmental trackers differ in scope, which matters for interpretation. One early August 2026 estimate from Global Energy Monitor put U.S. natural gas-fired capacity under development to directly serve data centers at 189 gigawatts, up from 97 gigawatts at the end of 2025. The Environmental Integrity Project reported on July 1, 2026 that at least 74 gas-fired power plants, including 71 new builds and 3 expansions, had been proposed exclusively for data centers, totaling about 143 gigawatts. Cleanview’s September 2026 project tracker described a broader universe: 877 natural gas power plants in development across the U.S., representing about 271,714 megawatts of planned capacity.
Those numbers should not be read as guaranteed additions to the grid. Projects can be delayed, downsized, denied, withdrawn, or redesigned. Still, the scale of the queue is large enough to affect air permitting, interconnection planning, pipeline approvals, and local siting decisions. Related infrastructure analysis from Natewin has tracked similar pressure points where power demand, construction timing, and regulatory review intersect.
Why Natural Gas Capacity Is Rising
Natural Gas Capacity And Data Centers
The data-center connection is the technical distinction in the 2026 buildout. These plants are not being discussed only as generic grid resources. In the Environmental Integrity Project count, the 74 identified gas-fired projects were proposed to power data centers exclusively. That siting logic changes the risk profile. A utility-scale plant built for broad grid service is assessed against system demand, dispatch needs, and ratepayer impacts. A plant built for a specific data-center load may hinge on private load forecasts, campus construction schedules, dedicated transmission arrangements, and contracts that are not always visible in public filings.
The Natural Gas Capacity tied to data centers also raises reliability questions that are narrower than the public debate often suggests. Gas turbines can provide dispatchable power, but dispatchability is not the same as fuel assurance, pipeline deliverability, cooling-water availability, or emissions compliance during high-load periods. If a data-center campus expects high uptime and steady electricity supply, the supporting generation plan must account for gas transport constraints, maintenance outages, extreme weather, and backup systems. Those details determine whether a proposed plant actually reduces grid stress or shifts it to another part of the energy system.
Pipeline And Plant Dependencies
Gas-fired generation capacity is also pipeline-dependent. The January 7, 2026 federal notice for the Texas-Arkansas Power Project described a proposal to transport up to 400,000 dekatherms per day of natural gas to supply existing and newly proposed gas-fired electric generation in northern Texas and Arkansas, with environmental assessments ongoing. That type of linked project shows why plant approvals and pipeline approvals cannot be treated as separate technical questions. A generating unit may clear one permitting stage while its fuel-delivery system remains under review, or a pipeline may advance before the full emissions burden of the dependent generation fleet is resolved.
Washington state’s TransAlta coal-to-gas transition proposal shows another variant. The Centralia Power Plant project in Lewis County sought a draft air permit and environmental review to shift one coal unit to natural gas and included a 1,500-foot, 12-inch diameter gas pipeline. As of September 4, 2026, the public comment period was scheduled to run through September 15, 2026. The project is not the same as a new data-center plant, but it illustrates the same permitting issue: a fuel switch can reduce some coal-related emissions while still requiring analysis of gas combustion, pipeline safety, methane-related impacts, and local air quality.
What The Permitting Shift Does
Blanket Certificates And Case Review
Federal permitting is moving at the same time as project volume is rising. On May 21, 2026, the Federal Energy Regulatory Commission issued a Notice of Proposed Rulemaking on blanket certificates for natural gas infrastructure. The proposal would expand the types and sizes of projects that can proceed without case-by-case authorization, increase cost thresholds, and loosen some landowner-notification and environmental review requirements, according to FERC’s notice.
The technical effect of a blanket-certificate reform depends on which projects qualify. Smaller pipeline upgrades, replacements, compression changes, or related facilities can be materially different from greenfield buildouts. A faster process can reduce administrative delay for routine work, especially where infrastructure is already disturbed. The risk is that a series of smaller authorizations can produce cumulative effects that are harder for local stakeholders to evaluate if review is fragmented.
For regulators, the central task is separating routine capacity work from projects that change emissions, land use, safety exposure, or community burden in a meaningful way. That requires clear thresholds and enough public information to test developer claims. Cost thresholds alone are an imperfect proxy because a lower-cost project can still affect sensitive land, disadvantaged communities, or constrained pipeline corridors.
- Project scope: whether the filing covers maintenance, expansion, new compression, or new pipeline segments.
- Load linkage: whether the gas supply is tied to specific power plants or data-center campuses.
- Cumulative effects: whether multiple small projects support a larger generation buildout.
- Public notice: whether affected landowners and communities have enough time and information to respond.
Emissions Controls And Community Exposure
EPA Standards In The Plant Design File
Air regulation is the second major constraint. The EPA is proposing greenhouse gas emission standards and guidelines under Section 111 of the Clean Air Act for new, modified, and reconstructed fossil fuel-fired electric generating units. The agency’s supporting material refers to approaches that include carbon capture, hydrogen co-firing, and low-greenhouse-gas alternatives, as described in the EPA supporting information.
Those technologies are not interchangeable from an engineering or permitting standpoint. Carbon capture adds equipment, energy demand, water and solvent management questions, carbon dioxide transport or storage requirements, and long-term monitoring needs. Hydrogen co-firing depends on fuel availability, burner compatibility, pipeline specifications, and emissions behavior that can vary by blend and turbine configuration. Low-greenhouse-gas alternatives can reduce the need for gas generation in some cases, but they require site-specific analysis of availability, dispatch, transmission, and storage.
The Environmental Integrity Project estimated that the 74 data-center-linked gas plant proposals could emit roughly 662 million tons of greenhouse gases per year if built. That figure is conditional, not a measured outcome. It depends on whether projects proceed, how often they run, what controls are installed, how efficiently they operate, and how emissions are counted. Even with that uncertainty, the estimate is large enough to make emissions control assumptions central to project review rather than a secondary filing issue.
Local pollutants also remain relevant. The research record identifies expected emissions of nitrogen oxides and fine particulate matter from many proposed projects, with health links including asthma and cardiovascular effects. These impacts are not distributed evenly. The Environmental Integrity Project reported that many proposed data-center-tied gas plants would be built in lower-income or disproportionately affected communities. That does not determine the legal outcome of any individual permit, but it raises the standard for transparent dispersion modeling, cumulative exposure analysis, and public participation.
Natural Gas Capacity Projects Under Review
The state distribution of proposed data-center power plants is uneven. The Environmental Integrity Project count placed 32 of the 74 proposed plants in Texas, more than any other state. Other clusters appeared in the Ohio River Valley, including 10 in Ohio, 6 in western Pennsylvania, and 4 in West Virginia. This concentration matters because regional gas supply, grid congestion, air quality baselines, and permitting capacity vary by state and county.
For developers, the near-term challenge is proving that a plant and its fuel path can be permitted, financed, connected, cooled, and operated within the timeline claimed by the associated data-center project. For regulators, the challenge is maintaining technical review depth while facing larger application volumes and pressure for faster approvals. For communities, the key issue is whether project documents describe not only stack emissions, but also pipelines, compressors, backup generation, water use, construction traffic, and cumulative pollution exposure.
Natural Gas Capacity growth linked to data centers is therefore best understood as an infrastructure coordination problem rather than a single fuel-choice debate. Gas plants may be proposed as a response to high-load digital infrastructure, but each project still has to clear engineering, environmental, and legal tests. The evidence available as of September 4, 2026 supports a cautious reading: the pipeline of proposals is large, the permitting rules are in flux, and the environmental impacts depend on site-specific design and operating choices that are not yet settled for many projects.
