Utility-scale solar and batteries made up most new U.S. power plant builds in early 2026, and that shifts how operators should buy capacity

A quiet operational reality is hiding behind the U.S. power sector’s loudest headlines: most new plants that actually entered service in the first half of 2026 were utility-scale solar arrays and battery projects, not big thermal builds.

Inside Climate News reviewed U.S. Energy Information Administration data and counted 368 utility-scale plants that began operating from January through June 2026. The list is dominated by solar on volume and batteries on both volume and capacity, a mix that changes what “new capacity” means for enterprise load planning and energy procurement.

What the first-half 2026 build list actually says

The Inside Climate News analysis, based on EIA’s monthly generator inventory dataset (EIA-860M), puts hard numbers on the mix. By project count, utility-scale solar led with 207 plants coming online in the first six months of 2026, followed by batteries with 95, according to Inside Climate News. Natural gas additions showed up mostly as peaking and smaller units: 22 combustion turbines and 21 internal combustion engine plants, with just 3 combined-cycle gas plants entering service in that window, Inside Climate News reported.

On a nameplate megawatt basis, utility-scale solar again led, with 11,458 MW added January through June 2026, followed by 8,207 MW of batteries, Inside Climate News wrote. Onshore wind added 5,473 MW, but that figure is heavily influenced by one long-gestating development: SunZia Wind South and SunZia Wind North in New Mexico, which Inside Climate News identified as the two largest new projects and described as the largest wind farms in the country, totaling 3,650 MW.

For operators, that “top-heavy” dynamic matters. Big plants make for simple talking points, but they can be poor predictors of what will be available to site a new data center, electrify a process line, or negotiate a multi-site power purchase agreement. Inside Climate News explicitly cautioned that SunZia is best read as an echo of earlier onshore wind development because U.S. onshore wind development has slowed for multiple reasons, rather than as a new wave.

The operational catch: nameplate MW is not the same thing as usable supply

Inside Climate News also flagged the core caveat procurement teams already wrestle with: technologies differ in how and when they deliver. Solar doesn’t produce at night, and some gas plants run only during a handful of peak-demand days or weeks. That’s not an academic point when an enterprise is trying to match a 24/7 load profile, meet internal reliability targets, or pass a board-level review of energy risk.

The Ars Technica forum discussion linked to the same reporting reinforces why “capacity” needs qualifiers in any internal briefing. One commenter argued that capacity factor is the missing term in a lot of public discourse, citing typical figures of about 25% for grid-scale solar and about 35% for onshore wind, and noting that many battery projects are designed around peak-hour needs with a nominal four-hour duration. The forum is not a primary dataset, but it reflects the operator-level questions that come up as solar-plus-storage becomes the default new-build pattern.

Taken together, the EIA-860M numbers and the capacity-factor/duration lens point to a planning implication: if a facility’s risk model assumes that “X MW of new builds” translates into “X MW of firm supply,” it will overestimate coverage during evening ramps, extended cold snaps, or multi-day heat events unless the contract stack explicitly accounts for storage duration, curtailment, and backup supply.

Where this shows up first: Texas, PJM, and the replicable build pattern

Geography is the other operator’s filter. Inside Climate News reported that three states, Texas, New Mexico, and Arizona, accounted for about half of the country’s new generating capacity in the first half of 2026. Texas led the list, and its mix leaned more heavily on batteries and gas than other states, according to Inside Climate News.

The Texas breakdown is unusually concrete for anyone doing ERCOT-adjacent capacity planning: batteries led with 3,067 MW, followed by 2,311 MW of utility-scale solar, then 1,154 MW of natural gas combustion turbines or internal combustion engines, Inside Climate News wrote. That blend suggests that new “reliability” tools in the state are increasingly dispatch rules, interconnection timing, and market design for storage, rather than fuel contracting alone.

Meanwhile, the combined-cycle gas story in the EIA-860M first-half list is a timing story. Inside Climate News pointed to the Trumbull Energy Center in Ohio, a 900 MW combined-cycle natural gas plant, as one of the largest new plants to enter service and tied it to an expansion of gas power in the PJM Interconnection region. Yet the broader count shows just three combined-cycle plants entered operation from January through June 2026, and Inside Climate News noted that more of the ‘high tide’ of new gas plants is expected later in the decade, not in the immediate half-year window.

For procurement and operations leaders, the repeatable signal is neither the occasional mega-wind project nor the occasional large combined-cycle plant. It’s the steady churn of mid-sized utility-scale solar projects, often paired with batteries. Energy Innovation senior fellow Eric Gimon told Inside Climate News that these projects are easier to site and build, with more diverse location options, and that’s why they’re appearing almost everywhere.

Questions energy buyers should put into specs and site models now

• When evaluating PPAs or retail supply offers that reference new solar and storage builds, what is the contracted product: nameplate MW, delivered MWh, or a time-matched shape (for example, evening coverage)? Ask bidders to show hour-by-hour delivery assumptions and curtailment treatment using the same load profile used in internal planning.
• For projects in Texas or other storage-heavy build regions, what battery duration and dispatch constraints are being assumed? Inside Climate News’ first-half Texas capacity mix (3,067 MW batteries vs. 2,311 MW solar) makes four-hour storage a common baseline, but enterprise risk models may need stress tests for multi-day events.
• For PJM-area expansions, separate “plant entered service” from “capacity is deliverable to the node.” The EIA-860M list shows large plants can come online while the broader combined-cycle wave is still pending. Interconnection upgrades, transmission constraints, and congestion exposure should be treated as first-order commercial terms, not footnotes.