Development could begin as early as 2027, with initial operating capability targeted by — or ahead of — 2030. Formal lease agreements are still in the early stages.

The Army isn’t buying the power plants

Here’s where the deal gets interesting for the power industry.

The Army isn’t paying for construction. The projects will use an Enhanced Use Lease, or EUL, under federal law.

The arrangement allows the Army to lease underutilized land to private companies while retaining ownership of the property. Developers take on the heavy lifting and the heavy capital costs, including financing, design, construction, operation, security and eventual decommissioning.

In exchange, the developers pay rent for the land.

The Army said it prefers that compensation be in the form of “in-kind” improvements, meaning developers could fund infrastructure upgrades at the host installation rather than simply writing a check. It also requires a decommissioning bond to ensure money is available to restore the property when a lease ends.

Power generation meets grid resilience

The strategy isn’t just about keeping the lights on inside the fence line.

The Army’s June solicitation called for commercial generation projects that can serve the grid under normal conditions while providing islanded power to installations during extended outages. In other words: Keep the electrons flowing, even when the grid doesn’t.

That creates an unusual business model: Build generation on military land, sell power into the grid, while maintaining the capability to serve a critical customer when the grid has a bad day. The Army gets access to private capital and infrastructure expertise, the installations get another layer of energy security, and the grid gets potentially more generation and storage.

The solicitation identified seven potential installations and allowed technologies including solar, natural gas, geothermal and nuclear, with nuclear proposals capped at 500 MW per unit. The Army did not include Fort Drum, New York, in its Sept. 18 announcement.

The winners

• Ameresco said its projects at Aberdeen Proving Ground and Picatinny Arsenal in New Jersey could include traditional firm power generation and nuclear technology.

For Ameresco, this includes proposed phased power generation campuses, potentially including firm natural gas and advanced nuclear generation at Aberdeen Proving Ground, and microreactor and small modular reactor technologies at Picatinny Arsenal, subject to final agreements and required approvals.

The company anticipates integrating commercial power generation to add market-based grid capacity while supporting resilient on-site capabilities for Army installations during qualifying grid outages.

“Ameresco is honored to be among the companies selected to support the U.S. Army’s Strategic Capital Initiatives and advance energy resilience across critical military installations,” said Nicole Bulgarino, Ameresco co-president, in a media release. “At Aberdeen Proving Ground in Maryland and Picatinny Arsenal in New Jersey, we look forward to deploying resilient, on-site energy infrastructure that strengthens mission readiness and enhances energy security.”

• Indelible-Bloom is partnering with Bloom Energy on projects at Fort Detrick in Maryland, Letterkenny Army Depot in Pennsylvania, and West Point in New York.

Silicon Valley-based Bloom Energy is known for its fuel cell systems that provide electricity for Fortune 500 customers worldwide, including data centers, semiconductor manufacturing, large utilities, and other commercial and industrial sectors, as well as mission-critical organizations in local communities.

The Army’s statement specifically said the partnership draws on Bloom Energy’s experience supplying reliable power to critical facilities such as hospitals, universities and research laboratories.

This is strategically important because military installations have unusually strong requirements for islanding, resilience and dependable onsite generation. Bloom’s fuel-cell systems are designed around exactly that use case.

• Frontier Power USA said it’ll use the Tobyhanna Army Depot in Pennsylvania for a project that combines American-made battery energy storage with power generation.

Project stakeholders said the partnership reflects Frontier Power USA’s approach to developing integrated energy infrastructure solutions that combine power generation, energy storage, and related technologies to meet critical customers’ needs.

By bringing multiple technologies together into a single resilient platform, FPUSA said it can deliver reliable, secure, and domestically sourced energy solutions that strengthen operational readiness and support broader national energy priorities.

The company also said the approach reflects a broader model of cooperation between the Army and private industry: leasing non-excess, underutilized land to American companies that finance, build, own, and operate the assets at no capital cost to the government. As power demand rises across the United States, projects like this add market-based generation and storage capacity to the grid while giving host installations ready, on-site capability.

“FPUSA is honored to support the U.S. Army and its mission,” said Frontier Power USA CEO Jay Bellows, in a statement. “We are proud to bring private capital and American innovation to Army installations, combining American-made battery energy storage with U.S.-compliant power generation to deliver resilient, mission-ready power at scale.”

Source: Tobyhanna Army Depot.

Who’s building what?

The Army limited eligibility to companies organized under U.S. law, with majority domestic ownership and control and a U.S. place of business.

Infrastructure improvements provided to the Army also must comply with Buy American Act and Davis-Bacon requirements.

These projects still have plenty of runway before they become power plants. Construction cannot begin until required environmental and regulatory reviews are completed, including reviews under the National Environmental Policy Act and applicable clean-air and clean-water requirements, along with federal, state and local permitting.

The announcement comes just over four months after Fervo’s May 2026 initial public offering. Following the successful deployment of Fervo’s EGS technology at Project Red, which has been supplying electricity to the grid since 2023, first power at Cape Station represents a “significant scale-up” of the technology in a greenfield setting, the company said.  

Fervo uses horizontal drilling techniques adapted from the oil and gas industry, allowing multiple wells to be drilled from a single pad. The project is an enhanced geothermal system that produces energy by injecting water into hot subsurface rock formations and then extracting the heated water to generate electricity, rather than depending on naturally occurring underground hot water like traditional geothermal systems. If fully developed, the project will cover approximately 631 acres, including 148 acres on public lands.

“This is a gamechanger for the geothermal industry. It establishes EGS as the defining new power generation technology of our time and we believe it shows that the commercial and technical maturity of EGS is ready to meet the urgent need for reliable, clean power,” said Tim Latimer, CEO and Co-Founder of Fervo Energy. “I could not be more proud of the years of hard work from Fervo’s employees, investors, suppliers, customers, and partners, which brought us to this moment. We are just getting started.” 

Cape Station Phase I is an approximately 100 MW installation comprising three 33-MW GeoBlocks, the first of which has achieved first power and is expected to reach its contractual commercial operations date (COD) by October 1, 2026. Fervo will continue ramping up the first GeoBlock as it brings the full well system online while commissioning the remaining two GeoBlocks, which are expected to reach contractual COD by January 1, 2027. The next phase, an additional 400 MW, is already under construction, with an expected COD in 2028. Fully contracted offtake at Cape Station sits at approximately 900 MW, enough to power the equivalent of nearly 1 million U.S. homes annually. 

Utah is home to immense geothermal potential. Researchers estimate that the southwest portion of the state contains more than 10 GW of high-quality geothermal reserves. Additionally, Cape Station will benefit from the Department of Energy’s Frontier Observatory for Research in Geothermal Energy (FORGE). Over the last several years, FORGE has completed research that has advanced geothermal development in the region.

“Cape Station works because we treated the subsurface like an engineering challenge,” said Jack Norbeck, Chief Technology Officer and Co-Founder of Fervo Energy. “Years of drilling, completion design, subsurface modeling, and flow testing led to this moment, and this is the validation that matters most. First Power is proof the science works.” 

Earlier this month, Fervo Energy announced a record-breaking 396-megawatt (MW) power purchase agreement (PPA) with Google to enable the continued development of the Cape Station EGS GeoCluster. Under the agreement, Google will purchase carbon-free energy designed to serve as a “foundational building block” for a potential data center in Utah. As part of the PPA, Fervo will offer Google an option to expand its offtake by approximately 600 MW, for a total of nearly 1 gigawatt (GW) by June 2030. Final data center plans remain subject to a variety of factors including engineering feasibility, state and local approvals, and commercial conditions, the companies said.

The Fervo-Google partnership began with Project Red, Fervo’s commercial pilot project in Nevada that came online in 2023. That pilot delivers power to the local grid, including Google’s data centers in the state. Following Project Red, Fervo signed a 115 MW PPA with Google and NV Energy in June 2024 that helped pioneer the Clean Transition Tariff, which Google argues enabled it to bring more geothermal energy onto the Nevada grid while insulating customers from the project’s costs.

Additional PPAs at Cape Station include Shell Energy North America, Southern California Edison, and an expanded deal with Clean Power Alliance that adds 18 MW of carbon-free geothermal energy to their existing PPA with Fervo.

“These projects will empower American innovators to unlock the tremendous geothermal resources beneath our feet,” said DOE Under Secretary of Energy Kyle Haustveit. “Under President Trump’s leadership, we’re advancing next-generation geothermal technologies that can lower costs, strengthen American energy dominance, and turn more of our vast domestic geothermal resources into reliable and affordable power.”

Five of the projects will conduct field-scale enhanced geothermal systems (EGS) tests to validate technologies under real-world conditions, while 16 additional projects will conduct exploration drilling to identify and characterize “promising” next-generation geothermal resources. DOE argues that these efforts can help reduce technical and development risk, as well as provide some of the information needed to support future commercial projects and development.

Data generated by these projects will be publicly available through DOE’s Geothermal Data Repository (GDR), giving industry, researchers and other stakeholders access to information from the field tests and geothermal exploration activities.

Pilots and demonstrations

Five projects are conducting field-scale EGS tests at depths and temperatures “appropriate for full-scale geothermal development,” per DOE:

• Fervo Energy Company (Houston, Texas) — Drill, complete, and stimulate EGS wells in Elmore County, Idaho, deploy the first high-temperature three-component geophone array at or above 200°C, and conduct advanced seismic and geochemical analysis in partnership with national labs. 
• Zanskar Geothermal and Minerals (Salt Lake City, Utah) — Design, drill, and hydraulically stimulate a short-lateral horizontal injection well at Lightning Dock Geothermal field in Hidalgo County, New Mexico, to deliver engineered pressure support to an actively producing hydrothermal reservoir and capture heat from impermeable, hot rock in the vicinity of the hydrothermal system.
• AlterG Resources (Salt Lake City, Utah) — Field validate technical performance of EGS at commercially relevant depths in Pershing and Churchill Counties, Nevada, using a deep horizontal EGS well pair and a comparative hydraulic stimulation program testing plug-and-perf and sliding sleeve completions.
• University of Utah (Salt Lake City, Utah) — Design, drill, stimulate, operate, and validate a high‑performance EGS doublet at the Dixie Valley West geothermal field in Nevada, with the aim to demonstrate that engineered reservoirs developed adjacent to existing hydrothermal systems can deliver sustained circulation, enhanced permeability, and commercially relevant thermal output. 
• Quaise Energy (Houston, Texas) — Conduct field-scale EGS demonstration on southern flank of Newberry Volcano in central Oregon, with bottomhole temperatures of 265°–365°C—above the range where most conventional EGS tools and materials have been validated.

Sixteen additional projects are working on exploration drilling to gather the subsurface data needed to characterize and potentially confirm “promising” next-generation and conventional hydrothermal geothermal resources:

• DAVINCI EP LLC (Denver, Colorado) — Drill and test two paired horizontal geothermal confirmation wells in the Brawley (California) Known Geothermal Resource Area to characterize a naturally enhanced hydrothermal reservoir without stimulation, identify parameters to support commercial-scale development of the resource, and provide a replicable framework for other naturally fractured geothermal basins.
• University of Utah (Salt Lake City, Utah) — Conduct targeted drilling and subsurface characterization of a large thermal anomaly in Utah’s Cove Fort geothermal field and use modern drilling, logging, and data integration techniques to better understand how heat and fluids are distributed in carbonate rock formations.
• Zanskar Geothermal and Minerals (Salt Lake City, Utah) — Test and document a confirmation methodology of directionally-drilled, deep wells from a single pad, using near-real-time stochastic model updates and logging-while-drilling to reduce anomaly-to-discovery timeline, and confirm a commercially viable hydrothermal resource at a Basin and Range site that has never been drilled to reservoir depth.
• Fervo Energy Company (Houston, Texas) — Execute a targeted appraisal campaign at a high-priority EGS prospect in Humboldt County, Nevada, drilling to confirm reservoir temperatures of 200–220°C.
• San Ildefonso Services LLC (Santa Fe, New Mexico) — Drill a pre-feasibility geothermal exploratory well in the San Ildefonso Pueblo, New Mexico, on the eastern border of the Jemez Mountains volcanic field in the Rio Grande Rift, and collect basement core and data from injection tests and logging.
• GeoAlaska LLC (Anchorage, Alaska) — Drill, core, log, and test a geothermal confirmation well on the southern flank of Mt. Augustine, a volcanic island in lower Cook Inlet, Alaska, and characterize the subsurface reservoir using a shallower well targeting >150°C and a deeper well targeting >350°C. 
• Hexagon Energy LLC (Charlottesville, Virginia) — Drill an exploration well in Washington’s Wind River Valley, targeting temperatures >150°C, and collect, test, and publish downhole data sufficient to design a commercial-scale geothermal drilling and power generation program.
• GreenFire Energy Inc. (Oklahoma City, Oklahoma) — Confirm and characterize deep geothermal resource potential in Sierra Army Depot (California) by siting and drilling a vertical well targeting 204°C and integrating legacy data, geophysical surveys, drilling, reservoir testing, and modeling.
• TLS Geothermics Corp (Reno, Nevada) — Drill a confirmation well in Nevada targeting >204°C, using polycrystalline diamond compact drill bit technology in faulted crystalline rock, and develop and calibrate a 3D model linking magnetotelluric resistivity to temperature and permeability.
• LiPower Geothermal, LLC (Gloucester, Massachusetts) — Construct two wells, a production well and a reinjection well, to complete a Definitive Feasibility Study on an under-explored geothermal resource in the South Brawley New River area (California).
• INTEK Inc (Arlington, Virginia) — Site and drill two reservoir-depth wells into the geothermal resource at Pisgah Crater (California) and collect data to characterize and confirm the resource, determine power production feasibility, and accelerate opportunities for commercial-scale geothermal power at the site.
• Oriah Geothermal, LLC (Fort Worth, Texas) — Evaluate a next-generation geothermal resource in eastern Oregon through staged drilling of thermal-gradient wells to constrain the thermal regime and refine subsurface targets, followed by a reservoir-depth exploration well, advanced diagnostics, and integrated reservoir modeling.
• XGS Energy Inc. (Palo Alto, California) — Drill deep vertical appraisal well in Socorro County, New Mexico, acquiring downhole temperature, logging, core, and continuous electronic drilling record data to validate subsurface datasets, update thermal and geologic models, and assess resource suitability.
• Raser Power Systems, LLC (Reno, Nevada) — Assess opportunity for increased power generation from Thermo No. 1 geothermal power plant in Utah’s Escalante Desert by drilling an exploratory well in the southern fault intersection zone and conducting wellfield interaction monitoring between the exploration well and an existing well.
• 400C Energy Inc. (Denver, Colorado) — Re-enter and deepen an existing geothermal well in Beaver County, Utah, to determine whether the Salt Cove area contains an extension of the nearby Roosevelt Hot Springs hydrothermal system, a high-temperature next-generation geothermal resource in hot crystalline rock.
• Invenergy Geothermal Development LLC (Chicago, Illinois) — Validate hybrid hydrothermal–EGS potential near Grand View, Idaho, by drilling a deep temperature confirmation well to confirm subsurface temperature profiles, evaluating permeability indicators, and collecting high-resolution geologic and geophysical data.

Originally published in Factor This.

It’s a major licensing step toward what the company said could become the world’s first gas-to-nuclear power plant. The site would combine GE Vernova 7HA.02 gas turbines with GE Vernova Hitachi BWRX-300 small modular reactors in a project designed to transition from natural gas generation to nuclear power.

The latest action places the Maryland-based developer among a small group of companies submitting construction permit applications in the latest wave of U.S. advanced nuclear development.

The company said it is one of just five companies to submit a construction permit application.

RELATED: First ‘gas-plus-nuclear’ plant for data centers, advanced manufacturing is closer to becoming a reality

Blue Energy expects a final investment decision in 2027.

The company said its approach — dubbed the “Blue Way” — uses standardized, prefabricated nuclear components that can be manufactured offsite, transported, and assembled at the project site. Blue Energy said the strategy could trim more than a decade from the traditional nuclear construction timeline while reducing costs and project risk.

That timeline is now facing its first major regulatory test.

Blue Energy seeks early construction approval

The first half of the NRC application involves the project’s first nuclear unit and requests approval to begin limited construction activities.

Those activities would include deep foundation and shaft work for the first BWRX-300 reactor.

The application also seeks approval for the company’s “Blue Bridge” approach, which would allow the company to construct the natural gas portion and balance of plant before installing the nuclear reactors. The filing includes a nearly 300-page environmental report covering the project’s natural gas facilities.

The same balance-of-plant infrastructure would later support the nuclear units.

The NRC previously approved the methodology behind the gas-to-nuclear construction sequence in a topical report submitted by Blue Energy last September, clearing a regulatory hurdle for the construction sequencing strategy that is now included in the company’s permit application.

“This is serious work done by serious people for a serious project,” said Jake Jurewicz, Blue Energy’s CEO and co-founder, in a statement.

The company said its regulatory team, which includes engineers, environmental specialists, financial experts, and other technical advisers, has more than 80 years of combined NRC licensing experience.

The filing also includes information on the company’s proposed financing strategy, in which the project is expected to be project-financed rather than financed through ratepayer-backed mechanisms.

The financing question could prove just as important as the technology itself as developers attempt to bring a new generation of nuclear plants online.

Earlier this month, Blue Energy announced it had retained Société Générale to help structure and arrange project financing for the Texas-based endeavor.

The company has also raised more than $400 million in financing, including strategic investment from Constellation Technology Ventures.

BWRX-300 selected just five months ago

The NRC filing comes only five months after the Blue Energy selected GE Vernova Hitachi’s BWRX-300 as its reactor technology.

The BWRX-300 is a 300-MW-class small modular reactor design that is being pursued for projects in the U.S. and Canada.

Blue Energy also has a project acceleration agreement with GE Vernova and GE Vernova Hitachi covering deployment of the 7HA.02 gas turbines and BWRX-300 reactors.

The company’s next major licensing milestone is already on the calendar. Blue Energy expects to submit the second half of its construction permit application in late-2027, after completing the NRC-required one-calendar-year of site-specific data collection.

If the project moves forward, the Victoria plant would put Blue Energy’s central bet to the test: Build the gas infrastructure first, use it as a bridge to nuclear generation – then rely on standardized manufacturing and construction to shorten the timeline for bringing new nuclear capacity online.

“We look forward to completing the licensing process so we can build this groundbreaking gas-to-nuclear plant and demonstrate the world’s first project-financed nuclear power product,” Jurewicz said.

Other nuclear innovation projects underway across the US

• TerraPower broke ground in Kemmerer, Wyoming, in April on its Natrium project, which the company says would become the first utility-scale advanced nuclear power plant in the U.S. if completed on schedule.
  
  Kemmerer-1 is the first project of its kind to receive Nuclear Regulatory Commission approval of a commercial nuclear reactor in nearly a decade, regulators said, and the first approval for a commercial non–light water reactor in more than 40 years. 
  
  The project is being driven through a public-private partnership with the U.S. Department of Energy’s Advanced Reactor Demonstration Program and is planned to feature a 345-megawatt sodium-cooled fast reactor with an integrated molten salt-based energy storage system.
  
  TerraPower had picked Bechtel as its EPC partner, but the relationship recently hit a snag. The companies announced this month they couldn’t agree on the next phase of the project.
  
  A TerraPower spokesperson said the disagreement, however, wouldn’t impact the Kemmerer-1 project timeline, which currently calls for construction completion in 2030 and commercial operation by 2031.
  
  TerraPower said it will rely on subcontractor teams and other EPC support services from commercial-grade firms that have been working on-site.

• The Nuclear Regulatory Commission is poised to approve a construction permit tied to the Clinch River Nuclear site project underway with the Tennessee Valley Authority this Fall.
  
  TVA now has NRC clearance to build a BWRX-300 reactor, the same reactor technology Blue Energy is eyeing.
  
  If plans move along as expected, TVA would be the first U.S. entity granted a construction permit for a BWRX-300.

• Dow and X-Energy are partnering to deploy a first-of-its-kind grid-scale advanced nuclear reactor project at Dow’s Seadrift Operations manufacturing site in Seadrift, Texas.
  
  The construction permit application was announced in early 2025 and covers a four-module, 320-MWe Xe-100 high-temperature gas-cooled plant.
  
  The NRC aims to finish its final safety evaluation by November.

• Holtec International is advancing plans to develop Pioneer Units 1 and 2, two SMR-300 advanced pressurized light-water reactors at the Palisades Energy Center in Michigan.
  
  The two units are designed to provide about 680 megawatts of additional electric generating capacity.
  
  The project is being developed with partners, including Hyundai Engineering & Construction.
  
  Holtec is targeting commissioning in the early 2030s, pending regulatory reviews and approvals.
  
  In recent weeks, the NRC approved an exemption allowing Holtec to install permanent support-of-excavation and cutoff-wall systems before the agency issues a Limited Work Authorization for the project.
  
  The approval also advances planned ground-improvement work at Palisades, located in Michigan’s Covert Township.
  
  The exemption is part of Holtec’s phased construction-permit application and allows certain early construction activities to move forward while the NRC continues reviewing the broader application.
  
  Blue Energy is pursuing a similar phased licensing approach, with plans for its own construction-permit application that includes a request for a Limited Work Authorization.

Bechtel is stepping away — but not immediately

Bechtel plans to remain involved long enough to complete the Natrium sodium test and fill facility, and help with an orderly transition, according to the companies.

Bechtel’s work on the project has included supporting TerraPower through the Nuclear Regulatory Commission’s construction permit process, completing the exterior enclosure of the sodium test and fill facility, placing major equipment procurements and beginning foundation work for Kemmerer Unit 1.

That work was performed under an earlier phase of the contract, before the full EPC scope was determined.

Now, that next phase is up for grabs.

About 200 jobs could be affected

The break-up also has a workforce component.

Bechtel filed a Worker Adjustment and Retraining Notification, often called a WARN notice, in Virginia earlier this month, indicating that about 200 jobs could be affected as its footprint on the Natrium project winds down.

The reductions are expected to begin on or around Nov. 16.

World Nuclear News reported that Bechtel does not expect the WARN notices to necessarily translate into actual layoffs.

Bechtel said it is working to find other assignments for impacted employees, including positions on other nuclear projects.

Non-nuclear activities already under construction

The timing of the Bechtel-TerraPower break-up is noteworthy.

TerraPower started construction at the Kemmerer site earlier this year after the NRC approved its construction permit in March.

That approval was historic: It was the NRC’s first construction permit for a commercial reactor in nearly a decade and the first for a commercial non-light-water reactor in more than 40 years, according to ANS.

The Natrium plant is designed around a 345-MWe sodium-cooled fast reactor paired with a molten-salt energy storage system.

The storage system is designed to temporarily boost the plant’s output to 500 MWe for more than five hours.

TerraPower has described Kemmerer Unit 1 as a potential blueprint for deploying additional Natrium reactors, and that’s where the EPC question gets particularly interesting.

Who will build the future Natrium reactors?

TerraPower has not suggested who it will tap as its next EPC contractor for Kemmerer Unit 1.

At the same time, TerraPower is already working to establish a broader supply chain for its Natrium technology.

According to World Nuclear News, TerraPower announced contracts with 12 equipment vendors to support construction of the Kemmerer plant and completion of the sodium test and fill facility in August.

And TerraPower has bigger ambitions than a single reactor in Wyoming.

Earlier this year, TerraPower said it reached an agreement with Meta that could support deployment of up to eight Natrium units, with the first two targeted for delivery as early as 2032.

TerraPower has also partnered with Hyundai Engineering & Construction, which has been selected as its EPC contractor for up to eight 345-MWe sodium-cooled fast reactors.

While that doesn’t answer the immediate question of who will take over the full EPC role at Kemmerer, it provides a sense of the scale of TerraPower’s plans for the Natrium design.

First-of-a-kind project faces a different challenge

The Natrium project has also been positioned as a test case for the next generation of nuclear construction in the U.S.

The reactor is being built on a greenfield site near the retiring Naughton coal plant in western Wyoming, with TerraPower expecting construction to involve about 1,600 workers and the completed plant to employ about 250 full-time positions.

The project also carries significance well beyond Wyoming.

If TerraPower can move Kemmerer Unit 1 from construction to commercial operation amid its current timeline, the plant could provide a real-world demonstration of how a commercial-scale advanced reactor can be licensed, built, and operated in the U.S.

For now, though, TerraPower has another piece of the puzzle to solve, with its EPC team is changing.

The agreement is still subject to approval by the Georgia Public Service Commission (PSC). The filings also include a new Nuclear Uprate (NU-1) tariff structure as well as a request to approve a new extended power uprate (EPU) for Plant Hatch 1 & 2. An EPU for Plant Vogtle 1 & 2 was previously approved by the Georgia PSC in the 2025 Integrated Resource Plan (IRP).

As part of this transaction, Google will subscribe to the new NU-1 tariff and receive the Zero-Emission Credits (ZECs), which represent the carbon-free attributes of nuclear energy, associated with the power generated by the nuclear unit uprates. This structure, which Georgia Power says helps protect non-participating customers from incremental costs related to the uprate work, was originally contemplated as part of the 2025 IRP.

An EPU is a process that increases the electrical output of a nuclear generating unit through targeted modifications and upgrades to key plant equipment, such as turbines, pumps, motors, and cooling systems. These enhancements enable the reactor to operate at a higher licensed thermal power level, producing additional heat that increases steam flow to the turbine generator. As a result, the unit generates more electricity while continuing to utilize the existing reactor fuel design and plant infrastructure.

In its 2025 IRP, Georgia Power again updated load growth projections in its territory due to projected electricity demand from data centers, among other factors. Over the next six years, Georgia Power projects approximately 8,500 megawatts (MW) of electrical load growth – an increase of approximately 2,600 MW in peak demand by the end of 2030 when compared to projections in the 2023 IRP Update. In its proposed IRP from earlier in the year, those numbers were 8,200 MW and more than 2,200 MW, respectively.

Vogtle Units 3 & 4, the most recent reactors built in the U.S., entered commercial operation in 2023 and 2024, respectively, capping a project that cost billions over budget and took years longer than originally projected. The reactors were originally projected to cost $14 billion and be completed by 2017.

Georgia Power owns 45.7% of the reactors. Smaller shares are owned by Oglethorpe Power Corp., which provides electricity to member-owned cooperatives, the Municipal Electric Authority of Georgia and the city of Dalton. Utilities in Jacksonville, Florida, as well as in the Florida Panhandle and parts of Alabama also have contracted to buy Vogtle’s power.

This article contains reporting from the Associated Press.

Hyperscale data center campuses are arriving on utility systems in 50 to 300 MW blocks and asking to be served in 12 to 24 months. Interconnection queues across much of the country run five to eight years. New combined-cycle capacity is quoting delivery in 2029 and beyond. Owners are pivoting to upgrades to squeeze out every possible megawatt.

The POWERGEN 2027 Conference Program agenda, released last week, presents a set of often competing answers to this backdrop. Reciprocating engine suppliers, fuel cell manufacturers, grid operators, EPC contractors, air permitting specialists and turbine owner-operators are all working the same constraint from different positions, and in several cases they reach different conclusions.

POWERGEN 2027 runs Jan. 18-21 at the Salt Palace Convention Center in Salt Lake City, with PacifiCorp serving as Host Utility. The program includes dozens of sessions selected from more than 200 submitted abstracts, with other offerings carefully curated by industry experts on the conference’s esteemed advisory committee. In all, hundreds of speakers will participate.

Speed to power and the gates that control it

One of the clearest expressions of the timeline problem comes from the reciprocating engine case. In Data Center Demand and RICE Power, Jaisen Mody, a power generation consultant who previously managed a multi-billion dollar plant portfolio for Portland General Electric, argues that sub-five-minute starts, wide turndown, cycling durability and modular architecture let capital deploy in phases that track data hall construction, in a way large frame turbines with 2029 lead times cannot.

That is the supply-side answer. The interconnection process-side answer comes from Texas. ERCOT is sending two supervisors, one from technical review and one from operations stability planning, to ERCOT Interconnection and IBR Integration: How the Connect-and-Manage Model Works in Practice. Under the ERCOT model, projects are studied individually on their own merits rather than in cluster queues. That structure changes how developers should think about schedule risk, technical requirements, and the conditions that can delay or reshape a project. The session addresses what the state’s large-load surge is doing to stability requirements for generators, not only for load.

The session most likely to change a development schedule is Development and Air Permitting of Behind-the-Meter Power Generation for Data Centers. Peter Belmonte, National Power Sector Lead at Intera, Inc., walks through what happens when prime power lands at a campus: a minor emissions source can become a major stationary source under the Clean Air Act, triggering New Source Review, Prevention of Significant Deterioration in attainment areas, and lowest achievable emission rate plus offsets in nonattainment areas. Permitting runs one to two years and is lengthening. Aggregation rules can push a campus over the major source threshold even when individual units look modest.

How new load affects the machines

Two sessions examine AI load as a physical problem for existing generation.

In AI Data Center Load Impacts on Synchronous Generators, an engineer from Sargent & Lundy’s transmission planning group presents electromagnetic transient analysis of rapid, large-magnitude load swings near a hyperscale data center campus. The analysis shows how these swings can excite torsional vibration modes in turbine-generator shafts, potentially causing cumulative mechanical damage. These effects may not be fully captured in conventional interconnection studies. The presentation also compares three mitigation options: E-STATCOM, battery storage and synchronous condensers, establishing a 25 MW step change as the reference threshold below which impacts stay acceptable.

A counterpart session takes the opposite approach by asking the load to move instead. Data Center Flexibility: Unlocking Value for the Grid and the Customer covers work between a major power producer and NVIDIA on making data center infrastructure responsive to grid conditions. The session rests on a distinction that appears across several presentations: training, inference and batch processing have meaningfully different curtailment profiles. Those differences affect the value of demand response for both sides of the meter.

Examining the existing fleet

For owner-operators who cannot buy their way out of the lead-time problem, the program devotes substantial space to output that already exists.

A National Pulse Check on the U.S. Gas Turbine Fleet brings fleet-wide analysis from Turbine Logic and EPRI, incorporating 2025 and 2026 operating data across simple-cycle and combined-cycle units to identify where the installed fleet underperforms its potential. The study also looks at construction phase data, permitting timelines and build durations to show how much the development environment has changed, which is the argument for optimizing what is standing.

RELATED: Power producers increasingly opt for gas turbine upgrades as demand surges

Case studies led by the utilities and IPPs that own the assets run throughout the agenda and remain the bread and butter of POWERGEN, and two of them speak directly to what a constrained owner can do now.

The gas side is about knowing what you already have. In Plant Optimization and O&M: Tennessee Valley Authority Generation Fleet Phasor Monitoring Unit Implementation, TVA and HDR describe replacing periodic on-site stage testing with continuous synchrophasor capture across more than 100 natural gas and fuel oil generators at 17 sites. Meeting NERC MOD-026 and MOD-027 requirements by sending third-party contractors to units one at a time is costly and operationally disruptive at that scale. Permanent monitoring consolidates generator data onto a dedicated network, builds a fleet-wide frequency model, and allows a single tested unit to satisfy the requirement for sister units with identical generator, exciter and power system stabilizer parameters. The approach reaches cost parity with the current stage-testing cycle by years six to seven, and the real-time visibility into grid injection behavior keeps returning value well beyond that.

The coal side of the same question comes from Australia. Thriving Within Australia’s Energy Market Transition brings the team from AGL’s Bayswater station, a large coal-fired plant that has spent five years becoming something it was never designed to be. Since 2021 the station has increased its level of plant flexing by 113%, through lower minimum generation levels, the ability to two-shift units, faster and more repeatable starts, and dispatch driven by spot price signals and renewable output forecasts rather than steady-state baseload assumptions. The session is candid about what that cost: more starts and stops, greater cycling wear, and heavier demands on operators running dynamic dispatch in compressed timeframes. For plant managers at U.S. coal and gas assets facing the same pressure from a different direction, the parallels are close enough to be uncomfortable.

Storage meets the code cycle and the neighbors

Several sessions in the 2027 program deal with battery safety, fire code or community acceptance.

Fire Safety and Code Compliance for BESS: Testing Standards, AHJ Engagement and Asset Protection covers the gap between evolving fire codes and on-the-ground implementation. This session draws on submissions from the Fire Research Association, covering UL 9540A testing methodologies, NFPA 855 compliance and AHJ (Authority Having Jurisdiction) engagement strategies.

Community opposition is one of the most underestimated risks in utility-scale BESS development, with projects facing delays or cancellations not due to technical or regulatory failure, but due to eroded public trust. Siting BESS: Building Community Trust walks through the human side of energy storage deployment, drawing on developers’ experiences with siting challenges, AHJ and community concerns, as well as EPA perspectives on emergency preparedness and public communication during incidents. The session will feature lessons from projects that successfully navigated, or struggled to overcome, community resistance, including pre-application outreach strategies, the role of local emergency responders as project allies, and what the industry can do collectively to strengthen public understanding and confidence around battery safety.

In 2027, more of the conference program sessions will be on the exhibit floor to make it easier for attendees to move seamlessly between education, networking and the technologies and solutions shaping the power generation industry.

Keynote programming spans the three main days of the show: an executive panel on the supply and build cycle on Tuesday, a grid cybersecurity panel on Wednesday with FirstEnergy, AEP and APPA, and a Thursday panel on the reliability and performance gap with ERCOT, NERC, Dominion and EPRI.

The full program and registration are now available on the POWERGEN website. Factor This Power Engineering will publish interviews with program speakers in the weeks ahead.

AtkinsRéalis’ share of the contract will be added to its nuclear backlog in the third quarter of 2026. The contract represents the first RFBR unit in the broader refurbishment project, with similar work anticipated on Units 6, 7 and 8 in subsequent phases, all pending regulatory approval. It covers engineering and design services, project delivery, as well as program and project management associated with the planned life-extension of four of Pickering’s CANDU reactors, including work that would enable Unit 5 to operate until the 2060s.

“AtkinsRéalis brings decades of hands-on experience refurbishing Canadian-owned CANDU reactors in Ontario, with a proven track record of delivering complex life-extension work safely, reliably and with the discipline required for on-time and on-budget performance,” stated Ian L. Edwards, President and Chief Executive Officer, AtkinsRéalis.

In 2024, the Ontario government said it would support OPG’s plan to proceed with the next steps toward refurbishing Pickering Nuclear Generating Station’s “B” units (units 5-8), providing 2,000 MW once work is complete. In 2022, the Ontario government said it asked OPG to conduct a feasibility study on the potential for refurbishing Units 5 through 8. The last feasibility study was conducted between 2006 and 2009. OPG said while there was support for the refurbishment at that time, it didn’t happen because of challenging economics, stagnant electricity demand and anticipated supply chain issues and costs.

In September 2022, Ontario announced that OPG would continue to safely operate the Pickering Nuclear Generating Station through September 2026, pending CNSC approval. Further operation of Pickering Nuclear Generating Station beyond September 2026 would require a complete refurbishment. In 2023 the Pickering Nuclear Generating Station recorded its highest generation output since 2019 and its second-highest output ever as a six-unit station.

Ontario’s Independent Electricity System Operator forecasts that electricity demand in the province will increase. A life-extension of Pickering’s CANDU reactors, to be completed by the mid-2030s pending regulatory approval, would allow the reactors to operate for 30 or more years, and is part of Ontario’s plan to maintain and increase the supply of baseload clean power. This will take place against the backdrop of ongoing life-extension of CANDU reactors at the Bruce Power site, the recently completed Darlington refurbishment, the building of four small modular reactors at Darlington, the planned build of new nuclear reactors totaling 4,800 MW at Bruce Power’s site, and exploration of three OPG-owned sites in southern Ontario which are zoned for power production, regarding their suitability for hosting new energy generation including new nuclear power.

Planted targets faster solar deployment

Planted said its latest project demonstrates how quickly its model can move from development to power.

The company recently deployed a 28-MW behind-the-meter solar installation serving a neocloud data center.

According to Planted, the project went from initial contact to power in 10 months, with field construction taking less than three months.

The timeline highlights the company’s broader pitch to power-hungry customers: Build generation closer to the load and use automation to shorten the construction timeline.

Planted deployed more than 10 MW of solar in 2025, and said they’re on track to reach 100 MW by the end of this year. Its project pipeline now exceeds 20 GW.

Robots enter the power construction race

The company’s robotic construction fleet is central to that growth strategy.

Planted says its existing robots are fully committed through 2027. Its next-generation Sage robot is expected to more than double field productivity compared with the company’s current fleet.

The company ultimately is targeting a 10-fold increase in field-crew output.

“We have robots in the field proving the model works today,” CEO Eric Brown said. “This round is how we scale it.”

The strategy comes as power developers face a growing list of challenges, from interconnection constraints and land availability to supply-chain bottlenecks and a shortage of skilled construction workers.

Planted is just one company deploying smart machines to help build utility-scale solar projects. Via a partnership with engineering, procurement and construction (EPC) firm Burns & McDonnell, venture-backed AI startup Gritt has spent the past year evaluating robotics and testing their effectiveness in real-world construction environments, including at multiple utility-scale solar sites.

A new model for powering data centers?

Planted plans to expand beyond conventional solar projects.

Next year, the company said it plans to offer powered land — sites equipped with compact clean-energy systems and storage that can provide power for large electricity users.

The concept could appeal to data-center developers and other energy-intensive customers facing long waits for additional grid capacity.

Instead of waiting years for new transmission or grid infrastructure, the model aims to bring generation and storage to the load.

For the power industry, that points to a potentially important shift in how new electricity infrastructure gets deployed.

The next question isn’t simply how much power the U.S. can generate. It’s how quickly developers can build it.

Planted is betting autonomous robots can help close that gap.

As coal-fired power plant retirements accumulated in the last two decades, a popular replacement became combined-cycle generation. Additional combined-cycle generation is on the drawing board to power some of the emerging data center campuses. At many combined-cycle plants, a mindset developed – which was perhaps a carryover from simple cycle operation – that “lean and mean” staffing is best, with operators handling most tasks, including water treatment and steam generation chemistry monitoring. Yet, just one major chemistry upset can potentially shut down a plant for months at enormous cost to the owner. 

In this series, we will look at recommended on-line instrumentation that can assist plant personnel in monitoring water/steam chemistry throughout the network and for detecting upset conditions, where prompt corrective action can save a unit from major damage. Part 1 focuses on makeup water treatment. We will then move along a natural progression from condensate/feedwater to the boiler and on to the steam network.

High-Purity Makeup Water Is a Requirement for Power Boilers

The high temperatures and pressures in utility boilers (conventional and heat recovery steam generators (HRSGs) both) require high-purity makeup water. Below is a common set of guidelines and on-line instrument recommendations for makeup system effluent purity.

• Specific conductivity (S.C.):  ≤0.1 µS/cm
• Sodium: ≤2 parts per billion (ppb)
• Silica: ≤10 ppb

The development of synthetic ion exchange resins in the 1930s enabled makeup water systems to meet these guidelines, particularly for the high-pressure power units that evolved in the middle of the century.

Common pretreatment for IX systems was raw water clarification/media filtration to reduce particulate loading on the resins. But this process did very little to remove dissolved ions. Even with “clean” supplies such as a recreational lake or municipal potable water source, sufficient dissolved ions are present to steadily exhaust demineralizer resins. Resin regeneration is expensive, requires a waste regenerant neutralization system, and presents complexities related to hazardous chemical storage and handling. Starting in the 1980s, RO retrofit ahead of existing demineralizers gained popularity as a method to (greatly) extend IX resin run times and lessen the costs of regeneration. RO has now become the core demineralization technology for many HRSGs, with micro- and especially ultrafiltration for upstream particulate removal and portable mixed-bed IX bottles or continuous electrodeionization (CEDI) for final polishing. Such arrangements eliminate the need for hazardous regenerant storage and handling.

A note to Power Engineering’s co-generation and industrial steam generating plant readers; RO is becoming increasingly popular as a makeup water treatment replacement for sodium softening at industrial facilities.  

By far the most common RO design employs spiral-wound membranes, with each membrane, plus spacer and support material, wrapped around a central, perforated tube, all placed in a cylindrical pressure vessel.  

In this crossflow filtration process, pressure from the RO feed pump forces purified water (permeate) through the membranes.  The permeate spirals down to the central tube for discharge from the pressure vessels.  The reject, aka concentrate, exits as a separate stream, carrying most of the impurities away. 

Standard spiral-wound element dimensions are 8 inches in diameter by 40 inches in length.  Five or six membranes per pressure vessel is common.¹  

Modern RO membranes can remove 99% of dissolved impurities, which explains why the process has become so popular for bulk demineralization ahead of polishing IX units.

RO Instrumentation

Beyond the makeup system effluent guidelines and instruments recommended earlier, critical for overall system reliability is accurate RO performance monitoring. RO units typically include a variety of instruments for this purpose, which are highlighted in Figure 5. 

A primary issue with RO is that the size of the membrane channels, which are angstroms in diameter, will shrink or enlarge with changing water temperature. This effect influences several operating and performance parameters, including water passage through the membranes, salt rejection, and pump pressure. Temperature changes can also mask membrane fouling, scaling, or other problems. Accordingly, reputable RO and membrane manufacturers usually offer computer-based “normalization” programs that track system instrument readouts to make “apples to apples” comparisons and alert operators to deviations.

Now, let’s examine a number of the primary reasons why online analyses are important. Reference 1 goes into greater depth regarding all of these items and numerous others.

* Pressure: Differential pressure across the cartridge filters provides data on how quickly the filters are accumulating material and when they need to be swapped out with fresh replacements. Inlet pressure, along with flow, to the RO feed pump indicates if the pump is receiving an adequate supply of water. If not, pump overheating and failure may result. Normalized pressure deviations, including those from the pump discharge, and the permeate and reject discharge streams, are often the first sign of fouling or scale formation, indicating that a clean-in-place (CIP) or other corrective action is necessary. A common guideline is to schedule a membrane clean-in-place (CIP) procedure when normalized pressure or permeate readings have declined by 10% from baseline. We will examine additional CIP details later in the article.

Regarding permeate discharge pressure, RO systems should be designed to minimize or eliminate the potential for backpressure. Backpressure excursions can damage membranes and seriously influence RO performance. Similarly, feed pump overpressure can also damage membranes.

* Flow: In “normal” applications, a basic RO unit will produce 75% permeate and 25% reject, and the system flow meters allow for accurate calculations in this regard. As with pressure, unexpected flow deviations may indicate fouling or scaling, or perhaps a mechanical issue somewhere in the system.  Reject flow monitoring is critical. An automatic alarm should be in place to ensure that flow has not been pinched or worse, that the discharge valve has somehow been completely closed. That scenario defeats the crossflow process and forces all impurities into the RO membranes; an action that can quickly cause scale formation and irreversible fouling.

* Specific Conductivity: A gradual rise in normalized permeate specific conductivity is often an indicator of membrane degradation. Even with good chemistry control and regular CIPs, membranes age and lose efficiency. Incursions of oxidizing biocides, and especially bleach/chlorine, will accelerate degradation. An important monitoring tool in this regard is chlorine residual (or perhaps oxidation-reduction potential (ORP)) to ensure that the reducing agent is consistently removing the oxidizer. But note that some microbes will go into hibernation when they detect an oxidizing biocide and then re-emerge once the biocide residual has been neutralized with a reducing agent. They then can cause rapid fouling of RO cartridge filters and membranes. I directly observed this microbiological activity at two former plants. In one case, we installed a supplemental chemical feed system to periodically inject a non-oxidizing biocide, which solved the problem. Researchers have also been developing oxidizing biocides that control microorganisms but are much less aggressive to membrane material.

A sudden rise in permeate conductivity often indicates a mechanical failure somewhere in the system. A common RO feature has a tap and valve to grab sample the permeate discharge from each pressure vessel. This arrangement allows the operator to zero in on the pressure vessel with the failed membrane. Further refinement is possible by use of a flexible probe that can be inserted into the permeate discharge line to identify the failed membrane.

* pH: A common material for early RO membranes was cellulose acetate, which had to operate in a rather narrow pH range of 4-7 to avoid membrane degradation. Sometimes required was a small acid feed upstream of the unit. Most RO membranes are now of polyamide material, which can operate in a much broader pH range. Regardless, pH is still a common on-line measurement to guard against excursions that could damage membranes. For the two-pass RO units integral to high-purity makeup systems; common is a small caustic (NaOH) feed ahead of second pass to convert any free carbon dioxide to bicarbonate alkalinity for removal in the second pass.

* Temperature: Raw water temperature ranges are normally well within limits that won’t affect RO integrity. However, for systems with inlet water heaters, a malfunction could raise the water temperature to potentially troublesome levels. Also, as the next section outlines, CIP systems often include a heater in the blend tank to raise the temperature of the cleaning solution. But the circulating pump will also raise temperature, so this data point requires monitoring to prevent temperature excursions during membrane cleanings.

Some Brief Notes About Membrane Cleaning

Even with well-designed pre- and process treatment programs, RO membranes gradually accumulate suspended and dissolved solids. If not removed, the solids will eventually cause irreversible fouling. Accordingly, periodic off-line cleanings are necessary to restore membrane conditions to near baseline.  Recommended guidelines to initiate a cleaning are:¹

• Loss of 10 to 15% in normalized permeate flow
• Increase of 10 to 15% in normalized differential pressure
• Decrease of 1 to 2% in salt rejection

Because fouling occurs on the concentrate side of the membranes, cleaning solutions are delivered through the reject circuit. Piping and connections should be plumbed such that each stage can be cleaned individually with a fresh cleaning solution. Figure 6 illustrates the basic schematic of a cleaning skid.

RO permeate is a requirement for cleaning solution makeup. Via the recycle line and tank heater, the solution is brought to an appropriate temperature range; 95-105°F is common. The operators then valve in the circulation loop to and from each stage. A standard component of the circuit is a cartridge filter to capture particles released during cleaning. Both acidic and alkaline solutions may be necessary for cleaning to remove mineral scales, organic compounds, and other materials. With the aid of feed water chemistry and operating data, a reputable RO equipment or membrane supplier should be able to recommend and/or provide formulations for optimal cleaning. As was mentioned earlier, most modern RO membranes can handle a relatively broad pH range during operation, but expert advice is important for cleaning solution selection and procedures to protect the membranes.

Conclusion

Reverse osmosis has become an integral piece of makeup water treatment systems at many industries including power, electronics, pharmaceuticals, and so forth. But the units require proper attention and monitoring to perform reliably. Modern online instrumentation combined with good computer software can provide the tools that operators and technical personnel need to keep things running smoothly.

Disclaimer: This article offers general information and should not serve as a design specification. Every project has unique aspects that must be individually evaluated by experts from reputable water treatment equipment and chemical firms.

References

1. Byrne, W., Reverse Osmosis: A Practical Guide for Industrial Users, Tall Oaks Publishing, Littleton, Colorado, 2002.

About the Author: Brad Buecker currently serves as Senior Technical Consultant with SAMCO Technologies. Buecker has many years of experience in or supporting the power industry, much of it in steam generation chemistry, water treatment, air quality control, and results engineering positions with City Water, Light & Power (Springfield, Illinois) and Kansas City Power & Light Company’s (now Evergy) La Cygne, Kansas, station. Additionally, his background includes eleven years with two engineering firms, Burns & McDonnell and Kiewit, and he spent two years as acting water/wastewater supervisor at a chemical plant. Buecker has a B.S. in chemistry from Iowa State University with additional course work in fluid mechanics, energy and materials balances, and advanced inorganic chemistry. He has authored or co-authored over 300 articles for various technical trade magazines, and he has written three books on power plant chemistry and air pollution control. He is a member of the ACS, AIChE, AMPP, ASME, AWT, and he is active with Power-Gen International, the Electric Utility & Cogeneration Chemistry Workshop (now co-located with the annual POWERGEN conference), and the International Water Conference. He can be reached at bueckerb@samcotech.com.