There is the economy you see on the screen. And then there is the economy that runs the screen.

For the last few years, the screen has been winning.

We have been living in a “Zero Marginal Cost” fantasy…

We convinced ourselves that adding another million users to ChatGPT was free. That scaling a startup was just a matter of writing better code. That “The Cloud” was an infinite, weightless place where business happened by magic.

But today, S&P Global Energy dropped a bucket of ice water on that fever dream.

I’ve spent the morning reading through their new Horizons Top Trends 2026 forecast. It’s dense. It’s technical. It’s written in the polite, measured language of analysts who don’t like to shout. But if you translate the data into plain English, they are shouting.

The bottleneck isn’t the code anymore. It’s the copper. It’s the transformers. It’s the sheer, raw wattage required to keep the simulation running. And right now? We don’t have enough of it.

Adding Another India to a Grid That Can’t Handle It

Let’s look at the raw numbers, because they are wild.

S&P projects that by 2030, global data center power demand could hit 2,200 terawatt-hours (TWh).

It is easy to glaze over a number like that. We hear stats about “energy consumption” all the time. But context is everything.

Last year, the IEA warned we might hit 1,000 TWh in 2026…

S&P has looked at the trajectory of AI adoption and effectively doubled that forecast.

To put that 2,200 TWh figure in perspective: that is roughly the total electricity consumption of India.

But the “India” comparison actually undersells the severity of the problem.

A country’s energy demand breathes. It fluctuates. People turn off lights… factories shut down for the night. The load varies.

AI demand is different. It is a flat line of maximum intensity. It is relentless, baseload hunger.

Data centers don’t sleep, they don’t take holidays, and they don’t throttle down.

We are attempting to bolt a rigid, massive, always-on load onto a global grid that was designed for the variable rhythms of the 20th century. And we are trying to do it while 38% of the companies driving this demand have no credible plan to get to net zero.

We are pouring $500 billion in capital expenditure into U.S. data centers for 2026 alone. That is half a trillion dollars going into the “engine” of the economy, while the “fuel lines” are rusted through.

We are building the most advanced computing infrastructure in human history, and we are expecting it to run on transmission maps drawn up during the Nixon administration.

The past year saw a multitude of factors driving up electricity prices, including rapid growth in electricity demand, supply chain tightness, deployment delays for transmission and production projects, and an uncertain political and permitting climate. We expect all of these trends to continue in 2026. Changing economics have turned the justification for project development in one of the most conservative market sectors on its head. And while new entrants and technologies entered the marketplace, we also saw technologies and developers favored by the prior presidential administration struggle to find their footing under the current one.

COMMENTARY

As businesses and policy makers strive to adapt, one underlying fact remains: the explosive growth of power demand has made energy one of the hottest sectors for investment. We expect that the whiplash and market disruption caused by policy variation in an era of massive load growth means significant opportunity for investors.

Trends Shaping the Market

Historically, companies entering the energy market faced constraints of legacy hydro and thermal generation setting the market rates below breakeven point for new technologies and products; there was little room for the cost of innovation. Flat electricity demand meant that a first-of-a-kind investment could only break even with significant government incentives and subsidies. Government grants and subsidies helped mature the solar, wind, and battery markets to the point where they could compete against fully depreciated legacy assets. With the energy grid as a static playing field, new opportunities arose only when legacy assets retired. In recent years, this dynamic has been upended. The market now faces more entrants, greater demand, and more regulatory uncertainty. First, the Biden-era massive infrastructure investments under the Infrastructure Investment and Jobs Act and the Inflation Reduction Act dangled billions of dollars of federal funds to cost-share development across energy production, transmission, and distribution. While largely unspent, these funds did spur great interest in development, largely from wind, solar and battery developers, though also in nuclear and geothermal sectors. Second, a surge in demand—largely driven by data center development, electrification, and the return of manufacturing to U.S. soil—is straining existing infrastructure and creating new opportunities. Third, the policy shifts of the current administration—including the rescission of granted permits, issuance of stop work orders, and freezes on infrastructure funding—have created regulatory uncertainty, especially with clean energy options like wind, solar and battery storage. Concurrently, the current administration is exercising rarely-used powers in favor of coal, natural gas, and nuclear projects. New technology developers continue to enter the market to address the misalignment of demand and supply. The current administration has prioritized fixed thermal resources such as fossil, nuclear, and geothermal and de-prioritized and reevaluated the place of wind and solar in the market. Related sectors have been affected both positively, e.g. mining for critical minerals, and negatively, e.g. transmission tied to grid-scale wind and solar projects.

In early 2020, the prevailing narrative in the power sector was a continuation story of the developments from the decade before: renewable buildout will keep compounding, thermal capacity will keep retiring (albeit at a slower rate), markets will evolve to compensate for flexible generation products, capital will keep moving earlier in the development value chain and the grid will gradually transition to a cleaner fuel mix. But soon enough, this “straight-line” version of that story started losing strength. COVID-era supply chain disruptions and the start of the U.S.-China economic decoupling across strategic industries exposed how fragile “just-in-time” global energy supply chains really were. Interest rates reset the cost of capital, changing what penciled out and what didn’t. Project development has become a riskier proposition. Reliability concerns around dispatchability and extreme events moved from theoretical to visible. The energy transition didn’t fizzle out but instead slowed down and stopped being linear.

COMMENTARY

As we reach year-end 2025, the narrative of the U.S. power sector has shifted drastically. Power is no longer just one chapter of the energy transition; it has become a strategic constraint on nationwide economic growth. Artificial intelligence (AI)-driven data center loads are now arriving fast and in clusters and require strict reliability. In that sense, this new race for “power for compute” resembles prior historic industrial and infrastructure turning points (e.g., the 19th century buildout of railroads, early 20th century mass electrification, the telecom network rollout) rather than the prior decade’s slow incremental load growth and thermal generation attrition. And because leadership in AI has significant geopolitical consequences, the electricity abundance needed to catalyze the new data center economy is now part of the great power competition between the U.S. and China. Data centers represent a once-in-a-century demand shock. Demand growth from U.S. data centers is hard to fathom. Data centers are expected to represent up to 12% of total U.S. electricity consumption by 2028 (up from ~4.4% in 2023). They are the key driver for forecasted 5.7% annual energy demand growth in the U.S. over the next five years (compared to a sluggish 0.2% annual average in the 2010s). The challenge is not just the absolute scale of growth, but the speed of incremental demand additions; each new data center can be akin to adding a large city’s electricity demand to the grid, often within just one to two years. This puts enormous pressure on power infrastructure. This pressure is often local before it is national: Pew Research notes extreme regional concentration already (e.g., data centers consuming more than 25% of Virginia’s electricity in recent years). This is why the AI load wave may feel like a national story but behaves operationally like a regional story. It shows up as a reliability and deliverability problem in specific places where power, fiber, land, incentives and workforce overlap. This is often not where interconnection is easiest. PEI expects that as premium locations are exhausted, data center demand will start moving to the next best alternative locations and will become more geographically spread in the mid-term. Tier II markets such as Austin-San Antonio in Texas, and Las Vegas-Reno in Nevada, are already displaying increased momentum, as well as rural locations like Louisiana and North Dakota. This is a “once-in-a-century” industrial demand shock: AI and data center growth is reshaping energy infrastructure in ways comparable to prior waves of large-scale infrastructure expansion and geostrategic races. Electricity is at the core of compute competitiveness because large-scale AI requires highly reliable baseload-like supply. The AI race turns electrons into a strategic input. AI’s role in demand means that the power sector is once again a national priority. As compute depends on electricity, U.S. AI competitiveness hinges on reliable electricity infrastructure to support data centers. Moreover, supply chains are no longer “just economics.” U.S. policy and trade enforcement have increasingly constrained reliance on certain China-linked energy inputs. The Department of Energy’s “Speed to Power” initiative is framed as a federal effort to accelerate large- scale generation and transmission development “to win the AI race.” This is a direct signal that electricity abundance is now being treated as a matter of global economic competitiveness and national security interest. This laser-focused federal action echoes the same catalysts behind prior large scale industrial buildout efforts in the U.S., supplementing private capital with strong policy directions and financial support. Global context matters as well. In 2024, China built more than 50 GW of thermal generation and 277 GW of utility-scale solar, according to government figures, versus the U.S. at ~2.5 GW thermal generation and ~40 GW solar, according to the U.S. Energy Information Administration (EIA). It is a sobering reminder that the ability to build power plants is itself a competitive advantage. This also demonstrates why, in the U.S., solar and storage will continue to be built out to the maximum feasible system capacity alongside thermal capacity—not because they solve reliability alone, but because they are among the fastest scalable sources of incremental energy in a build-constrained environment. If the power sector falls behind expectations of policymakers (i.e., if capacity deployment timelines remain incompatible with strategic AI load and reliability needs), federal support could plausibly increase accordingly. This could take many forms: expedited siting/permitting for large-scale resources, new approaches to transmission corridors and cost allocation or federal coordination around “national priority” projects. From cheap MWh to deliverable firm power: how AI demand is repricing reliability. Rapid improvements in wind and solar technology; supported by state and federal incentives drove a surge in intermittent generation, with solar levelized cost of energy (LCOE) falling by roughly 85% between 2010 and 2019. Grid operators were able to integrate this capacity largely because many markets still had significant thermal oversupply and anemic demand growth, allowing renewable growth and plant retirements to occur without immediate reliability pressures. Two key structural changes collided to shift the narrative: 1) Demand is back, and it is concentrated. Data center and AI loads are large and often need to locate near power, fiber networks, metro areas and talent. This demand is less sensitive to power prices as long as reliability needs are met. The result is localized scarcity: at certain nodes, the system doesn’t need more “clean annual MWh” as much as it needs more deliverable capacity. 2) There is scarcity of available firm power. Offtakers increasingly want firmness, not just energy attributes. They want firm service (based on location, shape of load and availability) because downtime and curtailment risk are existential to large AI workloads (resulting in the industry- standard 99.999% uptime requirement). This does not mean that offtakers are walking away from their prior environmental commitments, but rather that their key priority has shifted pragmatically towards reliability. Of course, if reliability is assured, hyperscalers still prefer clean solutions. These factors are creating a market dominated by two main themes: At a national level, renewables continue to add needed lowest-cost marginal energy to the system; and, constrained grids create a premium for firm, deliverable power, where pricing reflects the scarcity of capacity that can show up when and where it is needed. In other words, there has been an investment logic shift away from just a “cheap energy” ideal (where the goal is to produce MWhs at the lowest LCOE) to a “deliverability + firmness” ideal (where the goal is to provide assured service with MWhs delivered with high probability, plus the capacity attributes that keep the system stable). In this world, firmness is repriced, and the market is creating “reliability as a product.” One of the clearest signals of this regime change is the widening gap between what consensus market forward price curves imply for energy, and what large-load buyers are willing to pay for firm, long-dated, deliverable supply. One telling example is Vistra’s recent announcement of a 20-year power purchase agreement (PPA) with a large investment-grade counterparty for up to 1,200 MW of its Texas-based Comanche Peak nuclear plant. The implied PPA pricing was ~$90–$100/MWh, with PEI analysis backing into an implied reliability/capacity value of roughly ~$24/kW-month (~$790/MW-day). This is a clear expression of how valuable reliable capacity has become in ERCOT. The Comanche Peak deal suggests that forward forecasts and markets may not fully capture the gravity of inadequate reliable supply over the long run. This is what firm clean power looks like when it moves from aspiration to contracting reality, driven by price insensitive large-scale demand from data centers. The new power sector bottleneck is not capital, but execution: interconnection, equipment, and build-out rate. There is roughly ~2 TW of utility-scale solar and BESS projects in U.S. interconnection queues, according to published information. The U.S. built roughly ~40 GW utility-scale solar and ~10 GW utility-scale BESS in 2024, implying only ~2% of queued capacity gets built annually, according to EIA. The agency’s expectations for 2025 additions (~33 GW solar, ~18 GW storage) underline the same low throughput, which indicates structural constraints that are not related to project pipelines or capital. Equipment lead times for projects are now a key input in strategy, not just a procurement detail. Transformers are emblematic of this deep shift. Lead times for generation step-up transformers have been documented at ~143 weeks in 2025, and large transformer lead times are widely reported as stretching into multi-year territory. These dynamics suggest that next decade’s winners will not only be those with low cost of capital. Winners will be those with streamlined execution platforms who can seamlessly manage and integrate land, permits, queue position, contracting and construction resources, with the capacity and know-how to build generation where and when needed by data centers. What the next decade will look like for the U.S. power sector. Near-term (2026–2030): “Speed to power” is the name of the game. When capacity demand timelines compress, the system reaches for what can be delivered most rapidly, regardless of technology type:

• • Continued deployment of solar and BESS, given their fast speed to market and significantly shorter development timelines.

  • Utilizing existing generation more efficiently (uprates, extending run time, improving availability).

  • Leveraging existing interconnections.

  • Repowering and life-extending where feasible (including nuclear where permitted).

  • Self-build / behind-the-meter solutions where grid timelines are too slow.

  • Adding dispatchable generation where policy and permitting allow.

  This is also where scarcity emerges in unexpected places: access to turbine kits, interconnection equipment, transformers, EPC capacity and viable sites becomes a scarce commodity. Mid-term (2030–2035): The bulk of incremental firm supply is likely to consist of combined cycles (plus select peakers), but at a higher cost ceiling. As demand growth persists and interconnection/transmission reform remains slow, the system’s mid- cycle “workhorse” is likely to be combined-cycle gas turbine (CCGT) new builds, supplemented by peakers and uprates, because these remain the most scalable dispatchable mid-term option that can be replicated across many regions. However, the economics no longer resemble the “cheap gas build” era of the 2010s. Capital costs for new combined-cycle projects have increased materially as turbine supply tightens, EPC bandwidth remains constrained, and labor markets stay competitive. Industry leaders have noted that all-in CCGT costs, which were roughly ~$1,200/kW in 2022, can now approach ~$2,500–$2,800/kW, according to NextEra investor relations, depending on configuration and site conditions. Despite these higher costs, new CCGTs are becoming essential infrastructure for supporting large-scale data-center expansion. Many of these projects are likely to be developed under long-term, 20-year offtake arrangements that explicitly value deliverability, availability and speed-to-power. Higher capital requirements naturally imply upward pressure on electricity prices, but it is uncertain how much of this cost increase regulators will permit to be fully socialized; especially as a meaningful share of incremental demand originates from a single high-growth industry. This tension between system-wide reliability needs and ratepayer protection will shape procurement and cost-allocation debates across several regions. As a result, we expect increasing interest in microgrid and behind-the-meter solutions, particularly for large data center operators and industrial loads seeking to insulate themselves from regulatory uncertainty, interconnection delays and the rising cost of grid-supplied electricity. These alternative architectures may become an important release valve as policy makers balance affordability, fairness and the need to bring firm capacity online quickly. Ultimately, the rising capex environment reinforces why near-term reliability additions continue to favor existing interconnections, uprates and brownfield advantages: these are pathways that compress both cost and schedule risk while the system races to meet rapidly clustering load. Post-2035: New firm clean generation technologies could become meaningfully additive (SMRs and geothermal), but the timeline remains uncertain. Geothermal and SMRs can plausibly become real contributors after 2035, because they provide a dual solution for AI load: firm, carbon-free power with high availability. The critical question is not concept but buildability (repeatability, standardization and supply chain, as well as cost). Capital support is clearly building for both technologies. Amazon has publicly signed agreements to support SMR development (including projects expected in the early 2030s) and invested in X-energy; broader private capital and public support are also accelerating the space. On the federal side, the DOE recently announced up to $800 million of support for SMR projects (given to TVA and Holtec), explicitly linked to rising demand from AI and other loads. For geothermal, next-generation developers have raised substantial capital to scale enhanced geothermal projects aimed at bringing firm capacity online before the decade ends, laying the groundwork for larger contributions in the following decade. Power sector implications: the investment playbook evolves. The AI/data center cycle reshapes the opportunity set. It pulls forward new builds, but it also re-values operating assets. In a system where “getting built” is the binding constraint, operating assets with proven interconnection, permits, fuel access and deliverability increasingly behave like scarce inventory. The recent Three Mile Island nuclear plant revival is a clear indication of today’s need for near-term power. Existing thermal assets remain scarce and mergers and acquisitions (M&A) are active into 2026 (with non-CCGT moving to center stage). For much of the last two decades, thermal generation was underwritten as an attrition story. That is no longer valid. Regardless of technology or fuel, there is renewed demand for dispatchable generation, driven primarily by public IPPs; regulated utilities; and sponsor-backed scaled platforms. While assets in the highest-growth areas are trading at significant premiums, the valuation uplift is increasingly broad-based across the sector. A key accelerant is the private-to-public arbitrage: many private transactions are clearing around ~6–9x EV/EBITDA, while public IPPs trade >11x, creating room for accretive consolidation. We expect this M&A cycle to continue in 2026, with non-CCGT assets (older-vintage gas, peakers and coal) taking a larger share of processes as buyers prioritize near-term deliverability and time-to-power. In this environment, PEI is seeing these dynamics play out in real time across the M&A landscape. Processes are increasingly shaped by buyers who value deliverability, contractability and speed, and we are advising on a growing number of both buy-side and sell-side mandates where these themes are central. Market tension is rising as competitive investors, public IPPs, utilities, infrastructure funds and global strategics pursue scarce operating thermal assets, and we are leveraging our market insights to help clients frame and capture that value. Parallel financing activity has also become a defining feature of current processes: we are routinely running sell-side financing processes alongside M&A to establish valuation floors, sharpen valuation benchmarks and broaden the field of premium buyers. Across these live mandates, PEI’s combination of investor access, power-market expertise, specialist debt advisory and proprietary analytics is proving critical in positioning assets for maximum resonance with the market and delivering superior outcomes for clients. Thermal development is fashionable again and “speed-to-power” is the premium attribute. Thermal development is back in the investable set, supported by favorable contract terms and strong demand for shovel-ready, fully permitted projects. In this environment, the market is increasingly indifferent to technology labels and intensely sensitive to schedule. Projects that can credibly reach operations quickly (not just projects with attractive long-run economics) command a premium. That premium is even higher for developers who have secured equipment ahead of construction financing. Original equipment manufacturers (OEM) are increasingly asking for earlier payments to lock turbine slots, and the ability to de-risk equipment has become a core value-creation lever. PEI is observing this dynamic firsthand in the market. We are raising debt and equity for thermal development across markets and have been supporting developers with equipment and pre-construction financing solutions that de-risk procurement and compress timeline risk, often translating directly into better contractability and higher valuations. U.S. renewables continue to attract global capital. There is still deep global interest in U.S. renewables, but it is increasingly focused on large platforms with operating portfolios and credible pipelines. Global investors (including pension funds, sovereign wealth funds and other large institutions) are looking to write large checks for scalable opportunities that provide capacity deployment certainty and repeatability. This sets up a more active sell-down environment in 2026. While single asset or small portfolio transactions can be more challenging, there are strategic paths that can be pursued, including sell-downs and structured equity solutions, that provide floor valuations and create competitive tensions for maximizing transaction outcomes. PEI is the leading specialist investment bank in designing and running global M&A processes and bringing global investors into U.S. opportunities. PEI’s Hong Kong presence supports that outreach with boots-on-the-ground coverage of cross-border investors pursuing U.S. renewables platforms. Creative financing becomes a competitive advantage for renewable platforms (LCs + private credit + flexibility). As pipelines grow, LC and deposit requirements rise and development timelines extend, renewable platforms are increasingly turning to bank-financed pre-NTP LC facilities to keep projects moving through queues and procurement milestones. These facilities are becoming larger and more strategic as platforms scale. In parallel, more developers are using private credit and structured equity solutions for construction equity needs. Such options are often non-dilutive and more cost competitive compared to traditional equity, allowing platforms to recycle capital and grow faster. In short, creative financing solutions are now a structural part of the renewable ecosystem. Platforms that adopt them early will be one step ahead. PEI is actively structuring these solutions, designing tailored LC facilities, junior capital solutions and working across bank and private-credit markets to deliver flexible, developer-specific capital that supports growth and aligns with the new reality of longer timelines and higher collateral intensity. The U.S. power sector has entered a defining decade shaped by speed, reliability and electricity pragmatism. The slow-and-steady transition narrative of the last cycle has collided with today’s data center driven load growth and unexpected shifts in supply chains, cost of capital and system reliability. The coming years will require balancing these pressures while supporting U.S. economic competitiveness and geopolitical strength, with federal policy once again serving as an essential backdrop. In this environment, the most valuable power will not be the cheapest MWh but the power that is deliverable, firm, financeable and on time. —Adil Sener is partner at PEI Global Partners LLC. This material has been prepared by PEI Global Partners Holdings LLC and PEI Global Partners LLC (together, “PEI”) a U.S.-registered broker-dealer. This communication is being provided strictly for informational purposes only. Any views or opinions expressed herein are solely those of the institutions identified, not PEI.

  View Comments

  Terms and Privacy Policy

  Privacy Dashboard

  ---

• 

  Meeting the Moment: Industry Leaders Chart the Course for Power in 2026

  • 
  • 
  • 
  • 
  • 
  • 
  • 

  1 / 5

  2026-power-industry-forecast

  Aaron Larson

  Fri, January 2, 2026 at 8:20 AM EST 18 min read

  From artificial intelligence-driven efficiency to transmission bottlenecks, power industry insiders share their perspectives on the opportunities and obstacles shaping 2026 and beyond. The power generation sector enters 2026 at a critical inflection point. Electricity demand is surging—driven by data centers, manufacturing reshoring, and transportation electrification—even as the industry navigates an unprecedented transformation in how power is generated, transmitted, and consumed. Utilities and independent power producers face the dual challenge of maintaining grid reliability while transitioning to cleaner energy sources, all amid regulatory uncertainty and shifting market dynamics. This year promises to be pivotal for decisions that will shape the industry’s trajectory for decades. From the continued buildout of renewable capacity to the integration of artificial intelligence (AI) into grid management, 2026 will test the sector’s ability to balance innovation with pragmatism.

  AI Enhances Efficiency

  “The grid will become even more AI-enabled in the next year as AI becomes necessary for utilities to be more efficient,” Pradeep Tagare, head of Investments with National Grid Partners, told POWER. “In our recent Utility Innovation Survey, we saw this in real time: utilities are no longer treating AI as ‘nice to have,’ but rather as a strategic tool to manage load growth, enable reliability, and accelerate grid expansion. Due to the pressures that the electrification of heating, transport, and data center expansion place on the grid, AI-driven forecasting, risk mitigation, and infrastructure planning tools will become essential.” Albert Hofeldt, Chief Product Officer with Power Factors, said optimization has always been a priority for renewable energy operators, but it has taken on a new significance over the past few years. “Returns for renewable energy projects are shrinking, a sign of our current economic context and a signal of maturity for the market. Rising interest rates have increased the cost of capital, while financial support from state and federal governments disappear, pushing costs back onto developers. The renewable energy market is mature enough now to handle these changes, but it increases the pressure to build a bulletproof business model. In many cases, renewable energy companies are realizing that improving production and reducing future costs can actually deliver a greater and safer return over time than a new development,” he told POWER. “Machine learning and AI have a big role to play in increasing production from existing assets,” Hofeldt added. “AI can predict issues with underperforming assets earlier, making it possible to conduct maintenance with less downtime, ultimately maximizing energy production and revenue. It can also identify the root cause of underperformance, helping operators prioritize issues quickly and plan maintenance so that the highest-impact fixes are done first.”

• How Much Power Will AI Demand

  While AI is rapidly becoming indispensable for optimizing grid efficiency and predictive maintenance, this very success is fueling a new massive challenge, that is, meeting the unprecedented increase in power demand that is predicted from the data centers that house AI infrastructure. “For years, the conversation was about gradual electrification,” Sonya Montgomery, CEO of The Desoto Group, told POWER, referring to electric vehicles (EVs) and heat pumps, as examples of the slow but steady shift. “Recently, however, the need to find gigawatts, not megawatts, almost overnight due to AI-computing demands is something the industry must adapt to.” A report published in December 2024 by the Lawrence Berkeley National Laboratory under contract with the U.S. Department of Energy projected total data center energy use through 2028. Researchers presented a range of estimates to reflect various scenarios. Their low- and high-end projections were roughly 325 TWh and 580 TWh in 2028, which represented 6.7% to 12.0% of total U.S. electricity consumption forecasted that year. However, since that time, new developments have put in question almost all estimates of future data center energy demand. “The DeepSeek R1 model release was probably the most pivotal point in 2025,” Ryan Luther, director of Energy Transition Research with Enverus, reflected. “It made the entire market second guess the load growth from data centers.” Developed by Chinese startup DeepSeek, R1 is a powerful open-source reasoning model known for its high performance in coding and mathematics at a fraction of the cost of its American counterparts. DeepSeek R1’s efficiency breakthrough showed that algorithmic innovation could drastically reduce the energy footprint per AI task, casting doubt on the exponential power-growth curves currently driving data center planning. “I think we will continue to see moments like that in 2026 that will have major impacts on the outlook for power markets,” Luther continued. “For instance, Meta is clearly losing the AI race and wasting hundreds of billions pursuing it. If they don’t turn it around quickly, they might have to abandon the endeavor. That would cut a fifth of the data center build-out and reset expectations on power demand growth.”

  Utility-Scale Solar Surges On

  Regardless of what happens with data center demand, solar photovoltaic (PV) technology will remain the powerhouse of the energy transition (Figure 1). Its unprecedented growth will continue to reshape power markets in 2026 and beyond.

  1. The Hale Kuawehi Solar and Battery Storage Project reached commercial operations on March 25, 2025. The project, located on Hawaii Island, integrates 30 MW of solar photovoltaic capacity with 30 MW/120 MWh of battery storage, ensuring a stable and reliable supply of clean electricity to the Hawaiian Electric grid. Courtesy: Innergex Renewable Energy Inc.

  “Renewables and storage will continue to do the bulk of the heavy lifting,” said Peter Davidson, CEO and co-founder of Aligned Climate Capital. “Solar, wind, and batteries are the fastest, cheapest, most scalable way to add firm power—and anyone who thinks otherwise is clinging to an outdated narrative.” The International Energy Agency (IEA) projects in its Renewables 2025 report, which was released in October, that solar will account for about 80% of the total increase in global renewable power capacity over the next five years. It estimates about 3.68 TW of solar capacity will be added by 2030. This is an astounding figure. At least one detailed study suggests fixed-tilt PV systems require on average 2.85 acres per MW and that tracking systems take even more space—4.17 acres/MW. (Tracking systems use mechanisms to follow the sun, which increases energy yield but requires more spacing between rows to avoid shading as panels tilt and rotate.) Assuming 3.5 acres/MW are used on average for all the new solar PV being added in the next five years, nearly 13 million acres of land will be needed. To put that into perspective, that’s like covering the entire country of Slovakia or most of the state of West Virginia completely with solar panels. “The solar industry’s resilience is real, but it’s not guaranteed,” Jorge Vargas, CEO of Aspen Power, told POWER. “The biggest potential disruptor would be supply chain friction, especially if trade policies shift in the coming years. Another leading indicator is the health of the tax equity market. If there’s any contraction in available tax credit capital, that would slow growth more than any technical factor.”

  Residential Solar Has Gone Mainstream

  Driving the decentralized energy movement, U.S. residential solar installations are now measured in the millions, with the Solar Energy Industries Association (SEIA) reporting well over 5 million total installations across the country. However, recent economic headwinds and policy transitions have led to a notable contraction in the rate of new residential installations through the first half of 2025. Still, SEIA’s base case forecast projects that the residential solar market will grow by 3% annually on average from 2025 to 2030. SolarTech, a leading solar company with offices in California and Arizona, released its own study on the state of solar in December, which was created based on a survey of 2,000 homeowners. “What we’re seeing in the 2025 State of Solar report is a decisive shift in homeowner mindset,” Sonny Gonzalez, director of Marketing at SolarTech, told POWER. “Solar is no longer viewed as an experimental or early adopter investment—it’s becoming a standard part of long-term home planning. With roughly 70% of homeowners either already using solar or expecting to adopt it within the next five years, the market has clearly moved into the mainstream. This level of interest reflects not just improved technology and affordability, but a growing desire among homeowners to take control of their energy future.” “Right now, the biggest driver is still the fundamental economics,” said Vargas. “Solar is cost-effective, scalable, and increasingly flexible across customer segments. Utility rates are expected to rise sharply over the next several years, which makes solar an even more compelling hedge against those escalating costs. The Inflation Reduction Act [IRA] provided a significant tailwind for project development, but as those incentives begin to phase out, developers and customers will have to rely more heavily on intrinsic project economics and risk management. Going forward, development will increasingly be driven by cost savings, energy resilience, and load-side management. The projects that get built will be those that help customers control costs and mitigate exposure to rising utility prices, even in a world without IRA-backed subsidies.” “While utility savings remain the single strongest motivator for going solar, the data shows that homeowners are thinking far beyond monthly bills,” said Gonzalez. “Rising concerns about grid reliability, unpredictable outages, and increasing electricity rates are reshaping how people evaluate their energy options. Solar is now seen as a pathway to stability and independence. For many homeowners, adopting solar is as much about resilience and control as it is about reducing costs—and that marks a major evolution in the residential solar mindset.”

  FEOC Rules Take Hold

  The Foreign Entity of Concern (FEOC) requirements are a set of restrictions established by Congress across several laws, beginning with the Infrastructure Investment and Jobs Act and the IRA, and later expanded by the One Big Beautiful Bill Act (OBBBA, Figure 2). They are designed to prevent entities with strong ties to adversarial nations from benefiting from U.S. clean energy tax credits and to reduce U.S. reliance on these nations—specifically China, Russia, North Korea, and Iran, which are defined in statute as “covered nations”—for critical components in the clean energy supply chain, such as solar panels and electric vehicle (EV) batteries.

  2. President Trump delivers remarks at an event on the “One Big Beautiful Bill Act.” Source: White House

  FEOC and related “prohibited foreign entity” restrictions represent a significant shift in how clean energy projects qualify for federal tax incentives. Building on IRA and modified by OBBBA, these rules can deny technology‑neutral tax credits—including the Section 48E Investment Tax Credit, Section 45Y Production Tax Credit, and Section 45X Manufacturing Tax Credit—to power plants, energy storage projects, and U.S.‑made products that have too much ownership, control, or critical content traceable to entities linked to covered nations. Compliance analyses focus on several dimensions, including whether the taxpayer is a specified foreign or foreign‑influenced entity, the share of project or product costs attributable to covered‑nation equipment and materials, and whether contractual or financing arrangements give a prohibited foreign entity effective control or “material assistance” in the project. “We may hear a lot more about FEOC in 2026,” Jim Nutter, managing director with Stout, told POWER. “Developers have tried their best to prepare, but the new part of the law and its interpretation might be hard to navigate in practice.” Luther suggested uncertainty around the requirements remains. “For 2026, everyone is looking at FEOC to see where the rules end up,” he said. The stakes are high. For certain technology‑neutral investment tax credits, projects can face full or substantial recapture if, within 10 years of being placed in service, the taxpayer makes payments that give specified prohibited foreign entities effective control or otherwise trigger the new ownership and “applicable payment” restrictions, depending on how the U.S. Department of Treasury ultimately interprets those rules. As the industry awaits clearer Treasury guidance, many developers are working to start construction before key effective dates to take advantage of grandfathering provisions, while also navigating tariff exposure and trying to secure FEOC‑compliant supply chains in a constrained and rapidly evolving market.

  Supply Chain Issues Persist, but Some Improvement Seen

  Supply chain disruptions have plagued the power industry since at least 2020 when the pandemic upset the normal flow of products, creating cascading delays in everything from transformers to turbines to solar modules. For years, developers have grappled with extended lead times, price volatility, and equipment shortages that stretched project timelines and squeezed margins. Brandy Johnson, Chief Technology Officer with Babcock & Wilcox (B&W), suggested the problem was lingering long before the pandemic. “The power industry supply chain has retracted because power build-out in the U.S. has been so minimal for so long. But now, especially due to growing demand from AI and data centers, we have a projected build-out in the U.S. energy sector for more generation capacity in a shorter period of time than we’ve potentially seen in our lifetimes,” she told POWER. “But the lack of manufacturing capacity means supply chains are really showing weakness in their ability to meet the industry’s needs,” Johnson said. “We’re seeing very long lead times on gas turbines, for example, because demand is greatly outpacing the rate at which they can be manufactured. The capacity to build pressure parts for boilers in the U.S. also is diminished, which means suppliers have to look overseas. It’s not just one thing. It’s all of the pieces of the supply chain that are affected.” Still, it’s not all doom and gloom—the pressure is lessening in some areas. “We’re seeing some easing in the supply chain for basic construction materials, such as steel, which was a major bottleneck during the pandemic,” said Montgomery. “However, more complex materials are still facing delays, particularly those related to transmission line components and high-voltage equipment. High-voltage direct-current cables are an example of procurement that can take up to 24 months to source. The North American Electric Reliability Corp. states that the lead time, the wait between placing an order and the product being delivered, has reached around two years, with large power transformers [Figure 3] taking up to four years.”

  3. The demand for transformers has led to major investments by manufacturers such as Siemens Energy, which announced in September 2025 it was investing about €220 million to expand its transformer factory in Nuremberg, Germany. In 2024, the company invested $150 million to expand operations at its transformer factory in Charlotte, North Carolina. Courtesy: Siemens Energy

  “We will begin to see supply chain … challenges easing for sectors tied to electrification or distribution level, but continue to see long lead times for power transformers,” said Charles Murray, CEO and co-founder of Switched Source. “New investments in domestic transformer manufacturing at the distribution level are starting to have an impact, but do not expect prices to come back down for transformers or cables.” “There have been major investments in domestic manufacturing of grid-critical equipment,” Tagare said. “In fact, National Grid is investing £35 billion over the next six years to help strengthen and unlock long-term supply chain capacity and skills across England and Wales.” Luther suggested developers need to consider all of their options when sourcing components, not simply focus on one possibility while belaboring its lack of availability. “Everyone is talking about the backlog on utility-scale gas turbines, but there’s over 25 GW in annual global production of sub-100-MW generators that could be used,” he noted. “It’s a significant amount of smaller-scale generation that no one is talking about.” Johnson also suggested developers should think about alternatives when an originally contemplated form of generation is unavailable. “Especially with data centers, B&W is trying to offer alternatives to build new power generation capacity when supply chain issues limit developers’ options,” she said. “For example, if a developer can’t get gas turbines for a combined cycle plant, we have alternatives that can help them get their plant online much sooner. Alternative technologies are coming to light because they may have supply chains that are more readily accessible.” Another way companies can find a workaround for supply chain issues is through standardization. “One of the things B&W is doing to address this is more standardization in our boiler and plant designs,” Johnson said. “We’re telling customers that we can speed up the design and construction of a plant if we do less customization. With power producers wanting to bring projects online quickly, speed through standardization is critical.”

  The T&D Dilemma

  Even as supply chain constraints are overcome, or accounted for, a more fundamental bottleneck threatens to limit how quickly new generation can actually reach customers: the transmission and distribution (T&D) infrastructure itself. Interconnection queues have swelled to historic levels, with gigawatts of ready-to-build projects waiting years for grid connection studies and upgrades. While new solar farms and battery storage facilities can be constructed in months, the transmission lines needed to deliver their power often require a decade or more of planning, permitting, and construction. This growing mismatch between generation development timelines and grid infrastructure timelines has pushed T&D constraints to the forefront of industry challenges (Figure 4) for 2026 and beyond.

  4. Robert Blue, chairman and CEO of Dominion Energy, told attendees at the Data Center POWER eXchange event that his company had invested $2.1 billion last year in transmission—about 18% more than the year before—and it plans greater than $2.8 billion in annual transmission capital spending starting in 2027. Source POWER

  “Much of the focus on electricity growth is on new power generation, but power transmission and distribution constraints don’t seem to get as much attention in public discourse,” David Carter, industrials senior analyst at RSM US, told POWER. “Further, just like in generation, these transmission and distribution constraints need an ‘all-of-the-above’ approach. That includes not just new power lines, but expanding the capacity of existing ones by ‘reconductoring’ them with more modern cables and using technology solutions to make better use of what is already in place. In a time where permitting is such a challenge, having options to increase capacity without a years-long process to obtain new rights of way will be key.” “Queue reform and transmission buildout are make-or-break for the [solar] industry’s ability to scale sustainably,” said Vargas. Mark Feasel, vice president of Sales with Mission Critical Group, suggested limitations in power delivery could force developers to find other options. “Grid constraints will push organizations toward rapid adoption of distributed generation and on-site power strategies that bring energy closer to the load this year,” he said. “As data centers consume increasing portions of available capacity, hospitals, manufacturers, and other critical infrastructure operators will be forced to rethink how they secure not just reliable power, but resilient power, when the utility can no longer meet their needs.” Luther suggested self-generation is already a growing trend. “A lot of hyperscalers are looking to bring their own generation and go behind the meter to avoid the long lead times of getting grid connected, which means the load might not materialize on the grid but off of it,” he said. “At the same time PJM, MISO [Midcontinent Independent System Operator], and SPP [Southwest Power Pool] are all accelerating queue timelines for dispatchable assets. So, if both of these trends emerge, its very likely these markets end up being better supplied than they are today.” Meanwhile, leaders in Washington could also play a role this year. “The administration and Congress both seem to recognize the importance of permitting reform, and a streamlined process that reduces uncertainty would help power projects come online faster, with better project economics,” Carter noted.

  Closing Thoughts

  All of the leaders POWER spoke to expressed genuine optimism about the year ahead. Their confidence stems not from abstract hope but from tangible opportunities they see emerging in many areas. Yet, aside from the big talking points—AI, data centers, solar power’s future, supply chains, and infrastructure—three interesting items emerged. “There is a growing importance of collaboration across the utility-startup ecosystem. Too many promising technologies stall in the ‘pilot doom loop’ because utility leaders and innovators lack the shared frameworks, data access, and operational pathways needed to move from experimentation to real system impact,” Tagare said. “The most successful solutions emerge when utility leaders and startups collaborate on deployment models early—aligning on integration requirements, regulatory considerations, and measurable value from day one.” “With the increasing electrification of things such as transport, heating, and cooking, it’s important to recognize and plan for a shift in peak power from a summer afternoon to a winter evening as the heat gets going, dinner is cooking, and EVs are charging,” said Carter. “One leader for a distribution utility recently highlighted the importance of how ‘EVs come in gangs’ to neighborhoods as neighbors influence other neighbors on EV adoption, it can quickly overload local transformer capacity during peak periods without proper planning or load-shifting capabilities.” “Most people understand that storage is important for solar, but the reality is that it still does not always make financial sense. Making storage work requires careful project design, timing, and alignment with grid needs. That is why we are so focused on it. Projects that get storage right will capture real value, while others risk underperforming,” Vargas said. “Another misconception is that solar growth will stall once IRA incentives fade. Projects with strong underlying economics will continue moving forward, and developers who integrate storage effectively will lead the market.” The common thread running through these perspectives is clear: success in 2026’s power sector will belong to those who move beyond siloed thinking, whether that means utilities partnering earlier with innovators, grid planners anticipating neighborhood-level adoption patterns, or developers designing projects around actual grid value rather than subsidy availability. It’s a fascinating time to be involved in the power industry—the challenges are out there to be solved, and pioneering industry leaders are finding the answers.

  —Aaron Larson is POWER’s executive editor.

  https://finance.yahoo.com/news/meeting-moment-industry-leaders-chart-132018086.html