The North American Electric Reliability Corporation (NERC) guidance emphasizes that AI should serve as a decision-support tool rather than an autonomous controller.42 In line with this, the industry is beginning to put safeguards in place—such as model registries, audit trails, and risk controls. Human oversight (human-in-the-loop) is essential to ensure strong governance. Resource adequacy rules are expected to start recognizing flexible load paired with four-hour batteries as a dependable capacity resource. Some utilities and regulators now require hyperscalers to share costs, provide telemetry, and demonstrate flexibility for faster interconnection. Once viewed as inflexible mega-loads, hyperscalers are now potential operational partners.24 In 2025, US data centers emerged as one of the fastest-growing sources of electricity demand.
While O&M spending has “benefitted from productivity gains driven by smart-grid investment,” O&M costs rose for 31 of 42 companies that reported the line item. However, operations and maintenance (O&M) expenses rose 4.7%, reflecting “essential safety and reliability needs” even after years of pandemic-era cost discipline. “Total Electrical Generation Cost fell 6.2% as the average cost of natural gas for electricity generation was 18% lower year-to-year,” EEI wrote, adding that higher output from renewable generation “with zero fuel cost” further constrained aggregate fuel costs. Of course, prospects for higher demand growth come from more than AI and data centers. In 2024, notably, state regulators and electric companies http://romj.org/2025-0302 proposed new tariffs to address growth issues in several states, including California, Georgia, Indiana, Nevada, Ohio, Oregon, and West Virginia.
Manufacturing accounts for another 30 gigawatts, and the remainder comes from building and transportation electrification, mining, and other loads. Much of the higher estimate is due to data center development, which is expected to account for 90 gigawatts of the new peak demand growth. Utilities also benefit from a modernized grid, including improved security, reduced peak loads, increased integration of renewables, and lower operational https://open-innovation-projects.org/blog/where-open-source-software-thrives-exploring-its-impact-and-potential-across-industries costs. These solutions can enhance grid flexibility, ensure greater reliability and help reduce overall investment costs, and are already being applied across regions. These solutions can increase grid flexibility, ensure greater reliability and help reduce overall investment costs by relieving different types of binding operational constraints and improving utilisation of existing assets. The AI age is expected to require scaling data centers, grid capacity, and supply chains.
Support Independent Texas News
The approach reflects a comprehensive system upgrade required to accommodate both increased power flows and changing generation patterns. Distribution infrastructure, however, continues commanding the largest investment category within the projected $207.9 billion in functional capital expenditures for 2025, followed by generation, transmission, and gas-related infrastructure investments. Combined, the 579-GW pipeline represents 46.3% of the nation’s existing grid capacity, and it potentially positions utilities to expand the power system by nearly half its current size within the next half-decade. EEI added that “these investments will ensure reliability and enable companies to provide electricity at the lowest-possible cost for customers.”
- Partnership supports WeaveGrid’s expansion beyond EVs and creates new utility program opportunities for SolarEdge battery owners
- Originally designed to manage EV charging at scale, DISCO has evolved into a core utility platform for aligning customer energy use with distribution-level grid needs.
- If a utility sees conditions building near vulnerable equipment, it can move from watching the storm to preparing for a specific risk.
- The appendix also evaluates technologies including virtual inertia, grid-forming inverters and synchronous condensers as planners assess how a changing generation mix could affect future system reliability.
- Many utilities are now also planning new natural gas generation capacity in addition to renewable capacity to keep pace with growing demand.”
- At the same time, 112 GW of retirements are scheduled through 2029—about half from coal and one-third from natural gas (56 GW and 37 GW, respectively).
Segments we support
- Data center site selection is now fundamentally governed by power availability, forcing a geographic diversification away from historically dominant but power-constrained primary markets toward secondary and tertiary regions with available grid capacity.
- OpenAI CEO Sam Altman, in a letter to the White House’s Office of Science and Technology Policy, recommended that the United States add 100 gigawatts of energy production capacity a year to stay competitive in the artificial intelligence (AI) race.
- A lack of grid capacity is emerging as a critical bottleneck in many regions, driving higher levels of congestion and slowing the deployment of new electricity generation, storage and demand.
- This is because power consumption has grown by 20 gigawatts from the previous winter, without the supply keeping up.
- “Data centers must operate in ways that reduce costs for residential electricity customers, do not drain water needed for our communities, and take into consideration the needs of our neighborhoods,” said Governor Abbott.
- But AI data centers whose workloads—though extremely energy-intensive—are more intermittent, may be able to adjust without disrupting essential tasks, trial data suggests.
Almost 70 gigawatts (GW) of new solar generating capacity projects are scheduled to come online in 2026 and 2027, which represents a 49% increase in U.S. solar operating capacity compared with the end of 2025. The three main dispatchable sources of electricity generation (natural gas, coal, and nuclear) accounted for 75% of total generation in 2025, but we expect the share of generation from these sources will fall to about 72% in 2027. While the number of PV systems interconnected to the grid has increased significantly over the last decade, only recently have PV systems been installed in major metropolitan areas and tied to electric distribution secondary network systems (networks).
AI will enable real-time optimization of dispatch, asset performance, and outage response, while stronger supply chains support infrastructure. Utilities that set the pace will be those that embed financial, operational, and digital flexibility into their playbooks—delivering capacity where and when it’s needed while safeguarding affordability. Key inflection points will likely include the repeal or phaseout of certain clean energy tax credits, evolving tariffs, new foreign entity of concern–related procurement requirements, and the integration of AI into core operations. Utilities that can blend self-financed projects with partnerships, securitized financing, and outcome-based models will likely deliver more capacity, faster, without overburdening customers.
DOE emergency orders are incurring additional costs. What are the benefits?
Beyond non‑firm connections, standard congestion‑management tools and robust regulatory frameworks that support the co‑location of multiple power plants and battery energy storage systems (BESS) at a single connection point can further ease grid constraints. This built‑in flexibility helps unlock additional hosting capacity by allowing more assets to connect before major grid reinforcements are completed. Complementary measures, such as grid‑enhancing technologies and regulatory adjustments, can unlock near‑term grid capacity, delivering net system-wide economic benefits. This is especially relevant, as grids are built to serve peak demand, but often have substantial unused capacity during non‑peak periods.
New federal data shows solar and battery assets will comprise nearly 80% of all new utility-scale power additions planned for the U.S. electric grid this year. But unit one was already offline because of planned maintenance and refueling, according to EDF, the plant’s operator. So some planned outages happen in the spring and into the summer, which is affecting the supply right now. In the US, planned outages for maintenance and refueling tend to come in the spring and fall when demand falls below the summer and slightly smaller winter peaks. Generally, grid operators plan maintenance and outages at power plants around expected peaks in demand.
Grids are emerging as a bottleneck for connecting supply, demand and storage
More than 4,000 data centers now operate in the U.S., according to one estimate, with an additional 3,000 planned or under construction. The Edison Electric Institute, which represents investor-owned electric utilities, said FERC’s order builds on regional and state processes already underway while “supporting flexibility and innovation.” But that order can do little to address the tightening energy supplies that are driving up electricity bills in some areas and raising warnings of blackouts as the construction of data centers outpaces the speed of new power plants coming online to serve them.
Smart MRO: Using Technology and Assessment to Enhance Power Plant Performance
Cryptocurrency, hydrogen fuel plants and other large loads were also cited as contributing to increased demand, though NERC noted that additional factors are at play. Accenture named a Leader in IDC MarketScape for Energy Transition Services 2025, recognized for digital and AI-driven solutions supporting utilities clients. Empower your workforce with future-ready skills and AI-driven learning to unlock performance, adaptability and growth. A key example is the $20 billion partnership between Google, Intersect Power, and TPG to develop new clean energy projects specifically to create a dedicated power supply for future data centers. The sudden shift is driven by the exponential and unanticipated energy demand from AI workloads.
Winter is an especially risky time because solar generation is available for fewer hours, and battery operations may be affected. This is because power consumption has grown by 20 gigawatts from the previous winter, without the supply keeping up. In response, state lawmakers approved a measure requiring new data centers to disconnect from the grid during periods of peak stress — a mandate whose specific rules are still being developed by ERCOT and state regulators. Over the past year, “large-load” forecasts within the Electric Reliability Council of Texas (ERCOT) have nearly quadrupled, signaling a potential doubling of the grid’s peak demand. As reported by Canary Media, Dominion Energy is proposing new gas-fired power needed to maintain reliability due to data center growth in Virginia — the world’s largest data center hub. Via Canary Media, higher load forecasts have driven up capacity costs — the prices paid to power plants and other grid resources to meet peak grid demand — from $2.2 billion in 2023 to $14.7 billion in 2024, and to $16.1 billion in the summer’s capacity auction.
