U.S. electricity use grew steadily through the 1980s, 1990s, and early 2000s, then entered roughly 15 years of nearly flat consumption, a plateau the U.S. Energy Information Administration (EIA) says has now ended, with average growth of 2.1% per year over the most recent five years. In its Annual Energy Outlook 2026, EIA projects consumption to keep rising through 2050 at annual rates ranging from 0.9% to 1.6%.

Other forecasts, including those from the National Electrical Manufacturers Association and Energy Solutions by McKinsey, project even faster demand acceleration. On a national power system, each percentage point represents tens of billions of kilowatt-hours (kWh) and requires generation, fuel, transmission, substations, and operating reserves.

This is not a replay of the 20th-century buildout. Today’s growth is arriving on a grid simultaneously undergoing decarbonization, electrification, and climate adaptation, a combination that makes the engineering challenge far broader than simply adding megawatts. 

Artificial intelligence (AI) and data centers are the most visible catalysts. The Department of Energy’s (DOE) 2024 United States Data Center Energy Usage Report found that U.S. data center electricity use had tripled over the preceding decade and could double or triple again by 2028, with EIA’s 2026 outlook projecting server electricity use rising from roughly 105 billion kWh in 2020 to about 446 billion kWh in 2050, and as high as 818 billion kWh in the high-electricity-demand case. DOE has also reported connection requests for hyperscale facilities in the 300- to 1,000-megawatt range, loads comparable to a major industrial complex or a mid-sized city, often sought on development schedules much shorter than those for new power plants or transmission lines.

Data centers are not typical commercial loads. They seek high load factors, high reliability, and rapid energization, and unlike most commercial customers, they do not largely disappear at night. A single hyperscale customer can therefore alter a utility’s peak forecast, transmission plan, fuel strategy, and resource adequacy position.

Utility integrated resource plans (IRPs) show how the industry is responding. The details vary by geography, regulation, fuel access, and policy, but the emerging pattern is a portfolio rather than a single winning technology.

• Duke Energy’s 2025 Carolinas Resource Plan retained five new combined-cycle units and increased planned combustion turbines to seven, alongside nuclear, renewables, storage, and grid investment.
• Georgia Power’s approved 2025 IRP authorizes investment in generation, demand-side programs, hydro fleet modernization, and transmission to support extraordinary state growth.
• SMUD’s zero-carbon planning emphasizes renewable additions, storage, customer programs, and the continuing reliability value of its Upper American River hydro system and available thermal resources during the transition.

The common message is more important than the differences: utilities are planning to build, procure, upgrade, and interconnect simultaneously. They need new energy resources, but also capacity, ramping capability, inertia, reserves, black-start capability, and transmission deliverability. The next grid will be assembled from complementary resources, not from a single technology used in isolation.

Hydropower is exceptionally well-positioned in that portfolio. Hydropower is often described as legacy infrastructure; a better description is that it is strategic infrastructure equipped for a new operating mission. Its familiar attributes, including renewable energy, long asset life, and low operating emissions, remain important, but its operational capabilities are becoming more valuable as the system changes.

A hydro unit can provide rapid ramping, black-start capability, and spinning reserve on short notice. Pumped storage can move large quantities of energy across hours or days, delivering services that short-duration batteries may not economically replicate.

The opportunity is not limited to new dams. Much of the work will occur at existing facilities: turbine and generator replacements, uprates, digital governors, excitation systems, protection and controls, condition monitoring, penstock and valve rehabilitation, switchyard upgrades, sediment and water-passage improvements, fish protection, seismic evaluations, spillway work, and dam safety modifications.

In many cases, modernization can add dependable capacity or flexibility without constructing an entirely new water project. The constraints, including variable hydrology, complex environmental requirements, and lengthy licensing, are real, but they are also precisely why experienced multidisciplinary engineering is more valuable, not less.

Natural gas generation is also reemerging as a central element of resource planning. Combined-cycle plants can provide efficient, dispatchable energy, while fast-start combustion turbines supply capacity and flexibility during peaks, renewable shortfalls, transmission outages, and extreme weather. EIA’s baseline outlook illustrates the scale: from 2025 through 2050, it projects roughly 431 gigawatts (GW) of cumulative gas capacity additions, more than the modeled solar or wind additions individually. This is a scenario result, not a prediction that every announced unit will be built, but it demonstrates how strongly the model values dispatchable capacity under sustained load growth.

In most utility resource plans, gas generation is part of the reliability architecture that enables higher renewable penetration, providing the multi-hour and multi-day backup that batteries and demand response cannot yet consistently supply. The engineering agenda includes high-efficiency combined cycles, dual-fuel capability, black-start systems, and designs that preserve future optionality. Utilities will also have to manage fuel-price exposure, emissions, permitting risk, and the possibility of tighter future carbon constraints. Those tradeoffs are precisely why disciplined planning and engineering matter.

The California Independent System Operator’s (CAISO) January 2026 large-load issue paper cites the California Energy Commission forecast of an additional 1.8 GW of data center load in the ISO balancing area by 2030 and 4.9 GW by 2040. More broadly, CAISO states that the 2025 IEPR forecast adds about 30 GW of peak load by the mid-2030s, including data centers, electric vehicles, and manufacturing.

California already operates a grid with high midday solar production and steep evening net-load ramps. Batteries have become a major part of the response, but hydro, imports, demand response, transmission, and flexible thermal generation continue to matter. Further, natural gas still accounts for the largest share of in-state capacity and electricity generation in California. And because gas plants frequently serve as the marginal generating resource, their production cost sets the market-clearing price. To a large extent, the price of natural gas directly determines the price of wholesale power.

While every region will have a different resource mix, load forecasting, generation procurement, interconnection, and transmission planning must be synchronized. A large load that can be permitted and constructed in two years cannot be served reliably by infrastructure that requires seven to 10 years.

Across every engineering discipline, the demand for power generation expertise is tangible.

• Civil and geotechnical teams are busy investigating foundations, embankments, and seismic risk.
  
• Mechanical engineers are designing and rehabilitating turbines, valves, and gas-turbine auxiliaries. 
• Electrical engineers are developing generators, protection systems, and interconnections.
• Controls specialists are integrating digital plant systems, cybersecurity, and advanced monitoring.
• And environmental professionals are helping projects navigate water quality, air emissions, and community impacts. 

The scale of the task is daunting, but that is what makes it exciting. A rehabilitated hydro unit, a new combined-cycle plant, a pumped-storage development, or a transmission interconnection is not a short-lived product; it is the infrastructure of life!

That infrastructure is what GFT’s power generation engineers have spent more than 80 years helping to build and maintain. The work spans hydropower dam safety and rehabilitation, design of natural gas plants, turbine and generator upgrades, electrical and mechanical systems, and high-voltage transmission, alongside the civil, environmental, and regulatory coordination that each phase requires. That integration across civil, mechanical, electrical, controls, and environmental disciplines is what allows GFT to support projects from feasibility through construction, without losing the thread between planning and delivery. For utilities, owners, and developers thinking through the next investment in generation or grid capacity, GFT welcomes the conversation.