For the first two decades of the 21st century, electric load growth was relatively flat as efficiency gains offset economic growth and energy demand. Electric industry capital spending grew more moderately but increased sharply over the past decade as utilities retired legacy generation and invested in replacing aging infrastructure.

The historically moderate level of growth has now been displaced by three simultaneous, related forces:

• Aging Infrastructure and Reliability/Resilience Needs: A large share of baseload fossil generation is 40 years old or more, exceeding its initial design life.  Many have been retired or are planned to be replaced with newer technologies, including renewables. Additionally, transmission and distribution systems built for a different era are being modernized to incorporate new technologies and hardened against extreme weather, wildfires, and higher reliability expectations. 
• Policy and Market Changes: Electrification of transportation and heat, policy-driven emissions targets, and shifting wholesale market dynamics are changing both what resources are needed and where they must be delivered. 
• New Technologies: Solar, wind, battery storage, and grid technologies enable new operating capabilities and create new challenges to incorporate into the grid. Additionally, new computing and artificial intelligence technologies drive electric demand.  

Together, these factors are driving a significant increase in capital investments needed to replace aging infrastructure, modernize the electric grid, strengthen reliability and resilience to storms and wildfires, and meet growing energy demand. This has led to significant cost challenges, as the rate base has grown from capital programs and the replacement of aging generation, while inflation and supply-chain constraints increase costs and delay project delivery. 

The growth of data centers amplifies these pressures by adding a new, large, concentrated, and constant load profile. A data center is a specialized facility that houses powerful computer systems and equipment needed to run digital services. There are several common configurations of data centers, which differ by size, location, and function, and as such, the corresponding energy demand may differ:

• Hyperscale/Cloud: Large campuses operated by major cloud providers. 
• Enterprise: Owned by and used for a single organization.
• Colocation: Shared facilities with multiple tenants.
• Edge: Small data centers located near end-users to reduce latency. 
• Modular: Pre-fabricated, portable units. 

A single hyperscale facility can exceed 1,000 megawatts (MW), an order of magnitude larger than those of traditional large-load industrial customers. As a result, the pace and scale of data center development will have material implications for local transmission and distribution of infrastructure and resource adequacy needs. Developers often operate on aggressive, quarter-based speed-to-market timelines that often conflict with the realities of interconnection studies, permitting, equipment lead times, and construction schedules. Consequently, the true cost to serve is frequently driven not only by peak MW but also by reliability requirements and the ability to deliver under contingent conditions.

The pace and scale of data center development are compounding long-standing challenges facing the electric industry. The growing demand for power has driven up costs across the supply chain; for example, the cost of new gas-fired combined-cycle power plants has increased 2.5 times over the past three years. In response, utility commissions have considered and implemented novel regulatory approaches, including large-load tariffs, minimum contract terms, and exit fees, to mitigate the risk that electric customers bear the cost of development. At the same time, data center developers and utilities have used innovative agreements and technologies to bring scale to market as quickly as possible.

Even when the preferred resource path is clear, the execution of generation and transmission projects is increasingly constrained by long lead times. Interconnection studies and required network upgrades can take multiple years, particularly in congested areas with large queue backlogs. Major equipment — especially high-voltage transformers and switchgear — often has long procurement windows, and transmission routing and approvals can further extend schedules. As a result, “Can we procure and permit it?” is now just as determinative as “Can we finance it?” for both generation and delivery.

Long lead times are compounded by labor supply limitations. The industry is competing for a limited pool of specialized trade and engineering talent needed for substation operations, protection and controls schemes, system commissioning, and high-voltage construction, while simultaneously executing large reliability, resiliency, and asset-replacement programs. Meanwhile, affordability is a binding constraint: rising bills heighten regulatory scrutiny and public concern, which can slow approvals, reshape cost allocation, and influence the pace at which capital programs can scale.

A range of supply and delivery options can help address near‑ and long‑term capacity and reliability needs. Each option, however, involves distinct tradeoffs around timing, cost, risk, and policy alignment. These approaches will be explored in greater depth in future blog posts. 

• Existing Generation Life Extensions: Life extensions of existing assets can provide near‑term capacity and reliability with fewer greenfield development hurdles. Extending nuclear units preserves large volumes of firm, carbon‑free energy but typically requires major capital programs, stringent regulatory oversight, and disciplined outage and risk management. Extensions of fossil units can help bridge reliability needs where policy allows, though they face emissions constraints, fuel and maintenance risk, and potential stranded‑cost exposure.
• Gas-Fired Generation: Gas‑fired resources—including combined‑cycle units and simple‑cycle peakers—can add dispatchable capacity and essential grid services on comparatively practical timelines. However, siting and air permitting, fuel infrastructure availability, and long‑term decarbonization policies introduce material uncertainty around feasibility and utilization. 
• Solar and Wind: Solar and wind resources can scale quickly to deliver low‑cost energy, but their variability and deliverability constraints limit their ability to provide firm capacity on their own. Maintaining reliability requires pairing these resources with transmission expansion, flexible generation, storage, and careful interconnection planning. 
• Small Modular Reactors (SMRs): SMRs offer the potential for firm, carbon‑free baseload generation that could align well with large, steady loads such as data centers. Key challenges include first‑of‑a‑kind economics, licensing timelines, supply‑chain maturity, and public acceptance. 
• Battery Storage: Battery storage is increasingly effective for short‑duration balancing, ramping support, and peak shaving. However, cost, materials constraints, and the limited duration of most current deployments restrict its ability to fully replace firm capacity without complementary resources. 
• Pumped Hydro Storage: Pumped hydro can provide large‑scale, long‑duration flexibility and significant reliability value, but deployment is constrained by geography, permitting complexity, and long development timelines. 
• Hydro Resources: Hydro resources—including imports and large hydro such as Canadian supply—can provide firm energy and seasonal balancing benefits. Feasibility depends on transmission buildout, market integration, and cross‑jurisdictional coordination.

North America’s power market is entering a period of accelerated change, as surging demand, aging infrastructure, and evolving technologies reshape how electricity is planned, delivered, and regulated. Future posts will explore solutions in greater depth, examining their tradeoffs, implementation challenges, and implications for reliability, affordability, and speed-to-market in an increasingly constrained electric system.