US Power Generation Dashboard

This dashboard provides comprehensive data on United States electricity generation by source, installed capacity, and demand forecasts through 2030. As artificial intelligence and data center expansion drive unprecedented power demand growth, nuclear energy's role as a reliable, carbon-free baseload source is becoming increasingly critical to grid stability.

1,200 GWTotal US Installed Capacity
95.5 GWNuclear Capacity
470 GWCurrent Demand
+47 GWAI Demand Growth

Current US Power Generation by Source

The United States generated approximately 4,243 billion kilowatt-hours (kWh) of electricity in 2023. Natural gas is the largest single source, accounting for roughly 43% of total generation, followed by renewable energy sources and nuclear power. Nuclear remains the largest source of carbon-free electricity, generating more clean power than all US wind and solar installations combined.

Energy SourceGeneration (GW typical)Share of TotalCarbon-FreeCapacity Factor
Natural Gas~190 GW~43%No55–60%
Nuclear~95 GW~18–19%Yes92–93%
Coal~70 GW~16%No40–55%
Wind~55 GW~10%Yes28–35%
Hydropower~33 GW~6%Yes35–45%
Solar (Utility + Distributed)~25 GW~4%Yes22–28%
Other (Biomass, Geothermal, Oil)~12 GW~3%MixedVaries

Nuclear Energy's Contribution to US Electricity

Nuclear power plants in the United States generated approximately 778 billion kWh of electricity in 2023, representing about 18.2% of total US electricity generation. This makes nuclear the single largest source of carbon-free electricity in America — more than wind, solar, and hydropower combined. The 93 operating US reactors achieve an average capacity factor of over 92%, far exceeding any other large-scale generation technology.

Key Fact: Nuclear power generates more carbon-free electricity in the US than all wind farms and solar installations combined. The 95 GW nuclear fleet produces as much clean energy as roughly 300 GW of installed solar or 270 GW of installed wind capacity would at their typical capacity factors.

AI-Driven Power Demand Growth

The rapid expansion of artificial intelligence workloads is creating extraordinary new electricity demand that threatens to outpace current grid capacity. Large language models, generative AI applications, and AI inference engines require massive, continuous computing power — and computing power requires electricity.

Data Center Power Consumption Projections

YearUS Data Center Power DemandAI PortionGrowth from 2024
2024 (baseline)~25–30 GW~5–8 GW
2025~35 GW~12 GW+40%
2026~45 GW~18 GW+80%
2027~55 GW~25 GW+120%
2028~65 GW~33 GW+160%
2029~68 GW~40 GW+190%
2030 (projected)~72 GW~47 GW+230%

Why AI Demands Baseload Power

Unlike traditional computing workloads that can be distributed across time, AI training runs and inference serving are continuous, time-sensitive processes that require guaranteed power 24 hours a day, 7 days a week. A single large AI training cluster can consume 50–100 MW of continuous power — the equivalent of powering a small city. Power interruptions during training can waste weeks of computation and cost millions of dollars.

This need for reliable, uninterruptible baseload power makes nuclear energy an ideal partner for AI infrastructure. Nuclear plants maintain capacity factors above 90% and are not subject to weather-related variability. Unlike wind and solar, nuclear plants can be co-located with or near data centers, reducing transmission costs and improving reliability.

Tech Company Nuclear Power Agreements

Leading technology companies are signing long-term power purchase agreements directly with nuclear operators to secure carbon-free electricity for AI infrastructure:

CompanyNuclear AgreementPower VolumeAnnounced
MicrosoftThree Mile Island Unit 1 restart (Constellation Energy)835 MW, 20 years2023
AmazonSusquehanna Nuclear Plant (Talen Energy)960 MW capacity2023
GoogleKairos Power SMR (multiple sites)500 MW by 20302024
MetaNuclear RFP for US data centers1,000–4,000 MW2024
OracleSMR-powered data campus~1,600 MW2024

US Grid Capacity and Demand Gap

The United States currently has approximately 1,200 GW of total installed electricity generation capacity, with peak demand reaching around 770 GW during extreme weather events. However, the rapid growth of AI data centers is expected to create a significant new demand gap by 2030 that cannot be filled by intermittent renewable sources alone.

Projected Demand Gap Analysis

Metric202420272030
Total US Power Demand (avg)470 GW490 GW517 GW
AI/Data Center Demand25–30 GW55 GW72 GW
Nuclear Generation95 GW96 GW100+ GW
Baseload Gap (estimated)0 GW~15 GW~47 GW

Nuclear Energy's Role in Bridging the Gap

With US nuclear capacity currently at 95.5 GW and dozens of plants pursuing 20-year license extensions, nuclear energy will remain the cornerstone of US clean baseload generation through at least 2050. Additionally:

  • Three Mile Island Unit 1 added 835 MW back to the grid in September 2024
  • Palisades Nuclear Plant in Michigan is being evaluated for restart (~800 MW)
  • The Kemmerer Natrium SMR in Wyoming is under construction (345 MW, operable ~2030)
  • At least 6 other SMR projects are in the NRC licensing pipeline
  • The DOE's Advanced Reactor Demonstration Program is funding multiple SMR designs
  • Several existing plants are adding capacity through uprate projects

Power Generation Trends 2000–2024

Over the past two decades, the US electricity generation mix has shifted substantially. Coal's share has fallen from 52% in 2000 to under 16% in 2024 as plants retire due to economic pressure from cheap natural gas and increasingly competitive renewables. Nuclear's share has remained remarkably stable at 18–20% throughout this period, even as absolute generation increased slightly due to improved capacity factors and uprate projects at existing plants.

Source2000 Share2010 Share2020 Share2024 Share (est.)
Natural Gas16%24%41%43%
Nuclear20%20%20%18%
Coal52%45%20%16%
Hydropower7%6%7%6%
Wind<1%2%8%10%
Solar0%<1%3%4%
Other Renewables2%3%2%3%