The North American Electric Reliability Corporation (NERC) forecasts that peak electricity demand in the US will grow by 224 gigawatts in summer and 246 gigawatts in winter over the next decade. This shift is driven by the continuous power consumption of AI data centers, which eliminates the traditional seasonal lulls in grid stress.
The PJM Interconnection, a regional grid operator in the eastern US, anticipates an additional 30 gigawatts of demand from data centers by 2030. This surge has led to increased congestion costs, as transmission lines struggle to deliver power, forcing reliance on more expensive local generation sources.
On October 19, 2026, the Federal Energy Regulatory Commission (FERC) ordered six regional grid operators to justify their current connection tariffs. These rules have created significant bottlenecks, with utilities reporting wait times of 5–10 years for new grid connections, effectively limiting the expansion of data center capacity.
In response to these challenges, the Department of Energy authorized emergency operations for specific generation units in the PJM region during the fall of 2026. This measure provides operators with greater flexibility to manage demand spikes and maintenance requirements.
