ERCOT Faces Record AI Data Center Demand in Texas
We are closely tracking how AI data center demand is reshaping power infrastructure across Texas. This week, we observed a significant rise in large-load interconnection requests submitted to the Electric Reliability Council of Texas (ERCOT). The volume of new applications demonstrates how rapidly energy requirements are shifting as AI workloads scale. According to ERCOT, more than 230 gigawatts of requests have been filed in 2025, a dramatic increase compared to the 63 gigawatts recorded at the end of 2024.
Much of this demand is tied to new data centers designed to support AI training and inference operations. ERCOT reported that many individual sites now exceed one gigawatt in requested load, comparable to the output of a large gas-powered generation facility. More than 70 percent of these requests come from developers building high-density AI infrastructure, according to the System Planning and Weatherization Update.
Growth Trends Driven by AI Infrastructure
AI Data Center Demand Accelerates Grid Pressure
ERCOT records show that companies submitted 225 large-load requests through mid-November.. This figure represents meaningful growth from the 152 requests submitted across the 2022–2024 period. The second quarter alone saw 78 applications totaling 70,000 megawatts, highlighting how rapidly the landscape is evolving.
This pace of expansion continues to outmatch the timeline for transmission and generation development. ERCOT currently has 1,999 active generation interconnection requests totaling 432 gigawatts, with most projects focused on solar and battery capacity. Since last winter, the grid has added just over 11,000 megawatts of new resources.
Resource Adequacy and Infrastructure Planning
Utilities Evaluate Transmission Capacity for AI Data Centers
In preparation for this week’s ERCOT board meetings, executives outlined resource adequacy considerations tied to the surge in large-load applications. We noted that utilities must now evaluate both transmission capability and generation availability before approving new high-density industrial connections.
The scale of AI data center demand presents added complexity for future grid planning, especially as individual sites require accelerated assessments of infrastructure readiness. ERCOT highlighted that the timing of new resources will be a determining factor in the approval of large loads.
Texas Regulators Respond to AI-Driven Load Growth
New Standards Proposed for Large-Load Interconnection
The Public Utility Commission of Texas (PUCT) announced a draft proposal aimed at managing the rapid rise of large-scale energy users. Under this framework, any customer requesting 75 megawatts or more at a single location would need stricter coordination with utilities and ERCOT’s transmission entities.
We understand that this marks the first structured approach to handling the intersection of the state’s growing data center sector and the significant energy volumes required to operate AI infrastructure. The proposal seeks to reduce operational uncertainty for both developers and grid operators by establishing clear coordination procedures.
Outlook on AI Data Center Demand in Texas
As AI workloads expand, the energy landscape in Texas continues to transform at scale. Our ongoing analysis shows that load growth tied to AI combined with accelerated interconnection requests will require sustained evaluation of grid capability, resource adequacy, and regulatory clarity. We will continue monitoring ERCOT updates, industry filings, and infrastructure developments to provide accurate insights into this rapidly shifting environment.
