The grid operator that keeps the lights on for roughly 27 million Texans says electricity demand across its territory could more than quadruple by 2032, driven almost entirely by a wave of data center projects tied to the artificial intelligence boom. But in an unusual move, the regulators overseeing that operator are pushing back, warning that the projections may be built on shaky ground.
The Electric Reliability Council of Texas published a seasonal forecast projecting that peak demand could reach approximately 367,790 megawatts by 2032, up from a system that has historically peaked near 85,000 MW during the hottest summer days. The bulk of that growth comes from a single source: data centers, which ERCOT expects to consume roughly 7,400 MW in 2026 before ballooning to nearly 230,000 MW six years later.
That data center figure alone would represent more than twice the grid’s current all-time peak load. Building it out would require constructing power-hungry server farms at a pace no U.S. electricity market has ever attempted.
Regulators flag the numbers
The Public Utility Commission of Texas, which oversees ERCOT, reviewed the forecast and sent it back for revision, concluding the demand figures may be significantly inflated. That step is uncommon and signals real concern that ERCOT’s planning models are absorbing interconnection requests that lack firm financial backing or realistic construction timelines.
The problem is structural. ERCOT’s forecasting process counts what developers say they want to build, not what they can prove they will build. A company can file an interconnection request for 5,000 MW of capacity while still shopping for financing, land, or customers. There is no mechanism requiring applicants to demonstrate secured funding, signed power purchase agreements, or active construction before their load request gets folded into the grid operator’s demand outlook.
According to Utility Dive’s reporting, ERCOT officials have acknowledged this limitation publicly, emphasizing that the upper-end scenario is a planning case meant to capture the outer boundary of potential growth, not a prediction. But headlines and political reactions have often treated the 367,790 MW figure as a baseline expectation, creating a gap between technical nuance and public understanding that regulators are now trying to close.
Data center demand and the interconnection queue
Texas has become one of the most attractive markets in the country for large-scale data center development. The state offers relatively cheap land, no state income tax, a deregulated electricity market, and proximity to fiber-optic backbone routes. The explosion of generative AI has accelerated the trend, as training and running large language models requires enormous computing power, and the facilities that house those servers consume electricity around the clock at intensities that dwarf traditional commercial buildings.
That context makes the broad direction of ERCOT’s forecast plausible. Data center demand is growing, and Texas is a top destination. The question is not whether load will increase but by how much and how quickly. However, neither ERCOT nor the reporting outlets covering the forecast have published a breakdown showing which specific companies or projects account for the largest share of the projected 230,000 MW. Without that detail, independent observers cannot assess whether the demand is concentrated among a handful of well-capitalized firms or spread across dozens of early-stage ventures whose plans may never materialize.
Why the gap matters for Texans
Grid planning decisions flow directly from demand forecasts. If ERCOT’s projections hold even partially, Texas will need to build generation capacity at a pace that dwarfs any previous expansion in the state’s history. New natural gas plants, solar farms, battery storage facilities, and potentially nuclear units would all be on the table. Transmission planners would need to connect remote renewable resources to urban load centers while reinforcing local networks near major data center clusters. The investment required would be measured in tens of billions of dollars.
But if the forecasts prove inflated by a factor of two or three, the state risks approving infrastructure that becomes underutilized. Ratepayers would bear the cost of those stranded assets through higher electricity bills for years. Overbuilding could also crowd out investments in energy efficiency programs or distributed resources that might offer more flexible ways to meet actual demand.
Texas has recent, painful experience with the consequences of inadequate grid planning. The February 2021 winter storm that knocked out power to millions of homes exposed how thin the margin between supply and demand had become. The political and financial fallout from that crisis still shapes how regulators, lawmakers, and ERCOT approach reliability questions. Overestimating demand carries real costs, but underestimating it carries potentially catastrophic ones.
Revised forecast still pending as of May 2026
As of May 2026, ERCOT has not released detailed documentation of its revised methodology or indicated when a corrected forecast will be published. The regulatory feedback loop between the grid operator and the Public Utility Commission remains ongoing, and no timeline for a final, revised projection has been made public.
All of the load-bearing numbers in this article originate from ERCOT’s seasonal assessment as reported by Utility Dive and Fox 4. The raw ERCOT seasonal forecast document has not been independently located or linked in available reporting. No independent dataset or third-party audit of the figures has been published. Readers should treat the specific megawatt figures as ERCOT’s stated projections under its current methodology, not as independently validated predictions of what Texas will actually need.
For Texas businesses and households, the practical stakes are straightforward. Grid reliability during summer heat waves and winter cold snaps depends on whether ERCOT and its regulators can close the gap between speculative demand filings and realistic load growth. The revised forecast will shape investment decisions, rate cases, and reliability standards across the state for years to come. What regulators decide to believe will determine not just how many data centers get built, but how much Texans pay to keep the grid running.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
