Commentaries at the San Antonio Report provide space for our community to share perspectives and offer solutions to pressing local issues. The views expressed in this commentary belong to the author alone.
As anyone who lost power during Winter Storm Uri or watched a neighbor’s home threatened by wildfire smoke drifting in from a drought-dried Hill Country knows, Texas must harden, upgrade and modernize its grid to meet basic safety and reliability standards. These investments are not optional. And they carry real risk of raising electricity bills for households and small businesses in a city where rate pressure is already a live political issue.
The debate in San Antonio right now centers on a startling set of numbers. The roughly two dozen data centers operating in far western Bexar County already consume around 324 megawatts of power. By the end of 2028, an influx of new centers could add another 2,700 megawatts of demand, and by 2033, that figure could top 3,300 megawatts. To meet that growth, CPS Energy expects to spend approximately $1.3 billion on transmission projects over the next five years, adding roughly one gigawatt of load-serving capacity.
Those are real numbers, and the conversations happening at San Antonio City Hall and CPS Energy’s boardroom about who pays for them, and how, are exactly the right conversations to have. City Council members have emphasized that they are not anti-data center, but anti-unplanned, unbuffered growth. We agree. But while policymakers debate the financing, San Antonio may be overlooking a faster, cheaper solution hiding in plain sight: using the grid Texans already paid for more effectively.
The financial case is striking. Based on GridCARE’s financial analysis, one gigawatt of AI data center load serviced by the existing system in San Antonio would generate roughly $123 million per year of new revenue after pass-through costs, with zero incremental system costs. That $123 million could cut system-wide rates by up to 4.4% (saving the average customer roughly $64 per year), justify up to $1.45 billion in new infrastructure investment without a rate increase, or both reduce rates and fund new system investment.
Think of the electric grid as San Antonio’s highway system. Interstate Loop 1604 and I-10 are the transmission lines. The neighborhood roads carrying power to homes are the distribution lines. Just as the freeways and the “fiesta stack” interchange aren’t gridlocked at 2 a.m. on a Tuesday, much of the electric grid sits underutilized most of the time. Research from Stanford University on the Western grid and emerging analysis of ERCOT show that true system-wide peak conditions, the equivalent of rush hour with every lane blocked, occur for only a small fraction of hours per year. The rest of the time, substantial capacity sits idle.
This matters because the costs to maintain poles, wires, substations and control systems are borne by native customers. Adding large energy users to the existing system, rather than waiting years to build entirely new infrastructure, spreads those fixed costs across a broader customer base and reduces upward pressure on rates. Blocking or delaying load growth has the opposite effect: it concentrates rising costs onto a stagnant customer base — which is exactly what San Antonio families and small businesses cannot afford.
Importantly, the tools to make this work already exist in Texas law. Texas Senate Bill 6 pairs mandatory curtailment requirements with a voluntary demand response program under which large loads of 75 megawatts or more can ramp down or switch to backup generation at utilities’ request. In essence, the Legislature has already codified the principle: data centers can connect sooner if they agree to step back during the rare hours the grid is under maximum stress — the electrical equivalent of agreeing to avoid the freeway during rush hour.
We’ve seen this model work in practice. In Hillsboro, Oregon, Portland General Electric partnered with GridCARE to apply advanced artificial intelligence and engineering analysis to identify where power lines and equipment were underused, and when large loads could safely reduce consumption during the few peak hours when the grid is stressed. The result: the utility expects to energize more than 400 megawatts by 2029, with 80 megawatts of new data center load coming online in 2026, without waiting years to build new power lines and without putting reliability at risk.
This is not a giveaway to the tech industry. It is a pragmatic affordability strategy for the people of San Antonio. San Antonians have already invested in the infrastructure underneath this city. Using it more effectively honors that investment while providing time and resources to plan and build the next wave of upgrades affordably.
Texas has a head start that California does not. ERCOT’s market structure, combined with the flexibility framework established by SB 6, gives San Antonio a real opportunity to model responsible data center growth for the rest of the country. The San Antonio City Council is already asking the right questions. The missing piece is a technical framework that translates “flexibility” from a policy aspiration into verified, engineering-backed rules that CPS and its customers can rely on. GridCARE’s framework complements ERCOT’s Aurora Study findings by providing a formal, engineering-based methodology to qualify, qualify and activate flexibility as certifiable capacity at the point of interconnection.
Done right, San Antonio can attract the data center investment that diversifies its tax base, modernizes its grid, protects families from the next blackout and keeps electricity affordable, all at once. The path forward is not to block load growth, but to govern it intelligently.
