SB 6, ERCOT, and who pays when a data center joins the grid
In 2025, the Texas Legislature passed Senate Bill 6 — the "Large Load Interconnection" bill. It was written in direct response to data centers. The most important thing it does is not what most people think. It does not cap load. It does not require closed-loop cooling. It kicks the hardest question — who pays for new transmission — to the Public Utility Commission of Texas. That fight is ongoing, and the outcome will show up on your power bill.
The demand problem, in two numbers
Texas peak electricity demand in summer 2024 was about 85 gigawatts. ERCOT, the grid operator that serves about 90 percent of the state, published a 2025 capacity and demand forecast showing more than 150 gigawatts of new load requests in the interconnection queue over the next decade. Roughly half of that queue is data-center and large-industrial.
A gigawatt is a billion watts. Added together, the queued load is nearly twice today's peak. Not all of it will be built — some projects are speculative, some will lose out to others, and queue positions are traded — but the directional signal is real: demand growth that used to take 20 years to materialize is now being requested in five.
SB 6 was the Legislature's attempt to make that wave manageable. Whether it succeeds depends on rules the PUCT is still writing.
What SB 6 actually does
SB 6 has three moving parts. Each one matters for a different group of Texans.
1. A new "large load" category.
Any new interconnection request above 75 megawatts — roughly the load of a small city or a mid-size hyperscale data center — now falls into a separate review track. Applicants must disclose whether their request is firm (they have signed contracts and capital) or speculative (still shopping jurisdictions). ERCOT now has authority to prioritize firm requests and deprioritize speculative ones, reducing the queue gridlock that plagued 2023 and 2024.
2. Curtailment authority in exchange for faster queue access.
Large loads that accept "interruptible" status — meaning ERCOT can temporarily shed them during a grid emergency — get preferential queue treatment. This is a voluntary bargain. Data-center operators that accept curtailment flexibility move faster; those that demand 100 percent firm supply wait longer.
3. A cost-allocation rulemaking at the PUCT.
This is the big one. SB 6 directs the PUCT to open a rulemaking — now known as Docket 57579 — to decide who pays for the transmission upgrades needed to serve large new loads. That question was unsettled before SB 6 and remains unsettled now. SB 6 just forced it onto the commission's docket.
The cost-allocation question, in plain English
When a 500-megawatt data center interconnects, it rarely plugs into wires that already exist. New substations, transmission lines, and sometimes new generation have to be built. Those upgrades cost hundreds of millions of dollars each.
There are three ways that cost can be paid:
- The large load pays in full through a direct assignment of the upgrade cost. The operator's demand charges go up to cover it.
- All ratepayers share the cost because, in theory, a bigger, stronger grid benefits everyone. Residential bills go up modestly; the large-load bill stays lower.
- A hybrid — the direct-assignment portion covers the substation and tap; the shared portion covers the broader transmission backbone.
In practice, every state that has faced this question has chosen some form of hybrid. The fight is over the split. Industry files for a heavier shared portion. Consumer advocates and some utilities file for a heavier direct-assignment portion. The PUCT commissioners — three of them, appointed by the governor and confirmed by the Texas Senate — decide.
Why this shows up on a residential bill
If the PUCT rules that 60 percent of large-load transmission upgrades are "systemwide benefit" and shared across all customers, that cost is recovered through the transmission cost recovery factor on every Texan's bill. Analysts' estimates vary, but a 2025 ERCOT-area modeling exercise suggested that a heavily-shared allocation could add $15 to $60 per year to an average residential bill, depending on how much new load interconnects.
Where other states have landed
Texas is not first. Virginia, Ohio, Georgia, and Arizona have all run a version of this fight. A few patterns are worth knowing.
- Virginia (Dominion Energy territory) has approved hybrid allocations that put most direct-tap costs on the data-center operator but socialize the backbone. The Virginia State Corporation Commission is now reviewing whether to tighten that further; filings are public.
- Ohio took a more aggressive direct-assignment approach in a 2024 case involving AEP Ohio, ordering large loads to post collateral for their share of upgrades before interconnection.
- Georgia has allowed Georgia Power to pass on some large-load costs to residential customers, which triggered regulatory backlash and an ongoing review.
The Texas answer is not yet written. PUCT Docket 57579 is open for public comment, and the record is where this decision is being built.
What SB 6 does not do
Several things the public often assumes are in SB 6 are not.
- It does not cap data-center load in Texas.
- It does not require cooling-method disclosure — that is a county or city decision, not a state one.
- It does not regulate groundwater. Groundwater is governed by local Groundwater Conservation Districts under Chapter 36 of the Texas Water Code.
- It does not change the Chapter 312 tax-abatement framework. That lives in the Texas Tax Code.
- It does not give counties new authority to reject data-center projects on power-grid grounds. County authority over industrial land use comes from Chapter 211 and Chapter 212, not from SB 6.
Those gaps are why the fight over data centers in Texas is not a single fight. It is several fights happening at once — in the commissioners' court, at the groundwater district, and at the PUCT.
What a resident can do
Three things, in order of effort.
First, file a written comment to PUCT Docket 57579. The PUCT accepts filings from non-utility parties and reads them. A comment does not need to be legal-grade — a paragraph stating your position, your household, and your reasoning is enough to enter the record. Written comments shape what the commission has to address in its final order.
Second, contact your state senator and representative. SB 6 was the 2025 vehicle. The 2027 session will likely revisit the unresolved questions — cooling disclosure, groundwater impact, decommissioning. Legislators track constituent contact.
Third, use Stand for local alerts. The local data-center fight — the rezoning, the abatement, the water permit — happens at the county and city level. SB 6 sets the grid context; local votes set everything else.
Flag: what's still uncertain
The PUCT cost-allocation docket is expected to conclude sometime in late 2026 or 2027. Estimates of residential bill impact depend on assumptions about how much queued load actually gets built. Modeling from 2025 should be read as a range, not a forecast. Stand will update this article when a final PUCT rule is issued.
Get alerts for your county
One email when a data-center item lands on your commissioners' court agenda. Free. No spam.
Sources
- Texas Senate Bill 6, 89th Legislature Regular Session (2025). capitol.texas.gov
- ERCOT, 2025 Capacity, Demand, and Reserves Report (May 2025). ercot.com
- Public Utility Commission of Texas, Docket 57579 — Large Load Cost Allocation Rulemaking (ongoing). interchange.puc.texas.gov
- ERCOT Regional Transmission Plan (2024).
- Virginia State Corporation Commission, Case Nos. PUR-2023-00117 and PUR-2024-00072.