The scale of data center-driven electricity demand is no longer theoretical. PG&E has reported its data center pipeline has grown to 12.7 gigawatts (GW), more than double the 5.1 GW it saw in the first quarter, while CenterPoint Energy has submitted over 17 GW of large-load projects to ERCOT’s new Batch Zero process and expects 14 GW to be deemed eligible. Those numbers are staggering on their own, but the more telling detail is the gap between pipeline and reality: PG&E only forecasts 1.8 GW of new data center load actually coming online by 2030. This is the moment when utilities stop counting megawatts and start deciding which projects are real enough to plan around.
Context: The gap between announced demand and actual delivery
The utility disclosures reveal a two-tier system in the making. PG&E’s 12.7 GW pipeline represents potential demand from early-stage inquiries and interconnection requests, but the company has explicitly said it is applying more scrutiny to filter out speculative projects. The 1.8 GW forecast for 2030 suggests that roughly 85% of the pipeline may never materialize as firm load. That is not a failure; it is a filtering process. Utilities are learning that data center developers often submit multiple requests for the same site or test the waters before securing financing and power purchase agreements.
CenterPoint’s numbers tell a similar story, but with a regional twist. The 17 GW submitted into ERCOT’s Batch Zero process – a fast-track interconnection study for large loads – would, if fully realized, push Houston’s peak demand up by more than 65%. The company expects 14 GW to pass initial eligibility, but passing eligibility is not the same as signing a service agreement. ERCOT’s queue is notoriously congested, and the new Batch Zero process is designed to prioritize projects that are genuinely ready to move forward. The fact that CenterPoint is willing to name a 14 GW figure suggests it has already done some preliminary triage, but the final number of energized projects will likely be lower.
The Meta and BlackRock announcement in El Paso adds another layer of complexity. A 1 GW data center is a major load, but the press release did not identify a dedicated power source. In a region where ERCOT’s grid is already tight and transmission constraints are common, that missing piece is not a footnote – it is the central challenge. The project may end up relying on a mix of renewable generation and battery storage, but firm, dispatchable power is still the missing ingredient for 24/7 data center operations.
Cross-cutting analysis: How this reshapes generation and transmission planning
The growth in data center pipelines is colliding with two other trends: the retirement of fossil fuel plants and the slow buildout of new transmission. Utilities are now being forced to make decisions under deep uncertainty. If they build infrastructure for 12.7 GW and only 1.8 GW arrives, ratepayers eat the cost. If they build for 1.8 GW and data centers arrive faster, they face reliability crises and moratoriums. This is why PG&E’s “quality over quantity” approach is spreading across the industry.
The filtering mechanisms themselves are becoming a competitive tool. Utilities are increasingly requiring larger deposits, proof of financing, and milestone commitments before reserving capacity. Some are also charging higher standby rates for speculative load. These measures will likely weed out weaker projects, but they also add friction for legitimate developers. The result is a more conservative planning environment, which could delay some projects that would have been viable under looser rules.
On the generation side, the demand surge is shifting the calculus for new power plants. Intermittent renewables alone cannot back a data center unless paired with massive storage or firm backup. Industry estimates suggest that data centers could account for a significant share of new U.S. electricity demand in the coming decade – some analysts put it near 10% of total consumption by 2030. If even a fraction of the announced pipelines become real, the need for natural gas, nuclear, and long-duration storage will intensify. The tension is that data center developers often prefer clean power for corporate sustainability goals, but the grid cannot deliver it fast enough without new transmission and dispatchable resources.
There is also a geographic dimension. PG&E’s territory in Northern California has limited new generation capacity and a history of rolling blackouts during heat waves. CenterPoint’s Houston region is in ERCOT, which has no capacity market and relies on energy-only pricing. Both regions face different regulatory pressures, but the common thread is that interconnection queues are becoming the real battleground. The projects that secure queue position and transmission capacity first will likely determine which data center hubs emerge over the next decade.
Who this affects
- Utility planners: Need to refine load forecasting models that separate speculative pipeline from credible demand. The 85% gap between PG&E’s pipeline and 2030 forecast is a warning against overbuilding.
- Generation developers: Should focus on projects with signed interconnection agreements and clear fuel or storage plans. The Batch Zero process in ERCOT will reward projects that can demonstrate readiness.
- Data center operators and hyperscalers: Must secure power supply earlier in the site selection process. The Meta/BlackRock El Paso project shows that even a 1 GW nameplate can lack a defined power source.
- State and federal policymakers: Need to revisit transmission cost allocation and interconnection reform. The gap between announced demand and actual delivery is partly a symptom of a queue process that rewards speculation.
What to watch next
- PG&E’s next quarterly earnings call for updates on how much of the 12.7 GW pipeline converts to signed agreements and whether the 2030 forecast changes.
- ERCOT’s Batch Zero study results, specifically which of CenterPoint’s 17 GW projects pass eligibility and which drop out due to transmission constraints or lack of financing.
- Progress on the Meta/BlackRock El Paso project, particularly any power purchase agreements or new generation announcements that fill the supply gap.
- Regulatory filings at the California Public Utilities Commission and the Public Utility Commission of Texas regarding cost recovery for grid upgrades tied to data center loads.
Bottom line
The data center boom is real, but it is not a uniform wave. Utilities are now in a sorting phase, and the numbers that matter are not the headline pipeline totals but the conversion rates. PG&E’s 12.7 GW pipeline with a 1.8 GW forecast and CenterPoint’s 14 GW expected eligibility are early signals of a more disciplined approach. The next few years will separate projects with real power procurement from those that exist only on paper. For anyone involved in energy markets, the key metric to track is not how many gigawatts are announced, but how many actually reach the meter.
Read the full report at Energy Central
Note: facts and figures attributed above to reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply