Skip to main content

Insights

Behind-the-meter vs. the ERCOT queue: which is faster to power?

2026-07-08W Land Development

For AI and high-performance computing operators racing to energize, behind-the-meter (BTM) power is generally faster to bring online than waiting in the ERCOT interconnection queue. BTM generation sits on-site, past the utility meter, and never enters a public-grid queue that now exceeds 100 GW and can take several years to clear. Grid interconnection still matters for scale, but for raw speed-to-power, on-site supply leads.

What does "behind-the-meter" power actually mean?

Behind-the-meter refers to generation and storage on the customer's side of the utility meter, serving load directly rather than routing electricity through the public transmission and distribution system. Because the power is produced and consumed on-site, a BTM project does not have to secure and energize a grid interconnection before it can serve a data hall — the load and its supply are co-located.

  • On-site generation (commonly natural-gas turbines or reciprocating engines) sited directly next to the compute load.
  • Frequently paired with BESS (battery energy storage) and UPS systems for power quality, transient smoothing, and ride-through.
  • No dependence on clearing a public interconnection queue in order to begin serving load.

This is the model underpinning our approach to private grid power for large, latency-sensitive computing campuses.

Why is the ERCOT queue so slow right now?

ERCOT's large-load interconnection queue has swelled to well over 100 GW, with roughly 70% tied to data centers (ERCOT). Studying that load, approving it, and physically building the transmission upgrades to serve it all take time — typical interconnection timelines run several years, not months. The bottleneck is national, not only Texan: Lawrence Berkeley National Laboratory's Queued Up research shows interconnection queues across the U.S. growing far faster than projects can actually be built and energized.

  • Queue volume well over 100 GW, roughly 70% data-center driven (ERCOT).
  • Typical interconnection timelines measured in several years.
  • Congestion is systemic across U.S. markets (LBNL, Queued Up).

How much faster is behind-the-meter to energize?

Speed is the core reason BTM has surged. Because it sidesteps the queue, on-site generation can be energized on a construction-driven schedule rather than a transmission-study one. The market has voted with capital: as of 2026, nearly 90 GW of behind-the-meter gas generation has been announced nationwide to serve data centers (Cleanview). The clearest signal is Microsoft and Chevron's Project Kilby, structured as an explicitly behind-the-meter project that bypasses ERCOT interconnection entirely — a template many hyperscalers and GPU-cloud operators are now studying for their own first phases.

Is behind-the-meter a replacement or a bridge to the grid?

In practice, most large operators treat BTM as a bridge rather than a permanent substitute. On-site generation can carry early phases and de-risk time-to-power while a parallel grid interconnection matures in the queue, and many designs preserve the option to synchronize with the public grid later. Pairing generation with storage strengthens the case further: BESS and UPS assets smooth power quality and provide ride-through during transients, dispatch changes, or maintenance windows — the reliability profile AI training and inference workloads demand.

What does this mean for West Texas site selection?

West Texas is compelling precisely because it combines abundant land, nearby gas resources, and grid proximity — conditions suited to both BTM-first energization and eventual interconnection. W Land's West Texas AI energy campus is planned around this dual approach, with on-site generation targeted to shorten time-to-power while grid options remain open. All capacity figures, phasing, and dates are preliminary and subject to engineering, permitting, and financing.

For operators weighing options, the practical question is sequencing. To see how these timelines compare on the ground, read our companion analysis on time to power for an AI data center in West Texas, and evaluate parcels through the lens of data center site development.

The takeaway: with the ERCOT queue exceeding 100 GW and interconnection typically taking several years (ERCOT), behind-the-meter generation has become the fastest route to first megawatts — which is why nearly 90 GW of BTM gas capacity has been announced for data centers nationwide (Cleanview). To review how a BTM-first campus could be sequenced for your specific load, request an NDA briefing with our development team.

All insights

Request a 30-minute NDA briefing

Qualified inquiries for AI energy campus, private-grid power, and data center site development.

Request NDA Briefing