The ERCOT interconnection queue is the first place energy originators look for large load prospects. It's also increasingly the wrong place to look. A growing share of the largest power consumers in Texas — and across the US — are deliberately engineering their development pipelines to never appear in it.
This is not a loophole or an accident. It's a deliberate strategy. As grid interconnection lead times have ballooned past six years in some ISOs, data center operators have made a structural shift: instead of waiting for utility power, they're building it themselves. And when you build your own power behind the meter, you don't need to queue for grid interconnection at all.
For energy developers — solar originators, wind developers, power traders, and storage project teams — this creates a critical intelligence gap. The companies with the largest, most bankable power demand in the US are systematically absent from the data sources most teams rely on.
The Visibility Problem
Traditional interconnection queues track generation — the solar farms, gas plants, and wind projects seeking to connect supply to the grid. Load interconnection, where it exists at all, is inconsistently reported and rarely granular enough for origination intelligence. But the bigger issue now is that the largest loads are bypassing the grid connection process entirely.
Data center operators pursuing behind-the-meter power don't apply for utility service through normal channels. They don't show up in ERCOT's NOIE filings as a standard grid customer. And crucially, they often structure their land acquisitions through holding companies and subsidiaries specifically to avoid tipping off the market to their development plans.
Unlike industrial facilities, data center operators often use complex corporate structures and multiple land acquisitions to maintain secrecy about their development activities. Many projects designed for behind-the-meter generation never even appear in traditional interconnection queues.
The result is a fundamental mismatch. An originator prospecting for long-term PPA counterparties will screen the interconnection queue, run through PUCT dockets, and come up with a list that looks comprehensive — but is systematically missing the most capital-intensive, longest-duration power demand in the market.
Why BTM, Why Now
The economics driving the BTM shift are straightforward. Grid interconnection in major markets like PJM requires a non-refundable readiness deposit of $4,000 per MW or more just to hold a queue position. For a 250 MW campus, that's a $1M entry fee with no guarantee of approval. For a gigawatt-scale hyperscale deployment, costs exceed $4M — and remain sunken even if the application is rejected.
That's before the timeline problem. Substation interconnection lead times have surpassed six years in some ISOs. Data center construction timelines are 18 to 36 months. A developer who breaks ground today could have a completed campus sitting dark for years waiting for utility power to arrive.
"On-site power and BTM solutions now account for over 25% of all new data center capacity."
— LandGate 2026 Behind-the-Meter ReportInterconnection requests for gas generators jumped 160% year over year as developers seek to avoid traditional grid delays. The US gas fleet supporting data centers is expected to double in size. More than one third of all new gas-fired capacity under development in the US is slated to directly power data centers as BTM projects.
BTM Capacity by Power Type
Natural gas dominates because of its dispatchability and rapid deployment — mobile turbines can go live in 18 months as a bridge to permanent infrastructure. Solar is the fastest-growing segment, with deployment timelines as short as 6–12 months once site prep is complete. Nuclear operates at a 92.5% capacity factor — the only carbon-free source that truly matches the continuous load profile of a data center.
Where BTM Deals Are Concentrating
The BTM trend is concentrated in states with power-friendly regulatory environments, abundant land, and access to natural gas infrastructure or high solar irradiance. For originators, prospect density is heavily geographic.
| State | BTM Capacity | Primary Source | Priority |
|---|---|---|---|
| Texas (ERCOT) | 33 GW+ | Gas + Solar | A — Act Now |
| Pennsylvania (PJM) | 11 GW | Gas + Nuclear | A — Act Now |
| West Virginia (PJM) | 8.6 GW | Gas | B — Pipeline |
| Utah | 5 GW+ | Mixed | B — Pipeline |
| Louisiana | 4.4 GW | Gas | B — Pipeline |
| Wyoming | 3.9 GW | Wind + Gas | C — Watch |
| Tennessee | 2.4 GW+ | Mixed | C — Watch |
Texas is the standout. ERCOT faces over 40 GW of demand attributable to data center development. Hyperscalers including xAI, Meta, and OpenAI have all committed to on-site gas generation within the state. Data centers are projected to reach over 30% of total US power demand by 2028, with Texas carrying a disproportionate share.
What This Means for Origination
The BTM shift does not mean these companies don't need renewable energy agreements. It means the structure of the deal changes. Instead of a traditional utility-scale PPA for grid-connected power, BTM developers need co-located generation, direct interconnects, and hybrid power agreements — often structured as energy-as-a-service where a developer owns and operates the generation plant under a long-term offtake contract.
The PPA opportunity is still very much alive. It just requires getting to these companies earlier, before they've locked in a generation partner, and through channels that aren't the standard interconnection queue scan.
BTM data centers represent the highest-value, longest-duration power demand in the US market. The challenge is intelligence, not demand. They exist, they're building, and they need generation partners — they're just invisible to originators relying solely on queue data.
The signals that surface BTM offtakers
Because BTM developers bypass traditional interconnection, the signals that reveal their activity are scattered across other regulatory and public record systems. PUCT dockets capture large load service requests before any grid application is filed. TCEQ air permits are required for on-site gas generation and are often the first public record of a data center's power strategy. TDLR construction records show large-scale building activity months before queue filings. SEC EDGAR 8-Ks from public companies disclose capital commitments that never appear in grid data.
None of these sources is monitored systematically by most energy origination teams. That's the gap EnergyLeads was built to close.
We track every major BTM data center in ERCOT — scored, sourced, and contact-ready.
50+ active large-load offtakers. 28 GW of aggregate demand. Weekly updates. Every record traced to PUCT, TCEQ, EDGAR, or authoritative industry databases — with the procurement contacts you need to start the conversation.
The Bottom Line
Behind-the-meter is not a workaround. It is a structural shift in how the largest power consumers in the US are being built and powered. For energy originators, this creates both a challenge and an advantage: the teams that learn to surface BTM offtaker intelligence early will have access to deal flow that competitors simply cannot see.
The ERCOT market is the clearest proof point. More than 33 GW of BTM capacity is already in development in Texas. Data centers are on track to account for nearly a third of all ERCOT demand within two years. The origination opportunity is enormous — and almost entirely invisible to teams relying on interconnection queue data alone.
The signal is out there. It's in the permit records, the corporate filings, and the construction databases. You just have to know where to look.