Executive · Grid Interconnection
Grid interconnection realities and utility queue bottlenecks
Power availability, not land or labor, now decides where large data centers get built and when they open. Treat interconnection as the first critical path item, evaluate bridging power honestly, and keep nuclear and SMR in the long-term plan rather than the near-term schedule.
Figures and policy details change quickly. Verify queue positions, tariffs, and regulatory rules with the relevant utility and regional transmission organization before committing capital.
The core constraint
Utilities were not built for loads that arrive in 300 MW to 1 GW increments on a two-year build cycle. Large load requests have surged in markets such as Northern Virginia, Texas, Ohio, Arizona, and Georgia, and many utilities now report queued demand far exceeding what they can serve in the near term. Lawrence Berkeley National Laboratory's work on generator interconnection has shown multi-year waits for generation projects, and load interconnection is following a similar pattern as transmission, substation, and generation additions all have to land in sequence.
In many markets, a request submitted today yields energization dates in the late 2020s or early 2030s. Some utilities have paused new large load commitments entirely or introduced stricter requirements.
Why delays happen
- Transmission and substation capacity. New 230 kV and 345 kV lines, and the substations to support them, take years to permit and build.
- Generation adequacy. Even where wires exist, utilities must demonstrate they can serve the load reliably. Capacity markets, such as PJM's, have reflected this tightness in higher prices.
- Equipment. The same transformer and switchgear lead times described in Long-Lead Equipment Procurement Strategies for 2026 and Beyond constrain utilities too.
- Study backlogs. System impact and facility studies run sequentially, and restudies after other applicants withdraw can reset your timeline.
- Speculative requests. Developers file for multiple sites to hold options, inflating queues. Utilities have responded with deposits, minimum take agreements, and milestone requirements.
Site selection: power first
Reverse the traditional sequence. Instead of picking land and then asking for power, start with the utility.
A power-first screening framework:
1. Identify existing capacity. Look for substations with available headroom, retiring industrial or generation sites with intact interconnection, and areas where load has declined.
2. Talk to the utility early. Request a realistic energization timeline in writing and ask about phasing options (for example, 100 MW in year one, scaling to 400 MW).
3. Check the tariff. Large load tariffs increasingly require long-term contracts, minimum demand charges, and collateral. Understand your financial commitment before you commit.
4. Assess transmission proximity. Sites within a few miles of high-voltage lines and a viable substation location carry less risk.
5. Evaluate the regulatory environment. Some states are moving to prioritize or restrict large loads, and rules on cost allocation are evolving.
6. Consider ancillary constraints. Water availability for cooling, fiber routes, and permitting speed still matter, but rank below power.
Secondary and tertiary markets with available capacity can outperform primary markets with long queues, provided you solve labor. See Megawatt-Scale Labor Mobilization and Capacity Planning.
Reading the queue honestly
A queue position is not a delivery date. Interrogate it.
- What specific upgrades does your study require, and who is funding them?
- Which of those upgrades have permits, and which are still in planning?
- What equipment is on order, and what are the delivery dates?
- How many projects ahead of you hold deposits versus placeholders?
Build the program schedule from the utility's most conservative credible date and treat earlier dates as upside. Tie your own capital release gates to utility milestones, so you are not spending on a campus that cannot be energized.
Bridging and off-grid power options
When grid power is years away, on-site generation can bridge or, in some cases, serve as the primary supply. Each option has tradeoffs.
| Option | Typical role | Key constraints |
|---|---|---|
| Natural gas reciprocating engines | Bridge or primary; modular, scalable | Air permits, gas supply and pipeline capacity, emissions, noise |
| Aeroderivative and industrial gas turbines | Larger scale bridge or primary | Turbine slot availability (many manufacturers report multi-year backlogs), permitting |
| Fuel cells | Fast deployment, lower local emissions | Cost per kW, gas supply, vendor capacity |
| Solar plus storage | Supplement, not a firm standalone source at this scale | Land area, intermittency, cost of firming |
| Battery storage | Peak shaving, ride-through, grid support | Cost, duration limits |
| Grid-connected with on-site generation | Hybrid; bridge now, backup later | Complex interconnection and market rules |
Several developers have deployed temporary gas turbines or engines to start operations while waiting for grid service. This approach has drawn regulatory and community scrutiny over emissions and permitting, so budget time and cost for that process.
Key questions for any behind-the-meter plan:
- Fuel. Is firm gas supply available at the site, and at what pressure? Pipeline expansion takes years.
- Reliability. Off-grid generation needs its own redundancy. Achieving the equivalent of 2N on generation is expensive, and many designs retain a grid connection as the second source.
- Permits. Air quality permits for large combustion sources can take 12 to 24 months.
- Economics. Self-generation often costs more per kWh than grid supply. It is a way to buy time and revenue, and the model should say so.
- Exit. Decide at the outset whether this generation becomes permanent backup, gets relocated, or is sold once grid power arrives.
Colocation at existing power plants, including nuclear, has become a notable strategy. Regulatory treatment of these arrangements has been contested at FERC, so confirm the current rules and any pending decisions before structuring a deal.
Nuclear and SMR: set realistic expectations
Interest in nuclear from large technology buyers is real and growing. Power purchase agreements with existing nuclear plants, restarts of recently shuttered units, and development agreements with small modular reactor companies have all been announced.
The timelines deserve careful reading.
- Existing plants and restarts. Power from relicensed or restarted units can arrive in the near-to-medium term, with announced restart targets around 2027 to 2028 for some projects. Capacity is limited, and these deals are few.
- Uprates and long-term PPAs. Incremental additions at existing plants can add firm carbon-free supply, subject to regulatory approval.
- SMRs. Designs are advancing through NRC review and early demonstration, but commercial-scale deployment in meaningful volume is generally expected in the 2030s. First-of-a-kind costs, supply chain for fuel (including HALEU), and licensing remain open risks.
For a data center program starting now, SMRs belong in a long-term power strategy, perhaps as a second or third phase supply, and not in the energization plan for the first phase. Programs that plan around SMR delivery for opening-day power are taking a bet they cannot hedge.
A practical sequence for executives
1. Secure the utility path first. Obtain a written energization schedule, phasing, and tariff terms before spending heavily on site work.
2. Decide on bridge power early. Evaluate permitting and fuel before design freeze, since it affects site layout and electrical architecture.
3. Phase the campus to match power delivery. Build in blocks that can energize as capacity arrives, with shared infrastructure sized for the final state.
4. Stack long-term sources. Combine grid supply, contracted clean power, on-site generation, and storage, and add nuclear options on a realistic horizon.
5. Govern power as a program risk. Track utility milestones, equipment deliveries for the substation, and regulatory developments at the same level of attention as construction progress.
Takeaways
- Treat power as the first gate, and screen sites by utility capacity before land.
- A queue position is not a date. Build your schedule from the conservative case.
- Bridging power can work, but permitting, fuel, and economics need a full analysis.
- Nuclear belongs in the long-term strategy, with SMR timelines pointing to the 2030s.
- Phase the campus so each block energizes on a confirmed power milestone.
Related briefs: Hyperscale Program Governance and Capital Risk Mitigation | Long-Lead Equipment Procurement Strategies for 2026 and Beyond
---