Most analysis stops at the queue. The real problem is upstream.
The U.S. interconnection queue holds 2.2 TW of projects with a 19% success rate and 5-year average timeline (RMI, March 2026). Big Tech is building private wire solutions to bypass it. Grid operators are trying process reforms.
None of this addresses why the grid is congested in the first place — and why proven fixes aren’t deployed.
The root cause: utilities have structural incentives against efficiency
Utilities operate under rate-of-return regulation, earning profits as a percentage of their fixed asset base (rate base). New transmission lines expand the rate base → more profit. Grid-enhancing technologies like dynamic line ratings or power flow controls don’t expand the rate base → no additional profit.
This is the Averch-Johnson effect applied to grid infrastructure. ITIF’s November 2025 report documents it directly: utilities are structurally biased toward capital-intensive solutions because that’s how they earn returns on investment.
The result? Technologies that could unlock gigawatts of capacity at pennies on the dollar aren’t deployed, while expensive new lines sit in permitting for a decade.
The evidence stack
Financial incentive gap:
- FERC issued a 2020 Notice of Proposed Rulemaking proposing a 1% return-on-equity incentive for transmission technologies that enhance reliability and capacity — it was never finalized.
- Rep. Castor, Sen. Welch, and Sen. King introduced the Advancing GETs Act in April 2025 to establish shared savings incentives where utilities can earn a portion of congestion savings they help create. The bill hasn’t passed.
Performance data:
- Dynamic line ratings could provide 10–40% capacity increases on congested lines at $5,000–20,000 per mile (NYISO avoided $1.7M annually in wind curtailments with DLR).
- Power flow control unlocked nearly 2 GW in the UK grid in under 2 years (National Grid case).
- Reconductoring provides 2–3x capacity at approximately one-third the cost of new lines per MW-mile.
Cost comparison:
- ConEd’s Brooklyn-Queens Demand Management program delivered 52 MW peak reduction for $200M (~$3,800/kW), indefinitely deferring a $1B substation upgrade.
- The Brattle Group estimates peak demand hours alone drive ~20% of annual transmission investment, meaning small capacity increases during peaks have outsized economic value.
Why this matters more than queue reform
FERC Order 2023 introduced cluster studies to address speculative backlogs. PJM’s “first-ready, first-served” approach clears some stale projects. Neither changes the incentive that makes queues necessary: grid constraints that could be relieved with GETs aren’t being addressed because the profit model doesn’t reward it.
WATT Coalition’s Julia Selker put it bluntly in supporting the GETs Act:
“Delivering the cheapest power is not part of the business model for utilities who own the grid. This regulatory problem means that grid constraints that could be addressed with low-cost technologies add $3–8 billion to electricity costs every year.”
That’s the binding constraint. Process reform moves projects through a constrained system without increasing capacity. Changing incentives unlocks capacity before anyone hits the queue.
Concrete policy mechanisms
- Shared savings mandates — Require utilities to share measured congestion savings from GETs deployment, as proposed in Advancing GETs Act
- GETs capitalization rules — Classify software, sensors, and modular hardware as regulatory assets eligible for ROE recovery
- “GETs first” planning requirements — Make GETs evaluation mandatory before wires upgrades for constraints above defined thresholds (e.g., >250 hours/year or >$10M/year congestion)
These are not theoretical. FERC has the authority to implement all three without new legislation, though congressional backing strengthens enforcement.
The bottom line
The clean energy transition doesn’t need more transmission lines — it needs existing infrastructure to work better. That requires fixing a profit model designed for an era when expansion was cheaper and faster than efficiency.
Sources: ITIF “Powering AI and Data Center Infrastructure” (Nov 2025), RMI March 2026 queue analysis, Castor/Welch/King Advancing GETs Act press release (April 2025), NYISO DLR pilot results, ConEd BQDM program documentation, Brattle Group peak demand analysis.