Electricity can be cheap and still be stranded
A generator produces at a node, not everywhere. Its economic output is the energy that can cross the network after losses, operating limits, contingencies, and congestion—not the energy it could produce in isolation.
That distinction becomes decisive when wind and solar equipment gets cheaper and faster to deploy than the network around it. A low bid at a windy site is not a delivered price to a city. The missing terms include interconnection upgrades, substations, long-distance transfer, congestion, curtailment, reserves, and the years during which capital waits for access.
The unit that matters is commissioned transfer
Cost per mile hides voltage and power. Cost per megawatt-mile is better, but still ignores utilization and schedule. The most revealing ledger records dollars, firm transfer capability, congestion relieved, and years to energization together.
A waiting list became a market signal
Interconnection studies determine the network upgrades required to connect a proposed plant and who initially pays for them. As requests multiplied, a serial engineering process became a portfolio-selection problem.
Time is the price this market charges. Among projects that reached operation, the median request-to-operation span ran about 22 months for the 2008 cohort, 36 months for 2015, and 61 months for projects completed in 2025 (Measured, Berkeley Lab)—nearly a tripling across three cohorts. The stages split similarly: about 45 months from request to interconnection agreement, then a median 31 months from agreement to operation in 2025, against 21 months in 2008 (Measured).
Throughput explains why the queue is shrinking while the clock stays long. Active capacity stood at 2,061 GW at the end of 2025 (1,312 GW generation, 749 GW storage), down 10% year over year and 21% from the 2023 peak of 2,598 GW, as more than 750 GW withdrew during 2025 alone (Measured). The active mix is solar 773 GW, storage 749 GW, gas 253 GW—up 86%—and wind 220 GW (Measured). With 549 GW holding an interconnection agreement but not yet operating, a falling queue mostly records attrition, not faster delivery.
The queue is therefore neither a forecast nor pure noise. It records the option value developers place on network access. Withdrawal can mean a speculative proposal failed; it can also mean upgrade costs, study delays, land, procurement, financing, or power contracts made an otherwise buildable plant uneconomic.
The bottleneck has two clocks
| Clock | Earlier benchmark | Recent benchmark | What it measures |
|---|---|---|---|
| Generator interconnection | 22 months · projects built 2008 | 61 months · projects built 2025 | Request to commercial operation among projects that finished. |
| Federal environmental review | Fragmented agency processes | ~4-year historical average; two-year CITAP target | Notice of intent to record of decision, not the whole project. |
| New transmission | Project-specific | ~10 years average in DOE-cited review | Development through completion across 30-plus projects. |
| Existing-corridor upgrade | Static operation | Often 1–3 years where feasible | GETs or reconductoring without a new right-of-way. |
The wire is only one layer of a transmission line
Steel, aluminium, copper, insulation, concrete, and power electronics are tangible. The route, approvals, system studies, beneficiary agreement, construction windows, and operating rules determine whether those materials become useful infrastructure.
Choose a need
Forecast load, retirements, generation, reliability events, and trade across a network whose future is uncertain.
- Layer
- Planning
- Output
- Commissioned, operable transfer capability
Decide who benefits
A line can lower energy cost, improve reliability, and unlock generation in different places. Agreement over who pays often precedes engineering.
- Layer
- Cost allocation
- Output
- Commissioned, operable transfer capability
Secure a path
Rights-of-way, state and federal approvals, environmental review, tribal consultation, landowner agreements, and community design define the buildable route.
- Layer
- Siting
- Output
- Commissioned, operable transfer capability
Build the terminals
Substations, breakers, transformers, converter stations, protection, and controls determine whether conductor capacity can enter the operating grid.
- Layer
- Equipment
- Output
- Commissioned, operable transfer capability
String and commission
Towers, foundations, conductors, access roads, specialist labour, outages, testing, and market integration finally create usable transfer capability.
- Layer
- Construction
- Output
- Commissioned, operable transfer capability
The physical floor is lower than the delivered project
Ohmic loss is unavoidable: current heats a conductor in proportion to current squared and resistance. Higher voltage moves the same power with less current, which is why long-distance systems climb from distribution voltages to hundreds of kilovolts and why HVDC becomes attractive for particular long, controlled transfers.
But today's dominant gap is rarely between conductor performance and a thermodynamic limit. It is between a viable electrical design and a commissioned asset. The materials set a floor; coordination sets much of the premium and the variance.
An existing corridor is a platform
Once a route has towers, access, easements, known neighbours, and a place in operating models, its reuse can avoid the slowest parts of a greenfield project.
Dynamic line ratings measure weather and conductor conditions instead of assuming a conservative fixed limit. Power-flow controls redirect electricity around overloaded elements. Topology optimization finds better network configurations. Advanced conductors can tolerate higher temperatures with less sag and, on suitable structures, roughly double capacity.
These are not four versions of free capacity. A windy-day rating is variable. A rerouted flow consumes headroom elsewhere. A stronger conductor may expose a substation limit. Reconductoring requires outages and structures that can accept the design. Their value is speed: they buy time and transfer while the system decides where genuinely new corridors are unavoidable.
Upgrade before expansion—but do not confuse them
Measure and route
Reveal thermal headroom and use the network's spare paths more deliberately.
Fastest · variable or topology-dependent gainReconduct and uprate
Reuse a permitted route while replacing wires and, where needed, terminal equipment.
Medium term · potentially near-2× line capacityAdd new corridors
Create firm geographic transfer that optimization and existing rights-of-way cannot supply.
Slowest · largest structural network changeA regional plan is part of the machine
Transmission creates multiple products over decades: energy savings, reliability, resource adequacy, resilience, reduced losses, access to generation, and insurance against futures that never arrive exactly as forecast.
Planning only for the first generator that requests service tends to produce piecemeal upgrades and repeated studies. Planning for a portfolio can identify shared facilities, but immediately raises the political question: which future benefits count, and which customers should fund them?
FERC Order No. 1920 requires long-term regional transmission planning and consideration of a broader set of benefits. Order No. 2023 moved generator interconnection toward first-ready, first-served cluster studies. Rules can shorten duplicated work, but they cannot manufacture transformers, settle every state dispute, or create local consent. Reform succeeds when measured in energized capacity and lower congestion—not filings completed.
Planning is sequential
The business case, regional plan, beneficiary analysis, route, permits, equipment orders, and construction cannot all begin at once. Delay at one layer idles the rest.
The cheapest line can be unbuildable
A direct path may cross too many jurisdictions or sensitive landscapes. Route changes raise length and cost before material is purchased.
A line is a network intervention
Its value depends on generation, load, neighbouring constraints, dispatch, and contingencies. Dollars per mile alone cannot describe the useful output.
Transformers can become the clock
A conductor upgrade that outruns terminal equipment simply moves congestion into substations, breakers, and protection systems.
Queues contain options, not forecasts
Developers submit projects under uncertainty and most historical queued capacity has not reached operation. Queue volume measures demand for access, not future supply.
Local costs and broad benefits diverge
Communities host towers and land impacts while reliability and market savings spread across regions. Durable consent needs a distributional answer, not only a net-benefit model.
An optimistic view, with conditions
The grid has a faster lane and a durable lane
Measure, control, and reconductor existing lines now; plan new regional corridors before each individual project proves the need again. The technologies are complementary because one buys years and the other buys geography.
Expose headroom
Deploy ratings, controls, topology tools, storage, and queue reform where they can release capacity without new land.
Reuse corridors
Reconduct suitable routes and upgrade the substations and protection systems that would otherwise become the next bottleneck.
Plan across futures
Build high-capacity regional and interregional paths whose reliability and market value survive any single generator proposal.
Sources, method, and boundaries
This report treats queue entries as applications, nameplate uplift as conditional engineering potential, and targets as targets. It does not collapse overhead lines, underground cables, HVAC, HVDC, interconnection upgrades, or bulk regional facilities into one misleading cost curve.
