The Grid Is the Bottleneck: A Practical U.S. Agenda for Transmission, Interconnection, and Local Resilience

The bottleneck is coordination, not a single missing technology

Public debate often reduces the grid problem to a construction target: build more power plants, build more transmission, add batteries or manage new demand. Each can be necessary. None is sufficient when the institutions responsible for generation, load, transmission, distribution, land use and consumer protection work from different assumptions.

A generator can be financially viable but unable to connect. A transmission line can have regional benefits but no accepted method for allocating its costs. A data center can receive local economic-development support while its electricity requirements are absent from an older regional forecast. A community can be asked to host infrastructure after the commercially preferred route has already become difficult to change. A distribution system can experience outages even when the bulk system has enough generation.

These are governance failures expressed through engineering constraints. The remedy is not a command to approve every project. It is a planning system that uses common data, tests credible alternatives, assigns costs according to benefits, rejects speculative proposals earlier and preserves enough flexibility for forecasts to be wrong.

Demand has moved faster than the planning baseline

The North American Electric Reliability Corporation’s 2025 Long-Term Reliability Assessment projects 224 gigawatts of summer peak-demand growth and 246 gigawatts of winter growth over the following decade. NERC says the summer projection increased by more than 69 percent relative to its 2024 assessment, with data centers for artificial intelligence and the digital economy accounting for most of the added forecast. NERC also warns that uncertainty and delays in new-resource additions increase the risk that supply will not keep pace.

Forecasts are not commitments. Some announced data centers, factories and hydrogen projects will be delayed, reduced or never built. Treating every request as certain would cause consumers to finance unnecessary infrastructure. Treating rapid load growth as temporary noise would create the opposite risk: inadequate capacity, emergency procurement and deteriorating reliability.

The planning response should therefore use scenarios rather than a single point forecast. Scenarios should distinguish committed loads from speculative inquiries, test different rates of electrification and industrial growth, and identify projects that provide value across several plausible futures. This is consistent with the logic of FERC Order 1920, which requires long-term regional planning using at least three scenarios over a horizon of no less than 20 years.

A crowded queue is evidence of demand for access—not proof of available capacity

Lawrence Berkeley National Laboratory’s 2026 edition of Queued Up reports that approximately 8,200 projects were actively seeking U.S. grid interconnection at the end of 2025. They represented about 1,312 gigawatts of generation and 749 gigawatts of storage. Together, the active queue exceeded 2,060 gigawatts.

That number should not be described as electricity waiting to be switched on. Interconnection queues contain projects at different levels of maturity. Many are withdrawn when studies reveal network-upgrade costs, financing changes, permits fail or commercial assumptions no longer hold. Berkeley Lab explicitly cautions that most projects entering queues are not built and that completed projects have taken longer to move through studies and reach operation.

Even so, the queue is an important planning signal. It shows that developers seek access in locations and at volumes the inherited study process was not designed to handle. Studying projects sequentially can force the same network condition to be reexamined repeatedly. It can also assign a large upgrade to one project even when several later projects would use the same equipment.

FERC Order 2023 responds by moving public utility transmission providers toward cluster studies and a “first-ready, first-served” model, adding project-readiness requirements, study deadlines and reforms intended to account for technological advances. The reform is directionally sound: speculative requests should not indefinitely occupy scarce study capacity. But queue discipline cannot compensate for an undersized regional plan. If every cluster identifies the same transmission constraint without a process to consider a shared regional solution, the system becomes more orderly without becoming adequate.

Long-term planning must count the full value of a line

Transmission assets last for decades and serve multiple purposes. A line may reduce congestion in normal hours, transfer power during extreme weather, connect new resources, avoid repeated local upgrades and preserve options for future demand. Cost allocation becomes contentious because these benefits occur in different places and at different times.

Order 1920 and its rehearing orders require long-term regional transmission planning, transparent selection criteria, evaluation of multiple benefits and a formal role for states in discussing cost allocation. The rule does not require planners to select every project that scores well, and it does not eliminate disagreement over who benefits. It creates a more disciplined process for identifying needs that short-term reliability studies can miss.

The Department of Energy’s National Transmission Needs Study reinforces the underlying case. Its current findings identify a pressing need for additional transmission because of load growth, generation and load interconnection, reliability and congestion. DOE also identifies interregional links as a source of reliability, resilience and consumer value, particularly during stressed conditions.

A credible benefit framework should report results by category rather than collapsing every effect into one opaque score. At minimum, planners should separately estimate production-cost savings, avoided or deferred upgrades, resource-adequacy value, loss reduction, extreme-event transfer capability and the option value of serving uncertain future needs. Publishing both assumptions and sensitivity ranges would let states and consumers see which benefits drive a recommendation.

Interregional capability is reliability insurance

Weather, fuel supply and equipment failures do not respect utility boundaries. A region experiencing scarcity benefits when it can import from a region whose demand and resource conditions differ. Interregional transmission does not eliminate the need for local generation or reserves, but it expands the set of resources available during stress.

The economic case should be evaluated as insurance as well as average-hour congestion relief. A connection that is lightly used in typical conditions may be valuable during a small number of severe hours. DOE’s needs analysis notes that a large share of congestion is concentrated in a limited number of hours and identifies substantial potential value in additional cross-region links.

This does not justify unlimited construction. Interregional projects are expensive and can create difficult cost disputes. It justifies a recurring, transparent assessment of transfer capability under common extreme-weather and outage scenarios. Regions should identify where modest upgrades, advanced conductors, topology optimization or new lines produce the highest resilience value before assuming the answer is always the largest project.

A six-part implementation agenda

1. Publish comparable queue and large-load dashboards

Transmission providers should publish machine-readable information on active projects, cluster stage, study delays, withdrawal rates, network upgrades and available transfer capability. Large-load requests should be reported in maturity bands that protect commercially sensitive information while distinguishing executed service agreements from preliminary inquiries. Comparable definitions are essential; a national total built from inconsistent categories can mislead as easily as it informs.

2. Connect interconnection studies to regional plans

Repeated cluster findings should trigger review in the long-term planning process. When several ready projects require overlapping upgrades, planners should compare project-specific facilities with a shared regional alternative. The cost of the shared option should be allocated only after its broader beneficiaries and benefits are documented.

3. Standardize benefit categories, not predetermined outcomes

FERC should continue developing consistent reporting for major benefit categories while allowing regional models to reflect real system differences. Standardization should make assumptions comparable, not force identical project choices. Every major proposal should show base results, sensitivity cases and which parties receive each estimated benefit.

4. Establish an interregional reliability portfolio

Neighboring regions should run common stress tests and rank a portfolio of solutions: operational coordination, grid-enhancing technologies, reconductoring, targeted upgrades and new transfer paths. A portfolio approach reduces the false choice between doing nothing and approving one extremely large line.

5. Make permitting predictable and participatory

Lead agencies should establish one public schedule, one shared evidentiary record and early milestones for route alternatives. States, Tribes, landowners and affected communities should be engaged before a preferred route becomes commercially entrenched. Community participation should influence siting, mitigation and benefit design; it should not be reduced to a final hearing after the meaningful choices have closed.

6. Pair regional expansion with local resilience

Distribution automation, microgrids for critical facilities, demand flexibility, storage and building-level backup can reduce outage consequences and defer some upgrades. They cannot replace the energy-transfer and diversity benefits of the bulk system. Planning should measure where local resources reduce peak requirements or provide critical-load continuity and compensate them for verified performance.

Consumer protection must be built into cost allocation

A project can produce system benefits and still distribute costs unfairly. Residential and small-business customers cannot negotiate individual transmission contracts, and they may pay for forecast errors through regulated rates. Consumer protection therefore requires more than a finding that a line is broadly useful.

Planning records should identify which customer groups are expected to benefit, when benefits begin, which forecasts control the estimate and what happens if the large load or generation project driving the need does not appear. Where a discrete customer creates a substantial near-term upgrade, that customer should bear costs that are not supported by documented broader benefits. Where a portfolio supplies regional reliability or congestion value, beneficiaries can reasonably share costs according to a transparent method.

Large-load agreements also need protections against stranded investment. Milestone payments, collateral, minimum-demand commitments and withdrawal charges can reduce the chance that other customers finance facilities built for a project that disappears. These mechanisms should be calibrated to actual risk: terms so severe that every new industrial load becomes uneconomic can be as damaging as socializing all development risk.

Regulators should publish bill-impact ranges, not only total project costs. A low, central and high case can show how demand, financing and fuel assumptions affect customers over time. Review should consider whether construction reduces other expenses, including congestion, emergency procurement and repeated local upgrades. Affordability is the net effect on service, not the sticker price of one asset.

A staged implementation sequence

The agenda can be implemented without waiting for a new national grid authority. During the first year, FERC and DOE should define common reporting fields while regions publish queue and large-load data using maturity bands. States should identify representatives for Order 1920 planning and cost-allocation forums, and neighboring regions should agree on a small set of common stress scenarios.

During years two and three, regional plans should flag constraints that recur across interconnection clusters, compare shared solutions and publish benefit sensitivities. Interregional studies should rank operational changes, advanced technologies, reconductoring and new lines in a single portfolio. Lead permitting agencies should publish integrated review schedules and document how early consultation changed a route, mitigation measure or community-benefit arrangement.

By year five, regulators should be able to compare whether reforms reduced study time, produced more accurate upgrade estimates, increased completion by commercially ready projects, improved transfer capability and protected customers from abandoned-load costs. Metrics should not reward speed alone. A fast study that repeatedly underestimates network work or shifts costs without evidence is not successful reform.

The scorecard should also record uncertainty. Forecast accuracy, project withdrawal and cost variance are not embarrassing side notes; they are evidence needed to improve the next planning cycle. A mature system learns from proposals that fail as well as facilities that enter service.

Distribution systems need the same discipline

Most customers experience the grid through local distribution wires, not interstate transmission. Electrified transport, rooftop generation, batteries, heat pumps and severe weather can create neighborhood constraints that regional models do not see. Utilities should publish hosting-capacity information, planned feeder upgrades and reliability performance in forms that customers and local governments can use.

Distribution planning should test whether targeted efficiency, flexible load, storage or circuit modernization can defer a conventional upgrade without compromising service. When a non-wires alternative is selected, compensation should depend on availability and measured performance during the hours that created the need. When a physical upgrade is more reliable or economical, planners should say so directly rather than treating every distributed resource as an automatic substitute.

Critical facilities require special attention. Hospitals, water systems, emergency communications and shelters may need islandable power, prioritized restoration or onsite reserves even when broad reliability metrics look acceptable. State regulators can require utilities to identify these loads, evaluate cost-effective resilience options and coordinate with local emergency planning. This is a different task from keeping every building continuously powered, and the distinction helps target limited funds.

What this agenda would not solve

Transmission reform cannot guarantee that every announced load or generator will be economic. It cannot resolve all conflicts over land, rates or local development. It cannot substitute for resource-adequacy requirements, distribution investment, fuel security or competent utility regulation. Faster processes can still produce bad projects if their assumptions are weak; more participation can still fail if agencies cannot make decisions.

The principal uncertainty is forecasting. Demand, technology costs and project pipelines will change. That is why the recommended framework emphasizes scenarios, staged commitments and options that perform across more than one future. The standard should be adaptive adequacy, not a claim that planners can know the grid of 2045 in detail.

Conclusion

The grid bottleneck is real, but its most important feature is institutional fragmentation. The United States now has major federal reforms addressing interconnection and long-term transmission planning. The next task is implementation that joins those processes, makes their evidence comparable and gives states and communities a meaningful role before conflict becomes the default.

A capable grid is not defined by the number of projects approved. It is defined by whether reliable resources can connect, power can move when it is needed, costs follow demonstrable benefits and the system remains useful when forecasts are wrong.

That standard is deliberately technology-neutral and durable across administrations. It can support conventional generation, renewable resources, storage, advanced transmission, flexible demand and technologies that have not yet reached commercial scale, provided each demonstrates value under the same transparent planning assumptions.

Sources and methodology

This paper synthesizes current federal rules, official reliability assessments and national laboratory queue data. Recommendations are WSOI Institute’s analysis and are not statements by the cited agencies.

• North American Electric Reliability Corporation, 2025 Long-Term Reliability Assessment.
• Lawrence Berkeley National Laboratory, Queued Up: 2026 Edition.
• Federal Energy Regulatory Commission, Order No. 2023 explainer.
• Federal Energy Regulatory Commission, Orders 1920, 1920-A and 1920-B explainer.
• Federal Energy Regulatory Commission, Order 1920 compliance schedule.
• U.S. Department of Energy, National Transmission Needs Study.