Interconnection queue prediction for data center site selection

Every ISO and RTO maintains an interconnection queue — a ordered backlog of proposed generators, storage, and hybrid projects seeking grid connection. Projects move through feasibility, system impact, and facility studies. Capacity shown on utility hosting maps often assumes every queued project will succeed.

Lawrence Berkeley National Laboratory's Queued Up research finds roughly 75% of proposed U.S. interconnection capacity withdraws before reaching commercial operation. In GridSolve's dataset, the capacity-weighted national withdrawal rate is 61% across 5,660 active projects at 4,687 substations.

Without interconnection queue prediction, developers treat speculative megawatts as firm capacity. A substation that appears full may have headroom once likely withdrawals are removed; one that appears open may be dominated by cluster studies that stall for years.

For data center interconnection planning, queue reality matters even though large loads typically interconnect through separate utility processes. Deliverable generation-queue capacity at a substation is a strong signal of where grid headroom is likely to exist — a practical input to site selection before committing to interconnection studies.

GridSolve fits a Cox proportional-hazards survival model on historical generation-interconnection projects. Each active project receives a withdrawal probability based on attributes that predict dropout: cluster study involvement, service type, technology, queue position, and regional base rates.

Those probabilities roll up to every substation as likely withdraw MW. Subtracting expected withdrawals from queued megawatts yields adjusted available MW — withdrawal-adjusted interconnection capacity ranked by the GridSolve Site Score.

See the full interconnection queue model methodology for regional validation, data sources, and scope limits.

GridSolve covers National — 9 regions. Model validation varies by region. — 4,687 substations across 9 ISO/RTO regions and 48 states. Withdrawal rates are computed empirically within each region from resolved projects; model predictive quality varies by region and is surfaced as a confidence tier on each substation.

What is the interconnection queue?

The interconnection queue is the backlog of generation and storage projects waiting to connect to the electric grid. Each ISO and RTO maintains its own queue. Projects must complete feasibility, system impact, and facility studies before they can operate. Most proposed capacity in U.S. interconnection queues never reaches commercial operation.

What is interconnection queue prediction?

Interconnection queue prediction estimates which active queue projects are likely to withdraw before reaching commercial operation. GridSolve uses a Cox survival model trained on historical interconnection projects to assign withdrawal probabilities based on cluster studies, technology, service type, and queue position — then discounts substation capacity by expected withdrawals.

How does GridSolve help data center developers?

Data center developers need to know where grid power is realistically available, not just what hosting-capacity maps claim. GridSolve re-scores 4,600+ U.S. substations by withdrawal-adjusted deliverable megawatts — capacity that remains after likely queue withdrawals — so developers can prioritize sites with real interconnection headroom.

Does GridSolve predict data center interconnection requests?

GridSolve models generation-interconnection queue projects (solar, wind, storage, and similar), not large-load or data-center interconnection requests, which follow separate utility tariffs. Deliverable capacity per substation is a proxy for where grid headroom is likely to exist — not a load interconnection study.

What is withdrawal-adjusted interconnection capacity?

Withdrawal-adjusted capacity is the megawatts at a substation that are likely to remain deliverable after speculative queue projects withdraw. GridSolve subtracts each active project's predicted withdrawal share from queued megawatts to produce adjusted available capacity — a more realistic view than raw hosting-capacity figures.

What data sources does GridSolve use?

GridSolve combines ISO/RTO interconnection queue data from Lawrence Berkeley National Laboratory's Queued Up dataset with LBNL interconnection cost studies. Withdrawal rates are computed empirically within each region from resolved projects; the survival model validates against regional concordance metrics.