The $2 Trillion Interconnection Queue Bottleneck: What It's Costing Solar+Storage Projects
- 3 hours ago
- 17 min read
1. Baseline Conditions Have Changed
There is a line most developers have heard on some project call: "We’re just waiting on the interconnection study." A decade ago, this was a procedural hurdle. In 2026, it has become an existential constraint What was once a procedural inconvenience has become, in several major markets, the defining constraint on whether a project gets built at all — and if it does, whether it gets built on a timeline that still makes financial sense.
The data from Lawrence Berkeley National Laboratory's 2025 Queued Up report is worth sitting with. As of the end of 2024, approximately 2,290 gigawatts of generation and storage capacity were actively seeking grid interconnection in the United States. For context: the entire installed capacity of the U.S. power fleet is roughly 1,200 GW. The queue, in other words, is nearly twice the size of the infrastructure it is trying to connect to. The system was not designed for this volume, and it shows.
What shows more clearly is what happens to projects that enter that system. Of all capacity that submitted interconnection requests between 2000 and 2019, only 13% had reached commercial operation by the end of 2024. Seventy-seven percent had been withdrawn. The queue, by this measure, functions less like a pipeline and more like an attrition mechanism — a long, expensive process through which most projects do not emerge on the other side.
There is a particular kind of optimism that has historically surrounded the phrase "we're in the interconnection queue." It carries the implication of forward motion. The queue, in this framing, is a waiting room. The doctor has been double-booked for five years.
The timeline problem is compounding. The median duration from interconnection request to commercial operation has doubled over the past two decades: under two years for projects that reached COD between 2000 and 2007, over four years for those completing between 2018 and 2024, and a median of five years for projects that reached COD in 2023 specifically. These are not outlier cases. They are the central tendency of the market.
The 2024 data introduced a wrinkle that deserves careful reading. Total active queue volume fell 12% year-over-year — the first significant decline in years — driven partly by fewer new submissions and partly by a record wave of withdrawals: over 700 GW in a single year. Some analysts read the decline as a sign of queue rationalization. A more precise reading is that the attrition is accelerating. One-third of 2024 withdrawals occurred at the facility study or interconnection agreement phases — late enough to disrupt the study assumptions of neighboring projects, triggering re-studies and cascading delays for projects that remained. The queue got shorter. It did not get faster.
This is also a live story, not an archival one. PJM — the grid operator serving 65 million people across 13 states and Washington D.C. — reopened its standard interconnection process to new applicants at the end of April 2026, for the first time since closing it in 2022 to clear an accumulated backlog. The reopening attracted 811 new project applications almost immediately. For context: projects that reached commercial operation in PJM in 2025 had spent, on average, eight years navigating the process that just reopened. The institutional memory of that experience does not appear to have dampened enthusiasm for re-entry.
FERC Order 2023 mandates meaningful reforms: a shift to cluster-based study processing, deposit requirements designed to deter speculative submissions, and deadlines intended to reduce procedural drag. These are structural improvements, and they matter. But implementation is uneven across ISOs, the reforms apply to new queue submissions rather than the existing backlog, and — as LBNL notes — it is too early to assess their full impact on actual timelines. Developers building projects today are navigating a queue that was formed before most of those reforms took effect.
The situation varies significantly by region, but not in a way that offers much relief. CAISO, MISO, PJM, and SPP each carry backlogs measured in hundreds of gigawatts. ERCOT operates under a different regulatory structure but faces its own study capacity constraints. Non-ISO utilities — which collectively cover a substantial portion of U.S. territory — have widely inconsistent study processes, timelines, and cost allocation methodologies. There is no single queue problem in the United States. There are dozens of them, operating under different rules, producing different outcomes, and creating a patchwork of interconnection risk that no irradiance map captures.
This is the context in which every solar and solar-plus-storage project now sits. Not an exceptional development cycle with unusually bad queue luck. The baseline.
2. What Queue Position Costs
Interconnection is categorized, across most development pro formas, as a line item under soft costs. This is technically accurate and practically misleading. Soft costs, in the conventional framing, are the fees and studies that precede real money. What the queue has become, at current timelines and cost trajectories, is something closer to a slow-motion capital event — one that compounds quietly against a project's financial model while the development team waits for study results that arrive, on average, years after they were requested.
The cost escalation is documented and significant. LBNL's interconnection cost data — covering more than 3,000 projects across ISO and non-ISO regions — shows average interconnection costs for solar projects running at approximately $167 per kilowatt, with roughly 75% of that figure attributable to required network upgrades rather than direct connection costs.
For projects completing between 2018 and 2024 in non-ISO balancing authorities, average costs reached $194/kW. In PJM specifically, mean interconnection costs for projects active in the queue between 2020 and 2022 ran to $240/kW — against a baseline of $29/kW in the two years prior. That is not cost escalation. That is a different category of financial exposure operating under the same name.
The withdrawal data sharpens the picture further. Network upgrade costs for projects that withdrew from interconnection queues averaged 70% of total interconnection costs — and projects that withdrew from PJM's queue between 2020 and 2022 faced mean interconnection costs of $599/kW. Projects do not typically withdraw because the engineering is insurmountable. They withdraw because the cost of proceeding — often known only after years of study — exceeds what the financial model can absorb. The queue, in other words, is not just slow. It is an instrument through which interconnection cost uncertainty gets resolved at the worst possible moment in a project's capital lifecycle.
That moment matters because of what surrounds it. A project in the queue is not idle. It is carrying land option costs, development team overhead, insurance, and — if financing has been arranged — the cost of capital against committed but undeployed funds. For projects at 50 MW or larger, working capital requirements can climb from under 35% of total development costs to over 50% if development stretches beyond 15 months. At five-year median queue timelines, the compound effect of those holding costs against a fixed land option and a project that may or may not clear its interconnection agreement is not a rounding error. It is a material contributor to the withdrawal calculus.
Equipment price exposure sits alongside it. A project that entered the queue in 2022 modelled panel and inverter pricing against 2022 market conditions. If it reaches interconnection agreement in 2026 or 2027, it is procuring against a different tariff environment, a different supply chain, and in some cases a materially different cost basis. The financial model written at queue entry is not the financial model that closes financing — and in most cases, the gap between them is not explained by irradiance variability or O&M assumptions.
The ITC dimension adds a layer of deadline mechanics that the queue timeline does not accommodate naturally. The construction-start requirement for the ITC's safe harbor — satisfying the 5% incurred cost test or beginning physical work — creates a fixed calendar pressure that runs independently of where a project sits in its interconnection study cycle. A project waiting on a system impact study result cannot manufacture its own construction-start eligibility. Developers have become creative about this boundary: ordering long-lead equipment, breaking ground on access roads, satisfying the physical work test in ways that are technically compliant but commercially awkward for a project whose interconnection path remains unresolved. It is a form of project development performed in sequence by necessity but priced as if it could be planned in parallel.
Lenders have noticed. Projects with signed interconnection agreements attract materially stronger interest from investors and lenders, enabling faster progression toward financing close. The corollary — which any project finance team can confirm from recent deal flow — is that projects without executed IAs are increasingly being offered financing structures that price in interconnection uncertainty through wider spreads, more conservative debt sizing, or conditions precedent that defer commitment until IA execution. The interconnection agreement has become, in practical terms, a financing instrument as much as a regulatory milestone. Waiting five years to obtain it has financial consequences that begin accumulating on day one.
Queue delay duration mapped against the cost categories that most pro formas underweight at time of queue entry:
Queue Delay | Holding Cost Exposure | Equipment Price Risk | Financing Gap / Carry Cost | ITC Timing Pressure |
1–2 years | Low–moderate | Low | Manageable | Limited |
2–3 years | Moderate | Moderate | Increasing | Elevated |
3–4 years | High | High | Material | Significant |
4+ years | Very high | Very high | Project-threatening | Often unrecoverable |
Sources: LBNL Interconnection Cost Analysis; industry-reported development economics. Ranges are conservative and vary by market, project size, and capital structure.
3. Where the Engineering Decisions Actually Are
The interconnection study process has a formal sequence that most developers are broadly familiar with and most pro formas are not built around. The sequence runs from feasibility study through system impact study to facilities study — three distinct analytical exercises, each contingent on the last, each capable of producing a cost estimate that bears little resemblance to the one that preceded it.
The feasibility study tells you whether interconnection is physically possible at your proposed point. The system impact study tells you what the grid needs to accommodate your project. The facilities study tells you what that will cost and who pays. By the time the third document arrives, a developer has typically been in the queue for two to three years and has already made capital commitments that assumed a different answer.
This is not a process failure unique to any one ISO. It is the structural consequence of running a sequential, study-intensive review process at a volume it was not designed to handle — and of treating interconnection engineering as something that happens to a project, rather than something that shapes it from the beginning. The distinction between those two postures is where most of the controllable risk sits.
What the feasibility study does not tell you
The feasibility study is often treated as a green light. A project clears feasibility, the development team notes that interconnection looks viable, and the assumption gets baked into the schedule and financial model accordingly. What the feasibility study actually produces is a load flow and short circuit analysis at a specific point in time, against a grid snapshot that will change — sometimes materially — by the time the system impact study runs. Projects that joined the queue before or alongside yours affect your system impact results. Withdrawals affect them too, in the opposite direction.
The feasibility study result and the system impact study result are connected by a thread of assumptions, not a continuous analysis, and the gap between them is where upgrade cost surprises live. For projects in congested queue clusters — which describes most active development corridors in CAISO, PJM, and MISO — the system impact study is running a shared analysis across multiple projects simultaneously. How your project interacts with adjacent queue entrants determines a portion of your network upgrade allocation. That interaction is not visible at feasibility. A developer who has not engaged engineering to think through the likely cluster composition and its implications for upgrade cost allocation is not making a fully informed site selection decision. They are making a feasibility study decision and calling it due diligence.
Substation proximity as a primary filter — not an afterthought
A parcel close to a high-capacity substation with available headroom faces a shorter, cheaper interconnection path — and a lower risk that studies come back with deal-breaking upgrade requirements. This observation has been true for as long as utility-scale solar has existed. What has changed is the degree to which substation proximity and available capacity headroom now function as the primary site filters in competitive development markets, with irradiance becoming a secondary optimization once the grid access question is resolved.
But distance is only one variable. Voltage level matters — utility-scale projects connecting to 69kV or higher transmission infrastructure face different study assumptions and different network upgrade exposure than those connecting at distribution voltage. Available capacity headroom matters — proximity to a high-voltage line does not guarantee sufficient capacity to accommodate additional generation from a solar project. And the existing queue of projects ahead of you at the same substation matters most of all, because those projects will consume available headroom before yours does, and their study outcomes will affect yours.
Proximity to existing transmission corridors shortens new right-of-way requirements and can lower interconnection costs — but it also introduces constraints on clearances, outages for cutovers, and switching windows that an engineering team needs to assess before site control is exercised, not after. The point is not that proximity is a simple positive. It is that the engineering variables embedded in a site selection decision are too consequential to leave to a desktop review of publicly available substation maps.
Protection coordination: the study nobody schedules early enough
Protection coordination is among the more reliably underestimated elements of interconnection engineering, partly because it surfaces formally late in the study sequence and partly because its cost implications are not obvious until they are. A solar or solar-plus-storage project connecting to the transmission grid introduces generation into a system designed around different fault current assumptions, different relay coordination schemes, and in older infrastructure, different equipment capabilities than modern inverter-based resources require.
The protection coordination study determines whether your project's contribution to fault current is compatible with existing relay settings and protection devices at the point of interconnection and upstream. When it is not — and in congested, heavily queued corridors with aging infrastructure, the probability of incompatibility is not negligible — the required remediation typically involves relay replacements, setting changes, or in some cases physical infrastructure upgrades at the substation. These are not line items that appear on a feasibility study. They appear in the facilities study, at a point in the project lifecycle when the development team has already committed to a timeline, a capital structure, and in some cases a power purchase agreement with a fixed delivery date.
The remediation is solvable. The surprise is not.
Protection coordination scoped at pre-development — as part of the site assessment rather than as a study triggered by the facilities study output — converts a late-stage cost event into an early-stage design input. The upgrade cost does not disappear. The scramble does.
What BESS co-location changes — and what it does not
Battery energy storage co-location has become a near-universal consideration in utility-scale development, driven partly by the value stack and partly by the widely discussed observation that storage attachment can improve interconnection outcomes. The nuance is important. Storage co-location can reduce network upgrade requirements in specific grid configurations — where the project's net export profile is smoothed by the storage dispatch — and has in some cases accelerated queue review timelines in ISOs that prioritize hybrid resources for reliability reasons.
What storage co-location does not do is resolve the underlying protection coordination complexity. A hybrid PV-plus-storage project introduces additional inverter-based generation, additional protection coordination requirements, and in most cases additional facilities study scope. Developers who model storage attachment as a queue shortcut without running the engineering implications of the hybrid configuration through the full study sequence are optimizing for queue position at the cost of study outcome certainty. The interconnection queue is long enough without discovering mid-facilities-study that the BESS configuration requires relay upgrades that were not in the budget.
The pre-development interconnection checklist — what to assess before site control. Download the one-page reference guide.
4. How Developers Are Adapting — and What's Still Broken
The development community has not been passive in the face of a five-year median queue timeline. It has adapted — methodically, creatively, and in some cases expensively — in ways that deserve honest assessment. Some of the adaptations are genuinely improving project outcomes. Others are sophisticated-sounding responses to a structural problem that sophistication alone cannot solve. The distinction is worth making clearly, because the market currently rewards the appearance of strategic positioning almost as much as the position itself.
What is actually working
The most durable adaptation has been the shift in how experienced developers approach site selection. Interconnection timelines, not land readiness, now determine feasibility in most active development markets — and the developers who have internalized this are screening sites on grid access criteria before the land acquisition process runs in earnest. Hosting capacity maps, which a growing number of utilities now publish, give developers an early read on where available headroom exists at the distribution and transmission level. The maps are imperfect — frequently lagged, sometimes incomplete, occasionally optimistic — but used alongside substation capacity data and queue position analysis, they provide a meaningful pre-feasibility filter that was not systematically applied five years ago.
The second real adaptation is earlier utility engagement. Developers are engaging utilities earlier in the acquisition cycle and requiring defined energization milestones before committing capital. This sounds procedural. In practice it represents a meaningful shift in how interconnection risk is priced at the deal level — from a line item that gets refined post-site-control to a gate condition that shapes whether site control is exercised at all.
SPP's Surplus Interconnection Process — which allows a storage project to connect using existing unused interconnection capacity at an operating facility under a Surplus Interconnection Agreement rather than a full GIA — has emerged as a genuinely faster pathway, with timelines of six to twelve months rather than several years. It is geographically constrained, technically conditional, and dependent on finding an operating facility with available headroom. It is not a broadly replicable strategy. But for developers with the portfolio scale and regional relationships to identify and negotiate these opportunities, it represents a real timeline compression.
What is not working as advertised
FERC Order 2023 is the most significant structural reform to the interconnection process in two decades, and the instinct to credit it with future improvement is reasonable. The current evidence is more complicated. Early results in the first cluster cycles have been mixed: fewer restudies, but upgrade cost allocations have in some cases increased as shared network upgrades are split among fewer queue participants than expected. The cluster study model was designed partly to distribute upgrade costs more equitably across projects that collectively drive the need for grid improvements. When clusters thin out through withdrawals — which has been happening at record rates — the remaining projects bear a larger share of the allocated costs than the original study assumed. The queue gets shorter, the cluster gets smaller, and the survivors pay more.
The timeline picture is similarly complicated. The 150-day cluster study deadline mandated under Order 2023 is a meaningful accountability mechanism. What the 150-day window does not address is the period before the cluster study begins: the customer request window, the customer engagement window, and in many regions a transition process backlog that predates the new framework entirely. Developers who entered the queue before Order 2023's compliance dates in their ISO are navigating a hybrid process — part legacy serial study, part transitional cluster — with the attendant uncertainty that implies.
The broader limitation of the adaptation strategies the market has converged on is that they are, by definition, optimization within a constrained system rather than resolution of the constraint. Prioritizing substations with available headroom works until enough developers prioritize the same substations, at which point the headroom disappears and the strategy recycles. Sound portfolio management will be standard practice across the industry within two or three years, at which point its differentiation value approaches zero. The developers who move early capture the advantage. The advantage is not structural.
What is structural is the transmission infrastructure itself. Transmission expansion is the only long-term fix — a conclusion that the reform process, the cluster study methodology, and every sophisticated site selection strategy implicitly acknowledges by working around it rather than through it. The U.S. added approximately 1,000 miles of new high-voltage transmission in 2023. It needs, by most credible estimates, several times that annually to accommodate the generation pipeline currently in queue. The math does not close on optimization alone.
The market knows this. It has adjusted. It has not solved it — and the distinction between those two things is where the financial exposure described in the previous section continues to accumulate, regardless of how carefully a development team selects its next site.
5. What Belongs in Your Engineering Scope From Day One
The conventional project development model places engineering after site control, after preliminary financial modelling, and in many cases after a queue position has been secured. This sequencing made reasonable sense when interconnection timelines were measured in months and study outcomes were largely predictable from public substation data. It makes considerably less sense when the queue is five years deep, upgrade cost allocations are determined by cluster composition that changes with every withdrawal, and the facilities study result is the document your lender is waiting for before they will size your debt.
Engineering that begins after site control is not early-stage engineering. It is engineering that arrives at the second act of a three-act problem.
What belongs in the pre-development scope — before site control is exercised, before queue entry, before the financial model is locked — is a structured interconnection feasibility assessment that goes beyond desktop review of publicly available substation maps. Not a formal study. An engineering-led analysis that answers, with reasonable confidence, the questions that the feasibility study will eventually answer formally: What is the likely point of interconnection? What voltage level does the project connect at, and what does that imply for protection coordination scope? What is the approximate available headroom at the nearest viable substation, accounting for projects already in queue? What is the realistic upgrade cost range, and at what point does that range make the site financially unviable?
These questions are not unanswerable before queue entry. They are not answered before queue entry because they have not historically been asked before queue entry. The habit of treating interconnection as a study process rather than an engineering discipline is what creates the conditions under which a facilities study result arrives three years into a project's development cycle and reclassifies a viable site as an unviable one.
The protection coordination question
Ask your engineering team, before site control, whether the proposed point of interconnection is in a corridor with aging protection infrastructure and high inverter-based resource penetration. This is not a question that requires a formal study to answer directionally. A competent protection and control engineer can assess the publicly available one-line data, the vintage of the substation equipment, and the composition of projects already interconnected at or near that point, and give you a meaningful early read on protection coordination risk. If the answer suggests material relay upgrade exposure, that exposure belongs in your development budget and your financial model — not in a facilities study result that arrives after you have committed to a construction timeline.
The ITC timing question
If your project is subject to ITC safe harbor requirements — and most utility-scale and C&I solar projects in active development are navigating some version of this — the construction-start timeline needs to be engineered backward from the interconnection agreement milestone, not forward from a queue entry date. The two calendars do not naturally align. Making them align requires knowing, with engineering-informed confidence, what your interconnection path looks like and what the realistic IA execution timeline is in your ISO and at your specific point of interconnection. This is not financial modelling work. It is engineering work that informs financial modelling — and it is most valuable at the point in the development process when financial models are still adjustable.
The BESS co-location question
If storage attachment is in the project scope — and increasingly it is, for both financial and queue-strategic reasons — the hybrid configuration needs to be engineered as a unified interconnection problem, not as a solar project with storage added. The protection coordination implications of a hybrid PV-plus-BESS system are distinct from those of a standalone solar project. The interconnection study scope is broader. The facilities study will reflect that scope regardless of when in the development process the BESS was added to the design. Adding it early, with the full engineering implications modelled, is cheaper than adding it late. This is not an argument that has ever been seriously disputed. It is an argument that the development timeline routinely overrides in the interest of keeping the initial feasibility analysis clean and the early-stage numbers attractive.
What this requires of the engineering relationship
None of this is achievable if engineering is engaged as a study execution function rather than as a development partner. The pre-development analysis described above is not a standard deliverable in a typical engineering scope of work. It requires engineers who understand the interconnection process well enough to reason about it before the process has started — who can read a grid configuration, assess a queue composition, and form a defensible view on interconnection risk without the formal study apparatus that most engineering work is organized around.
It also requires a development team willing to pay for that analysis at a stage in the project where the site has not yet been controlled and the capital commitment has not yet been made. Pre-development engineering on a site that may not be acquired feels like spending money on an option that has not been exercised. The alternative — acquiring the site, entering the queue, and waiting three years for a study result that reprices the project — is also spending money on an option. It is simply a more expensive and less reversible version of the same bet.
The interconnection queue will get shorter, eventually. FERC Order 2023 will work its way through the system. Transmission infrastructure will be built, incrementally and expensively, in the corridors where it is most needed. The structural constraint will ease — on a timeline measured in decades rather than development cycles.
In the meantime, the projects that close financing, meet construction-start deadlines, and reach COD on a schedule their pro formas can recognize are, with striking consistency, the ones where the engineering was in the room before the land was under control. Not because engineering is magic. Because five years is a long time to discover you asked the wrong questions about the wrong site.
If your project is in or approaching the interconnection queue, we can review your current engineering scope and identify where the exposure sits. No proposal required. Get in touch.
Comments