Editor’s note: This is the second in a series of articles from ASG discussing power, market design and policy.
In our previous blog, we laid out the numbers: data center demand is surging, capital is available, and projects are ready to move. But if you’ve spent any time inside this market, you already know that supply alone does not explain why so many well-financed, well-sited projects are still stuck.
The deeper problem is that the system built to deliver power was not designed for this moment. It was built for a slower, smaller, more predictable era of electricity demand growth. Today’s volume, pace, and scale expose the flaws in that design.
This piece diagnoses those flaws. Not the headline supply gap, but the structural barriers embedded in grid processes, market rules, and regulatory timelines that determine whether a project actually gets built. Understanding that distinction, system design versus raw supply, matters for anyone operating in the digital infrastructure space right now. It changes what you plan for, what you negotiate, and where you spend your effort.
The Supply Interconnection Queue Is Broken
The generation interconnection backlog has been a problem long in the making. The first-come, first-served process was never built for today’s volume or scale of new interconnection requests. Lawrence Berkeley National Laboratory 2026 Queued Up report estimates about 2,060 GW of capacity sat active in U.S. interconnection queues at the end of 2025 across 8,200 projects, close to double the capacity of the entire existing U.S. power plant fleet. Total queue volume actually fell 10% in 2025 but was driven by withdrawals rather than by projects getting built. Wait times kept climbing. The median project reaching commercial operation in 2025 spent 61 months in the queue, up from 36 months in 2015 and 22 months in 2008. Another 549 GW holds a draft or executed interconnection agreement and still has not reached commercial operation.
Speculative projects clog the queue alongside legitimate ones. When speculative entries drop out, remaining projects often get restudied, resetting timelines that were already years long. FERC’s Order 2023 was a direct response to this dynamic. It replaced the old serial, first-come, first-served study process with a first-ready, first-served cluster study model, requiring commercial readiness deposits and imposing penalties on transmission providers that miss study deadlines. It’s a meaningful reform, but it’s also not a complete fix. Physical upgrades, permitting for new lines and substations, and jurisdiction-specific approvals still sit downstream of the study process itself, and those steps are where much of the real time gets consumed.
The practical implication for developers: even a fully financed, well-sited power project can realistically face a four-to-seven-year wait to interconnect. That timeline doesn’t move because a project is well capitalized or is in high demand.
Cost Allocation: Who Pays for Grid Upgrades?
Cost allocation is among the most consequential and least visible barriers in the market. Large load connections routinely trigger transmission and distribution upgrades. While some costs are clearly tied to and allocated to a large load connection, responsibility for broader system grid upgrades may be spread across all of a utility’s customers or even administratively unresolved.
Two broad approaches exist: socializing costs across a utility’s ratepayer base, or assigning them directly to the interconnecting customer. Utilities take different approaches subject to different rules and there are tradeoffs in most circumstances. Broadly sharing costs spreads risk but raises equity questions about who benefits versus who pays. Direct assignment aligns cost with cause but often results in free-riding beneficiaries or if too high, can render marginal projects uneconomical or push developers toward markets with more favorable rules. Under FERC Order 2023’s cluster process, network upgrade costs are allocated proportionally among interconnection customers being studied in the same study cluster, which helps within a cluster but does not resolve the broader socialization-versus-assignment debate playing out at both the state and federal levels.
The result is a patchwork of utility and regional approaches that creates inefficiencies and uncertainty. Uncertainty about which rules will apply, and whether they will hold, delays investment decisions across the board, independent of any single project’s merits.
Market Signals Aren’t Translating Fast or Strongly Enough
In deregulated markets, price signals are supposed to drive new generation investment. That theory is not manifesting into reality. For a myriad of reasons, forward capacity and energy markets have not provided consistent signals to drive meaningful new construction. Administratively, changing regulations like market caps further limit the strength of those signals. Of course, even if the market signal was strong enough to incent new builds without a long-term contract, challenges in the planning and interconnection queues still slow them down.
Regulated markets face a different kind of bottleneck. Integrated resource planning cycles and rate cases are deliberate by design. That deliberateness served the system well when demand grew predictably but is now misaligned with the current pace of demand growth and faces cost allocation challenges that didn’t previously exist.
The throughline across both market structures: capital is available. But market structure is not converting that availability into a willingness to deliver supply on a timeline that matters to the businesses waiting on it.
Regulatory Timelines Operate on a Different Clock
Permitting, siting, environmental review, and certificate of need processes are layered, jurisdiction-specific, and largely sequential. Each layer has its own clock, and those clocks rarely run in parallel.
The structural mismatch is stark. Data center developers typically work on 18 to 24 month delivery timelines, while these other regulatory processes can run five to seven years. The mismatch doesn’t necessarily exist because of misaligned interests. They are incompatible because they were built to solve different problems on different time horizons. This isn’t a criticism of regulators. It’s a genuine structural incompatibility, and any real reform has to address it directly rather than treat it as a communication or coordination issue.
This reality reinforces what we said before: you can’t buy your way through a system that moves at a regulatory pace. Capital accelerates financing and construction. It doesn’t accelerate a five-year IRP cycle or a multi-agency environmental review.
Then Came the Data Centers
The planning, supply interconnection, and cost allocation challenges we consider above were rendering our electricity system increasingly unfit for purpose for a long time before exploding data center demand emerged as today’s central challenge. But the need to implement solutions wasn’t as acute in the low- or no- load growth period this country has experienced, generally, for the last two decades. Market participants, regulators and legislators have been dealing with these challenges incrementally, and on the standard regulatory timeline. Suddenly with the advent of rapidly increasing demand largely attributable to data center development, these challenges seem to be colliding on center stage – commercially and politically.
The System Is Stressed, Not Broken
Encouragingly, the current state of affairs does not need to be permanent. These challenges are the symptoms of a system under pressure it was not designed to handle, which means reform is both possible and already underway. In June, for example, the Federal Energy Regulatory Commission (FERC) took a direct step into this problem, at least as relates to large load interconnection. It issued a set of orders that require the six RTOs under its jurisdiction to build a new framework to manage large load interconnection. Cost-shift protection is central to the directive: RTOs must publicly post, by utility, the aggregate large load additions in their queues, the specific network upgrades planned to accommodate them, and the associated cost estimates. RTOs must also establish a minimum revenue contribution that large load customers pay toward the utility’s cost of service, backed by financial security. The FERC action is a meaningful step toward transparency in a part of the process that has largely operated without it, though it applies only to RTO territory. Vertically integrated utilities outside RTO footprints are not subject to the same requirement, which means the patchwork gets sharper, not flatter, in the near term. The RTOs have 60 days to file proposed rules, and implementation, not the order itself, will determine whether this actually changes outcomes for developers.
These FERC orders are one live example: real movement, still early, with outcomes that depend on how each RTO’s stakeholder process actually plays out over the next several months. Similar work is underway in state commissions and inside RTO processes more broadly. It takes time, and it requires sustained industry engagement to succeed rather than stall out as another well-intentioned proceeding.
In the meantime, developers and investors are not waiting for the system to catch up. The ones winning right now are building strategies around the market’s current realities rather than the market they wish they had. Our next piece examines what that looks like in practice: power-first versus site-first thinking, utility partnership over pure procurement, and the difference between power on paper and power that is actually deliverable.
Navigating these barriers takes experience working inside all of these processes at once, interconnection, regulatory approval, cost allocation, and market design. If you are working through these challenges on a live project, we can help. Let’s connect.


