The finding
Eighty-nine per cent of proposed energy capacity in America's largest grid never gets built. Analysis of 16,014 interconnection projects across four FERC-jurisdictional grid operators — PJM, CAISO, ISO-NE and NYISO, together serving roughly 60 per cent of the US population — shows a system in which failure is the norm. In PJM, 10.9 per cent of proposed capacity has reached commercial operation. Withdrawal rates run from 63 per cent in PJM's newer cycle queue to 83 per cent in NYISO. Eighty per cent of the speculative surge of 2021, when 1,352 applications were filed in a single year, has already been withdrawn.
The queue is not a queue. It is a map of regulatory failure — and reading that map is the highest-value capability an infrastructure investor can develop.
Why it matters now
The technologies the transition depends on fail most.
PJM conversion rates: solar 10 per cent, battery storage 3 per cent, offshore wind 1 per cent — against nuclear at 59 per cent and gas at 34 per cent. A system built to process hundreds of projects is processing thousands, and it is crushing the technologies it was supposed to enable. This is a design failure, not an engineering one.
Data centres are rewriting demand from the inside.
In NYISO, 48 active load projects represent 12.2 GW — 49 per cent of all active capacity in the region. These are consumption projects, not generation: hyperscale campuses requesting substation-level access. Virginia alone hosts 1,789 projects representing 136 GW, and developers have warned Dominion of 40 GW of coming data-centre load. Surging demand is colliding with a collapsing supply pipeline.
Regulatory geography decides survival.
The same 200 MW solar project faces a 2.6× gap in regulatory favourability between Loudoun County, Virginia (Regulatory Risk Score 7.15 out of 10) and Greene County, Pennsylvania (2.70) — quantified across six measurable dimensions: legislative direction, permitting timelines, utility-commission posture, zoning, environmental constraints and legal-challenge history.
The frameworks
| Regulatory Risk Score | Six-dimension, county-level scoring of policy risk, queue rules, cost-allocation exposure, permitting friction and grid capacity — a pre-acquisition screen for infrastructure private equity and lenders. |
|---|---|
| Convergence Zones | Queue-survival data overlaid on the regulatory map, to identify where demand, functional regulation and grid capacity align — and the dead zones where capital goes to die. |
| T3E (Alethia) | The Third-Order Effects Engine: cascading regulatory, macro and geopolitical consequences for asset valuation mapped across a 0–36-month horizon, as used in investor briefings. |
Implications for capital allocation
Underwrite regulatory risk explicitly. Most infrastructure models treat permitting as a line item or a sensitivity case. The queue data argues that this is backwards: regulatory risk is the primary determinant of whether a project exists at all. An investment committee approving a 200 MW solar project in a high-friction jurisdiction without a quantified assessment is taking a bet with an 80–90 per cent probability of failure, whether or not the model says so.
Concentrate capital in convergence zones. Returns to completion are highest where demand, functional regulation and supply constraints coincide — and the data identifies those places empirically. Virginia's PJM footprint is the clearest current example; the Hudson Valley, parts of New England and ERCOT's constrained West Texas corridor show similar dynamics.
Build the intelligence layer. No commercial product integrates queue data, state legislation, county zoning and utility-commission dockets into one continuously updated view. The fund that builds or procures that capability holds a structural information advantage. In a market where 89 per cent of projects fail, the ability to predict which 11 per cent will succeed is not merely useful; it is the entire game.