Verify · the pickaxe, in the open
How the LOADSTAR Number is computed
Every figure below is pulled live from official U.S. ISO interconnection queues and recomputable from the raw JSON. This is the whole derivation — no hidden step. Don't trust; recompute.
The one-line claim, straight from the single source that every LOADSTAR surface reads:
Where U.S. ISO generation interconnection queues publish a machine-readable interconnection-agreement status, the signed share our reading finds varies widely by queue: CAISO 78% · ISONE 66% · SPP 39% · PJM 12% · MISO 0% · ERCOT 0% — a spread that reflects each queue’s disclosure and market design as much as project maturity. In the other 1 queues (NYISO), no usable machine-readable status field appears in our reading, so their 25.2 GW count as unverified — with no claim made about them. Across all 7 queues (50 MW and up), 944.1 GW is requested and 146.9 GW (about 15.6%) is machine-verifiable as signed — an inventory floor, not a national rate: queues are mostly early-stage by nature (historically, 13% of 2000–2019 requests reached operation and 77% withdrew — LBNL).
The derivation, step by step
1Pull the live queues.
Active interconnection-queue rows, 50 MW and up, from the seven live U.S. ISOs — PJM, ERCOT, CAISO, MISO, SPP, ISO-NE, NYISO. This is the generation queue (power plants & storage seeking to connect), not data-center load requests. Any fuel type. Bulk annual datasets without a status field (e.g. LBNL) are reported separately, never mixed in.
2Sum requested, per ISO.
Requested = ∑ capacity of those active rows. Signed = the subset showing a signed agreement or construction status in that queue's own published data (how we read it per ISO is in the next section).
| ISO / RTO | Requested (GW) | Signed (GW) | Requests | Status coverage |
|---|
| ERCOT | 299.0 | 0.0 | 1,117 | 100.0% |
| MISO | 229.8 | 0.2 | 1,006 | 15.7% |
| SPP | 150.7 | 59.4 | 692 | 39.4% |
| PJM | 148.7 | 17.7 | 647 | 11.9% |
| CAISO | 78.8 | 61.7 | 247 | 78.3% |
| NYISO | 25.2 | 0.0 | 137 | 1.1% |
| ISONE | 11.9 | 7.9 | 37 | 96.2% |
| ALL SEVEN | 944.1 | 146.9 | 3,883 | — |
3Divide.
6.4× requested vs verified-signed · ~16% verified (a floor)
The Number is the ratio; the floor is the honesty. 944.1 GW requested ÷ 146.9 GW verified = 6.4×. ‘Verified’ is a floor by construction: in 1 of the 7 queues our reading finds no machine-readable agreement-status field, so those GW count as unverified — never as speculative. The ratio is therefore an upper bound on the true requested-to-signed gap. An unverified request is not a phantom.
How we read “signed”, per ISO
‘Signed / under construction’ is our reading of each ISO's published queue-status data, applied consistently across all seven queues — it is not a single certified field (there is no national one). Exact rule per ISO, mirrored from the collector:
| ISO / RTO | What we count as “signed” |
|---|
| PJM | our keyword read of the queue's published status text — the one ISO that publicly tiers signed agreements, and the source of ~99% of all confirmable ‘signed’ MW. A floor, not a boundary. |
| CAISO | column «Interconnection Agreement Status» == “Executed”. |
| ISO-NE | column «Interconnection Agreement Status» == “Executed”. |
| SPP | column «Status (Original)» begins with “IA FULLY EXECUTED”. |
| MISO | column «Post Generator Interconnection Agreement Status» in the executed/post-GIA set. |
| ERCOT | column «IA Signed» present (a signature date = a signed agreement). |
| NYISO | column «S» (stage code) ≥ 11 — IA Completed, Under Construction, or In Service. |
The caveat travels with the number
This is the GENERATION interconnection queue — power plants and storage seeking to connect to serve future load growth, of which AI data centers are one driver — not data-center (large-load) requests, which most ISOs do not publish. ‘Verified’ means OUR consistent reading of each queue’s own published status field (e.g. CAISO’s ‘Interconnection Agreement Status: Executed’, MISO’s post-GIA stage), including projects already under construction — which require an executed agreement; it is not a certified or audited figure. Status taxonomies differ by ISO, so per-queue shares are not directly comparable, and a near-zero share can be design rather than weakness: ERCOT runs a connect-and-manage regime outside FERC’s interconnection jurisdiction, under which its published active queue rarely shows executed-agreement dates, and MISO’s post-GIA stage field appears on only a small slice of its queue. Where our reading finds no usable machine-readable field (currently NYISO), that says nothing about those ISOs’ disclosure — it may be our parsing limit — and those GW are counted as unverified, never as phantom. An unverified request is NOT a phantom: queues are unsigned by nature and status is graded (FERC Order No. 2023 replaced serial first-come with first-ready cluster studies). The raw queue is not deduplicated; the same project can appear more than once (part of what FERC’s RM26-4 show-cause targets). Historically most queued generation never reaches operation: of the capacity that requested interconnection from 2000 to 2019, 13% reached commercial operation by the end of 2024 and 77% was withdrawn (LBNL, ‘Queued Up: 2025 Edition’) — that cohort ran under the serial regime Order 2023 has since replaced. Descriptive of today, not a forecast.