Where grid power arrives, and when.
Hyperscalers, colocation providers, fabs and other large loads now choose sites by when they can get power, not by the price of land. Time-to-Power gives every tracked site an energisation window based on public records, and shows each market's new supply year by year, so you can see which sites can realistically be powered by 2028–2030 and which won't be until the 2030s.
Every window is graded by the evidence behind it, from an announcement to construction on site. This public edition is anonymised. Subscribers see the site, its owner, the exact location, the substation and the contact route.
How much new firm power each market adds by a given year.
Move the year. Blue bars are new supply each market's interconnection queue is expected to deliver by the end of that year, discounted by that market's completion odds and counted at accredited, not nameplate, capacity. Dark blue is under signed agreements; light blue is not yet signed. Orange is the disclosed load of the sites we track whose P50 window falls by then.
The squeeze in one line: Southeast US (non-ISO): published large-load requests are 9× the accredited supply its queue is expected to add by 2030; Texas (ERCOT): published large-load requests are 6.9× the accredited supply its queue is expected to add by 2030; Great Britain: published large-load requests are 3× the accredited supply its queue is expected to add by 2030. Most of those requests will never connect, and that's the point: a buyer has to find the minority of sites that will.
What becomes available in each year, in one market.
- Typical request → energised
- A site filing today, P50
- Requests that get built
- …once an agreement is signed
- Active queue
- Tracked sites
Where a new site waits least, and where it waits most.
On outcome history, West US (non-ISO) has the shortest typical wait from request to energisation (2.2 years) and California (CAISO) the longest (6.7 years). Read the wait with the odds: a fast market where few requests get built is a lottery, not a shortcut.
| Market | Typical wait | A site filing today | Requests built | Built once signed | Accredited supply to 2030 | Published large-load requests | Tracked sites |
|---|---|---|---|---|---|---|---|
| Great Britain | 3.1 yrsmiddle half 1.7–6.0 | 2029 | — | — | 16.9 GW | 50 GW · 3× | 2023 with construction/power evidence |
| Mid-Atlantic & Ohio Valley (PJM) | 3.8 yrsmiddle half 1.9–5.6 | 2030 | 18% | 86% | 12.2 GW | — | 9614 with construction/power evidence |
| Texas (ERCOT) | 3.6 yrsmiddle half 2.5–4.6 | 2029 | 28% | 79% | 69 GW | 474 GW · 6.9× | 4219 with construction/power evidence |
| Southeast US (non-ISO) | 3.8 yrsmiddle half 2.7–5.0 | 2030 | 16% | 15% | 8.7 GW | 77.7 GW · 9× | 3511 with construction/power evidence |
| Central US (MISO) | 3.8 yrsmiddle half 2.5–4.6 | 2030 | 13% | 35% | 16.4 GW | — | 3514 with construction/power evidence |
| West US (non-ISO) | 2.2 yrsmiddle half 0.9–4.5 | 2028 | 14% | 71% | 28.1 GW | — | 318 with construction/power evidence |
| Southwest Power Pool (SPP) | 4.1 yrsmiddle half 2.2–5.3 | 2030 | 12% | — | 8.5 GW | — | 226 with construction/power evidence |
| California (CAISO) | 6.7 yrsmiddle half 4.2–8.3 | 2032 | 10% | 91% | 19 GW | — | 70 with construction/power evidence |
| New York (NYISO) | 4.9 yrsmiddle half 3.9–6.3 | 2030 | 13% | 70% | 1.8 GW | — | 51 with construction/power evidence |
| New England (ISO-NE)default range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | 2030 | 23% | 75% | 1.9 GW | — | 40 with construction/power evidence |
| Norwaydefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | — | 1816 with construction/power evidence |
| Germanydefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | — | 31 with construction/power evidence |
| Czechiadefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | — | 10 with construction/power evidence |
| Spaindefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | 23.7 GW | 10 with construction/power evidence |
| Croatiadefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | — | 10 with construction/power evidence |
| Italydefault range: no outcome history yet | 4.5 yrsmiddle half 3.0–7.0 | — | — | — | — | — | 11 with construction/power evidence |
Where Spain's grid can take a large new load today.
Spain's grid operator publishes, node by node, how much new demand each transmission substation can still accept. As of 2026-10-01, 443 of 937 nodes can take a new load of 100 MW or more, while 23.7 GW of demand requests are waiting for a decision and 18.5 GW has already been granted. The data-centre hub of Madrid has only 18 such nodes out of 91; Castilla y León has 70. 50 nodes are reserved for capacity tenders.
Filter the market for sites that can be powered when you need them.
Every large-load site we track, placed on one evidence ladder with a P10–P50–P90 energisation window. Filter by market, stage, size and the year you need power. Names, owners and exact locations are hidden here; they are in the subscriber edition.
194 sites located only to a state or country aren't drawn; they're in the list below. Sites are drawn at about 50 km. The exact location is in the subscriber edition.
The queue keeps growing faster than it delivers.
US interconnection requests, by year
Capacity entering the queue, withdrawing from it and reaching operation each year, across all US markets above.
One utility's data-centre pipeline, by quarter
The only large-load queue a US utility files project by project (Southeast, state commission docket): data-centre load requested, and the share under contract.
One evidence ladder for every site.
A site moves up only on evidence a buyer could check. Each rung keeps a set share of its market's typical wait still ahead; a date the record itself states (a first-power date, a building completion) becomes the optimistic edge of the window.
Signal
Announced, rumoured or land-assembly only
100% of the market wait still aheadFiled
Land-use, planning or connection application lodged
85% of the market wait still aheadApproved
Land-use or planning consent, building registration or public award
65% of the market wait still aheadPower-evidenced
Connection licence, utility agreement, on-site generation permit or stated first-power date
40% of the market wait still aheadConstruction
Construction started on site
20% of the market wait still aheadSupply, year by year
Every active request in each market's interconnection queue, placed in its proposed year (plus a year's slip before an agreement is signed), weighted by that market's historical odds of reaching operation, and counted at an accredited-capacity proxy: gas 90%, storage 50%, solar and wind 15%.
Waits and odds
US markets: how long requests filed since 2012 took to reach operation, and the share of the 2015–2019 cohort that got built. Great Britain: the register's stated connection dates, with completion odds by register status, because its outcome history isn't published yet.
What it is not
A screen, not a utility study. Large-load availability at a specific substation and the cost of network upgrades are in utility studies, not the public record. Subscribers get the sources for each site and a route to the people who hold that study.
Sources: interconnection queue data for 7 US grid operators and ~50 other US balancing areas through 31 December 2025 (LBNL Queued Up, 2026 edition); GB Transmission Entry Capacity register (daily); US and UK planning and permitting records; state building registrations; public award and licence registers; local hearing and press records. Published demand figures: ERCOT, Texas Senate briefing, Jul 2026; ERCOT RPG, Sep 2026; NESO Demand Queue call for input; NESO evidence to Parliament; REE, demand access capacity per node (monthly); State utility commission docket, quarterly large-load report. Model ttp-0.2, built 22 October 2026.
See the site, the owner, and the person who can give you an answer.
The public edition shows the market and the window. Subscribers get the rest, plus alerts when a site moves up the ladder or a queue, docket or rezoning changes its date.
Public edition
- Power clock and market ranking
- Anonymised site windows and filters
- Method and sources
Time-to-Power subscription
- Site name, owner and exact location
- Substation, utility and the record behind every rung
- Alerts on stage, queue and docket changes
- CSV export and introductions through Setcoin
Site & Power report
- One site or shortlist, studied in depth
- Queue contention and stranded positions nearby
- Owner verified with Setcoin Verify
Time-to-Power
About Time-to-Power
What is Setcoin Time-to-Power?
An intelligence product that estimates when grid power will be available for large-load sites such as data centres and fabs. It covers 504 tracked sites across 16 markets in the US, Great Britain and Europe, each with an evidence-graded P10–P50–P90 energisation window, and forecasts year by year how much accredited new supply each market's interconnection queue is expected to deliver.
Why does the energisation date matter more than the land?
Because power is now the constraint. Grid operators have published large-load requests far above the supply their queues are expected to add this decade, so a site that can be energised by 2028–2030 is worth far more than one that waits until the mid-2030s.
Where does the data come from?
Public records only: interconnection queues and their outcome history, the GB connection register, planning and permitting records, building registrations, public award and licence registers, utility commission dockets, and local hearing records. Every window states the evidence it rests on.
Are site names and owners shown?
Not on this page. The public edition is anonymised: no site, company, person, county, substation or source link appears. Subscribers see the site, the owner, the exact location, the substation and the record behind each rung, and can be introduced through Setcoin.
How accurate is a window?
It is a screen, not a utility study. Windows come from each market’s own history and the evidence on the site’s record, and are graded A to D by how strong that evidence is. Substation-level capacity and upgrade cost come from utility studies, which a Site & Power report addresses for a specific site.