120 GW is queued at connection nodes that do not exist
The connections reform replaced named substations with strategic connection nodes in large parts of the register. 122 of them, across 35 regions, carry 120,085 MW of contracted storage. None of them physically exists. Between them they hold no Gate 2 capacity at all.
West Anglia alone accounts for 14,719 MW across nine nodes. Trent Valley South 13,569 MW across six. North Anglia 12,041 MW. South Anglia 9,443 MW. Derbyshire 8,959 MW across twelve.
They are not missing from the map, they are missing from the ground
The same is true of a second group. 172 connection points in the register are named after places that carry no substation in any published gazetteer: Navenby, Biggleswade, Coddington, Woolavington, Mablethorpe, Dunstable, Lavendon, Appleford, Cottered. These are proposed 400 kV substations under the transmission build programme, holding another 62,371 MW.
Put together, 182 GW of the queue is contracted at infrastructure that has not been built. That is not a data quality problem. It is the clearest single statement of delivery risk available anywhere in the published register.
How the tool draws something that has no location
NESO publishes no coordinate for a node that has no site, so inventing one would be dishonest. Each node family is drawn instead as a dashed circle over the region its name refers to, at a centre set by hand with a radius reflecting the extent of that region. It is labelled as inference about the name rather than a claim about a site, and the layer can be switched off. Capacity sits on the region, never on a point.