# Denmark has about 70 GW of grid connection requests for a grid that peaks around 7 GW. How should planners treat a queue like that? — retained record

> Retained reddit record from reddit:energy.

Canonical: https://highsignal.app/data/records/f114aad187a9eb1b

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:energy)  published 2026-10-28 05:21 ingested 2026-10-29 02:41

[original source ↗](https://www.reddit.com/r/energy/comments/1xhfi57/denmark_has_about_70_gw_of_grid_connection/)

## retained text



```
I was just listening to the Volts episode on Denmark from 16 Oct. david roberts put it to Kristian Jensen (CEO of Green Power Denmark) that the country has about 70 GW of requests to connect to a grid with roughly a 7 GW peak. Jensen's answer was basically: that's not real demand.

His explanation is the familiar one. data centre developers are "throwing darts at a board", applying at three or four locations when they only want one, plus batteries hoping to get on. he thinks something like 20 to 25 GW in ten years is the realistic target, which is still a huge increase.

He also made a point I found more interesting.. some existing data centres already have a big connection they barely use, his example was a 1 GW connection running about 200 MW. so there's headroom sitting inside contracts that are already signed.

Env consultant here, and this is the bit that bugs me from the planning side. a queue is a count of applications, not a load forecast. if a network plan or a cumulative impact assessment takes the queue as an input, it inherits every duplicate, and you end up sizing substations, lines and land take for load that never shows up. and going the other way, nobody's counting the unused capacity in existing connections.

the numbers I'd actually want are what share of requests ever reach energisation, and requested vs actually used capacity on existing connections. as far as I can tell neither gets published routinely, in Denmark or elsewhere.

For people who do grid planning or interconnection work: what do you actually use instead of the raw queue? historical attrition rates, deposits/milestones, something else?

sources:  
https://www.volts.wtf/p/whats-up-with-denmark
```



document fetched 2026-10-29 00:40 · canonical https://reddit.com/r/energy/comments/1xhfi57/denmark_has_about_70_gw_of_grid_connection
