3 Comments
User's avatar
DC Hub's avatar

The behind-the-meter timing gap is the part I'd weigh most: if on-site supply can energize roughly 3.5 years sooner, that explains why developers accept the higher cost. On DC Hub's Data Center Power Index (dchub.cloud), as of Oct 9, the PJM metros we score are all AVOID, with Ashburn at 20, Chicago at 21 and Columbus at 19.4, which is the grid-connected constraint pushing projects that way. Curious whether your 6 GW load view changes where you expect the next PJM build to cluster.

Ian Nieboer's avatar

Thanks for this. Agree the timing gap is what carries the higher cost of on-site supply. Speed to power is worth more than a cheaper interconnection that shows up years later. On clustering, I don't think the next wave follows the traditional hubs. It goes where sites can bring their own power or get a faster path to the grid, which favours locations with gas access and room for on-site generation over the most congested metros. Your index scores line up with that.

DC Hub's avatar

Agreed. Ashburn's live PJM queue wait reads about 39 months on DC Hub right now, and that's the cost congested metros carry. It's a big part of why sites with gas access and room for onsite generation look better on speed to power. Source: DC Hub (dchub.cloud), as of Oct 13, 2026.