A campus that had chips and no power
2024 – 2026- Setup
- Silicon allocation secured, land secured, permits in motion, and an interconnect queue position that put firm grid power roughly three years out. Every part of the project was ready except the only input that could not be bought forward.
- What happened
- The developer bridged the gap with behind-the-meter gas generation at a substantial premium to eventual grid pricing, plus the capex of the plant itself. That premium — not the hardware, not the site, not the tenant — determined the project's return, and it was the line the original model treated as a rounding error.
- Where it stands
- The comparable project a hundred miles away, without on-site generation, is still in the queue. Its anchor tenant has renegotiated once already and holds an exit if energisation slips again. The developer with the bridge is earning less than it hoped and considerably more than nothing.
- Interconnect queue at underwriting
- ~3 years
- Return driver
- bridge power premium
- Comparable project without bridge
- 1 renegotiation
The scarce asset is a site with firm power on a timeline that matters, not the chips. Underwrite the bridge cost explicitly, because that is where the return actually lives.
