From Bottlenecks to Breakthroughs: Rewriting the Grid Planning Playbook in the Southwest Power Pool and Texas

In brief

Arushi argues that many regional grids still plan AI data campuses as if they were inflexible industrial loads, which makes planning itself the bottleneck. She examines the Southwest Power Pool’s proposed High-Impact Large Load Generation Interconnection Assessment (HILLGA) and Conditional High Impact Large Load Service (CHILLS), which study co-located generation alongside new load and let loads connect quickly on interruptible terms. ERCOT, she finds, has strong modeling tools but no defined way to credit co-located resources in large-load studies. She closes by treating planning as a competitive variable.

Key points

1. 01Data campuses with their own generation and storage shouldn’t be modeled like inflexible steel mills.
2. 02SPP’s HILLGA and CHILLS proposals study co-located generation with the load and offer fast, interruptible service.
3. 03ERCOT has the tools but no non-firm transmission product that credits co-located resources.