Residential solar aggregation enters data-center power stack while gas cements role as grid baseloadâDCPI build markets correlate with gas-rich ISOs
## Natural gas is capturing 63% of new utility capacity additions in 2026 as AI data-center load reshapes the US grid fuel mixâa structural shift now visible in DCPI market rankings.
While wind and solar dominate headline announcements, gas-fired generation is winning the reliability race. Utilities across ERCOT, SPP, and PJM are contracting gas peakers and combined-cycle plants to backstop intermittent renewables and meet 24/7 AI workload demand. The correlation is measurable: the top 10 DCPI excess-power markets sit inside ISOs with active gas pipeline infrastructure and sub-18-month interconnection timelines for fossil capacity.
**MidlandâOdessa** (ERCOT, 85.7 excess) and **Oklahoma City** (SPP, 76.5 excess) lead the build rankings in part because both connect to Permian Basin gas and can activate peaker capacity on 12-month schedulesâfaster than any renewable-plus-storage greenfield.
Meanwhile, a quieter shift: residential solar aggregation is entering the data-center power stack. Sunrun's deal with Voltus to bundle distributed solar-plus-storage for AI facilities marks the first utility-scale arbitrage of behind-the-meter residential capacity. The economics are marginal today but the precedent mattersâif 2% of US rooftop solar (roughly 900 MW) can be aggregated and dispatched, it becomes a new procurement channel for hyperscalers chasing Scope 2 offsets without building their own arrays.
## What it means
Site selection: gas-rich markets with excess DCPI headroom can guarantee power delivery timelines renewable-only corridors cannot. Constraint scores below 30 signal both gas availability and grid slack.
Capex planning: expect utilities to layer residential DER aggregation into RFPs by Q1 2027, creating a pricing floor for behind-the-meter capacity that competes with PPAs.
## Methodology
Source: DC Hub Data Center Power Index (https://dchub.cloud/dcpi). Updated daily. Excess and constraint metrics derived from ISO reserve margins, interconnection queue depth, transmission headroom, and fuel-mix data across 280+ markets and 7 US ISOs.
Press inquiries: [email protected] · DC Hub MCP API: https://dchub.cloud/mcp