Speed to power.
Barker Infrastructure helps datacenter developers and investors evaluate sites in ERCOT. We model time-to-power, power cost at the site's node, and the grid buildout nearby. We are also building an AI pipeline that tracks the interconnection queue, market rules, and legislation.
We help datacenters decide where to build. Revenue starts at energization. Time-to-power, nodal power cost, and surrounding buildout set a site's economics, and we model each at the settlement-point level.
Power is the constraint on the datacenter buildout.
Power availability
New datacenter capacity is limited by electricity availability more than land or fiber. Announced capacity far exceeds what gets energized, and the interconnection queue shows where and when power will actually exist.
Timelines vary by site
Revenue starts at energization. Time-to-power differs between candidate sites, and those differences are visible in current filings.
Cost varies by node
Two sites forty miles apart can face different congestion, basis, and buildout exposure. Power cost is set at each site's node and shifts as nearby generation and load are added.
Siting analysis, and the models behind it.
For datacenter developers and investors making location decisions in ERCOT.
Site screening
Candidate screening across ERCOT: interconnection activity around each site, nodal price history, congestion exposure, and the buildout planned nearby.
Time-to-power assessment
How fast a specific site can energize: queue position, required upgrades, equipment lead times, and whether non-firm or flexible interconnection service can pull the date forward. Timelines drawn from current filings.
Power cost & offtake
Forecast power cost at the site's node, and the structures that firm it: hedges and revenue collars priced against simulated hourly scenarios, including the basis risk that remains after hub settlement.
Project valuation
Revenue and cashflow models built from hourly price simulation, including valuation of interconnection positions, which now trade as standalone assets. Our models have supported investment evaluations of projects in the ERCOT market.
Capacity expansion
Scenarios for generation, storage, and load entering the system around a site, and their effect on congestion, basis, and power cost over the asset's life.
AI ingestion pipelineIn development
Machine-learning ingestion of the interconnection queue, market rules, and legislation, feeding the capacity expansion models used in screening and cost forecasts.
The machine-learning program.
We are combining traditional capacity expansion modeling with machine-learning ingestion of the documents that drive its assumptions: developer planning materials, interconnection filings, ERCOT rule changes, and legislation. The goal is siting and power-cost estimates that stay current as the grid is built out.
Inside the stackSiting a datacenter in ERCOT?
We're a small shop and we answer our own email.