Today’s Industrial AI Daily Signal · Smart Infrastructure · Data-center power
Data-center speed is locking in costlier on-site gas.
U.S. data-center developers are accelerating behind-the-meter gas generation to bypass grid delays, creating a 29.6-gigawatt pipeline built around faster but costlier equipment.
Enverus Intelligence Research projects 29.6 GW of behind-the-meter gas generation will be added in the United States through 2030, with data centers accounting for roughly 88% of the demand. Developers are favoring modular aeroderivative turbines, industrial turbines and reciprocating engines because large combined-cycle plants can take as long as six years to deliver.
Crusoe has secured 29 GE Vernova 35 MW aeroderivative units, including ten for its Abilene campus, plus about 750 MW of Bergen reciprocating engines for multiple U.S. sites. The speed premium is material: EPRI says capital cost per unit of electricity can be up to 50% higher for smaller turbines, which also emit more per MWh than modern combined-cycle plants.
AI-assisted content may contain errors. Verify important facts and decisions against the linked original sources.
01
What changed
Individual hyperscalers had already ordered on-site generation. What changed on September 29 is a quantified market-wide operating picture: 29.6 GW through 2030, data centers responsible for 88% of demand, and disclosed cost and delivery tradeoffs across turbine classes.
02
Why it matters
AI infrastructure is being shaped as much by power-equipment lead times as by chips. Behind-the-meter generation can energize campuses years before a grid interconnection, but it also transfers fuel, maintenance, emissions and stranded-asset risk onto the operator and host community. The emerging fleet is large enough to influence turbine factories, gas infrastructure and utility planning. For executives, the key comparison is no longer simply cost per installed megawatt; it is time to power, lifetime cost per MWh, emissions, availability and whether temporary campus generation can later support the grid or becomes an expensive island.
03
What to watch
Watch contracted versus commissioned megawatts, turbine delivery schedules, air permits, campus load factors, fuel arrangements and whether batteries, demand response or later grid connections reduce run hours. Developers should publish both time-to-power and lifetime cost per MWh.
Reuters Events combines Enverus market estimates with named orders and interviews from Crusoe, Siemens Energy, EPRI, Global Energy Monitor and RMI. GE Vernova independently documents the Crusoe turbine fleet. Confidence is high in the disclosed orders and engineering tradeoffs, while the 2030 market total remains a forecast.
Impact score
92/100
Reader reaction
How does this development make you feel?
Sources used for this edition
Behind today’s selection
Today’s two runners-up
Runner-up 1 · Robotics · Manufacturing capacity
Amazon doubles robot-making capacity with a fourth plant.
Amazon plans a $100 million-plus Indiana manufacturing hub that will bring its robot-making network to four plants and support North American fulfillment automation.
Why it was not selected: The factory adds concrete robot-making capacity and 300 skilled jobs, but it is scheduled for 2028 and has no disclosed unit-output target, making the live campus-power constraint more immediate.
View scoring details
- Industrial relevance
- 25/25
- Operational or economic impact
- 18/20
- Technology significance
- 13/15
- Evidence of real-world adoption
- 11/15
- Strategic significance
- 9/10
- Novelty
- 7/10
- Source confidence
- 4/5
- Source reliability
- 29/30
- Independent corroboration
- 23/25
- Primary or official evidence
- 22/25
- Evidence consistency
- 18/20
Runner-up 2 · Industrial Software · Fleet operations
Fleet AI moves from demos into dispatch and maintenance.
Trimble released customer-ready AI for route planning, trailer orchestration, repair invoices and transport-management workflows across physical fleet operations.
Why it was not selected: The products are available across live fleet software versions and one workflow has measured savings, but customer-level operating outcomes and independent benchmarks remain limited.
View scoring details
- Industrial relevance
- 25/25
- Operational or economic impact
- 17/20
- Technology significance
- 14/15
- Evidence of real-world adoption
- 12/15
- Strategic significance
- 8/10
- Novelty
- 6/10
- Source confidence
- 4/5
- Source reliability
- 28/30
- Independent corroboration
- 10/25
- Primary or official evidence
- 25/25
- Evidence consistency
- 20/20