the WattHomeNewsEspañol
Análisis

Data Center Forecasts Are Overstating Gas and Transmission Demand in the US

IEEFA finds data center forecasts inflate gas demand by 5x. Latitude Media reports a transmission slowdown for the same reason. A direct warning for Mexico.

Por REDACCIÓN THE WATT · 26 jun 2026 · 2 MIN READ
Data centers and transmission lines in an industrial landscape with energy demand forecast charts
Imagen generada con inteligencia artificial

The Institute for Energy Economics and Financial Analysis (IEEFA) documented on June 25 that natural gas consumption forecasts for data centers, used to justify two new pipelines totaling $5.2 billion in the southeastern United States, exceed grid operators' own base scenarios by a factor of five.

The same day, Latitude Media reported that the Mid-Atlantic Resiliency Link (MARL) transmission line, a $960 million, 160-kilometer NextEra project connecting Pennsylvania, West Virginia, Maryland, and Virginia, faces a regulatory setback. Consumer advocates are challenging whether the concentrated demand from data centers in northern Virginia justifies spreading costs across all users of the PJM regional grid, the mid-Atlantic transmission operator. Maryland filed a formal complaint with the Federal Energy Regulatory Commission (FERC) in May 2026, arguing that costs should be assigned directly to the zone where large consumers are located.

The IEEFA report puts numbers to the gap. Kinder Morgan projects, in its high-growth scenario, a natural gas deficit for electricity generation of 2.45 billion cubic feet per day (bcf/d) by 2035, to justify the Mississippi Crossing (MSX, $1.7 billion) and South System Expansion 4 (SSE4, $3.5 billion) pipelines. The base scenario shows a deficit of just 0.46 bcf/d, less than 15 percent of the proposed new capacity. Regional natural gas demand also declined by an average of 0.78 percent annually between 2017 and 2025.

On the transmission side, the MARL episode exposes a problem that extends well beyond this single project. The United States went from building 6,400 kilometers of high-voltage lines in 2013 to an average of just a few hundred kilometers per year in recent years. The slowdown is not only about permitting: it is an interstate dispute over who pays when new load concentrates in a single zone.

For Mexico, the pattern is a direct warning. As nearshoring accelerates data center installations along the northern industrial corridor, the Centro Nacional de Control de Energía (CENACE) and the Secretaría de Energía (SENER) will face the same forecasting dilemma: modeling demand with their own verifiable assumptions, not with developers' projections. The update to the Programa de Desarrollo del Sistema Eléctrico Nacional (PRODESEN) will be the first real test of that rigor.

This article was written with artificial intelligence assistance based on verified sources and reviewed by a human editor before publication.

This article was drafted with AI assistance from verified sources and reviewed by a human editor before publication.

Frequently asked questions

¿Por qué los pronósticos de demanda de centros de datos están sobredimensionados en Estados Unidos?
Según IEEFA, las proyecciones de Kinder Morgan exceden hasta por un factor de cinco los escenarios base: proyectan un déficit de gas de 2.45 bcf/d en 2035, pero el escenario base arroja solo 0.46 bcf/d. La demanda regional de gas natural cayó 0.78 por ciento anual entre 2017 y 2025.
¿Cómo afectan los centros de datos a la expansión de la red de transmisión en Estados Unidos?
La demanda concentrada de centros de datos en el norte de Virginia genera disputas interestatales sobre quién paga las nuevas líneas. En mayo de 2026, Maryland presentó una queja ante FERC para que los costos de la línea MARL se asignen directamente a la zona de los grandes consumidores, no a todos los usuarios de la red PJM.
¿Qué implica este fenómeno para la planeación energética de México?
Conforme el nearshoring acelere la instalación de centros de datos en el norte de México, CENACE y SENER enfrentarán el mismo dilema: modelar la demanda con supuestos propios y verificables, no con las proyecciones de los desarrolladores. La actualización del PRODESEN será una prueba clave de ese rigor.
Share:✓
← All news