
AI's Grid Strain Presents Opportunity for Bakken Gas, Infrastructure
Surging, inflexible electricity demand from data centers could accelerate natural gas power generation and grid projects, benefiting North Dakota.
The rapid growth of artificial intelligence is creating a massive, concentrated demand for electricity that could significantly impact energy markets and infrastructure development in North Dakota, according to a report from OilPrice.com.
While AI data centers are expected to account for only about 3% of global electricity demand by 2030, their impact is highly localized. The International Energy Agency (IEA) estimates that data-center electricity consumption will roughly double between 2025 and 2030, reaching approximately 950 terawatt-hours. Nearly half of existing U.S. data-center capacity is concentrated in just five regional clusters, and half of new developments are planned for these same areas.
This concentration creates a critical "inflexibility problem," OilPrice.com reported. Technology companies can build massive computing infrastructure in two to three years, but the transmission lines needed to power them can take four to eight years to complete. The IEA estimates that about 20% of planned data-center projects could face delays due to these electricity-sector bottlenecks.
For North Dakota's Bakken region, this strain represents a dual opportunity. First, it reinforces the role of natural gas as a dispatchable power source. The IEA expects renewables to meet around half of the additional data-center electricity requirement through 2035, but natural gas and nuclear power are also projected to make significant contributions. This sustained demand for reliable generation supports the market for Bakken natural gas, which is often flared or needs additional pipeline takeaway capacity.
Second, the urgent need for grid expansion and new power generation to serve these large, sudden loads could accelerate infrastructure projects. The report notes that waiting times for key components like transformers, cables, and gas turbines have increased, highlighting a need for investment and manufacturing.
The development timeline mismatch—where data centers are built much faster than the grid can adapt—means regions with existing energy infrastructure and generation assets may see increased investment. For Bakken operators and royalty owners, the growing, inflexible power demand from AI underscores the long-term value of the region's natural gas resources and the importance of connecting them to broader energy markets.
Source
OilPrice.com report "AI’s Electricity Demand Is Not the Real Problem. Its Inflexibility Is" published August 1, 2026.


