- One reason the AI infrastructure story is hard to dismiss now is that it is starting to show up in public electricity data, not just company spending plans and developer presentations.
- That matters because Virginia is not just another regional market.
- The second layer is what happens at peak.
- Section
- Markets
- Read time
- 6 min read
- Data included
- Virginia load snapshot
Virginia load snapshot
The public-data signal is no longer subtle. Virginia is showing both measurable electricity-sales growth and a forward-looking peak-demand warning in the same story.
Virginia commercial electricity sales
| Signal | Earlier baseline | Current / next wave | Why it matters |
|---|---|---|---|
| Commercial electricity sales | ~84 million MWh (2019) | ~114 million MWh (2024) | State-level load data is now visibly bending with data center concentration. |
| Dominion summer peak outlook | Already one of PJM’s major load zones | Largest absolute peak-growth increases in PJM’s 2026 long-term view | AI-era demand is becoming a planning and reliability problem, not just a growth narrative. |
Source context: U.S. Energy Information Administration analysis published May 5, 2026, and PJM 2026 long-term load forecast references cited in the article.
One reason the AI infrastructure story is hard to dismiss now is that it is starting to show up in public electricity data, not just company spending plans and developer presentations. In a May 5, 2026 analysis, the US Energy Information Administration said Virginia commercial electricity sales rose from about 84 million megawatt-hours in 2019 to nearly 114 million megawatt-hours in 2024, with data center growth driving much of the increase.
That matters because Virginia is not just another regional market. Northern Virginia is the largest data center cluster in the world, so shifts there are a useful signal for how AI and cloud infrastructure begin to register inside real utility and grid numbers. Once the impact appears in state-level electricity sales, the conversation moves beyond hype. It becomes a measurable load story.
Once data center growth starts showing up clearly in public electricity-sales data, the AI power story has moved past theory and into system reality.
The second layer is what happens at peak. PJM’s 2026 long-term load forecast points to Dominion zones carrying the largest absolute summer peak increases in the coming years, with data centers acting as a major driver. That does not mean every megawatt of growth is AI-specific, but it does reinforce the bigger point: the power system is being asked to absorb a class of commercial demand that is unusually large, concentrated, and urgent.
This is why the Virginia story matters for more than one state. Investors, utilities, developers, and policymakers are all trying to figure out when AI demand becomes visible enough to change planning assumptions. In Virginia, that threshold has already been crossed. The electricity-sales data gives the growth story a harder edge, and the PJM forecast gives that story a forward operating consequence.
The practical implication is that the AI buildout is no longer just competing on chips and capital. It is competing on whether the surrounding power system can keep up. Virginia now looks less like an isolated exception and more like an early warning of what happens when data center concentration becomes large enough to bend real energy statistics.
By Nawaz Lalani
The Grid Report is written by Nawaz Lalani and focuses on source-backed coverage of AI infrastructure, grid power demand, automation systems, and market signals.
Follow the signal, not just the headline.
Get the daily Grid brief for source-backed coverage on AI power demand, infrastructure timing, automation, and market signals.