$1.1 trillion in grid investment and AI data centers still face decade-long connection waits

In Northern Virginia's data center corridor, a new facility can wait up to 14 years just to get a grid connection, according to Forbes. That friction is playing out against one of the largest capital commitments in U.S. utility history: $208 billion in grid investment planned for 2025 alone, with more than $1.1 trillion projected over five years, per Edison Electric Institute figures cited by Hitachi Energy. The numbers illustrate a structural tension that procurement and operations leaders at enterprise technology companies now have to plan around.

Electricity demand across advanced economies had been relatively flat for years. The rapid build-out of AI infrastructure changed that calculus fast. Utilities are revising load forecasts upward, and the queue of large data center projects competing for interconnection has grown to the point that some jurisdictions are rejecting them outright. British Columbia, for example, has moved to prioritize other industrial loads over data centers and cryptocurrency mining operations, according to Forbes.

A coordination problem, not just a capacity problem

Ali Ghorashi, Senior Vice President at DNV, told Forbes that the current wave of data center demand is different from prior energy investment cycles because it sits at the intersection of industries that have historically operated in separate lanes: real estate, technology, electricity markets, utilities, and infrastructure finance. Most participants, he noted, come from one of those worlds and are unfamiliar with how the others operate.

That knowledge gap shows up directly in interconnection negotiations. Large technology companies arrive with capital and firm demand signals, but utilities work on infrastructure timelines measured in years or decades. Regulators are beginning to respond by raising application costs and exploring penalties for projects that reserve grid capacity without advancing, according to Forbes. The practical result is that the bottleneck is as much organizational and regulatory as it is physical.

Ghorashi's framing, as reported by Forbes, is that solving the grid access problem requires different players to understand each other's constraints. Transmission upgrades and new permitting pathways matter, but so does better coordination between technology buyers, utilities, and project finance teams from the earliest stages of site selection.

Efficiency at every conversion step

The other side of the equation is how efficiently electricity is used once it reaches a facility. Franziskus Gehle, Vice President at Delta Electronics, argued in Forbes that efficiency analysis has to start at the system level, not the chip level. Electricity passes through multiple conversion steps before it performs any computation, and each step carries losses. Tightening those conversion inefficiencies reduces the total load a facility places on the grid, which matters both operationally and in interconnection negotiations.

Hitachi Energy has made similar arguments in the context of broader grid modernization, noting that electrification is accelerating across sectors simultaneously, not just in data centers. That compound demand growth means utilities and their enterprise customers alike have to pursue efficiency alongside new supply, not treat them as alternatives.

Forbes also reported that modern power systems have more levers than simple generation expansion: short- and long-term storage, demand flexibility programs, and better utilization of existing transmission infrastructure all factor into how quickly the grid can absorb new large loads. For data center operators, that means the technical and contractual design of a facility, including whether it can participate in demand response or shift non-critical workloads, affects how quickly it can get connected and at what cost.

What this means for your team

• Audit interconnection timelines early: site selection teams should engage with utilities on grid capacity and queue position before committing to a location, given connection waits that can now reach 14 years in constrained markets.
• Build demand flexibility into facility design: data centers that can participate in demand response or shift non-critical workloads have a stronger negotiating position with utilities and may access capacity faster.
• Evaluate power conversion efficiency at the system level, not just at the IT load: gains across UPS, cooling, and distribution infrastructure reduce total megawatt demand and lower the cost basis of interconnection requests.
• Engage infrastructure finance and utility teams together from the start: the coordination gap between technology buyers and utility planners is a documented bottleneck; closing it early reduces project risk.