TL;DR

  • The grid is a structural bottleneck: US interconnection queues hold over 2,600 GW of projects with median wait times of 4.5 years, stretching to 8 to 12 years for large-scale data centers.
  • Lessons from mobile electrification: Data center operators face the exact power delays that EV fleet managers experienced a decade ago, where relying on promised utility timelines stalled operations.
  • Immediate bridge power deployment: On-site solutions like Battery Energy Storage Systems (BESS) and microgrids can be commissioned in weeks to bring compute capacity online without waiting for formal grid connections.

# # #

By Josh Aviv, CEO and Founder of SparkCharge

The data center industry is facing a power problem that looks familiar to me.

For the past decade, fleet operators, autonomous vehicle companies, and logistics hubs have sought solutions at the intersection of commercial energy demand and grid infrastructure restraints. The grid has been unable to keep up with the pace at which businesses need power. These industries cannot wait years for utility connections.

Data center developers are now standing at the same crossroads. The timelines are longer, the loads are larger, and the financial stakes are higher. But the fundamental problem is identical: enterprises need power now, and the grid delivers it in years.

The lessons the mobile energy industry learned the hard way over the past decade are directly applicable to what data center operators are navigating today.

The Grid Is a Structural Problem, Not a Temporary One

The most important thing I can tell any data center developer is this: do not plan your business around the grid catching up. It will not catch up on your timeline.

The US interconnection queue holds more than 2,600 GW of projects waiting for grid connection, a figure representing more than twice the total installed generating capacity of the country, according to Lawrence Berkeley National Laboratory’s 2025 Queued Up report. The median wait time for a new grid interconnection is 4.5 years. For large-scale data center projects in constrained markets, timelines of eight to twelve years are not unusual.

This is not a temporary backlog caused by a spike in applications. It is a structural mismatch between the pace of commercial energy demand and the pace at which utility infrastructure can be built, permitted, studied, and connected. AI workloads, fleet electrification, and industrial reshoring are all drawing from the same constrained grid simultaneously. The queue is not clearing. It is growing.

The EV fleet operators who struggled most in the early years of electrification were the ones who believed the utility’s promised timeline. They signed leases, ordered vehicles, and told their operations teams to expect a charging infrastructure that arrived 18 months late. The data center operators who will struggle most in the next five years are the ones making the same assumptions.

Facilities Managers Are Being Asked to Solve an Infrastructure Problem

One of the most underappreciated dynamics in the current data center power crisis is where the pressure lands inside an organization. Prefabricated and modular data center systems have compressed construction timelines significantly. Containerized edge units can be operational weeks after on-site delivery. Room-scale prefabricated modules typically run 12 to 24 weeks from contract to installation. The build problem is largely solved.

That means the person handed a completed data center building is now also handed a power gap that no one in the procurement chain has an answer for. Facilities managers and infrastructure directors are being asked to make operational something that has no grid power and no confirmed delivery date for grid power. That is an unsustainable operating model, and it is becoming the norm rather than the exception.

In the electric vehicle (EV) and autonomous vehicle fleet world, we saw the equivalent dynamic play out at the depot level. Fleet managers were handed EVs and asked to charge them at facilities that had not been upgraded. The ones who found solutions did not wait for the utility. They found energy infrastructure that could be deployed immediately and scaled as the fleet grew.

The data center industry needs to make the same mental shift. The grid is not a planning assumption. It is a variable with a very wide confidence interval.

Mobile Energy Infrastructure Is Not a Workaround. It Is a Strategy.

At fleet depots, mobile battery energy storage has often been framed as temporary bridge power to cover the period before permanent infrastructure becomes available. In practice, these systems can serve more than one role: they can support operations during infrastructure delays and, if grid power arrives, remain part of a broader energy expansion.

Battery Energy Storage Systems deployed for bridge power do not become obsolete when a grid connection arrives. They become more valuable. A BESS system that powers a data center during the interconnection gap transitions into a permanent grid-tied asset if the utility connection goes live. It shaves peak demand, eliminates demand charge exposure, and in markets with active wholesale energy programs, participates in demand response and capacity markets, generating revenue from capacity that would otherwise sit idle.

The financial framing shifts entirely. Bridge power is not a capital expenditure liability. It is the first phase of a long-term energy infrastructure investment that continues to generate value well beyond the interconnection date.

Large-scale BESS deployments can be commissioned in weeks, delivering immediate behind-the-meter power that allows compute infrastructure to come online without waiting for formal grid interconnection. At a moment when AI compute capacity is among the most commercially valuable infrastructure on earth, every month of delayed operation represents revenue that does not materialize. Bridge power converts that waiting period into a generating period.

Mobile distributed generation systems add continuous runtime to the equation. Paired with BESS, they create a hybrid energy architecture where generation runs at its optimal load point, BESS handles peak demand and power conditioning, and intelligent energy management software coordinates both autonomously. The result is unlimited runtime, lower fuel cost per kWh delivered, and a system that scales as rack density increases.

Build Ahead of the Grid, Not Around It

The organizations getting this right are not treating the power problem as someone else’s responsibility to solve. They are treating it as a first-order infrastructure decision that gets made before the site selection, before the lease is signed, and before the equipment is ordered.

The grid is not the enemy. I want to be clear about that. On-site energy systems, including battery storage and dispatchable generation, can complement rather than replace the grid. They can provide power while utility interconnections and upgrades are pending and, where the technical and economic conditions support it, continue to help manage demand after grid service becomes available.

But the grid is also not a planning assumption any serious data center developer can rely on in 2026. The interconnection queue data, the transformer lead times running two to four years according to Wood Mackenzie’s Q2 2025 survey, and the regulatory shifts requiring data centers to fund their own grid upgrade costs all point in the same direction. The enterprises that deploy on-site energy infrastructure now will be generating revenue while their competitors are still in the interconnection queue.

That is not speculation. It is what happened in the EV fleet industry. The operators who moved first on microgrid energy infrastructure are the ones running the most competitive, most reliable, most cost-efficient fleets today. The ones who waited are still waiting.

Data center developers do not have to make the same mistake. The infrastructure exists. It deploys in days. It works before the grid arrives, and it keeps working after it does.

Do not wait for the utility to solve a problem the utility has not been able to solve for fifty years. Build ahead of the grid.

# # #

About the Author

Joshua Aviv is the Founder and CEO of SparkCharge, an intelligent microgrid provider. SparkCharge deploys microgrids that deliver grid-scale power in days, not years, for enterprises that can’t wait for the grid. Each turnkey microgrid combines SparkAI and SparkEMS with Battery Energy Storage and PowerHub generation. SparkCharge powers data centers, autonomous vehicle operations, ports and logistics hubs, construction and industrial sites, live events, and EV fleets across North America and the United Kingdom.