How Xcel's Proposal Can Cut Data Center Costs

Utilities don't usually make headlines for doing the right thing. But when a major power company steps up to keep data center costs from landing on ordinary ratepayers, it's worth paying attention — especially as the energy demands of the tech industry reshape grid infrastructure at a pace that regulators are still struggling to absorb.

Xcel Energy's recent proposal does exactly that. While the details are still working their way through regulatory channels, the core idea is straightforward: data centers should pay their own way. Danny Katz, a consumer advocate, put it plainly: "Xcel's proposal looks like a good first step to ensure these data center costs don't fall on consumers."

That framing matters more than it might seem.

The Problem With How We've Been Paying for Data Center Growth

Data centers are not subtle infrastructure. A single hyperscale facility can draw anywhere from 20 to 100+ megawatts of continuous load — the equivalent of tens of thousands of homes running 24/7, with almost zero flexibility in when or how they consume power. When a utility has to upgrade substations, build new transmission lines, or add generation capacity to serve that kind of load, those capital costs traditionally get socialized across the entire ratepayer base.

That model made sense when large industrial customers were relatively rare and their grid needs were modest. It doesn't hold up when a single tech campus requires as much infrastructure buildout as a small city.

The data center boom has changed the math entirely. Between AI workloads, cloud migration, and cryptocurrency (in some regions), electricity demand projections that utilities filed just five years ago are already obsolete. Xcel and other major utilities are scrambling to build ahead of demand — and someone has to pay for that scramble.

Without a specific cost-allocation policy, the default answer has been: everyone pays. Your monthly bill absorbs a fraction of the substation upgrade that was built primarily to serve a hyperscale tenant who negotiated a favorable commercial rate. That's the status quo Xcel's proposal aims to disrupt.

What the Xcel Data Center Proposal Actually Does

The specifics of Xcel's proposal center on cost responsibility — making data center developers bear the infrastructure costs that their projects directly generate, rather than distributing those costs across residential and small commercial ratepayers.

This is a structural shift in how utilities think about large load interconnection. Traditionally, utilities have used "socialized" cost models where grid upgrades are treated as shared public infrastructure. The alternative — often called a "beneficiary pays" or direct assignment model — requires the entity creating the need to fund the solution.

The beneficiary-pays approach isn't new in energy circles, but applying it specifically and aggressively to data center development represents a meaningful recalibration of how utilities handle the current wave of tech-driven load growth.

For regulators, this is a genuinely difficult balancing act. Data centers bring jobs, tax revenue, and economic development arguments that make them politically attractive. But grid upgrades are expensive — often running into tens or hundreds of millions of dollars per major project — and those costs have a nasty habit of outlasting the economic benefits if the original tenant downsizes, relocates, or goes dark.

Xcel's proposal, even as a "first step" in Katz's characterization, signals that the utility is trying to get ahead of that risk before it becomes a ratepayer liability.

What This Means for Infrastructure Developers

If you're developing data center projects in Xcel's service territory — which spans parts of Colorado, Minnesota, and several other states — this proposal changes your pro forma.

The blunt reality: infrastructure interconnection costs that might have previously been partially absorbed by the utility (and thus spread across all customers) could now land directly on your project budget. Depending on the scale of the facility and the condition of the local grid, that could mean millions of additional dollars in upfront capital requirements.

That's not necessarily a dealbreaker. Sophisticated data center developers already model interconnection risk as a variable — particularly after watching projects in PJM and MISO territories sit in interconnection queues for three to five years before getting cost estimates. What changes is the certainty of who pays, and earlier certainty, even if the number is higher, is often preferable to late-stage surprises.

There's also a competitive siting angle here. If Xcel's territory becomes known for clear, developer-pays cost allocation rules, it could attract better-capitalized developers who prefer regulatory predictability over the ambiguity that comes with utilities that haven't yet addressed this question. Serious players want clear rules. What kills projects is uncertainty.

Smaller or more speculative developers, on the other hand, may find the numbers harder to pencil. Projects that were marginally feasible under a socialized cost assumption may not survive direct cost assignment. That's a feature, not a bug, from a grid planning perspective — it filters out undercapitalized projects before they consume utility planning resources.

The Consumer Equation

For residential customers and small businesses, the logic is almost elegantly simple: if you didn't create the need for the infrastructure, you shouldn't fund it. Xcel's proposal, if adopted, puts that principle into practice.

The near-term benefit is protection from cost pass-through. The longer-term benefit is a cleaner grid planning process. When infrastructure investment is tied directly to the customers generating the demand, utilities have better information about which projects are real, which developers are committed, and which load forecasts to trust.

Ratepayer protection and grid planning discipline are two sides of the same coin here — and Xcel's proposal advances both simultaneously.

There's a broader signal worth noting for anyone watching the energy sector. Utilities across the country are facing versions of this same question. PJM recently overhauled its interconnection queue process partly in response to the explosion of new load applications. FERC has been pushing transmission cost allocation reform for years. State regulators in Virginia — home to the largest concentration of data center capacity on the planet — are wrestling with identical dynamics.

Xcel moving proactively to address data center cost allocation puts it ahead of utilities that are still waiting to see how the regulatory dust settles elsewhere. That proactive posture tends to produce better outcomes for both utilities and their customers, because it avoids the reactive scramble that comes from trying to fix cost allocation after major infrastructure commitments have already been made.

Where This Goes Next

Proposals like this don't become policy overnight. Xcel's plan will face scrutiny from data center industry lobbyists, economic development advocates, and potentially from developers who've already begun site selection in Xcel territory under different cost assumptions.

The outcome isn't guaranteed. But the direction is right, and the momentum behind it — consumer advocates like Katz lending public support, growing awareness of the ratepayer equity issue, and federal-level pressure on utilities to modernize their interconnection practices — suggests this kind of framework is becoming harder to resist.

Watch for how other utilities respond. If Xcel's proposal clears the regulatory process with its core cost-assignment structure intact, it becomes a template. Utilities in other high-growth data center markets will face immediate pressure from regulators and consumer groups asking why they haven't adopted similar protections.

The data center industry isn't going to slow down. Demand for AI compute, cloud infrastructure, and digital storage is compounding faster than most utility integrated resource plans anticipated. That means grid infrastructure investment is going to continue accelerating — and the question of who funds that investment is only going to get more consequential.

Xcel's answer: the people who benefit from it should pay for it. As policy principles go, that one's hard to argue with.

Call to Action: Discover more about how Xcel's proposal can impact your data center projects and the marketplace at InfraSale Marketplace.

[INTERNAL LINK: data center costs]

[INTERNAL LINK: energy sector trends]

[INTERNAL LINK: infrastructure investment]