Minnesota's Capacity*Connect Gamble: A Utility-Owned VPP Without Competition

On April 2, 2026, the Minnesota Public Utilities Commission approved a utility-owned battery program unlike any other in the United States. One utility. One deployment partner. No competitive solicitation. Up to $430 million in ratepayer-backed capital to build a 200-megawatt distributed battery fleet across Xcel Energy's Minnesota service territory by 2028. No other state had said yes to that shape of a deal.

The program is called Capacity*Connect, and the headline of every trade-press account has been the same word: controversial. The controversy is not really about virtual power plants, which are broadly popular on both sides of the aisle. It is about who gets to build them, who gets to own them, and who absorbs the risk if the economics turn out worse than the filing predicted. Minnesota just answered all three of those questions in a way that breaks with how the rest of the country has structured the same category of investment — and the decision is worth unpacking on its own terms, not as a referendum on storage.

What Capacity*Connect Actually Is

Strip away the branding and Capacity*Connect is a utility capex line item wearing VPP clothing. Under the Phase 2 approval issued by the MPUC in docket 25-378, Xcel may deploy up to 200 MW of battery energy storage across its distribution grid in increments of 1 to 3 MW each, positioned near communities rather than at centralized substations. The utility owns the hardware. The utility controls the dispatch. The utility books the rate base.

That last sentence is the one that matters. In most U.S. virtual power plant programs, a third-party aggregator signs up customers, installs or enrolls behind-the-meter batteries, and is paid by the utility for the dispatchable capacity it can deliver. The customer earns an incentive, the aggregator earns a margin, and the utility avoids lumpy capital spending. Xcel's model inverts that arrangement. The distributed batteries are in-front-of-the-meter utility assets, deployed through a single sole-source contract with the infrastructure services firm Sparkfund, and financed the same way a gas peaker or transmission upgrade would be. Customers pay through their bills.

The Minnesota PUC's own bulletin, published on April 3, 2026, characterized the program as a "first-of-its-kind" utility-owned VPP. Commissioner Hwikwon Ham described it as "a vital step toward modernizing the energy grid" and a path toward "a more equitable energy future." The Energy Mix's reporting makes the structural novelty explicit: this is the first utility-owned-and-operated VPP in the country, whereas most other programs involve third-party companies managing customer-sited resources.

The Google Wrinkle

Any serious reading of Capacity*Connect has to factor in a second Minnesota story running in parallel. Xcel is building a Google data center in Pine Island, and the data-center deal is entangled with the battery program in ways that are easy to miss if you only read the PUC filing.

According to The Energy Mix, Google is putting $50 million into Capacity*Connect as part of the broader commercial arrangement covering the data center's power needs. Separately, Energy-Storage.News reported in April 2026 that Xcel's Pine Island plan includes a 30 GWh iron-air battery deployment from Form Energy, announced in March 2026, and that Xcel has filed for a 600 MW BESS at its Sherco Energy Hub. All three projects sit within the same Upper Midwest Energy Plan approved by the MPUC in February 2025, which sets an Xcel storage target of 600 MW by 2030.

In other words, Capacity*Connect is one slice of a much larger storage portfolio. That matters when evaluating the $430 million figure, because the program is not a standalone gamble on VPPs — it is a distribution-grid complement to centralized battery builds at Sherco and Pine Island, and at least a portion of the funding is backstopped by a hyperscaler customer. The portfolio context does not answer the competitive-procurement objections, but it does change how a reader should read the dollar figure.

Why the Industry Objected

The joint April 2 statement from the Solar Energy Industries Association, its Minnesota branch, and the Coalition for Community Solar Access was unusually direct. Sarah Webbe of MnSEIA told Energy-Storage.News that "giving control to just one partner leaves out Minnesota's experienced solar and storage developers." Andrew Linhares of SEIA, speaking to the same outlet and also quoted by The Energy Mix, argued that "competitive markets for energy storage deployment ensure that ratepayers get the best, most affordable deal possible." Nick Bowman of CCSA went further, saying that "by severely limiting third-party developer opportunities…Capacity*Connect misses the mark."

Kevin Cray, a vice president at the Coalition for Community Solar Access, distilled the structural objection in a Latitude Media interview: the sole-source model amounts to the utility "effectively extending its monopoly and crowding out the industry from really participating," with the program "putting all their eggs in one basket." The basket in question is Sparkfund — a competent infrastructure services firm but a single point of contracting concentration that independent developers, by design, do not get to bid against.

The critique is not that distributed batteries are a bad idea. All of the opposing groups are, broadly, storage boosters. The critique is procurement. pv magazine USA summarized the shared objection cleanly: the $430 million program budget "shifts financial risks to ratepayers rather than leveraging private capital from third-party developers." Minnesota, in this framing, picked the expensive path because the expensive path is the one that preserves the utility's balance-sheet control.

The Cost Question, Handled Carefully

Whether the expensive path is actually more expensive is the argument that will outlast the approval. The $430 million budget for up to 200 MW is a defensible headline number. Dividing one by the other gives a per-kilowatt figure that sits above several published VPP benchmarks — but per-kW comparisons across VPP programs are genuinely difficult, because incentive structures, hardware ownership, and software costs rarely map cleanly between schemes.

Latitude Media noted the point without pinning a multiplier to it, writing that stakeholders observed "Xcel's cost estimates are higher than the costs of many existing VPP programs." A useful contrast sits next door in Xcel's own footprint: the Colorado Renewable Battery Connect program offers a $500-per-kilowatt incentive to customers who install and share control of their own batteries, per Latitude Media's reporting. But Colorado's model pays customers for behind-the-meter assets customers own; Minnesota's model pays a contractor to install assets the utility owns. The dollars land in different places and build different balance sheets. The comparison is suggestive, not dispositive.

The more concrete ratepayer number is cleaner. Per Xcel's filing as summarized by The Energy Mix, Capacity*Connect will add between 67 cents and $1.50 per year to a typical residential customer's bill through 2030. That is a small figure in absolute terms. The debate is less about the monthly line item and more about whether the same capacity could have been procured for materially less by opening the work to independent developers. That counterfactual is, by construction, unobservable. A competitive procurement was not run.

The Guardrails the Commission Actually Imposed

For all the focus on what the PUC approved, the conditions it attached to the approval deserve equal attention. Per the MPUC bulletin and Utility Dive's April coverage, Xcel must:

• Submit a comprehensive evaluation plan within 180 days, with explicit metrics for measuring cost savings and grid benefits.
• Develop specific grid-benefit estimates for distributed energy resources by November 2027.
• Commission an independent third-party program evaluation.
• Investigate and report on lessons from Xcel's Colorado behind-the-meter VPP pilot for Minnesota applicability.
• Consider placement of batteries in underserved communities.
• Partner with Building Strong Communities, a multi-trade apprenticeship preparatory program, to expand access to construction careers.

These conditions do two things worth naming. First, they create an evidentiary trail for the next docket. If Xcel's per-kilowatt costs do not reconcile to third-party benchmarks, the commission will know — and will have a contemporaneous record to point at in a future cost-disallowance proceeding. Second, they defer rather than decide the behind-the-meter question. pv magazine USA flagged that the PUC "did not move forward with a behind-the-meter virtual power plant program," and the ruling "establishes framework for future valuation of local clean energy resources." The behind-the-meter half of the VPP conversation — the half where third-party developers have their clearest opening — lives to fight another day.

The Bigger Picture for Minnesota's Grid

Minnesota's solar industry is mid-sized rather than dominant. Per pv magazine USA's reporting, the state sits 20th nationally with 3,264 MW of installed solar capacity, supports 4,793 jobs across 172 solar companies, and is projected to add roughly 3,297 MW over the next five years. That growth trajectory, combined with the Google-scale data-center load coming onto Xcel's system, gives the storage conversation unusual weight. The company has a 600 MW storage target by 2030 under the Upper Midwest Energy Plan, and Capacity*Connect is a visible down payment.

There is also a political-economy angle that Jigar Shah, former head of the U.S. Department of Energy's Loan Programs Office, captured in comments reported by Latitude Media. Shah framed the decision as Minnesota commissioners "affirming the value of VPPs as a core part of near-term capacity and grid infrastructure" — a signal to other states considering whether to treat distributed storage as ordinary utility rate base or as something that must always be procured from third parties. Whichever way the cost question resolves, the precedent has been set.

What Could Go Wrong

Four distinct risks shadow the program and deserve enumeration rather than a generic caveat.

Sole-source execution risk. Capacity*Connect's delivery rides on one contractor. Sparkfund is competent, but a 200 MW distributed build across dozens of sites in a compressed two-year window is a different operational animal from any single project the firm has executed previously. Permitting, interconnection, and siting variability in underserved-community locations can compound schedule slippage quickly.

Cost-benchmarking risk. Because no competitive solicitation was run, there is no contemporaneous market-clearing price against which to measure the program's per-unit cost. If the third-party evaluation in 2027 finds that equivalent capacity could have been procured for materially less via a competitive RFP, the PUC will have to decide whether to disallow costs retroactively — a politically fraught move that regulators are generally reluctant to make.

Ratepayer-risk asymmetry. The SEIA/MnSEIA/CCSA critique that Minnesota is the only state "to adopt a distributed storage model that forces everyday ratepayers to cover the investment risk instead of leveraging private capital" is structurally correct, regardless of how one feels about utility ownership. If dispatch performance undershoots, the ratepayer, not the developer, holds the bag.

Precedent-portability risk. Other commissions watching Minnesota may read Capacity*Connect as permission to let their incumbent utilities bypass competitive procurement on the next generation of distributed assets. Whether that is the right lesson depends on whether Xcel's costs prove reasonable — and that evidence will not arrive until 2027 at the earliest.

What This Does Not Tell Us — Yet

Four questions remain genuinely open after the April approval.

1. The competitive counterfactual. Nobody will ever know what a full RFP for 200 MW of distributed Minnesota storage would have cleared at. That is the unavoidable epistemic cost of declining to run one.
2. The per-dispatch economic value. The PUC required Xcel to develop grid-benefit estimates by November 2027. Until then, the peak-shaving, transmission-deferral, and reliability benefits are projections, not observations.
3. The Colorado-to-Minnesota transfer. Xcel's Colorado Renewable Battery Connect is customer-sited and behind-the-meter; Capacity*Connect is utility-sited and in-front-of-the-meter. The commission asked Xcel to report on Colorado lessons, but the two programs are structurally different enough that the mapping is nontrivial.
4. The Google cost-allocation. The $50 million Google contribution is reported in trade press but the regulatory allocation of that capital — what share offsets ratepayer cost versus data-center-specific rate-class cost — has not been publicly disclosed in detail.

Implications for the Next Two Years

For utility planners, Capacity*Connect is a live demonstration of distribution-grid battery ownership at a scale no other U.S. utility has attempted under a single program approval. The operational data will be valuable regardless of whether the procurement model is repeated.

For independent storage developers, the message is uncomfortable but clear. Utility-owned distributed storage is now a regulatory option in at least one jurisdiction. Convincing other state commissions to preserve competitive-procurement defaults will require harder evidence than prior-year cost curves — it will require a demonstrated delta between Minnesota's utility-owned costs and what third-party developers can deliver elsewhere.

For investors, the interesting instrument is not Xcel itself but the behind-the-meter VPP infrastructure that the PUC explicitly deferred. That is the part of Minnesota's distributed-storage market that remains contestable. The commission's November 2027 framework for valuing local clean energy resources will be the single most important regulatory milestone to watch.

For ratepayers, the near-term impact is modest — roughly a dollar a year — but the longer-term question is whether utility-owned distributed storage, built outside competitive procurement, delivers grid benefits proportional to its cost. That answer arrives later, on a docket not yet filed.

Key Takeaways

• Minnesota's PUC approved up to $430 million and up to 200 MW of utility-owned distributed batteries through 2028 — the first U.S. virtual power plant of this ownership shape at this scale.
• The decision broke with competitive-procurement norms by routing the entire buildout through a sole-source contract with Sparkfund, drawing joint opposition from SEIA, MnSEIA, and CCSA.
• A $50 million Google contribution ties the program to Xcel's Pine Island data-center deal and sits alongside a 600 MW Sherco BESS filing and a 30 GWh Form Energy iron-air project.
• Ratepayer impact is reported at 67 cents to $1.50 per year through 2030; the larger argument is about whether competitive procurement would have been cheaper, which is now an unobservable counterfactual.
• The commission imposed a 180-day evaluation plan, a November 2027 grid-benefit-estimate deadline, an independent third-party evaluation, and an obligation to report on Colorado VPP lessons — creating the evidentiary record for any future cost dispute.
• The behind-the-meter half of Minnesota's VPP conversation was deferred rather than decided, leaving the clearest opening for independent developers alive.