DOE Picked 31 SPARK Projects. The Federal Share Is Only 36 Percent — And Nothing Is Signed Yet

Read the Department of Energy's September 24 announcement quickly and you see a $1.9 billion grid award. Read the second number and the story inverts: recipients are bringing $3.35 billion of their own capital to the same 31 projects. The federal government is supplying 36 percent of a $5.25 billion program. Xcel Energy alone is matching at 80 percent on the Colorado project.

That ratio is the most instructive fact in the entire release, and it explains more about who won than any of the technology language in the press materials.

The Office of Electricity selected 31 projects across 26 states under SPARK — Speed to Power through Accelerated Reconductoring and other Key Advanced Transmission Technology Upgrades — from the funding opportunity DE-FOA-0003580 issued February 25, 2026. Applications closed May 5. Selections landed 142 days later. Collectively the projects are expected to reconductor or rebuild more than 1,500 miles of transmission line, deploy grid-enhancing technologies across nearly 21,000 miles, and add more than 23 gigawatts of transfer capacity serving roughly 100 million Americans. Energy Secretary Chris Wright framed it as "commonsense energy addition policies that lower electricity prices and strengthen our grid."

Selection Is Not an Award, and the Difference Matters Through January

DOE's own timeline puts award execution in the window of October 2026 through January 2027. Everything announced on September 24 is an intention to fund. Selected applicants now enter negotiation, where scope, budget, milestones, and cost-share commitments get pinned down before any money obligates.

This distinction has teeth. Under the predecessor rounds of the Grid Resilience and Innovation Partnerships program — the Infrastructure Investment and Jobs Act authority that funds SPARK, sized at up to $10.5 billion over five years — selected projects have been renegotiated down, restructured, and in some cases withdrawn when a partner's board declined to commit the match. DOE has not published a schedule for converting these 31 selections into executed agreements.

If you are on the list, the next 90 days are the real competition. The practical risks are mundane and fatal: a cost-share letter that was conditional in the application and has to become binding; an interconnection study that moved; a utility commission approval that now has to be filed; a prevailing-wage and apprenticeship plan that was described in prose and now needs a staffing schedule attached. Applicants who treated the application as the finish line tend to discover the gap in November.

What the Named Projects Tell You About the Winning Profile

DOE has not released per-project federal amounts for all 31, but enough have surfaced to read the shape of the portfolio.

The Colorado Energy Office is the prime recipient on CO2TX, a $1.2 billion project receiving $250 million in federal funds — a 21 percent federal share. The project relocates the Western-to-Eastern interconnection tie from Lamar, Colorado to the higher-capacity May Valley substation and upgrades existing conductors, lifting transfer capability from 210 megawatts to 700 megawatts with a path to 2,000. Xcel Energy Services, Public Service Company of Colorado, Southwestern Public Service, and Southwest Power Pool are partners; Xcel is carrying an 80 percent match against total project cost.

Oklahoma's Office of Management and Enterprise Services is leading the $832 million Three Corners Connection, a roughly 290-to-300-mile high-voltage DC line running at up to 525 kilovolts from the Western Electricity Coordinating Council near the Comanche Generating Station in Pueblo County, Colorado to the Southwest Power Pool at the Optima substation near Guymon, Oklahoma.

PPL Electric Utilities took up to $71.5 million for the Montour Grid Resilience and Advanced Reconductoring Project, rebuilding approximately 29.3 miles of an existing 230-kilovolt corridor serving the Susquehanna Valley, Lehigh Valley, and northeastern Pennsylvania.

Other named recipients include Alabama Power, Duke Energy Carolinas, Eversource Energy, and Kit Carson Electric Cooperative.

Two patterns jump out. First, the two largest projects are led by state agencies, not utilities — the Colorado Energy Office and Oklahoma's OMES. That is Topic Area 3, the Grid Innovation Program, which is open to states, tribal nations, local governments, and public utility commissions rather than to utilities directly. If you are a state energy office reading this as a utility story, you are reading it wrong: the structure rewards a state entity that can convene multi-jurisdictional partners and bring a utility's balance sheet as the match.

Second, Kit Carson Electric Cooperative sitting on the same list as Duke Energy is not decoration. Topic Area 1 eligibility runs to electric utilities, generators, storage operators, transmission and distribution providers, and fuel suppliers — a rural electric co-op with a credible reconductoring scope and a financeable match competes in that pool. Small applicants lose GRIP competitions on cost share, not on merit.

Reconductoring Won Because Permitting Did Not

The technical thesis behind SPARK is that the cheapest new transmission capacity in America is inside rights-of-way that already exist. Replacing conductors on standing towers with advanced composite-core cable can roughly double a line's capacity. Grid-enhancing technologies — dynamic line rating, power flow controllers, topology optimization — squeeze more out of the same steel without touching the conductor at all. The 21,000-mile GETs figure dwarfs the 1,500-mile reconductoring figure for exactly that reason: GETs are cheap per mile and fast to deploy.

Greenfield transmission, by contrast, takes a decade or more between siting, eminent domain, environmental review, and state approvals. The administration's grid policy in 2026 has consistently favored capacity that can be added inside existing corridors, a posture that also shaped the August 2026 bulk power system emergency declaration and the USDA Rural Utilities Service ARC program.

For applicants, the strategic read is blunt: a proposal whose critical path runs through a new right-of-way is competing against proposals whose critical path runs through a procurement order. In a program named for speed, that is a structural disadvantage no amount of narrative fixes.

What This Means for the Next Round

When we covered SPARK's launch in March, the open question was whether DOE would spread the money thin across many small resilience projects or concentrate it in a handful of interregional transfers. The answer is both — and the concentration is heavier than the project count suggests. Two projects account for more than $2 billion of the $5.25 billion total value.

Three things to carry into the next GRIP solicitation:

Build the match before you build the narrative. A 36 percent average federal share means DOE is functionally underwriting projects that a utility or state has already decided to do. Applications that made the federal award the enabling condition for the project performed worse than applications where the federal award accelerated or expanded something already financeable. Get a binding-in-principle commitment from the cost-share partner before the concept paper, not after selection.

Look at who is eligible for which topic area. The three SPARK tracks had genuinely different eligible-applicant lists — Topic 2 alone opened to universities, for-profits, nonprofits, government entities, and tribes. A nonprofit or university that reflexively skipped SPARK as a "utility program" skipped a track written for them. Read the eligibility section per topic area, every time.

Assume a 10-to-14-week selection lag and plan the negotiation staffing now. February 25 issue, May 5 close, September 24 selection, awards October through January. That cadence lets you calendar backward: if the next round opens on a similar schedule, your cost-share documentation needs to be board-approved roughly six weeks before the application deadline, not the week of.

The 31 selected teams have until January to convert a press release into an obligated award. Everyone else has until the next notice of funding opportunity to fix the thing that actually decided this round, which was never the technology — it was whether somebody else was already willing to pay for two-thirds of it. Matching that kind of readiness against a live solicitation calendar is the part Granted was built to shorten.