In January, Sparrow Mahoney signed a precedent agreement that few in Alaska thought would become a contract. On 5 November, it did. STAK Energy, the Anchorage company he founded, closed a 15-year gas sale agreement with ExxonMobil Alaska Production for a power campus on the North Slope designed to generate nearly 3 gigawatts and connect to no public grid at all.
The campus has a name: Project Aaka. It sits in an industrial corridor with year-round road and airport access, roughly 26 miles south of Deadhorse, on 715.4 acres of state land. Gas comes from North Slope production that has been reinjected or used in the field for decades because there was no route to a larger market.
STAK's pitch is direct. Put the gas into turbines where it is produced, sell the electricity to computing operators, and skip the interconnection queue entirely.
A plant built around a missing wire
That absence is the product. US data-center operators now wait years for grid connections in the largest markets, and the queue itself has become a bottleneck with its own economics. Moody's estimates that US data centers consumed about 224 TWh in 2025, more than all of New York State. Behind-the-meter capacity is shifting toward gas turbines, reciprocating engines and fuel cells.
STAK's plan fits that template. The company says the campus will run entirely off-grid on high-efficiency gas turbines, with no anticipated impact on Alaska's public electric grid or ratepayers. First power is targeted for 2030, with phased expansion through 2032.
Alaska's Department of Natural Resources has published a preliminary decision on the lease. It proposes a 50-year term, describes a power system with a capacity range of one to three gigawatts, and estimates about $500 million in capital expenditure and roughly 60 permanent operating positions. The same review flags potential cumulative effects on air quality, wetlands, permafrost, caribou and subsistence resources.
Those caveats define the critical path.
The customer has not been named
What STAK has not disclosed matters as much as what it has. The announcement did not name a computing customer, a turbine supplier, a gas volume, a price or a delivery start date. ExxonMobil is supplying fuel under a 15-year agreement. The buyer of the electricity is still an open slot.
For an infrastructure project, that is normal at this stage. It is also the difference between a supply contract and a business. A 3 GW campus needs anchor tenants willing to sign long-term offtake before lenders will fund turbines, substations and a gas pipeline. STAK's leadership has built large Arctic projects before, and the team's backgrounds run through ExxonMobil, ConocoPhillips, Santos and the Alaska Gasline Development Corporation. That history helps with permits. It does not create demand.
For fifty years, the North Slope has produced gas without a way to put it to work at scale. Today, that changes. Our agreement strengthens the case for bringing more power, and lower energy costs, home to Alaska families.— Sparrow Mahoney, founder and CEO, STAK Energy
Behind the meter is becoming the default
The market STAK is entering is no longer experimental. Omdia projects that data centers could deploy 35 GW of behind-the-meter power by 2030. GE Vernova expects its gas equipment order backlog to reach 125 GW by the end of 2026 while annual turbine output rises toward 20 GW, according to Moody's. Turbines are scarce, and the queue for them is measured in years.
Competition is forming around the same logic. Caterpillar has received a 2 GW equipment order for American Intelligence & Power's Monarch Compute Campus in West Virginia, with room to expand to 6 GW. Siemens Energy and Eaton have partnered on an off-grid gas plant design aimed at the data-center sector. VoltaGrid and Vantage Data Centers have a 1 GW off-grid gas deployment across North America.
Each of those projects sells speed. The pitch is electrons this decade, on a site the operator controls, priced above grid power.
Alaska adds two variables that lower-tier states cannot match. The North Slope holds an estimated 96 Tcf of undiscovered, technically recoverable natural gas, nearly three times its roughly 35 Tcf of proven reserves, according to the USGS. Ambient temperatures reduce cooling load. And the gas is already at the surface, separated from the Cook Inlet supply that heats Alaskan homes.
That last point is politically useful. Project Aaka can argue it is not competing with residential ratepayers for the same molecule.
What the state gets, and what it carries
STAK frames the project as a fiscal and employment story. The gas sale agreement and the campus are expected to generate new state and local tax revenue. Construction is projected to support thousands of peak workers, with a permanent operating crew in the dozens rather than the hundreds. For a state that has watched North Slope gas struggle to find a market, that is a credible political offer.
The environmental file is less tidy. Permits from federal, state and local authorities are still required. The DNR review names air quality, wetlands, permafrost, caribou and subsistence resources as areas of potential cumulative impact. A gas-fired campus lowers grid risk for its customers and raises emissions questions for everyone else.
At a moment when the country urgently needs more power, Alaska again has the resources, the land, and the will to provide it.— STAK Energy statement, 5 November 2026
The 2030 test
The investment case rests on four dated checkpoints. Final permits. A final investment decision. Turbine and pipeline orders. A named anchor customer. None of those has happened yet, and the 2030 first-power date is four years away.
That makes Project Aaka a 2030 option, not a 2026 cash flow. The 15-year ExxonMobil agreement de-risks the fuel side and gives the project a credible sponsor. It does not de-risk demand, equipment lead times or the Alaska permitting calendar.
For investors, the interesting signal is structural. Stranded gas plus stranded grid capacity plus AI demand has produced a new asset class: the off-grid compute campus. STAK is one of the first to attach a major oil company's supply contract to it in the Arctic.
The precedent matters beyond Alaska. If Project Aaka reaches first power, the model travels to every gas field with no pipeline and every data-center market with no queue capacity. If it stalls, the reason will be visible early: no customer, no turbines, no permits.