On October 20, 2026, Deep Fission announced that it had signed a binding 25-year master energy purchase agreement with the Great Plains Development Authority (GPDA), which owns an industrial park in Kansas, for power generated by underground reactors. The deal moves the project from a December 2025 letter of intent to a legally binding contract, but the amount of power purchased will depend on demand from tenant companies, and the agreement sets no minimum purchase obligation. The plan is to supply up to 150 MW in the initial phase and then expand capacity to as much as 2 GW as demand grows. Meanwhile, an economic model for the mass-production stage, published the same day, assumes that all of the expected output is sold. To judge how much recurring revenue this contract can actually generate, it is not enough to look at the contract term and supply capacity; the actual purchase volume and sale price also need to be examined.
25-year power purchase agreement, with purchase volume tied to tenant demand
The Great Plains Industrial Park, owned by GPDA, is located in Parsons, Kansas, on the site of a former U.S. Army ammunition plant. It covers more than 14,000 acres and has existing infrastructure such as rail access, and GPDA hopes a stable power supply will help attract businesses. The agreement formally establishes a long-term power supply framework between Deep Fission, which is developing the generating facilities, and GPDA, which is responsible for attracting companies that will use the electricity. According to Deep Fission's announcement, the contract covers a planned commercial project at the industrial park.
Under the contract, Deep Fission will supply power after it carries out development under the specified conditions, obtains the necessary approvals, and begins operations. GPDA, for its part, will purchase the available power within the contracted capacity to meet tenant demand. The purchase obligation may also be transferred to tenants that meet certain conditions. However, purchase volumes will depend on actual demand at the industrial park, and no minimum purchase obligation is set. Even with a 25-year term, therefore, there is no guarantee that a fixed amount of power will be purchased continuously over that period.
The initial supply capacity is up to 150 MW, with a target commercial operation date of September 30, 2028. The subtitle of the announcement refers to expanding to as much as 2 GW by 2031, and the main text describes a plan to increase supply capacity in stages, up to a total of 2,000 MW, as demand rises. These figures, however, are capacity targets for facilities yet to be built. They do not represent the output of equipment already in operation, nor the amount of power that tenants have committed to buy.
Note also that page 12 of the business presentation released the same day gives the signing date of the master energy purchase agreement as October 16. October 20 is the date the contract was announced, and the two should be distinguished. The move from a letter of intent to a binding contract is clear progress toward commercialization, but assessing the revenue outlook requires further confirmation of actual purchase volumes and sale prices.
$110/MWh and 93% utilization: how the mass-production model differs
Page 16 of the business presentation that Deep Fission published the same day includes an economic model for mass-produced reactors. It is not a concrete financial forecast for the first unit; it is an estimate assuming a stage at which efficiency has improved through standardized design and repeated construction. The model assumes a 40-year design life and a target output of 15 MWe, or 8 MWe in some configurations, and it also sets certain conditions for the sale price and volume of electricity.
The initial sale price assumed in the model is $110/MWh, which is not a figure for the cost of generation. The explanation of this price is also based on 2024 economic conditions. Utilization is assumed to be about 93%, but this is a share of operating time in the model, not operating performance demonstrated by a completed underground reactor. Compared with the terms of the GPDA contract, the assumptions about purchase volume in particular differ.
| Item | This GPDA contract | Mass-production illustrative model (business presentation, p. 16) |
|---|---|---|
| Purchase volume | Purchased up to contracted capacity according to tenant demand; no minimum purchase obligation | Assumes a contract to purchase all expected output |
| Sale price | Not stated in the announcement | Assumes an initial $110/MWh; price explanation based on 2024 economic conditions |
| Utilization | Does not indicate operating performance | Assumes about 93% |
While this contract has no minimum purchase obligation, the mass-production economic model assumes that all expected output is sold. It also cannot be confirmed that the model's $110/MWh price is the sale price that applies under this contract. Even if a reactor reaches the assumed utilization, it does not necessarily follow that all the power it generates can be sold. The 25-year term and maximum supply capacity alone cannot support the sales volumes or revenue the economic model assumes.
That said, the announcement does not state a total contract value, so not all of the transaction terms are public. Even without a minimum purchase obligation, continuous power sales could result if tenant demand grows sufficiently. It would therefore be inappropriate to conclude, on the basis of this contract alone, that the mass-production model's revenue targets cannot be met. Likewise, multiplying the 150 MW or 2 GW capacity figures by the model's sale price and treating the result as the total contract value is not valid.
Reactors about one mile underground; the demonstration test reached 100 feet
Deep Fission is developing pressurized water reactors fueled with standard low-enriched uranium, to be installed about one mile underground. According to page 7 of the business presentation, the design sets both the water column leading underground and the pressure inside the reactor at about 160 atmospheres, and carries heat to the surface through three independent closed loops. By using the geology and the water column for pressure maintenance, shielding, cooling and other functions, the company aims to reduce the equipment needed on the surface. The company's explanation that underground installation can cut construction costs is based on this design feature. However, the economics have not been confirmed through actual commercial operation.
On installation technology, the company conducted a test with a full-scale canister on September 3 and announced the results on September 10. In the test, a canister 20 feet (about 6.1 m) long was lowered through a borehole 34 inches (about 86 cm) in diameter to a depth of 100 feet (about 30 m). The canister's position was then adjusted and it was pulled back to the surface. The significance lies in confirming the whole sequence, from installation to retrieval, using existing commercial drilling equipment. However, the canister used was a non-nuclear, full-scale mock-up containing no nuclear fuel, and the test did not involve starting a nuclear reaction about one mile underground, extracting heat and generating power.
The reactors also have to pass through several licensing stages. According to Deep Fission's September 15 announcement on its safety design agreement, approval by the U.S. Department of Energy (DOE) is an early-stage step in the safety review of the pilot reactor. Obtaining operating authorization from DOE requires additional document submissions and review, and for commercialization the company must also obtain authorization from the U.S. Nuclear Regulatory Commission (NRC). The regulatory plan the company has published lays out a path of proceeding with demonstration under DOE and then obtaining NRC commercial authorization. It also states a policy of transitioning the demonstration reactor directly into commercial operation.
The technology to install and retrieve a canister underground, the technology to sustain a nuclear reaction deep underground and generate power, and the licensing needed for commercial supply are separate verification challenges. The results confirmed in the current installation test cannot be equated with a demonstration of commercial power generation from underground reactors. And when comparing a contract in which purchase volume depends on demand with an economic model that assumes all output is sold, it is necessary to look not only at generating capacity but also at the amount of power that can actually be operated and sold.
Conditions for turning 150 MW of commercial operation into recurring revenue
Meeting the September 30, 2028 target for commercial operation will require satisfying multiple conditions, including not only NRC authorization but also Kansas state approval and the arrangements needed to deliver the power. Expansion to as much as 2 GW assumes that power demand at the industrial park grows, and starting operations at the initial facility and later expanding supply by adding reactors are separate stages. How far the effort to attract companies to the park progresses, and how much power demand results, will affect both the plans to add capacity and actual sales volumes.
On the demand side, the amount of power tenants buy continuously and the actual sale price will be key to assessing the profitability of the mass-production model. On the supply side, the focus is on whether the company can carry out integrated demonstrations, including heat recovery and power generation deep underground, and obtain the necessary authorizations. The roughly 93% utilization assumed in the economic model also cannot be judged from the canister installation test alone and will need to be verified through actual reactor operation.
The 25-year master energy purchase agreement moves Deep Fission's commercialization plan forward, but it does not in itself guarantee a certain level of sales over the long term. Only if tenant demand translates into continuous power purchases, and stable underground operation and commercial authorization are achieved, can this long-term contract framework develop into an actual power supply business.