Deep Fission

Table of Contents

⚠ Disclaimer: This entry may be incomplete, out of date, or inaccurate. It is AI-maintained on a best-effort basis. Do not rely on it as a sole source — verify claims independently using the sources listed below.

Summary

Deep Fission is a Berkeley, California company (Nasdaq: FISN) founded in 2023 by Liz Muller and her father Rich Muller, Professor Emeritus of Physics at UC Berkeley — the same pair behind deep-borehole waste-disposal firm Deep Isolation. Its Gravity Reactor is a deliberately conventional 45 MWt / 15 MWe pressurized water reactor running four standard LEU fuel assemblies. The novelty is not the reactor but where it goes: one mile underground in a borehole, where the hydrostatic head of the water column supplies roughly 160 atmospheres — the operating pressure a surface PWR obtains from a thick steel pressure vessel — while surrounding rock provides shielding, security and isolation. If it works, the two most expensive structures in a conventional nuclear plant simply are not built.

It is the most interesting engineering idea in this section and, on the record so far, the weakest execution. Deep Fission missed its own July 4, 2026 criticality target entirely — no reactor has gone critical, no nuclear fuel is on site, and the only hardware delivered to Parsons, Kansas is a non-nuclear prototype canister. Its 2025 “go-public” reverse merger produced a stock that never traded. Its June 2026 IPO was cut from a marketed ~$157M at a ~$1.66B valuation to $40M gross at $16/share, and the shares trade around $7 — roughly 57% below issue. Its Q2 2026 10-Q carries a going-concern warning, ineffective disclosure controls, and prior restatements. Its 18.5 GW “customer pipeline” consists, by the company’s own disclosure, entirely of non-binding LOIs with no purchase commitments — and Kansas law currently prohibits the direct-to-datacenter power sales the model depends on.

Key Facts

Founded: 2023
HQ: Berkeley, CA (2705 Webster Street)
Type: Public — Nasdaq: FISN, listed June 18, 2026
Headcount: 61 at IPO
Co-Founder, CEO & Board Chair: Liz Muller — previously co-founded Deep Isolation (deep-borehole nuclear waste disposal) and Berkeley Earth
Co-Founder & CTO: Rich Muller, PhD — Professor Emeritus of Physics, UC Berkeley; MacArthur Fellow; 80+ patents. Liz Muller’s father. Not on the board
Reactor: “Gravity Reactor” / NRC design name Deep Borehole Pressurized Water Reactor (DB-PWR), “Deep Fission Borehole Reactor 1 (DFBR-1)”
Power: 45 MWt / up to 15 MWe per unit (thermal figure from the NRC page; most press cites only 15 MWe)
Fuel: Low-Enriched Uranium — four standard PWR fuel assemblies. Not HALEU, not TRISO
Borehole: ~1 mile / 1.6 km deep; minimum ~30-inch diameter per NRC (commercial range cited as 30–50 inches). Core temperature ~315 °C
Claimed economics: ~6-month construction per reactor; ten reactors = 150 MWe on a quarter-acre; scalable to 300 MWe–1.5+ GWe; LCOE 5–7 ¢/kWh — all company-asserted, pre-revenue, unaudited
Site: Great Plains Industrial Park, Parsons, Labette County, Kansas — 14,000+ acres, former Kansas Army Ammunition Plant. Groundbreaking December 9, 2025
Regulatory — DOE: Selected for the DOE Reactor Pilot Program (Aug 2025). Nuclear Safety Design Agreement approved August 6, 2026 — Gate 1 of a 4-gate sequence that does not authorize reactor operation. Remaining: PDSA (Gate 2), DSA + Technical Safety Requirements (Gate 3), Startup Approval requiring the Energy Secretary’s signature (Gate 4). The Q2 10-Q describes the NSDA as “conditionally approved.” Pursuing a NEPA categorical exclusion (published Feb 2026) — i.e. no Environmental Impact Statement
Regulatory — NRC: Pre-application only. Docket 99902126, engagement since May 2024. Conceptual Design Review complete; Conceptual Design Description under review; planned application type Combined License, targeted H1 2027. Nothing filed
Financing history: $4M founder funding (2024); a sought $15M seed in April 2025 reportedly failed; Sept 8, 2025 reverse merger into shell Surfside Acquisition with a $30M private placement at $3.00/share — the stock never actually traded; Feb 10, 2026 $80M private placement at $15.00/share (Ed Eisler/EE Holdings, Mark Tompkins/Montrose Capital, Blue Owl Capital Real Assets; Goldman Sachs advising); IPO priced June 17, listed June 18, 2026 — $16.00 × 2,500,000 shares = $40.0M gross / $34.3M net
Market (late Aug 2026): ~$7.13/share**, ~**$420M market cap on ~58.9M shares; 52-week range $7.08–$19.00; −57% from the IPO price. Wall Street Zen downgraded to Sell on August 23, 2026
Financials (Q2 2026 10-Q, filed Aug 3): **$0 revenue, ever.** Q2 net loss $31.0M; H1 net loss $52.4M; accumulated deficit $119.1M; cash $93.0M. Going-concern warning; disclosure controls “not effective,” with material weaknesses in ICFR and prior restatements (SAFE note valuation, stock-based compensation)
Fuel supplier: Urenco USA (Eunice, NM), binding agreement signed February 25, 2026; $9.2M paid against the LEU commitment
EPC: Day & Zimmermann (announced June 2, 2026) — pre-construction planning and above-ground construction. Drilling contractor not publicly named

What It Is / How It Works

The core idea: gravity replaces the pressure vessel. A conventional PWR needs about 155–160 atmospheres in the primary loop to keep water liquid at ~315 °C. Getting that pressure requires a forged steel vessel several inches thick, and containing the consequences of losing it requires a reinforced concrete dome. Both are among the most expensive and schedule-critical items in a nuclear build. Deep Fission’s insight is that a one-mile column of water produces roughly the same pressure for free. Lower the reactor down a borehole on a string, and the water above it does the work of the pressure vessel; the kilometre of rock above it does the work of the containment building, plus shielding and physical security.

The primary loop circulates by natural convection — no primary pump. A secondary loop with a surface-mounted pump raises steam through a heat exchanger at the surface. The reactor sits in a canister that can be raised and lowered. The fuel is four standard PWR assemblies at standard LEU enrichment, which means Deep Fission is entirely outside the HALEU and TRISO constraints that bind most of this section. On paper the combination is elegant: proven fuel, proven reactor physics, and the capital-intensive structures deleted rather than optimized.

What is not described. Public materials do not explain refueling or retrieval mechanics at depth, waste handling, how a 30-inch canister is serviced a mile down, or what happens if the string fails. These are not incidental details — they are the operational core of the concept, and they are absent.

Where the demonstration actually is. The well drilled at Parsons is 8 inches in diameter, a data-acquisition hole for geological, hydrological and thermal characterization. The reactor needs 30–50 inches, and those specifications are not finalized. The canister delivered on July 8, 2026 contains no nuclear components — it is commercial-grade proof-of-concept hardware that underwent hydrostatic testing. Between the current state and a critical reactor sits a wider-diameter borehole requiring a state permit, fuel that trade coverage describes as “a 2027 problem,” and three further DOE authorization gates.

The regulatory bet. Deep Fission’s stated plan is to build the Parsons unit under DOE authorization and then convert it into a fully NRC-licensed, commercially operating unit. There is no precedent for a DOE-authorized demonstration reactor converting to an NRC commercial licence, and the company asserts it without describing the mechanism.

Notable Developments

2026-08-06: DOE approves the Nuclear Safety Design Agreement for Gravity — Gate 1 of 4; permits continued design and testing, does not authorize operation. (Deep Fission; POWER)
2026-07-08: Prototype reactor canister delivered to Parsons after fabrication and hydrostatic testing — non-nuclear commercial-grade components only. (WNN)
2026-07-04: ⚑ Self-imposed criticality target missed. No reactor critical, no fuel on site. By May 2026 the company had stopped giving a criticality estimate. (TechCrunch)
2026-06-24: Announces an 18.5 GW “customer pipeline” — the company’s own release states the LOIs “are non-binding and do not contain commitments to purchase electricity, finance projects, construct facilities, grant exclusivity, or deploy a specified number of reactors.” No customer named. (Data Center Dynamics)
2026-06-17/18: IPO priced and listed — $16.00 × 2.5M shares, $40.0M gross, $34.3M net; Benchmark Company and Seaport Global joint leads, Maxim co-manager. **Downsized ~75% and priced ~38% below the marketed $24–26 range.** (BusinessWire; Renaissance Capital)
2026-06-02: Day & Zimmermann EPC partnership for pre-construction planning and above-ground construction at Parsons. No value, timeline, or binding designation disclosed. (PRNewswire)
2026-03: Drilling begins on an 8-inch-diameter, 6,000-ft data-acquisition well for site characterization.
2026-02-25: Urenco USA LEU fuel agreement — binding; enrichment, tonnage, term and price undisclosed. (WNN)
2026-02-10: $80M private placement at $15.00/share; Blue Owl Capital Real Assets among investors. (Orrick)
2025-12-09: Groundbreaking at Parsons, Kansas. 2025-12-04: Site announced; original criticality target July 4, 2026.
2025-09-08: Reverse merger into shell Surfside Acquisition with a $30M placement at $3.00/share — the stock was never publicly traded. (TechCrunch)
2025-08: Selected for the DOE Reactor Pilot Program. 2025-04: A sought $15M seed round reportedly failed.
2025-01-07: Endeavour Energy “commits to co-develop 2 GW” for edge-datacenter portfolio; first reactors “expected operational 2029” — no PPA, no pricing, no take-or-pay.
2024-05: NRC pre-application engagement begins.

Key People

Liz Muller — Co-Founder, CEO & Board Chair

LinkedIn: linkedin.com/in/lizmuller
Background: Co-founder of Deep Isolation (deep-borehole nuclear waste disposal) and Berkeley Earth; B.Math UC San Diego; M.Intl Mgmt ESCP; named inventor on multiple patents
Notes: ⚑ The Deep Isolation lineage is both the source of the borehole expertise and a specific point of local friction — Parsons residents have raised the concern that a failed reactor project could become a waste-disposal site. Muller also hedges commercialization publicly: “we haven’t yet committed to how big this commercial project could potentially be.”

Rich Muller, PhD — Co-Founder & CTO

LinkedIn: linkedin.com/in/richard-muller-0b34b31
Background: Professor Emeritus of Physics, UC Berkeley; MacArthur Fellow; 80+ patents, 100+ papers; 30+ years of national-security advisory work; co-founded Deep Isolation with his daughter
Notes: ⚑ Not a member of the board of directors, despite being co-founder and CTO

Michael Brasel — COO

Background: Former VP Nuclear Programs at NuScale Power; plant manager at Dairyland Power; US Navy nuclear operator

Mark Pérès — Chief Nuclear Officer

Background: 40+ years; Kairos Power, NuScale, Fluor, Los Alamos

Mark Schmitz — CFO

Background: 40 years; former CFO of Goodyear, Itron, Plug Power, Alghanim

Board of Directors

Liz Muller (Chair); Jonathon Angell (Angell Investments — ⚑ also Audit Committee Chair at Deep Isolation, a related-party overlap worth noting); Thomas Glanville (Eschelon Advisors; Audit Committee Chair); Leslie Goldman Tepper (Nominating & Governance Chair); Blake Janover (J2 Labs; Compensation Committee Chair).

People — Last Reviewed: 2026-08-27

Supply Chain Position

Layer	Detail
Fuel	Urenco USA (Eunice, NM) — binding LEU agreement, Feb 25, 2026; four standard PWR assemblies per reactor. No HALEU, no TRISO — outside both of this section’s principal fuel chokepoints
Fuel commitment	$9.2M paid against the LEU commitment (per Q2 10-Q)
EPC (above ground)	Day & Zimmermann — pre-construction planning and aboveground construction only
Drilling	Not publicly named. Methods described as borrowed from oil/gas and geothermal; $4.1M borehole casing commitment on the balance sheet
Canister fabrication	Not named
Site	Great Plains Industrial Park, Parsons, Labette County, KS (former Kansas Army Ammunition Plant)
Permitting	Kansas Department of Health and Environment (borehole); DOE authorization (reactor); NRC pre-application
Customers	18.5 GW of non-binding LOIs; Endeavour Energy 2 GW framework; Great Plains Industrial Park non-binding LOI

⚑ The one supply chain that genuinely is de-risked. Standard LEU in standard PWR assemblies from an operating, licensed US enricher is the most conventional fuel position of any company in this section — no BWXT TRISO queue, no HALEU allocation, no first-of-a-kind fabrication. It is a real advantage that the rest of the programme has not yet been able to use.

⚑ The drilling contractor is the unnamed critical path. The entire concept depends on drilling a 30–50-inch-diameter hole a mile deep and emplacing a reactor in it. No drilling partner has been announced, the diameter specification is not finalized, and Day & Zimmermann’s disclosed scope is explicitly above ground.

⚑ A legal blocker, not a technical one. Kansas law prohibits direct power sales to customers such as data centers. Deep Fission is working with utility Evergy on a workaround; it is unresolved. The business model — sell power to a co-located datacenter — is currently illegal in the state where the company is building.

Claim Verification

Claim: The IPO raised ~$157M at a ~$1.66B valuation

Status: False as stated — those were proposed terms that were never achieved

Supporting: The May 2026 S-1/A marketed roughly 6 million shares at $24–$26, seeking ~$156–157M at up to a $1.66B valuation. Those figures circulated widely.

Refuting / questioning: The executed deal was **2,500,000 shares at $16.00 = $40.0M gross, $34.3M net**, priced June 17 and listed June 18, 2026 — a ~75% reduction in proceeds and ~38% below the low end of the marketed range. Renaissance Capital's "$40M market cap" field is the deal size, not market capitalization; implied market cap at the $16 IPO price was ~$955M, and the shares now trade at roughly $420M.

Summary: Cite $16.00/share, $40M gross, June 17–18, 2026. The $24–26 / $1.66B figures describe a deal that did not happen.

Claim: Deep Fission has signed datacenter agreements

Status: No — all non-binding

Supporting: The company announced an 18.5 GW “customer pipeline” spanning data centers, co-developers, industrial parks and strategic partners, plus a 2 GW Endeavour Energy co-development commitment.

Refuting / questioning: Deep Fission’s own June 24, 2026 release states the LOIs “are non-binding and do not contain commitments to purchase electricity, finance projects, construct facilities, grant exclusivity, or deploy a specified number of reactors.” No customer is named. The Endeavour arrangement discloses no PPA, price, take-or-pay or capacity payment. The Great Plains Industrial Park arrangement is an LOI whose own president says “there can’t be any future commitments at this time until we know what the commercialization status is.” The Blue Owl relationship is an investment plus a non-binding MOU.

Summary: Zero binding offtake. The 18.5 GW figure should never be cited as contracted or committed capacity.

Claim: The August 2026 DOE approval is a major regulatory milestone

Status: Overstated — it is the first of four gates and authorizes nothing operational

Supporting: DOE did approve the Nuclear Safety Design Agreement on August 6, 2026, and it is a real step in the Reactor Pilot Program authorization sequence.

Refuting / questioning: The NSDA is Gate 1 of 4: it sets the safety framework and permits continued design and testing. Gates 2 (PDSA, at ~50% design completion), 3 (DSA + Technical Safety Requirements) and 4 (Startup Approval, requiring the Energy Secretary’s signature) all remain. The company’s own Q2 10-Q calls it “conditionally approved.” Beyond Nuclear’s framing — headlined “Deep Fission ‘approval’ is for a drawing, not a ‘milestone’” — captures the objection, though the article body could not be retrieved to quote directly.

Summary: A design-stage approval, not an authorization to build or operate a reactor.

Claim: 5–7 ¢/kWh LCOE and ~6-month per-reactor construction

Status: Company-asserted, unaudited, pre-revenue

Summary: No reactor has been built, no borehole of the required diameter has been drilled, and the company has never generated revenue. Treat both figures as marketing.

Risk & Skepticism Notes

Going concern. The Q2 2026 10-Q states “substantial doubt about the Company’s ability to continue as a going concern within one year”; current cash will not fund twelve months of planned operations. The same warning appeared in December 2025 and May 2026 filings. Disclosure controls are stated as “not effective,” with material weaknesses in internal control over financial reporting and prior restatements relating to SAFE note valuation and stock-based compensation.
A pattern of missed and repriced milestones. The July 4, 2026 criticality target was missed outright. The 2025 reverse merger produced a stock that never traded — TechCrunch noted the $3.00/share price against the customary $10 SPAC price “suggest[s] that Deep Fission wasn’t able to raise cash from new or existing shareholders.” The 2026 IPO was cut ~75% and priced ~38% below range. The stock is down ~57% within ten weeks of listing.
Local opposition. Parsons (population ~9,600) is divided. Marjorie Reynolds of the Prairie Dog Alliance: “My fear is that this project will not work and then they’re going to decide that they’re going to encapsulate nuclear waste and fill up the borehole” — pointed, given the founders’ Deep Isolation background. County Commissioner Vince Schibi publicly doubted the July timeline was feasible. Paul Gunter of Beyond Nuclear calls the NEPA categorical exclusion “unacceptable,” citing high-level waste beneath the town.
Engineering gap between demo and product. 8-inch characterization well vs a 30–50-inch reactor borehole whose specification is not final; a delivered canister containing no nuclear components; no named drilling contractor; refueling and retrieval mechanics undescribed.
Comparative benchmark. TechCrunch contrasted the $1.66B ask against X-energy, which has revenue and materially more NRC progress — arguing the valuation was disconnected from development stage. The market subsequently agreed.

The Infinite Unknown