Applied Atomics

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⚠ Disclaimer: This entry may be incomplete, out of date, or inaccurate.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

Applied Atomics is a seed-stage company, founded around 2025 by ex-SpaceX operators Ben Kellie and Paul Keutelian, that in June 2026 secured an exclusive license from BWX Technologies to deploy the mPower SMR in land-based commercial facilities in the US, Canada and elsewhere. mPower is a 195 MWe / 575 MWth integral pressurized water reactor on standard LEU fuel with a two-year-plus refueling cycle and no primary coolant piping — technically a credible, conventional, well-characterized design.

The transaction is genuinely interesting: BWXT retained the mPower engineering archive and test facilities after shelving the program, and licensing it out is a low-cost way to see whether a smaller, faster operator can carry it further. The concern is scale. Applied Atomics has raised roughly $12 million and has one to ten employees.** The mPower program it intends to complete consumed approximately **$500 million — around $375–400M private plus $111M disbursed by DOE — before Bechtel and BWXT terminated it in March 2017, and B&W’s own accounting estimated a further ~$650M was needed to reach construction. Against that, the company states it is “working with early customers to secure locations and long-term power agreements for the first 10 GW of planned capacity.”

There is also a material regulatory misstatement in the company’s framing. Applied Atomics says it will “resume mPower design certification activities” and “re-engage the NRC.” mPower never had a design certification application submitted — it was in pre-application review only, and the application expected in Q4 2013 was never filed. Applied Atomics itself does not appear on the NRC’s current advanced-reactor pre-application list, holds no DOE authorization, and has no Reactor Pilot Program slot. See Claim Verification.

Key Facts

Founded: ~2025 (inferred — a March 2026 release describes $12M raised “in its first 12 months”; no primary source states an incorporation year)
HQ: Anchorage, Alaska with a design studio in Los Angeles per the June 2026 release; March 2026 materials and local coverage describe it as Los Angeles-based (Arts District). Reads as an HQ-of-record relocation between March and June 2026
Type: Private, seed stage
Co-Founder & CEO: Ben Kellie — ex-SpaceX (led infrastructure including the Falcon 9 droneship); founded The Launch Company, sold to Voyager Space; lives in Anchorage; BS Mech. Eng. University of Alaska Fairbanks
CTO: Paul Keutelian — ex-SpaceX (technical operations / regulatory compliance); co-founder of Radiant Nuclear (see Radiant Industries)
Co-Founder & CFO: Mark Lambert — 17+ years in startups and regulated-industry B2B SaaS
Headcount: 1–10 employees per Crunchbase. The website’s “112-person delivery team” refers to a plant-delivery staffing model, not headcount
Funding: Pre-seed (~March 2025, Alpaca VC); Seed $8.3M announced March 25, 2026**, co-led by Transition (Ari Helgason) and Morpheus (Howard Ko) with Alpaca VC. **Total ~$12M
Licensed technology: mPower — integral PWR, 195 MWe / 575 MWth, integrated steam generators inside a single vessel (no primary coolant piping, eliminating large-break LOCA), standard LEU (<5%), refueling cycle ≥2 years, factory-manufactured and conventionally transportable
License terms: Applied Atomics gets exclusive rights to develop and deploy mPower in land-based commercial nuclear facilities in the US, Canada and elsewhere. BWXT retains full ownership of the mPower IP, exclusive manufacturing rights for all mPower components, and royalty rights on components made by Applied Atomics or third parties, and provides paid technical support. No royalty rate, license fee, equity component, duration, or reversion terms disclosed
Business model: Vertically integrated independent power producer — designs, builds, owns and operates 100 MW–1 GW plants co-located with industrial and datacenter campuses, selling power under long-term PPAs; ~$100M recurring revenue per plant claimed
Regulatory: No NRC docket. Not on the NRC’s advanced-reactor pre-application list. No DOE authorization. Not in the DOE Reactor Pilot Program. Stated intent to pursue an NRC Part 50 pathway (per the March 2026 release) — note the tension with the June release’s “design certification,” a Part 52 instrument
Customers / sites: None named. Only “first 10 GW of planned capacity” and “early customers.” The only concrete geography is a search for a New Orleans site for a non-nuclear testing and training simulator centre

What It Is / How It Works

mPower, briefly. The reactor is an integral PWR: the core, steam generators and pressurizer all sit inside a single pressure vessel, so there is no large-diameter primary coolant piping and therefore no large-break loss-of-coolant accident to design against. Fuel is standard sub-5% LEU in conventional assemblies, drawing on the existing commercial fuel supply chain rather than the HALEU or TRISO bottlenecks that constrain most companies in this section. At 195 MWe it is substantially larger than the microreactors in the 2026 criticality cohort and comparable in scale to NuScale’s multi-module plants. The technology is mature and thoroughly documented — B&W and Bechtel spent the better part of a decade on it.

Why it stopped, accurately. The history matters because Applied Atomics’ pitch is that a shelved design can be picked up cheaply:

2009 — B&W launches mPower; NRC pre-application interactions begin July 2009.
2010 — Generation mPower LLC formed: B&W ~90%, Bechtel ~10%.
By May 2012 — B&W had spent >$200M of its own money.
Nov 2012 — mPower wins the first-ever DOE SMR Licensing Technical Support award, beating Westinghouse, Holtec and NuScale: up to $226M over five years on a cost-share basis. Agreement signed April 2013 with a target of commercial demonstration by 2022.
2013 — B&W attempts to sell down its JV stake from 90% to ~20% and fails.
April 2014 — B&W shelves the program: annual spend cut to ~$15M/yr, 200+ layoffs. Stated reason: it could not secure additional investors or customer EPC commitments.
End 2014 — DOE halts payments. $111M of the $226M was actually disbursed and is not repayable.
March 3, 2017 — Bechtel notifies BWXT it cannot secure funding; the program is terminated and BWXT pays Bechtel a $30M settlement.

Total spend was roughly **$500M** (~$375–400M private plus $111M DOE), with B&W's accumulated losses by Q4 2014 around $400M and an estimated ~$650M more required to reach construction. Bechtel’s Fred deSousa summarized the failure precisely: the project “needed a plant owner with an identified location and an investor willing to wait a significant period of time for a return, and these were not available.” Shale gas had collapsed power prices, utilities were cautious post-Fukushima, and an unequal 90/10 JV between two large firms had no natural champion.

What changed, and what didn’t. The demand side has changed materially — datacenter load growth is exactly the “plant owner with an identified location” that mPower lacked, and the capital environment for nuclear in 2026 bears no resemblance to 2014. That is the real thesis, and it is not unreasonable. What has not changed is the cost of finishing the design and licensing it. Applied Atomics has ~$12M and fewer than ten people; the license leaves manufacturing and IP with BWXT; and the NRC pathway has to be started, not resumed.

A note on the carve-out. The license is specifically land-based. On the same day — June 17, 2026 — BWXT signed a separate feasibility study with Core Power to assess mPower for floating nuclear power plants. BWXT is monetizing the archive along two independent tracks while retaining the IP and the manufacturing.

Notable Developments

2026-06-17: Exclusive land-based mPower license from BWX Technologies announced. BWXT retains IP ownership, exclusive component manufacturing rights, and royalties. Applied Atomics will “re-engage the NRC” and develop site-specific engineering. (GlobeNewswire; Applied Atomics; ANS)
2026-06-17: Same day — BWXT signs a separate feasibility study with Core Power on mPower for floating nuclear power plants, the non-land-based counterpart to the Applied Atomics carve-out. (World Nuclear News)
2026-05: Kellie discusses seeking a New Orleans site for a non-nuclear testing and training simulator centre. (The Advocate)
2026-03-25: Seed round — $8.3M, co-led by Transition and Morpheus with Alpaca VC. Stated 12-month goals: finalize first host sites and customer agreements, activate engineering test stands, advance an NRC Part 50 licensing pathway. (GlobeNewswire)
2026: Joins the Launch Alaska 2026 portfolio, listed as Anchorage-based. (Launch Alaska)
2025-03: Pre-seed led by Alpaca VC. (Alpaca VC)

Key People

Ben Kellie — Co-Founder & CEO

LinkedIn: linkedin.com/in/benkellie
Background: Ex-SpaceX, where he led infrastructure work including the Falcon 9 droneship; founded The Launch Company, acquired by Voyager Space; BS Mechanical Engineering, University of Alaska Fairbanks; Entrepreneur in Residence at the University of Alaska
Notes: ⚑ No company statement explains the Alaska headquarters. The circumstantial explanation is founder residence plus the Launch Alaska accelerator — Kellie lives in Anchorage, and no Alaska project has been announced. Do not read the Alaska HQ as a market or siting strategy.

Paul Keutelian — CTO

LinkedIn: linkedin.com/in/paulkeutelian
Background: Ex-SpaceX technical operations and regulatory compliance; co-founder of Radiant Nuclear
Notes: ⚑ Overlap: Keutelian’s Radiant co-founding connects Applied Atomics directly to Radiant Industries, documented separately in this section — one of several SpaceX-alumni lineages running through the microreactor sector

Mark Lambert — Co-Founder & CFO

Background: 17+ years in startups and regulated-industry B2B SaaS

Other

Chris Newton and Leigh D’Angelo appear as founding team on the company website; backgrounds not disclosed.

People — Last Reviewed: 2026-08-27

Supply Chain Position

Layer	Detail
Reactor technology	Licensed from BWX Technologies; BWXT retains IP ownership
Manufacturing	BWXT holds exclusive manufacturing rights for all mPower components and royalty rights on any third-party-made components — Applied Atomics cannot manufacture independently
Fuel	Standard LEU (<5%) in conventional assemblies — no HALEU or TRISO exposure; supplier not named
Engineering support	Paid technical support contracted from BWXT during licensing
EPC	None named
Sites / customers	None named
Regulatory	No NRC docket; no DOE authorization

⚑ The license is narrower than the headline suggests. Applied Atomics has the right to deploy mPower on land; BWXT keeps the design, the factory and a royalty on every component. Applied Atomics is structurally a project developer and operator with an exclusive channel to a BWXT product, not a reactor vendor. That is a coherent business — it matches its stated IPP model — but it is not the same thing as owning an SMR design.

⚑ BWXT’s position strengthens either way. BWX Technologies already sits at the centre of this section’s TRISO chokepoint and won an Army Janus award at Fort Campbell. The mPower licensing structure adds a third revenue path — royalties and manufacturing on a design it had written off — with none of the deployment risk.

Claim Verification

Claim: Applied Atomics will “resume mPower design certification activities” / “re-engage the NRC”

Status: Materially misleading — there is no mPower design certification to resume

Supporting: The phrasing appears in the June 17, 2026 release and was reproduced verbatim by ANS. mPower did have genuine NRC pre-application interactions beginning July 2009.

Refuting / questioning: Per the NRC’s own archived mPower status page, the program was in pre-application review only, with a design certification application “expected 4th quarter of calendar year 2013” that was never filed. There is no docketed mPower design certification. Separately, the Clinch River Early Site Permit (applied May 2016, issued December 2019, ESP-006) was design-neutral — a Plant Parameter Envelope for “two or more SMR modules, up to 800 MWe,” with no vendor selected and no mention of mPower or BWXT; TVA is now pursuing GE Hitachi’s BWRX-300 there. And as of August 2026, Applied Atomics does not appear on the NRC’s current advanced-reactor pre-application activities list, which does include Hadron Energy and Terrestrial Energy USA.

Summary: Applied Atomics would be starting an NRC licensing process, not resuming one. The company’s own March 2026 release describes a Part 50 pathway while the June release says “design certification” (a Part 52 instrument) — the regulatory strategy is not stated consistently.

Claim: A ~36–48 month NRC licensing timeline using proven Gen III LWR technology

Status: Company-sourced aspiration with no precedent

Supporting: The argument that a mature LWR design should license faster than a Gen IV first-of-a-kind is reasonable in principle.

Refuting / questioning: No US SMR has achieved anything close. NuScale’s design certification took roughly seven years from application. X-energy’s Seadrift construction permit — a permit, not a licence — was docketed May 2025 with a final safety evaluation expected November 2026 and a decision after that. Applied Atomics has not filed anything.

Summary: Treat as marketing.

Claim: The company is pursuing “the first 10 GW of planned capacity”

Status: Aspirational; no customer, site, or agreement identified

Refuting / questioning: No customer, host site, PPA, or LOI has been named. Ten gigawatts at 195 MWe per unit implies roughly 51 reactors. The company has ~$12M and fewer than ten employees.

Summary: A statement of ambition, not a pipeline.

Claim: Headquarters location

Status: Unresolved / recently changed

Supporting: The June 2026 release states Anchorage, Alaska with a Los Angeles design studio; Crunchbase lists Anchorage; the company is in the Launch Alaska portfolio.

Refuting / questioning: The March 2026 release and May 2026 local coverage describe a Los Angeles (Arts District) company. Two domains are live — applied-atomics.com (older, with /about and /news now 404) and appliedatomics.com.

Summary: Most likely an HQ-of-record move to Alaska between March and June 2026, driven by founder residence and accelerator participation rather than operations.

Risk & Skepticism Notes

Capital-to-claims ratio is the central issue. ~$12M and ≤10 people against a design program that consumed ~$500M and was abandoned for lack of a further ~$650M. Even granting a transformed demand environment, the gap is one to two orders of magnitude.
No independent journalism exists. Every article traceable on this company originates from the GlobeNewswire releases; ANS’s coverage is a one-paragraph restatement. There is no technical or financial scrutiny in the record.
The license leaves the hard parts with BWXT. IP, manufacturing and royalties all stay on BWXT’s side; Applied Atomics carries the licensing cost, the site development, the financing and the operating risk.
No regulatory position at all. Not on the NRC pre-application list, no DOE authorization, no Reactor Pilot Program slot — the weakest regulatory standing of any company in this section.
Recorded here as one to watch, not one to count on. The mPower design is real and the founders have relevant hard-tech operating experience. Nothing beyond the license itself has yet been demonstrated.

The Infinite Unknown