Talk:Radiant Nuclear/GA1

GA review

Article (edit | visual edit | history) · Article talk (edit | history) · Watch

Nominator: H2so4aq (talk · contribs) 18:42, 22 September 2026 (UTC)

Reviewer: Operator 16 (talk · contribs) 14:18, 27 September 2026 (UTC)

This review is part of my pledge for the review at Talk:Letchworth-Love Mounds Archaeological State Park/GA1. 🌴Operator 16 (talk) 14:18, 27 September 2026 (UTC)

Thank you for the detailed review and comments! I'll get to fixing these; should be done this week. H2so4aq (talk) 18:44, 28 September 2026 (UTC)

@Operator 16: Thank you again for doing such a detailed review; you left a lot of great feedback and the article is looking a lot better. I'm ready for another look; I fixed the problems you identified and checked all the sources. Most fixes were done the way you suggested, but I would like you to look over a few of them (marked with  ). H2so4aq (talk) 03:35, 3 October 2026 (UTC)

GA review – see WP:WIAGA for criteria

---

This is a tidy, well-organized article on a young company, built quickly from industry press and clean on copyvio. I spot-checked 30 of the 38 cited sources against the article text. The main problems are as follows: (1) two of the load-bearing citations are LA Times Studios B2B content rather than LA Times newsroom journalism, one of them explicitly written from information Radiant provided; (2) a handful of places where the article states more than its source (the DOME "first", the Army award, Launch Pad support for R-50, the Wyoming critics); and (3) the lead stops before the Army contract, the funding history and the Wyoming controversy. All of it is fixable in a week hold period though so I think we can get this knocked out and listed as a Good Article! Also congratz on the DYK. Details under each criterion.

1. Is it well written?

   A. The prose is clear, concise and understandable to an appropriately broad audience, and the spelling and grammar are correct:

   Mostly clear. Things to fix before I'd pass it:

   • Tense drifts inside §Kaleidos reactor: uses / It will use integrated fans / is to be packaged / will be temporary. Pick one register (present tense for design features, future only for things that have not happened) and keep to it.

   Y Done.

   • Each particle is composed of a kernel of uranium oxycarbide coated with silicon carbide and carbon, and packed into a graphite compact: I looked this up and individual particles are not each packed into a compact. I'd suggest ...coated with layers of carbon and silicon carbide; the particles are then pressed into graphite compacts. but you're more read into this than I am so just something along that line.

   Y Yup that's correct, my phrasing was a bit awkward. Your suggestion looks good as-is.

   • Radiant selected Bar Nunn, Wyoming as their preferred site: the company is referred to with "its" everywhere else.

   Y Done.

   • significant opposition ... significant concerns in one sentence.

   Y significant concerns

   • It intends to replace diesel generators (lead): "It" can be read as the company or the reactor.

   Y Done.

   • dropped into place is informal; "installed" would be better.

   Y Done.

   • §History runs as one six-paragraph block and loses chronology: this Army/Centrus paragraph (August–September 2026) comes before the factory paragraph (2024–2026), and inside the Wyoming paragraph the Freedom Caucus opposition is described after the cancellation it caused. Suggest subheadings (e.g. Founding and funding / DOE demonstration / Military contracts / Factory) and put the Wyoming paragraph in order: shortlist → Bar Nunn selection → spent-fuel objections and Freedom Caucus campaign → withdrawal and stated reason → reaction.

     Done. Also moved the Equinix announcement into the ANPI paragraph, since it was announced about a week after the ANPI selection. Let me know if you think it should be moved elsewhere.

   • §Corporate affairs opens by repeating §History (who founded Radiant in 2020). Merge that sentence into the funding paragraph.

   Y Done.

   B. It complies with the manual of style guidelines for lead sections, layout, words to watch, fiction, list incorporation and table accessibility:

   Lead, infobox and MOS points:

   • Lead (MOS:LEAD). It does not summarize the Army Janus award (the largest contract the company has), the funding history (over $575 million raised), the Wyoming site controversy and move to Oak Ridge, or the Equinix order. One more paragraph would cover it; not too far off honestly.

   Y Done.

   • The lead says Buckley will be the company's first commercial unit; the body never says this, and the sources say "first Kaleidos reactors by 2028" (POWER) or give it as the company's aim (LA Times Studios). Either support it in the body with an independent source or drop the phrase "first commercial" here.

   Y Changed "first commercial unit" to "a commercial unit".

   • Lead/body mismatches: lead 1 MW vs body 1 MW_e; lead transportable by truck vs body truck, cargo aircraft or ship; lead designed to last 5 years without refueling vs body five years between refuelings with a 20-year life. "Last" is the wrong verb. Spell out "five" per MOS:NUMERAL; I always mess that one up so I've been lookin for it lately.

   Y Done.

   • Infobox. Key people Tori Shivanandan (president) and Mike Starrett (CRO) and founder Bob Urberger appear nowhere in the body. The body also says Bernauer founded the company with two other colleagues, which makes three founders, but the infobox lists two. Fix the two and give the officers a sentence or two in §Corporate affairs.

     Done. Couldn't find another source for Bob Urberger or the other founder so I changed to "with two colleagues". Cited Shivanandan as being both president and COO to a video from CNBC. The Los Angeles Business Journal does mention Urberger, but not any other founders. Not sure what to do here, since it says it was founded by Bernauer and Urberger, but other sources say Bernauer and two others but don't mention them by name.

   • MOS:DUPLINK: TRISO fuel is linked in both §History and §Kaleidos reactor, and so is spent nuclear fuel. Keep the first body link of each.

   Y Done.

   • MOS:EGG: [[K-25|K-27 and K-29 site]] sends the reader to a different building's article under another name. Unpipe it and explain, e.g. the former site of the K-27 and K-29 gaseous diffusion buildings, near the K-25 plant.

   Y Done.

   • MOS:SAID: decrying that → saying that.

   Y Done.

   • External links: the WyoFile news article should be a source or removed (WP:ELNO #1). It is also Earwig's top match (see 2d), so it may already be informing the Wyoming paragraph.

   Y Moved to body and added details from the article.

2. Is it verifiable with no original research, as shown by a source spot-check?

   A. It contains a list of all references (sources of information), presented in accordance with the layout style guideline:

   Reference list is present and properly laid out. Fix these citation errors while you are in there:

   • CSD_2024-03-26: the date is wrong. The URL and article are 26 March 2025, and the sentence it supports (the Bar Nunn selection) is 2025.

   Y Done.

   • E&E News: author is Camacho, not "Camancho".

   Y Done.

   • The Sundance Times piece is a Casper Star-Tribune story (Dylan Farrell and Allison Allsop) syndicated via the Wyoming News Exchange. Add via= or cite the Star-Tribune original. Its date is 16 October 2025, per the URL and the page, not the 15th.

   Y Done.

   • NYT and Business Insider: add url-access=subscription / limited as appropriate.

   Y Done.

   • Optional: the same outlets are cited with different templates (NEI and Nuclear Newswire as both {{cite web}} and {{cite magazine}}). GA does not require consistency, but it's a two-minute tidy.

   Y Done.

   B. Reliable sources are cited inline. All content that could reasonably be challenged, except for plot summaries and that which summarizes cited content elsewhere in the article, must be cited no later than the end of the paragraph (or line if the content is not in prose):

   Source reliability:

   • The two Los Angeles Times citations are not LA Times newsroom journalism. Both sit under latimes.com/b2b, in the "Business by LA Times Studios" section. The byline is David Nusbaum, described on the page as a contributing reporter and research consultant to LA Times Studios, and both pages carry sponsored-content flags in their metadata. The June 2026 piece ends Information for this story was provided by Radiant. Treat them as company-sourced, not independent:
     • LATimes_2026-06-12 (infobox footnote): remove; use CBS Colorado / POWER for the Air Force facts.

     This one was only used to source Mike Starrett as CRO, and Tori Shivanandan being appointed president, for the infobox. I'm not sure we need an independent source for this sort of basic information per WP:SELFSOURCE, but could seek out company press releases or an independent source if you think so. Content now cited in the article; no longer needed.

   • • LATimes_2025-07-20 is the sole source for: Bernauer trying to start a nuclear program inside SpaceX; the dance studio and Hughes warehouse; Bernauer self-funding; the Baranwal hire; about 100 employees (infobox); and the $6.50/gallon diesel break-even. That last one is currently attributed in-text to the Los Angeles Times, which it is not in any meaningful sense. It is Radiant's claim. Replace with independent sources where they exist (NEI covers the Series C; CBS Colorado names Baranwal as CNO); otherwise attribute to the company or cut.

     Attributed break-even price and Bernauer info to him or to the company, and replaced for Series C. Infobox and Baranwal info (specifically that she was hired in 2025) should be okay for the same reasons above (using the company as a source for basic company info), but could dig for other sources if you really think it should be replaced.

   • • Note also that this piece says Bernauer decided to create Radiant in 2019, against 2020 everywhere else. If you keep it, footnote the discrepancy; better to rely on POWER (Sept 2020) and Forbes.

     The first B2B source says Bernauer "decided to create Radiant in 2019", and Forbes says, "In 2019, the Department of Defense issued...he left SpaceX to found Radiant that same year". The second B2B source (the one using info "supplied by Radiant") says "Founded in 2020,...". I took this to mean he decided to create Radiant and left SpaceX in 2019, but incorporated the company in 2020. I'll trust the Forbes source here though, and change it to 2019.

   • The Cool Down (infobox footnote) is a consumer-green content site, not a suitable source for anything here. Remove.

   Y Done.

   • Interesting Engineering is cited five times. It is fine for uncontroversial technical details, but where a fact rests on IE alone (the Equinix order, the 12 August shipping date), pair it with WNN, NEI, POWER or ANS if you can.

   Y Done. Found Equinix press release and replaced August 12 ship date with WNN.

   • See 2c for places where the cited source is reliable but doesn't say what the article says.

   C. It contains no original research:

   Spot-check failures and overstatements (source quoted or paraphrased in each case):

   • The DOE later selected Radiant to be the first company to test at DOME. POWER (3 July 2025): Radiant and Westinghouse were jointly selected on 1 July 2025 to lead the first test campaigns. Redo this as a joint selection and give the date.

   Y Added info on joint selection. While the joint award with Westinghouse was the selection to lead the first campaigns, POWER_2026-04-30 explains that Radiant is the first company to test. I've now cited that for the "first".

   • The company was awarded $750 million, out of a total of $2.2 billion, to build 15 Kaleidos reactors at Benning. Defense News: The Army did not disclose the contract amounts; it was Radiant that disclosed an agreement "totaling $750 million for 15 Kaleidos nuclear microreactors". Nothing I read places all 15 at Benning, and the Army expects "more than 20" reactors across all five bases combined. CBS: the $2.2 billion is "up to", over five years, contingent on milestones. Suggest: Radiant said its agreement was worth $750 million for 15 Kaleidos units; the five Janus awards together are worth up to $2.2 billion over five years, contingent on performance milestones. This also removes the clash with at least one microreactor at Fort Benning in the previous sentence.

   Y Seems I misinterpreted the sources; nothing says all 15 are at Benning. The NYT source does seem to confirm the $750 million award, ("'Operating on military installations first gives us the reliability track record to unlock commercial deployments, and the resulting, much larger commercial volume drives down costs,' said Mike Starrett, chief commercial officer at Radiant, a start-up that was awarded $750 million to build up to 15 of its 1-megawatt microreactors."), although this also might be really quoting the company so I've attributed it to them and reworded.

   • R-50 is also supported by the DOE's Nuclear Energy Launch Pad initiative, which Radiant was selected for in April 2026. POWER (28 April 2026) reports Radiant's selection for Launch Pad on the strength of its Kaleidos/DOME program; it mentions R-50 only as a separate facility "under development". I couldn't access the Reuters piece; if it says Launch Pad supports R-50 specifically, please quote it here. Otherwise move the Launch Pad sentence to the DOE paragraph.

     Reuters quotes Baranwal: Launch Pad will provide authorization to build, fuel, and test Radiant’s first Kaleidos reactors at its new Oak Ridge factory in Tennessee, Baranwal said. I might be misinterpreting this source though; as it doesn't quite say the factory itself is "supported". Let me know if you think I should simply state that Radiant itself was selected for Launch Pad.

   • provoked criticism from Bar Nunn mayor Peter Boyer, as well as Wyoming governor Mark Gordon and several state legislators. The cited Cowboy State Daily piece (14 October 2025) has Gordon (the "Club No" remark) and one legislator, Sen. Ed Cooper. I found no mention of Boyer, and it does not support "several legislators". Source Boyer elsewhere or remove him, and name Cooper.

     The cited piece names two legislators, Sen. Ed Cooper and Rep. Steve Harshman as criticizing the decision. Seems I forgot to cite the details for Boyer, that's from CSD_2025-10-13.

   • examining Wyoming, Utah, Tennessee, and Idaho: Cowboy State Daily (15 March 2024) lists five finalists and includes Montana.

   Y Done.

   • In 2023, Radiant was awarded a grant by the US Department of Energy: POWER describes this as front-end engineering and experiment design (FEEED) funding, $3.9 million shared among three developers, to design the experiment. Say that; "grant" understates what it was for.

   Y Reworded.

   • In 2023, the company raised $40 million in a funding round: Forbes calls it a Series B. Say so, since the next sentence names the Series C.

   Y Done.

   • By mid-2025, Radiant had raised $165 million in a Series C funding round. Later that year, it hired Rita Baranwal: the cited July 2025 source already reports her as CNO in the same breath as the Series C, so "later that year" misorders them. Rephrase the Series C with NEI's facts: closed June 2025, led by DCVC, bringing total venture funding to $225 million.

   Y Done.

   • The reactor uses passive cooling ... combined with TRISO fuel to achieve a high degree of nuclear safety. In Forbes this is an interviewee's assessment ("she also says"), not a finding. Attribute it. Forbes also quotes Todd Allen (Michigan) comparing the risk to university research reactors, which would give you an independent voice.

   Y Reworded.

   • In August 2025, the US Air Force signed an agreement: E&E says the contract is with the Air Force and the Defense Innovation Unit. Add DIU.

   Y Done.

   • The Equinix sentence has no date; the IE source does not give one. Find the announcement and date it.

   Y August 2025 according to World Nuclear News.

   D. It contains no copyright violations nor plagiarism:

   Earwig's highest match was 18.7% (WyoFile, which sits in External links rather than References), then Cowboy State Daily at 13.8% and 11.5%. The overlap is names, titles and the Rodriguez-Williams quotation. No concerns. Earwig cannot see the paywalled sources (E&E, NYT, Oakridger, Axios), and I did not spot-check those for paraphrase either.

3. Is it broad in its coverage?

   A. It addresses the main aspects of the topic:

   Covers history, product and business reasonably, but three gaps are main aspects for a nuclear company:

   • Commercial licensing pathway. The DOME test runs under DOE authorization; commercial units will need NRC licensing. The body is silent on this, although an NRC pre-application page sits in External links. One or two sentences on where Kaleidos stands with the NRC.

   Y Done. Added detail about licensing for KDU, ANPI, and Janus, since it appears all three use different pathways (DOE, NRC, and DOD).

   • Series D. The Los Angeles Business Journal (already cited) reports a Series D of "more than $300 million" announced in December 2025. That round is most of the $575 million total, but the article skips from Series C to the total.

   Y Done.

   • Fuel supply. The fuel for the demonstration was fabricated by Standard Nuclear (WNN, 3 July 2026). IE (20 August 2026, already cited) reports a Standard Nuclear agreement to fabricate Kaleidos TRISO fuel into the early 2030s. Together with the Centrus HALEU contract, that is the company's fuel chain, and it deserves a sentence.

   Y Done.

   • Currency, not a blocker: the KDU test program was targeted for Q3 2026. I could find no report of criticality as of 27 September 2026. If it happens during the review, add it.

   Y Unfortunately not critical yet, so nothing to add.

   B. It stays focused on the topic without going into unnecessary detail (see summary style):

   Stays focused; good work. The Wyoming episode gets a long paragraph, but it is the most substantial independent coverage of the company, and it is proportionate once tightened after 1a.

4. Is it neutral?

   It represents viewpoints fairly and without editorial bias, giving due weight to each:

   Wording and balance:

   • Several sentences carry the company's framing in wikivoice: to achieve a high degree of nuclear safety (see 2c), allowing for highly compact design, dropped into place with minimal setup. Neutral wording: "which Radiant says...", "compact", "installed".

   Y Done.

   • Because much of the technical and business description comes from company-supplied material (see 2b), the article presents the company's plans (5-year cycles, 48-hour setup, spent fuel "entirely within the container") without independent comment. Where the sources frame something as a design goal or company statement, the article should too. §Kaleidos para 3 already does this (The company has stated...), and that is the model.

   Y Done.

   • Suggestion, not a blocker: if reliable sources discuss the economics, fuel security or waste questions for microreactors like Kaleidos, a short sentence of independent assessment would help balance. The Wyoming spent-fuel dispute is currently the only non-company perspective.

     I'm very hesitant to add statements talking about microreactors in general, unless they mention Kaleidos/Radiant specifically (or are in sources discussing them), due to the potential for WP:SYNTH. The Wyoming sources do quote from Prof. Alan Kuperman at UT-Austin (The vast majority of companies like Radiant that are proposing to build advanced nuclear reactors are going to go out of business before building a single reactor) when discussing the Bar Nunn dispute, but adding this seems a bit out-of-place given KDU's shipment, ANPI, Janus, and Equinix. I'm undecided but if you think it should be included then I'll add it in.

5. Is it stable?

   It does not change significantly from day to day because of an ongoing edit war or content dispute:

   Created 23–24 August 2026 (moved from draft) and edited almost solely by the nominator since, with one outside copyedit. No disputes or reverts. The subject is fast-moving and new.

6. Is it illustrated, if possible, by images?

   A. Images are tagged with their copyright status, and valid non-free use rationales are provided for non-free content:

   Only image is the non-free logo, which has a complete {{Non-free use rationale 2}} and {{Non-free logo}} tag. The "non-free files for NFUR review" category it sits in is routine: it clears when a file reviewer adds image_has_rationale=yes. Not a concern.

   B. Images are relevant to the topic, have suitable captions, and have alt text:

   Logo is appropriate in the infobox. No image is required for GA, but a freely licensed image of the Kaleidos unit, the DOME, or the El Segundo facility would help readers. A photo taken by DOE staff would be PD-USGov.

7. Overall:

   Pass or Fail:

   On hold for seven days. The article is close. The fixes are specific and the article is short, but please treat the 2c list as a sample, not an exhaustive audit: I checked 30 of 38 sources and found around ten places where the text says more than, or other than, the source. Before you ping me, re-read every sentence against its citation, especially in the funding and contracts material.

   In priority order:

   1. 2b: replace or attribute every claim resting on the LA Times Studios pieces, and remove The Cool Down.
   2. 2c: the DOME "first", the Army $750M / 15 units / Benning sentence, the Launch Pad–R-50 link, Boyer and "several legislators", Montana, FEEED, Series B/C details, and attribution of the safety claim.
   3. 1b: expand the lead to cover the Army award, funding, Wyoming/Oak Ridge and Equinix; reconcile the lead and infobox with the body (founders, officers, "first commercial unit", 1 MW_e, five-year refueling cycle).
   4. 3a: add a sentence each on the NRC commercial pathway, the Series D, and Standard Nuclear fuel.
   5. 4 and 1a: de-promote the wikivoice phrasing; fix tense; reorder and subdivide §History.
   6. 2a housekeeping: the date, author and via fixes.

   Good work on the page overall! I like the new developments in nuclear so this was an interesting read. I'm happy to discuss any of these; push back if you think I've misread a source.