DJ-2026-09-28Monday, September 28, 2026OSINT // UNCLASSIFIED
Blackgrove Global Risk · client intelligence product
The Daily Journal · DJ-2026-09-28 · Monday 28 September 2026 · Revised edition · Articles, not digests
Revised edition · New lead analysis · Full articles · Grades = sourcing postureBBN-NA-2026-009 · 008 · 008b
LATE WIRE New lead: heat rejection and thrust have contracts; net energy only bench-scale · PUMP is undersea · Sedwick patent allowed, grant number unconfirmed · Senate back 3:00 p.m.; S.4668 vote 5:30 · Release 06: 71 ours vs 75 secondary.
The cooling and the thrust are being paid for. The power plant is not.
Our review of public records finds real contracts behind superconducting magnets and electromagnetic drives—a DARPA seawater thruster, a NASA-funded magnet design, a magnet test on the space station. The fusion source that would let such a vehicle power itself exists on paper: one allowed patent, with no program found behind it.
Blackgrove Daily News · 28 September 2026 · DOCUMENTED USASpending / SBIR.gov / CORDIS / NASA NTRS / arXiv / patent record · DOCUMENTED-AS-REPORTED company and institute releases · Pulsar · AIAA paper · job posting · UNESTABLISHED patent grant number · Air Force intent
Editorial illustration — not evidence
Before anyone funds or briefs a vehicle that makes its own power from fusion, turns it straight into electricity and pushes with magnets instead of propellant, three engineering loops must close. Heat: the machine must shed its waste heat—hard in vacuum, harder in a hot hypersonic airframe. Energy: the power source must return more electricity than it takes to run. Momentum: electricity must become thrust, by pushing on plasma, seawater or air with magnetic force. As of this morning, the public record shows real money in the first and third. The second, which decides whether the vehicle can supply its own power, is worked at bench scale and on paper, with no funded vehicle behind it.
Two of the three loops have contracts behind them. The loop that decides whether the vehicle can power itself has a patent.
The parts, in plain terms
A high-temperature superconductor (HTS) carries current without resistance at temperatures still cryogenic but far warmer than older superconductors need, allowing strong, light magnets if they can be kept cold. A magnetohydrodynamic (MHD) device uses a magnetic field on a conducting fluid—plasma or seawater—to push it (a drive) or draw electricity from it (a generator), with no propeller or turbine. Aneutronic fusion fuels such as proton-boron-11 (p-B11) or deuterium-helium-3 (D-He3) release most of their energy as charged particles, which fields can in principle turn straight into electricity: direct conversion.
Where the money is: magnets that push
The best-funded work is under water. The Defense Advanced Research Projects Agency’s PUMP program (Principles of Undersea Magnetohydrodynamic Pumps) pays to drive seawater with magnetic force and no moving parts. Federal spending records show more than $12 million in contracts, by our sum, led by HRL Laboratories ($4.0 million), Tai-Yang Research ($2.5 million) and SRI International ($2.0 million). General Atomics is designing the superconducting magnets and in October 2025 subcontracted Tokamak Energy to build them from REBCO, a rare-earth superconducting tape. PUMP is not an aerospace program, but its magnet, cryogenics and electrodes are what an air or space electromagnetic drive would need.
In space, superconducting magnets are moving into electric thrusters. On 29 September 2025 NASA gave Ad Astra Rocket Company a $4.0 million Small Business Innovation Research (SBIR) award to mature VASIMR (the Variable Specific Impulse Magnetoplasma Rocket), including a critical design—not yet hardware—of an HTS magnet to replace its older one and a high-temperature heat-rejection architecture. New Zealand’s Robinson Research Institute is developing an applied-field magnetoplasmadynamic (AF-MPD) thruster, which accelerates plasma with current and an external magnetic field supplied by an HTS magnet above 1 tesla. Its precursor, Hēki, an HTS magnet, flew to the International Space Station (ISS) in September 2025; the institute reported in January that it met its primary-mission goals and cycled to 500 millitesla. We have seen no result from a planned quench test, which deliberately knocks the magnet out of its superconducting state, and Hēki is a magnet, not a thruster. Europe’s SUPREME, a €2.5 million European Innovation Council grant led by the University of Stuttgart, worked on an HTS AF-MPD thruster until it closed in November 2025.
Where it isn’t: the power plant
The most complete design we found is a U.S. patent application (US20240347218A1) by Raymond Sedwick and four co-inventors from the University of Maryland and UMBC centrifugal-mirror fusion group. It claims a spinning-plasma fusion reactor open to space, with HTS coils and fuels including D-He3 and p-B11; direct conversion of charged fusion products into electricity; thermionic capture of the X-rays the plasma sheds (bremsstrahlung) in a shroud radiating at high temperature, “eliminating the need for any additional radiators”; and a magnetic nozzle for thrust. It is the only live filing in our review that addresses all three loops. The patent office mailed a notice of allowance on 23 March 2026 and the issue fee was paid 27 April, so a patent has probably issued; we have not confirmed a grant number. The underlying experiment was awarded $5.18 million in 2020 by the Advanced Research Projects Agency–Energy (ARPA-E), for confinement physics. We found no funding for the propulsion system the patent describes. (The only other filings touching every piece, two Navy patents by Salvatore Pais, are dead; the Navy’s own test reportedly could not prove the effect.)
Below the patent, funded work is small or dated. The only U.S. award we found pairing fusion propulsion, HTS and a magnetic nozzle is Princeton Satellite Systems’ $749,981 NASA award from 2018. The University of Alabama in Huntsville holds a $649,845 NASA grant for magnetic nozzles that make thrust while recovering electricity—using pulsed plasma, not fusion fuel. Air Force fusion-propulsion SBIRs run from $49,999 to $749,685. Pulsar Fusion announced first plasma in its Sunbird drive on 25 March; we have seen that only in a press release, it involves no fusion yet, and we found no government award. Beneath the money gap sits a physics gap: a p-B11 plasma radiates bremsstrahlung faster than fusion replaces the energy unless electrons are held cooler than ions or the ions kept out of thermal balance. Papers from 2024 to 2026 propose routes; none demonstrates net energy.
Defense: integration signals, and one inference
PUMP is the clearest case of systems integration, with contracts, a named magnet builder and an independent test team. The Air Force record is thinner. A January 2026 American Institute of Aeronautics and Astronautics (AIAA) paper co-written by three Air Force Research Laboratory (AFRL) authors sizes superconducting magnets for aerospace MHD power generation. That suggests live Air Force interest in power without rotating machinery; that is our inference, and we found no award record behind it. An AFRL request for information posted in July on power and thermal management for reusable hypersonic aircraft does not mention MHD. A NASA-funded study finds a large superconducting magnet could produce MHD force within an order of magnitude of aerodynamic drag above 10 kilometers per second on planetary entry—analysis, not hardware. We found no job posting combining a clearance, plasma physics, superconducting magnets and aerospace integration; the nearest, a cleared Lockheed Martin plasma physics role in Palmdale, California, does not mention superconductors. For active hulls—metamaterial skins or superconducting walls that push on their surroundings—we found no credible funded program.
How to read these records
Titles and patents are not proof of working hardware. An allowed patent is not a funded program. A press release is not a contract record. Where we connect dots, we call it inference. By those rules, nothing in our review closes all three loops with money behind it.
Blackgrove assessment
The convergence is real for parts and pairs, not for a vehicle. The heat and momentum loops are being engineered with real money, mostly for space thrusters and undersea drives rather than aircraft. The energy loop is not: its funding is bench-scale grants under $1 million and ground fusion research, its fullest vehicle design is a patent with no program found behind it, and its preferred fuel’s radiation losses remain unsolved. For policy, that marks where missing money would have to go; for markets, the common layer is the superconducting magnet and its cooling, where DARPA, NASA, European and New Zealand money overlaps; for defense, the hardware money is under water. Anyone claiming a closed fusion-electric-electromagnetic vehicle is ahead of the public record. Effect size: moderate for the magnet and cryogenics supply chain; low for near-term vehicles.
What we could not verify
Whether the patent has issued, and its number; Hēki’s quench result; Pulsar beyond its press release; reports that a SUPREME partner, Neutron Star Systems, is in liquidation; the Air Force’s intent behind the AFRL paper; RocketStar’s claimed thrust gain. The federal contracting portal and the Defense Technical Information Center could not be read; one reported cleared hiring lead could not be confirmed and is not used.
What to watch
Patent office: a grant number for US20240347218A1—and, separately, any funding for its propulsion design. Robinson Institute: Hēki’s quench result. PUMP: magnet deliveries and any aerospace spin-off. AFRL: a solicitation naming MHD or superconducting power. 2027: Pulsar’s and Avalanche Energy’s planned orbital demonstrations—watch for direct conversion or HTS on flight hardware. Falsifier: one funded award naming a fusion source, direct conversion and an electromagnetic thruster in a single vehicle would overturn this assessment.
Sourcing.DOCUMENTED — USASpending: PUMP contracts N6600124C4018 (HRL, $4,008,815), N6600124C4016 (Tai-Yang, $2,500,000), N6600124C4017 (SRI, $2,014,522), HR001124C0478 (Diversified Technologies, $549,003), N0002424F8352 and N0002423F8408 (Penn State ARL, $2,566,899; $676,000); UAH 80NSSC23K0232 ($649,845); ARPA-E DEAR0001270 (UMBC, $5,177,342.82, 2020). SBIR.gov: Ad Astra 80NSSC25C0436 ($3,999,650, 29 Sep 2025); Princeton Satellite Systems 80NSSC18C0218 ($749,981, 2018); RocketStar FA8649-22-P-0401 and -1117 ($49,999; $749,685). CORDIS 101113107 (SUPREME, €2,499,995.75). Patent record US20240347218A1 (claims; allowance 23 Mar 2026; issue fee 27 Apr 2026). arXiv 2509.10558; p-B11 papers 2405.13260, 2502.13300, 2511.10885. NASA NTRS 20240015387. DOCUMENTED-AS-REPORTED — General Atomics–Tokamak Energy magnet subcontract (company release); Hēki results (Robinson Institute blog, 16 Jan 2026); Pulsar Sunbird (press release, seen in search results only); AIAA paper (abstract via index); AFRL request for information RFI-RAP-26-01, 15 Jul 2026 (contracts mirror; federal portal unreadable); Lockheed Martin posting (job board); Pais patent status and Navy test (patent records; The War Zone); Avalanche orbital target (Defense Innovation Unit). TESTIMONIAL — none; company performance claims are printed as claims. UNESTABLISHED — the patent’s grant number; Air Force funding or intent behind the AFRL paper (inference); Neutron Star Systems liquidation; Hēki quench outcome; RocketStar thrust gain; net energy from p-B11; working hardware for any all-loop design. PUMP total is Blackgrove arithmetic. Don’t claim: the patent as a program; PUMP as aerospace; Hēki as a thruster; Pulsar as fusion; the AFRL paper or request as a contract.
Plenty of dates this week. None of them belong to the clocks that matter.
The Senate returns this afternoon to vote on a college-sports bill; the patent gazette owes its next issue; a counter-drone contract starts Wednesday. The decisions that govern drone permissions and UAP disclosure—80 days at the White House regulatory office, 96 days without a recorded action on the disclosure amendment—carry no public due date at all.
Blackgrove Daily News · 28 September 2026 · DOCUMENTED Senate Daily Press / congress.gov / House Clerk / reginfo.gov / Federal Register / USPTO / USASpending · DOCUMENTED-AS-REPORTED Release 06 · waiver · SAM award status · Falcon Peak · Hormuz
Illustration — editorial · not a photograph · not evidence
If you are deciding this week whether to commit money to drones flown beyond the pilot’s line of sight, whether to budget for keeping drones off a facility you own, or what to tell a client about the government’s release of files on unidentified anomalous phenomena (UAP), the useful Monday question is not “what is scheduled?” It is “which of my clocks has a date on it?” This morning the answer is none. The week is full of dated events—a Senate vote, a House session, a gazette issue, a contract start—and every one is a calendar event. A calendar records that an institution met. A clock records that a decision advanced.
Many calendar events this week. No scheduled movement on a single clock that prices this sector.
What is actually on the calendar
Today. The Senate reconvenes at 3:00 p.m. with one printed vote at 5:30: passage of S.4668, the Protect College Sports Act, as amended, on which it invoked cloture 74–25 on Thursday. The House convenes at noon. Nothing printed carries defense-authorization (NDAA) language or UAP text. Floor activity is not disclosure.
Tuesday and Wednesday. The U.S. Patent and Trademark Office’s weekly Official Gazette still ends at Week 38, dated 22 September (Vol. 1550, No. 4); the next issue would carry 29 September. On Wednesday, the federal spending database USASpending lists a counter-drone (“C-UXS”) task order to MITRE, valued at $2.68 million and approved 22 September, with a start date of 30 September—an existing contract reaching its start date, not new policy money.
Every weekday at 08:45 ET. The Federal Register’s Public Inspection desk, where rules appear the day before formal publication, opens its daily cycle; it was dark overnight. It is where a final drone rule would surface first—a place to look, not a due date. The House Clerk lists sessions through 24 September but roll calls only to 16 September; we will not infer the missing votes.
Illustration — editorial · not a photograph · not evidence
The clocks with no date
Each count starts from the dated public record named in the line. The arithmetic is ours; the start dates are not.
Disclosure amendment: 96 days. Senate Amendment 6344, the UAP statutory-text amendment Blackgrove tracks, was submitted 24 June 2026; its Actions page still reads zero—more than three times the 30-day mark at which we call language stalled. Its possible vehicles are as still: S.4784’s latest action is a 27 July motion to proceed (63 days); H.R.8800’s is “Received in the Senate,” 14 September (14 days). Check: congress.gov.
Flying beyond sight—Part 108: 80 days. The Federal Aviation Administration’s (FAA) final rule normalizing beyond-visual-line-of-sight (BVLOS) drone operations, RIN 2120-AL82, has sat at the White House Office of Information and Regulatory Affairs (OIRA) since 10 July, “Pending EO 12866 Regulatory Review.” Its docket, FAA-2025-1908, holds four proposed-rule documents, the latest from 10 February. Final rules published 1–28 September matching Part 108, Part 74 or BVLOS: zero. Read the docket number before the headline. Check: reginfo.gov (rrid 1457213); federalregister.gov.
Drones off fixed sites—Part 74: 54 days. Docket FAA-2026-4558 is still a notice of proposed rulemaking (NPRM): published 6 May, extended 30 June, comments closed 5 August. For facility owners, this is the clock on formal site restrictions, and it has not moved. Check: federalregister.gov.
Records-center contract: 27 days inactive. A sole-source notice of intent, NUFOHRC_20260731, for the Pentagon’s All-domain Anomaly Resolution Office (AARO) to buy a data subscription from the National UFO Historic Records Center went inactive on 1 September with no award notice, per secondary reads of the System for Award Management (SAM). USASpending shows no award dollars to the center. A notice of intent is not a contract. Check: SAM.gov; USASpending.
Waiver and tranche: 14 and 10 days. The Department of War’s legal waiver, Article/4600020, dates from 14 September, with no public follow-on and no named designee from the Office of the Director of National Intelligence (ODNI). Release 06 of PURSUE, the department’s public UAP-records program, posted 18 September; no Release 07 is visible from sources that answer us. The second article takes up both.
Not this week. Expect 30 September to be confused with the expiry of federal counter-drone authority. Under 6 U.S.C. 124n as amended by P.L. 119-60, that date is 30 September 2031—a fixed statutory date, not a live clock.
The one real event in the airspace lane
A press-wire release dated 27 September, citing the Defense Visual Information Distribution Service (DVIDS), reports that Joint Interagency Task Force 401 and North American Aerospace Defense Command–U.S. Northern Command (NORAD-NORTHCOM) concluded Falcon Peak 26.2, counter-small-drone experimentation at Yuma Proving Ground, and established enduring counter-drone test lanes there. The originals could not be opened, so it is graded as reported. It is capacity—a standing place to test drone defenses—not new legal authority (no amendment to 6 U.S.C. 124n or its rules posted), a rule or a grant. Nor is it UAP news. Blackgrove defines Unattributed Airspace Exposure (UAX) as “the operational, legal, and financial consequence to an organisation of objects operating in or near its airspace that cannot be attributed to a responsible operator within actionable time.” UAX is not UAP.
Where the windows are shut
A Monday “nothing moved” is only as good as its windows. The Department of War’s UFO portal, the waiver page and AARO’s site refused our requests (HTTP 403, access denied), as did the Federal Emergency Management Agency’s counter-drone grant page and the Securities and Exchange Commission’s filing indexes for 25–28 September; the DVIDS and NORTHCOM originals could not be reached. Structural is not empty: a refusal or timeout is not evidence that nothing is there. USASpending shows the difference. Late last week it errored and told us nothing; this morning it answered, and “no records-center award” is a real empty. One is a wall; the other is a result.
Blackgrove assessment
Dated events fill the first three days of the week; none of the clocks that price this sector has a scheduled trigger. Keep three numbers in view: 80 days at OIRA if you fly drones, 54 days since Part 74 comments closed if you own sites, and 96 days of zero actions on Amendment 6344 if you watch disclosure. None is a forecast; all are checkable. Effect size: high for briefing discipline, near zero for new rule, money or custody claims. For bandwidth desks: Middle East Eye, citing Reuters, reported on 25 September that nine commodity vessels crossed the Strait of Hormuz the previous day, down from 14 and below a roughly 18-vessel ten-day average.
What to watch
Today: S.4668 passage at 5:30—confirm no NDAA or UAP language is added live. Daily after 08:45 ET: Public Inspection for a Part 108 or 74 final; reginfo.gov for RIN 2120-AL82 leaving “Pending.” Tuesday: Gazette Week 39. Wednesday: the task order’s start date. Any day: a first action on S.Amdt.6344; House roll calls after 16 September; a records-center award; an ODNI designee. Falsifier: a claim that a decision clock moved, without the primary record, is rumor until the record posts.
Sourcing.DOCUMENTED — Senate Daily Press (28 Sep schedule; Thu cloture 74–25); congress.gov S.Amdt.6344 Actions (0) since 24 Jun, S.4784 (27 Jul), H.R.8800 (14 Sep); House Clerk (sessions through 24 Sep; roll index to 16 Sep); reginfo.gov rrid=1457213 (received 10 Jul); federalregister.gov FAA-2025-1908, FAA-2026-4558, final-rule searches 1–28 Sep; USPTO Gazette (Week 38); USASpending (no records-center award; task order start 30 Sep); access status 28 Sep. DOCUMENTED-AS-REPORTED — Release 06 and waiver status (primary refused); SAM NUFOHRC_20260731 inactive, no award; Falcon Peak 26.2 (EIN Presswire, 27 Sep, citing DVIDS); Hormuz (Middle East Eye citing Reuters). TESTIMONIAL — none. UNESTABLISHED — NDAA/UAP floor action; any Release 07; OIRA timing. Day counts are Blackgrove arithmetic to 28 Sep from the start dates shown. Don’t claim: votes as disclosure; Part 108/74 “about to clear”; records center awarded; Falcon Peak as UAP news; refusals as empties.
News Analysis · Custody · BBN-NA-2026-008b
71 or 75? How a count drifts when the primary record is behind a locked door.
Blackgrove’s read of the Pentagon’s sixth UAP release lists 71 titles. The secondary ledger we can reach now says 75—up from 72 last week, for the same batch. The portal that could settle it still refuses us. When the primary is walled off, the record becomes whoever counted last.
Blackgrove Daily News · 28 September 2026 · DOCUMENTED Blackgrove title inventory · DVIDS PR152 · access status · DOCUMENTED-AS-REPORTED portal state · UAP Ledger · waiver terms
Illustration — editorial · not a photograph · not evidence
Anyone who writes “the Pentagon released X files” this week—in a client note or a pitch deck—is citing a number with a source and a date, whether they say so or not. For Release 06 of PURSUE, the Department of War’s public UAP-records program, we read the government’s page by browser on release day, 18 September; every check since has been answered “access denied.” In that vacuum, numbers move. Ours has not. The one we can compare it with has moved twice.
Two numbers, one batch
Our number comes from that read of war.gov/UFO/?release=06, stamped “cleared for release Sept. 18, 2026”: 71 titles, each tagged by type—55 documents, 15 videos, one audio recording. The other comes from UAP Ledger, an independent tracker. Last week its Release 06 page listed 72 files (56 documents, 15 videos, one audio). This morning it lists 75: 59 documents, 15 videos, one audio. It is the same batch, not a new release: the higher figure sits on the Release 06 page, and the ledger’s Release 7 page still returns “not found.” Until the primary settles it, we print both: 71 is ours, 75 is the secondary’s, and they are unreconciled.
71 is ours. 75 is the secondary’s. The page that could settle it answers “access denied.”
Where the gap lives—and what we cannot explain
Side by side, the breakdowns locate the difference. Videos: 15 and 15. Audio: one and one. All four extra files are documents—our 55 against the ledger’s 59. So the gap is not the usual culprit, video and audio counted differently; there the two agree. Our list shows no internal hole: its documents are an unbroken numbered run, D102 through D154, plus two unresolved-report files issued as documents. Ours is a list of titles and the ledger’s a count of files, but whether that difference explains any of the four, we cannot tell. Nor can we tell why the ledger’s document count rose from 56 to 59 in a week, or whether the portal changed after our read. The reason is unknown, and we will not supply one. The drift spreads, too: our read showed an archive of 446; at 75, it would be about 450—our arithmetic, unverifiable while the portal is shut.
Why counts drift behind a wall
When the primary is reachable, a count is a lookup. When it is walled off, a count becomes a chain of custody: who read the page, when, by what rule. Three figures now circulate for one release—71, 72 and 75—and none can be checked against the page it describes. The danger is a secondary becoming the record by default: numbers cited without dates, then treated as the government’s own. A tracker that revises is behaving honestly. The reader who cited last week’s 72 as “the Pentagon’s count” is now wrong without having done anything, because the citation never said whose count it was, or when.
What the record can still prove
Strip out the count and a core remains. One primary page is reachable: DVIDS video DOW-UAP-PR152, an unresolved UAP report from the Yellow Sea in 2023, time-stamped 09.18.2026 07:15 and live this morning. Our list shows the release is mostly paperwork: 44 titles from the Advanced Aerospace Weapon System Applications Program (AAWSAP), 2008–2011, including its contract paperwork and 37 Defense Intelligence Reference Document (DIRD) study titles, beside 16 unresolved-report files and 1950s Blue Book–era material. A study headed “antigravity for aerospace applications” shows someone commissioned a paper on the subject, not that the effect exists. Titles are not proof of physics.
The channel built to be quiet
Article/4600020, announced 14 September, is, as reported, a targeted legal waiver letting current and former Department of War personnel disclose UAP-related national defense information to official PURSUE representatives, with enforcement of their nondisclosure and special access program agreements (NDA/SAPIA) set aside for that channel only. A waiver is not a public dump; it is an intake channel, not a tranche. Fourteen days on there is no follow-on, no named representative, no ODNI designee—the silence the design predicts, proving neither failure nor success. Likewise “no Release 07” rests on the ledger’s 404 and a quiet DVIDS while the primary is walled: none visible, not none. And Release 05 (about 7 August) to Release 06 was 42 days. Two points make a gap, not a cadence.
Blackgrove assessment
The gap is trivial as news and serious as method: four documents, cause unknown, unsettled until someone reads the primary again. Cite with source and date—“71 titles (Blackgrove read of war.gov, 18 September)” or “75 files (UAP Ledger, 28 September)”—and never “the Pentagon released 75,” or 71, as if the portal had been checked this week. Effect size: low for disclosure substance; high for anyone whose work carries a count.
What to watch
Portal: a file-by-file browser read of war.gov/UFO/?release=06, including whether any documents fall outside D102–D154. Ledger: further revisions and the identifiers behind its 59. DVIDS: any new DOW-UAP posting. Waiver: named representatives or an ODNI designee. Falsifier: a primary count of 75 means our inventory missed four documents, and we will say so; 71 means the secondary erred.
Sourcing.DOCUMENTED — Blackgrove browser inventory of war.gov/UFO/?release=06, 18 Sep (71 titles: 55/15/1; documents D102–D154 plus PR130–PR131; archive 446)—titles only; DVIDS DOW-UAP-PR152 (posted 09.18.2026 07:15; live 28 Sep); access status 28 Sep. DOCUMENTED-AS-REPORTED — current portal state; UAP Ledger Release 06 (75: 59/15/1 on 28 Sep; 72: 56/15/1 last week; Release 7 page 404); waiver terms (newsroom mirrors). TESTIMONIAL — none. UNESTABLISHED — reason for the 71/75 gap; any Release 07; waiver intake; physics from titles. Arithmetic (42 days; ~450) is Blackgrove’s. Don’t claim: 71 or 75 as the primary count; titles as physics; waiver as dump.
