Quiet Breakout: A US Company Will Mine the Pacific Floor in 2027. Almost None of the Patents Are American.

On May 1, NOAA quietly told TMC USA that its application to dig nickel and cobalt out of the seabed four kilometers under the Pacific was “in full compliance” with the Deep Seabed Hard Mineral Resources Act, a 1980 statute most people in Washington forgot existed. The company expects the rest of the process (environmental impact statement, public comment, final NOAA sign-off) to wrap before the end of the first quarter of 2027. If it does, the first US-permitted industrial mine on the abyssal plain will go to sea less than two years after Donald Trump signed Executive Order 14285, ending a four-decade industrial pause the United States, technically, never joined.

The political win has been loud. The engineering win has not. Run the patents and you find something strange: the machine that will pull 3 million wet tonnes a year of polymetallic nodules out of the Clarion-Clipperton Zone is almost entirely a foreign invention. A search of the 9.3 million US utility grants issued since 2020 for patents specifically claiming nodule mining, seabed mining vehicles, deep-sea ore lifting, sediment plumes, or seafloor harvesting returns 21 documents. Strip out the ones with no assignee and you are left with roughly four nationalities of inventor: Chinese, Belgian, Norwegian, and Korean. The American share is one patent.

That patent belongs to a small engineering shop in Stafford, Texas, called Deep Reach Technology. Its CEO is John Halkyard, who between 1971 and 1980 was Chief Ocean Engineer for the Kennecott Consortium — the original American attempt at deep-sea mining, the one that ended when commodity prices collapsed and the UN Convention on the Law of the Sea changed the rules of the game. Halkyard built Kennecott’s nodule lab in La Jolla, hired its 25-person engineering team, and in 1978 became the consortium’s secretary, managing the bankable feasibility study that Bechtel ran for the project. When the consortium dissolved, almost everyone in the industry went home. Halkyard kept going. Forty-six years later, his firm holds US patent 12,606,982, “Methods for suppression of seabed mining plumes,” granted on April 21 of this year. It is the only US-assigned patent on the critical path of the operation TMC is about to permit.

The plume problem is the part of seabed mining the public actually cares about. Hydraulic collectors pull nodules off the seafloor by sucking up a slurry of seawater, sediment, and rock; the sediment, if it escapes the riser system, drifts as a turbidity current across kilometers of seabed and re-deposits itself on slow-growing animals that have not been disturbed in a million years. The most-cited paper in the field on this dynamic is a 2022 in situ study by an MIT-led team that measured a sediment plume from a “pre-prototype” collector vehicle 4,500 meters down in the CCZ. It has been cited 71 times in three years. The vehicle in question was Patania II, owned by Global Sea Mineral Resources, a subsidiary of the Belgian dredging giant DEME. A 2024 follow-up paper in Marine Pollution Bulletin reported “complex and variable disturbances of meiofaunal communities” from Patania II’s run. The Patania family of suction-head collectors is patented in the United States, but not by Americans. US 12,553,212 and US 12,516,497 are both assigned to Deeptech NV, the Belgian holding entity for DEME’s mining IP. Same three inventors on both: Kris De Bruyne, Harm Stoffers, Stéphane Flamen.

What Deep Reach holds, in other words, is the part of the mining stack that determines whether the regulator lets the rest of the stack go to sea. The patent’s specification describes an “inverse hydrocyclone,” a separator that spins centrifugal flow backwards so that nodules drop into a closed riser while sediment is discharged at depth, never reaching the surface vessel. Read the inventor list and the timing gets stranger. Co-inventors include Robert D. Blevins, John Halkyard, Michael Rai Anderson, James Wodehouse, and Samuel Holmes. The cover page notes a confirmatory license to the US Department of Energy, dated November 2024. That clause typically appears when invention rights are owed back to the federal government because the work was funded under a Department contract. The single American patent on the critical environmental subsystem of the first American deep-sea mine was paid for, in part, by the American taxpayer. The DOE owns a piece of it.

The rest of the stack lives somewhere else. The riser, the four-kilometer vertical pipe that lifts the nodules from collector to ship, has been re-patented in the US three times since mid-2025, twice by China Merchants Deepsea Research Institute in Sanya (US 12,371,993 and 12,378,749, granted in July and August 2025) and once by Central South University in Hunan. The Chinese designs are mechanically distinct from anything in the Anglo-American literature: a “single high-pressure silo feeding device” that batches ore into a pressurized lock between the surface riser and the seabed line, a layout intended to handle the abyssal pressure differential without continuously spinning a slurry pump. Ocean University of China holds the sediment-monitoring patent (US 12,061,304). A separate Chinese filing on a “deep-sea mining vehicle” includes a startling subsystem: a two-component grouting system (Slurry A and Slurry B), mounted on the crawler tracks, that injects a setting mixture under the vehicle’s belly when sinkage sensors detect that the machine is being swallowed by abyssal mud. That patent, US 12,565,838, was granted on March 3 of this year.

The nodule-processing patents follow the same pattern. Loke Marine Minerals AS, a Norwegian start-up, holds US 12,065,715, granted in August 2024, on the hydrometallurgical separation of nickel, cobalt, copper, and manganese from polymetallic nodules. Umicore, the Belgian battery materials giant, holds US 12,168,812 (December 2024) on the same problem. The single broad US-issued metallurgy patent specifically for nodule chemistry is Belgian. The closest US analogue is held by a San Jose venture called Impossible Metals (formerly Impossible Mining), whose 2025 patent on “Seafloor harvesting with autonomous drone swarms” (US 12,442,297) describes a fundamentally different architecture: not a tethered crawler, but battery-powered underwater drones, dispatched from Underwater Smart Hubs, picking nodules one at a time under computer-vision control. Whether that approach can scale to TMC’s stated 3 million wet tonnes per year is the obvious question. The fastest hydraulic collectors achieved 86.4 tonnes per hour at pilot. Drone-by-drone harvesting is, mechanically, a different industry.

Why this matters: critical minerals policy in Washington has been framed as a question of where the metal will come from. The patent data suggests the harder question is whose machine will pull it out. TMC’s commercial partner for the actual lift operation is Allseas, the Swiss-Dutch pipelay contractor that owns Hidden Gem, the converted drillship that did the 2022 pilot run and lifted roughly 3,000 tonnes of nodules up a 4.3-kilometer riser. Allseas has 14 US patents going back to 1996, all in pipelay, lifting, and welding. None of them claim a nodule riser. The actual riser engineering is either trade-secret or filed in jurisdictions the US database does not see. A confidentiality choice, not an absence of work.

There is one way to read all this that is generous to the American story. The US never ratified UNCLOS, never joined the International Seabed Authority, never paid into the regime that gave companies like GSR and Loke their exploration contracts. The firms with the engineering have it because their governments wrote checks. Belgium underwrote Patania II through GSR’s federal sponsor. China runs its nodule program through state research institutes that file their US patents under names like “First Institute of Oceanography, Ministry of Natural Resources.” Norway’s Loke bought Lockheed Martin’s UK Seabed Resources subsidiary in 2023, briefly becoming the largest contract holder in the CCZ, then filed for bankruptcy this spring and handed the exploration licences back to Lockheed. The US private sector mostly walked away in 1981 and has let the rest of the world tinker.

The exception is John Halkyard’s office in Stafford, Texas, the four-person engineering JV that kept the lab notebooks open. If TMC gets its permit next year, it will go to sea on a Swiss-Dutch ship, dragging a Belgian-derived collector through a Belgian-monitored sediment plume, lifting Chinese-style riser slurry to a deck that will process its nodules using Belgian or Norwegian metallurgy. The one piece of the system that is unambiguously American — the way the plume gets kept out of the riser — has a federal license clause stapled to its cover page. The first US deep-sea mine is going to be a mostly imported machine with a single domestic safety valve.

Whatever you think of the venture, the strategic accounting is worth doing now, before the first nodule lands on deck.

Method note

This brief uses 9.3 million issued US utility patents from the USPTO bulk grant feed, 357 million scientific records from OpenAlex, and the WordPress-public news pages of TMC, NOAA, Allseas, Deep Reach Technology, DEME, Loke, and Impossible Metals. Patent counts in the “21 narrow filings” figure reflect grants since January 2020 whose title or abstract claims one of: polymetallic nodule, seabed mining, deep-sea mining, seafloor mining, abyssal mining, nodule collector or lifting, deep-sea ore, or mining-generated sediment plume. Roughly one third of the matching documents in the public USPTO grant feed have no resolved assignee at the time of writing — the named inventors and Google Patents bibliographic data were used to identify the assignee where possible. Citation counts on the scientific literature are as reported by OpenAlex in May 2026.