Greenland Rare Earths: A Resource Map, Not a Supply Chain
DividendChase LTD | Institutional Research
As of 20 September 2026
Greenland does not supply rare earths today. It hosts undeveloped rock, a uranium law that killed the largest deposit, one project with an exploitation licence and Western offtake letters, and a new U.S.–Denmark–Greenland security pact that would let Washington screen “sensitive investments.” That is a permitting and processing story wearing a geology headline.
China still separates and metals most of the world’s magnet rare earths. Greenland changes that only if concentrate leaves the fjord and is cracked, separated, and metalled in a non-Chinese plant. The mine is the easy slide. The plant is the chain.
Where the Rock Sits
Southern Greenland’s Ilimaussaq / Gardar province holds the two names everyone quotes, plus a third that closed a U.S.-aligned acquisition this month.
| Project | Operator | What is in the ground | Binding constraint | Status (Sept 2026) |
|---|---|---|---|---|
| Tanbreez (Kringlerne) | Critical Metals (NASDAQ: CRML), ~92.5%; European Lithium 7.5% | Eudialyte-hosted REE; ~27% of the REE basket as HREE (Dy, Tb matter); Zr, Nb; low U/Th | Capex, metallurgy, fjord logistics, refinery JV | Exploitation licence; Greenland approved the last ownership transfer; first-ore talk late 2028 / early 2029; concentrate exports targeted 2H 2029 |
| Kvanefjeld / Kuannersuit | Energy Transition Minerals (ASX: ETM), formerly Greenland Minerals; Shenghe historically a large holder | ~1 bn tonnes @ ~1.4% TREO; Nd/Dy rich; uranium ~250–350 ppm, zinc | Uranium Act (>100 ppm U ban) | Exploration licence not renewed June 2026; arbitration narrowed; no production path |
| Sarfartoq | Greenland Mines (U.S.-aligned, deal closed 2 Sept 2026) | Carbonatite NdPr story; company claims ST1 could equal ~34% of non-China NdPr oxide over nine scheduled years | Exploitation licence not yet in hand; ESIA; offtake | Initial Assessment / S-K 1300 2026; next ask is an exploitation licence |
Tanbreez is the only one that is both licenced to mine in principle and not trapped by the uranium threshold. That is why it is the Western talking point. Kvanefjeld is larger on paper and dead as a mine until Nuuk changes the 2021 law — which the post-March 2025 coalition has not. Sarfartoq is a licence-application story dressed in magnet-market arithmetic.
The Chain That Does Not Yet Exist
A rare-earth supply chain is five steps. Greenland today owns step one, incompletely.
- Mine and crush — still future tense.
- Concentrate — Tanbreez plans a eudialyte concentrate at the fjord. Eudialyte is not bastnäsite. Flowsheets are fussier; Western pilots exist; commercial scale does not, on this island.
- Crack and separate into individual oxides (Nd, Pr, Dy, Tb, others). This is the Chinese bottleneck. CRML has pointed at a Romania JV refinery with study-level revenue of $1.8–2.2 bn a year — a slide, not a plant.
- Metal and alloy.
- Magnet (sintered NdFeB, Dy/Tb grain-boundary diffusion).
Offtake letters already try to skip the China step:
- REalloys (U.S.): 15-year binding offtake for 15% of Tanbreez concentrate, priority on high-Dy/Tb parcels, ROFR on more (expanded from a 2025 LOI).
- Ucore: LOI on the order of 10%.
- Combined marketed offtake still only ~25% of planned output. The other 75% has to find a separator that is not Baotou.
Until a European or North American separator is financed, permitted, and running, Greenlandic concentrate is one customs form away from the same midstream it was meant to escape. That is the CSIS point, and it is still right: China’s advantage is separation, not only mines. A Greenland pit that ships to Shandong is a Chinese supply chain with better PR.
Why Kvanefjeld Cannot Be Willed Open
Kuannersuit would have been the volume answer — Nd/Dy at scale, plus uranium that made the economics work. Greenland’s parliament set a 100 ppm uranium ceiling in 2021. The ore sits at roughly three times that. In June 2026 Nuuk refused to renew the exploration licence because further work was “not deemed likely” to find an exploitable deposit inside the Uranium Act. ETM’s arbitration already lost the “give us a licence” question to Greenlandic courts. Shenghe’s historical shareholding is the political scar tissue: even a Western operator on that deposit reads as a uranium-and-China problem in Narsaq.
Investors who model Kvanefjeld tonnes as 2028–30 supply are modeling a statute change, not a mine plan.
Logistics and the Physical Island
Southern Greenland gives Tanbreez something African and Australian inland deposits do not: deep-water fjord access. It does not give roads, power at industrial scale, a housing stock, or a winter construction season that looks like Texas. Every tonne of concentrate is a shipping, energy, and workforce problem. Company acceleration budgets ($30 million class programs, 2026 drilling in the 6–10 km range) are study money. A producing REE mine is a billion-class decision once you add the separator.
Environmental and social licence is not a footnote. Kvanefjeld died in Narsaq as much as in Inatsisartut. Tanbreez’s low U/Th is why it still has a path. It is not why the town will automatically accept a concentrate plant.
How Friday’s Security Pact Hits the Chain
The unpublished U.S.–Denmark–Greenland deal would bar adversaries from bases and “sensitive investments” without U.S. written approval, and U.S. officials say that screen is meant to survive independence. For rare earths that means:
- Chinese offtake, equity, and construction finance on Greenlandic REE projects become treaty-constrained, not just politically discouraged.
- Western offtake (REalloys, Ucore, any DoD/DPA-backed separator) becomes the default, which helps bankability if the U.S. also writes cheques — and raises the cost of capital if Washington only writes vetoes.
- A mine that needs a Chinese separator to clear the last-mile magnet market has a new political problem: the concentrate may be allowed to leave; the control of the midstream is what the pact is for.
Do not confuse a security veto with a offtake contract. One stops Beijing. The other pays for a mill.
Who Actually Competes
Non-Chinese REE supply that exists in 2026 is still MP Materials / Mountain Pass (U.S., light-REE biased, still midstream-dependent), Lynas (Australia/Malaysia), and a thin set of recyclers and nascent separators in the U.S., Japan, and Europe. Greenland is not in that list. If Tanbreez hits 2029 concentrate, it is a heavy-REE flavored addition — useful for Dy/Tb in magnets — not a replacement for Chinese NdPr metal.
Sarfartoq’s “34% of non-China NdPr” line is a resource-to-market share claim on a project without an exploitation licence. Treat it as an option on Nuuk’s next permit, not as 2027 oxide.
Investor Map
| Ticket / theme | What you actually own | Risk that matters |
|---|---|---|
| CRML / Tanbreez | Licence + HREE geology + partial Western offtake + Romania refinery slide | Metallurgy of eudialyte at scale; 75% uncommitted tons; 2028–29 timeline slip; equity dilution |
| ETM / Kvanefjeld | Litigation and a banned orebody | Uranium Act; community veto; Shenghe optics |
| Greenland Mines / Sarfartoq | Early resource + U.S. alignment + NdPr narrative | Exploitation licence; ESIA; offtake not yet a chain |
| U.S./EU magnet & separator names | The scarce step | They capture Greenland value if concentrate shows up |
| Chinese REE majors | Still the midstream | A Greenland veto does not empty Baotou |
| Denmark Inc. | Almost no REE beta | Novo and Maersk are not Tanbreez |
Rules.
- Resource tonnes ≠ supply. Greenland’s share of mined TREO in 2026 is zero.
- HREE share is Tanbreez’s only scarce geological edge. Light-REE volume is not.
- A binding offtake for 15% is not a fully contracted mine.
- The new security pact is a China screen, not project finance.
- The separator — Romania or Ohio or Western Australia — is the investment that makes Greenland Western. Without it, the fjord feeds the old chain.
DividendChase Stance
Greenland is a 2029–32 option on heavy rare earths, gated by Nuuk, metallurgy, and a non-Chinese crack plant. It is not a 2026 supply-chain solution to magnet dependence, and it is not Kvanefjeld. Size CRML-class names as development equity with political tailwinds after Friday’s pact — not as a substitute for MP, Lynas, or a funded U.S. separator. The island’s rock is real. The chain is still a drawing.
Intelligence for the Discerning Investor
DividendChase LTD
