U.S. Geological Survey data released in February 2026 puts China's 2025 mine output at 67,000 metric tons out of a global 85,000-ton total, and direct imports still accounted for 26% of U.S. tungsten supply between 2021 and 2024 [S1].
For a buyer, that combination is the whole problem: 80% of upstream concentrate and ammonium paratungstate (APT) starts in China, while the U.S. has not shipped tungsten from a domestic mine since 2015, leaving only conversion capacity and finished-parts output [S1][S5]. Procurement teams that want to keep Chinese material in the bill of materials now have to clear a paper trail plus a regulatory test, while teams that want to leave it have to qualify new suppliers, new APT, and new tooling geometries on a tight clock.
Where China's tungsten actually sits in 2026
China produced 67,000 t of tungsten in 2025, with Vietnam a distant second at 3,000 t and no other country above 2,500 t, according to the USGS Mineral Commodity Summary cited in S1. The country also mines and refines more than 80% of global APT, the precursor to tungsten carbide powder and most downstream tungsten chemicals [S3][S6].
Import shares for the U.S. between 2021 and 2024 broke down as 26% China (including Hong Kong), 14% Germany, 8% Bolivia, 8% Vietnam, and 44% all other sources, and the 26% direct figure excludes material melted or processed in third countries before re-export [S1]. Net import reliance has stayed above 50% of U.S. tungsten consumption every year since at least 2021, with the only U.S. activity being downstream conversion of concentrate, APT, oxide, or scrap into powder, carbide, and finished parts [S1].
What the regulatory stack looks like for a U.S. buyer
DFARS 252.225-7052 covers tungsten metal powder and tungsten heavy alloy, including finished or semi-finished parts, and prohibits delivery of any covered material melted or produced in China, Russia, Iran, or North Korea; through the end of 2026, the test is simply "melted or produced," with a sourcing-traceability expansion taking effect in 2027 [S1].
Parallel pressure runs through trade policy. The U.S. raised Section 301 tariffs to 50% on several Chinese tungsten products at the end of 2024, and in February 2025 China imposed new export licensing on selected tungsten items, a step that the U.S. Defense Logistics Agency responded to with a 2,041 t potential acquisition list for fiscal year 2025 and zero planned disposals [S1][S5]. The 50% Section 301 rate is the operative tariff on covered Chinese tungsten products as of end-2024, and the U.S. military faces a 2027 deadline to halt procurement of tungsten mined or manufactured in China and Russia [S5].
What 2025 did to the price of APT and concentrate
Rotterdam 65% tungsten concentrate went from $266 to $551 per metric ton unit (mtu) across 2025, and APT ran from $331 to $675 per mtu on the same track, both more than doubling in a single year per USGS data cited in S1. Floor-price discipline is already visible: Almonty Industries' offtake with Global Tungsten & Powders on South Korean APT locks in a $235/mtu minimum against a then-current $342.50/mtu spot, an explicit hedge against Chinese supply surges [S5].
Volatility transmits downstream. Constrained APT shows up as longer lead times, working-capital strain, qualification work for alternative sources, and customer-delivery renegotiations, with Hyperion Materials & Technologies listing material planning, working capital, production scheduling, and delivery commitments as the four internal failure points when APT tightens [S3].
Specifying Chinese vs. non-Chinese APT and powder
The technical specification is only the first filter; the origin question is the binding one. For tungsten carbide and finished parts, the discriminating criteria are: origin and traceability paperwork (mine, concentrate, APT, oxide, powder stage), APT chemistry and impurity profile (Mo, Si, P, S, As typical spec-line items), particle size distribution on the finished powder (FSSS typically 0.8 to 30 micrometres depending on grade), and binder content for cemented carbide (commonly 6 to 15 wt% Co for cutting-tool grades, higher for wear parts). DFARS 252.225-7052 binds the first criterion to a yes/no test; the rest still have to meet drawing requirements [S1].
Buyers that want to leave Chinese material behind should plan for a qualification cycle on new APT and powder before locking production. A practical comparison for a procurement team: Chinese APT is widely available, lowest cost in normal years, but now subject to export licensing and a 50% Section 301 tariff on downstream products; Vietnamese and Bolivian concentrate is available at small tonnage with limited APT conversion outside China; South Korean APT from the Sangdong restart will carry a contract floor, long-term offtake, and ex-China documentation once qualified; and North American sources are still pre-production, with Defense Production Act Title III awards funding projects in Nevada, New Brunswick, and Yukon [S1][S5]. Pricing on long-term contracts can be structured as a floor-plus-spread on the mtu, which is the model Almonty has signed with Global Tungsten & Powders [S5].
Operating limits, failure modes, and what auditors actually check
DFARS 252.225-7052 through the end of 2026 is a binary melt-and-produce test: if the covered material was melted or produced in a covered country, it cannot be delivered under a covered contract, and the contractor is responsible for keeping the documentation that proves otherwise [S1]. The third-country reroute is the failure mode auditors look for, meaning tungsten that entered Germany or Vietnam as concentrate, was processed there, and then re-entered the U.S. without an obvious China label [S1].
Mechanical failure modes do not change with origin, but procurement risk does. Hyperion's listed downstream effects of APT volatility, including lead-time slip, pricing instability, and forced requalification of alternative sources, are the operating symptoms a buyer should track internally rather than only monitoring spot APT prices [S3]. For comparison, an industrial process specifier weighing origin versus mechanical performance should treat origin documentation as a hard go/no-go gate before any of the standard spec items (pressure transmitter range, flow meter accuracy, industrial valve trim) are even reviewed, because a part that fails DFARS 252.225-7052 cannot ship regardless of how well it meets the rest of the drawing.
Standards, sourcing signals, and what to track next
Three verifiable signals will move this market between now and end of 2026. First, the 2027 expansion of DFARS 252.225-7052 to a full sourcing-traceability test (i.e. mine-to-powder documentation, not just melt-and-produce), which is a hard date already written into the rule [S1]. Second, the Defense Logistics Agency's stated 2,041 t fiscal year 2025 stockpile acquisition, with zero planned disposals, which sets a floor under U.S. government tungsten demand independent of commercial cycles [S1][S5]. Third, the production ramp at Almonty's Sangdong mine in South Korea, which the company has stated is targeting 50% of ex-China global tungsten supply once at full output [S2].
For procurement teams that still need Chinese-origin material, the immediate steps are: confirm the Section 301 tariff exposure at 50% on the specific HTS code, line up the mill's melt-and-produce certificate to the DFARS 252.225-7052 wording, and pre-clear the paperwork with the prime contractor's DCMA office before acceptance [S1]. For teams that want to leave Chinese material behind, the practical lever is qualifying non-China APT now, while Sangdong output is still building and Bolivian and Vietnamese concentrate is still available at small tonnage, and before the 2027 DFARS expansion raises the documentation bar [S1][S5]. Tracking these three signals, plus the next USGS Mineral Commodity Summary release and any new Chinese export-licensing updates, is the only way to keep a tungsten bill of materials both compliant and buildable through the end of 2026.
Related analysis: Food-Grade Caliper and Micrometer Spec Map for 2026.